Special Issue; Efforts for the Environment December 2009...¡Advance zero emissions: seek a...

December 2009

ISSN 0915-0528 OSAKA, JAPAN

Special Issue;Efforts for the Environment

R

TECHNICAL REVIEWNo.77 Special Issue:

Efforts for the Environment

corporation

Round-table talk

Our Line of New Products

106Advance and Future of Analysis Technology for Constant Velocity Joints

1Preface

Market Trend of Wind Turbine and Bearing TechnologiesNobuyuki NINOYU, Souichi YAGI and Tsuyoshi NIWA

Eco-friendly Products Development SNR RoulementsSiegfried RUHLAND, Ludovic SAUNIER, Cynthia TSEN, Bernard LIATARD and Gérald MIRABEL

Actuator for Electromechanical Brake Tatsuya YAMASAKI, Masaaki EGUCHI and Yusuke MAKINO

Initiatives for Environmental Conservation Masakazu HIRATA and Noboru UMEMOTO

Approach on Environment by SNRDidier SEPULCHRE DE CONDE and Jean-Hervé BULIT

Yoshikazu FUKUMURA

28

42

Electric Ball Screw Actuator for Automotive Yoshinori IKEDA and Hirakazu YOSHIDA

47

Plastic Bearings of Electric Pump for the Next Generation Cell Norio ITOH and Takuya ISHII

51

Low-torque Roller Lifter Unit Masashi NISHIMURA

56

Hub Bearing for Automobiles for Saving Resources Isao HIRAI

62

V-series Hub Joint Mitsuru UMEKIDA and Yuuichi ASANO

67

Extreme High Load Capacity Tapered Roller Bearings Takashi UENO and Tomoki MATSUSHITA

73

New Spherical Roller Bearings “ULTAGE” Series Type EA and EMYukihisa TSUMORI

81

Compact Free type Torque Diode Shoji ITOMI

86

Characteristics and Applications of DLC films Kouya OOHIRA

90

Optimized Radius of Roller Large End Face in Tapered Roller Bearings Hiroki FUJIWARA, Takashi TSUJIMOTO and Kazuto YAMAUCHI

96

34

2

10

114

Environmentally-friendly Products and Technology

An Approach the Reduction of Environmental Loads in the Material ProcessingAkira SERA, Hisatsune SAITO, Kenichi NAKANO and Syunsuke MAKINO

Improvement the Work Environment with Factory RobotShigeki MATSUSHITA and Genji NAGURA

22

18

Environment-responsive Production Engineering

Activity for Reducing Environmental Impact

"2009 'CHO' MONOZUKURI Innovative Components Awards" Encouragement Award High Capacity Tapered Roller BearingTakashi UENO

"2008 'CHO' MONOZUKURI Innovative Components Awards" Automotive Component Award Constant Velocity Steering Joint (CSJ)Kenta YAMAZAKI

"2008 Resource Recycling Technology and Systems" Encouragement Award Development of Briquetting Machine and Accomplishment Recycling System for Steelmaking DustShozo GOTO, Ikuo YAMADA and Katsutoshi MURAMATSU

110

111

112

Our Line of Award Winning Products

NTN TECHNICAL REVIEW No.77CONTENTS

-1-

NTN TECHNICAL REVIEW No.77（2009）

[ Preface ]

The concentration of greenhouse gases in the atmosphere has increased by about 30% since theonset of the Industrial Revolution. If the emission of greenhouse gases continues at this rate, the rise intemperature is expected to reach almost twice what it was before the Industrial Revolution by the end ofthe 21st century.

At the end of 2009, the 15th United Nations Climate Change Conference (COP15) was held inCopenhagen where medium and long-term reduction goals were discussed. In every country, efforts arebeing made to reduce greenhouse gases. Developments in electric cars and other next-generationautomobiles, solar and wind power generation, and other new energy fields must also be activelyadvanced.

Given these circumstances, on July 1, 2009 we established a New Products and Intellectual PropertyStrategy Department at NTN as a division that keeps future growth in mind. It is responsible for long-term business strategies that emphasize the global environment, energy conservation and resourceconservation in the fields of natural energy and electric cars, for example. In addition to utilizing ourcompany’s fundamental technologies, this department will promote collaboration between industry,academia and government, as well as with different business fields. The department will also seek toaccelerate the development of new products for automobile parts and industrial machinery that meet therequirements of new energy fields, for example.

In our efforts at NTN, we are prioritizing the development of fundamental technologies with the goal ofreducing environmental burdens and the development of products that are considerate of theenvironment in our efforts at NTN.

With the protection of the global environment, the creation of a recycling-based society and theenhancement of environmental management systems as key issues, we are working to raise the levelsof our environmental management activities at our production facilities. In particular, we are makinggreat efforts to reduce CO2 emissions in our industrial activities.

In this publication, we introduce “Activities for reducing Environmental Impacts,” “Engineering that isResponsive to the Environment” and “Environmentally-Friendly Products and Technologies” of NTN andSNR, a company that has joined the NTN Group and is headquartered in France. After a roundtablediscussion with Dr. Kimata, a former NTN director, ” we also introduced three developed products thathave received prizes in the current and previous fiscal year (two ‘CHO’ MONODZUKURI InnovativeComponents Awards and one Resource Recycling Technology and Systems Encouragement Award).

The NTN corporate philosophy is “For New Technology Network: Our contribution to the globalcommunity lies in our creation of new technologies and development of new products.” In our new “NTN2010 for The Next Step” medium-term management plan for the next two years from April 2010, we willseek to realize a management core that is not dependent on scale. In order to achieve “harmony withthe global environment,” we will promote the development of products that are better for theenvironment and contribute to the sustainable development of society.

Yoshikazu FUKUMURAManaging Director

To the New Issue Featuring Effortsfor the Environment

-2-


[ Technical Article ]

Initiatives for Environmental Conservation

At NTN, our entire group is working together toreduce the emissions of CO2 in order to prevent globalwarming and decrease the production of waste.Bearings and constant velocity joints, which reduce theamount of energy loss caused by friction, arefundamentally green products, These products are alsocontributing to the preservation of the environmentthrough their application in fields such as wind powergeneration, which uses natural energy, and rollingstock, which is a green means of transportation. Whenwe build new factories, we are effectively using naturalenergy through the incorporation of solar powergeneration and wind power generation, and adoptingmanufacturing processes that include environmentally-friendly manufacturing methods that eliminate the useof oil. We are also building more compact productionlines and reducing waste through the installation ofgrinding sludge briquetters. As a result of these efforts,we are establishing factories that are considerate ofand good for the global environment.

In ;past publications, we have primarily introducedour new products and technologies. Recently,however, preservation of the global environment andthe depletion of petroleum energy resources havebecome major concerns. For these reasons,businesses are also expected to reduce theiremissions of greenhouse gases and adopt activitiesthat are good for the global environment. In thisarticle, we explain NTN’s efforts for the environment,focusing on our efforts to reduce CO2 emissions inparticular.

1. Fundamental environmentalpolicies and goals

We have established the NTN Environmental Policy(Table 1), making “harmony with the globalenvironment” the top priority of our company, and weare advancing efforts to preserve the environment withthe following four policies as guides for our activities.

i) Develop and sell products that are good for theenvironment

ii) Reduce impacts on the environment: work toimprove energy efficiency and carefully managesubstances that have environmental impacts inevery aspect of our business, from manufacturing todistribution and sales, in order to reduce impacts onthe environment

iii) Build comprehensive systems for legal complianceand environmental management

iv) Contribute to society and protect the naturalenvironment

In order to advance and realize this environmentalpolicy, we have built an environmental managementstructure for our entire group that is centered on ourCompanywide Environmental ManagementCommittee. We are setting environmental targetsevery year for three areas—preserving the globalenvironment, creating a recycling society, andmaintaining and improving environmentalmanagement systems (Table 2), and we areadvancing our efforts for the environment as planned.

*Innovation and Intellectual Property Strategy Headquarters Environment Strategy Dept.

At NTN, we have been conducting our corporate activities basedon the NTN Environmental Policy and seeking harmony with theglobal environment through systematic and continuous initiativesto reduce impacts on the environment. In this article, we introduceour efforts to reduce CO2 in manufacturing and other efforts for theenvironment in research and development, as well as in ourcontributions to society.

Masakazu HIRATA*Noboru UMEMOTO*


-3-

Table 1 NTN Environmental Policy

¡Enforce thorough compliance with our green procurement　standards in our supply chain¡Handle investigations of substances with environmental　impacts reliably (begin the application of a REACH-　compliant system)¡Advance auditing of substances with environmental impacts　(internal auditing, auditing of suppliers)¡Continue inspection of incoming materials

Goal

Pre

serv

ing

the

glob

al e

nvir

onm

ent

Crea

ting

a re

cycl

ing

orga

niza

tion

Mai

ntai

ning

and

impr

ovin

gen

viro

nmen

tal m

anag

emen

t sys

tem

s

2009 fiscal year targets 2010 fiscal year targets

Reduction ofCO2 emissions

Prevention of air,water and soilpollution

Participation in effortsto protect the naturalenvironment

Promote environmentalmanagement in oursupply chain

Reduction ofwaste production

Preservation ofresources

Internal systems

Promote greenprocurement

Comply with new lawsand regulations

¡Total CO2 emissions: 250,000 tons/year

¡Conversion from halon to CO2 fire extinguishers: 70 units　(54 unchanged units at the end of FY2009)

¡Reduction in the handling of substances subject to the　PRTR Law: –38% compared to FY2006

¡CO2 emissions per basic unit: –8% compared to FY1997

¡Total CO2 emissions: 317,000 tons/year


¡CO2 emissions per basic unit: +3% compared to FY2006

¡Soil testing and purification after the closing of our　Takarazuka facility

¡Elimination of all halon fire extinguishers

¡Reduction in the handling of substances subject to the　PRTR Law: –40% compared to FY2006

¡Compliance with the revised Soil Pollution Countermea-　sures Law

¡Advance zero emissions: seek a recycling rate of 99.5%

¡Reduce landfill waste to 25 tons/month or less

¡Promotion of local revegetation efforts　(corporate forest stewardship, etc.)

¡Advance zero emissions: seek a recycling rate of 99.5%

¡Reduce landfill waste to 25 tons/month or less

¡Achieve zero emissions: seek a recycling rate of 98.3%¡Advance zero emissions: seek a recycling rate of 96.3%

¡Weight of purchased paper: –50% compared to FY1998¡Weight of purchased paper: –46.2% compared to FY1998

¡Reduce styrene foam: –96.1% compared to FY1999¡Reduce styrene foam: –92.4% compared to FY1999

¡Acquire ISO14001 certification at all locations worldwide

¡Certification of all suppliers

¡Prepare system for the acquisition of certification at　new locations

¡Support supplier efforts to obtain ISO14001 and similar　certifications: 90% certification rate

¡Revised Law Concerning the Rational Use of Energy,　revised PRTR Law


Japan

Japan

Japan

Japan

Japan

Japan

Japan

Japan

Japan

Over-seas

Over-seas

Over-seas

－

－

－

At the NTN group, we make harmony with the global environment our top priority, and we are continuously reducingthe impacts on the environment caused by our business activities. We are also striving to constantly contribute to thesustainable development of society.1. Develop and sell products that are good for the environment ¡We will utilize high-precision manufacturing and tribological technologies and develop environmentally-friendly products. 　By providing these products to society, we will also contribute to promoting energy conservation around the world.2. Reduce impacts on the environment ¡As we strive to help prevent global warming, we will seek to improve energy efficiency in every aspect of our business 　activities, from the procurement of raw materials, parts and other goods to manufacturing, distribution and sales. ¡We will focus on the effective use of resources and strive to reduce waste by thoroughly implementing the 3Rs (reduce, 　reuse, recycle). ¡In addition to comprehensively managing substances with environmental impacts that are used in products and 　manufacturing processes, we will actively seek to replace chemical substances that are suspected of being harmful with 　alternative substances, and we will strive to prevent pollution of the environment.3. Build comprehensive systems for legal compliance and environmental management ¡Starting with a foundation of strict compliance with laws and regulations, we will also establish and abide by voluntary 　standards and in-house regulations that are more stringent in order to respond proactively to the needs of communities 　and our customers. ¡In addition to preparing systems and operation methods for environmental management, by conducting environmental 　audits we will seek to maintain and improve our environmental management systems. In addition, we will publicize 　information about the status of our efforts toward the environment and promote communication with the public. ¡Through environmental education, internal bulletins and other efforts, we will seek to make every person working at an 　NTN location aware of our Environmental Policy, and we will seek to increase their awareness of environmental 　protection.4. Contribute to society and protect the natural environment ¡By actively participating in programs that contribute to local communities and support efforts to protect the natural 　environment, we will promote interaction with neighboring communities and reduce our environmental impact.

Table 2 Environmental targets in fiscal 2009 and fiscal 2010


-4-

established a plan to reduce the CO2 emissions of allour locations by about 20,000 tons, which isequivalent to about 6% of our total emissions in the2007 fiscal year.

[Changes in the amount of CO2 emissions andthe basic unit]

We can say that the basic unit of CO2 emissions forour Japanese locations by fiscal year tended todecrease, and our energy efficiency, which affectsCO2 emissions, had been improving annually through2007. The basic unit for the 2007 fiscal year was 1.46,which was an 18% improvement compared to the1997 fiscal year, so we achieved our target.

Between the 2003 and 2007 fiscal years, our CO2

emissions continuously increased each year as wealso increased our production volume. However, in the2008 fiscal year, as a result of the rapid decrease inproduction from the second half of the fiscal year, thebasic unit increased because our energy use did notdecline correspondingly with production volume.

Our CO2 emissions and energy consumption bypower source for the 2004 and 2008 fiscal years areshown in Figs. 2 and 3, as well as Table 3. Theseillustrations make clear that the majority of our CO2

emissions are the result of the use of electric power,and that the amount resulting from oil use decreasedfrom the 2004 fiscal year to the 2008 fiscal year.Compared to the 2004 fiscal year, kerosene usedecreased, while natural gas use doubled in the 2008fiscal year (Table 3). The use of thermal processingfurnaces and other fuel conversion, as well as thereplacement of air conditioning equipment, were thecauses of this switch from kerosene to natural gas andelectricity.

Fig. 1 Change in total CO2 emissions and CO2

emissions rate (At operating sites in Japan)

2. Efforts for the environment in ourproduction activities

The majority of our efforts toward our environmentaltargets, including preservation of the globalenvironment and creation of a recycling society, occurat our production sites. In this article, we introduce ourefforts for the environment in our production activities,focusing on efforts to reduce CO2 emissions inparticular.

2.1 Efforts to reduce CO2 emissionsAt the Third Conference of the Parties (COP3) of

the Framework Convention on Climate Change held in1997, when consensus was reached on numericalCO2 reduction values for each country, considerationin Japan about efforts for countermeasures againstglobal warming began immediately. The JapaneseFederation of Economic Organizations established avoluntary environmental action plan for industry in1997.

At NTN, we have participated in this voluntaryenvironmental action plan and the CO2 emissionstargets set by the Japan Bearing IndustrialAssociation, which set the target basic unit* foremissions for the 2010 fiscal as 13% less than that ofthe 1997 fiscal year. We voluntarily set an even higherreduction target of 25 in an effort to further reduceCO2 emissions.* The basic unit is the total amount of CO2 emissions (in tons)

per million yen of added value of production.

The first commitment period of the Kyoto Protocolbegan in 2008, and the importance of setting targetsfor total CO2 emissions has increased. The CO2

emissions for our corporate group have increasedyearly, but in March 2008 we added a target for thetotal emission of CO2 from our locations in Japan.

The CO2 generated by our group is the result of ourproduction activities. In order to achieve our target fortotal emissions, in April 2008 we initiated a CO2

Reduction Project to investigate the potential to furtherreduce CO2 emissions from our factories and initiateefforts to do so. The activities of this project wereconducted for one year, and included the acquisitionof information about the latest energy conservingtechnologies and their incorporation, the exchange ofinformation about examples of improved energyconservation and the incorporation of conservationequipment being undertaken at each manufacturinglocation, as well as the discussion of suchimplementation plans at other locations. Additionally,we examined the thermal processing furnaces of othercompanies and invited outside experts to introducetechnologies to us. Based on these efforts, we

Emission amounts(in units of 10,000 tons)

Basic unit(tons of CO2 per million yen)

60.0

45.0

30.0

15.0

0.0

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2.0

1.5

1.0

0.5

0.0

Emission amountsBasic unit

Year

-5-

[Efforts to reduce CO2 emissions at productionplants]

With the goal of reducing CO2 emissions caused byour production activities, since 2005 we havecontinuously received energy conservationdiagnostics assisted by the New Energy and IndustrialTechnology Development Organization (NEDO) toreceive advice about problem points and proposals forsignificant reform measures from the perspectives ofexperts. With this guidance, we have been draftingenergy conservation improvement plans.

Fig. 2 CO2 emissions by power source (2004)

Fig. 3 CO2 emissions by power source (2008)

Fig. 4 Wastewater pipeline and heat exchangerat Iwata Works

Fig. 5 High-efficiency screw compressor at Iwata Works

We will describe examples of energy conservationimprovements that we have implemented as a resultof these efforts, as well as examples of our ownindependent efforts.

At our Iwata Works in 2006, in addition to installingin-house power generation equipment, we are alsousing a pipeline (Fig. 4) to carry the hot water emittedby this equipment to our constant velocity joint factory.Through heat exchange, we are able to heat thefacility in the winter, for example, reducing CO2

emissions by about 1,500 tons in a single winterseason. We have also been able to reduce the fuelcosts from using kerosene boilers.

In 2008, through an energy service company(ESCO), we undertook the replacement of an aircompressor with the latest model of turbo compressor(Fig. 5), as well as adopted a system to control thenumber of units and an intake cooler. We alsoimplemented countermeasures for air pipe leaks andimproved air flow. As a result of these efforts, wepredict CO2 reductions of 4,000 tons per year startingin the 2009 fiscal year.

Oil, 8%

Electricity, 80%

Gas, 12%

Total CO2

emissions:310,000

tons/year

Oil, 6%

Gas, 13%

Total CO2

emissions:310,000

tons/year

Electricity, 81%

Table 3 Energy consumption by power source

Type of energy 2008 fiscal year2004 fiscal year

Electricity (GWh)

Natural gas (km3)

LPGButane (tons)Propane

Heavy oil (kL)

Kerosene (kL)

Light oil (kL)

Gasoline (kL)

Gas

Oil

510 520

7,310 6,810

3,220 6,330

3,520 2,620

4,250 2,590

160 160

130 130


-6-


At our Kuwana Works, we are advancing theconversion to natural gas as the fuel used for heattreatment facilities. We made the conversion of 22facilities in 2005, reducing the amount of CO2

generated due to heat treatment by about 10%. Weare continuing this fuel conversion at present.

At our Okayama Works, in 2008, we received agrant from the Ministry of Economy, Trade andIndustry (2008 Leading Model Business for theAdoption and Popularization of Load LevelingEquipment). We used this to install a new thermal air-conditioning system (Fig. 6). We also replaced aconventional absorption water cooler and heater thatwas powered by kerosene with a high-efficiencyelectric turbo refrigerator. By using these with athermal storage system that uses electricity late atnight, which makes for operation that is even moreefficient and as inexpensive as possible, we are ableto achieve both CO2 and cost reductions. By usingclathrate hydrate slurry, which is a thermal storagemedium that has twice the thermal storage capacity ofwater, we have increased thermal storage efficiencyand reduced energy use in the transportation of themedium. By incorporating this system, we expect toreduce CO2 emissions by 1,000 tons per year.

In 2008, NTN Kongo Corporation received a grantfrom the Ministry of the Environment, and installedhigh-efficiency equipment, including air-conditioning(Fig. 7), lighting, compressors and air blowers at itsfour factories through an ESCO business. As a part ofreceiving subsidies for these projects, we joined theMinistry of the Environment’s voluntary DomesticEmissions Trading System in the 2009 fiscal year, andwe have agreed to a CO2 reduction target of 3,000tons per year.

To increase energy conservation in lighting, in 2007we replaced 185 mercury lamps with high-efficiencymetal hydride lamps in the heat treatment plant of ourKuwana Works (Fig. 8), resulting in a reduction of 185tons of CO2 per year. At NTN Casting Corp., weinstalled inverter fluorescent lamps and made otherefforts to increase the efficiency of the lighting there.By doing this, we have reduced CO2 emissions by 690tons per year. At Hikari Seiki Industry Co., Ltd., in2007 we replaced all the existing lighting within thefactory with inverter fluorescent lamps, which arebrighter and better for the environment, resulting in areduction of the total number of lamps from 520 to 392and increased efficiency.

At every NTN location, we are also implementing airleak countermeasures (Fig. 9). Through the utilizationof leak detectors, we are conducting regularinspections of air pipes for leaks and makingimprovements as necessary.

Fig. 7 Turbo refrigerator and cooling tower for the air-conditioning system at NTN Kongo Corporation

Fig. 8 Metal hydride lamps at a heat treatment plantof Kuwana Works

Fig. 9 Check by air leak detector

Fig. 6 A thermal storage tower for the new airconditioning system installed at Okayama Works

-7-

Fig. 10 Wind power generator at the first factoryof NTN Mie Corporation

Through the combination of large-scale energyconservation improvements and small-scale energyconservation activities, we are striving to reduce CO2

emissions at NTN.

2.2 Waste reduction and resource preservationIn order to reduce waste products, we are seeking

to achieve a final target rate of 100% recycling, andwe are promoting the recycling of resources. In the2008 fiscal year, our results were mixed as weachieved a recycling rate in Japan of 99.6%, which ishigher than our target of 99%, but we failed to achieveour reduction target overseas. At the same time thatwe are reducing the production of waste, we are alsoadvancing the development of recycling equipment. Inthe 2008 fiscal year, the technology for “thedevelopment of a briquetting device for dust from ironand steel and the creation of a recycling system” thatwe developed jointly with Daiwa Steel Corporation, anelectric arc furnace manufacturer, received anencouragement award for Resource RecyclingTechnologies and Systems from the Clean JapanCenter, which is a foundation that is affiliated with theMinistry of Economy, Trade and Industry.

2.3 Promotion of green procurementWe have established Green Procurement

Standards, and we are striving to protect theenvironment and improve the environmentalperformance of our products in cooperation with ourbusiness partners. In order to comply fully withregulations and laws related to substances withenvironmental impacts, including the RoHS and ELVdirectives and the REACH regulations of theEuropean Union (EU), we have created systems forthe analysis of incoming parts and materials, andspecialized internal auditing systems for substanceswith environmental burdens.

The EU’s REACH regulations, which came intoeffect in June 2007, require manufacturers andimporters to register and evaluate the safety of allsubstances, compounds and chemicals that areknown to be emitted from products that are to be usedin the EU. At our group, we have prepared a list of 19chemicals that are included in grease and otherproducts that we export to the EU in amounts of oneton or more annually, and we confirmed that themanufacturers of all the raw materials conductedpreliminary registration. 15 substances of very highconcern (SVHC), which are substances that arecarcinogenic or bioaccumulative, were listed by theEuropean Chemicals Agency in October 2008. Sinceone of these substances, phthalate ester, has beencontained in rubber tools and CVJ boots, our

development divisions, in cooperation with ourbusiness partners, are now advancing thedevelopment of materials that do not contain thissubstance.

2.4 Environmental Guidelines for theconstruction of new factories and structures

We have established guidelines for the constructionof new factories and structures in order to contributeto the prevention of global warming by reducing CO2

emissions and with the goal of reducing energy costs.

[Environmental Guidelines](1) Use solar and wind power generation to provide

electricity for offices and kitchens (Fig. 10)(2) Adopt all-electrical kitchens and EcoCute energy-

conserving hot water supply equipment(3) For office buildings, plant vegetation on roofs and

install film that reflects light and heat on windows(4) Ensure that building roofs have heat reflecting

paint or two layers(5) Use lighting with motion sensors in places where

people pass infrequently and there are no safetyconcerns

(6) Utilize natural light(7) Ensure the use of use stratified (displacement) air-

conditioning systems(8) Use air-conditioning systems with thermal storage

that take advantage of low outdoor temperaturesand low electricity costs at night

(9) Investigate the incorporation of in-house powergeneration systems that use natural gas and theuse of the excess heat generated by suchequipment (in cases where heat-use needs arehigh and natural gas infrastructure is available)

(10) Prohibit the use of incandescent light bulbs (useinverter florescent lamps and compact fluorescentbulbs)


(11) Prohibit the use of mercury lamps (adopt the useof high-efficiency metal halide lamps)

(12) Prohibit the running of vending machines at night(install timers)

3. Efforts for the environment in thedevelopment of products

Bearings, constant velocity joints and our othermain products make great use of tribologicaltechnologies. They serve to reduce friction and wearin rotating parts and the parts that transmit force fromrotation. As such, these can be said to beenvironmentally-friendly products or “green” goods.In other articles in this journal, we introduce high-efficiency CVJs and DLCs that are the fruits of ourfriction reduction technologies, resource-conservinghub bearings that utilize our technologies fordesigning to conserve resources and otherenvironmental products and technologies.

As indicated in the previous section, the eliminationof substances with environmental impacts from ourproducts is an important issue. We have alreadyachieved compliance with the RoHS and ELVdirectives, so now we will rapidly advancecountermeasures for the substances of very highconcern (SVHC) defined in the REACH regulations.

In the future, it can be said that the most importantissue from the perspective of product development isthe extent that we can reduce environmental impactcaused by products in their manufacture, use anddisposal. From the development stage we will strive to(1) design to suppress the generation of CO2 in themanufacturing process, (2) design to reduce materialloss, (3) reduce the quantity of rare metals used, and(4) decrease energy loss during use (conserveenergy). As a part of these efforts, we have begun topromote life cycle assessment (LCA), which is amethod of quantitative evaluation. We are currentlyinvestigating manufacturing processes, but the nextstep can be thought to be LCA evaluation of productswhen they are in use.

4. Efforts to contribute to society

At every NTN location we are promoting efforts tocontribute to society in cooperation with localcommunities.

One such effort is tree-planting. For example, at ourIwata Works, we agreed in July 2008 to join theShizuoka Mirai No Mori Supporter (supporters ofShizuoka’s future forests) system with the Iwata Cityand Shizuoka Prefecture. In this system, we areundertaking activities at an NTN Kigyo no mori

-8-


(corporate forest) where people can interact withnature. At our Okayama Works as well, we signed aKigyo to kyodo no mori zukuri (public-corporate forest-planting) agreement with Mimasaka City, OkayamaPrefecture in August 2008. For five years until March2013, we are being allowed to borrow about 3.5hectares of forests owned by the city, and we areadvancing efforts to improve the forest there as anNTN Kigyo No Mori. At both of these Works, we havebeen undertaking efforts to preserve forests incooperation with local governments until now, but withthese agreements, we are further expanding the scaleof our activities. These restored forests will provideplaces of comfort not only for local residents, but theyshould also allow our employees and their familiesgood opportunities to interact with nature.

5. Conclusion

At NTN we are focusing our abilities onenvironmental preservation in everything frommanufacturing processes to products themselves,including their distribution, use and disposal. When weare building new factories and facilities, we make theincorporation of natural energy, including solar andwind power, a fundamental policy. We are activelyincorporating natural energy as one means ofreducing CO2 emissions. In the 2008 fiscal year, weinstalled 94-kW of solar power generation equipmentat our Kuwana Works (Fig. 11) and 62-kW at our MieWorks. So far, we have installed 10 solar powergenerators, 14 wind power generators and 13 hybridgenerators that combine wind and solar power.Together, these generators save the equivalent of 290tons of CO2 emissions per year (Fig. 12).

At NTN, our dream is to use the power of the windand the sun to manufacture bearings, and providebearings that conserve resources and energy to thegreatest extent possible. We will continue our dailyefforts to reduce environmental burdens in everythingfrom manufacturing processes to the productsthemselves, including their distribution, use anddisposal.

-9-

Masakazu HIRATA

Innovation and IntellectualProperty Strategy Headquarters

Environment Strategy Dept.

Photo of authors

Noboru UMEMOTO

Innovation and IntellectualProperty Strategy Headquarters

Environment Strategy Dept.

Fig. 11 Solar power generation system at Kuwana Works

Fig. 12. Natural energy output

700

600

500

400

300

200

100

02003 2004 2005 2006 2007 2008

Cumulative natural energy

Solar

Wind power

Fiscal year

Cum

ulat

ive

outp

ut

kW


-10-



Efforts for the Environment by SNR

1. EMS of SNR

Since its foundation, SNR has kept an establishedISO 14001-complying environmental managementsystem (EMS) for its all operation sites. This EMS isbased on a particular portion of the qualitymanagement system of SNR commonly applicable tofunctions and business operations (documentation,education, audit management, etc.).

This EMS is the very basis of the environmentalpolicy of SNR, and SNR wishes to fulfill itsenvironmental objective through continuedimprovement activities according to the basic principleof PDCA (Plan, Do, Check, Action). The corporateannual guideline of SNR is established according to itsenvironmental policy, guidance for its shareholders,possible environmental impacts resulting from itsbusiness activities, new ecology-related legalregulations, and results of internal and external audits.This annual guideline constitutes the basis forplanning with all the SNR operation sites, therebySNR is maintaining follow-up activities for progressand performance of projects by referring to variousindices.

Since 1999, all the SNR operation sites have beenregistered as “multi-sites” defined in ISO 14001,where the latest update took place in 2008.

2. Corporate organization

In SNR, the Environmental Management Divisionbelongs to the Quality Management Department, andis under the directly supervision of the President. Fig.1 schematically illustrates the organizational chart ofthe Environment Management Division of SNR.

In each SNR base within France, a responsibleengineer functioning as an environment manager isresponsible for smooth operation of EMS in his/hersite, by enhancing knowledge about his/her specialtyand expanding the scope of his/her role andresponsibility.

SNR Annecy Head Office & Plant includes aproduction plant and research division for bearingproducts, product and machine test center, including aproduction engineering division and an engineeringservice division (including laboratory and precisionmeasurement center). In each division, the “ecologypromoter” as a coordinator plays the role of “operationsite environment manager”.

In SNR overseas operation sites, the responsibilityabout environmental issues is assigned to qualitymanagers.

The “Environment Manager in the Head Office”chairs the regular meetings in which environmentmanagers from all the SNR operation sites participate.He/she always checks that the operation and activitiesabout SNR EMS are executed as planned.

**NTN-SNR ROULEMENTS President**NTN-SNR ROULEMENTS Environment Div. Manager

When we think about it bearings are present everywhere in our dailylives, from household appliances to aerospace applications not tomention in the automotive market. By reducing the energy needs ofmany things around us, bearings play a major role in supportingsustainable development. Since 1994, at SNR Roulements, we haveintegrated environmental concerns into our development.Environmental policy is a major focus for SNR Roulements for not onlyecological reasons, but also for economic and image reasons. That iswhy our environmental policy is being implemented from our head officedown to all our workshops and suppliers.

Didier SEPULCHRE DE CONDE**Jean-Hervé BULIT**

-11-


3. Production of bearings and resultantenvironmental impacts

Fig. 2 schematically illustrates factors in bearingproduction activities that can lead to environmentalimpacts.

As a result of investigation at each SNR operationsite, the following aspects of the bearing productionprocess that may lead to environmental impacts havebeen identified:qEnergy consumption resulting from production

activities and building heatingqConsumption of water and chemical substances

used for production process and coolant supply orchilling equipment

qRelease of evaporated water and volatile chemicalsubstances into atmosphere

qSolid waste typically occurring from turning andgrinding processes

qWastewater from rinsing machine; waste liquidassociated with the production process or processliquid

[Steps in Bearing Production Process](1) Production of bearing rings through forming of

blanks (tubes or bars of particular lengths) bymachining or press process

(2) Heat treatment(3) Grinding, and super-finishing(4) Assembly (bearing ring, rolling elements, and

cage)(5) Prefilling with grease or fitting of seal, depending

on bearing type.Fig. 1 Organization of SNR Environmental Management

Department

Fig. 2 Environmental impacts

SNR has developed excellent environmental systemtools (environmental database, dedicatedenvironmental analysis software, and ecology-relatedlegal regulations monitoring software), and has beenutilizing these tools to support its routineenvironmental analysis activities and help implementcommon use of its environmental data by all theoperation sites.

Person Responsible in Each Base

Assistant

CustomerQuality Div.

EnvironmentDiv.

Productionplant

EnvironmentalTechnologies

Quality &Environment Div.

EcologyPromoter

EcologyPromoter

EcologyPromoter

Ecology Promoter(doubles as Quality Manger)

Quality Management Dept.

Division Chief Meeting

QualityManagementSystem Div.

Resources

Waste

Water

Atmosphere

Impacting Factors

Steel material: 1.27 kgEnergy: 3.65 kWhElectricity: 3.03 kWhNatural gas: 0.62 kWhWater consumption: 3.38 L* Others: 0.12 g

* Others: Examples include grease, seal, resin, coolant, and rinse liquid.

Industrial waste (hazardous): 0.18 kgGrinding sludge: 0.07 kgIndustrial waste (non-hazardous): 0.07 kgMetal waste: 0.32 kg

Industrial wastewater: 0.21 LSanitary sewage: 0.9 L

Evaporation

1 kg of bearing product


-12-

past several years. Fig. 4 provides the type of basicunit for energy consumed for building heating at SNRas well as the trend in energy consumption for buildingheating at SNR over past several years.

5. Reduction in CO2 emissions

To be able to assess the contribution of SNR’s CO2

emissions reduction efforts to mitigation ofgreenhouse effect, we have translated the amounts ofour energy consumptions in electricity, oil and gas intocorresponding “CO2 emissions”.

In terms of CO2 emissions, the reduction in energiesconsumed by SNR is equivalent, if its productionoutput remains unchanged, to approximately 3,400tons reduction in CO2 emissions. Please note that thisreduction is limited only to the SNR operation sites inFrance. Fig. 5 graphically plots the trend in CO2

emissions at SNR over past several years.

Fig. 3 Energy of processing

0.76

0.74

0.72

0.72005

20062007

2008

Bas

ic u

nit

MW

h/kO

Year

Fig. 5 Discharge of CO2

CO

2 em

issi

ons

Ton

s

17000

16000

15000

14000

130002005 2006

Year

20072008

Fig. 4 Energy of heating

2.6

2.4

2.2

1.8

1.6

1.4

20052006

20072008

2

Bas

ic u

nit

MH

z/av

erag

e te

mpe

ratu

re

Year

4. Reduction in energy consumption

Typical examples of energy consumption reductionefforts of SNR are as follows:(1) An existing conventional air-compressor has been

replaced with a high-efficiency screw-type aircompressor; the compressed air line is inspectedfor air leakage at regular intervals (Meythet Plant).

(2) An exhaust system has been installed in a facilitywhere oil mist of higher concentration was present.During winter in particular, fresh air is supplied tothe workplace, thereby significantly improving theenvironment of the work station. As a result, theoverall electricity consumption previously neededfor the exhaust system in the entire works as wellas boiler gas consumption have been reducedconsiderably (Meythet Plant).

(3) Improved employee awareness for reduction inenergy on idle or inactive equipment (all SNRplants and offices).

(4) Two aged fuel-firing boilers have been replacedwith one novel gas-firing boiler, thereby improvingthe environment in the workplace and reducing theenergy consumption (Annecy Head Office-Plant).

Fig. 3 shows the type of basic unit for energyconsumed for processing in SNR as well as the trendin energy consumption for processing in SNR over

6. Reduction in water consumption

Typical efforts so far made by SNR in reducingwater consumption are presented below:(1) A certain portion of the aged water-cooled salt bath

heat-treatment plant has been replaced with a newhardening plant, and for the remaining portion ofthe existing heat-treatment plant, the rinsingsystem situated downstream the heat-treatmentprocess has been modified into an overflowdesign, thereby reducing water consumption(Annecy Plant).

(2) A new cooling tower has been introduced into thecoolant water system for metal working process tohelp reduce the risk of bleeding bacteria anddecreasing water consumption (Meythet Plant).

Fig. 6 illustrates the type of basic unit for waterconsumption at SNR as well as the trend in waterconsumption at SNR over past several years.

-13-

Fig. 6 Consumption of water

1

0.96

0.92

0.88

0.84

0.82005

2006

Year2007

2008

Bas

ic u

nit

m3 /

kO


Fig. 7 Organic compound

1.50

1.00

0.50

0.002006

2007

Year2008

Bas

ic u

nit

kg/k

O

Fig. 8 Emission of organic compound to atmosphere

Em

issi

ons

Ton

s

250

200

150

100

502006

2007Year 2008

7. Reduction in organic compoundsemissions into atmosphere

When exposed to sunlight, pollutants such asvolatile organic compounds (VOCs) in atmospherenear ground will generate photochemical substancesthat can adversely affect human health andvegetation. In bearing production processes at SNR,VOCs can typically occur from use of volatile oil, CFCor alcohol employed for rinsing and degreasing ofbearing parts.

Some of our efforts for reducing emissions oforganic compounds are described below:(1) Oil solvent used in final rinsing step within bearing

production has been superseded with a lessvolatile type (at all operation sites of SNR).

(2) Previously, CFC was used for rinsing bearingsbefore non-destructive inspection. This substancehas been superseded with a detergent (ArgonayPlant).

Through these efforts, all SNR operation sites withinFrance alone achieved reduction of about 152 tons interms of VOCs emitted into atmosphere.

Figs. 7 and 8 show the type of basic unit for organic

compounds used at SNR as well as the trend inorganic compound emissions into atmosphere at SNRover past several years.

8. Reduction in amount of industrialwastewater

Ordinary municipal wastewater treatment facilitiesare intended to treat household wastewater, and willnot able to treat all pollutants occurring from industrialwastewater (a pollutant may pass the soil in awastewater treatment facility and can be released intothe environment). To address this problem, it isnecessary that wastewater from the productionprocesses at SNR (in particular, the barrel polishingprocess) is preprocessed within the plant before beingreleased to the nearby municipal wastewatertreatment facility). Typical examples of our efforts forreducing emissions of our industrial wastewater tooutside facilities are as follows:(1) Heavily contaminated rinse water from heat-

treatment process is treated in special wastewatertreatment equipment rather than being directlyreleased into the municipal wastewater treatmentfacility (Annecy Plant).

(2) Previously, wastewater was directly released into awastewater treatment plant. Now, wastewaterhaving undergone treatment with a high-performance filter is returned to a reservoir andreused. This arrangement positively contributes toreduction in wastewater emission and freshwaterconsumption (Meythet Plant).

-14-


9. Management of wastes

Industrial wastes occurring from worksites ofproduction plants can be classified into “non-hazardous wastes” that can be treated together withhousehold wastes; and “hazardous wastes” that mustbe treated in a special plant. Non-hazardous wastesfrom industrial processes are more frequently recycledcompared to ordinary household wastes. Because thistype of wastes typically consists of paper, carton,plastic and glass materials, 90% of them can berecycled. At the same time, SNR has been proceedingwith research that will help convert certain hazardousindustrial wastes into recyclable-materials. Figs. 11and 12 show the amounts of non-hazardous andhazardous industrial wastes occurring from SNR overseveral years.

Fig. 9 illustrates the type of basic unit for reuse ofindustrial wastewater at SNR as well as the trend inreuse of industrial wastewater at SNR over pastseveral years.

Amount of industrial wastewater can be convertedinto amount of household wastewater in terms ofpopulation (number of residents). Thus, theachievement by efforts of SNR in mitigatingenvironmental impacts in this period through reductionin emissions of wastewater including industrialwastewater is equivalent to reduction of householdwastewater consumed by 590 residents.

Currently, the amount of wastewater released fromSNR is as small as the amount of householdwastewater from only 160 residents. Fig. 10 illustratesthe trend of amount of industrial wastewater from SNRover several years, wherein the amounts of industrialwastewater are shown as converted into the amountsreleased by residents of ordinary households.

Fig. 11 Innocuous waste

38

36

34

322005

2006

Year2007

2008

Bas

ic u

nit

kg/

kO

Fig. 12 Harmful waste

55

50

45

402005

2006

Year2007

2008

Bas

ic u

nit

kg/

kO

Fig. 9 Reuse of industrial wastewater

0.15

0.1

0.05

02005

2006

Year2007

2008

Bas

ic u

nit

m3 /

kO

Fig. 10 Conversion to household use (number of people)from industrial use (wastewater)

Equ

ival

ent u

sage

Pop

ulat

ion

800

700

600

500

400

300

200

100

20052006

Year2007

2008

0

-15-

Fig. 13 Principle of VACUDEST

M

(2)

(1)

(4)

(3)

(5)

(6)

Waste liquidinlet

Depositionliquid outlet

Condensateoutlet

(1) Recuperator (2) Evaporator (3) Water separator(4) Compressor ROOTS (5) Immersion type exchanger(6) Deposition storage


10. Ecology-improving technologies of SNR

SNR has long been committed to development andimprovement in bearing designs, as well asimprovement in bearing product manufacturingprocesses: each commitment positively contributing tobetter ecology. Some ecology-improving techniquesunique to SNR are summarized below:[Examples of development and improvement]• Improvement in performance of steel material

Reduction in bearing size helps reduce amount ofraw material consumed.

• Optimization of internal structureReduced friction in bearing helps reduce the energyneeded to drive the associated machinery.

• Improved grease qualityReduction in amount of prefilled grease helpsreduce amount of chemical substanceconsumption.

• Optimized seal performanceLonger bearing life and reduced grease loss.

• Compliance with REACHSNR assures that it uses the registered chemicalsubstances only.

[Examples of improved production processes]• Turning process was replaced with forging process:

consequently, the amount of metal mass removedfrom blank has been much reduced.

• Improvement in heat-treatment process has led tomitigated deformation of bearing rings.Consequently, grinding allowance and energyconsumption in grinding process have been muchreduced.

• Having been supported by CETIM (CentreTechnique des industries Mécaniques), SNR has

Fig. 14 Biological purification unit

Bio-reactor

Compressed air Detergenttank

Rinsing reservoir

Bio-activatingagent tank

pH-adjustingliquid tank

been conducting various surveys about industrialwaste.

10.1 Test for steam condenserWaste liquid essentially comprises water. For this

reason, a steam condenser separates water fromcontaminated liquid through evaporation.

SNR Alès plant has tested an advanced steamcondenser. Thanks to the resultant findings, SNR cannow calculate data about performance andenvironmental impacts of a given steam condenser,and has begun assessment of positive effectsobtained from a particular steam condenser. Fig. 13schematically illustrates the operating principle of thesteam condenser tested.

10.2 Test for biological purification unitWaste liquid from a rinsing reservoir contains a very

large portion of waste oil. On our biological purificationunit, the oil-containing waste liquid is transferred to abiological reactor where the oil content in the liquid iseliminated by bacteria, and then the purified liquid isreturned to the rinsing machine. The effectiveness ofthis demonstration plant has been verified in the SNRAnnecy Plant. Fig. 14 shows a schematic illustrationof this biological purification unit.

10.3 Grinding sludge briquettingSludge occurring from the grinding process contains

not only steel powder but also coolant. Thus, this typeof sludge is categorized into hazardous wastes. In theSNR Seynod Plant, grinding sludge is divided intocoolant and grinding powder, and then the grindingpowder portion is compressed into solid briquettes.Being recyclable, these briquettes are reused in asteel manufacturer.

11. Investments into ecology-improvingactivities

In order to continue its ecology-improving activities,SNR has invested approximately 5 million euro in the2005-2008 period. Fig. 16 shows a breakdown of theinvestments by objective for ecological improvement:

Incorporating an NTN proprietary technology forsludge briquetting (Fig. 15), the process has alreadybeen implemented in the SNR’s mass-production line.In addition, SNR has been testing techniques forbriquetting other sludge types.

-16-


Fig. 18 Fuel-efficient vehicle with SNR-made wheelbearings at the Shell Eco-Marathon®

Fig. 17 SNR booth at local exhibition

Fig. 16 Investments

Considerations abouttreatment and storageof waste4.4%

Considerations aboutrelease into atmosphere7.4%

Decontami-nation of soil15.4%

Risk avoidance:prevention ofLegionnaires’ disease5.6% Energy-saving

measures6.5%

Survey about applicablelaws and regulations,and necessary actions9.3%

Control over releaseof industrialwastewater8.1%

Risk avoidance:total ban on use ofPCB-containingtransformers39%

Noise control1.7%

Considerations aboutdischarge of rainwater1.3%

Risk avoidance:Safe storage oforganicsubstances1.2%

Fig. 15 Grinding sludge briquetter

12. Promotion for Eco-consciousness

To help the SNR employees become moreconscious about their ecology-associated activities,SNR has been keeping a basic eco-educationprogram for its newly hired employees. In particular,SNR provides special eco-education curriculum for itsemployees working in sites of higher environmentalrisk within SNR.

SNR is also committed to support of eco-consciousness promoting activities outside SNR. Forexample, SNR engineers and technicians have beenactively presenting ecology-related lectures atuniversities (Ecole Centrale de Lyon, PolytechSavoie), local science festivals, and ecology-centeredspecial seminars (World Bearing Association (BWA),and ADEME). Fig. 17 shows a picture of an SNRbooth at a local exhibition. Also, SNR has beensupporting the development of a specialized vehiclefor an inter-college eco-run race (vehicles compete fora maximum range run with 1 liter of fuel). Fig. 18shows a vehicle competing in the eco-run race. Thisvehicle utilizes SNR-made wheel bearings.

-17-

Didier SEPULCHRE DE CONDENTN-SNR ROULEMENTS President

Jean-Hervé BULITNTN-SNR ROULEMENTS Environment Div. Manager

Photo of authors


13. Conclusion

The achievements in ecology-improving activitiesdescribed above have resulted from the long years ofeco-conscious commitments of SNR employees, anddemonstrate that the EMS system of SNR has beenbeneficial.

Beginning in 1999, all the operation sites of SNR inFrance, as well as those in Brazil and Romania, haveacquired ISO 14001 certification. By closelymonitoring its carbon footprint, SNR is making stridesto improve its effects on the global environment.

-18-



Efforts for the Reduction of Environmental Impactsin Material Processing

1. Use of white lubricant in forging processfor constant velocity joint parts

Recently, every manufacturer has been increasinglyrequired to combat the challenge of environmentalissues. For example, in the forging process forconstant velocity joint parts within Japan’s bearingmanufacturers, graphite lubricants (lubricantsprepared by dispersing graphite powder into water ormineral oil) have been superseded with whitelubricants (water solution primarily comprising organicacid salt and polymer) in an attempt to mitigateenvironmental impacts and improve workingenvironments in production plants.

Generally, graphite lubricants have been used in thehot forging process because of their betterperformance in terms of lubricity, die-releasingcapability, and die life. However, these lubricants havedrawbacks: a work site involving a graphite lubricantcan readily get dirty; contaminating oil deriving from apress slide gib may not be readily separated: graphitelubricant mixed with an excessive amount of oil canexhibit deteriorated lubricating performance andneeds to be replaced, and the contaminated lubricantwill make an industrial waste. To sum up, use of thislubricant type can lead to a heavy environmentalimpact.

In contrast, white lubricants can help improve workenvironments in forging process, be readily separatedfrom oil, and readily recovered. Thus, use of thislubricant type can help greatly reduce occurrence ofindustrial wastes.

Wishing to help conserve the global environment,NTN has been increasing the use of white lubricants.This paper describes its efforts to mitigateenvironmental impacts through use of white lubricants.

1.1 Features of white lubricantsWhite lubricants are water-soluble lubricants each

essentially comprising of organic acid salt and apolymer. Their features are summarized in Table 1below.

*Production Engineering R&D Center

We developed a forging processing technique that uses a whitelubricant to achieve both the reduction of environmental loads andthe improvement of the operating environment in the forgingprocess.

We introduce a technology to reduce forging billet weightvariation and thea technology that improves yield rate by changingthe product shape and the forging process.

Akira SERA* Hisatsune SAITO*Kenichi NAKANO* Syunsuke MAKINO*

Table 1 The characteristic of white lubricants

Lubricant

Major components

Coolingperformance

Drying readiness

Oil recoveringquality

White lubricant

Polymer-based

Alkyl maleatePhthalatePhosphateSulfonate

Silicate glassColloidal silica

Carbonicacid-based

Waterglass-based

×

：Excellent　　：Good　　：Acceptable　　×：Poor

NOTE 1) Various white lubricant types are available, including polymer-based, carbonic acid-based, and water glass-based types. Manufacturers of these lubricants have been committed to development, aiming at the optimization of advantages of each lubricant type.NOTE 2) Aqueous solution of organic acid salt is transparent. After applied to a set of dies and heated to an appropriate temperature, an aqueous solution containing such an organic acid salt will form a whitish lubricant film after its water content has evaporated.

Cha

ract

eris

tics

-19-

Efforts for the Reduction of Environmental Impacts in Material Processing

1.2 Performance of graphite lubricants ascompared with that of white lubricants

NTN has compared performance of graphitelubricants with that of white lubricants, in terms offorging-lubrication performance, workability-improvingcapability, and safety. These results are summarizedin Table 2. Graphite lubricants excel in forging-lubrication performance. In contrast, white lubricantsboast better workability-improving capability andsafety. Therefore, so that white lubricants are morecommonly used in forging process, their forging-lubrication performance needs to be improved to alevel equivalent to that of graphite lubricants.

air blowing time to determine a temperature rangewhere a dry lubricant film readily occurs (Table 3). Anattempt was then made to quantify necessarylubricating conditions.

Fig. 2 shows a view of lubricant film formed on apunch used to forge an outer race of fixed constantvelocity joint.

To be able to determine an optimal set of lubricantfilm forming conditions, we have first definedconditions about duration and timing for applyinglubricant as well as conditions of blowing air. Therelevance of these conditions was verified byperforming a series of tests on work-pieces beingforged. Consequently, we have achieved good resultsabout quality of forging-formed products and extendeddie life.

As described above, we have adopted a whitelubricant in our forging process for NTN constant jointouter race primarily to improve work environmentsthrough the selection of an appropriate forginglubricant, control for die temperature, and lubricationconditions. Consequently, we have achievedimprovement in work environments. We wish to furtherextend forging die life and expand a line of applicablebearing numbers and sizes.

Table 2 Comparison of capability Graphitewith white lubricant

Forging-lubricationperformance

Workability-improvingcapability

Safety

Lubricity, die release property

Function Graphitelubricant

Whitelubricant

×

×

Die life

Working conditions

Die setup, die change

Putrefaction resistanceMaintainability againstseizure on sliding surfacesand lubricant path

Immunity against fire

：Excellent　　：Good　　：Acceptable　　×：Poor

Table 3 The film situation by die temperature during the test

Room temperatureto low temperature

Lowtemperature

Appropriatetemperature

Hightemperature

Die temperature State of lubricant film formed on die

Water content in lubricant does not evaporate,and the lubricant remains in liquid state.

Water content in lubricant evaporates quicklyand a lubricant film occurs promptly.

Droplets of lubricant rolling on the die surface arerepelled from the surface, not being deposited onthe die surface to form a lubricant film.

Water content in lubricant evaporates and alubricant film is eventually formed: however,this process takes an extended time.

1.3 Lubrication conditions setup and test resultAny white lubricant exhibits a satisfactory lubricating

effect only when it has formed a dry lubricant film. Tobe able to form an appropriate dry lubricant film,relevant control of die temperature is critical.Therefore, we have studied various working conditionsthat can help achieve formation of optimal lubricantfilm and evaluated these conditions.

Fig. 1 shows an example plot of die temperaturemeasurements taken during hot forging processes.First, we have adjusted lubricant application time and

High

0 60 120 180 240 300 360 420

Low

Time sec.　

Measurements of die surface temperature

PunchDies

Tem

pera

ture

˚C

Fig. 2 The film situation of the Punch for the Outer Race Joint

Fig. 1 Example of the result for die temperature check


2. Minimization of variation in weightof forging blanks

In order to cope with an increasing rise in rawmaterial prices and help conserve the globalenvironment, demands have been rising for lighterproduct designs, improved yields with parts thatconstitute final products, and introduction of near-net-shape forging techniques.

The forging process for constant velocity joint partsbegins with a shearing step as illustrated in Fig. 3.Since the shearing is the first step in the entire massreduction process, it is critical that the blanks beingprocessed have a much more precise weight. Ourefforts exercised for this objective are hereunderdescribed.

Incidentally, to be able to minimize variation in theweight of forging blanks, it is necessary to minimizevariation in feed length and outside diameter of thesteel bar work piece being sheared.

Fig. 3 The Forging process

Forging process

She

arin

g

Bon

derit

etr

eatm

ent

Col

d fo

rgin

g

Pre

coat

ing

Pre

forg

ing

Turn

ing Fig. 5 The image drawing of Shear Tooling

Fixed tool

Sizing stopper

Shearing blade

Steel barwork piece

Movable tool

Fig. 4 The outside drawing of a shear press with FeedForward System of Bar O/D

Shear pressproper

Pinch rollers

Bar diametermeasuring rig

Steel barwork piece

2.2 Countermeasure against variation in billetweight

(1) Countermeasure against variation in entirebillet length

(Possible cause) At the beginning of shearingoperation, the steel bar was bent by the downwardmotion of the movable tool. The bar was shearedeven though its positioning was, at the time,unstable.

(Remedy) We have modified the form of sizingstopper and shear tooling so that any steel bar isreliably sheared in a stable horizontal position.

(2) Method for correcting variation in steel baroutside diameter

(Cause) The outside diameter can vary even steelbars belong to a same lot (Ex.: if the true diameterof dia. 70 x 100 (nominal) steel bar work piece is70.1 mm, then the true weight of this work piecewill be 11 g greater).

(Remedy) Using the bar diameter measuring rig inFig. 4, we scanned the entire area of the steel barwork piece to determine the variation in the outsidediameter of the steel bar. Then, we determined thedeviation in bar diameter relative to the targetedoutside diameter and attempted to make weightcompensation for the billets by adjusting the lengthof billets so that the as-obtained billets weigh thesame as specimens of targeted outside diameter.

2.1 Overview of shear press (bar diametermeasuring rig) and shear tooling

Fig. 4 schematically illustrates the appearance ofour shear press equipped with a bar diametermeasuring rig. This machine comprises of the shearpress proper, a steel bar feeder, and a bar diametermeasuring rig. Before shearing, the bar diametermeasuring rig scans the entire area of the steel bar tomeasure the outside diameter and supplies thesemeasurements to the controller. When a change in theoutside diameter is detected, the sizing stopper isshifted accordingly to trigger a weight compensationoperation.

Fig. 5 shows the constitution of the shear tooling.The steel bar shearing procedure consists of insertionof the steel bar from the fixed side of the tooling to thesizing stopper, lowering the movable side of thetooling to shear the steel bar, and removing theseparate work pieces from the tooling. The separateblanks obtained from shearing are hereunder called“billets”.

-20-

-21-

Taking a dia. 69.5 mm steel bar material type as amodel, we have investigated currently achievabledimensions and attempted to minimize variation in theentire length and outside diameter with the billetsobtained from the steel bar material. Consequently,we have succeeded in limiting the variation in weightof the billets by approximately 60% as shown in Fig. 6.

We, NTN, have been not only attempting to realizelighter bearing designs but also committed to variousdevelopment works that help mitigate environmentalimpacts through reduction in the amount of materialsused through improvement in machining techniquesfor our bearing products. One representative exampleof our commitment is the above-mentioned near-net-shape technique adopted in our forging line, whereinthe feed stroke for the steel bar work piece isregulated based on the trend in the measurementsobtained from the outside diameter measuringequipment in order to minimize variation in the weightof resultant billets. NTN people in charge of forgingtechnology have been making efforts to develop noveltechnologies to help realize better ecoconsciousmanufacturing practices.

3. Efforts of NTN for improved yield withgood constant velocity joint productsand reduction in materials used

As a part of our commitment to reduction inenvironmental impacts, NTN has been challenged witha variety of objectives that include reduction in theamount of materials used in components of NTNproducts, reduction in the energy consumed duringmanufacture of these components, and wastereduction. This challenge also applies to themanufacturing of inner races and outer races for NTNconstant velocity joint products. To fulfill theserequirements, NTN has been involved with efforts forimproved yield for constant velocity joint products of

Fig. 6 Weight variation before and after used by FeedForward System of Bar O/D.

0

0

100 200 300 400 500 600 700Number of data sets

Bill

et w

eigh

t

Greaterweight

Lowerweight

Comparison in terms of variation in weightwith and without compensation

W/o compensation W/ compensation Permissible range

Fig. 7 Variation of Constant Velocity Joint parts

Akira SERAProduction Engineering

R&D Center

Hisatsune SAITOProduction Engineering

R&D Center

Kenichi NAKAJIMAProduction Engineering

R&D Center

Syunsuke MAKINOProduction Engineering

R&D Center

Photo of authors

Efforts for the Reduction of Environmental Impacts in Material Processing

acceptable quality, Typical examples of these effortsinclude:(1) Improved product design that can help delete a

machining step for the forged surface while the forgedwork piece positively satisfies needed functions

(2) Realization of not only extended die life but alsothe introduction of near-net-shape techniques bysuperseding the conventional forging process witha novel forging process (this new forging processtechnique is realized through introduction of newtechniques including a novel analysis technique.)

(3) Realization of more compact and lighter productdesigns whose performance is equivalent to that itsof conventional design, beginning with the reviewof designs of conventional constant velocitybearing products

Fig. 7 shows some examples of new NTN constantvelocity joint products that reflect our efforts in (1)through (3) above. NTN has positively achievedmitigation in its environmental impacts throughreduction in material used and alleviation inmachining-derived environmental impacts. Thesecontributing techniques include more common use offorging technique to delete certain machining stepsincluding turning, reduction in mass possibly removedby turning, and introduction of more compact andlighter product designs.

We will further remain committed to develop andmanufacture ecofriendly bearing products through thedevelopment of new techniques aimed at reducing theamount of materials used, mitigating environmentalimpacts from postforging process, reducing or deletingmachining steps, and creating lighter weight productdesigns by introducing multi-material designs.

-22-



Improvement of Work Environmentswith the Use of Factory Robots

1. Introduction

Constant velocity joints (hereinafter called CVJs)are a major product of NTN and are used in manyautomobiles as a means of power transmission.

In a typical CVJ assembly process in NTN, 12 to 16parts are assembled in sequence to form the finalproduct. Fig. 1 schematically illustrates a typical CVJassembly process.

NTN regards steps 1 through 3, which arecommon to various CVJ types, as the start-upassembly process and steps 4 through 6, which canvary depending on the CVJ type processed, as thecompletion assembly process.

NTN mass-produces a variety of its CVJ products.However, a larger portion of its CVJ assembly isachieved through manual operation because ofreasons described below. Thus, these jobs are difficultto execute and pose a heavy burden on the operators.(1) Unique shapes of CVJ components

Many CVJ components are prone to deformationbecause of their material and/or shapes, including therubber boot, resin boot, and ring-shaped boot bandmade of a thin metal strip.

To be able to automatically install these difficultcomponents, there will be orientation and alignmentchallenges during component feeding and handlingthroughout assembly. In conventional practice,position (orientation) of components is visuallychecked, and components are manually transferredand assembled.

(2) Unique shapes of finished CVJ productsBecause each CVJ is designed for a particular

function, the individual components of CVJ’s and thefinal products are diverse in shape. Thus, ordinarytransfer systems are not readily capable of handlingfinished CVJ products creating a necessity for humanoperators to transport these products.

With a larger finished CVJ product weighingapproximately 10 kg, and the operators involved willexperience heavy physical burden while transferring awork piece from a particular process to a downstreamprocess or while changing the orientation of a workpiece.

(3) Precision phase-matching procedureMany CVJ components (rubber boot, shaft, inner

ring, outer ring, etc.) have to be assembled into finalproducts, being fully phase-matched betweencomponents.

These components each have unique geometricalfeatures including cut-outs and projections in snap-fitjoints and boot receiving joints and these features areused for convenience to achieve an optimal phase-match between components during assembly of CVJproduct. The allowance for snap-fit is very smalland/or snap-fit joints consist of soft and a hardmaterial components forcing NTN to rely onexperienced operators to manually assemble thecomponents and remain conscious of phase matchingbetween components involved.

During manual assembly work of a CVJ product, anoperator assembles components into a final product

*Production Engineering R&D Center

Human sense and experience are necessary for assembling processes.Therefore, automation (mechanization) has been difficult. We used sensor,robot, and simulation technologies, which have made remarkable advancesin recent years, and succeeded in improving working environments in theassembly process for constant velocity joints (CVJ).

We introduce this case in this article.

Shigeki MATSUSHITA*Genji NAGURA*

-23-

Improvement of Work Environments with the Use of Factory Robots

while matching phases of the components involvedbased on visual detection and handling touch. Incontrast, to be able to utilize an automatic assemblysystem for a CVJ product, it is necessary to first detectthe phases of all the components involved and relaythe data to downstream processes so that thecomponents can be assembled in correct phases.

Fig. 1 Standard CVJ assembly processss

I/B boot+ small band

I/B inner ring

I/B large band

NOTE) "Inboard (I/B)" means the engine-side of CVJ having been installed onto a car body and "outboard (O/B)" means the wheel-side of installed CVJ.

I/B outer ring

Snap ring

Clip

O/B boot+ small band

O/B sub-assy

O/Blarge band

1

2

3

4

5

1 Boot and small band are inserted over the shaft. 2 I/B inner ring and O/B clip are fitted. 3 I/B snap ring and O/B sub-assy are installed.

4 OB large band is inserted over the sub-assy. 5 I/B outer ring is inserted, and the boot is fitted. 6 I/B large band is inserted, and both bands are　 form-fitted.

7 Functions and appearance of product are inspected.

Start-upassemblyprocess

Completionassemblyprocess

Inspectionprocess

Creating such a system that produces reliable resultsposes a great challenge for producing reliable results.

To address the above mentioned challenges, NTNhas committed to improving work environments in CVJassembly facilities by deleting high difficulty and highburden steps. For this purpose, NTN has developedan automated CVJ assembly process incorporating a6-axis vertical multi-joint robot that is virtually capableof human motions (hereinafter simply called the“robot”).

2. Approaches to automation throughuse of robot

The objective of the robot-assisted CVJ assemblyline recently developed is “the elimination of humanactivities by automation with robot.” It is intended tomitigate the work load on human operators, and to doso, NTN has executed the following four approaches:(1) Analysis of work flow

In considering the assembly process involvingrobots, each manual process step within the wholeprocess has been analyzed and numericallyassessed. Based on this obtained knowledge,challenging jobs done by experienced operators havebeen replaced with automated industrial robots.(2) Greater scope of applicability

A coordinated operation is performed using of arobot featuring greater degree of freedom of motionwhile maintaining a simple station for the operator toassist the robot. Consequently, delicate work, whichcannot be achieved with a robot alone, is possible,leading to a greater productivity through cooperation.(3) More reliable work piece transfer operation

Through the introduction of a more compact, lighterrobot hand as well as optimization for speed andrange of robot hand motion, the reliability of robotassisted word has been improved. Consequently,product transfer operation, which was once a largelyphysical burdening manual process, has now beenreduced with robot use.(4) Numerical control of force-sensitive operation

Where the worker used to do operations by feel orexperience, the robots allow for much more controlledoperation and gather data during the process asfeedback for further engineering.

To help proceed with these approaches, NTN hasexecuted robot motion simulation by gathering loadinganalysis data from manual operation and using a 3Dmodel. As a result, NTN is now capable of checkingequipment layout and potential interference onequipment and running cycle times before equipmentinstallation. Thus, NTN has achieved not only optimal


assembly procedure and equipment layout but also areduction in the development period for thisequipment.

Fig. 2 shows a view of the 3D robot motionsimulation system.

Fig. 3 Intelligent parts supply device

Fig. 4 Tripod and phase information

3. Challenges and solutions aboutautomated CVJ assembly process

NTN has identified the following four majorchallenges with the robot-assisted CVJ assemblyprocess. This section describes these challenges andtheir corresponding solutions.(1) Detection of component orientation(2) Automated snap-fitting work for serrated

components(3) Automated assembly work for difficult-to-handle-

shaped components(4) Improved efficiency in robot-assisted operation

3.1 Detection of component orientationTo run a problem-free automated assembly,

orientation-related data of each component, includingposition (location), phase, and orientation (front orback), must be correctly sensed.

To address this challenge, NTN has developed aunique intelligent parts feeder capable of correctlyorienting/aligning parts. This parts feeder systemcaptures images of CVJ components, makesjudgment about position (location), phase, andorientation (front or back) of individual CVJcomponents based on the captured images andtransmits the resultant judgment information to therobot. The system always places individualcomponents arranged in a same orientation into asame location. By this development, reliability in partstransfer with a robot has been much improved, and, atthe same time, has eliminated work issues for theoperator with respect to aligning parts. Fig. 3 shows aview of the intelligent parts feeder. Fig. 4 illustrates acaptured image of parts.

Phase information

-24-

Fig. 2 Robot simulation

3.2 Automated snap-fitting work for serratedcomponents

The majority of CVJ components must beassembled with their phases matching. Like manysteps in CVJ production, the snap-fitting step forjoining serrated components poses a challenge tomake automated assembly work for CVJ productionpossible.

As shown in Fig. 5, the outer circumference of bothends of the shaft as well as the inner bore of the innerring, which receives one end of the shaft, are serrated.Manual phase-matching with the two serratedcomponents is a delicate procedure where mismatchof phases is prone to occur.

To address this issue, NTN has developed twotechniques for snap-fitting with these two serratedcomponents, and is applying one technique accordingto the shape of the components.

-25-

Fig. 5 Serration fitting portion

Innerring

ShaftSerrated outer circumference of shaft

Serrated bore of inner ring

possible only when the phases of both componentsare correctly matched. In other words, the rubber bootcannot be fitted over the outer ring simply with press-fitting force. Fig. 6 shows the profiles of TJ boot andouter ring—two members that will constitute a joint.

Previously, an experienced NTN operator has beencommitted to this work, making fine-adjustment effortsbased on his/her experience and including informationon angle of force and location of pivot point.

Therefore, NTN has analyzed boot fitting motions ofits experienced operators to develop numerical data,and has a robot to trace the trajectories of motions ofthese operators. Consequently, NTN has realized areliable robot-assisted rubber boot fitting work for itsCVJ assembly process.

Fig. 6 Shape of TJ boot bore and outer ring O.D.

Boot bore

Fig. 7 Shape of one-touch boot band

Outer ring outer circumference

Small dia. quick-lockingboot band

Large dia. quick-lockingboot band

(1) Exploration-based methodTwo serrated components are brought into contact,

then, while the state of phase-matching is beingexplored, they are snap-fitted with each other. Adelicate sensor reacting to correct phase-matchingbetween both components sends an electrical signalby another sensor which outputs the signal as aninsertion trigger signal.

(2) Measurement-based methodFirst, the profile of serration on each of both

components is measured with a sensor. By referringto the resultant phase information, the serrated shaftend is fitt into the correspondingly serrated bore ofinner ring.

3.3 Automated assembly work for difficult-to-handle-shaped components

(1) Boot band insertion workIn this work, the boot band is fit over the outer

circumference of a particular small diameter section ofrubber boot. Reliability of this operation poses achallenge as two components of very differentshape—rubber boot and sheet metal boot band—arejoined together.

To address this issue, NTN has invented a uniqueassembly system in which a guide jig dedicated toband insertion is installed onto an auxiliary station sothat the band and rubber boot after undergoing shapecorrection are joined together.

Because both components are joined together onlyafter correction for shape, this system positivelyimproves reliability and repeatability of the boot bandinsertion work.

(2) Rubber boot fitting workIn this work, the outer ring is fit into a particular

large diameter section of rubber boot. Note that theboot bore and outer ring outer circumference are eachuniquely shaped: therefore, successful insertion is

(3) Boot band form-fitting workThrough form-fitting, the quick-locking boot band

used in NTN's CVJ secures the rubber boot when itsretainer is appropriately set. Previously, a delicateform-fitting process was necessary for the boot bandmade of thin metal strip (this component can readilyget deformed); therefore, manual operation wasmandatory in order to reliably ensure better qualityand productivity.

Fig. 7 shows shapes of the small and large bootbands.


NTN has attempted to supersede this complicatedmanual operation with a robot-assisted automatedarrangement, and has learned that a special robothand is necessary and exact tracing of necessaryhuman operators’ motions is impossible. Thoughbeginning with the original motions of humanoperators, NTN has revised/reorganized thenecessary motions to develop motions suitable forrobot-assisted automation. Consequently, a robot-assisted rubber band fitting operation has beenrealized which boasts more efficient and reliableoperation compared with procedure that could havebeen achieved by simply tracing human operators’motions.

3.4 Improved efficiency in robot-assistedoperation

To be able to fully utilize robot technology in its CVJassembly process, NTN has developed a uniquecombined robot hand system.

When an industrial robot system is used forautomated production, the robot itself uses a pluralityof robot hands that can be changed between.However, this practice of changing hands leads tounavoidable time loss.

To address this problem, NTN has developed aunique combined multi-functional robot hand system.This robot hand system not only combines a pluralityof robot hands but also is designed such that eachindividual hand when active does not interfere with aCVJ component or peripheral equipment.Furthermore, the individual hands are driven by acommon single drive, allowing the entire system to becompact and light-weight.

Fig. 8 shows the combined robot hand system thatis capable of chucking a boot and a small band at onetime.

To sum up, one robot unit can be applied to aplurality of processes, and previously unavoidabletime loss for hand change has been eliminated.

4. Positive effects of development ofrobot-assisted CVJ assembly line

NTN has adopted its newly developed robot-assisted CVJ assembly technology to all the CVJassembly lines within its factory; consequently, fullautomation of CVJ assembly has been realized. Fig. 9shows a view of the robot-assisted CVJ assembly lineinstalled in NTN factory.

The effects of this development are elimination ormitigation of the burden of highly difficult operationsand high-load operations associated with the transferof heavy work pieces, each being the challenge withconventional assembly practice. They also providedimproved work environments, as well as 60% laborsaving compared with conventional assemblypractices.

In addition, the intelligent part feeder checks theparts while they are loaded into the assembly robot,thereby misassembling the CVJ is prevented,positively helping improve the quality of NTN CVJproducts.


Fig. 9 CVJ robot assembly line

-26-

Fig. 8 Operation of combined robot hand(Upper: boot chuck, Lower: small band chuck)

5. Conclusion

NTN believes that robot-assisted assembly lines willcontinue to be further sophisticated, diversifying intovarious configurations to meet contemporary needs.

While performance of industrial robots has beenincreasingly enhanced, NTN thinks that robot usersneed to deepen their knowledge about robottechnology and improve their skills in utilizing robots.

Thus, NTN will remain committed to thedevelopment of robot technology that can be appliedto not only CVJ assembly processes but also all theproduction processes for NTN products, and wishes toestablish a production system that is safe and posesthe least burden to human workers and helps improvework environments.

-27-

Shigeki MATSUSHITAProduction Engineering

R&D Center

Genji NAGURAProduction Engineering

R&D Center

Photo of authors


-28-



Market Trendsfor Wind Turbine and Bearing Technologies

1. Introduction

The total electricity generation capacity of windpower generation plants worldwide reached 122 GWas of the end of 2008 business year, and according toa report, this value means a 30% increase over theBY2007 level.

Fig. 1 graphically plots the trends in the total windpower electricity generation capacities of the world andmajor participating nations in the BY2004–2008 period.During this period, three major nations—USA,Germany and China—accounted for approximately50% of the total installed capacity. In particular, thetotal installed capacity significantly increased in USAand China in the same period. 50% of the increase innewly installed capacity in the world in BY2008 isaccounted for by USA and China.

Wind power generation, which contributes to the preventionof global warming, is being actively introduced to theworldwide market as a clean energy source. Bearings areincreasing in size accordingly with the growth in the size of

wind power generation equipment. Evaluation tests of bearings that simulate actual performance areimportant for selecting suitable bearings, but they are becoming difficult with equipment that is verylarge, so conducting structural analysis, including for shafts and bearing housings, for example, isnecessary. This article introduces some of these analysis techniques.

Being a clean energy source that does not emit CO2,wind power generation plants are widely appreciatedaround the world, and their installed capacity has beenincreasing by more than 20% every year 1).

As the market for wind power generation plants hasbeen steadily expanding, more advanced wind turbinetechnologies have been increasingly developed. Thecurrent development efforts are focused on scale-upand offshore production of wind turbines. This reportdescribes the current market trend for wind turbinesand typical examples of the newest bearingtechnologies for wind turbines.

2. Market trend

2.1 Scale-upIn 1985, the rated output of a typical wind turbine

with rotor diameter 15 m stood at 50 kW. About twentyyears later, mass-production began with a larger windturbine that boasts a rotor diameter of 126 m and israted at 5,000 kW (5 MW).

In the engineering aspect, the contributing factorsthat have helped realize scale-up of wind turbines areimproved mechanical strength for rotor blades byintroduction of carbon fiber monolithic impregnatedstructure, advancement in mechanical strength designand analysis technique for the structures of tower andnacelle, and progress in design and productiontechniques for producing larger wind turbine bearings.

2004

(MW)

Year

140000

120000

100000

80000

60000

40000

20000

02005 2006 2007 2008To

tal e

lect

ricity

gen

erat

ion

capa

city

Others

ChinaGermanyThe United States

Fig. 1 Cumulative global wind power development

Nobuyuki NINOYU*Souichi YAGI**

Tsuyoshi NIWA***

***Application Engineering Dept. Industrial Sales headquarters***Industrial Sales headquarters***Product Design Dept. Industrial Sales headquarters

Market Trends for Wind Turbine and Bearing Technologies

-29-

Fig. 2 Under construction of offshore wind turbine

Offshore wind farmprojects in Europe (up to 2020)

The United Kingdom : 40 GWGermany : 15 GWNetherlands : 6 GWDenmark : 4 GWBelgium : 2.5 GWSweden : 1.5 GW

Fig. 3 Plan of offshore wind farms in north Europe

Fig. 4 Main shaft model

Research works for super-large wind turbine designsare in progress in particular for offshore wind turbinepower generators that are rated at 10 MW to 20 MW.

2.2 Offshore site productionFew sites are available on land where the wind

profile is appropriate for wind power generation. Forthis reason, offshore wind farms site production hasbeen increasing, especially in Europe. Fig. 2 shows anexample of typical offshore wind farm in Europe.

2.2.1 Offshore wind power generation programsin Europe

In Europe, there is a vast spread of continentalshelves down to a depth of 40 m, and wind farms areoften constructed on seabed foundation. In particular inthe north Europe, there are large scale offshore windfarm projects 2). Fig. 3 maps the currently presentoffshore wind farm projects in the north Europescheduled in the timeframe up to the year 2020.

at seabed) and a sailing type (floating platform structurecan freely move on the ocean surface). Research into asailing type platform is attracting attentions as thisplatform type appears to be promising because offshoreareas in Japan are often much deeper and Japan’slong-established shipbuilding technologies can beapplied to these sailing type structures.

3. Applicable bearing technologies

3.1 Efforts for analysis techniques for large bearings3.1.1 Main-shaft bearing

To cope with the scale-up of wind turbines, the sizesmust be larger for the bearings that support the mainshaft, speed-up gearing and generator. In particular,the larger main-shaft bearings that support the largerrotors often measure 2,000 mm OD or greater, andtheir types include self-aligning roller bearing, doublerow tapered roller bearing, and cylindrical roller bearing.

To be able to determine the final specifications for amain-shaft bearing, an intended bearing needs to betested by running on a test rig that can simulate theintended wind turbine so that the bearing in questioncan stably perform for an extended period. However, asimulation test of the bearing in question on an actualwind turbine requires a large test facility, higher costand longer preparation period before actual test, and,therefore, cannot be readily executed. To address thisissue, NTN has been performing not only simulationtests on an actual wind turbine but also finite elementmethod-based structural analysis for the bearings andauxiliaries including the bearing housing and bearingbed so that a more reliable bearing design is achieved.

Figs. 4 and 5 show a result of the analysis with amain-shaft model. By an external force, the bearing aswell as the bearing housing and shaft get deformed.

A larger wind turbine develops greater deformationowing to the greater external load it receives, andgreater this deformation on the wind turbine moreadversely affects the bearing clearance. Therefore, it isnecessary to take into account this deformation for

2.2.2 Offshore wind power generation programsin Japan

In Japan, application of seabed foundation for windfarms is difficult owing to topographical limitations.Therefore, joint industry-university teams have beendeveloping floating platform wind power generationplants. Note that floating platform plants can becategorized into an anchor-secured type (floatingplatform structure is secured with the anchoring point


-30-

correct analysis of the main-shaft system. NTNperforms design reviews after taking into account theeffect of deformation on the entire wind turbinestructure, and then finalizes the bearing specificationoptimized for the wind turbine in question. Fig. 5 showsan example of the structure for a bearing that supportsthe main shaft in nacelle, wherein the bearing structurecomprises a double row tapered roller bearing in theblade side and a cylindrical roller bearing in the speed-up gearing side. Fig. 6 provides structural diagrams ofthe bearings used in this analysis model.

The diagrams in Fig. 7 show the results of theanalysis of the effect of deformation on the bearinghousing and other structural members. In Fig. 7, theload distribution pattern over the rolling elements isshown for each of two assumptions—wheredeformation is taken into account and where it is nottaken into account.

In FEM analysis procedures, an “elastic body”element is used for the deformation analysis. If theelement is assumed to be a “rigid body”, thencalculations are possible without taking into accountdeformation.

The diagrams in Fig. 7 show that when deformationon a bearing housing and bearing are taken intoaccount, the load zone expands and, consequently,load peak is mitigated, meaning the overall loadimposed over the rolling elements is reduced.Compared with the “solid body” scenario, the bearing

Double row taperedroller bearings

Cylindricalroller bearings

Fig. 6 Bearing structures for analysis

Tapered roller bearing row 1 in Fig. 5: 34 rollers

Tapered roller bearing row 2 in Fig. 5: 34 rollers

1500

400

300

200

100

0

2 34

5

67

89

1011

1213

1415

161718192021

22

2324

2526

2728

2930

3132

33 34Load [kN]

Elastic bodyRigid body

elastic bodyLife ratio = = 1.2

rigid body

Rolling elementNos. (1–34)

1500

400

300

200

100

0

2 34

5

67

89

1011

1213

1415

161718192021

22

2324

2526

2728

2930

3132

33 34Load [kN]

Cylindrical roller bearings row 3 in Fig. 5: 27 rollers

1 23

4

56

78

910

11

1213141516

17

1819

2021

2223

24

2526

271200

800

400

0


elastic bodyLife ratio = = 1.1

rigid body


Load [kN]

Fig. 7 Load distribution of each rolling elements

Blade side

Speed-upgearing side

12

3

Double rowtapered roller bearings

(ID 750, OD1,100)

Cylindricalroller bearings

(ID 710, OD1,050)

Fig. 5 Analysis example of main shaft model

life resulting from the “elastic body” scenario isapparently longer—the life of tapered roller bearing isapproximately 20% longer and that of cylindrical rollerbearing is 10% longer.

It is true that deformation helped increase bearinglife in the example above. However, deformation in

-31-

Fig. 8 Planet gear

other scenarios can lead to reduced bearing life.Therefore, NTN is executing deformation analysis forthe entire wind turbine system in order to design anoptimal bearing for the main shaft on wind turbine.

3.1.2 Planet bearing for speed-up gearingFigs. 8, 9 and 10 provide information obtained from

structural analysis with bearings incorporated intospeed-up gearings in larger 2-2.5 MW class windturbines. Note that through this analysis, thespecification for relevant planet bearing has beendeveloped while taking into account the magnitude ofpossible deformation on the carrier, pinion shaft andplanetary gearing.

Fig. 9 shows an analysis example of the deformationmode with planetary gearing wherein gear load andcentrifugal force are imposed onto the planetarygearing.

In the main mode, the force resulting from twotangential loads Ft shown in Fig. 7 acts on the bearingas a radial load. Consequently, the outer ring isdeformed by two radial loads Fr, thereby the load zoneexpands.

Input torque

Pinion shaftDouble row cylindrical roller bearings

Planet gear

Rows 12 34

Carrier

Fig. 9 Analysis example of planet gear

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Ft

Ft

Fr

Fr

Centrifugalforce

Fig. 10 Load distribution of each low

1

1 1

Load [kN]


First row Second row

Third row Fourth row

23

45

67

8

91011

12

1314

1516

17

181980


elastic bodyrigid body

Total life ratio = = 1.5

60

40

20

0

23

45

67

8

91011

12

1314

1516

17

1819

1 23

45

67

8

91011

12

1314

1516

17

181980

60

40

0

80

60

0

40

23

45

67

8

91011

12

1314

1516

17

181980

60

0

40

20

20

20

Fig. 10 schematically illustrates the calculationresult. If the deformation on the carrier and planetarygear resulting from the input torque is taken intoaccount, the load acting on the rolling elements in rows1 through 4 on the two double row cylindrical rollerbearings is mitigated, compared with a case wheredeformation is not taken into account. Compared with acase where deformation is not taken into account, thecalculated bearing life with deformation being takeninto account more closely matches the actual lifeobtained from the bearings installed to an actual windturbine. In our test, the life of the two double rowcylindrical roller bearings with deformation taken intoaccount was approximately 50% longer compared witha case where deformation was not taken into account.


3.1.3 Thermal analysisWhen assessing the performance of bearings

incorporated into larger wind turbines, thermaldeformation needs to be taken into account. Fig. 11shows an example of a bearing model that wassubjected to our thermal analysis. In the analysis work,the temperature distribution over the entire model wasdetermined based on the heat release performance ofthe bearing as a heat source. Through analysis for acombination of heat and structure, it will be possible todevelop bearing specifications that better reflectoperating conditions of bearings installed on actualwind turbines. Note that in the thermal analysis work,the thermal characteristics including the heat transfercoefficient and heat release coefficient are affected byvarious factors including surface properties of thebearings, the weather, the ambient temperature, thehumidity and the wind velocity. Therefore, we haveattempted to improve precision of thermal analysis byfeeding back data obtained from field tests. This effortis later described.

-32-


Table 1 Test result

Analytical value

Displacement on point A (mm)

Measured value

Difference

Pre-run Post-run

0.403 0.326

0.424 0.259

5% 25%

3.2 Simulation on actual wind turbinesNTN owns and operates a unique test facility that is

capable of simulating performance of super largebearings for the main shaft of a 2–2.5 MW class windturbine while subjecting the bearing to a load that willoccur on an actual wind turbine. Utilizing this facility,NTN has been performing tests on the bearingssimultaneously with FEM analysis to improve precisionof FEM analysis, so that the bearing developmentperiod is shortened.

Fig. 12 shows an example of results of the FEManalysis with a bearing model (double row taperedroller bearing: 2,000 mm pitch diameter with roller set)for 2–2.5 MW class wind turbine, wherein a momentload of 1700 kN-m was applied to the inner ring.

Table 1 summarizes the information about acomparison between analytical results of FEM and

measured value on a particular point (point A in Fig.12) where the inclination of the inner ring is greatest.

In the test result data in Table 1, the differencebetween the pre-run analytical value and measuredvalue is 5%, which qualifies as a fairly good match. Incontrast, the difference between a post-run analyticalvalue and a measured value is as great as 25%. Aresult of a post-run analysis is provided below. For thisanalysis, the heat transfer coefficient was determinedbased on the temperature rise on the inner and outerrings, and the coefficient was applied to the thermalanalysis about the increase in bearing preload andchange in bearing rigidity resulting from operation ofthe bearing.

The bearing housing in Fig. 13 houses a double rowtapered roller bearing. This analysis model wascontinuously run thereby resulting in a temperature

Fig. 11 Example of heat transfer analysis

Blade side

Reducer side

15

20

25

30

35

40

45

50

Fig. 12 Example of FEM analysis

0.50mm

0.40

0.30

0.20

0.10

-0.10

-0.20

-0.30

-0.40

-0.50

0.0

Point A

Fig. 13 Analysis model

Temperature distribution onsurface of bearing housing50

45

40

35

30

25

20

15

Directly above double rowtapered roller bearing

distribution on the bearing and bearing box to beanalyzed.

Incidentally, for an estimation of temperaturedistribution on the bearing through thermal analysis,tuning (a technique that approximates analytical resultsto measured values obtained by experiment) wasemployed. More specifically, with this technique,temperatures measured on various test points on thebearing of an actual wind turbine are incorporated intoanalytical results to adjust the heat transfer coefficientthereby increasing the reliability of the temperaturedistribution analysis.

The example of bearing temperature distributioninside a bearing housing in Fig. 14 reflects measuredvalues for ambient temperature, nacelle insidetemperature, and temperature on the outercircumference surface of the outer ring. Table 2summarizes information about the comparisonbetween analytical result of thermal analysis andmeasured values.

Based on the estimated analytical value oftemperature difference across the inner ring and outerring, as given in Table 2, the decrease in bearingclearance on the running bearing was estimated, andthen, adequate initial bearing clearance has beenaccordingly determined. As a result, the differencebetween analytical value and measured value hasdecreased to 1%.

Considering the findings obtained from theseanalytical results, NTN has developed an optimalbearing design that has been installed in an actualwind turbine and is reliably running.

4. Conclusion

Wind power generation as a clean energy sourcehas been increasing its output. The quantity of installedoffshore wind turbine plants will further increase. At thesame time, wind turbines will increasingly featuregreater output and larger size, while they are requiredto withstand severer natural elements more reliably.

In this context, the bearing designs adopted for windturbines need not only to satisfy conventionalstandards but also to be optimized for intended windturbines through wind turbine-specific detailed analysistechnique and review of analysis result.

Wishing to contribute to development of wind powergeneration which is a new energy source possiblyhelping prevent global warming, NTN will furtherdevelop and stably supply highly reliable and durablebearing products.

References1) BTM Consult ApS International Wind Energy

Development World Market Update 20082) SCOTTISH Development International: Offshore Wind

Power Generation in Scotland and the UK

-33-

Photo of authors

Nobuyuki NINOYUApplication Engineering Dept.Industrial Sales headquarters

Souichi YAGIIndustrial Salesheadquarters

Tsuyoshi NIWAProduct Design Dept.

Industrial Sales headquarters

Table 2 Bearing temperature distribution

Analytical value

Measured value

Difference

Temperature differenceacross inner ring

and outer ring (˚C)

Temperature on outercircumference surface

of outer ring (˚C)

2.2 50.5

Not measured 50.0

－ 1%(Post-tuning difference)

Fig. 14 Example of thermal analysis

Double row tapered roller bearingHousing

Shaft


-34-



Eco-friendly Product Developmentat SNR Roulements

1. Preface

CO2 emissions control has been posing a challengeto any manufacturer in attempting to addressenvironmental issues including global warming. To beable to contribute to CO2 emissions reduction, theautomotive industry has been committed to design andproduce lighter, lower-torque automotive parts to helpimprove the fuel efficiency of cars they manufacture. Inthis article, SNR will present its technologies for lighterweight and lower torque in their automotive hubbearings transmission bearings and suspensionbearings.

2. Efforts for lighter bearing designs

This section describes the efforts of SNR for itsautomotive hub bearing (hereinafter H/B) products andwater pump bearing products.

2.1 Automotive hub bearingsBased on a novel concept, SNR has developed

unique H/B designs that can be applied to flanged 2ndgeneration and 3rd generation hub bearings forsupporting brake disks and wheel rims. Throughoptimized shape design for H/B, the newly developedH/B products boast 15% weight reduction whilekeeping excellent strength and durability.

The reduction of CO2 emissions is being promoted in Europe, and every automobile company isconsidering various measures to achieve this goal. "Mass reduction" and "torque reduction" areeffective ways to decrease CO2 emissions, and automobile and parts makers are challenging thelimits every day, including awith existing products. SNR products that contribute to "mass reduction"and "torque reduction" are introduced in this article.

High stress region

Low stress region

Fig. 1 Stress analysis on the hub bearing

Siegfried RUHLAND***Ludovic SAUNIER***

Cynthia TSEN***Bernard LIATARD***Gérald MIRABEL***

*SNR Product Innovation & Mechatronic **SNR Automotive Transmission Engineering ***SNR Automotive Equipment Engineering

2.1.1 Design activities for automotive hub bearingsUnder the loading conditions (rotational bending

force, hub bolt tightening torque, etc.) possibly actingon bearings installed in actual cars, we have analyzedthe stresses occurring on H/B. Fig. 1 shows ananalysis model used. As can be understood from thisdiagram, we have learned that the maximum stressoccurs on both the flange tightening area and thechamfer area at the base of flange. Therefore, whendeveloping the new bearing design, we have shifted aportion of the mass of the hub from a low stress regionto a high stress region to mitigate stress concentrationon the hub while not altering the design of critical areasincluding the raceway surface, disk and wheelreceiving area.

Eco-friendly Product Development at SNR Roulements

-35-

Fig. 2 Conventional design (left) and new optimizeddesign (right)

35% reductionin stress

Conventional design New optimized design

Str

ess

Fig. 3 Stress analysis results

Fig. 4 View of Water pump

Load from belt

Integral shaft bearing

Fig. 5 Water pump structure

Fig. 2 shows shapes of a conventional design andthe new optimized design. The new hub designfeatures a unique flange shape, each of its bolt-downsegments being reinforced with two ribs. With thisdesign, the mass of outer ring has been reduced by20% and that of entire H/B by 15%. This weightreduction will mean a maximum weight reduction of1,300 g per automobile axle.

2.2 Water pump bearing for automobileFig. 4 shows an appearance of a conventional water

pump for cooling an automotive engine, and Fig. 5schematically illustrates the structure of the waterpump. Many present-day water pumps used for coolingmodern automotive engines incorporate integral shaftbearings that have a raceway surface on their shaftswith their inner ring deleted.

With a conventional structure in Fig. 5, the belttension poses an offset load onto the bearing andcauses the contact pressure to be greater. Thisproblem in turn poses a challenge in providing longerlife for the bearing. At the same time, resultantmisalignment of the bearing relative to the shaft cancause water to pass the seal and enter the bearing.Consequently, the grease can leak out of the bearing,leading to poor lubrication of the bearing. The bearingcan then develop abnormal noise and/or exhibitexcessively short life.

20% weightreduction

2.1.2 Performance of new optimized designAs described above, the new optimized design

boasts a much reduced weight, while its rigidity isequivalent to that of conventional designs.Furthermore, because the bolt-down segments arereinforced with ribs, the stress occurring at the flangebase of the new design is 35% smaller as shown inFig. 3.

Drain hole

Pulley

Drive shaft

Hub (pulley sheet)

Seal Pump rotor

Pump body


-36-

2.2.1 Structure of new bearingIn order to achieve lower running noise, lighter

weight and longer bearing life, we have developed anovel external bearing that has a shaft integrated witha flange. Fig. 6 shows a view of this novel externalbearing, and Fig. 7 schematically illustrates thestructure of this bearing. This structure ensures thatthe tension from the drive belt acts on the center ofbearing, helping prevent misalignment of the bearingrelative to the shaft. Compared with conventionaldesigns, the contact pressure between the balls andraceway surface on this bearing is lower, leading toimproved bearing durability. In addition, because ournew bearing is free from bearing-shaft misalignmentwhich will develop moment load, the balls will roll at thebottom of groove, resulting in lower running noiseoccurrence.

Fig. 8 Process of water pump bearing

Fig. 6 View of external bearing

Load from belt

Flange

External bearing

Shaft

Fig. 7 Structure of external bearing

Improvements• Greater load carrying capacity with bearing• Pump capacity• Bearing life• Noise level• Weight

451g

571g

2004 2008

Op

tim

izat

ion

of

wat

er p

um

p

Greater pump capacity Drive belt for auxiliaries is directly engaged with the outer ring; this arrangementhelps realize a space-saving pump design.Use of larger balls helps reduce contact pressure, mitigating adverse effect ofpress-fitting operation.21%

Longer life

Weight reduction

Years

Through improvement in our production technology,we have integrated the bearing and flange with theshaft, thereby we have successfully achieved anincrease in load carrying capacity of bearing, increasein pump capacity, longer bearing life, lower noiseemission and lighter weight. Fig. 8 summarizeschanges in appearance and weight of water pumpbearing over several years.

2.2.2 Determination of contact pressure occurringbetween balls and raceway surface

So that a water pump on an automobile remainsoperative without any need for repair until theautomobile reaches the end of its useful service life,the life of water pump bearing needs to satisfy 3×105

-37-

km of automobile’s total travel distance. In order for agiven water pump bearing design to satisfy this liferequirement, its internal design needs to be finalized sothat the contact pressure at any portion in it does notexceed fatigue limit of that portion. To this end, it isalso important to optimize tightening allowance of theseal, amount of prefilled grease, and location ofprefilled grease.

We have studied various parameters includingbearing life, temperature, parts tolerances, minimumand maximum clearances, and tightening allowance,executed functional test, and various tests with thepump lifetime tester shown in Fig. 9, thereby we haveestablished the design standard that satisfies thebearing life requirement.

Fig. 9 View of pump lifetime test

Fig. 10 Autonomous bearings using solid lubricant

3. Efforts for bearing products of lowertorque

This section describes our efforts for low torquedesigns with automotive transmission bearings, hubbearings and suspension bearings.

3.1 Automotive transmission bearingsVarious improvement activities have been performed

by people involved in automotive technologies to helpimprove operability of automotive transmissions andreduce CO2 emissions from automobiles. To promotevarious automotive transmission systems, SNR hasbeen committed to lighter weight, higher efficiency(lower torque) transmission bearing designs.

Approx. 50% of torque loss occurring on a givenautomotive transmission results from stirring-inducedresistance with lubricating oil. Therefore, anappropriate choice of lubricating oil is one of importantconsiderations in achieving lower torque on anautomotive transmission, and a currently more favoredcombination is a use of minimum amount of lowviscosity lubricating oil. Incidentally, efficiency of an Fig. 11 Transmission simulation

automotive transmission is greatly affected by torqueloss occurring from an inactive idler gear: therefore,bearing torque poses a critical factor in the torque losson the transmission.

The efforts SNR has so far made with itstransmission bearings in order to reduce CO2

emissions are as follows:(1) The inner and outer rings and rolling elements of

our bearings made of the standard steel material100Cr6 are now subjected to a special heattreatment process. Consequently, thesecomponents boast longer life even under severelubricating conditions.

(2) Development of unique self-lubricating bearings forautomotive gearboxes (Fig. 10 shows examples ofself-lubricating bearing lubricated with solidlubricant).

(3) Decreased torque by superseding sliding bearingsfor idler gear with rolling bearings such as cagedneedle roller bearing.

(4) Optimized bearing design through analysis for loadand torque on transmission bearing by using newlydeveloped engineering computation software (asimulation model is shown in Fig. 11).

(5) Reduced frictional torque through improved internaldesign for bearings.


3.2 Low torque ball bearingsTo reduce friction-induced torque loss on an

automotive transmission, use of a ball bearing ratherthan a tapered roller bearing is advantageous.However, since load carrying capacity of a ball bearingis smaller compared with a tapered roller bearing, thesize of ball bearing tends to be larger so that thebearing can satisfy targeted life. Incidentally, theordinary fracture mode of transmission bearing issurface damage starting from a dent mark on theraceway owing to lubricant contaminated with foreignmatter, rather than fatigue failure of bearing material.Therefore, to prevent ingress of foreign matter into abearing to ensure longer bearing life, a bearing havingfilter seals shown in Fig. 12 is used. Filter seals needto be optimally designed so that they not only providegood filtering performance but also allow lubricating oilto be introduced into the bearing to lubricate and cooldown the bearing.

Many bearings with filter seals feature an increasedbearing width dimension to accommodate an additionalelastomer filter seal; however, this configuration leadsto increased frictional torque on bearing. A noveldesign developed by SNR, as shown in Fig. 13 has aclipped double polyamide cage that has filteringfunction. The advantages of this design are as follows:

(1) Compact and light-weight design is possible.(2) Labyrinth clearance between the cage and

raceway can be controlled at higher precision.(3) Provision of two oil inlets helps increase

lubricating oil flow into the bearing.(4) Reduced cost

The result of comparison between our newlydeveloped double seal design and a conventional sealdesign is described below.

Fig. 14 illustrates results of measurement of oil flowthrough a bearing under oil bath lubricationenvironment. The data mapped along the vertical axisrepresent filtering effect and oil flow that passes a filterand contribute to lubrication. Compared with aconventional filter seal, our newly developed doublefilter seals boast better lubrication efficiency as moreoil flows through the bearing though the filtering effectis roughly same.

Fig. 15 illustrates a result of comparing runningtorque of our newly developed bearing with doublefilter seals to that of conventional bearing. Torqueoccurring inside the bearing including rolling resistanceis virtually the same with both bearing types;notwithstanding, the total torque on our new bearingdesign has been reduced by 65% because the torquesoccurring on the seals has been greatly reduced. Also,heat buildup in our bearing is mitigated; consequently,

-38-


Fig. 12 Bearing with filter seals

Fig. 14 Efficiency against pollution

Filt

erin

g ef

fect

and

oil

flow

Filtering effect

Oil flow

Filter seal (conventional

design)

Doublefilter seals

(new design)

Fig. 15 Torque test results

Run

ning

torq

ue

N・m

65%reduction

Bearing interior

Seal

Conventionaldesign

New design

0.350

0.300

0.250

0.200

0.150

0.100

0.050

0.000

Fig. 13 Bearing with double filter seals into a clippeddouble polyamid cage

Oil flow

Clipped doublepolyamide cage

15.5

our bearing boasts improved seizure resistance.Furthermore, filtering function of our bearing is virtuallyequivalent to that of conventional bearing design, andthe width of our bearing can be smaller. Our bearingdesign can be applied to produce compact, light-weighttransmission bearing products.

3.3 Low torque tapered roller bearingsTapered roller bearings are often used as

transmission bearings because they can carry agreater combined axial and radial load compared withother bearing types of similar size. However, reductionin frictional torque occurring on this bearing type posesa challenge.

The SNR tapered roller bearings for automotivetransmissions boasts decreased frictional torquethrough adoption of NTN’s design and productiontechnologies for low torque tapered roller bearings.The advantages of these SNR products can be definedas follows:(1) Special heat treatment process helps improve

fatigue strength and resistance against dent markthat can result from foreign matter in lubricant.

(2) Adoption of polyamide cage in which shear frictionwith oil is lower, compared with that in standardsteel sheet cages

(3) Improved design for bearing interior in order toimprove load carrying capacity and rigidity andmitigate misalignment and frictional torque(optimized number and size of rollers, and specialcrowning on raceway surface and rolling surface)

(4) Optimization of contact locations between inner ringflange surface, rollers, and raceway surface

(5) Reduction in surface roughness of inner ring flangesurface and roller end faces to reduce frictionaltorque on roller end faces

Fig. 16 graphically plots information aboutcomparisons of SNR's tapered roller bearing fortransmission and a competitor’s tapered roller bearing,wherein the running speeds data of these bearings areplotted along the horizontal axis. As can be understood

from this diagram, the bearing products of SNR boast25 to 50% reduction in torque loss, compared withcompetitor’s products.

25% reduction in torque loss with transmissionbearings means 2 g/km reduction in CO2 emissions pervehicle. In addition, use of low torque tapered rollerbearings helps improve seizure resistance as well asquality of shift-change operation.

3.4 Automotive hub bearingsAbout 50% of frictional torque occurring on a wheel

bearing results from sliding friction on contact surfaceof seal(s). Therefore, the use of hub bearings featuringreduced sliding resistance is effective in reducing theCO2 emissions from the vehicle that incorporates lowfriction wheel bearings. However, reduced frictionalresistance on a given seal often means jeopardizedsealing performance of that seal. Therefore, there hasbeen a challenge of meeting conflicting needs forreduced frictional resistance and improved sealingperformance.

SNR has developed a unique tribological effect-capable seal that boasts approx. 11% reduction insliding resistance and 22% improvement in sealingperformance, by improving lubrication quality on thesliding surface of seal lip. As SNR has applied theabove-mentioned development concept to its currentlypresent seal products, cost reduction and reduceddevelopment lead time have been achieved.

3.4.1 Advantages of tribological effect-capableseal products

On our newly developed tribological effect-capableseals, the lip sliding surface of the slinger is uniquelyshaped so that oil film is readily formed on it.

-39-

Fig. 16 Torque measurements with rotations

00 200 400 600 800 1000 1200 1400 1600 1800 2000

0.2

0.4

0.6

0.8

1.0

Running speed min-1

Run

ning

torq

ue

N・

m

Competitor’s tapered roller bearing

SNR low torque tapered roller bearing


Consequently, lower torque and mitigation of heatbuildup have been achieved. Fig. 17 illustrates across-sectional plan of our seal, and a view of thecontact area of the slinger. Micro-pores present in thiscontact area help form an oil film between the seal lipand slinger, and the oil film reduces friction on the seallip. Furthermore, these micro-pores prevent wearcaused by ingress and trapping of foreign matters.

3.4.2 Performance of tribological effect-capable sealFigs. 18 and 19 show running torque and test

results of a muddy water bearing life test withconventional seal and our newly developed tribologicaleffect-capable seal. Owing to reduced resistance onseal, the running torque has decreased by 11%, andthe muddy water bearing life has increased by 22%.

3.5 Automotive suspension bearingsTo be able to operate without problems under

severe operating conditions (environments such asmuddy water), a sealed strut bearing needs to featureimproved reliability and sealing performance. Inchallenging this issue, SNR has developed a uniquelow torque sealed strut bearing complete with a floatingseal in order to satisfy requirements for both animprovement in running torque and sealingperformance (Fig. 21).

3.5.1 Features of sealed strut bearingsMuddy water resistance of a bearing is governed by

lip performance of its seal. Therefore, it is important todevelop a seal that is capable of satisfying both bettersealing performance and stabler running torqueperformance. To address this challenge, we haverecently developed a novel low torque sealed strutbearing complete with a floating seal. Fig. 20 shows aconventional design, and Fig. 21 illustrates our newdesign. For comparison purpose, Fig. 22 shows astructure of a different design, that is, an overmoldedlip seal.

-40-


Fig. 17 Seal shape on the slinger

Micro-pores

Enlarged view ofcontact area

on slinger

Fig. 20 Usual design

Fig. 21 SNR design: floating seal

Fig. 22 Competitor design: overmolded lip seal

Fig. 18 Comparison of rotational friction betweendevelopment seal and current one

Conventionaldesign

Tor

que

N・

m

Tribologicaleffect-capable seal

11% reductionin runningtorque

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

Seal areaBearing area

Fig. 19 Comparison of muddy water resistance betweendevelopment seal and current one

Conventional designMud

dy w

ater

bea

ring

life

h

Tribologicaleffect-capable seal

700

600

500

400

300

200

100

0

22%

3.5.2 Test resultsFig. 23 illustrates the contribution of a seal onto the

running torque of a strut bearing. Compared with aconventional design, the running torque on our newlydeveloped floating seal (SNR design in Fig. 21) is 18%greater; compared with an overmolded lip seal, therunning torque on our new seal design is 50% smaller.

Fig. 24 provides results of muddy water test.Compared with the conventional design, our newlydeveloped floating seal (SNR design in Fig. 21) boastsgreatly improved sealing performance that helpspositively reduce water ingress into the bearing.Though excellent in initial sealing performance, theovermolded lip seal experiences wear of theovermolded lip: the SNR’s newly developed designboasts 38% smaller water ingress.

4. Afterword

This article has provided an overview about lighter-weight, lower-torque, and eco-friendly automotivebearing technologies of SNR. We have beencommitted to development of new technologies andproducts, aiming at further progress in light-weight andlow-torque bearing designs, and, at the same time,remains involved in long-term projects for hybridelectric vehicles and pure electric vehicles.

-41-

Photo of authors

Siegfried RUHLAND

SNR Product Innovation& Mechatronic

Ludovic SAUNIER


Cynthia TSEN


Bernard LIATARD

SNR AutomotiveTransmission Engineering

Gérald MIRABEL

SNR AutomotiveEquipment Engineering

Fig. 23 Torque test results

Seal area

Bearing area

Run

ning

torq

ue

50% reductionin running torque

Usualdesign

SNRdesign

Overmoldeddesign

Fig. 24 Muddy water test results

Red

uctio

n in

wat

er in

gres

s %

38% reductionin water ingress

Usualdesign

SNRdesign

Overmoldeddesign


-42-


[ New Product ]

Actuator for Electromechanical Brakes

1. Introduction

The brake system plays a critical role in the safeoperation of any automobile. Recently, carmanufacturers have been improving the safety of theircars through improvements in ABS, ESC and brakeassist systems that have resulted from improvementsin hydraulic control technologies1), 2).

Incidentally, while concerns about the globalenvironment have been mounting an increasingnumber of electric motor-driven cars such as HybridElectric Vehicles (HEV), which boast better fueleconomy, and Electric Vehicles (EV), which do not useany fossil fuels, have been marketed and will see everincreasing demand.

On HEV’s and EV’s a regenerative braking systemrecovers energy during the deceleration process byallowing an electric motor to function as a generator.To be able to recover energy more efficiently, it isnecessary to further improve control over mechanicaland regenerative braking systems. In addition, on anHEV or EV, the negative pressure occurring on theengine is insufficient or cannot be used; therefore, if aconventional braking system is used a separatenegative pressure generating system will be needed.

To sum up, it is difficult with conventional brakingsystems to readily solve difficult issues, such as brakeswith more sophisticated functions and improved eco-

Electromechanical brakes (EMB) have some advantages, such asvehicle safety improvement and simple system. In electric vehicles andhybrid electric vehicles, which have received much attention in recentyears, applying EMB and controlling them in coordination withregenerative braking is expected to improve fuel economy. On the otherhand, it is desirable that the brakes be compact and lightweight, becausethey are installed under the springs of the vehicles. NTN has developeda small actuator for EMBs with our original linear motion device.

friendliness. There has been a mounting need in themarket for an electrically actuated electromechanicalbraking system, as a means for solving these issues.However, conventional electromechanical brakingsystems, which use a ballscrew or ball-ramp for a linearmotion device, require a reducer mechanism to obtaina greater reduction ratio; thus, a compact actuatordesign for this purpose has been difficult to realize.

By using its propriety linear motion mechanism, NTNhas developed a compact actuator forelectromechanical brakes 3). In this paper, we provideinformation about our actuator for an electromagneticbrake, whose performance has been improved byoptimization of its internal design.

2. Structure of actuator for electro-magnetic brake

2.1 Linear motion device(planet roller screw mechanism)

The constitution and operating principle of our linearmotion device are hereunder described. As shown inFig. 1, our linear motion device comprises of a sunroller, planet rollers, outer ring, carrier, support pins,springs, and bearings. The planet rollers arecircumferentially arranged at equal intervals betweenthe sun roller that functions as an input means and theaxially sliding outer ring that functions as an output

Tatsuya YAMASAKI**Masaaki EGUCHI**Yusuke MAKINO**

**Automotive Module Product Development Dept. New Product Development R&D Center**Mechatronics Research Dept. New Product Development R&D Center


-43-

Support pin support bearing

Revolutionsupport bearing

Rotation

Projection

Caliper (housing)

CarrierSprings

Friction-typetransmission

Rotation of sun roller

Revolution of planet rollerRotation of planet roller

Planet roller

Sun roller (input)

Outer ring (output)

Fig. 1 Schematic of linear motion device

Thrust

Load exerted by the sun roller

Load exerted by the outer ring

Fig. 2 Load acting on planet roller

means. Torque on the sun roller is transmitted to theplanet rollers through friction. The outer circumferenceof each planet roller has an external thread orcircumferential groove that engages with an internalthread on the bore surface of the outer ring; whereinthe pitch of the internal thread on the bore surface ofthe outer ring is the same as that of the external threador circumferential groove on each planet roller but thelead of internal thread on the outer ring is reversed tothat of the external thread or circumferential groove oneach planet roller. The carrier that supports the planetrollers is supported by the caliper (housing) such that itcan rotate, but does not slide in the axial direction.With this construction, when the sun roller rotates theplanet rollers revolve while rotating, thereby the rotarymotion of the sun roller is finally translated into theaxial slide motion of the outer ring.

Next, let us describe a means for providing a normalload that is necessary to permit a friction-typetransmission of torque from the sun roller to the planetrollers. In a report from a previous issue of the NTNTechnical Review, the necessary normal load wasprovided by positioning the planetary rollers betweenthe sun roller and outer ring by a shrink-fittingtechnique 3). With our new development, the planetaryrollers are forced into contact with the sun roller bymeans of a spring force being applied by the springsconnected to both ends of the support pin. At the sametime, the contact areas between the planetary rollersand the outer ring are each designed to form a ramphaving a particular flank angle; consequently, when anaxial load acts on the outer ring as shown in Fig. 2, theramp-type contact surface exerts a load that forces thecorresponding planet roller toward the sun roller.

This structure helps mitigate the effect of adimensional change owing to a worn torquetransmission surface to the normal load; thereby, a

stable load acts along a normal line. In addition,through optimization of the flank angle between theplanet roller and the outer ring, reliable torquetransmission is achieved without causing excessiveslippage between the sun roller and the planet rollerseven when a very high load is acting on the sun rollerand the planet rollers.

2.2 Electromagnetic brake unitFig. 3 schematically illustrates our electromagnetic

brake unit designed for the front wheels of a 1500 cc-class vehicle. To configure this brake unit, we haveincorporated our linear motion device, described inSec. 2.1, into the caliper together with an electric motorsuch that the entire axial length of the brake unit isshorter. The linear motion device and the driving motorare arranged in parallel, where the motor transmitsdriving power to the linear motion device through agear train. Main specifications of this new EMB unit aresummarized in Table 1. Note that the driving motor forthis actuator has been designed by NTN Fig. 4illustrates the structure of this motor, and Table 2summarizes major specifications of this motor.


-44-

3. Performance

3.1 EfficiencyFig. 5 illustrates the interrelation between motor

torque, thrust force and efficiency. Note that thecalculated values in Fig. 5 have been determined byapplying the efficiency calculation method presented ina previous issue of NTN Technical Review 3). Theexperimental values fairly match the calculated values.Thus, our EMB unit performs as designed.

Table 1 Specifications of EMB unit

Table 2 Specifications of motor 3.2 Thrust variation rateBy adopting the caliper shape shown in Fig. 3 and

by using a pad and disk used on an actual brakesystem, the variation rate of thrust was assessed. Fig.6 shows the results of the assessment obtained byapplying a constant voltage (12 V) to the EMB unitplaced in an ambient temperature of 20˚C. Table 3summarizes the response times and thrust variationrates defined by expressions (1) through (4).

Brake pad Linear motiondevice

Brake disk

Wheel

Caliper (housing)Gears

Motor84.5mm 137mm

70m

m

Fig. 3 Schematic of EMB unit

Load retainingfunction.

Specification RemarksCharacteristic

Max. thrust force

Size(excluding pad clamp)

Mass(excluding pads)

Lubrication system

30kNEquivalent to that of front wheelsfor 1,500 cc-class vehicleThis setting may be changed to“Yes” by appropriately alteringthe specifications for threadingfor planet rollers and outer ring.

None

137mm×70mm×84.5mm －

5.6kg －

Grease lubrication －

Fig. 4 Structure of motor

Thr

ust

kN

40

30

20

10

0

80

60

40

20

00 0.2 0.4 0.6 0.8 1

Effi

cien

cy

%

Torque Nm

ThrustEfficiency

Calculated Experimental

Fig. 5 Efficiency of EMB unit (developed)

Stator

Output shaft

Rotor

54.5

φ70

Size

SpecificationCharacteristic

Type

Voltage applied

Max. running speed

Max. torque

DC brushless

φ70×54.5

12V

5000min-1 (w/load)

0.8Nm

-45-

Thr

ust k

N, V

olta

ge a

pplie

d v

40

30

20

10

00 0.1 0.2 0.3-0.1

Fa

0.75Fa

t0 t1 t2t1'

Time sec

Thrust Voltage applied Thrust Voltage applied

Thr

ust k

N, V

olta

ge a

pplie

d v

40

30

20

10

00 0.1 0.2 0.3

Time sec

t3 t4 t5

(a) During thrust increase (initial clearance: 0.0 mm) (b) During thrust decrease

Fa

Fig. 6 Thrust variation rate of EMB unit (developed)

ΔtINC＝t1－t0 ……………………………………… (1)

ΔtDEC＝t5－t3 ……………………………………… (2)

0.75・FaVINC＝―――― …………………………………… (3)

t1'－t1

FaVDEC＝――――………………………………………(4)

t5－t4

Fa ：Max. thrust force = 30 kNt0 ：Actuation start point (during thrust increase)t1' ：Thrust increase start pointt1 ：75% maximum thrust reached point

(during thrust increase)t2 ：Maximum thrust reached pointt3 ：Actuation start point (during thrust decrease)t4 ：Thrust decrease start pointt5 ：Brake release end pointVDEC ：Thrust variation rate (thrust decrease rate)VINC ：Thrust variation rate (thrust increase rate)ΔtDEC：Time (during thrust decrease)ΔtINC：Time (during thrust increase)

Table 3 Response time and thrust variation rate

During thrustdecrease

Thrust variation rateTimeCharacteristic

During thrustincrease

0.172 s(=ΔtINC)

0.146 s(=ΔtDEC)

170 kN/s(=VINC)

221 kN/s(=VDEC)

Table 4 Durability test

High thrust durability

Normal temperatureactuation durability

High temperatureactuation durability

Vibration durability

Test parametersCharacteristics tested

Torque durabilityBraking torque Equivalent to 0.6GNumber of braking cycles 20×104

Magnitude of thrust 30 kNNumber of thrust applications 1×104

Temperature 4～35˚CMagnitude of thrust 15 kNNumber of thrust applications 50×104

Temperature 120˚CMagnitude of thrust 15 kNNumber of thrust applications 7×104

Vibration acceleration ±20G (vertical direction)Vibration frequency 60 HzNumber of vibration applications 500×104

3.3 DurabilityTable 4 summarizes the characteristics tested in the

durability test that has been performed in accordancewith JASO C 448-89 that specifies the bench testmethod for the disk brake caliper assembly of apassenger car. Our new design has satisfied thedurability requirements for all the characteristicssummarized in Table 4.


4. Conclusion

This paper has presented our unique actuator, whichincorporates NTN’s propriety linear motion device for acompact electromagnetic brake.

As electric motor-driven cars such as HEV and EVbecome more commonly used, there will be anincreasing need for electromechanical brake systems.NTN will further improve durability and response speedof the actuator so that it can be reliably used onelectromagnetic brakes.

References1) Tomohiko Adachi: Trends in Electronic Stability

Control (ESC) Systems, Journal of the Society ofAutomotive Engineers of Japan, Vol. 60, No.12 (2006)28-33 (Japanese)

2) Bo Cheng, Tetsuo Taniguchi, Tadashi Hatano, ToshiyaHirose: Effects of Brake Assistance Systems inEmergency Situations, Proceedings of the Society ofAutomotive Engineers of Japan Annual Congress No.20065894 (2006) (Japanese)

3) T. Yamasaki, M. Eguchi and M. Makino: Developmentof an Electromechanical Brake, NTN TechnicalReview No. 75 (2007) 53-61

-46-


Photo of authors

Tatsuya YAMASAKI

Automotive Module ProductDevelopment Dept.

New Product DevelopmentR&D Center

Masaaki EGUCHI

Automotive Module ProductDevelopment Dept.

New Product DevelopmentR&D Center

Yusuke MAKINO

Mechatronics Research Dept.New Product Development

R&D Center

-47-


[ New Product ]

Electric Ball Screw Actuators for Automobiles

1. Introduction

Control-by-wire technologies for automobiles arebecoming more commonly adopted with one typicalexample being the by-wire throttle control system.Recently though, certain cars have adopted brake-by-wire technology. Control-by-wire technologies havebeen ever evolving in an effort to achieve bettercomfort and safety in operating vehicles.

Recently, NTN has developed a high-response,high-thrust, electrically driven ballscrew actuator forautomobiles (hereinafter referred to as “actuator”) thatcan be adopted for control-by-wire systems.

This paper hereunder describes the structure andfeatures of this actuator.

In the automotive market, many new hybrid cars and low -fuel -consumption and low -emission engines have been developed toreduce CO2 and to produce cleaner exhaust. At the same time,many projects are advancing to achieve more car amenities and

safer driving using quicker and more reliable electric motor drives. NTN had already developed anew electric ball screw for use in automatic manual transmission and engine control. Building onthis, we have developed an electric ball screw actuator with modularized peripheral parts. Thisarticle introduces the structure and the features of this ball screw actuator unit.

DC motor

Shaft cap

Oil sealOil seal

Boot

Coupling member

Interlocking arm

Double row angular contact ball bearingsLock nut

Aluminum case

Ballscrew nut Ballscrew shaftInvolute spline

Actuating shaft

Linear ball bearing

Fig. 1 Structure of actuator

Yoshinori IKEDA*Hirakazu YOSHIDA*

*Automotive Sales Headquarters Automotive Engineering Dept.

2. Structure of the actuator

Fig. 1 shows the structure of our actuator.The ballscrew is coupled with an electric motor via

an involute spline formed at one end of the ballscrewshaft to transmit torque from the motor.

At the other end of the ballscrew shaft, a double rowangular ball bearing is situated securely in analuminum case. The ballscrew and actuator aresynchronized by means of an interlocking arm. Acrossboth ends of the actuating shaft, linear ball bearingsand oil seals are arranged symmetrically.

The linear ball bearings support the actuating shaftwhile developing only minimum friction so that theactuating shaft can slide in the axial direction. The oilseals situated around the linear ball bearings preventingress of foreign material into the aluminum case. Anactuating wire is connected to the coupling member atthe end of the actuating shaft.


3.2 Corrosion resistance, dust-proof &water-proof performance

Any automotive actuator needs to perform asdesigned for a prolonged period under variousenvironmental conditions. Our actuator, in particular,has an actuating shaft that is exposed outside thecase; therefore, both ends of the shaft are providedwith reciprocating motion-capable oil seals to preventingress of foreign material.

For enhanced corrosion resistance, the actuatingshaft is provided with special plating. Furthermore, theactuating shaft is equipped with a boot and shaft capso that the shaft is not directly exposed to the ambientair (Fig. 6).

-48-

Fig. 2 Ballscrew actuator

Fig. 3 Ballscrew for actuator

Fig. 4 Efficiency of ballscrew

Fig. 5 Liner ball bearing

3. Advantages of our actuator

Fig. 2 shows our actuator, and Fig. 3 shows theballscrew for our actuator.

The actuator has undergone various testssimulating various environmental conditions includingwater, snow, mud, gravel, and dust along withvibration and impact testing. Through these tests, theactuator has been progressively improved and nowboasts a higher degree of reliability. The actuatorboasts the following advantages:

(1) Highly efficient ballscrew and lower frictionmoving parts help achieve a higher degree ofresponse and greater thrust force.

(2) Sufficient corrosion resistance, and dust-proof& water-proof performance

These advantages are described below:

as 32%; in contrast, the efficiency of the ballscrewadopted in our actuator is very high, standing at 92%.

Fig. 5 shows the linear ball bearings that supportboth ends of the actuating shaft. The actuating shaft issupported by the linear ball bearings in rolling contact,thereby the resultant lower friction helps the actuatingshaft to move smoothly.

Table 1 Spec. of ballscrew

Lead Lead angleShaft diameter

14.5mm 4mm 4˚47'

92%1009080706050403020100

0 1 2 3 4 5 6

Effi

cien

cy

%

Lead angle deg

Ball screw(Friction coefficient: 0.005)

(Benchmark)Sliding screw(Friction coefficient: 0.1)

Efficiency in forward rotation

Efficiency in reverse rotation

Efficiency in forward rotation

3.1 Higher efficiency and lower frictionWe have set up the specifications for our actuator

so that the conversion efficiency of motion of theballscrew adopted, from both forward and reverserotation directions, is 90% or greater. The actuatingshaft can be readily moved by hand thanks to its low-friction design.

Table 1 provides the specifications for theballscrew. Fig. 4 graphically plots the theoreticalefficiency of the ballscrew. At a lead angle of 4°47’,the efficiency with a benchmark sliding screw is as low

-49-

4. Evaluation test

In an effort to develop a highly reliableelectromechanical actuator, we have performedvarious tests to verify that our actuator has sufficientperformance and reliability. The major characteristicstested are described below:

4.1 Thrust force testThe load torque on a ballscrew can be determined

with the following expression:

P・LT＝―――― ………………………………………(1)

2π・η

where,T ： Load torque (N-m)P ： Axial load (N)L ： Lead of ballscrew (m)η ： Efficiency 0.92

Using the result from expression (1), the thrust forcecan be determined with expression (2) below:

T・2π・ηP＝――――― ……………………………………(2)

L

For the thrust force test, a load cell was insertedbetween the coupling member and a fixed part asshown in Fig. 7 to measure the load.

The result of the thrust force measurement isplotted in Fig. 8. We have verified that the theoreticalcalculated value of 1,200 N is achieved with ourballscrew. The thrust force available from a slidingscrew whose size is similar to that of our ballscrew isapproximately 1/3 that of our ballscrew.

4.2 Durability test in severe environmentWe have subjected the actuator samples to a

durability test under severe environmental conditionswith a high-temperature ambience while applying apredetermined load to each actuator sample. So that aconstant load is applied to an operating actuator, we

Shaft capOil seal

Boot

Actuating shaft

Linear ball bearing (2 pos.)

Fig. 6 Structure of sealing of shaft

Actuator

Direction ofload application

Load cell

Fixed part

Moving part

Fig. 7 Appearance of thrust force test

3000

2000

1200N1000

2500

1500

500

3.5 4.50

Time sec.

Sliding screw

Gen

erat

ed a

xial

thru

st fo

rce

N

Fig. 8 Thrust force measurement

have fabricated a special durability tester that uses anadditional ballscrew responsible for applying aconstant load to an actuator sample. Fig. 9 shows aview of this tester. We performed a durability test for anumber of loading cycles that is equivalent to the lifeof a car while applying the maximum expectedworking load to each actuator sample and, aftercompletion of the durability test, we performed anactuator response test.

Electric Ball Screw Actuators for Automobiles

MotionActuator

Load cellBackward end sensor

Backward end sensor

Load applying device

Fig. 9 Appearance of durability test

-50-

Fig. 10 Example of operating time measurement

Fig. 11 Internal condition after cyclic corrosion test

20

-200 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

15

-15

10

-10

5

-5

0

Time sec.

Afterdurability test

0.16 sec

Beforedurability test

Loca

tion

mm

Fig. 10shows results of the response test obtainedfrom actuator samples before and after the durabilitytest. Our actuator having undergone a severedurability test still boasts a higher level of responsecompared with the pre-test level and its post-testresponse shows no deterioration.

From these findings we have determined that evenafter having undergone a durability test under severeenvironmental conditions the performance of ouractuator is as good as the pre-tested level.

Photos of authors

Yoshinori IKEDA

Automotive Sales HeadquartersAutomotive Engineering Dept.

Hirakazu YOSHIDA



4.3 Combined cyclic corrosion testOur actuator was placed in the combined cyclic

corrosion tester and each sample was subjected torepeated environmental cycles. Each cycle consistedof a salt water spray, high-temperature drying, and ahigh-temperature high-humidity condition to verifycorrosion resistance of our actuator under severeenvironmental conditions.

Fig. 11 shows the appearance of our actuatorhaving undergone the combined cyclic corrosion test.The interior of our actuator having undergone thesevere test does not show any problems such as rustor water ingress; the characteristics of the post-testactuator do not show any signs of problem.

5. Conclusion

We have presented information about our unique,automotive electromechanical ballscrew-drivenactuator that adopts various mechanical element-related technologies for realizing higher response andgreater thrust force.

With electromechanical moving parts increasinglybeing used in recent automobile designs, moreautomobile designers have been considering use ofballscrew-driven mechanisms that can be actuatedwith smaller electric motors. Thus, our ballscrewactuator products will find a greater market share.

To be able to further improve the lighter-weight andenhanced functionality designs for its automotiveelectromechanical ballscrew actuator products, NTN iscommitted to improvements in not only elementaryparts such as the ballscrew but also auxiliarycomponents such as sensors and electric motors.Through these efforts, NTN will help expand the scopeof applications of its ballscrew actuator products.

References1) K. Tateishi and K. Kazuno: Ball-screw for Automated

Manual Transmissions, NTN Technical Review No. 73(2005) 72-75

2) K. Kazuno: Ball-screw Unit for Variable Valve Eventand Lift Systems, NTN Technical Review No. 75(2007) 72-77

-51-


[ New Product ]

Plastic Bearings for Electric Pumpsfor Next Generation Battery Cells

1. Preface

To help reduce CO2 emissions, unconventionalapparatuses that do not burn fossil fuels have beenincreasingly marketed recently, and the need forapparatuses incorporating next generation battery cellshas been mounting. The most common applicationsintroducing these novel apparatuses include fuel-cellcogeneration systems in the industrial and homeelectric appliances fields and electric vehicles andhybrid electric vehicles in the automotive engineeringfield. Each of the cooling systems that mitigate heatbuildup in these apparatuses has a built-inelectromechanical pump that circulates cooling water.This article provides information about resin slidingbearings suitable for electromechanical pumps and thetypical applications of such bearings are included.

2. Electromechanical pumps forvarious systems

2.1 Fuel-cell cogeneration system1)

By the electrochemical reaction between hydrogenand oxygen, a fuel-cell generates electricity. Theoxygen it uses is derived from air, while the hydrogen ituses is obtained by reforming the air with a reformer:the so-obtained oxygen and hydrogen are introduced

In order to reduce CO2 emissions, devices with next generationbattery cells are being introduced or are already on the market.Fuel cell co-generation systems, electric evehicles and hybridvehicles have electric pumps to circulate cooling water in thecooling systems for the cells and motors. BEAREE AS5704

bearings, which have excellent self-lubricity, chemical resistance and high flexibility in design, areused in these electric pumps. This article introduces the characteristics and applications of BEAREEAS5704 bearings.

into a fuel-cell stack that generates electricity byelectrochemical reaction. Electrochemical reactiontaking place in a fuel-cell stack is an exothermicreaction; therefore, the cooling system (whichcirculates cooling water) that regulates the temperatureinside the cell stack to a constant level is driven by anelectromechanical pump. The typical requirements forthis electromechanical pump are high efficiency andcompact size. In order to satisfy these requirements,the electromechanical pump adopted is often a magnetdrive-type centrifugal pump.

2.2 Cooling systems on electric vehicles (EV)and hybrid electric vehicles (HEV)

On conventional automobiles driven by internalcombustion engines (ICE), a centrifugal pump is usedto circulate radiator coolant to cool down the ICE,wherein rotary motion on the engine is input to thepump shaft via a drive belt.

Incidentally, any EV does not have an ICE, and anICE on HEV is shut down in idling stop mode:therefore, an electromechanical pump is needed toactuate a cooling system for battery cells and/orelectric motor. The electromechanical pump used forthis purpose is often a magnet drive-type centrifugalpump.

Norio ITOH*Takuya ISHII*

*Engineering Dept. NTN Engineering Plastics Corporation


-52-

Electricmotor

Motor shaft

MagnetThrust bearing

MagnetImpeller

Stationary shaft

Bearing

Casing

Fluid OUT

FluidIN

Casing

Fig. 1 Structure of magnet drive centrifugal pump

Retaining feature(D-shaped cutout)

Lubricating slot(4 pos.)

Fig. 2 Bearing

3. Construction of electromechanical pump

Generally, pumps transport fluids such as water,liquid fuels, lubricating oils, organic solvents, and acidicand alkali liquids. Three major pump operation typesare available: centrifugal, volumetric rotary andvolumetric reciprocating types, wherein their rotaryshafts are supported by bearings. When the fluidtransported by a pump is water, or acidic or alkaliliquid, an ordinary metal bearing can develop acorrosion problem: to address this issue, carbon slidingbearings and resin sliding bearings may be used 2).Resin sliding bearings boast better self-lubricatingperformance and chemical resistance. In particular,injection-molded components of sliding bearing boastgreater freedom of design for factors including shape.

3.1 Magnet drive-type centrifugal pumpFig. 1 shows a structure of a typical magnet drive-

type centrifugal pump. With this structure, the magnetsinstalled onto the motor shaft rotate to cause theirmagnetic force to rotate the impeller integrated with themagnets in the casing, thereby causing the pump totransport the fluid.

3.2 Resin sliding bearingConventional magnet drive type centrifugal pumps

often use carbon sliding bearings. A carbon slidingbearing can be molded together with an impeller;however, because this bearing component is obtainedonly by machining a molded blank, the form factorflexibility of this work piece is limited. Therefore, thisbearing type requires improvement in its resistance toimpact-induced cracking and reduction of costs.Through an injection molding process, geometricalfeatures can be readily formed on a resin-made slidingbearing, and the examples of such features include

lubricating slots on the bore surface and end face ofthe bearing and a retaining feature (D-shaped cutout,projection, etc.) on the outer circumference of thebearing to help securely engage the bearing with theimpeller (see Fig. 2).

Without undergoing any machining process, a resin-made sliding bearing can be formed together with anintegrated impeller.

When a resin sliding bearing runs, a radial load andan axial load act on it, thereby causing sliding contactbetween the bearing bore and stationary shaft andbetween the bearing end face and thrust force carryingface . The material used for this resin sliding bearing isNTN BEAREE AS5704.

4. BEAREE AS5704 resin bearing

4.1 FeaturesThe BEAREE AS5704 resin sliding bearing consists

of PPS (polyphenylene sulfide) blended with a specialfiller.

<Advantages>(1) Compared with a general-purpose PPS resin

sliding bearing, wear of our bearing in water is20% or lower.

(2) Low wear or abrasion on a mating material suchas stainless steel

(3) BEAREE AS5704 bearing boasts greaterfreedom of design in shape due to injectionmolding.

(4) Can be used even in anti-freeze liquid, or acidicor alkaline liquid

4.2 Basic physical characteristicsTable 1 summarizes the basic physical characteristics

of BEAREE AS5704 material.

-53-

Table 2 Wear test condition

Test equipment

Bearing pressure

Sliding velocity

Mating material

Environment

Test time

Characteristics

Underwater ring on disk type

0.4 MPa

25 m/min（Hv200，0.4μmRa）

SUS304

Water (normal temperature, no temperature control)

50h

Content

Table 3 Comparison of features various bearings

Process used

Wear resistance(underwater)

Chemicalresistance

Dimensional stabilityof material havingabsorbed water

Resistance to crackingdue to impact

Resistance to defor-mation due to impact

Degree ofFreedom of design

Injection compre-ssion molding

Injectionmolding

×

×

×

Machining Machining

Price ×

×

×

×

×

×

Bearing type BEAREEAS5704

Phenol resin (including graphite)

PTFE resin (including graphite)Carbon

：Excellent　：Good　：Acceptable　×：Not acceptable

Table 3 provides a comparison of performance ofsliding bearings made of these materials. Using PPSresin as a base material, the BEAREE AS5704 bearingexcels in chemical resistance and dimensional stabilityafter having absorbed water.

BEAREEAS5704

Carbon

Spe

cific

wea

r ×

10-8

mm

3 /（N・

m）

Phenolresin

(including graphite)

PTFEresin

(including graphite)

500

400

300

200

100

0

Fig. 3 Specific wear of various bearings

Plastic Bearings for Electric Pumps for Next Generation Battery Cells

Table 1 Basic characteristics of BEAREE AS5704

Specific gravity

Characteristics

Tensile strengthElongation

Bending strengthBending elastic modulus

Linear expansioncoefficient

Rockwell hardness

Izod impact strength

ASTM D792

ASTM D638

ASTM D790

ASTM D785

ASTM D256(Notch type)

TMA method

― MPa

MPaGPa

%

1/˚C

R scale

J/m

1.64540.7

10310

MD : 2.0×10-5

CD : 4.5×10-5

112

27

Test method Unit ofmeasurement

Characteristicvalue

※The values given above are representative values.

4.3 Comparison of various bearings(resin materials)

Examples of materials of typical resin slidingbearings used in liquid include phenol resins and PTFE(polytetrafluoroethylene) resins. To compare wearresistance characteristics of these materials with thoseof BEAREE AS5704 and carbon material, we haveperformed an underwater wear test. Table 2summarizes the test conditions applied, and Fig. 3shows specific wear NOTE) of various bearing materials.

BEAREE AS5704 boasts superior wear resistanceand has lower friction compared with the carbonmaterial.

NOTE) “Specific wear” means a wear volume per unit slidingdistance or per unit load calculated from wear before andafter the test. Smaller specific wear means smaller wear.

4.4 Comparison to PPS resin bearingUnder the test conditions summarized in Table 2,

comparison test for friction wear characteristics hasbeen performed for BEAREE AS5704 and three PPS-resin based bearing materials (PPS + fiber glass, PPS+ carbon fiber, and PPS + PTFE). Fig. 4 shows thetime-dependent variation in dynamic frictioncoefficients of these materials, and Fig. 5 provides thespecific wear of these materials.

BEAREE AS5704 boasts the lowest dynamic frictioncoefficient, and the value of this coefficient remainsstable. Furthermore, the specific wear of BEAREEAS5704 is less than 20% of the PPS resin blendedwith carbon fiber coefficient. The wear depth on themating material (SUS 304) that underwent the test hasbeen measured. As a result, the wear depth on themating material tested with the PPS resin blended withcarbon fiber was approximately 5 µm. In contrast, themating material tested with BEAREE AS5704 did notdevelop any wear (Figs. 6 and 7).


-54-

Table 4 Limit PV test condition

Characteristics Content

Test equipment

Sliding velocity

Mating material

Environment

Underwater high-speed ring on disk type

100，300 m/min

SUS304（Hv200，0.4μmRa）

Antifreeze (normal temperature, no temperature control)

Table 6 Wear test condition

Characteristics Content

Test equipment

Bearing pressure

Sliding velocity

Mating material

Underwater high-speed ring on disk type

1.15 MPa

170 m/min

SUS304（Hv200，0.4μmRa）

Environment Antifreeze (normal temperature, no temperature control)

Test time 500h

Table 5 Limiting PV value

Sliding velocity (V)Limiting bearing

pressure (P) Limiting PV value

100m/min

300m/min

7 MPa

3 MPa

700 MPa・m/min

900 MPa・m/min

Fig. 6 Surface roughness of mating material after test

Sliding area

1mm

Fig. 7 Surface roughness of mating material after test

Sliding area

1mm

4.5 Limiting PV value of BEAREE AS5704Two sets of velocity conditions summarized in Table

4 were tested. Under each set of conditions, thebearing pressure was increased by 1 MPa per hour.The limiting bearing pressure was recorded as thepressure at which the wear exceeded 20 µm or thesliding surface melted. Table 5 summarizes limiting PV

values calculated from the test velocity and limitingbearing pressure.

4.6 Wear curve of BEAREE AS5704Under the test conditions summarized in Table 6,

the time-dependent wear on BEAREE AS 5704specimen were measured; then the wear curve in Fig.8 was plotted. This plot shows an initial wear phasebeginning at the start of test and ending at 50operating hours. After that point, the wear curveexhibits steady wear state and shows virtually noincrease in the wear.

BEAREE AS5704

PPS resin + fiber glass

PPS resin + carbon fiber

Dyn

amic

fric

tion

coef

ficie

nt

PPS resin + PTFE resin

0 10 20 30 40 500.0

0.2

0.4

0.6

0.8

Time h

Fig. 4 Coefficient of dynamic friction of PPS bearings

BEAREEAS5704

PPS resin +fiber glass

PPS resin +carbon fiber

Spe

cific

wea

r ×

10-8

mm

3 /（N・

m）

PPS resin +PTFE resin

500

400

300

200

100

0

Fig. 5 Specific wear of PPS bearings

5. Considerations about use of resinsliding bearings

When setting up the clearance between a resinsliding bearing and a mating shaft, it is necessary toconsider the linear expansion coefficient of the resinmaterial. Generally, linear expansion coefficients ofresin materials are greater than those of metalmaterials. For example, in a case where the material ofan impeller integrated with a bearing is a metal, orthere is a temperature gradient across the bearing andthe impeller, resulting in heat generation at the slidingarea, the bearing expands toward its bore side. In thiscase, incorrect bearing clearance design can lead toelimination of bearing clearance to the mating shaft,and this in turn can lead to abnormally severe wear onthe bearing and mating shaft 3).

6. Special materials for difficult applications

Electromechanical pumps have been used invarious industrial fields and environments. In additionto BEAREE AS5704, NTN offers the BEAREE FL3700and FL3642 lines of products made from PTFE resinas base material for difficult applications (Table 7).

-55-

Table 7 Bearings under special condition

Bearing Base material Advantages

BEARI FL3700

BEARI FL3642

PTFE resin

PTFE resin

Chemical resistance

Non-black color: may be used incontact with food or drinking water.

Because its base material and filler are highlyresistant to chemicals, FL3700 remains impervious tovirtually all chemicals, and therefore is used inchemical pumps.

For food machinery applications, the bearing resinmaterial must not be black in color. Therefore, FL3642,which is pale yellow, is adopted 4).

7. Afterword

EVs and EHVs will be more commonly used in thefuture. Consequently, demand for electromechanicalpumps used on these vehicles will further expand.These electromechanical pumps need to be moreefficient, have longer service life and feature muchmore compact size. Therefore, performance of resinsliding bearings for these pumps need to be furtherimproved. To address these challenges, we are goingto continue our research and development efforts forour resin sliding bearings so that we can furtherimprove the sliding mechanism with our bearingproducts to improve their functionality. We believe thatdemands for resin sliding bearings will further increasein applications that require low friction and low wearcharacteristics. We hope this report may help deliveruseful information to people involved in thisengineering field.

References1) O. Tajima: Development Status of Proton Exchange

Membrane Fuel Cells (PEFC) 2, Denki (2004) 35(Japanese)

2) T. Hayashi: Machine Parts and Elements, JapanPlastics Vol. 57, No. 4 (2006) 50 (Japanese)

3) NTN Corporation Editing Team: UnderstandingBearings (2007) 154

4) Oki and T. Ishii: Tribological Properties of PTFEComposite Materials in Water, Monthly Tribology Vol.13, No. 12 (1999) 48 (Japanese)

Photos of authors

Norio ITOH

Engneering dept.NTN Engneering Plastics Co.,ltd

Takuya ISHII

Engneering dept.NTN Engneering Plastics Co.,ltd

Fig. 8 Wear curve

Wea

r μ

m

30

20

10

00 100 200 300 400 500

Time h

Plastic Bearings for Electric Pumps for Next Generation Battery Cells

-56-


[ New Product ]

Low-torque Roller Lifter Unit

1. Introduction

EU has decided to enact in 2012 its CO2 emissionsregulation according to average CO2 emission perpassenger car of 130 g/km or lower, which means20% reduction relative to the present emissions level.Also, USA has established targets for improved fueleconomy which include average 40% improvement infuel economy by fiscal year 2020 in terms of theamount of gasoline consumed by passenger cars orcompact trucks 1). Thus, NTN believes that the needfor more fuel-efficient cars will increase and morestringent emissions standards will be introduced.

In Japan, a new fuel economy standard (see Table1) has been legislated whose deadline is set at fiscalyear 2015: accordingly, the fuel economy ofpassenger cars in Japan in FY2015 will be 23.5%

The demand for improved fuel consumption and exhaust emissionscontrol in cars has increased on a global basis in recent years. Automobilecompanies are working to develop gasoline direct-injection engines as onemeans to achieve new fuel efficiency targets. Other demands includeincreasing the fuel pressure and reducing the friction in the fuelcompression process in the high-pressure fuel pumps used for direct-

injection engines. NTN has developed a low-torque roller lifter unit that responds to these demands.This report introduces the features of this NTN low-torque roller lifter unit.

better compared with the FY2004 level 2).To address this situation, automakers have been

committed to development of hybrid electric vehiclesand electric vehicles; as well as fuel economyimprovement through techniques including variablevalve system and idling stop. Direct-injection gasolineengine also appears to be a particularly promisingmeans.

Direct-injection engines are characterized by thefact that air alone is drawn into cylinders andcompressed. Fuel is compressed by a fuel pump isdirectly injected into the cylinders where the fuel isfired. In contrast with conventional engines that drawair-fuel mixture generated in an intake pipe into thecylinders (port injection), direct-injection engines aresaid to develop better fuel economy owing tooperation of the following reasons 3), 4).

(1) Improved anti-knock properties to allow for ahigher compression ratio (greater engine power).

(2) Accurate control of the air/fuel ratio(3) Reduced pumping loss at a lower load range

Masashi NISHIMURA*

*Needle Roller Bearing Engineering Dept. Automotive Sales Headquarters

Table 1 2015 year fuel-efficient target(Improvement rate compare with 2004) 2)

Passenger car

Actual valuefor FY2004

Estimated valuefor FY2015

Improvement in fueleconomy overFY2004 level

Compact bus

Compact truck

Car type

13.6（km/R） 16.8（km/R）

8.3（km/R） 8.9（km/R）

13.5（km/R） 15.2（km/R）

23.5%

7.2%

12.6%


-57-

Fig. 1 Appearance of roller lifter unit Fig. 2 Direct injection engine, fuel pump

Fig. 3 Application of roller lifter unit

Fuel pump

Fuel pump driving camIntake camshaft

Roller lift unit

Mating guide hole

Fuel pump

2. Peripheral structure around direct-injection engine and typicalapplications of roller lifter unit

Fig. 1 shows an appearance of a roller lifter unitwhile Fig. 2 illustrates a structure around a fuel pumpon a direct injection engine. Fig. 3 provides a view oftypical application of a roller lifter unit.

In combination with a cam, a roller lifter unit situatedin a fuel pump drive on a direct injection enginetransmits rotary motion of the engine shaft to areciprocating plunger.

Previously, most commonly used roller lifter unitshad a sliding surface in contact with a cam. Now,certain automakers are adopting rolling type roller lifterunits in order to reduce friction loss from contact witha cam and to allow a cam to be able to run at a higherspeed.

Usually, on a direct injection engine, the fuel isinjected with a pressure of 4 to 13 MPa. On certaindirect injection engines, the injection pressure is as

high as 20 MPa 5) in order to improve combustionefficiency. It appears that much higher injectionpressure will be increasingly adopted.

A higher fuel injection pressure means a greaterload exerted onto the lifter. A conventional lifter, inparticular, whose contact surface to the cam is asliding surface can experience deterioration in fueleconomy owing to greater friction with the cam.

At the same time, to be able to adopt higher fuelpressure, it will be necessary to increase a number ofcam-operated compression cycles (increased camcyclic rate and increased number of high points oncam) as well as a cam lift. To cope with thesechanges, the outer circumference length of cam needsto be longer if the base circle dimension of camremains unchanged. Consequently, the bearing usedin this type of application needs to be capable ofhigher running speeds.


-58-

complement roller type bearing.Furthermore, NTN has optimized wall thicknesses

of various areas of the casing of its roller lifter unit sothat the casing features lighter weight whilemaintaining necessary mechanical strength. Thus,NTN has reduced the inertial force occurring on itsroller lifter unit in reciprocating motion so that its rollerlifter unit can more accurately follow the motion of camrunning at higher speed.

Fig. 4 shows the specification for NTN low-torqueroller lifter unit.

4. Evaluation by performance

Higher fuel temperature means a greater loadapplied to the lifter. In particular, on conventionallifters whose contact surface to the cam is a slidingsurface. The friction on the contact surface will begreater and can cause deterioration in fuel economy.Furthermore, to help achieve higher fuel pressure, thelifter needs to be more durable at higher speed.

In order to evaluate the performance of our rollerlifter unit, NTN measured torques and temperatureincreases on the bearing. The section below presentssome portion of information about our tests forevaluating our roller lifter unit.

Bearing size: ID 8 ×OD 16×W18 (in mm)

[Outer ring]

[Rollers]FA treatment

[Cage]

[Casing]

[Detent pin]*2

[Shaft]

φ26

30

Special inductionheat-treatedHollow shaft

[light-weight design]

Fig. 4 Specification of NTN low-torque roller lifter unit

*1: “Full complement roller bearing” means a roller bearing type not having a cage. Though this bearing type boasts greater load ratingbecause of a greater number of rollers for a given bearing size, but can develop problems such as skew proneness of rollers becausethe rollers are not guided by a cage.

*2: The detent pin slides in the groove formed in the mating guide hole to prevent the roller lifter from rotating in the circumferentialdirection on the casing.

3. Structure and advantages of NTNLow Torque Roller Lift Unit

As mentioned above, more common roller lifter unitspreviously used have sliding contact surface that is incontact with a cam: certain automakers are usingroller lifter units incorporating rolling bearings in orderto mitigate friction loss. The rolling bearings used forthis purpose are full complement roller bearingsbecause of their excellent load bearing performanceand longer life*1. However, full complement rollerbearings can pose problems such as heat buildup andincreased running torque. Intermittent loading by thecam as well as inclination occurring from the gapbetween the roller lifter unit and mating guide hole willlead to skew on the rollers. This skew will cause thebearing to develop lateral runout , which triggers theabove-mentioned problems.

In order to mitigate lateral runout on the bearing forroller lifter unit, NTN has adopted a caged rollerbearing to inhibit skew occurrence on the rollers,reduce running torque on the bearing, and improvehigh-speed durability of the roller lifter unit. Also, NTNhas adopted FA treatment which is NTN’s proprietyheat treatment technique to improve bearing life,thereby NTN has achieved a calculated bearing lifethat is equivalent to or better than that of full

-59-

Fig. 5 Diagrammatic illustration of test machine

4.1 Torque measuring operationNTN measured torques on bearings in lifter units to

determine the reduction of friction loss using NTN lowtorque caged roller lifter compared to conventionalsliding type lifters and full complement roller type lifterunits.

4.1.1 Torque measuring conditionsTable 2 summarizes the conditions for our torque

measuring test and Fig. 5 schematically illustrates thetest rig used.

Load

Electric motor Torque meter

Samples being tested

Temperature measuring point(end face of shaft)

Drive roll

Fig. 6 Result of rotational torque measurement

Running torque (at load 500 N)

Running torque

1.4 1

0.8

0.6

0.4

0.2

0

1.2

1

0.8

0.6

0.4

0.2

01000 3000 6000 9000

1000 3000 6000 9000

Bearing speed min-1

Bearing speed min-1

Run

ning

torq

ue

N・

m

1.4

1.2

1

0.8

0.6

0.4

0.2

0

Run

ning

torq

ue

N・

m

1.4

1.2

1

0.8

0.6

0.4

0.2

0

Sta

rtin

g to

rque

N・

m

Rel

ativ

e to

rque

rat

ios

of c

aged

type

1

0.8

0.6

0.4

0.2

0

Rel

ativ

e to

rque

rat

ios

of c

aged

type

Running torque(full complement roller type)Torque ratio(relative to full complement roller type)

Running torque(full complement roller type)Torque ratio(relative to full complement roller type)

Running torque (caged type)Running torque(relative to sliding type lifter)

Running torque (caged type)Running torque(relative to sliding type lifter)

Torque ratio (relative to sliding type lifter)

Torque ratio (relative to sliding type lifter)

Caged type

Full complement roller typeRelative to sliding type lifter

Running torque (at load 1000 N)

Table 2 Rotational torque measurement conditions

Samplesbeing tested

Testequipment

Torque tester (outer ring rotation tester complete with a torque meter)

Loading conditions

Bearingspeed*4

Lubrication

*3 Because the contact surface of sliding type lifter in contact with the cam is a flat face of cylindrical tappet, the diameter and height of the tappet are given here.*4 Means running speed of outer ring of rolling bearings (caged type, full complement roller type). Measurement on sliding type lifter was taken while allowing the drive roll to run at a running speed equivalent to that for measuring bearing speed of rolling bearing.*5 A test method where the drive roll is brought into contact with the outer circumference of outer ring to allow the outer ring to turn in response to the rotary motion of the drive roll. (The outer ring is allowed to rotate on the roller lifter unit installed on an actual automotive fuel pump.)

• Rolling bearing (caged type) Size: ID 8×OD 16×W18 (in mm)• Rolling bearing (full complement roller type) Size: ID 7.67×OD 16×W16 (in mm)• Sliding type lifter Size: D 30×H 25 (in mm)*3

500 N, 1,000 N (Measurement was performedwith starting torque set at 500 N.)

1000min-1，3000min-1，6000min-1，9000min-1　The starting toque at start-up phase, whereacceleration from 0 to 3,000 min-1 takes place,was measured (with outer ring rotated*5).

Lubricating oil: engine oil with dynamic viscosity of 0W-20Lubricating temperature: ordinary temperatureLubricating system: splash lubrication (oil level: centerline of drive roll)


4.1.2 Result of torque measurementFig. 6 Graphically plots the resultant torque

measurements. From Fig. 6, our findings are apparent; in

comparison with a sliding type lifter, NTN’s standardtorque roller lifter unit (caged type) boasts an 85%reduction in starting torque and 73 to 86% reduction inrunning torque. Also, in comparison with a fullcomplement roller type, NTN’s standard torque rollerlifter unit exhibits 8 to 29% torque reductions invarious measuring conditions.

4.2 Temperature-rise testTemperature rise tests for bearings on roller lifter

units have been performed to verify the excellent highspeed durability of NTN’s caged low torque unitcompared to complement roller type lifter units.

4.2.1 Temperature-rise test conditionsTable 3 summarizes the test conditions adopted.

-60-


4.2.2 Results of temperature-rise testFig. 7 graphically plots comparison of temperature

rise on the samples tested and Fig. 8 shows a view ofseizure on a rolling bearing (full complement rollertype) having undergone the test.

The full complement roller type has developedseizure as a result of lateral runout of its bearing at18,000min-1. In contrast, the NTN’s standard lowtorque roller lifter unit (caged type) has not developedseizure up to 36,000 min-1. Thus, we have verified thatthe high speed durability, in terms of maximumallowable running speed, of the caged type is morethan twice as much as that of the full complementroller type.

Fig. 7 Comparison of temperature-rise(Full complement roller type)

Bearing speed min-1

Full complement roller type hasdeveloped seizure at 18,000 min-1.

Caged type

Full complement roller type

Tem

pera

ture

ris

e ˚

C

04,500 9,000 13,500 22,500 31,50018,000 27,000 36,000

10

20

30

40

50

60

Fig. 8 Appearance of test sample(Full complement roller type)

Table 3 Temperature-rise test condition

Samplesbeing tested

Tester Outer ring rotation tester (See Fig. 5.)

Loadingconditions

Bearingspeed

Lubrication

• Rolling bearing (caged type) Size: ID 8×OD 16×W18 (in mm)• Rolling bearing (full complement roller type) Size: ID 7.67×OD 16×W16 (in mm)

P/C=0.22*Relative to rolling bearing (caged type)

Eight incremental steps: 4,500, 9,000, 13,500,18,000, 22,500, 27,000, 31,500 and 36,000 min-1

(with outer ring rotated)

Lubricating oil: engine oil with dynamic viscosity of 0W-20Lubricating temperature: ordinary temperature,lubricating oil flow rate: 150 ml/minLubricating system: circulating lubrication

Table 4 Life test conditions

Samplesbeing tested

Testequipment Outer ring rotation tester (See Fig. 5.)

LoadingconditionsBearingspeed

Lubrication

Calculatedlife per JIS

• Rolling bearing (caged type) Size: ID 8×OD 16×W18 (in mm)

P/C=0.22

28000min-1 (with outer ring rotated)

Lubricating oil: engine oil with dynamic viscosity of 0W-20Lubricating temperature: ordinary temperature,lubricating oil flow rate: 150 ml/minLubricating system: circulating lubrication

97h (Information) Calculated life for full complementroller type is 203 hours.

4.3 Bearing life testThe bearing of the NTN’s low torque roller lifter unit

is a caged type. Accordingly, the number of rollers inthis type is unavoidably smaller compared with the fullcomplement roller type causing its load bearingcapacity to be lower and expected service life is to beshorter. In order to address a problem of shorterbearing life on the NTN’s low torque roller lifter unit,NTN is subjecting the rollers to our unique FAtreatment and the shaft to special induction heatingtreatment as previously described in Sec. 3. NTN hasperformed a life test on the bearing section in questionin order to verify the effectiveness of these lifeextending measures.

4.3.1 Test conditions for bearing life testTable 4 summarizes the test conditions applied for

the life test.

5. Conclusion

This paper has presented information about NTN’slow torque roller lifter unit and associated technology.

The size of market will further expand for directinjection engines as one of potential measures forimproving fuel economy for automobiles. To addressthis trend, NTN will further market its low torque rollerlifter products as well as enhance their functionality.

References1) Website of Japan Automobile Manufacturers

Association, Inc. (JAMA)2) Website of Ministry of Land, Infrastructure, Transport

and Tourism3) H. Ando: Technical Prospects of Gasoline Direct

Injection Engines, Engine Technology Review Vol. 1,No. 1 April 2009, 18-23

4) T. Sena and Y. Katsuragi, Introduction to EngineScience, first edition, Grand Prix Publishing, (1997)173–183 (Japanese)

5) Direct Injection, Car Mechanism Encyclopedia EngineVolume, August 2009 special issue 111–115(Japanese)

-61-

4.3.2 Result of bearing life testFig. 9 shows graphical plotting of the result of our

bearing life test.NTN has have subjected the rollers to FA treatment

and the shaft to a unique induction heat treatmentprocess. The results have verified that the standardrating life of our bearing stands at 1,370 hours, whichis 14 times as long of a life value calculated accordingto a JIS method (97 hours).

When assuming that the average engine speed is3,000 min-1 (equivalent to average running speed of7,500 min-1 at the bearing of roller lifter unit*6), theaverage load acting on the bearing is 1,000 N*7, andthe above-mentioned life value calculated per a JISstandard is multiplied by 14, then the resultant bearinglife for our roller lifter unit stands at 5×104 hours andthis life coincides with a life of 27 years*8. Assumingthat the useful life of an average passenger car is 15years, the NTN’s bearing well satisfies this useful liferequirement.

*6 Assumptions: This speed is equivalent to cam running speedof 1,500 min-1, and the cam outside circumference length isfive times as long as the bearing outside circumferencelength.

*7 Assumptions: The fuel pressure is set at 13 MPa, and thepump plunger diameter measures 10 mm.

*8 Assumption: The average daily use time for an averagepassenger car is 5 hours.

Photo of author

Masashi NISHIMURANeedle Roller Bearing

Engineering Dept.Automotive Sales Headquarters

Fig. 9 Life test result

Life h

Cum

ulat

ive

failu

re p

roba

bilit

y %

5

10

203040506070809095

8 9 1103 103

1.5 2 2.5 3


-62-


[ New Product ]

Hub Bearings for Automobiles that Conserve Resources

1. Preface

One of the leading environmental issues today isglobal warming. To help in the prevention of globalwarming it is necessary to reduce the emissions ofgreen house gases - in particular, CO2. Therefore, inthe auto industry around the world there is an everincreasing need for light-weight, low friction designsthat will lead to lower energy consumption on carsduring travel as well as a reduction in the amount ofmaterials consumed during car manufacturing.

Believing that coexistence with the globalenvironment is its most urgent challenge, NTN haslong been committed to efforts for designing lighter,lower-friction car components. We have developed aunique hub bearing, which is one of the underbody carcomponents. In this paper we are going to presentinformation about this novel “resource-savingautomotive hub bearing design” that features areduction in the amount of materials consumed inproduction while boasting functions and characteristicsequivalent to or better than those of conventional hubbearing products.

Automotive parts are being expected to respond to environmentalissues such as global warming and to effectively use naturalresources. This article introduces our hub bearings for automobilesthat conserve resources in order to meet these demands. They aredesigned to be lighter than conventional products and to reduce theamounts of materials used in order to improve the yield from thematerials during processing.

2. Structure and features

In developing our “resource-saving automotive hubbearing design”, we have adopted the Gen3 hubbearing for non-driven wheels (rear wheels on frontdrive vehicles) as a model case. In this developmentwork we have attempted to meet the following targets:• Compatibility with conventional designs in terms

of mounting dimensions• 10% or more weight reduction compared with

conventional designs• At least 20% reduction in the materials used for

the hub and outer rings whose masses accountfor a large portion of the bearing product

• Functions including rolling fatigue life, mechanicalstrength and rigidity of our novel hub bearingdesign are equivalent to or better than those ofconventional designs, even though our hubbearing design does not adopt special materialsand/or heat treatment technique.

• Production cost equivalent to that of conventionaldesignsTo fulfill these targets, we have improved forging

techniques in the manufacturing aspect, and fullyutilized structural analysis techniques including FEM inthe design aspect. We have optimized the bearingshape so that the amount of material scrapped in theforging process and the amount of material removed inthe machining process are minimized, while still

Isao HIRAI*

*Axle Unit Engineering Dept. Automotive Sales Headquarters


-63-

Hub bolt

Hub flange

Outerring

Inner ring

Outer ring pilotHub pilot

Hub ring

Rollingelements

Fig. 1 Structure of Gen3 hub bearing for non-driven wheel

Non-commonflange shape

Fig. 4 Cross shaped hub flange

Hub flangeConvex on outercircumference

Concave on outercircumference

Fig. 5 Shape of developed hub bearing

Hub flange

Outer ring flange

Fig. 2 Round shaped hub flange

Lightening hole

Hub bolt hole

Fig. 3 Lighter holes in hub flange

maintaining functionality of the hub bearing. Thus, wehave achieved a light weight hub bearing product aswell as a reduction in the amount of materialsconsumed in production.

2.1 Hub flangeFig. 1 shows a typical example of the structure of a

Gen3 hub bearing for a non-driven wheel. The hub ringas a rotary body has a hub flange that functions as amount for installing the wheel and brake disk. The mostcommonly used hub flanges are disk type hub flangessuch as a one shown in Fig. 2. Light-weight hub flangeproducts are available; examples of which include ahub flange having lightening holes as shown in Fig. 3and a cross shaped hub flange (non-common shape)shown in Fig. 4.

A hub ring is prepared by obtaining a work blankfrom a round steel bar and then subjecting the blank toforging, turning, heat treatment and grindingprocesses. The ratio of the mass of the steel blankwork loaded into the forging process to the mass of thefinished hub ring is known as “material yield”.

The lightening holes shown in Fig. 3 are formed bypunching during the forging process. Despite itslightening effect, this technique leads to deterioratedmaterial yield. When a cross shaped hub flange ismanufactured, a work blank is first forged into a shape

somewhat larger than the flange design, and then thework piece is punched to the final flange shape, andthis technique also leads to deteriorated material yield.

In order to achieve not only improved material yieldbut also a lighter hub ring, we have developed a hubbearing shape shown in Fig. 5. In finalizing this shape,we have adopted an FEM analysis technique tooptimize the mechanical strength and forgeability of thehub flange. More specifically, concaves of a largerradius are included in the outer circumference of thehub flange so that the convexes on the outercircumference are readily formed during the forgingprocess and the number of portions which can laterdevelop intermittent cutting during the turning process(which means deteriorated workability).


-64-

2.3 Outer ring pilotWhen a hub bearing is installed to the car body

(suspension), the outer ring of the hub bearing offers amounting point to the car body. The outer ring has aring pilot that serves as a guide.

We have again chosen an intermittent configurationfor the outer ring pilot on our hub bearing in order toreduce the area being machined and realize a lighterouter ring design. However, unlike the hub pilot, theinner bore of the outer ring pilot is fitted with a seal thatprevents ingress of water and dust into the hubbearing. Therefore, as can be understood from theviews given in Fig. 9, the inner edge of the outer ringpilot is continuous while the outer edge features adiscontinuous form.

Incidentally, we have adopted a continuousmachined surface around the flange of outer ring pilotin order to improve accuracy for installing the hubbearing and prevent corrosion of the flange surfacedue to the ingress of muddy water. Like in the case ofthe hub pilot, areas subject to intermittent cutting on

2.2 Hub pilotThe hub ring is provided with a hub pilot that serves

as a guide for the wheel and brake disk installation onthe hub. Previously, the outer circumference of the hubwas machined to generate a hub pilot in the form of acylindrical-sectioned belt as shown in Fig. 6, and themass removed by the machining process wasscrapped as chips.

In the case of our newly developed hub bearing, toachieve lightening as well as reduce the portionremoved by machining, the previous pilot (cylindrical-sectioned belt) has been superseded with anintermittent hub pilot shown in Fig. 7 that consists of aplurality of teeth formed by forging.

An intermittent pilot shape can contribute tolightening as well as improved material yield owing to areduction in surface being machined. However, thisfeature necessitates intermittent cutting, which leads toshorter tool life. At the same time, because this pilotshape involves an increased number of edges on thework piece, a chamfering process is needed to removeburrs.

Fig. 6 Example of conventional shaped hub pilot

Hub pilot

Fig. 8 Measure for intermittent cutting of hub pilot

Intermittent hub pilot

Ramps are formed at both ends on eachpilot tooth along the circumferential direction.

Fig. 7 Discontinuous shaped hub pilot

Intermittent hub pilot

Convex on outercircumference

Concave on outercircumference

To address this problem, ramps are formed during theforging process at both ends on each pilot tooth alongthe circumferential direction as shown in Fig. 8. Thisarrangement helps not only mitigate the impact thatoccurs when the cutting tool comes into contact withthe pilot teeth but also inhibits the occurrence of burrs,thereby extending the life of the cutting tool anddecreasing the number of machining steps.Furthermore, as can be understood from Fig. 7, thepilot is situated in a location whose phase is the sameas that of a concave on the outer circumference of ourhub flange; consequently, the difference in cross-sectional area between a convex and a concave on theouter circumference of our hub flange is minimized andforgeability of the hub flange is improved.

-65-

Table 2 Results of fatigue strength test

New design

Conventionaldesign

Number of revolutions at least 1.9 timesas many as the targeted value

No fracture on flange

Number of revolutions at least 1.7 timesas many as the targeted value

No fracture on flange

Subject of evaluation Result

Fig. 9 Shape of outer ring pilot

Convex on outercircumference

Machined surface onouter circumference

Seal

Discontinuous outerring pilot

Continuous machined surfaceadjacent to the flange, aroundentire circumference

Enlarged view

Outer circumfe-rence surface

Inside bore surface

the outer ring pilot have outward inclined ramps formedby forging in order to help extend the useful life of thecutting tool used.

As can be understood from Fig. 9, the portions beingmachined on the outer circumference of the outer ringpilot are situated in the same phases as those of theconvexes on the outer circumference on the hubflange. This arrangement is intended to promote easeof installation of the hub bearing to the car body andimprove mechanical strength and rigidity of theinstallation. Because projection of the outer ring flangeis small, we can give priority to this objective ratherthan improvement in its forgeability as described inSec. 2.2.

2.4 Lightening and reduced material consumptionAs a result of the reduction in material consumption

and improved material yield, our new design hasachieved the following improvements overconventional designs: approximately 10% weightreduction of the hub bearing, approximately 20%reduction in material consumption with the hub ringand outer ring, and approximately 30% reduction in thetotal amount of material scrapped as a result ofmachining for the hub ring and outer ring.

3. Evaluation test

We first performed the theoretical review by FEManalysis technique of our novel “resource-savingautomotive hub bearing design” that reflects thepreviously mentioned lightening technique and materialconsumption reduction technique. Then, we havesubjected the prototype to a bench test looking at fivecharacteristics, whereby we can verify that our newdesign is equivalent to the conventional design interms of functionality:(1) Bearing rolling fatigue life: life (durability) test for

bearing subjected to turning load on the car(2) Fatigue strength of hub flange: fatigue strength test(3) Rigidity of hub bearing: rigidity measurement(4) Mechanical strength of hub bearing: static strength

test(5) Rotational balance of non-common shape flange:

unbalance measurement

3.1 Bearing life (durability) test for bearingsubjected to turning load on the car

The findings from the bearing life test are given inTable 1. The measured rolling fatigue life of our newdesign is more than six time as long as the targetedrating life. Thus, it has been proven that our newbearing design has sufficient durability.

Table 1 Results of durability test with curving moment

Conventional design At least six times as long as rating life

Subject of evaluation Result

3.2 Fatigue strength testAs summarized in Table 2, the samples have been

run for a number of revolutions in excess of a targetnumber. Though having undergone a test whose runtime was longer than that of a conventional hub flangedesign, our newly developed hub flange has notdeveloped any fractures, thereby we have determinedthat our new design has sufficiently high fatiguestrength.


3.4 Static strength testThe mechanical strength required for our bearing in

the static strength test is that the bearing does notdevelop a fracture when subjected to a lateral G loadof 2.0 in the vehicle turning mode (side collisionsituation where the wheel has hit a curb). As can beunderstood from the information given in Table 4, ournew bearing design satisfies the load requirement andhas mechanical strength equivalent to that ofconventional bearing designs.

-66-


Fig. 10 Image for comparison

Lighterweight

New designConventional design

Cost

Wastereduction

Reducedmaterial input

Rotationalbalance

Mechanicalstrength

Rigidity

Bearing lifeBetter

3.5 Unbalance measurementTo be able to determine the magnitude of effects of

a non-common hub flange shape on the rotationalbalance of the hub ring, the unbalance of the hub ringneeds to be measured. The resultant unbalance is ofthis design is better than G16 per JIS B 0905 and thislevel well satisfies G40, which is the levelrecommended by the same JIS (Japanese IndustrialStandard) for quality of balance with automotivewheels.

From these findings, we are sure that our novel“resource-saving automotive hub bearing design” hasfunctions equivalent to those of conventional hubbearing designs. The chart in Fig. 10 provides acomparison between our new hub bearing design andconventional design in terms of functionality andadvantages in manufacture.

4. Afterword

To be able to contribute to the reduction of CO2

emissions, NTN has designed a unique hub bearingproduct that boasts lighter weight through improveddesign as well as a reduction in materials consumedthrough an improved forging technique.

The applicable specification for individual NTN hubbearings can vary depending on the vehicle typeand/or location of use. It is possible that all theconsiderations described in this document may not beapplicable for the intended usage; notwithstanding,NTN will as necessary expand the scope ofapplicability of this product line. NTN is going tointegrate its already proven “hub bearing with built-inhigh-sealing low-torque seal” technology with its novel“resource-saving automotive hub bearing” technologydescribed above. Through these efforts, NTN willpromote the reduction in weight and friction for hubbearings and thereby will further contribute toreductions in energy consumption by cars in travel aswell as proceeding with development of novel eco-conscious technologies.

References1) K. Shibata and T. Norimatsu: Technical Trends in Axle

Bearings and the Development of Related Products,NTN Technical Review No. 75 (2007) 29-35

Table 4 Result of static strength test

New design

Conventional design

Subject of evaluation Lateral G in turning mode equivalent to fracture load

2.4

2.4

Photo of author

Isao HIRAI

Axle Unit Engineering Dept.Automotive Sales Headquarters

Table 3 Measurement of stiffness with curving moment

New design

Conventional design

Subject of evaluation Rigidity of tilted hub flange kN/˚

21

22

3.3 Rigidity measurementTable 3 summarizes the results of the rigidity

measurement. The rigidity of our newly developed hubflange is virtually equivalent to that of a conventionaldesign.

-67-


[ New Product ]

V-series Hub Joint

1. Preface

A typical environmental issue to be addressed todayis global warming. Recently, emissions reduction ofCO2, the most abundant greenhouse gas, has beenposing a challenge that needs to be addressedurgently. The car-manufacturing industry around theworld has been committed to develop cars with betterfuel efficiency in order to help reduce CO2 emissionsfrom their products. As a result of this, the globalmarket for hybrid electric vehicles has beendramatically expanding.

In this paper, we will provide information about our Vseries hub joint products (hereinafter, may be simplyreferred to as V series H/J) that are modularizedproducts each obtained by joining a hub bearing(hereinafter, may be referred to as H/B) with a constantvelocity joint (hereinafter, may be referred to as CVJ)by a “press-cut fitting” technique that is NTN’s proprietyjoining technique; consequently, the V series hub jointsboast improved ease of assembly and productivity forour clients as well as a 12% weight reductioncompared with conventional products.

2. Structure and advantages

Hub joints (hereinafter, may be referred to as H/J)are products each integrally comprising H/B and CVJ;

To improve fuel efficiency, reducing the weight of axle units isdesirable. NTN has developed various hubbearing and CV jointproducts to meet this requirement. This report introduces the V-series hub joint, which is an integrated hub-bearing and CV jointcrafted using NTN’s original press-cut fitting. This productachieves a lighter weight through integration of a hub- bearingand a CV joint

in particular, H/J products adopting press cut fittingtechnique are referred to as V series H/J products.NTN has developed two types of V series H/Jproducts: “non-separable type” that boasts the lightestweight, and a “separable type” that features separableconstruction that allows a bearing user to be able torepair the purchased bearing product easily.

This report hereunder describes the informationabout the press cut fitting technique as well as thestructural features of “non-separable type” and“separable type” of V series H/J products.

2.1 Previous fitting techniqueA previous NTN fitting technique for H/B’s with CVJ’s

is hereunder described. Fig. 1 illustrates the structureof NTN’s previous serration fitting joint. Both the outercircumference of the outer ring stem of the CVJ andthe inside bore surface of the H/B are serrated. Torqueis transmitted via the serrated joint between bothserrated members. Also, the CVJ is secured to the H/Bwith a hub nut tightened onto the threaded end of CVJouter ring stem.

Torque transmission from CVJ to H/B takes place atthe contact surfaces between tooth flanks of bothserrated portions. However, because of the limitation ofmachining accuracy with both members (difference intooth pitch between both serrated parts), it is difficult torealize a tight fit between both members.

Mitsuru UMEKIDA**Yuuichi ASANO**

**C.V.Joint Engineering Dept. Automotive Sales Headquarters**Axle Unit Engineering Dept. Automotive Sales Headquarters


-68-

Serration on outer circumference of outer ring stem ofCVJ is engaged with serration on inside bore of H/B.

CVJ is secured with hub nut.CVJ outer ring stem

H/B

Fig. 1 Fitting by serration (previous method)

Variation in gap between tooth flanks

Gap betweentooth flanks

Fig. 2 Contact condition of teeth (previous method)

[CVJ outer ring stem]• Special serration (grooves and ridges)• Hardened through heat-treatment

[Inside bore surface of H/B]• Cylindrical form (before fitting)• Non-hardened (not heat-treated)

Fig. 3 Press cut fitting (before fitting)

No gap between tooth flanks

Boundary of serration

Fig. 4 Contact condition of teeth (Press cut fitting)

Consequently, variation in the gap between toothflanks can occur as shown in Fig. 2, and excessiveplay can occur on the joint in the circumferentialdirection with the previous fitting technique.

There has been a method available for eliminatingcircumferential play between CVJ and H/B, that is, theprovision of a helix angle on the serration of the CVJ.However, this method has its drawbacks. As non-uniformity in contact pressure occurs across the tipside and the root side of the serrated shafts, the entiretooth surface on the serrated shaft may fail to uniformlybear the load owing to the difference between the toothpitch of the CVJ side serration and that of the H/B.

As explained above, the serrated joint of theprevious method fails to evenly bear the torque andtherefore, the design length of the mesh between bothserrations has to be longer.

2.2 Joint formed by press cut fitting techniqueWith the previous fitting method, both the CVJ and

H/B are serrated. In contrast, with our press cut fittingmethod the CVJ outer ring alone is provided with aspecial serration as shown in Fig. 3, while the insidebore of H/B remains cylindrical.

With our press cut fitting system, the CVJ outer ringstem is hardened through heat-treatment, while theinside bore surface of the H/B is not hardened so thatthere remains a difference in hardness between theCVJ outer ring stem and inside bore surface of theH/B. Consequently, when the CVJ outer ring stem isfitted into the H/B inside bore, the serration on the CVJouter ring stem generates serrations on the inside boresurface of the H/B. NTN calls this unique fitting method“press cut fitting”.

During the press cut fitting process, the inside boresurface of the H/B is serrated as this surface is underelastic deformation. As a result, the ridges found on theCVJ outer ring stem fit into the grooves formed on theinside bore surface of the H/B while a tighteningallowance is present between the ridges and grooves.Thus, a preload is exerted onto the resultant joint, andthe CVJ outer ring stem is tightly fitted within the insidebore of the H/B as illustrated in Fig. 4.

-69-

As a result of this unique fitting method, the torqueinput is uniformly borne by all the teeth and toothflanks, and the allowable torque capacity of this joint isgreater, thereby the necessary length of mesh betweenserrations can be much shorter compared with theprevious fitting method.

2.3 Structure of V-series hub jointAs previously mentioned, the V-series H/J products

subjected to press cut fitting method can becategorized into “non-separable type” and “separabletype”. The structures of these product types arehereunder described.

2.3.1 Non-separable H/JThe “non-separable type” products boast lighter

weight through integration of the H/B with the CVJ. Fig.5 shows a cross-sectional view of a V-series non-separable H/J product.

As previously mentioned, both members in a jointremain in full a contact state, and do not come apartwhile the hub joint is running because a preloadremains applied to the joint. Nevertheless, to preventaccidental disengagement of the CVJ outer ring stemin a non-separable type from the H/B, the outer end ofthe CVJ outer ring stem is coined to secure the CVJ tothe H/B.

The V-series H/J products boast lighter weightbecause the length of meshed serration is shortercompared with previous products and they do not havea hub nut.

With the previous fitting method (Fig. 1), whentorque from the CVJ is input to the H/B and sudden slipdevelops on the contact surfaces between theoscillation-capable coined area of the H/B and the rearface of CVJ outer ring, a unique noise known as “stick-slip sound” can occur. Reduction in bearing pressure

Caulked area on stem end

Oscillation-capablecaulked area

Rear face of outer ring

Fig. 5 V-series H/J (Non-separable type)

Fig. 6 Structure and characteristics of V-series H/J Non-separable type

on the contact surfaces is effective in preventingoccurrence of the stick-slip sound.

Unlike the previous structure, the joint formed by ourpress cutting fitting method is resistant againstaccidental disengagement, and is capable ofcontrolling the axial pressing force during the presscutting fitting process; thereby, the bearing pressure onthe contact surfaces between the CVJ and the H/B canbe reduced. Thus, our press cut fitting method is usefulin preventing the stick-slip sound.

[Structure and features of non-separable H/B]Fig. 6 shows the structure and features of NTN non-

separable H/B.1 Press cut fitting: No gap between contact surfaces

of serration on H/B and that on CVJ2 Deletion of flange on H/B outer ring, that is, a flange

for installing a knuckle: Adoption of a press-fittingtechnique for weight reduction

3 Deletion of hub pilot: Lighter weight (Area A onflange outer circumference shown in Fig. 6 servesas a pilot)

6 Coining at stem end.

3 Deletion of hub pilot

A 2 Deletion of H/B outer ring flange

5 V series CVJ

1 Press cut fitting

4 Reduced bearing pressure on contact surfaces between H/B and CVJ

V-series Hub Joint


-70-

Fig. 7 V-series H/J (Separable type)

4 Reduced bearing pressure on contact surfacesbetween H/B and CVJ: Structure to prevent stick-slip sound

5 V series CVJ6 Coining at stem end

2.3.2 Separable H/JA hub joint consists of a constant velocity joint

integrated with a hub bearing. To help realize easyrepair work for a marketed hub bearing, NTN hasdeveloped a “separable” hub joint design that iscapable of disassembly and reassembly. Figs. 7 and 8show the structure of our separable hub joint.

To allow the H/B to be separated from the CVJ, theCVJ outer ring stem is secured by fastening a bolt,instead of coining at the stem end on the non-separable H/J. This bolt is also used to retract the CVJouter ring stem into the inside bore of the H/B duringthe reassembly procedure for the separable H/J. To

allow the CVJ outer ring stem to be pulled in, the H/Bhas an inside wall and the CVJ outer ring stem has athreaded hole along its centerline to accept this bolt.

This unique structure allows the H/B to be separatedfrom and reinstalled to the CVJ as necessary.

As mentioned previously, a joint formed by our presscut fitting technique is preloaded and as a result theH/B cannot be readily separated from the CVJ, unlikethe joint obtained from the previous method describedin Sec. 2.1; this drawback remains true when the H/Bis reinstalled to the CVJ. The unique advantage of ournewly developed “separable” hub joint is being able todisassemble and reassemble the joint, which theprocedures for our “separable” design are brieflydescribed below.<Disassembly procedure>

Referring to Fig. 9, fit the stem push-out bolt into thethreaded hole on the CVJ outer ring, and then mountthe pulley extractor onto the H/B. Turn the stem push-out bolt to push out the CVJ outer ring stem. Then, theH/B can be separated from the CVJ.

HB inside wallThreaded hole

Fig. 9 Separate method of separable type

Stem push-out bolt

Pulley extractor

Bolt

Fig. 8 Structure and characteristics of V-series H/J (Separable type)

Stem pull-in bolt

Press cut fitting

Phase-matching guide slot for reassembly

<Reassembly procedure>The joint formed by our press cut fitting technique is

a tight-fitting joint that’s being preloaded. If a H/B isseparated from a CVJ, it will be impossible to readilyreestablish circumferential phase-matching betweenthe ridges on CVJ outer ring stem and the grooves onthe H/B. To address this drawback, the entrance toinside bore of the H/B in our “separable” design has aphase-matching guide slot that is geometricallysituated in a circle greater than the one defined by theridges of the serration on the CVJ outer ring stem.

For reassembly, the CVJ outer ring stem is guidedalong the guide slot on the H/B, and the stem pull-inbolt is tightened to a recommended tightening torque.

-71-

3. Function assessment test

The functions required of a joint formed by press cutfitting technique are sufficient static torsional strength,torsional fatigue strength against torque transmittedfrom the CVJ to the H/B, and durability of the jointagainst a bending moment that occurs when the CVJ isoperating at a greater angle.

In the assessment test, we used a CVJ whose presscut fitting joint that has sufficiently great static torsionalstrength and tested the CVJ samples under the testconditions that led to a fracture of the press cut fittingjoint.

(1) Static torsional strength test(with non-separable type)

Fig. 11 graphically plots the test result from thestatic torsional strength test with non-separable

specimens.The non-separable product has sufficiently high

torsional strength as the specimens of non-separabledesign satisfy the strength target while the press-fitallowance (difference between the outside diameter ofserration on CVJ outer ring stem and the insidediameter of H/B) remains in the standard range.

Fig. 10 Re-assemble method of separable type

Phase-matching guide slot for reassembly

Serration on CVJ

Stem pull-in bolt

Guide slot on H/B

Fig. 11 Static torsional strength test

Static torsional strength

Frac

ture

torq

ue

Standard press-fitting allowance

Target

Press-fitting allowance for press cut fitting technique

Fig. 12 Bi-directional torsional fatigue strength test

Bi-directional torsional fatigue strength

Torq

ue lo

adin

g

Number of repeated cycles

θ＝0˚

θ＝43˚Target

V-series Hub Joint

(2) Bi-directional torsional fatigue strength test(with non-separable type)

Fig. 12 graphically plots the test result from the bi-directional torsional strength test with non-separablespecimens.

Assessment has been performed with two types ofCVJ samples whose operating angles are 0 degreeand 43 degrees respectively. Both sample types havewithstood a targeted number of load applications andhave exhibited sufficient durability.

(3) Static torsional strength test (separable type)Fig. 13 shows the test result from the static torsional

test with separable specimens.Static torsional strength has been assessed with

samples having undergone disassembly/reassemblycycles. The samples having undergonedisassembly/assembly cycles do not show anyevidence of deterioration in mechanical strength, andboast sufficient mechanical strength.

-72-

Photo of authors

Mitsuru UMEKIDA

Axle Unit Engineering Dept.Automotive Sales Headquarters

Yuuichi ASANO

C.V. Joint Engineering Dept.Automotive Sales Headquarters


Fig. 13 Static torsional strength test (Separable type)

Static torsional strength: Separable

Torq

ue lo

adin

g

Target

No separation/reassembly cycle

One separation/reassembly cycle

Five separation/reassembly cycle

4. Afterword

NTN has successfully developed the V series hubjoint products that feature lighter weights by integratinga hub bearing with a constant velocity joint throughNTN’s propriety press cut fitting technique. The Vseries hub joints are categorized into two types: “non-separable” type boasting a light-weight design and a“separable” type featuring easy repair work. This paperhas presented information about the structures andadvantages of both types. By fully utilizing thetechnology described above, we will continue to remaincommitted to the development of novel corner moduleproducts for hybrid electric vehicles and electricvehicles.

-73-


[ New Product ]

Extremely High Load Capacity Tapered Roller Bearings

1. Foreword

People around the world have been increasinglyconcerned with prevention of global warming andimprovement in air quality. In this context, the level ofallowable vehicle emissions has been increasinglymore stringent, and a Japanese target value forimprovement in fuel economy for cars has beenestablished—average 23.5% improvement in fueleconomy by the end of fiscal year 2015 relative to thefiscal year 2004 level 1). While automakers have beenmore deeply committed to efforts for improved fueleconomy, engineering people specializing inautomotive transmission and differential gearing havebeen attempting to use less viscous oil and morecompact, lighter weight designs aimed at reducedfriction of automotive transmissions and differentials.To help contribute to this trend, the bearing industryhas been challenged with a life and rigidity requirementfor compact, light-weight bearings. To address thesechallenges, NTN has developed and marketed a line ofextreme high load capacity tapered roller bearingproducts 2), which result in lighter, more compactbearings at the same bearing life or a longer bearinglife bearing size.

These new products feature longer rollers: the lengthof rollers has been maximized by further incorporatingnew engineering improvements into the cage and innerring of the NTN’s already developed high load capacitytapered roller bearings. As a result of this new

NTN developed a tapered roller bearing with extremely high loadcapacity to improve the fuel efficiency of cars. This new bearinghas improved capacity from increasing the number and length ofthe rollers. This bearing is used for transmissions and differentialgears. This article introduces the design and performance of thisproduct that we have developed.

development, the NTN extreme high load capacitytapered roller bearing products boast much decreasedbearing pressure with a bearing under load; longerbearing life under severe lubricating conditions orcontaminated lubricating conditions; and higher rigidity.This paper describes the structure and advantages ofthis newly developed bearing product.

2. Structure of NTN extreme high loadcapacity tapered roller bearing

Previous NTN high load capacity tapered rollerbearings are characterized by a smaller clearancebetween the cage and outer ring as well as a greatercage pitch circle diameter—these features areintended to ensure presence of spaces sufficient forthe width of cage bars and reduced distances betweenrollers. Consequently, the number of rollers in any NTNhigh load capacity tapered roller bearing is virtuallyequivalent to that of a similarly sized full complementroller bearing.

In addition to these structural features of theprevious high load capacity bearing products, ournewly developed extreme high load capacity taperedroller bearings (the example in the right in Fig. 1) havea resin cage and maximized-length rollers, and areavailable in two structural types: (1) bearing withoutinner ring small side rib, and (2) bearing with inner ringsmall side rib.

Takashi UENO*Tomoki MATSUSHITA*

*Automotive Engineering Dept. Automotive Sales Headquarters

Standard taperedroller bearing

Extreme high loadcapacity bearing


-74-

Standard tapered roller bearing

Clearance betweenouter ring and cage

Smaller clearancebetween outer ring and cage

Cage bar width

Clearancebetween rollers


Smaller clearance between outer ring and cage

Roller length extended bylength of small side rib

1 Structure without inner ring small side rib

2 Structure with inner ring small side rib

High load capacity bearing Extreme high load capacity bearing

Cage bar width

Steel plateLock

structure

Without inner ringsmall side rib

Resin

Resin

With inner ringsmall side rib

(Smaller)

Fig. 1 Structure of extreme high capacity tapered roller bearing

(3) Lock structure with teeth on cageDimension of

cage protrusion

(2) Roller length is maximized to maximum allowable inner ring raceway width.

(1) Smaller dimension (thinner) of cage small diameter side rib

Fig. 2 Structure of extreme high capacity tapered rollerbearing without inner ring small side rib

Fig. 3 3D model of Extreme High Capacity tapered rollerbearing

2.1 Bearings without inner ring small side ribFig. 2 schematically shows the structure of a bearing

without an inner ring small side rib, and Fig. 3illustrates a 3D model of this bearing.

A bearing design without an inner ring small side ribis intended to be used in space constrainedapplications where the additional geometry required forthe rib would make the overall bearing too large. Thefeatures of this bearing structure are as summarizedbelow:(1) Cage: The small side rib has a minimum width just

sufficient for satisfying the mechanical strengthrequired of it.

(2) Roller length: The length of rollers may beincreased to the maximum width of inner and outerring raceways provided that the maximum

allowable cage protrusion is not exceeded.(3) Lock structure using teeth on cage: So that an inner

ring/rollers/cage ASSY can remain non-separable,a lock structure is provided which consists of teethon the large diameter side on cage, wherein theseteeth are engaged with the grooves formed on theouter circumference of inner ring large side rib.

2.2 Bearings with inner ring small side ribFig. 4 shows one example of a bearing structure

with an inner ring small side rib.There may be cases where the dimension of inner

ring width can be increased even though the allowabledimension of protrusion of cage is limited. Then, thebearing with inner ring small side rib may be adopted,and its features are described below.

-75-

(1) Cage: The cage is designed with the minimum ribwidth necessary to satisfy the mechanical strengthrequired of the cage small diameter rib under theintended bearing operation conditions.

(2) Roller length: The length of rollers are maximizedwhile leaving sufficient room for the small rib andhaving a cage small end protrusion which isacceptable.

3. Features of NTN extreme high loadcapacity tapered roller bearings

Thanks to increase in quantity and length of rollers,the NTN extreme high load capacity tapered rollerbearings boast functions improved over the NTNstandard tapered roller bearing products, and theexamples of improvement in functions are as describedbelow:

(1) Greater load rating¡Basic dynamic load rating: 16% increase at

maximum (64% increase at maximum in calculatedlife)

¡Basic static load rating: 21% increase at maximum(21% increase at maximum in safety factor)

(2) Greater rigidity¡Bearing rigidity: 14% improvement at maximum

(14% reduction at maximum in elastic displacement)

(3) Longer bearing life¡Useful life under clean oil lubrication condition is

extended.The use of an increased number of longer rollers

helps reduce the maximum bearing stress, leading toan increased oil film thickness and alleviation of stressthat occurs in a metal-to-metal contact mode.Consequently, occurrence of surface initiated flaking,

which can occur from metal-to-metal contact under acondition where formation of oil film is difficult, hasbeen prevented and bearing life has been extended.¡Useful life under contaminated lubrication condition

is extended.The use of an increased number of longer rollers

helps reduce the maximum bearing stress, limiting thesize of dent mark caused by trapped foreign matter.The stress occurring on the raised material around adent mark is also reduced thereby resulting in longerlife under contaminated lubrication conditions.

4. Performance of NTN extreme highload capacity tapered roller bearings

The NTN extreme high load capacity tapered rollerbearing differs from other NTN tapered roller bearingsin having an increased number of longer rollers and alock structure on the large side diameter of the cage.To study the effects of these special features on thebearing, we have performed a function assessmenttest comparing the standard tapered roller bearing andextreme high load capacity tapered roller bearing(without inner ring small side rib) shown in Table 1.

(1) Result of pumping performance testThe NTN extreme high load capacity tapered roller

bearing features a unique lock structure on the largerdiameter side of cage as well as increased number ofrollers and roller length. Fig. 5 provides visualcomparison between both bearing types in terms of thespace volume. It is apparent that the clearancebetween rollers is smaller on the NTN extreme highload capacity tapered roller bearing owing to increasein the number of rollers and as a result, the spacevolume in the bearing is smaller.

We have compared the flow rates of lubricating oilflowing through both bearing types in running mode.Fig. 6 schematically illustrates concept of the pumpingperformance test. When a bearing is run with theinterior of bearing and housing filled with lubricating oil,a pumping function takes place on the bearing, therebythe lubricating oil flows from the large diameter side ofthe bearing. Fig. 7 graphically plots the measuredamounts of lubricating oil pumped through the bearingsafter operating for a fixed duration. Though having asmaller space volume within its interior, the NTNextreme high load capacity tapered roller bearing offersa flow rate comparable to that of the standard rollerbearing sample.

(2) Roller settling performance comparison testWhen an NTN extreme high load capacity tapered

roller bearing without inner ring small rib is used, the


(1) Roller length is maximized to maximum allowable outer ring raceway width.

Dimension ofcage protrusion

(3) Inner ring small rib is added.

(2) Smaller dimension (thinner) of cage small diameter side rib

Fig. 4 Structure of extreme high capacity tapered rollerbearing with inner ring small side rib


-76-

Fig. 5 Space volume of bearing inside

roller settling performance in the bearing may differfrom that of other tapered roller bearings because ofabsence of inner ring small side rib that otherwiselimits the location of rollers.

Let us describe roller settling performance. When aninner ring assembly is fitted into an outer ring fromabove, a clearance will occur on the inner ring largeside rib since the length of rollers is usually smallerthan the inner ring raceway width (see Fig. 8).Because the rollers in a running bearing rotate whilebeing guided by the inner ring large side rib surface,the bearing needs to be subjected to a seatingoperation in order to allow the rollers to shift to normal

1.1mm0.6mmSpace

(a) Standard taperedroller bearing

(b) Extreme high loadcapacity bearing

Fig. 7 Test result of oil flow

Am

ount

del

iver

ed

ml

20

15

10

5

0

Loading conditions: Axial load Fa=500NBearing speed: 1500 min-1

Lubrication: ATF



Fig. 6 Evaluation of oil flow through bearing inside

N

Fa

Initial level oflubricating oil


Schematicview

Bearing size

Number of rollers

Length of rollers

Cage material

Load rating(comparison withstandard tapered

roller bearing)


φ25×φ52×16.25

Basic dynamic load rating Cr =31.5kNBasic dynamic load rating Cr =34.0kN

(8% increase)Basic dynamic load rating Cr =36.5kN

(16% increase)

Basic static load rating C0r =34.0kN

16 rollers

Basic static load rating C0r =37.0kN(9% increase)

Basic static load rating C0r =41.0kN(21% increase)

17 rollers 17 rollers

Steel plate Steel plate Resin

10.5mm 10.5mm 11.8mm

16.25

φ52

φ25

Dimension of cage protrusion1.5max.



16.25

φ52

φ25

16.25

φ52

φ25

Table 1 Comparison of bearing internal design

-77-

revolutions. The reason the NTN extreme high loadcapacity tapered roller bearing achieves good rollersettlement performance can be explained as follows:

Fig. 11 provides schematic diagrams for situationswhere the rollers in both NTN extreme high loadcapacity tapered roller bearing and standard taperedroller bearing are in settled state.

On the NTN standard tapered roller bearing, thedimension of roller displacement is limited by the endface of inner ring small side rib. In this state, the cageremains in contact with the roller large diameter sideend faces of rollers: consequently, as soon as break-inoperation begins, working forces occur betweenassociated components and generate a thrust thatforces each roller toward the inner ring large side rib;thereby while this thrust continues to lift up the cage,the rollers shift toward the inner ring large side rib.

In contrast, on the NTN extreme capacity taperedroller bearing, the dimension of roller displacement islimited by the end face of cage pocket small diameterside. As soon as break-in operation begins, the rollersshift toward the inner ring large side rib by a distanceequivalent to the distance between the rollers and cagepockets, causing them to come into contact with thecage pocket large side end faces; then like in amovement on the standard tapered roller bearing, the

Fig. 9 Testing method of roller settling

Dial gageWeight

Bearingbeing tested

Outer ringhousing

Loading conditions: Axial load Fa=302 NMeasuring method: Dimension of inner ring displacement resulting from revolutions of bearing is measured.

Fig. 10 Test results of roller settling

Dim

ensi

on o

f sin

king

of i

nner

ring

μ

m

Revolution

20

15

10

5

00 5 10 15 20

Extreme high load capacity bearing


Completion ofroller settlement Completion of

roller settlement

Fig. 11 Clearance to determine roller seating characteristics

Clearance between roller and cage pocket Clearance between roller and cage pocket

Position of settled roller

Clearance between rollerand inner ring large side rib

Clearance between rollerand inner ring large side rib

(a) Standard tapered roller bearing (b) Extreme high load capacity bearing

locations where they remain in contact with the innerring large side rib surface (considered fully seated).The lower the number of bearing revolutions neededfor the bearing to be fully seated means easier preloadsetting for the bearing.

Fig. 9 schematically illustrates the test method usedfor investigating roller seating performance on the NTNextreme high load capacity tapered roller bearing, andFig. 10 graphically plots the test results for rollerseating performance. The test result in Fig. 10 showsthat the rollers in the NTN extreme high load capacitytapered roller bearing fulfill settlement in about fiverevolutions while the rollers in the standard taperedroller bearing reaches settlement in about 13

Fig. 8 TClearance change between inner raceway width and roller

Some clearance No clearance

(a) Before seatingoperation

(b) After seatingoperation



-78-


Schematic view

Load rating

Bearing sizeBearing mass

(reduction in comparison withstandard tapered roller bearing)

Calculation resultfor running torque※ (reduction in comparison with

standard tapered roller bearing)


Basic dynamic load rating Cr=31.5kN Basic static load rating C0r=34.0kN

0.293Nm

0.129kg(16.2% reduction)

0.115kg(25.3% reduction)

0.287Nm(2.0% reduction)

0.271Nm(7.5% reduction)

16.25

φ52

φ25

Dimension of cage protrusion 1.5 max

15.4

φ49

.3

φ25


15

φ49

φ25


φ25×φ52×16.25

0.154kg

φ25×φ49.3×15.4 φ25×φ49×15

※Does not include stirring resistance on lubricating oil.

Table 2 Study of down sizing under same load capacity

Table 3 Calculation condition of rotating torque

load capacity tapered roller bearing: Table 3summarizes the parameters adopted for calculating therunning torque of bearing samples.

The high load capacity tapered roller bearing boasts16.2% reduction in bearing weight and 2% reduction inrunning torque.

Adoption of the NTN extreme high load capacitydesign helps achieve 25.3% reduction in bearingweight and 7.5% reduction in running torque. Thus thistapered roller bearing technology contributes to thegoal of reduced size (more compact), lighter weight,and reduced torque.

Radial load Fr 5000N

Axial load Fa 5000N

Bearing speed 5000min-1

Lubrication ATF

Lubricating oil temperature 90˚C

rollers shift to the inner ring large side rib end facewhile lifting up the cage.

As described above, the dimension of rollerdisplacement in the extreme high load capacity taperedroller bearing is greater owing to the structure of thisbearing: at the earlier stage of break-in operation, therollers alone, independent of the cage, shift towardinner ring large side rib; thereby the shift per revolutionof the bearing is greater and the rollers will settle in ashorter time.

5. Compact, light-weight designtechnique for NTN extreme high loadcapacity tapered roller Bearing

Demands will continue to grow for improved fueleconomy and riding comfort with automobiles, as wellas higher engine power and greater number oftransmission speeds. To help satisfy these demands,automotive bearings have to be capable ofwithstanding greater loads with no changes to theenvelope dimensions, or feature compact size andlighter weight with a given load bearing capacity. In thispaper, Table 2 below summarizes technical data forstudies into compact size, lighter weight design for asame set of a given load bearing capacity and a givenshaft diameter. Because a compact size, lighter weightdesign helps a tapered roller bearing to achieve lowerrunning torque, this paper includes the facts about thebearing torque reduction with the NTN extreme high

-79-

6. About special cage structure

By adoption of toothed lock structure on the largediameter side and narrower rib on the small diameterside on the cage of the NTN extreme high loadcapacity tapered roller bearing, the length of rollers canbe maximized. The shape of this cage is very unique,and the cage cannot be shaped with steel plate:therefore, it is made of a resin. Information about theresin used for this purpose and mechanical studyabout the special cage structure are hereunderdiscussed.

(1) Resin materialLubricating oil in a transmission and/or differential

sometimes contains phosphorus or sulfur content asextreme pressure additive which are known to bedetrimental to some resin materials. Therefore, theresin material for the cage needs to be resistantagainst oils containing these additives.

When considering ease of assembly into bearingand durability in bearing, each needed for a material oftapered roller bearing cage, the resin material of thecage also needs to have excellent physical propertiesincluding mechanical strength, toughness and heatresistance. Beginning with these considerations, NTNhas performed a necessary research and hassuccessfully developed a PPS (polyphenylene sulfide)resin cage that features physical properties needed forthe cage on the NTN extreme high load capacitytapered roller bearing.

Fig. 12 graphically plots the results of oil resistancetest with various resin cage materials. The PPS resindoes not exhibit deterioration in tensile breakingstrength even after undergoing immersion for 2,000hours: it is apparent that this resin material has

sufficiently high oil resistance performance.If our extreme high load capacity tapered roller

bearing is used in a lubricating oil that does not containa lot of phosphorus or sulfur content, it is possible toadopt PA46 (polyamide 46) or PA66 (polyamide 66)each often used as a resin material for a bearing cage.

(2) Dynamic analysis of extreme high loadcapacity cage

We have attempted to determine the mechanicalstrength needed for the cage small diameter side ribthat is a unique feature of the cage for the NTNextreme high load capacity tapered roller bearing, andhave found a minimum necessary rib width. To be ableto determine the stress occurring on the cage in therunning bearing, we have used an NTN-developed 3Ddynamic analysis tool for tapered roller bearings 3).

Fig. 13 shows one typical example of model fordynamic analysis of tapered roller bearing. Thisdiagram shows that the largest stress occurs at thesmall diameter side rib among various areas on thecage. In designing the small diameter side rib on thecage, we have determined the size of rib that hasnecessary mechanical strength by reflecting thefindings obtained from the dynamic analysis.

Fig. 13 Model of dynamic analyzed bearing

Inlet to load zone(small diameter side rib)

Area subject tomaximum stress

Fig. 12 Test result of oil resistance

PA66 PA46 PPS

Tens

ile b

reak

ing

stre

ngth

%

Immersion time h

4.0

3.5

3.0

2.5

2.0

1.5

1.00 500 1000 1500 2000

Oil bath: differential oil GL-5, oil temperature 150˚C



7. Conclusion

As the challenge to prevent global warming andimprove air quality continues, people committed toautomotive technologies have been more strenuouslyinvolved in improving fuel economy. Thus, demand forsmaller, lower torque, longer-life automotive bearingswill continue to increase.

This paper has presented information about the“NTN extreme high load capacity tapered rollerbearing” that boasts compact size, lighter weight andhigher rigidity by reflecting novel bearing designtechniques—“maximized roller length” and “adoptionof newly developed resin cage”—in the NTN high loadcapacity tapered roller bearing that has been alreadyrunning on actual automobiles.

By remaining committed to development of bearingtechnologies and products, NTN will further cope withincreasingly demanding operating conditions anddiversifying structures where NTN bearing products willbe used.

References1) Website of Ministry of Land, Infrastructure, Transport

and Tourismhttp://www.mlit.go.jp/kisha/kisha07/09/090702_.htm

2) T. Tsujimoto and J. Mochizuki: High Capacity TaperedRoller Bearings, NTN Technical Review No. 73 (2005)30-39

3) K. Harada and T. Sakaguchi: Dynamic Analysis of aHigh-Load Capacity Tapered Roller Bearing, NTNTechnical Review No. 73 (2005) 20-29

-80-

Photo of authors

Takashi UENOAutomotive Engineering Dept.

Automotive Sales Headquarters

Tomoki MATSUSHITAAutomotive Engineering Dept.


-81-


[ New Product ]

New ULTAGE SeriesEA and EM Types of Spherical Roller Bearings

1. Introduction

In contribution to realization of the ecology-orientedsociety, novel technologies have been developed inevery technical field for industrial machinery.

The bearings used on rotary mechanisms inindustrial machines need to satisfy requirements forimprovements in “bearing life”, “load capacity”, “high-speed performance”, and “handling quality”. NTN hasbeen marketing the “ULTAGE*” Series bearingproducts that boast the world’s best performance tomeet these requirements. The “ULTAGE” Series line ofbearing products already includes product series forprecision angular ball bearings and cylindrical rollerbearings intended for use on machine tool mainspindles.

Recently, NTN has added new standard productseries to the ULTAGE Series: they are new self-aligning roller bearings, EA Type (pressed steel cage)and EM Type (machined copper alloy cage) eachboasting higher load capacity and longer life.

*“ULTAGE” (a name created from the combination of“ultimate,” signifying refinement, and “stage,” signifyingNTN’s intention that this series of products beemployed in diverse applications) is the general namefor NTN’s new generation of bearings that are noted fortheir industry-leading performance.

For spherical roller bearings that are used in industrial machinery,including construction equipment and iron and steel facilities, we haveachieved the world’s best load capacity and possible rotation speedthrough major revisions to the internal design. We began sales of thesebearings as the new ULTAGE series of models in March 2009, allowingus to contribute further to creating a society that is more harmoniouswith the environment.

Pressedsteel cage

Induced thrust

Inner ringcenter rib

Asymmetricroller

Fig. 1 Spherical roller bearings (type B)

Yukihisa TSUMORI*

*Product Design Dept. Industrial Sales headquarters

2. Structure of new self-aligning rollerbearings in ULTAGE Series

Fig. 1 shows the structure of the NTN conventionalType B self-aligning roller bearing. The structure of thisroller bearing is unique in that the bearing hasasymmetric rollers, and its inner ring center rib bearsthe thrust induced by the running bearing. Due to beingguided by the inner ring center rib, the running rollersare maintained in stable attitude. This structure isappropriate for larger bearings and high-speedbearings.

At the same time, recent needs for self-aligning rollerbearings, other than high-speed capability, are longerlife and greater rigidity. To cope with these needs,bearing manufacturers have been developing bearingproducts with improved cage shape and more amplysized rollers.

-82-

Outer rings･ With oil inlets and oil groove･ Optional curvature

Inner ring･ No rib required･ Optional curvature

Rollers･ Larger rollers･ Maximized roller length･ Maximum number of rollers

Cage･ Window-type pressed steel cage･ Cage end faces function as guides･ Special surface treatment

Fig. 2 Ultage series type EA spherical roller bearings

Fig. 4 Design of cage pocket tabs Fig. 5 View of oil path

Oil path

Cage end faces function as guides

Four cage pockettabs stabilize theroller position.

Fig. 3 Cage guide type ofEA spherical roller bearings

Table 1 provides information about comparison ofseveral self-aligning roller bearing types. In order tocope with the above-mentioned needs for higher loadcapacity bearing products, NTN has been series-producing unique self-aligning roller bearings with theinner ring center rib deleted. They are the Type C self-aligning roller bearings having a guide ring and theType E self-aligning roller bearings whose rollers arestabilized in terms of attitude with the resin cage. NTNwill include these types in the ULTAGE Series bearingproducts.

Fig. 2 shows the structure and features of the newlydeveloped ULTAGE Series Type EA bearing. BothTypes EA and EM employ symmetric rollers, whereinthe diameter and length of the rollers have beenmaximized so that the bearings can withstand a much

greater load.In addition, by adoption of novel cage featuring

higher rigidity and simplified form, it has becomepossible to delete the inner ring center rib andincorporate a maximum number of rollers of maximizedlength.

As shown in Fig. 3, the pressed steel cage in theType EA bearing guides the rollers with its both endfaces. In addition, four tabs (protrusions) situated in thecage pocket shown in Fig. 4 help stabilize the attitudeof rollers while the bearing is running and generatesmooth lubricating oil flow in the bearing (Fig. 5).Furthermore, the entire surface of pressed steel cageis subjected to a special surface treatment to improvewear resistance of the cage.

Fig. 6 shows the structure and features of the Type

Tble 1 Comparison of spherical roller bearing type

Cross-sectional

plan

Rollers

Cage

Center rib

Asymmetric roller

Pressed steel

Yes

Symmetrical rollers

Pressed steel

None (w/guide ring)

Symmetrical rollers

Resin molding

None

Type CType B Type Type E


New ULTAGE Series EA and EM Types of Spherical Roller Bearings

-83-

EM bearing and Fig. 7 illustrates the machined hightensile brass cage. This cage adopts roller guide typein order to stabilize the attitude of rollers while thebearing is running. The right half and left half of thecage constitute a non-separable body and the innerring has the outer rib in order to prevent loss of rollers.The Type EM bearing is particularly advantageous inapplications subject to severe vibrations and impacts.

The dimensions in a Type EM bearing are same asthose of an equivalently sized Type EA bearing.

3. Advantages of ULTAGE Series self-aligning roller bearings

3.1 Load capacityMaximized roller diameter and maximized number of

rollers help the Types EA and EM bearings achieve theworld’s highest load capacity.(1) Basic dynamic load rating: Comparison with

competitors’ products is summarized in Fig. 8.(2) Basic static load rating: Comparison with

competitors’ products is summarized in Fig. 9.

3.2 Allowable speedEmploying a high-rigidity, pressed steel cage type,

the Types EA and EM bearings boast the world’shighest allowable speed as compared with thecompetitors’ catalogue values given in Fig. 10.

Fig. 7 Cage type of EM spherical roller bearings

One-piece machined cage

Bas

ic d

ynam

ic lo

ad r

atin

g k

N

300

250

200

150

100

50

0EA

(New design)Company

ACompany

BCompany

C

Fig. 6 Ultage series type EM spherical roller bearings

Fig. 8 Catalogue comparison of basic dynamic loadrating (Cr)

22216(φ80×φ140×33)

300

250

200

150

100

50

0

Bas

ic s

tatic

load

rat

ing

kN

EA(New design)

CompanyA

CompanyB

CompanyC

Fig. 9 Catalogue comparison of basic static load rating(Cor)

22216(φ80×φ140×33)

Allo

wab

le s

peed

m

in-1

5,000

4,000

3,000

2,000

1,000

0

Oil

Grease

EA(New design)

CompanyA

CompanyB

CompanyC

Fig. 10 Catalogue comparison of allowable speed22216(φ80×φ140×33)

Outer rings• With oil inlets and oil groove• Optional curvature

Rollers• Larger rollers• Maximized roller length• Maximum number of rollers

Inner ring• W/outer rib• Optional curvature

Cage• One-piece machined cage• Roller guide type


-84-

Fig. 11 graphically plots test results for bearingtemperature rise under circulation lubrication condition.It is apparent that the temperature rise on the TypesEA and EM bearings is limited compared with thecompetitors’ bearings, while the running torque is low.

Tble 2 Example, compact and lightweight

22220B(Conventional design)

22218EA(New design)

Model No.Load rating

(kN)Boundary

dimensions(mm)

Mass(kg)

Bearingvolume(cm3)

φ100×φ180×46

φ90×φ160×40

Cr Cor

315 810 4.95

600 3.28

25% 34%

415

385 398

Out

er r

ing

tem

pera

ture

˚C

90

80

70

60

50

400 2000 4000 6000 8000

Bearing speed min-1

EA (New design)Company ACompany B

Fig. 11 Test results of temperature rise22216(φ80×φ140×33)

Fig. 12 Grease application

Fig. 3 Assembly / disassembly

3.3 Operating temperature rangeThanks to adoption of a special heat-treatment

technique, the maximum allowable operatingtemperature is set as high as 200°C.

3.4 Compact and lightweight designBy developing the Types EA and EM bearings, NTN

has achieved the following improvements over theprevious designs: 65% increase at maximum in basicdynamic load rating, 35% increase at maximum instatic load rating, and 20% increase at maximum inallowable bearing speed.

A higher load capacity design technique helpsachieve a line of compact and lightweight bearingproducts. Table 2 summarizes an NTN’s attempt torealize such an improved bearing product. By applyingthis design technique, the size of conventional bearingNo. 22220B (dia. 100×dia. 180×46 in mm) can bereduced to that of a product of the new series—22218EA (dia. 90×dia. 160×40 in mm): this size

reduction means an approximately 30% reduction inboth bearing volume and mass. NTN believes that thisachievement will positively help machine designersdesign lighter, much compact industrial machines.

3.5 Improved handling qualityIntroduction of a simple-shaped window type

pressed steel cage helps simplify assembly,disassembly, and greasing works.(1) Easy grease application to roller surfaces (Fig. 12)

(2) Projection of rollers from the cage is small, androllers smoothly return to their normal state. Thus,the disassembly/reassembly procedure for thebearing is simple (Fig. 13).

-85-

4. Inclusion into standard ULTAGESeries products

The ULTAGE Series Types EA and EM self-aligningroller bearing products are now available in all theseries range with outside diameter 420 mm or smaller.

NTN is going to modify the conventional types (B, C,and E) to the new ULTAGE standards. For details, thereaders are encouraged to contact the NTN.

For the bearing number system for the newULTAGE Series products, see an example of bearingnumber shown below:

5. Conclusion

NTN has developed an improved functionalityversion of its standard self-aligning roller bearingproducts, and added these improved products to itsULTAGE Series—that is NTN’s next generationbearing product series.

Industrial machinery technologies have beenincreasingly innovated to help realize the ecology-oriented society. NTN believes that the new bearingproducts in the ULTAGE Series that boast greater loadcapacity and higher speed positively satisfy demandingmarket needs for lighter weight, smaller size and

longer life.NTN will remain committed to

development of bearing products of anext generation, through furtherimprovement, realization of advancedfunctionality and lower cost in order tocope with diversifying market needs.

Photo of author

Yukihisa TSUMORI

Product Design Dept.Industrial Salesheadquarters

Bearing Number

222 20 EA K D1 C3

Type code EA: Window-type pressed steel cage (EM: One-piece machined high-tension brass cage)

Dimension series code

Bore diameter code

Raceway configuration code No code: Cylindrical bore K: Tapered bore

Oil inlet/oil groove code D1: With oil inlet/oil groove (D1 specification only) W33: With oil inlet/oil groove (European market specification)

Internal clearance code C3: C3 clearance No code: Normal clearance

New ULTAGE Series EA and EM Types of Spherical Roller Bearings

-86-


[ New Product ]

Compact Free-type Torque Diode

1. Preface

Free type torque diodes (TDF) are often used as anautomatic switching mechanism to switch overbetween electromechanical mode and manual modeon rotary equipment and as a safety-protectionmechanism for electric motors. To be able to expandthe scope of applications of this product line, NTN hasdeveloped the “Compact Free Type Torque Diode” thatcan be used on compact precision equipment.

2. Dimensions and features

Figs. 1 and 2 provide information for comparison insize between the NTN newly developed product andconventional product.

The size of the newly developed product is 1/3 assmall as that of the conventional product.

NTN developed a free-type torque diode that rotates the output shaftby turning the input shaft , but does not rotate the input shaft by turningthe output shaft. Recently, NTN has developed a small and light free-type torque diode that we call a “compact free-type torque diode.” Thisreport introduces the structure and function of this product.

New design

Conventional design

Output shaftInput shaft

Cam faceOuter ring 10

16.7

φ30

.5

φ10

φ10φ

4φ4

Fig. 2 Comparison of size of developed product and current product

Fig. 1 Comparison of size of developed productand current product

Shoji ITOMI*

*Application Engineering Dept. Industrial Sales headquarters

New designConventional

design

The features of the newly developed compact freetype torque diode product can be summarized asfollows:

-87-


Fig. 3 Case of input shaft rotating

Fig. 4 Case of output shaft rotating (Input shaft not rotating)

[Features](1) A portion of the output shaft also serves as a

cam face: consequently, the size of output shafthas been greatly reduced to OD 10 mm××10 mm(1/3 reduction, compared to the conventionalproduct).

(2) Now, the housing and input shaft are made ofresin: consequently, the newly developedproduct weighs mere 28 g (2/5 reductioncompared to the conventional product).

(3) The area of sliding surfaces has been reducedto decrease the running torque* (resultanttorque is 5 mN-m or smaller; 1/25 reductioncompared to the conventional product).

(*The running torque mentioned here is a torque on theproduct in unloaded state.)

(4) Now, the housing is made of a resin; therebythe shape of mount area of the housing can bedesigned with greater freedom.

3. Operating principle of free typetorque diode

“Torque diode” can be defined as a reverse inputblocking mechanical clutch that transmits rotationaltorque from the input shaft to the output shaft but doesnot transmit rotational torque from the output shaft tothe input shaft. The reverse input blocking mechanical

clutch can be categorized into two clear-cut typesaccording to the behavior of the clutch in response toreverse input situation: “locked rotation type” thatprevents torque transmission to the input shaft, and“free-rotation” type whose output shaft turns freely anddoes not transmits torque to the input shaft.

Now, Table 1 summarizes the relation oftransmission of rotational torque from the input shaft tothe output shaft on a free type torque diode.

Figs. 3 and 4 illustrate the structure and operatingprinciple of the newly developed compact free typetorque diode.

The outer ring is coupled to the input ring, remainingcapable of transmitting the rotational torque from theinput shaft; and the inner bore surface of outer ring hasa cam face. The cage retains the rollers and balls,wherein the balls are forced to the outer ring end faceand bearing cover side face by spring force.

Table 1 Relation of input and output shaft rotationof Free type Torque Diode

(When the input shaft is turned, the output shaft rotates too.)

(When the output shaft is turned, the output shaft alone rotates.)

Input shaft Output shaft

Rotation

Does not rotate.

Rotation

Rotation

Rotational input from theoutput shaft

The output shaft remains at a standstill: the input shaft does not rotate.

No contact

Roller

Rotational input from the input shaft causes theouter ring to rotate, thereby the output shaft rotates.

Contact

Bearing cover side faceOuter ring end faceOuter ring

RollersSprings

Input shaft

Output shaft

Cage

BallCapable of

bidirectional rotation Capable ofbidirectional rotation


Variation in the torque transmission path isdescribed below between a case where the torque isinput from the input shaft and another case where thetorque is input from the output shaft.

3.1 Case where rotational torque is input fromthe input shaft

When the input shaft turns, the outer ring coupledwith the input shaft rotates; and the cage rotates in adirection same as that of the outer ring at a speedabout 1/2 as fast as that of the outer ring. The ballsretained by the cage roll on the outer ring end face andbearing cover side face; the rollers shift to the outerring end face owing to difference in rotational speedbetween the outer ring and cage, and then lockedbetween the outer ring cam face and output shaft; andthe output shaft rotates together with the outer ring(Fig. 3).

3.2 Case where rotational torque is input fromthe output shaft

When rotational torque is input from the output shaft,the rollers are held in location where they do not comeinto contact with the outer ring cam face; thereby, therollers are not engaged with the outer ring cam face;consequently, even when the output shaft rotates, theinput shaft does not rotate (Fig. 4).

NOTE) After having stopped rotation, the input shaft isturned slightly in the reverse direction: consequently,the rollers come off the cam face and take a relationshown in Fig. 4; and then rotation of the output shaft isnot transmitted to the input shaft any more.

Fig. 5 Forward and reverse rotation test machine

Compact Free Type Torque Diode

Bidirectional torque limiter (load)

Electric motor (input)Output shaft Input shaft

4. Endurance test

By using the forward-reverse rotation test machineillustrated in Fig. 5 and applying the endurance testconditions summarized in Table 2, durability of thenewly developed Compact Free Type Torque Diodehas been investigated by turning the input shaft inforward and reverse directions to turn the output shaft.

Table 3 summarizes the tests result obtained. Thereis not much difference found in the running torque,state of internal structure. And backlash between,before, and after operation for 11×104 cycles with thenewly developed Compact Free Type Torque Diode;this new product has proved its good durability.

Table 2 Test condition of endurance test

Running speed

Test equipment

Operating pattern

Torque loading

Temperature

Output shaft Dia.

100min-1

Forward and reverse rotation test machine(see Fig. 5)

50mN･cm (Loading is governed with a bidirectionaltorque limiter.)

Room temperature

φ4 mm needle pin is used(hardness number: HRC60 or greater).

One operating cycle consists of one forward rotation(1 sec.), one standstill (0.5 sec.), one reverse rotation(1 sec.) and one standstill (0.5 sec.).

Table 3 Test results

Running time

Rotational torquewith no load applied

State insidethe product

Backlash

110,000 cycles (61 hours)

Remains at 3 mN･cm, no change between beforeand after test operation.

Very minor wear on contact surfacesNo apparent variation in the amount of residual grease

No variation between before and after test operation

-88-

-89-


5. Intended application

The product will be conveniently used for switchingover between motor-driven mode and manual mode forlens zooming functions on still cameras and videocameras.

6. Conclusion

Previously, switch-over between electromechanicalmode and manual mode for rotary mechanism hasbeen achieved by switching over between paths fortransmitting rotational motion by actuating an electro-mechanical component such as an electromagneticclutch or a manually operated component such as ahand lever. Use of a free type torque diode helpsrealize automatic switch-over between these modeswithout using an electrical switch.

The newly developed NTN Compact Free TypeTorque Diode helps realize automatic switch-overbetween electromechanical mode and manual modefor small-sized mechanisms. NTN will not only remaincommitted to efforts for cost reduction and functionimprovement, but also propose other applications to beable to support activities to develop low energy-consumption equipment.

Photo of author

Shoji ITOMI

Application Engineering Dept.Industrial Salesheadquarters

-90-


[ New Product ]

Characteristics and Applications of DLC films

1. Introduction

NTN Group strives to reduce its environmentalfootprint and thus remains committed to continuedmitigation of environmental impacts caused by itscorporate activities through improving energy efficiencyand utilization of resources in its production activities.As an example, bearing engineers have beenproposing reduced use of rare materials, such as Mo,included in bearing lubricant, as well as reduction inthe amount and viscosity of the lubricating oil usedwhich, result in lower bearing torque. However, theseattempts can result in deteriorated lubrication of thebearing components, and conventional techniquesmay fail to sufficiently lubricate the bearingcomponents involved.

Surface modification appears to be a promisingtechnique that can dramatically improve the tribologicalcharacteristics of bearing components. Bearingengineers have been considering the adoption ofvarious hard films in order to alleviate friction and wearon sliding surfaces within bearings. In particular,Diamond-Like Carbon (DLC) film boasts higherhardness, excellent wear resistance, and lower frictioncoefficient and has been applied to automotive parts,dies, and tools.

There are various reports available aboutapplications of DLC films: examples state that when

From the view point of environmental and energy issues,vigorous efforts have been put into research and developmentactivities for tribological coating films in order to reduce friction andwear of sliding surfaces. In particular, diamond-like carbon (DLC)coatings have superior low-friction and anti-wear propertiescompared to other coating films, and they are attracting a great

amount of attention from tribology researchers and engineers. he friction and wear properties of DLCfilms, however, depend on their use conditions as well as their production processes, and therefore itis necessary to choose appropriate DLC films according to their use conditions and engineeringpurposes. This article introduces various properties of DLC films.

sliding in nitrogen gas or a vacuum, a-C:H(hydrogenated amorphous carbon), which is ahydrogen-rich DLC, exhibits an extremely low frictioncoefficient of 0.001 1), 2); another report says that ta-C(tetrahedral amorphous carbon), which is a hydrogen-free DLC, exhibits low friction when lubricated with oilin an automotive gasoline engine 3). A variety of DLCfilms are available, and their sliding properties can varysignificantly depending on coating method and/orconditions and the operating environment. Therefore,the optimal DLC film design must be selected based onthe expected operating conditions. The following articledescribes various properties of DLC films andexamples of their application to molding dies.

2. Features of DLC films

2.1 What is DLC film?A DLC film can be defined as an amorphous film

(Fig. 3) consisting of an irregular mixture of diamondatoms that form sp3 bonds (diamond structure shownin Fig. 1) and diamond atoms that form sp2 bonds(graphite structure shown in Fig. 2). To provide aconcept diagram that helps to better explain DLC films,Frrai and Robertson 4) have proposed a ternary phasediagram such as given in Fig. 4. In this diagram, thecorner at the top represents diamond, the lower leftcorner graphite, and the lower right corner hydrogen.

Kouya OOHIRA*

*Elemental Technological R&D Center

-91-

Fig. 1 Structure of diamond

Diamond

Organic polymer

GraphiteGlassy carbonPyrocarbon

Organic gas

sp3

sp2

ta-C:H

ta-C

ta-C ：Tetrahedral amorphous carbon a-C ：Amorphous carbon a-C:H ：Hydrogenated amorphous carbon ta-C:H ：Hydrogenated ta-C

a-C

a-C:H

H

Fig. 4 Classification of amorphous carbon

Fig. 2 Structure of graphite

Fig. 3 Structure of DLC films

Table 1 summarizes the general physical propertiesof carbon materials including graphite, diamond, andDLC films. Data for Young’s modulus, hardnessnumber, and the electrical characteristics of DLC filmare similar to those of diamond, and the thermalconductivity of DLC film is similar to that of graphite.The variation in physical property values of DLC filmsexist due to variation in proportion of sp3 bonds to sp2

bonds and hydrogen content 5, 6, 7).

sp2

sp3

Table 1 Comparison of characteristics of carbon material

Specific gravity

Specific electricresistance (Ω・cm)

Thermal conductivity(W/cm/K)

Lattice constant(nm)

Young’s modulus(GPa)

Hardness number(Hv)

Oxidation starttemperature

2.25

10-3

0.4～2.1

a=0.2456c=0.6708

(interlayer)

…

…

400～450

3.52

1012～1016

1000～2000

a=0.3567

…

1000～2000

10000～12000

600

1.0～3.0

109～14

0.2～30

…

…

100～800

1000～8000

300～500

Graphite Diamond DLC


2.2 Coating methods for forming DLC filmsDLC films can be created from solid carbon or

hydrocarbon gases such as methane, acetylene andbenzene and are usually formed in a vacuum chamber.Depending on the carbon source used, the DLC filmfalls within one of the following two categories:

(1) PVD: Physical Vapor DepositionA process that uses solid carbon as a starting

material, wherein solid carbon is evaporated and thenthe resultant carbon vapor is allowed to deposit on asubstrate to form the DLC film.(2) CVD: Chemical Vapor Deposition

A process that uses hydrocarbon gas as a startingmaterial, wherein the gas is decomposed in a vacuumchamber and then the resultant carbon vapor isallowed to deposit on a substrate to form the DLC film.

By using solid carbon as an evaporation source, thePVD process can produce a DLC film solely composedof carbon. In contrast, since the CVD process useshydrocarbon gas as a starting material, the resultantDLC film unavoidably contains hydrogen atoms ofapproximately 15 to 50 atm%. Either the PVD or CVDprocesses can be applied in various methods. Fig. 5summarizes typical coating processes applied to formDLC films.


Table 2 lists several general physical characteristicsof DLC films formed by various coating processes. Ithas been known that characteristics of synthesizedDLC films can vary greatly depending on startingmaterials, atomic elements added, and coatingconditions 5, 6, 7).

The UBMS (Unbalanced Magnetron Sputtering)process listed in Table 2 is one type of sputteringprocess, and is characterized by an unbalancedmagnetic field that is employed to expand a plasmalayer in the vicinity of the substrate. Because theplasma density near the substrate is increased, the ionassist effect (ions are allowed to collide the substrate)is enhanced; consequently, the film characteristicsincluding density and adhesion are improved 8).

The AIP (Arc Ion Plating) process uses a solidcarbon cathode and the chamber inner wall as ananode in order to trigger an arc discharge and allowcarbon on the target surface to evaporate. The

Table 2 Characteristics of various DLC films

Coating material

Coatingtemperature (˚C)

Hardnessnumber (Hv）

Surface roughness

Coverage

Adhesion

Wear resistance

Cost

PVD process

UBMS process AIP process Plasma CVD process

Solid carbon Solid carbon Hydrocarbon gas

＜250 ＜300 ＜200

2000～5000 5000～8000 1000～2000

CVD process

Fig. 5 Classification of coating methods

Sputtering process

Arc ion plating process

Ion vapor deposition process

Ion beam process

Laser abrasion process

Plasma CVD process

Thermal CVD process

Photo CVD process

PVDprocess

CVDprocess

resultant carbon vapor is ionized and is allowed todeposit on the sample surface to which a negative biasvoltage is applied. As a result, a DLC film with a higherdegree of adhesion is formed 5).

In a plasma CVD process, hydrocarbon gas (suchas methane, acetylene, benzene, etc.) is decomposedin a vacuum chamber to generate hydrocarbon ionsthat are accelerated and allowed to collide into asubstrate to which negative voltage is applied resultingin the DLC film. Because of using gas as a startingmaterial, the CVD process features better “coverage”and is suitable for forming a DLC film on samples withcomplex geometry. Also, with this process, a DLC filmof uniform thickness can be formed without the need toreorient the sample.

2.3 Comparison of DLC films with various otherhard coatings

2.3.1 Characteristics in generalTable 3 summarizes the characteristics of DLC films

and other hard coatings. DLC films boast particularlyhigh hardness and excel in anti-seizure quality.However, DLC films have an unavoidable drawbackthat is, poor adhesion to a substrate. Poor adhesionoccurs because stresses within a DLC film are high,and carbon atoms, which are chemically stable, do notreadily coagulate with a dissimilar material. Therefore,improved DLC film adhesion poses an engineeringchallenge.2.3.2 Friction wear resistance

Using the NTN Savin type wear test rig, NTN hasevaluated the sliding characteristics of DLC films(formed with UBMS process) and of various othernitride coatings.

This test rig is designed to maintain sliding contactbetween the flat test piece and a driven rotatingcylinder which is crowned. Fig. 6 illustrates theconfiguration of the NTN Savin test rig, and Table 4summarizes the test conditions applied. The test riguses a crowned cylinder so that uneven contact doesnot occur between the test piece and cylinder, thusachieving a much more accurate assessment of thefilm’s wear resistance. Fig. 7 summarizes the testresults, where the specific wear has been calculatedbased on the width of wear marks found on the testpiece and the friction coefficients that have beendetermined based on friction values measured with theload cell.

The specific wear of DLC film is 1/6 or lesscompared with that of TiN, which was the best of thenitride films tested. In addition, the friction coefficient ofDLC film is about 0.2 while the friction coefficients ofother hard films fall in a range of 0.4 to 0.8, indicatingthe DLC film features excellent friction characteristics.

-92-

-93-


Thus, the DLC film appears to be a promising meansfor creating a highly wear-resistant surface treatmentused on machine parts/components, dies, and tools.

Table 3 Characteristics of DLC films and other hard coatings

DLC

TiN

ZrN

CrN

TiC

TiCN

TialN

Gray to black

Gold

White gold

Silver white

Silver white

Purple to gray

Purple to black

1000～8000 Cutting tools, dies, functional films

Cutting tools, dies, jewelry

Jewelry

Machine parts/components, dies

Cutting tools, dies

Cutting tools, dies

Cutting tools, dies, jewelry

2000～2400

2000～2200

2000～2200

3200～3800

3000～3500

2300～2500

Film type Color Corrosionresistance ApplicationsOxidation

resistanceSeizure

resistance AdhesionHardnessnumber (Hv)

Table 4 Test conditions

Testpiece

Matingmaterial

Load

Speed

Time

Environment

Hardened SUS440C stainless steel, OD 48 mm,thickness 7mm, surface roughness 0.005 mmRa

50 N (max. contact pressure 0.5 GPa)）

0.05 m/s

3 min (sliding distance 9 m)

Under blowing of dry air, temperature 0–20%RH

Quenched and tempered SUJ2 steel, curvature R60 mm,hardness Hv784, surface roughness 0.01 mmRa

Fig. 6 Wear test rig

Load cell Weight

Test piece

Matingmaterial

Fig. 7 Wear property and coefficient of friction of eachspecimen specimens

0 0

1

Fric

tion

coef

ficie

nt a

t the

end

of t

est

40000

32000

24000

16000

0.2

0.4

0.6

0.8

8000

DLC TiN TiCN TiAlN CrN CrAlN CrSiNTiSiAlN

Spe

cific

wea

r ×

10-1

0 mm

3 /(N・

m)

3. Coating method-dependent variationin characteristics of DLC films

3.1 Sliding wear characteristicsFig. 8 plots the NTN Savin type wear test results of

three DLC film samples, wherein the DLC films ofthese samples were prepared by UBMS process, AIPprocess, and plasma CVD process previouslydescribed in Sec. 2.2. Table 5 summarizes the coatingmethods, hardness numbers and thicknesses of thesesamples. The test conditions are the same as thosespecified in Sec. 2.3.2 except in that the test durationwas changed to 30 minutes.

It is apparent that among three DLC film types, theUBMS process boasts best wear resistance and theAIP process features the lowest friction coefficient.

Fig. 8 Wear property and coefficient of frictionof each specimen

0Spe

cific

wea

r ×

10-1

0 mm

3 /(N・

m)

0

Frict

ion

coef

ficie

nt a

t the

end

of t

est

500

400

300

200

100

A（UBMS） B（AIP） C（CVD)

0.1

0.2

0.3

0.4

0.5

Table 5 Coating method, hardness, thickness of DLC films

Test piece Coating method Hardness, Gpa Thickness, μm

A

B

C

UBMS

AIP

CVD

26

53

15

1.2

1.0

2.0

3.2 Power of adhesion to substrateNTN performed an adhesion evaluation test per the

Rockwell indentation test method to determineadhesion of the DLC films which were tested in Sec.3.1 to substrates. The Rockwell indentation test forcesa cylindrical diamond indenter into each DLC filmsample under a particular load (1,470 N) to form anindentation mark on the film. The quality of adhesion ofthe DLC film to a substrate has been evaluated basedon the evidence of cracking or flaking of the filmaround the indentation mark. Figs. 9 through 11 showthe resulting test samples.

In this evaluation, DLC film (A) has not developedcracking or flaking: DLC film (B) has developed radialcracks, though not flaking: DLC film (C) has exhibitedsevere flaking around the indentation.

Generally, DLC films are prone to peeling becausetheir hardness is very high and internal stress exists:therefore realization of reliable adhesion to a substrateposes an engineering challenge. Techniques forimproving film-to-substrate adhesion include:


Fig. 9 Indentation of DLC film(A)

Fig. 10 Indentation of DLC film(B)

Fig. 11 Indentation of DLC film(C)

(1) Higher substrate hardness (created by nitriding,shot peening, etc.)

(2) Surface pre-treatment (by chemical etching,abrasive blasting, etc.)

(3) Adoption of an intermediate layer that is highlyaffinitive to both the substrate and DLC film

Through adoption of techniques best suited for theintended operating conditions, a DLC film will positivelyadhere to a substrate.

As described above, DLC films can have diversecharacteristics depending on coating method andconditions adopted. The optimal coating method andconditions for forming a DLC film need to be selectedaccording to the intended application and performancerequirements.

4. Applications to dies-DLC film applicationto dies for sintered metal moldings

In molding/forming work for sintered copper alloyparts, tungsten carbide dies are usually used becauseof their excellent wear resistance. However, in this typeof work, excessive wear is occasionally observed onthe dies even though the tungsten carbine materialboasts significantly higher hardness compared with thecopper alloy material. As a result, dies of much longerlife have been increasingly needed. To address thischallenge, NTN has attempted to adopt DLC films thatprovide much greater wear resistance.

Using the NTN Savin type wear test rig, NTN hasevaluated wear resistance of non-coated tungstencarbide test pieces and tungsten carbide test piecescoated with a DLC film. The DLC film used was formedusing the UBMS method. Table 6 summarizes the testconditions applied; Figs. 12 and 13 show wear marksfound on the sliding surfaces; and Figs. 14 and 15 plotthe resultant profile measurements on the wornsurfaces of the test pieces.

The tungsten carbide-only test piece has exhibited anoticeable wear mark on its sliding surface eventhough it has undergone sliding contact with a sinteredcopper alloy member (a relatively soft material). Themaximum depth of wear has reached 0.8 mm. Incontrast, the test piece consisting of tungsten carbide

0.1mm

0.1mm

-94-

Table 6 Test conditions

Testpiece

Matingmaterial

Load

Speed

Time

Environment

Tungsten carbide (WC-Co), OD 48 mm, thickness 7 mm,surface roughness 0.005 mmRa

50 N

0.1 m/s

30 min (sliding distance 180 m)

Under blowing of dry air, temperature 0–20%RH

Sintered copper alloy (Cu58%, Fe40%),surface roughness 0.3 mmRa

-95-


coated with DLC film exhibited a wear depth of 0.04µm which is 1/20 that of the uncoated test piece. TheDLC film has helped dramatically reduce wear on thetungsten carbide test piece.

The reason tungsten carbide (WC-Co) wears as aresult of sliding against the softer sintered copper alloymay be because the copper atoms diffuse into thecobalt particles at the contact interface. The cobaltparticles act as a binder in the tungsten carbidestructure. As a result of this diffusion, the Cuconcentration between the WC and Co particlesincreases, thereby deterioration the mechanicalstrength of the tungsten carbide structure. By coatingthe surface of a tungsten carbide die with a DLC film,which does not readily react with copper atoms whilemaintaining wear resistance, wear on the die is muchreduced. In addition, based on endurance test resultswith a die used commercially for sintered copper alloy,it has been verified that the life of the die coated withthe DLC film is much longer compared with a non-coated die of a same geometry.

5. Conclusion

As described above, the DLC film treatment is atechnique that is effective in extending life of dies andtools. Boasting a higher degree of wear resistance andlower friction, the DLC coating provides a promisingtechnology that helps address environmental andenergy issues. NTN will attempt to further develop thistechnology and expand the scope of its applications.

Fig. 12 Wear of sliding part (WC)

Test point for shape measurement

Fig. 13 Wear of sliding part (WC + DLC film)

Fig. 14 Surface profile of wear (WC) Fig. 15 Surface profile of wear (WC + DLC films)

References1) A. Erdemir et al., SAE Paper, 2000-01-05182) J. Andersson et al., Wear, 254 (2003), 1070-10753) M. Kano et al.: Proceedings of The Third Asia

International Conference on Tribology, October 2006,Kanazawa, 399

4) A.C. Ferrari, J. Robertson, Phys. Rev., B61, 14095(2000)

5) M. Ikenaga (ed.), Film Technologies for AdvancedFunctionality, Nikkan Kogyo Shimbun (2007)(Japanese)

6) N. Otake, Surfaces, Vol. 45, No. 9, 22-35 (2007)(Japanese)

7) M. Ikenaga and H. Suzuki, Principles of Super HardFilms Made through Dry Processing and theirIndustrial Applications, Nikkan Kogyo Shimbun (2000)(Japanese)

8) N. Otake: Application and Technology of DLC Films“Evolving Industrial Applications and FutureTechnologies of Diamond-Like Carbon,” CMCPublishing Co., Ltd. (2007) (Japanese)

0.3mm

Test point for shape measurement

0.3mm

Photo of author

Kouya OOHIRA

Elemental TechnologicalR&D Center

-96-


[ Technical Paper ]

Optimized Radius of Roller Large End Facein Tapered Roller Bearings

1. Preface

A basic tapered roller bearing consists of acompliment of tapered rollers that are arrangedbetween inner and outer rings having conicalraceways. The inner and outer ring of a tapered rollerbearing is commonly referred to as a cone and a cuprespectively. Because tapered roller bearings cansupport both a relatively high radial load and a single-directional axial load, they are widely used in drivesystems in automobiles and railway cars, steel millmachines, and other industrial applications. The apexof the inner and outer ring raceways and the rollercontact surfaces of a tapered roller bearing convergeon the bearing centerline. The rollers develop purerolling motion without sliding on the conical racewaysurfaces. However, when a radial load acts on thebearing, an axial component force occurs on the rollersdue to a disparity in angles between the inner andouter ring raceways. As a typical means for bearingthis force, a cone flange is provided. As a result, thelarge end face of each roller is guided while beingforced to the cone flange. Though each rollerundergoes pure rolling motion on the raceway surface,there remains rolling-sliding contact between the roller

Tapered roller bearings can support heavy combined radial and thrustloads, and are widely used for automobiles, railcars and industrialmachines. The bearings have an inner ring rib into which the large endsof rollers are thrust, and the contacts are accompanied with rolling-slidingmotions. The arge ends of rollers and the rib surfaces are spherical andconical in shape, respectively. The typical lubrication regime of the

contact is elastohydrodynamic lubrication (EHL). However, some surface damage may occur if the oil film formationis insufficient due to low-rotational and/or heavy-load operations. The film thickness should be thick enough toprevent damage in a given operating condition. In this paper, an EHL numerical model is developed inconsideration of both asperity contact and roller skewing. A parameter study shows that to form thicker oil films theoptimum radius of the large end face of a roller is about 85% that of the rib face conical surface.

large end face and the cone flange.Fig. 1 (a) shows a cross-sectional view of a tapered

roller bearing. The cone flange face usually consists ofa part of a conical surface perpendicular to the bearingaxis, while the roller large end face comprises a part ofspherical surface. This condition results in roller end-flange face point contact resulting in the highestpressures. The cone flange face may possess aconcave surface matches the contour of the rollerlarge end face. In this situation, the pressure betweenthese two faces will be smallest. However, even smallrotations of the roller on the y-axis (see Fig. 1 (a)) inthis condition can cause the edge of the roller largeend face to come into contact with the cone flangeface, possibly damaging the roller or the cone flangeface. To avoid this problem, the radius of curvature onthe roller large end face is designed to be somewhatsmaller than that of cone flange face. Note thatrotational motion on the y-axis is referred to as“skewing” while rotational motion on the x-axis is called“tilting”.

At the roller end-flange contact a sufficiently thick oilfilm may not be readily formed when the bearing runsat lower speeds or is insufficiently lubricated.Consequently, frictional torque on the bearing will

Hiroki FUJIWARA***Takashi TSUJIMOTO***

Kazuto YAMAUCHI***

*Elemental Technological R&D Center **Automotive Sales Headquarters Automotive Engineering Dept. ***Graduate School of Engineering Osaka Univercity

Modify the Japan Society of Mechanical Engineers. C. Journal (Vol. 75, 2009), pp-2319-2326.

-97-

Fig. 1 Tapered roller bearing

rml r

lg

r B

(a) Radial cross-section and legend

(b) Mode of contact between roller large end face andcone guide flange, and associated coordinates system

increase and under higher loads sliding will occur atthe contact surfaces. This will increase the possibilityof surface damage or seizure at these areas. Note thatthe situation where mating components haveinsufficient lubrication is known as “starvation”.

So far, the mode of contact between the roller largeend face and the cone flange face has been reviewedthrough experimental methods and theoreticalanalysis. The experimental methods include thefollowing examples. Yamada performed a series ofexperiments to study the effects of surface roughnesson oil film formation at lower speeds. This resulted inachieving lower torque at these contact surfacesthrough reduction in surface roughness of the rollerend face and flange face 1). Okamoto, et al. attemptedto reduce the distance from the roller large end-flangeface contact point to the cone raceway. The intent wasto reduce the slip ratio of these components in order toimprove their anti-seizure properties 2).

There are many examples of applying EHL(Elastohydrodynamic Lubrication) analysis. Jiang, et al.performed thermal EHL analysis with non-Newtonianfluids and discussed the relationship between theshape of the contact area and the oil film thickness 3).Nishida, et al. performed EHL analysis under a mixedlubrication condition 5) by introducing the Patir-Cheng

mean flow model 4).Incidentally, a tapered roller has a tendency to skew

even if factors such as misalignment are not present,as pointed out by Harada et al. 6). However, so far,there have been few examples of EHL analysis thatconsider roller skewing. Furthermore, there have beenno examples of experiments to optimize the radius ofthe roller large end face. We recently performed EHLanalysis considering roller skewing. Based on theresults obtained from this work, we now propose anoptimal radius for the roller large end face in order toform a reliable lubricating oil film. This optimal radiuswill be verified through a series of experiments.

2. Analysis of elastohydrodynamiclubrication (EHL) between roller endand flange faces

2.1 Isothermal EHL analysisEHL analysis was conducted for the contact area

between the roller large end and cone flange face oftapered roller bearings. The velocity, bearing load, andlubricating oil viscosity used for this analysis wereselected in order to minimize heat generation at thecontact area. Therefore, the temperature condition forthis analysis was assumed to be isothermal.Specifically, the following assumptions were made:

• Adequate lubrication exists.• Heat generation in the lubricating oil due to

shearing is ignored.Equations involving the following variables are

solved simultaneously:• Reynolds’ equation• Elastic deformation derived from semi-infinite elastic

contact theory• High-pressure viscosity properties of the lubricating

oil• Balance between force and momentAs shown in Fig. 1 (b), in the entire region of

contact, both roller end face and flange face have avelocity distribution in the x and y directions.Therefore, the Reynolds’ equation needs to incorporateconsiderations about stretch terms ( 3rd and 4th termsin the right-hand side in expression (1)). Consequently,the Reynolds’ equation will be expressed as:

（）（） ∂x

∂

∂x

∂(ρh)

∂x

∂p

12η

ρh3

um∂y

∂(ρh)vm

∂y

∂

∂y

∂p

12η

ρh3

＋

＋ ∂x

∂umρh＋

∂y

∂vmρh＋ …（1） =

where, h stands for clearance, ηviscosity oflubricating oil, p pressure, ρdensity of lubricating oil,um and vm x-direction component and y-direction


-98-

component of the mean velocity at the contact areabetween roller large end face and flange face,respectively.

To investigate the high-pressure viscosity property ofthe lubricating oil, Houpert converted the formula givenby Rolelands into the SI system 7). This formulaincorporates considerations for temperature change.However, because our study was intended to determinethe isothermal viscosity of the lubricating oil, we haveignored the terms associated with temperature andhave adopted the following expression:

2.2 Effects of surface roughnessWe adopted the mean flow model of Patir-Cheng 4)

to be able to determine the effects of surfaceroughness, and introduced the mixed fluid lubricationtheory 10) of Greenwood-Tripp. Incidentally, with thesetwo models, surface roughness is handled in aprobabilistic manner. In other words, the above-mentioned methods have a low probability ofcalculating results for pressure and film thicknessdistribution that agree with results obtained from thoseof a fixed contact area.

In our study, we only considered isotropic roughnessin the Patir-Cheng’s mean flow model 4). Whenassuming that the pressure flow factor is defined as:

｛｝［］ …（2）

………………………（3）

……………（4）

η=η0 exp［lnη0＋9.67］

z=

α0=（0.1657＋0.2332log10v）m×10-8

－1＋（l＋5.1×10-9p）z

where, η0 stands for atmospheric viscosity andis expressed as:

where, α0represents the viscosity-pressure coefficient,and is given by the following Wu-Klaus-Duda formula 8):

where, v means the kinematic viscosity (mm2/s), and mviscosity slope constant appearing in Walther-ASTMformula 9).

5.1×10-9（lnη0＋9.67）

α0

provided that, when

then, the Reynolds formula (1) will be modified asfollows:

σ

h＞5

A1=1.899

A2=1.126

α1=0.98α2=0.92α3=0.05

（） F2.5　　　 =

（）（

（）

） ∂x

∂

∂x

∂(ρh)

∂x

∂p

12η

ρh3

um2

ur－ui

t－ f *（t）dt

∂y

∂

∂y

∂p

12η

ρh3

＋

＋ ∂x

∂（ρφsx）

…………（7）

……………………………（8）

……………（9）

…………………（10）

…………………（11）

=

φx φy

σ

∂y

∂(ρh)vm

2

vr－vi＋

∂y

∂（ρφsy）＋ σ

∂x

∂umρh ρσφsx

2

ur－ui＋

∂x

∂ ＋

（）

（）

∂y

∂vmρh ρσφsy

2

ur－ui＋

∂y

∂

σ

hc

（）（） σ

hc

σ hc

（） F2.5σ

hc

（）｛ F2.5σ

hc （） σ

hc

σ hc

σ

hc 2

2

5

＋

If surface asperities of mating components come intocontact then the resultant contact pressure can be,based on Greenwood-Tripp theory 10), expressed as:

For function , Patir-Cheng offers the following

approximation 12):

pa= kc E'F2.5

∫ ∞

kc= π（Nβσ)15

4.4086×10-5

0

4－

β σ 8√ 2

6.804

＜4

σ hc＞4

=when

when

（）｛｝（）（）

………………（5）

………………（6）

φx=φy=1－0.90exp

φs= A1

－0.56（） σ

h

α1 2

σ

h

σ

h

σ

h

σ

h

and that the roller large end face is a smooth surface; and when adopting the shear flow factor 11) that resultsfrom these assumptions and is defined as:

exp －α2 ＋α3 ≤5

（） φs= A2 σ

h

σ

hexp －0.25 ＞5when

when

where, hc stands for the oil film thickness at themidpoint. For values of N, β and σ, availableresearch so far has attempted to define them in theform of Nβσand σ/β. Greenwood-Tripp haveassumed that Nβσ= 0.03 to 0.05 10). We also assumeNβσ= 0.05. Incidentally, Patir-Cheng assumes that σ/β= 100 12): however, through our experience, webelieve thatσ/β= 20 is a reasonable assumption.

Bearing engineers believe through experience thatthe boundary friction coefficient μa related with contactsurface between roller large end face and flange facefalls in a range of 0.12 to 0.15. For this study, we haveassumed μa = 0.12.

-99-

Optimized Radius of Roller Large End Face in Tapered Roller Bearings

2.3 Boundary conditionsIn ordinary EHL analysis works, researchers choose

an evaluation region whose size is several times largerthan the area of projected Hertzian contact surfaces sothat the projected pressure at the outer fringes in thisregion is zero. By contrast, when studying roller largeend-flange contact, the region of generated oil film(hatched area in Fig. 1(b)) is not sufficiently largerelative to the size of Hertzian contact surfaces.Therefore, we have introduced the boundary conditionsdefined below.

The contact pressure occurs only in the regionwhere the roller large end face contacts the flat surfaceof the cone flange face. The contact pressure in otherregions is assumed to be zero.

2.4 Assumptions about motionsWe developed the following assumptions about the

motions of the rollers, inner and outer rings:• Angular velocities of the rollers and inner and outer

rings occur due to the mechanical relationshipbetween them.

• Centrifugal force and gravitational force can beignored.

• Interface between the rolling surfaces of the roller andraceways and between the rollers and the cage donot affect roller skewing.

• The rollers do not tilt.• Skewing occurs due to the oil film pressure

distribution between the roller large end face andcone flange face, traction, and metal-to-metal frictionbetween surface roughness asperities.

3. Skewing on rollers in EHL mode

Generally, under an EHL (ElastohydrodynamicLubrication) condition, the pressure on the oil filmgradually increases at the fluid inlet side, and suddenlydrops at the outlet of the Hertzian contact area.Consequently, the pressure gradient along the oil filmis asymmetric relative to the center of contact.Because of this, a moment is created between thecontacting objects. This moment will cause the rollersto skew as shown in illustrated in Fig. 2. While slip ofapproximately 20% is occurring between the rollerlarge end face and cone flange, a moment of forcederiving from traction is also occurring. Furthermore, ifmetal-to-metal contact is present, the resultantfrictional force also contributes to the occurrence of amoment. Each roller turns such that balance isachieved among the pressure on the oil film, thetractional force from the oil film, the frictional force frommetal-to-metal contact, and the outer force (forceacting on the cone flange). The moments taking part in

Fig. 2 Skewing of roller

φ

r mr B

………………………………（12） ∫∫xpdxdyMh=

this rotational motion of rollers are described below,wherein the center of rotation is assumed to be at thecenter of gravity of the roller.

When referring to the coordinates system given inFig. 1, the moment of force Mh deriving from pressureon the oil film p can be calculated by the expressionbelow:

wherein, the origin is defined as the intersectionpoint of the x-axis and y-axis in Fig. 1 (b).

The magnitude of the moment caused by thetractional force can be determined in the followingmanner: from the experiment by Ono et al. 13), themaximum tractional coefficient μt, max is determined asfollows:

μt, max=25

0.09×

……（13）

where, α0 stands for the viscosity-pressure coefficient.Since the maximum projected skewing angle is assmall as approximately 0.05 degrees, the slip ratio s can be calculated with the following expression, ignoring minor variation by skewing:

α0 p whenα0 p≤25

μt, max=25

0.09whenα0 p＞25

-100-


Table 1 Analysis condition

Cup tapered angleRoller tapered angleR ratio of roller large end faceRoller lengthYoung's modulusPoisson ratioOil kinematic viscosity at 40˚COil kinematic viscosity at 100˚COil densityViscosity - pressure coefficientRotating speedRib loadOil temperatureRoot mean square roughnessof roller large end faceRoot mean square roughnessof cone back face rib

deg.deg.

mmGPa

mm2/smm2/skg/m3

Pa-1

min-1

N˚C

µm

µm

57.628.50.8

13.522080.332.25.45850

1.93～10-8

200020040

0.025

0.1

…

When a roller begins skewing, the contact pointbetween the roller large end face and cone flangeshifts, and the vector of the contact force no longerintersects the center of the roller. As a result, theflange generates moment Mr that helps reduceskewing. When the distance in the x-direction betweenthe roller flange contact point and center of the roller(i.e. the distance the contact point shifts), is taken aslshift, Mr can be defined as follows:

Mr= lshift P

rB－rm

rB（rm－lg）sinφ lshift =

Mh ＋ Mt ＋ Ma ＋ Mr = 0

…………………（21）

…………………（22）

…………………………………（20）

where, P is load on rib. Then, the geometrical relationship between lshift and the skewing angleφcanbe expressed as:

where, rm, rB and lg are lengths shown in Figs. 1 and 2. Finally, skewing angleφcan be determined by solvingthe expression below:

s（x, y）=

μt = μt, max

………………………（14）

………………（15）

……………………………（16）

……………………………………（17）

where, ur and ui represent the velocity at the roller large end face and the velocity at the cone flange facerespectively. If the slip ratio leading to the maximumtractional coefficient is taken as smax, then the tractionalcoefficient μt can be determined, by adopting thecircular-model of Lee-Hamrock 14), with the followingexpression:

Therefore, the tractional force Ft is defined as:

Then, when assuming that the distance between the ontact point to the center of gravity of the roller is lg, the tractional moment Mt can be expressed as:

Ft =μt ∬pdxdy

Mt = lg Ft

ur－ui

（ur＋ui）/2

s / smax

1＋（s / smax）2

Fa=μa A0 pa

Ma= lg Fa

………………………………（18）

…………………………………（19）

The frictional force resulting from metal-to-metal contact can be investigated using the Greenwood-Tripptheory 10). When assuming the boundary frictional coefficient as μa and the apparent contact area as A0, then the frictional force Fa on the asperity contact surface can be expressed as:

The moment Ma deriving from Fa is:

4. Calculation results

4.1 One detailed example of calculationsCalculations were performed under the analysis

conditions summarized in Table 1, and the results areshown in Fig. 3. The R ratio of roller large end facegiven in Table 1 can be defined as rm/rB, using theassociated codes in Fig. 1. The analysis conditions inTable 1 correspond to a situation where lubricating oil

Fig. 3 Result of EHL analysis between roller large endface and cone flange face

(a) Oil film pressure distribution

(b) Oil film thickness distribution

μm

-101-


0.07

0.060.05

0.04

0.03

0.02

0.01

0

-0.010 50 100 150 200

Ske

win

g an

gle

deg

Rib load

Exp.1000 min-1

Exp.2000 min-1

Exp.3000 min-1

EHL1000 min-1

EHL2000 min-1

EHL3000 min-1

HL1000 min-1

HL2000 min-1

HL3000 min-1

of viscosity grade ISO VG32 is kept at 40˚C, pure axialload of 9.8 kN is applied to the tapered roller bearing30306D, and the inner ring is allowed to run at 2,000min-1.

Fig. 3 (a) provides an oil film pressure distributiondiagram. Fig. 3 (b) shows an oil film thicknessdistribution diagram. The x and y -axis in thesediagrams are the same as those in Fig. 1 (b), and theregion enclosed with solid lines is the zone wherelubricating oil film formation can occur. The two facesgenerally shift in the positive direction along the x-axis.As shown in Fig. 3 (a), a pressure spike indicative ofEHL occurs at the location near the exit of the fluidflow. A skewing angle of -0.019˚ of the roller is present,an angle at which oil film tends to occur due to a wedgefilm effect. The point of maximum oil film pressure hasshifted by 0.2 mm in the positive direction along the x-axis due to skewing. As a result, the inlet oil flow areais greater than if the roller was not skewed. As shownin Fig. 3 (b), the deformed region constitutes an ovalhaving its long axis in the x-direction, which is the majorflow direction of the lubricating oil. There is ahorseshoe-shaped neck at the oil flow outlet. Unlike aone-way oil flow situation, the oil film profile is notsymmetric along the x-axis at y = 0.

4.2 Experimental verification of skewing angleHarada et al. measured the skewing angle on a

running 32310 tapered roller bearing 6). The resultantreport states that a skewing angle was calculated usingthe roller large end face/cone flange face contactprogram. This program simulates an isoviscosity-rigidbody region but there was a discrepancy in magnitude.Harada et al. think this is because in the experimentalstudy, a high viscosity-elastic body region was used incontrast to the calculation-based study, whichassumed an isoviscosity-rigid body region.

We compared the experimental result of Harada etal. with the calculation results obtained from ourmethod described in this paper. The results aregraphically plotted in Fig. 4. In this chart, “Exp.” are

Fig. 4 Comparison between experimental and calculatedof skewing angle

experimental results of Harada et al., “EHL” stands forcalculation results obtained from our EHL analysistechnique previously described in this paper, and “HL”indicates a results of fluid lubrication analysis on rollerlarge end flange performed by Harada et al. assumingan isoviscosity-rigid body region. Compared withvalues associated with “HL”, the values obtained from“EHL” are more similar to the experimental results.This is because with “HL” results, the force as anintegral of pressure is significantly shifted toward thefluid inlet path since elastic deformation is notconsidered. In contrast, with “EHL” results, the oil filmpressure is approximately the same as the Hertzpressure; thereby, the force (i.e. integral of pressure)acts on a point situated near the contact center.

The calculated effects of the rotational speedqualitatively differ from the experimental results. In thecalculated results, a greater rotating speed leads to agreater magnitude of the skewing angle for both “EHL”and “HL”. However, at lower rotational speeds (e.g.1,000 min-1), the experimental skewing angle closelymatches that obtained from EHL analysis. From thesefindings, we believe that at low rotational speeds,where starvation seems not to be present, the resultsof our EHL analysis technique well represents theexperimentally determined lubrication state.Furthermore, we reason that the experimentallyobtained skewing angle is smaller than that obtainedfrom the EHL technique because experimentally athigh pressure, the pressure at the inlet oil flow pathdrops due to starvation causing the moment derivedfrom the oil film pressure to drops.

4.3 Effects of surface roughnessBehavior of fluid flow is primarily governed by the

pressure gradient across and velocity between twofaces. Roughness asperities within the pressuregradient will result in localized pressure variation. Inaddition, fluid in the troughs of a rough surface willmove along with the surface: affecting the oil flow ratecaused by relative motion between the faces, resultingin a substantial decrease of the oil film thickness.Consequently, the expected flow will vary. By adoptinga pressure flow rate coefficient and a shear flow ratecoefficient in the Patir-Cheng mean flow model 4), wecan quantitatively determine the effects of 2D (two-dimensional) surface roughness.

The “pressure flow rate coefficient” can be definedas the ratio of the mean pressure flow on a roughsurface to that on smooth surface. The “shear flow ratecoefficient” is similar but with additional flow caused bya shift of the rough surface. Both the pressure flow ratecoefficient and the flow rate are affected by directionalparameters associated with roughness. For

convenience of discussion, we will deal with isotropicsurface roughness.

Taking the values in Table 1 to define surfaceroughness, we have determined changes in the oil filmthickness caused by the presence/absence ofroughness. Fig. 5 plots the resultant change of the oilfilm thickness.

Using the same conditions, the samples includingfactors for surface roughness have exhibitedsomewhat larger minimum oil film thickness.

In the roller large end-flange contact mode, the coneflange has higher surface roughness and surfacevelocity. As a result, it seems that the additional oil flowcaused by the shifting of the surface roughness of therib is greater than the inward oil flow restricted by thesurface roughness of the roller large end face. Also,the amount of oil flowing into the contact area is higher,leading to a thicker oil film.

-102-


5. Optimal end face R ratio resultingfrom oil film forming capability

So far, proposed factors triggering seizure ofcontacting objects include critical oil film thickness,temperature and frictional loss. Note that regardless ofthe underlying cause, metal-to-metal contact must beavoided to prevent seizure. Therefore, a maximumpossible oil film is preferred. In order to maximize theoil film thickness in a bearing, we have attempted tooptimize the roller end face R ratio. Because theoptimal roller end face R ratio appears to be governedby bearing operating conditions, we have calculatedthe oil film thickness using velocity, viscosity and loadas parameters.

Fig. 6 includes examples of plotting the relationshipbetween roller large end face R ratio and oil filmthickness, with several sets of operating parameters.These charts show the following trends:(1) Regardless of the combination of operating

parameters, there is a particular R ratio that leadsto a maximum oil film thickness. However, the oilfilm thickness can suddenly decrease when the Rratio is increased above the optimum value (thoughR ratios smaller than the optimum value do not leadto as drastic a decrease).

(2) The R ratio that leads to a minimal oil film thicknessappears to be least affected by rotating speed andviscosity (except for cases with a high viscositylubricating oil).

Now, let us discuss the relationships we havelearned so far concerning velocity, viscosity andoptimal R ratio with our tapered roller bearings.

5.1 Effects of velocityAs shown in Fig. 6 (a), at a rotational speed of 500

min-1, the minimum oil film thickness is largest when theR ratio is 93%; with a higher rotating speed, the optimalR ratio is somewhat smaller, standing at 88% at 4,000min-1. However, at the lower rotational speed rangewhere the oil film thickness is smaller, the variation inminimum oil film thickness near the optimal R ratio issmaller. Therefore, we think it is not necessary toseriously consider the effects of rotating speed foroptimal R ratio at low rotational speeds.

5.2 Effects of viscosityIn Fig. 6 (b), the minimum oil film thickness is largest

when the R ratio is 78% and the kinematic viscosity is335 mm2/s. Because the magnitude of the oil filmthickness is large in this case, the viscosity of thelubricating oil is not an important factor when pertainingto oil film formation and R ratio. On the other hand, atlower viscosities where the oil film thickness is small, the

Cen

ter

oil f

ilm th

ickn

ess

µm

1

0.8

0.6

0.4

0.2

00 500 1000 1500 2000

Smooth

Rotating speed min-1

Rough

Min

imum

oil

film

thic

knes

sµ

m

0.5

0.4

0.3

0.2

0.1

00 500 1000 1500 2000

Smooth


Rough

(a) Central oil film thickness

(b) Minimum oil film thickness

Bearing: 30306DRib load: 200 NKinematic viscosity of lubricating oil: 32.2 mm2/s

Fig. 5 Effect of surface roughness on oil film thickness

-103-


effect of the viscosity of the lubricating oil on the optimalR ratio is small.

5.3 Effects of loadFrom Fig. 6 (c), it is apparent that increasing the load

causes the optimal R ratio to be smaller. Furthermore,under higher loads, the minimum oil film thicknessresulting from an R ratio greater than optimal will besmaller. Therefore, when the oil film thickness is smallerand the rib load is higher, the R ratio needs to besmaller.

In summary, based on our results of the optimal Rratio and oil film formation, it is not necessary to giveparticular considerations to the effects of velocity and/orviscosity, assuming the operating conditions in ourstudy. It is reasonable to set the R ratio to approximately85% for a high load situation and to approximately 95%for a low load situation. At the same time, it is desirably

　B

earin

g te

mpe

ratu

re ˚

C

160

140

120

100

80

60

40

20

00 2000 60004000


R ratio 97%

R ratio 91.7%

R ratio 80%

Bearing: 30306DRadial load: 19.6 kNAxial load: 6.9 kNLubricating oil viscosity grade: ISO VG 56

Fig. 7 Results of measured bearing temperature: Effectsof R ratio of the roller large end face

Min

imum

oil

film

thic

knes

sµ

mM

inim

um o

il fil

m th

ickn

ess

µm

Min

imum

oil

film

thic

knes

sµ

m

0.5

0.6

1.5

1

0.5

0

2

0.4

0.3

0.2

0.1

0

0.5

0.4

0.3

0.2

0.1

0

70 80 90 100

70 80 90 100

70 80 90 100

R ratio of roller large end face %



4000 min-1

335 mm2/s

117 mm2/s

32 mm2/s

20N

100N

200N

350N

500N

11 mm2/s

5.5 mm2/s

3000 min-1

2000 min-1

1500 min-1

1000 min-1

500 min-1

(a) Effect of rotating speed Rib load: 200 N

Kinematic viscosity: 32 mm2/s

(b) Effect of oil kinematic viscosity Bearing speed: 2,000 min-1

Rib load: 200 N

(c) Effect of rib loadBearing speed: 2,000 min-1

Kinematic viscosity: 32 mm2/s

Fig. 6 Relationship between R ratio of roller largeend face and minimum oil film thickness

to design the R ratio smaller than the optimal value sincethe oil film thickness decreases dramatically when the Rratio exceeds the optimal value, but much less so whenthe R ratio is less than the optimum value.

5.4 Experimental verificationWe produced tapered roller bearing samples with

varying roller end face R ratios—97%, 91.7% and 80%.We analyzed the temperature rise of the outer ring withincreasing outer ring rotational speeds. Fig. 7 shows aplot of the measured bearing temperatures. Typically,flange faces are super-finished. However, the flangefaces in these samples were ground so that the effectsof lubrication would be more apparent. Excessive heatbuildup occurred in the samples with 97% R ratiosrunning at 2,000 min-1 and 91.7% R ratios running at3,000 min-1. It was determined that these samples weregenerating an excessive amount of heat, too high forrecommended operation. In contrast, samples with 80%R ratios proved to be able to run at 5,000 min-1 withoutexcessive heat generation. In the previous sections, wedemonstrated through analysis that oil film thickness isoptimized by selecting an R ratio of 85% or smaller.Through experimental verification, we proved theappropriateness of our analytical results.

6. Conclusion

Studying roller large end-flange face contact ontapered roller bearings, we performed isothermal EHLanalysis while considering roller skewing. Wediscovered the following characteristics aboutlubrication at the roller large end-flange face:• Assuming isotropic surface roughness, the magnitude

of oil film thickness will be greater when surfaceroughness is taken into account.

• If an EHL condition is assumed, the calculated valueof the roller skewing angle is smaller than thatobtained without considering deformation of thecontacting elements; it more closely matches theactual measured value.In addition, we discovered the following design

guidelines for end face R ratios to maximize thelubricant oil film thickness:

• The optimal end face R ratio that helps maximize thelubricating oil film thickness can vary depending onthe rotational speed, lubricating oil viscosity and loadconditions. Basically, it should be approximately 85%for high load situation. However, once the end face Rratio exceeds the optimal value, the oil film thicknesssuddenly decreases. Therefore, the design end faceR ratio should be smaller than the optimal value.

• Tapered roller bearings can experience operationalproblems when the oil film thickness is very low. If thelow oil film thickness is due to low speed or low oilviscosity, then the optimal end face R ratio has littleaffect.From these studies it is desirable to design the roller

with an end face R ratio of approximately 85% at most.This will allow the tapered roller bearing to operateproperly under heavy loads.

In preparing this technical paper, the authors have revisedTransactions of the Japan Society of Mechanical Engineers(Trans. Jpn. Soc. Mech. Eng.), Book C, Vol. 75 (2009), pp. 2319-2326.

-104-


Photo of authors

Hiroki FUJIWARA

Elemental TechnologicalR&D Center

Takashi TSUJIMOTO


Kazut0 YAMAUCHI

Graduate School of EngineeringOsaka Univercity

References1) T. Yamada, Torque Control and Operational Improvement of

Tapered Roller Bearings, SAE Technical Paper Series,(1986), 860726.

2) Y. Okamoto and T. Tsujimoto: ECO-Top Tapered RollerBearings, NTN Technical Review No. 68 (2000) 34-38.

3) X. Jiang et al., Thermal Non-Newtonian EHL Analysis of Rib-Roller End Contact in Tapered Roller Bearings, Transactionsof the ASME, Journal of Tribology, 117 (1995), 646-654.

4) N. Patir and H.S. Cheng, An Average Flow Model forDetermining Effects of Three-Dimensional Roughness onPartial Hydrodynamic Lubrication, Transactions of the ASME,Journal of Lubrication Technology, 100 (1978), 12-17.

5) H. Nishida et. al., Analysis of mixed lubrications between theLarge End Faces of Tapered Roller Bearings and Ribs,Proceedings of the Fall 2001 Tribology Conference inUtsunomiya, (2001), 277-278. (Japanese)

6) K. Harada and T. Sakaguchi, Rolling Element Skew inDynamic Analysis for Tapered roller Bearings, Proceedings ofThe Third Asia International Conference on Tribology, (2006),647-548.

7) L. Houpert, New Results of Traction Force Calculations inElastohydrodynamic Contacts, Trans. ASME, J. Tribol., 107-2(1985), 241-248.

8) C.S. Wu, E.E. Klaus and J.L. Duda, Development of a Methodfor the Prediction of Pressure-Viscosity Coefficients ofLubricating Oils Based on Free-Volume Theory, Trans. ASME,J. Tribol., 111-1 (1989), 121-128.

9) ASTM, Annual Book of ASTM Standards, Section 5, Vol.05.01Petroleum Products, Lubricants, and Fossil Fuels, (1987),197.

10) J.A. Greenwood and J.H. Tripp, The Contact of TwoNominally Flat Rough Surfaces, Proceedings of IMechE, 185(1970-71), 625-633.

11) N. Patir and H.S. Cheng, Application of Average Flow Modelto Lubrication Between Rough Sliding Surfaces, Trans.ASME, J. Lub. Tech., 101 (1979), 220-230.

12) N. Patir and H.S. Cheng, Effect of Surface RoughnessOrientation on the Central Film Thickness in E.H.D. Contacts,Proc. of 5TH Leeds-Lyon Symp. on Tribol., (1978), 15-21.

13) N. Ohno et al.: Effect of Bulk Modulus of Solidified Oils underHigh Pressure on Tractional Behavior, Journal of JapaneseSociety of Tribologists, Vol. 38 (1993), 927-934 (Japanese)

14) R.T. Lee and B.J. Hamrock, A Circular Non-Newtonian FluidModel: Part I - Used in Elasto-hydrodynamic Lubrication,Transactions of the ASME, Journal of Tribology, 112 (1990),486-496.


[ Discussion ]

Advances in and the Future of Analysis Techniques for Constant Velocity Joints

Tomogami: The three young engineers here todayhelped Dr. Kimata with the numerical analysis usedin the completion of his thesis. As they advanced thisnumerical analysis, they were able to study histheories, so this was a valuable experience for them.We are very grateful for this.

Currently, we are also working on applying thisanalysis to the development of constant-velocityjoints (hereafter, “CVJ”) in order to acceleratedevelopment. MSC. Adams* (hereafter, “Adams”),which is a commercially-available general-purposesoftware application for dynamic mechanismanalysis, is being used as one means for analysis. Iwould like to ask Dr. Kimata for his opinionsregarding the differences between his own “ball-typeconstant-velocity joint analysis method” and Adams.

(*Registered trademark of MSC. Software)

Differences between the “ball-type constant-velocity joints analysis method” and Adams

Kimata: The main contents of my thesis are related tothe static analysis and dynamic analysis of ball-typeCVJs. As equations of motion for dynamic analysis, Iused Euler’s equations of motion. Adams is general-purpose software for the dynamic analysis ofmechanisms, and it is supposed to use Lagrange’sequations of motion, so I think that theirfundamentals for dynamic analysis are the same.

I believe that in order to correctly interpret datafrom numerical analysis, it is necessary to haveknowledge about analytical theory. Even when usingAdams, I think that analytical theory related to CVJsis useful when interpreting the results.

Static analysis is very useful in understanding thefundamental characteristics of CVJ. I think that it isalso useful in understanding the results of dynamicanalysis.

Participants

Kei Kimata Former DirectorMasakazu Hirata (Intellectual Property Strategy Department)Yoshinori Terasaka (C.V.Joint Engineering Dept.)Hisaaki Kura (C.V.Joint Engineering Dept.)Shin Tomogami (C.V.Joint Engineering Dept.)Kisao Yamazaki (C.V.Joint Engineering Dept.)Teruaki Fujio (C.V.Joint Engineering Dept.)Masashi Funahashi (C.V.Joint Engineering Dept.)Kazuhiko Yoshida (New Product Development R&D Center)

-106-

Kei Kimata Former Director

Masakazu Hirata

Hirata: Former Director Kimata received a doctoral degree in engineering fromNagoya University in December 2008 with a thesis entitled “Research related toanalysis methods for ball-type constant-velocity joints.” Before Dr. Kimata retiredfrom our company in 1998, he had, for example, been the chief of our bearingtechnology research laboratory and the president of NTN Engineering PlasticsCorporation. He also served as a corporate director. Since retiring, he hascontinued his research, which culminated with his dissertation at the age of 72.Today, Dr. Kimata, who is one of our great predecessors, joins us in a discussionabout the “Advances in and the future of analysis techniques for constant velocityjoints” with some of our young engineers.


-107-

About the application of analysis technology

Tomogami: In other words, regarding the results thatcome from analysis using Adams, it is important tohave the sense to always verify whether or not theyare valid in consideration of analytical theory. In ourdevelopment—particularly in the development ofparts for automobiles—I believe the main approachhas been a method that involves the creation ofprototypes, their evaluation and the consideration ofthe results, or, in other words, trial and error. Iwonder, however, how our young colleagues feelabout this way of proceeding.

Yamazaki: Through analysis it is possible to evaluate theperformance of a product without creating aprototype, so I think this has the effect of reducingdevelopment time and cutting costs. On the otherhand, since there is uncertainty as to whether theresults are accurate and whether phenomena havebeen suitably simulated, I think it is necessary toproceed by using both analysis and experimentationskillfully.

Fujio: In the process of development, we conductanalysis, of course. In particular, when we aredesigning a new CVJ, for example, to investigatehow the internal contact load differs in comparisonwith an existing product, we analyze using Adams.Analysis is convenient because it allows conditions tobe changed easily. Adams is also useful because itallows the visualization of the internal force vectorsfrom calculation results.

I believe that even though development should beadvanced with analysis in the lead, at present we stillcannot say that we are able to accurately simulateevery phenomenon, so in the future I want toinvestigate new methods.

Funahashi: I think that analysis using Adams is valuableduring the initial investigation of new structures.However, the fact is that there are many cases wherenumerical analysis is difficult when plasticdeformation and destruction occur. The pursuit ofboth mechanism analysis and finite element analysisis necessary. Particularly in miniaturization, the ratioof plastic deformation volume increases makinganalysis complex. I want to apply the power of youthto resolving this.

Terasaka: That is a reassuring statement (laughs).

Improving analysis technologies for the future

Kimata: For the improvement of analysis technologies,efforts in elastic-plastic analysis are necessary. Areefforts in the engineering division advancing?

Fujio: I think that, for CVJ development in particular,analysis must be undertaken related to the failure ofCVJs caused by plastic deformation where inner andouter rims contact balls. I want to acceleratedevelopment by applying dynamic finite elementanalysis in the analysis of these parts.

Hirata: At the Elemental Technology Research &Development Center, as one fundamentaltechnology, we are also striving to develop coupledanalysis that combines the dynamic analysis ofmechanisms comprised of multiple parts anddeformation analysis of those comprising parts.

Kimata: Being able to process mechanism analysisbased on dynamics and elastic-plastic finite elementanalysis in one program would be good. However, atthis point realizing this is going to be very difficult.High-speed computers with great processingcapacity are necessary for dynamic analysis.

From the left, Yoshida, Tomogami, Terasaka, Fujio, Funahashi, Yamazaki and Kura

-108-


Seeking the strength conditions with dynamicanalysis and using these as input conditions toconduct elastic-plastic analyses with a finite elementsapproach and then applying this procedurerepeatedly might be the only way.

Funahashi: I also feel this way. I think that at presentconducting analyses at each small step is probablythe most suitable.

Yoshida: In CVJ testing, there is a semi-torsional test inwhich it is twisted and fractured in low-speed fixed-revolution conditions. In order to clarify thisdestruction mechanism, we are working on dynamicelastic-plastic finite element analysis. This model isat a level that allows the semi-torsional fracturemechanism to be simulated qualitatively. Currentlywe are striving to improve this model to allowquantitative simulation.

Kimata: I hope that you will continue your efforts for CVJelastic-plastic analysis as a manufacturer thatspecializes in CVJs.

How to advance efficient development

Tomogami: If we want to further increase the speed ofdevelopment, I believe that we must apply the latesttechnologies for analysis. Could our young engineerswho are advancing our development work receivesome advice from you, Dr. Kimata?

Kura: We also discussed this earlier, but wheninterpreting data, I wonder if the way to advance tothe next step will differ depending on the approach tothe interpretation of data from dynamic analysis andexperimentation.

Kimata: Well, accurately evaluating the data isnecessary. In addition, I think that grasping themechanisms that produce that data is important. Agraph of the results of dynamic analysis iscomplicated, but the results of static analysis aretransformed by the impact of the friction generatedby rotational movement. Looking at the data fromthat analysis and the dynamic analysis together cansometimes make it easier to understand. Byconsidering and investigating how the frictionworked, I have had the experience of being able toeasily interpret a dynamic analysis graph that hadappeared complicated at first glance.

In analysis, it is easier to understand if we start bythinking about a model that has been simplified asmuch as possible. In some senses, I think that inanalysis, while considering the necessary conditions,the crucial point is how much it can be simplified.

The future of constant-velocity joints

Tomogami: Now I would like to change the topic anddiscuss future technologies for constant-velocityjoints. As you all know, environmental considerationshave become an extremely important theme even forautomobiles. As parts for the driving force ofautomobiles, CVJs are said to be mature products. Inorder to meet the needs of customers, however, weare still working on issues that include making themlighter and more compact and further increasing theirtorque transfer performance.

Terasaka: Make them lighter, make them smaller, andmake them thinner. In every case, we must considerrigidity and deformation, but I think that it is difficult toinvestigate thoroughly enough to be able to identifythe limits. Do you have any suggestions for how toapproach this?

Kimata: I think that in design it is important that you focuson reducing the concentration of stress. I think theapproach of increasing the number of balls, whichNTN is advancing, is also one solution (E SeriesCVJs).

By making the balls smaller in size and bearing theforce with numerous balls, the stress at the point ofload can be decreased. By reducing theconcentration of stress, the possibility of usingmaterials other than steel also arises. Incidentally,have you been able to utilize resin materials inconstant-velocity joints yet?

Yoshida: In propeller shaft assemblies, by using CFRPshafts, weight has been reduced and the use of twojoints has been made possible where three joints hadbeen used before.

Yamazaki: In the case of drive shafts, since the axialdiameter is fine and the input torque is greatcompared to propeller shafts, contact surfacepressure and shear stress are higher. At present, wehave not yet utilized resin materials in high-strengthparts such as drive shafts, but we are investigatingthe possibility of application.

Hirata: Among non-ferrous materials, ceramic also has ahigh modulus of elasticity and high strength, so weare considering its application in strong parts.

Tomogami: It is a light and appealing material, butpliability, processing characteristics and cost areissues at this time.

Kimata: For these reasons, I think that the age of steelcontinues…

-109-

Hirata: Considering the long-term, I believe that thereprobably is a need to work with new materialsbesides steel, and the current state is that we areadvancing efforts from the standpoint of elementaltechnologies. Looking to the future, I am convincedthat we must work with these new materials.

Kura: Have you ever investigated sliding type constant-velocity joints, for example, which are completelydifferent from ball-type CVJs?

Kimata: I have investigated TJ-type CVJs, but it did notgo well in the end. With two-way sliding, I think that itis probably difficult to achieve an oil film. Frictionincreases obviously. For bearing large loads, with theexception of parts like cages that must be madesliding, rolling or rolling-sliding is probably better. Thereason for this is that the lubrication characteristicsare better.

Hirata: In the application of sliding for CVJs, I haveinvestigated surface modifications and greases, but,as you say, with rolling-sliding, the grease isreplenished well, while with sliding only, the grease isexpelled, reducing performance in my experience.

Kura: So we should think about new joints as we userolling-sliding skillfully.

Kimata: It would be good if there was a sliding materialwith low friction and high load capacity, but…

To further increase performance

Tomogami: Please give us some advice about how tofurther improve the performance of CVJs.

Kimata: It might be obvious, but I think it is minimizingthe coefficient of friction of the contact surface asmuch as possible. I think that inside CVJs, thecoefficient of friction is not fixed, but rather dependson the relative movement (sliding or rolling-sliding)conditions of the contact parts. They changedepending on the conditions of the contact surfacesand the grease.

For this reason, I think that surface modification isa good method.

Yamazaki: Recently, HL treatment of balls for CVJs thatwe have developed is an example of this. Bycombining HL treatment of balls for CVJs with greasethat has been matched to this, we are able toachieve a great reduction in torque loss.

Encouragement for young engineers

Terasaka: We have enjoyed discussing a variety oftopics, but we have come to the end of our time.During your many years of research anddevelopment work, including analysis related toconstant-velocity joints, I expect that you have alsorun into a few walls. One of the things you said wasto start by simplifying things to think about them. Attimes when you have a problem and think that thingsare difficult, are there any other kinds of things thatform the basis for your thinking?

Kimata: I believe that it is returning to the fundamentalsto think about the problem. By practicing returning tothe fundamentals, I believe that you will gain avariety of cognitive abilities.

Terasaka: For example, when test results disagree with atheory that you have devised, what is a good way toproceed?

Kimata: You should probably check and confirmeverything thoroughly. Was there a mistake in thetest procedures? You have to return to thefundamentals and reconsider even when conductinganalysis. The greatest danger is becomingoverconfident and believing that there are nomistakes in what you have done. I have also hadmany failures, but the important thing has alwaysbeen returning to the basics to think and reviewing toconfirm that there are no contradictions in the theory.

Hirata: Today you have given us the opportunity toparticipate in this valuable and lengthy discussion.Thank you very much.

Everyone: Thank you very much.


-110-

High Capacity Tapered Roller Bearing

1. Introduction

NTN “High Capacity Tapered Roller Bearing” won anEncouragement Award in “CHO MONOZUKURI InnovativeParts and Components Award 2009” sponsored by TheConference for the Promotion of MONOZUKURI andNikkan Kogyo Shimbun Ltd., Japan’s leading business andindustry newspaper publisher (and backed up by Ministryof Economy, Trade and Industry in Japan, and The JapanChamber of Commerce and Industry).

With bearing size unchanged, this novel high capacitytapered roller bearing design boasts a 10% increase atmaximum basic dynamic load rating (37% increase atmaximum calculated bearing life), 15% increase atmaximum basic static load rating (15% increase in safetyfactor) and 10% increase at maximum bearing rigidity(10% reduction at maximum elastic displacement). Thisnovel tapered roller bearing also features lower contactpressure and has exhibited longer life even in a difficultlubricating environment where the bearing is lubricatedwith a low viscosity lubricating oil 1).

2. Structure of NTN high capacitytapered roller bearing

The NTN high capacity tapered roller bearing featuresa unique structure which include: a smaller clearancebetween the cage and outer ring combined with a largerpitch circle diameter with the cage, helps the cage to besituated nearer to the outer ring. Therefore, the roller-to-roller clearance has been reduced to maximize thenumber of rollers (a number same as that on a fullcomplement roller bearing of a same size) while the cagebar width dimension remains same as that on the NTNstandard tapered roller bearing (see Figs. 1 and 2).

3. Conclusion

Boasting longer life and higher rigidity, the NTN highcapacity tapered roller bearings will help solve challengesoccurring from needs for much improved fuel economywith automobiles. By additionally adopting NTN’spropriety FA heat treatment technique to help designunique tapered roller bearings of longer life and lowertorque, the new NTN high capacity tapered roller

Takashi UENO

Fig. 1 Structure of high capacity tapered roller bearing

Fig. 2 Figure of high capacity tapered roller bearing

[Left: Standard tapered roller bearing Right: High capacity tapered roller bearing]

Takashi UENOAutomotive Engineering Dept.


Clearance between outer ring and cage

Smaller clearance betweenouter ring and cage

Cage bar width

Cage bar width(Same dimension)


Smaller clearancebetween rollers



Photo of author

NTN TECHNICAL REVIEW No. 77（2009）

Presentation of award-winning articles

“2009 ‘CHO’ MONOZUKURI Innovative Components Awards”Encouragement Award

bearings will feature much reduced size, weight, andrunning torque. Thus, NTN is going to more positivelymarket this line of bearings as products that willcontribute to better fuel economy on automobiles.

References1) T. Tsujimoto and J. Mochizuki: High Capacity Tapered

Roller Bearings, NTN Technical Review No. 73 (2005)20-29

-111-

1. IntroductionNTN “Constant Velocity Steering Joint (CSJ)” won an

Automotive Component Award in “CHO MONOZUKURIInnovative Parts and Components Award 2008” sponsoredby The Conference for the Promotion of MONOZUKURIand Nikkan Kogyo Shimbun Ltd., Japan’s leading businessand industry newspaper publisher (and backed up byMinistry of Economy, Trade and Industry in Japan, andThe Japan Chamber of Commerce and Industry).

The CSJ satisfies otherwise conflicting requirementsfor “smooth rotation” and “reduced play on steeringwheel” through provision of a special preloadingmechanism within the joint to eliminate backlash on thejoint and optimization of internal design.

Compared with the double Cardan joint, which is aconventional constant velocity steering joint (hereinafterreferred to as “D-CJ”), NTN CSJ are about half the weightand volume which can run at a constant speed in theoperating range (0-48 C˚) without developing backlash.Thus, the NTN CJS helps car steering system designersflexibly design the steering system layout.

2. Structure of constant velocitysteering joint CSJ

The basic structure of CSJ 1) comprises as a basis, theBJ, (ball type constant velocity joint) that is frequentlyused for driveshaft a plunger, and a spherical plateincorporated into the BJ (see Fig. 1). The spring force ofthe plunger causes the inner ring to shift rightward in thisdiagram; thereby the balls always remain in contact withthe surfaces of ball grooves on the inner and outer rings;as a result, clearances between related parts areeliminated in order to inhibit rotational backlash.

3. Features of constant velocitysteering joint CSJ

Results of comparison of performance of the NTN CSJwith conventional constant velocity steering joint (D-CJ)are summarized below 2):1) Result of comparison in terms of rotational backlash

As shown in Fig. 2, compared with the D-CJ, the CSJboast smaller hysteresis and higher rigidity.2) Result of comparison in terms of rotational torque

Fig. 3 graphically plots the relation between rotationalangle and rotational torque. There is no significantvariation in the rotational torque of CSJ. In other words,the NTN CSJ has characteristics needed for anautomotive steering joint.

Fig. 1 Structure to reduce rotational play

Fig. 2 Rotational bach lash Fig. 3 Rotational torque

Photo of author

Kenta YAMAZAKI

C.V. Joint Engineering Dept.Automotive Sales Headquarters

CSJD-CJ

+

+

0

0- -

Tor

que

load

ing

Torsional angle

Higher torsionalrigidity

Low torsional rigidity,and high hysteresis

CSJD-CJ

+

00 60 120 180 240 300 360

Rot

atio

nal t

orqu

e

Rotational angle

Spherical plate

Axial displacement

Plunger

Ball

Outer ring

Inner ring

Cage

4. ConclusionBeing the world’s first ball type constant velocity steering

joint product, the NTN CSJ boasts not only greatfunctionality but also significant reduction in both weightand size—approximately 50% reduction compared with theNTN’s conventional D-CJ. This novel constant velocitysteering joint will positively contribute to mitigation ofenvironmental impacts possibly caused by automobiles.

References1) K. Yamazaki: Constant Velocity Steering Joint (CSJ),

NTN Technical Review No. 73 (2005) 84-87

Constant Velocity Steering Joint (CSJ)Development

Kenta YAMAZAKI



“2008 ‘CHO’ MONOZUKURI Innovative Components Awards”Automotive Component Award

-112-

Development of a Briquetting Machine and theConstruction of a Recycling System

for Steelmaking Dust

Overview

In Japan, approximately 500,000 tons of steelmakingdust annually occurs in steelmaking electric furnaceplants. This type of “waste” has been traditionallydisposed of by landfill, dezincification or reuse.

Now, disposal by landfill poses problems—increasinglylimited availability of landfill sites and mounting costs.Incidentally, costs for dezincification are high, includingexpenses needed for transportation. Therefore, it isnecessary to reduce the amounts of steelmaking dust tobe disposed of.

Various recycling techniques for steelmaking dust havebeen proposed. However, to be able to improve meltingefficiency and cost-effectiveness, NTN has chosenDAIWA STEEL CORPORATION as a partner.Consequently, the team has successfully developed aunique steelmaking dust recycling method that realizesboth mitigation in environmental impacts and reduction incosts.

In the newly developed recycling system, which isschematically illustrated in Fig. 1 below, carbon andwater are added to steelmaking dust, and then the

mixture is pelletized. The resultant pellets are formed intocylindrical briquettes by pressure, without using glue, andthen the resultant briquettes are loaded into an electricfurnace together with steel scraps.

Because the carbon content included in the briquettesserves as a fuel/reducing agent, NTN’s new recyclingsystem boasts higher degree of melting efficiency. Inaddition, the density of steel content in the briquettesobtained from the NTN’s novel steelmaking dust recyclingsystem is high compared with the previous methods.thus, more iron content will be melted in the molten bathand recovered as an iron. Note that with previoussteelmaking dust disposal methods, steelmaking dusteither unmodified or pelletized was fed into an electricsteelmaking furnace.

This new development has already resulted inimproved economy, total elimination in amount ofdisposal by landfill and improved electricity use in theNTN Group. The resultant recycling ratio of steelmakingdust (increase in recovered iron material) has helpeddecrease the amount of steelmaking dust beingsubjected to dezincification. Thereby, NTN has beencontributing to mitigation in environmental impacts

resulting from its manufacturingactivities.

While environmental conservationposes challenges to manufacturers,this novel steelmaking dustbriquetting system has beenpositively contributing to realizationof the recycling-oriented society.

Shouzo GOTO* Ikuo YAMADA** Katsutoshi MURAMATSU***

“2008 Resource Recycling Technology and Systems”Encouragement Award

Fig. 1 Briquetting Machine and Recycling System for Steelmaking Dust

Photo of author (Representative)

Shouzo GOTOH

Fomer Production Engineering.R&D Center

*Fomer Production Engineering R&D Center **Production Engineering R&D Center ***Elemental Technological R&D Center



(4) NTN steelmaking dust briquetting machine

(1) Electric furnace

(3) PelletizerPellets

Briquette(φ150×120H)

Steelmaking dust

(2) Filter

(5) Dezincification plant

Metal zinc

Iron component

Carbon

Scrap

Recycling of steel products

Recycling of steelmaking dustoccurring from electric furnace

Steel products

Dust Exhaust gas

Water

-114-

Features

Construction

In-wheel motorElectromechanical brake

Hub bearing

Load sensor

Sensor informationSteering angle

Vehiclecontroller

Accelerator opening

Motor control

Brake control

Steering angle control

Suspension control

In-wheel motor

Electromechanicalbrake

Load sensor

Wheel speed (motor speed)

Driving power (motor current)

Braking power

Cornering force

Ground load


[ New Products Information ]

Intelligent In-wheel Unit

Integration of electric motor, brake and sensor as well as size reductionin in-wheel unit contribute to optimal energy regeneration and vehicle

stability control, and improved driving performance.

(1) Improvement in both fuel economy anddriving safety

(2) Much improved freedom in designing vehicleitself and automotive components

(3) Car maneuverability can be tailored accordingto users’ needs.

Applications

¡Electric vehicles, fuel cell vehicles

-115-

Wide-Range High-Resolution Small SensorIntegrated Hub Bearing

Two sets of control signals are available to contribute to improvementin precision and reliability in vehicle safety control.

Features Construction

Magnetic encoder High-resolution sensor

Wheel hub bearings for passenger cars¡Car attitude control¡Hill hold control

Applications

(1) The built-in sensor covers the entirespeed range from extremely low to high-speed regions: will contribute tomitigation of load in signal processing.

(2) Running speed detection at higherresolution (40 times as accuratecompared with a previous detectionsystem): now, the motion of wheel isdetected to accuracy in steps of mm.

(3) Now, judgment on whether the wheel isrunning forward or reverse is possible:also, fall motion in hill launch mode ispossible.

(4) Use of a compact sensor contributes tosize reduction of hub bearing product.

High-resolution signal (40 times as high comparedwith conventional signal) and ABS signal

Magnetic poles onmagnetic encoder

ABS signal for high-speed travel

High-resolution signalfor low-speed travel

New Products Information


-116-

Positively assists in fuel efficiency of cars thanks to 30% reduction in torque transmission loss!Compact, light-weight constant velocity joint products capable

of higher operating angle (50 degrees).

High Efficiency Fixed TypeConstant Velocity Joint (EUJ-S)

- Advantages over conventionalproducts (EUJ) -

¡Design of internal structure has beenoptimized to reduce friction loss thatresults from contact between internalcomponents in joint.

¡Use of newly developed low-µ grease andadoption of surface modificationtechnique for metal components helpdecrease friction loss.

(1) Torque transmission loss: Reduced byapprox. 30%

(2) Strength/Durability: Same level(3) Max. working angle: Same level (50˚)

Applications

¡Automotive drive shaft

Construction

BallCage Inner ringOuter ring

Sliding type constant velocity joint

Fixed Type Constant Velocity Joint(EUJ-S)

Tire

Features (comparison with conventional products)

-117-

High Efficiency and Low Heat Generation Ten BallConstant Velocity Joint (HELJ) for Propeller Shafts

¡ Automotive drive shaft(Propeller shafts on FR and 4WD vehicles)

Realization of reduced torque loss, limited heat generation and light weight!Contributes to higher functionality and improved fuel economy!

Features (comparison with conventional products)

Applications

Construction

Comparison usingcross-sectional view

Ball CageInnerring

Outerring

Conventionalproduct (HLJ)Number of balls: 6

Ball slots inclined inopposite directions

New product (HELJ)Number of balls: 10

Propeller shafts

(1) High efficiency: Torque loss reduced by50%

(2) Low heat generation: Temperature risereduced by 50%

(3) Light weight: Mass reduced by 5%

- Advantages over conventional products (HLJ) -

¡Optimally angled ball slots helps reduce frictionalresistance occurring on balls in axially shifting motion;consequently, heat generation on balls is mitigated.

¡ Smaller diameter balls are adopted: number of balls hasbeen increased from six to ten: the size and weight ofinner ring, outer ring and cage have been muchreduced.



-118-

Long Life Ball Screw for Actuators

Longer life in spite of compact, light-weight design. Contributes to realization ofelectromechanical design that contributes to better fuel efficiency.

Features (comparison with conventional products) Construction

[New product]

Lead < ball size

[Conventional product]

Lead > ball sizeLead Lead

Ball screw

Actuator section

Locker arm

LinkLink

Driveshaft

Output cam

Valve lifter

Valve

Control shaft

¡Automotive continuous variable valvemechanism and electromechanical actuator

Applications

(1) Life: Improved by 6 times(compared to the same size product)

(2) Size: Down sized axially by 7%(compared to products with similar life)

(3) Mass: Reduced by 30%(compared to products with similar life)

¡Balls whose size is greater than that of leadMaximized ball size helps increase loadcapacity of ball-screw.

New product

Conventional product

-119-

Much reduced torque contributes to improvement in productivityand alleviation in environmental impact!

Ultra Low Torque Back-up Roll Unitfor Tension Levelers

Features (comparison with conventional products) Construction

Novel seal structure (spacer and O-ring)

Low torque needle roller bearing

¡Steel sheet production equipment, Tensionlevelers

Applications

(1) Rotational torque: Reduced by 40%(2) Temperature rise: Reduced by 20%

Back-up roll unit

Intermediate roll unit

Work roll unit

Steel sheet

Wind up



-120-

Two Way Sorting Feeder

Parts feeding mechanism derived from new concept contributesto improvement in productivity!

Construction

Parts feedingmechanism

Forward direction: Feed direction with bawl feeder

Reverse direction: Feed direction with parts feeding mechanism

Increasing speed/Decreasing speed

Sloped chute

Features

¡Aligning and feeding of automotive parts,parts for electric equipment, and parts formedical equipment, etc.

Applications

(1) Feeding in two opposing directionsA well conceived mounting system for apair of leaf springs can realize partsfeeding in two opposing directions.

(2) Increasing speed/Decreasing speedSpeed can be varied for the attachmentsection that aligns and feeds parts.

(3) Transfer of vibration to the sloped chuteCongestion of parts within the chute isprevented.

-121-

New-Type One-Way Clutch for Office Equipment(NCT)

This one-way clutch can be readily built into a resin-built component, whilemaintaining highly reliable clutching function!

Features

¡Paper feeding section in printers andphotocopiers

Applications

Construction

Thermal caulking(5 pos.)

Outer ring flange

Detent

Outer ring Cage with built-in cover

7

φ12

φ6

Paper output tray

Applicable areas

Paper feeding tray

(1) Can be integrated with a resin-builtcomponent.(Detents are situated on the outercircumference of outer ring.)

(2) Improved handling ease• Cage cover is thermally caulked to the

outer ring to form a single component(non-separable type).

• Orientation for assembly can beidentified by color-coding (lockdirection is clearly indicated on theproduct).

• Lock direction can be altered by simplyreversing the orientation for installingthe clutch.

(3) Higher reliability and durability• Equivalent with the unit type (NCZ)

one-way clutch(4) Capability for bearing radial load

• The cage bore surface on the clutchbears radial load.


Special Issue; Efforts for the Environment December 2009...¡Advance zero emissions: seek a...

Documents

Transcript of Special Issue; Efforts for the Environment December 2009...¡Advance zero emissions: seek a...