10FTM02

AGMA Technical Paper

Improving HeatTreating Flexibility forWind Turbine GearSystems ThroughCarburizing, Quenchingand Material HandlingAlternativesBy W. Titus, AFC--Holcroft

Improving Heat Treating Flexibility for Wind Turbine GearSystems Through Carburizing, Quenching and MaterialHandling Alternatives

Wallace (Jack) Titus, AFC--Holcroft

[The statements and opinions contained herein are those of the author and should not be construed as anofficial action or opinion of the American Gear Manufacturers Association.]

AbstractPart handling and processes for heat treating large gears have created challenges for decades. Growth inwind energy technology has focused more attention on this issue in recent years. The vast majority ofinstallations processing such large parts utilize conventional methods via pit furnace systems. Suchequipment has inherent limitations with respect to quench flow and part handling, making true improvementsin areas such as distortion control difficult due to physical limitations of this processing approach. Thispresentation will explain alternative methods for heat treating large components that allow part distortion to beminimized. Benefits will be quantified regarding cost savings to produce such gearing and quality.

Copyright 2010

American Gear Manufacturers Association500 Montgomery Street, Suite 350Alexandria, Virginia, 22314

October 2010

ISBN: 978--1--55589--977--6

3

Improving Heat Treating Flexibility for Wind Turbine Gear Systems ThroughCarburizing, Quenching and Material Handling Alternatives

Wallace (Jack) Titus, AFC--Holcroft

Introduction

Commercial size wind turbines shown in Figure 1and their gearboxes Figure 2, are designed tosurvive in extreme environmental conditions, mostnotably high wind forces, desert heat and arcticcold. However robust the design and use of materi-als, the typical gearbox struggles to meet theirdesign life. Studies by the National RenewableEnergy Laboratory (NREL) [1] have shown that onaverage a gearbox’s time to repair can be as shortas five years give or take when the economic pay-back model requires 20 years. Many failures seemto occur first in the bearings, also outlined in theNREL report. Studies are continuing to identify theroot causes for these and other failures. Otherpreliminary studies indicate that the gearbox and orplanetary carrier systems are subjected to high-er-than-anticipated stresses and that the materials

used may not have the strength to resist thestresses encountered. Contaminated lubricationhas also been targeted as a potential problem area.Since ferrous alloys are themost economical choicefor drive train components, heat treating is a majorconsideration when designing for material strengthand fatigue resistance.

Historically pit furnaces and the pit quench havebeen employed to case harden these and otherlarge gears because for treaters there just has beenno other choice. They are used primarily becausethey exist and alternative equipment has not, andthere is the comfort factor – we’ve always done itthat way. There are some who will say that pitfurnaces offer a higher quality product but that’s notso, as data presented later will show. In addition, pitfurnaces can be energy hogs especially whencarburizing cycles of 35 hours and longer arecommon.

Figure 1. Commercial size wind turbine

4

Figure 2. Heat treated parts of a typical wind turbine gearbox

Easy site preparation

Common sense suggests that eliminating the needfor a pit is a very desirable concept, offering morefreedom for heat treat site selection. Often the deepexcavations needed for pit furnaces and quenchtanks pose environmental issues in brown as wellas green field sites. Further, if expanding an exist-ing facility is contemplated, what facilities engineerhasn’t experienced a flooded pit or power outagetrying to install a pit? Who wants to accommodatean extra-high bay and additional large crane?Lower installation costs are achieved by reducingthe overall furnace equipment height and using apitless quench tank. Emerging economic realitiesfor commercial as well as captive heat treatersdictate that the need for a more flexible and cost ef-fective method of carburizing and quenching isoverdue for large gears. Wind energy has attractedthe attention of commercial heat treaters but theydon’t want to put all of their eggs in one basket, like apit furnace; they need equipment flexibility if thewind turbine market fails to live up to expectations.

Cost effective material

Case hardening is necessary to achieve the toothstrength and wear resistance in ferrous alloys andthe process allowing the most economical use ofmaterials is carburizing. Adding carbon to the wearsurfaces of bearings and gear teeth can be morecost effective than induction hardening where theentire part contains the alloy, even areas that playno role in the stressed application. Both processesrequire quenching to achieve the hardnessnecessary and both result in distortion that must be

corrected by grinding to bring the part to operatingdimension.

Anyone associated with manufacturing and heattreating precision drive systems knows grinding is anecessary process required to obtain final size andsurface finish after quenching. In addition, grindingremedies minor machining errors resulting frommanufacturing very large gears. Grinding, althougha critical requirement, is very expensive and can beminimized if quench distortion can be controlled andheld to a minimum. If distortion can be predictedthen designers can plan for it, thereby providingonlythe amount of material to be removed. Helical bevelgears are especially troublesome because of theirtendency to “unwind”. The unwinding if predictablecan be accommodated in the design andmachining.

Quenching options

Obviously quenching is a very difficult process tocontrol and one that can ruin a perfectly successfulcarburizing process. Only recently with thedevelopment of sophisticated computerized fluiddynamics (CFD) has liquid quenching, specificallyoil quenching, been targeted for investigation;maybe that’s because most heat treaters wish itwould just go away – no chance. Since heat treatingferrous material began eons ago, oil has no matchfor the range and control of quenching response.No other medium works better than oil when cor-rectly applied knowing its limitations.

S Polymer has been tried and in some applicationsit can provide acceptable results in smaller

5

gears but concentration maintenance is an issueand it too has disadvantages, creating a need formore investigation into controlling distortion.

S Salt is a great quenching medium but has foundalmost no application with large gears. Historic-ally, however, salt has been used extensively toreduce distortion in smaller gears.

S Helium high pressure quenching is very good atcontrolling distortion but cannot match thequench severity of liquid that’s required for verymassive parts.

S Press quenching with oil is the most acceptedand practiced method for distortion control withgears and bearing cups & cones, but to date ithas seen little if any application for smaller windturbine gears.

S IntensiQuench is a patented high intensitywater quench method designed to impartcompressive stresses onto the part surface inaddition to allowing the use of lower alloy steels.Proprietary software computes the quench timeto form martensite and the time required for thecore heat to temper the martensitic case whenthe part is raised from the water thenre-immersed into the water for final quenching.

Although diverse in their markets and application,all of the above quenching options can be imple-mented by AFC-Holcroft’s modular heat treatingsystem: EZ-Lynks.

Unlike smaller automotive drive trains, large gearcomponents by necessity employ very highhardenability steels due to their size. However, highhardenability is a two-edged sword, easier toharden but at greater risk of distortion in quenching.Again, a better solution would allow several quench-ing techniques as documented above. Open pit oilquenching and the required fixturing offer little in theway of flexibility for quenching options. Many gearscan be press quenched or otherwise individuallyquenched but not from pits where the awkwardhanging fixture makes it impossible to removeindividual parts.

Large is a relative term when applied to gears andfor this paper is defined as 24” to 36” diameterplanetary gears and up to 65” to 84” diameter spurgears. Quenching sun gears, shafts, pinions andplanetary gears by their vertical orientation can becontrolled somewhat easier but still requiresattention to flow uniformity and velocity. In manyapplications, fairly large gears can be press

quenched with oil or water but only if the furnacetype and material handling can provide the accessfor removal.

The IQ furnace has been the preferred furnace forcommercial and captive heat treaters alike, primar-ily due to the flexibility offered in process, load sizeand efficiency. With electric as well as gas firedradiant tubes, the IQ furnace has providedunmatched utility and quality for all industrial,aerospace and military markets worldwide.AFC-Holcroft’s UBQ series of IQ furnaces are thepreferred product for many of the world’s largestdrive train suppliers. Today AFC-Holcroft’s batchfurnace as shown in Figure 3 is carburizing andquenching planet gears used by a major windturbine manufacturer. Flame Metals ProcessingCorp. in Rogers, MN is a certified heat treat suppliermeeting all of the AGMA and wind turbine qualityrequirements. Even though AFC-Holcroft has soldIQ furnace designed for very large loads, up to 72”(1829 mm) square and 12,000 lbs (5454 kg), theyhave not been applied to large gears probably be-cause of rear handler (RH) capacity restrictions andthe quench tank design. EZ-Lynks, AFC-Holcroft’salternative heat treating system shown inFigure 4a,Figure 4b and Figure 4c eliminates the pit disad-vantage and overcomes the material handlingconcerns by separating the material handling fromthe furnace proper allowing more flexibility in the hotzone and quench tank designs.

Figure 3. UBQ integral quench batch furnace

6

Figure 4a. EZ--Lynks modular heat treatingsystem

Figure 4b. EZ--Lynks modular hot zone

Figure 4c. EZ--Lynks modular system can belocated anywhere and easily expanded as

needed

The EZ-Lynks cell illustrated in Figure 5a andFigure 5b consists of multiple hot zone modules,enclosed floor mounted oil quench tank, wash, tem-per and transfer shuttle. All equipment is rated to16,000 lbs (7,272 kg). A manufacturer or heat treat-er with an existing facility does not have to invest inapit or extra high bay to accommodate EZ-Lynks.Doors can be designed to open sideways, radianttubes can be removed from the rear and quenchtanks can rest on the floor. Process flexibility is theresult of a traversing and rotating transfer car,Figure 6, capable transferring loads automaticallythroughout the cell according to a preconfiguredrecipe.

Figure 5a. EZ--Lynks modular heat treat floor plan

7

Figure 5b. EZ--Lynks modular heat treat elevation

Figure 6. EZ--Lynks rotating transfer shuttle

The Achilles heel of pit quenching is the huge flamecreated when the load is immersed in the oil.Flames rage to engulf the crane, cables and truckand risk igniting the soot-covered ceiling. Figure 7and Figure 8 illustrates the typical pit quenching ofgears. Who hasn’t been tempted to run for the exitswhen pit quenching a massive load in oil? A morecommon sense solution like that provided by EZ--Lynks would protect the building from oil ignition

within an enclosure much like the IQ furnace. Apneumatic elevator can provide consistent motioninto the oil even during an electrical interruption.

Figure 7. Typical pit stacked gear load

8

Figure 8. Resulting flame from pit oil quenchof stacked gears

Recipe driven quench distortion control

Martensite, the transformation product providingthe hardness and strength required in ferrousalloys, creates a volumetric expansion that contrib-utes to distortion especially when formed at unequalrates in the carburized case. A major key tocontrolling distortion is creating a heat transfer rateuniformly over the part surface, in this case over thegear teeth profile and root profile. In addition, andperhaps as important, is the heat transfer over thegear hub or side areas that can cause oil canningand bore tapering. Typically in pit quenching fix-tured gears are stacked with little or no spacebetween them, resulting in trapped oil vapor andcreating slower cooling than at the top or bottom ofthe gear stack, Figure 9.

Figure 9. Vertical oil flow in typical pitquench

Traditional pit quench systems circulate oil frombottom to top leaving the oil between gears tostagnate, vaporize and reduce heat transfer. Atypical method to counteract or reduce the differen-tial cooling effect of vapor in pit quenching is toreduce the overall agitation which increases ratherthan reducing oil vapor over the entire gear or stackof gears in an attempt to reduce distortion byslowing the heat transfer of the entire gear. Onlydue to the high hardenability of expensive alloys canheat treaters even consider this approach. Unfortu-nately this technique fails to account for theexposed top of the part surface that is free torelease the oil vapor via gravity. CFD modelingstudies by others have shown that with little or noagitation, the top surface of quenched parts coolsfaster due to the gravity effect of vapor bubblesreleasing. In addition, depending on the gear webdesign vapor can be trapped under the gear, furthercausing nonuniform cooling. A new approach tomore uniform quenching is needed. Oil can becirculated from side-to-side, flushing the oil vaporfrom between gears or from under gears, Figure 10.Vertical oil flow works well for and is required forpinions, sun gears and stacked planetary gearswhere vertical straightness is critical. EZ-Lynks’quench system shown in Figure 11a and Figure 11b

9

can provide recipe-selected oil flow in either direc-tion, side-to-side or vertical, or a combination ac-cording to process requirements. For improved ODroundness control while employing side-to-sideflow, gears can be rotated 360 exposing the entireOD to very uniform and high velocity oil flow directlyat the gear teeth. This feature also assures themaximum quench penetration into the tooth profileand root. As a result, any problematic NMTP can beeliminated.

Figure 10. EZ--Lynks side--to--side oil flow, oilquench options

Figure 11a. EZ--Lynks tank design for verticaloil flow, oil quench option

Figure 11b. EZ--Lynks tank design forside--to--side oil flow, oil quench option

Carburizing, as stated above, is the mostcost-effective process for improving the wear resist-ance and strength of ferrous alloys, especiallygears. Atmosphere (endothermic) carburizing isthe most applied process throughout the worldsimply because it is so predictable even withoutadvanced simulation models. However, withoff-lineand on-line case profile modeling the process isalmost foolproof. Still, due to the very high vestedcost of large precision gears, current wind turbinespecifications demand that samples be removedfrom the pit furnace and evaluated. At strategic in-tervals during the carburizing process prior toquenching, three (3) samples are hand quenchedand checked for case depth and hardness to offerthe chance to save the load if the hardenedproperties don’t meet expectations. Thus is the fearof failure in heat treating – using pit furnaces. It’s theout-of-sight/out-of-mind mentality. Other than shimstock tests to calibrate oxygen probes, no othercarburizing market requires the evaluations ofsamples to confirm process success. Public trans-portation applications such as aerospace gearmanufacturers have the majority of their drive traincomponents carburized in UBQ furnaces -hundreds at a time - without the need for periodicin-process sample evaluation.

10

EZ-Lynks retains the best of UBQtechnology

For those who say only pit furnaces can success-fully carburize wind turbine gears, Flame MetalsProcessing Corp. of Rogers, MN (mentionedabove) has been an approved supplier for wind tur-bine gears for several years. The argument bysome against refractory lined furnaces (furnaceswithout the stainless steel retort) was the inability tocarburize gears without excessive IGO. ISO stand-ard 6336 part 5 issued 2002 (Calculation of load ca-pacity of spur and helical gears - part 5: Strengthand quality of materials), outlines the acceptableIGO limits depending on carbide distribution. FlameMetals utilizing the 36 x 72 x 54 (914 mm x 1829 mmx 1372 mm) UBQ batch furnace can meet or exceedthe applicable standards as referenced in the6336-5 specifications. Figure 12 contains a photoof a carburized and oil quenched planetary gearload. Typical results from such a load are shown inFigure 13a, Figure 13b and Figure 13c. [2]

Figure 12. Flame metals planetary gear load750 lb (341 kg) each

IGO has long been a point of contention incarburizing wind turbine gears since fatigue stressis a significant factor affecting traditional gear lifebecause general practice has been to grind thetooth flank but not the tooth root. Since the toothroot retains its virgin surface, IGO becomes apotential failure risk. Having said that, it’s beenreported that no tooth roots have been identified asinitiating gearbox failure sites. If gears have failed, itis presumed to be caused by bearing wear resultingin accelerated tooth wear from misalignment of the

gearbox and/or carrier. It follows that there needs tobe more intensive auditing of the bearing heat treat-ing process and material selection. Assuming thatbearing failures will eventually be eliminated or atleast reduced the next weak link then may well betooth fatigue failure and to address those issues astronger tooth hardened case profile will berequired.

To further illustrate the point regarding IGO and itscauses, we conducted our own internal investiga-tion at AFC-Holcroft (in addition to data provided byFlame Metal on 18CrNiMo7-6) to determine theeffect steel chemistry has on IGO formation. Weconducted a series of bench scale tests with three(3) common U.S. carburizing steels listed as a, band c:

a. 20MnCr5, C-0.2%, Si-0.25%, Mn-1.25%,Cr-1.15%, S-<0.25%

b. 9310, C-0.10%, Si-0.22%, Mn-0.55%,Cr-1.20%, S-<0.35%, Mo-0.10%, Ni-3.25%

c. 8620, C-0.20%, Si-0.31%, Mn-0.82%,Cr-0.50%, Mo-0.20%, Ni-0.52%

d. 18CrNiMo7-6, C-.18%, Si-0.20%, Mn-0.70%,Cr-1.65%, Ni-1.55%, Mo-0.30%

Table 1 is a description summary of the test para-meters, Simultaneously all samples were carbur-ized at 926C (1700F) for 24 hours with endo-thermic gas with varying concentrations of carbonmonoxide (CO) and carbon potentials resulting inthe following tabulation. These steels were chosenfor testing because they represent a large harde-nability range and where higher manganese, an in-expensive but IGO contributor and an alternative tomoly and nickel for hardenability is gaining interestto reduce material cost.

To investigate further the effect temperature has onIGO formation since high temperature carburizing isgaining favor due to finer grain steels, a carburizingtest was conducted using the bench set up on steelsa, b, and c at 1.04% CP for seven (7) hours at1850F (1010C). Although the IGO depth is great-er, it’s not significantly so. The exception andperhaps an anomaly is the depth for 20MnCr5 that isless than the 24 hour test and that’s extremelyunlikely and will be re-evaluated. Carburizing at1850F for seven (7) hours will produce approxim-ately the same case depth as carburizing 24 hoursat 1700F. Carburizing at elevated temperaturesprovides the most dramatic reduction in time of anyprocess control parameter.

11

Figure 13a. Typical flame metals planetary gear process report

12

Figure 13b. Typical flame metals planetary gear process report

13

Figure 13c. Typical flame metals planetary gear process report

14

Table 1. Test parameter summary

Steel

Carburized for 24 hrs @ 1700F (925C)

Carburizedfor 4.5 hrs. @

1700F(925C)

Carburizedfor 7 hrs @

1850F(1010C)

Carburizedfor 38 hrs @

1725F(940C)

CO 18.3%CP 0.82%

(Alloy)

CO 14.8%CP 1.16%

(Alloy)

CO 14.8%CP 1.16%

(IFB)

CO 18.3%CP 1.04%

(Fiber)

CO 18.3%CP 1.05%

(IFB)

CO 18.3%CP 1.05%

(IFB)

CO 20.0%CP 1.18%

(IFB)

20MnCr5 23 20 22.9 25.4 15.3 17.2 Flame metal

9310 18 12.7 12.7 15.2 7.6 20.3

8620 25 25 24 22.8 10.2 27.9

18CrNiMo7--6 20.0

NOTE:25.4 microns = 0.001”

To complete the investigation a short 4.5 hourcarburizing test at 1700F (925C) was conductedat 0.75% CP to try and establish when IGO penetra-tion became a significant risk factor. There’s somethought that a major portion of IGO is formed at thebeginning of carburizing. Results of the 4.5 hourtest are shown above.

From these short term tests, one could argue thatindeed well over half the depth of IGO can form incertain steel grades but more testing will be requiredto identify a trend.

In addition to comparing a steel grade’s propensityto form IGO, we wanted to know if furnace construc-tion had any bearing on IGO formation. As we allknow, the traditional pit furnace has alloy retorts thatseparate the refractory from the load and carburiz-ing atmosphere. Thereby, it is felt in some circlesthat this represents a more pure environment result-ing in lower IGO values. Still others believe that theendo atmosphere alone causes IGO formation.These theories were evaluated by (1) conductingthe test in the alloy pipe simulating the alloy retort.(2) Adding ceramic fiber and IFB into the alloy pipewithin the cold to hot transition simulating the re-fractory in conventional carburizing furnaces. Theresults above speak for themselves. No differencein IGO penetration regardless of the presence ofceramic fiber or IFB refractory.

Commentary by Flame Metals ProcessingCorp. [3]

Flame Metals Processing Corporation, being acommercial heat-treating facility, had ap-proached the problem of unacceptably deep

IGO penetration in a series of deep-case carbur-ized products. Considering the need to maintainflexibility and economical competitiveness,Flame Metals decided not to invest in special-ized equipment and technology but rather to im-prove on the carburizing process based onexist-ing flexible-purpose IQ furnaces andeconomically feasible natural gas generated at-mosphere.

Because the total carburizing times for eachproduct were already optimized in terms of pro-cess temperature and that of the sequence ofthe boost and diffuse cycles, the slow IGOgrowth was inevitable and had to continue withinthe prescribed carburizing time. However,Flame realized that the large portion of the IGOgrowth is produced during the initial stage of theprocess and is driven by the events that could becontrolled and improved such as loading pro-cedure and bringing the load to the heat. Inotherwords, the main thrust of the effort was directedat reducing the initial oxidation. The effort paidoff resulting in dramatic decrease in IGO penet-ration depths.

Total furnace time in both cases was 6 hourslonger and included the identical procedures in-volving charging the load into furnace (0.5hours), heating to the carburizing temperature(3 hours), and cooling down to lower temperat-ure before quench (2.3 hours). In general, theheating time to the temperature may be reducedbut in this case slow, controlled heating was ap-plied in order to reduce distortion of the parts.

The steel alloying plays some role in IGO devel-opment in long-time carburizing process, but it

15

seems to be relatively minor when compared tothe other factors. For example, after 34.5 hoursof carburizing process conducted under thesame conditions as given above, the IGO depthmeasured on 18CrNiMo7-6 steel was 24 mm and28 mm on 8620 steel (see Figure 14).

It is our conviction at Flame Metals that the ma-jor event influencing development of IGO occursin the initial stage of process. When the partsare loaded into the furnace, inevitably oxygen isintroduced as well. Flame Metals was able to re-duce the oxidation risk by purchasing exception-ally gas-tight IQ furnaces and implementingspecial bringing up to heat procedures. As a res-ult, the steel surface exposure to the oxidationby residual oxygen at high temperatures is min-imized.

Historically, pit furnaces, from the writer’sperspective based on decades working in the heattreating industry, have been known as soot generat-ors due to the difficulty in accurately controllingcarbon potential. Currently the only way pitfurnaces in the U.S. and Europe used for windturbine carburizing can be successfully managedeven employing the sampling method referencedpreviously is with very expensive process controlsystems with multi-gas analysis. Costing severalhundreds of thousands of dollars all by themselves,these control systems manage pits for the samereason computers keep the B2 bomber in the air –from an aerodynamic standpoint they can’t fly.

EZ-Lynks brings a cost effective common senseapproach to carburizing large awkward compon-ents to the heat treating industry. As with the UBQbatch furnace cell, EZ-Lynks can be programmed toautomatically retrieve loads without manualintervention, transfer to the available furnace,quench, wash and temper the load and transfer theload to the output queue, all the while tracking theload location within the cell and trending processdata for archiving to an industrial-strength computerwith automatic dual hard-drive back up.

Batchmaster II standard control systemdeveloped by AFC-Holcroft manages the carburiz-ing process with PLC logic providing consistent andrepeatable results. Associated software tracks allloads through the cell providing real-time in/out dataplus appropriate information to a company’s hostallowing SPC analysis if required.

Operating complexity and cost when consideringlean manufacturing is becoming an important para-meter in the purchase decision for captive andcommercial heat treaters as utility costs continue torise. Both depend on operating efficiency andup-time to achieve the lowest possible cost perpound processed per square foot occupied. Pitfurnaces in principle seem to be fairly simpledevices (the furnace alone can be the lowest costinvested) until all of the associated components areconsidered to process with just one pit furnace.Table 2 is a comparison of items required for pit andEZ-Lynks operation.

IGO penetration

Carburizing time, hrs

IGO

,mm

Figure 14. Flame metals (IGO) graph depth vs. time and material

16

Table 2. EZ--Lynks and pit item comparison

Item requirement Pit system EZ--LynksVery expensive multi--gas processcontrol system

Yes No, standard controls only

High capacity bridge crane systems Yes No, only for single partloading

Anti--sway crane capability Yes, to make loading easier No

Inner and outer alloy retort replacementcost

Yes No

Water--cooled lid seal maintenance Yes No, water cooling required

Hanging fixture assembly Yes, can consume 30% ofgross load weight

No (tray only or no trayrequired)

Seal maintenance, depending on howthe retort ends are sealed -- sand orexpansion joints

Yes No

Elaborate washing systems Yes No (standard washer)

Quench tanks that lack versatility plusopen tank fire hazard

Yes, open tank fire hazard Enclosed tank with flexiblequench

Extra high facility bay Yes No

Elaborate fire suppression systems Yes No, vestibule can be nitrogenpurged

Large and deep pit required Yes No

Finally, to further the comparison, from aneconomicstandpoint, Figure 15a and Figure 15b highlights anoperating cost comparison between a typical pitsystem and EZ-Lynks when carburizing a maximumload to 3.3 mm ECD at 940C (1725F). This cyclerepresents a typical case depth for a 2 to 2.7 mega-watt (MW) wind turbine 24” (609 mm) diameterplanetary gear and other gear components in theturbine gearbox. Calculations shown in the Excelworksheet take into account the average utilitycosts for natural gas, electricity and nitrogen in Ohioand Michigan. EZ-Lynks can be supplied withelectric or gas heat, but this comparison uses gasheating with metallic alloy radiant tubes. The pitfurnace is calculated with electric heat since mostsystems used throughout the U.S. and especiallyEurope are electrically heated. In addition, the typ-ical pit furnace fixture whether hanging or bottomsupported is very substantial and contributes about30% of the gross weight of the pit load. This is anobvious disadvantage since the fixture is quenched

with the load and as such has a useful life of aboutthree (3) years under ideal conditions. EZ-Lynks incontrast employs a standard tray system and issubjected to much less overall stress during carbur-izing and quenching.

As stated in the opening of this paper, wind turbinegearbox failures may have less to do with gears andmore the result of premature bearing wear, lubrica-tion issues and environmental forces still not fullyunderstood. However, if gearbox wind turbines areto ever attain the popularity once envisioned, theseproblems must be overcome. Rare-earth magnetwind turbines although more expensive may seemore offshore applications but they too will employheat treated bearings among other components.No matter what direction the alternative energyindustry takes, gears and/or bearings will be heattreated and manufacturers and heat treaters byeconomic necessity will benefit from a largecomponent flexible heat treating system.

17

Figure 15a. Operating cost comparison between a typical pit system and EZ-Lynks

18

Figure 15b. Operating cost comparison between a typical pit system and EZ-Lynks

19

References

1. National Renewable Energy Laboratory(NREL), Kathleen O’Dell, July/August 2009,p. 22.

2. Planetary Gear report provided by Mike

LaPlante, General Manager, Flame Metals Pro-cessing Corp., April 28, 2010.

(3) Commentary provided by Dr. Jarek Haftek,Metallurgist, Flame Metals Processing Corp.,April 23, 2010.