Polyurethane Foam Molding Technologies for ImprovingTotal Passenger Compartment Comfort

R. BROOSDow Benelux N.VP.O. Box 484530 AA Terneuzen,The Netherlands

J.-M. SONNEY, H. PHAN THANH and F.M. CASATIDow Europe S.AInternational Development Center13, rue de Veyrot - P.O. 3CH-1217 Meyrin 2 (GE)Switzerland

ABSTRACT

Improvements in passenger compartment comfort con-tinue to be one of the key needs of the global transportationindustry. Since their introduction more than 40 years ago,flexible molded polyurethane foams have successfully con-tributed to the comfort provided by all forms of transporta-tion seating. Initially required to provide just a wide rangeof load bearing, seating foams are now being designed forlonger service life and better vibration damping and areconsidered to be a functional part of the overall acousticalpackage. New performance requirements are being placedon the NVH grade of foams and all interior components ofpassenger compartments must contribute to a reduction inodor and emissions.

To address the ongoing and newer challenges, The DowChemical Company has developed several new chemicaltechnologies that will be detailed in this paper. Specificdata will be presented to illustrate new foam chemistriesthat offer low emissions, reduced density, improved vibra-tion management, noise abatement, durability and process-ing. Attention will be focused on how the choice ofpolyurethane raw materials impacts foam resilience, me-chanical damping and resonance as well as acoustic absorp-tion and noise transmission.

The results presented in this paper will help foam pro-ducers, seat assemblers, acoustic part manufacturers andOEM interior designers in their efforts to further improvethe comfort performance of transportation systems basedon flexible molded polyurethane foams.

INTRODUCTION

Polyurethane molded flexible foams [1] are key compo-nents of automobile interiors and contribute to passengercomfort in many different ways. More than 40 years after

their introduction, the foam use per vehicle is still growingworldwide. Indeed, by virtue of their superior vibrationdamping capability over a wide range of frequencies at lowmaterial density, polyurethane foams are found today notonly in automotive seats but also in various acousticalparts.

Comfort experience is a combination of many differentfactors, including aesthetics. During a journey, car occu-pants are subjected to both mental and physical stresseswhen exposed to road vibrations, dense traffic, noise anddifferent weather conditions. Polyurethane foams found inseats and in sound insulating packages are instrumental inreducing those stresses. Design for comfort is intimatelylinked to vibration research. Vibration isolation in the lowfrequency range by the seat assembly and the attenuation ofhigh frequency vibrations from the running engine, or othersources, transmitted into the passenger compartment, withNVH (Noise, Vibration, Harshness) components, allowsthe construction of comfortable cars. In addition, polyure-thane foams show high durability and perform well fromthe beginning to the end of a journey and over a vehicle lifeof more than one hundred thousands miles. OEMs world-wide are relying on flexible polyurethane foams today morethan ever to optimize automotive comfort.

The trend towards density reduction whilst maintainingtechnical performance specifications continues. In additionof that, elimination of all types of chemical emissionsand/or odors is becoming an important issue. Dow Chemi-cal is developing new raw materials to respond to theseneeds.

Figure 1: Damped harmonic oscillator model for seatperformance

BACKGROUND

SEATING COMFORT

Human exposure to mechanical vibrations causes fatigueand discomfort. The magnitude of the effect depends on theintensity, duration and directions of the excitation. In seatcushion design, the most important vibration frequency isthe “occupant-cushion-seat base” mass-spring-mass systemresonance frequency, but the filtering at higher frequenciesis important as well. The basic design requirement is toensure that the resonance motion of the seated person oc-curs in a frequency range where the force input from theroad is low (2-5 Hz) and outside of the body high sensitiv-ity range (4-8 Hz). In practice, for a predetermined loadbearing requirement, foam development is striving to shiftthe body motion resonance to lower frequency. The ampli-fication factor at resonance needs to be kept as small aspossible to minimize the total force transmitted to thedriver.

In use, the seat foam core is subjected to multi-axialstresses of variable amplitudes and frequencies and theimposed strains may be large. Automotive force transmis-sibility tests, such as the Japanese Automobile StandardsOrganization (JASO) B 407-82 test for example, are al-ready a simplification as the applied static force and strainamplitude are constant and a quasi-sinusoidal frequencysweep is applied. In a previous paper [2], we have demon-strated that the resonance frequency of the “Tekken” testassembly (JASO test) and the amplification at resonanceare readily approximated by the equations for the dampedharmonic oscillator [3] (Figure 1). The foam’s cushioningperformance is modeled by the combination of a springwith spring constant (k) and a dash pot with damping con-stant (C). The spring represents the foam dynamic stiffness.The total foam damping capability is the sum of visco-elastic, pneumatic and frictional effects (emphasized sepa-rately in the center of the figure).

The steady state solutions for the resonance frequency, fr,and the amplification factor of vibration at resonance, ar

(also called Transmissibility, Tr), for the model are de-scribed by the equations below:

Thus, for a fixed test assembly, the resonant frequencyvaries only with the foam dynamic stiffness k in the de-formed state (from the static weight M). The dynamic stiff-ness (= foam spring constant, k) is dependent on the foampad thickness (= length of the spring) and the foam hard-ness (modulus). To ensure that the resonant frequency islow, the foam should be tuned such that the use defor

(1)2

1=fr

M

k⋅π

2)(.C

1=ar Mk⋅

mation is in the plateau of elastic buckling where the dy-namic stiffness is near its minimum [4].The amplification factor (a) is proportional to the foamstiffness and inversely proportional to the damping con-stant (C). The amplification at resonance ar is reduced byincreasing the foam damping constant (hysteresis loss);higher damping however tends to considerably broaden thetransmissibility peak, such that the point where the systembecomes vibration isolating is reached later and the damp-ing capacity at high frequency is diminished. In practice, acompromise needs to be sought.

NOISE CONTROL

In the automotive domain, acoustical comfort relatesboth to the comfort of the passengers and the comfort ofthe pedestrians and people living close to the traffic. Noiseemissions by motor vehicles were set in Europe by the1995 EU “pass-by” legislation. Comparable regulations areplanned in other geographical regions. Interior car noisequality has become a major criterion in the highly competi-tive automotive market. Because of the inherent ability ofpolyurethane to produce foams with very different physicalcharacteristics, it has been the material of choice in manyapplications in Automotive Noise and Vibration Controltreatment.

Hirabayashi and al. reviewed the mechanistic aspects ofNVH control [5]. A noise system can be analyzed in termsof three basic elements: the noise source, the noise trans-mission path and the receiver. The three dominant noisesources in a car are the drive train, the tires, including sus-pension system, and the wind. The noise paths are complexbut, in principle, sound is propagating through air (airbornesound) or through the car body (structural noise). Polyure-thane foams are an integral part of acoustic barriers placedto attenuate the sound wave before it reaches the occupant(receiver). In the cabin, foams act as absorbers reducing thetotal sound pressure around the passengers.

Sound absorption

The design goal for all soft trim parts in the automobileinterior is to provide as much sound absorption as possible.The seats account for about half of the total absorptionpossible in a typical car [6]. The headliner is the next im-portant absorber. For a seat assembly with a given surface,the sound absorption characteristics depend on the struc-ture of the seat cover materials and the type of foam usedinside the cushion and backrest. Cheng and Ebbitt [7]measured noise absorption of seats and covers. Test resultsshowed distinct differences between textiles, leather andvinyl, with cloth seats having superior properties. Foamcore densities affected the performance as well.

In this section we will more closely examine foams, withand without skin, and compare the absorption coefficientsof different foam chemistries. The aim is to correlateacoustic performance with foam structural parameters, inorder to optimize the seat acoustic performance.

The propagation of sound waves through porous mediaas described by the theory of Biot has been applied topolyurethane foams by Lauriks [8]. Theoretical [8] andempirical [9] models exist, but are complex with respect toinput parameters and computing mathematics. The impor-tant foam parameters also can be introduced intuitively.

When an airborne sound wave hits a rigid surface, forexample a metal sheet or a closed cell rigid foam, it will bereflected substantially, like light in a mirror, and only aminor fraction of the sound energy enters the material. Thesituation is quite different when the foam cells are open andthe air borne wave continues its way into the foam struc-ture. Upon propagating through the pores, the sound pres-sure is reduced by friction of air against the pore walls andeventually by moving the cell membranes when the foam issufficiently soft. The longer the wave travels through thefoam, the more damping occurs. Therefore, foam thicknessis an important parameter of absorption performance. Atconstant foam layer thickness, the next parameter to opti-mize is the amount of friction between the medium (air)and the frame (foam skeleton). Intuitively, it can be felt thatpore diameter or flow resistivity will play a determinantrole. To transfer energy from the medium to the skeleton,the frame modulus should be sufficiently low, which is thecase with soft foam (measured by IFD, CFD for example);and the surface interaction should be large.The absorption coefficient αααα of a material is defined as:

αααα = 1 – [Rr/Ri] (3)

with Ri = incident sound wave pressureRr = reflected sound wave pressure

It can be understood that the absorption increases: as theamount of accessible pores increases i.e. with increasingacoustic porosity or open cell content and within certainranges, increases with decreasing pore diameter (D) or in-

creasing flow resistivity and finally is influenced by theframe flexibility.

Interestingly, these foam parameters have been histori-cally measured and used to characterize foams (open cellcontent, airflow and load bearing). In this paper, we willcorrelate such parameters with the absorption coefficientmeasured for different foam chemistries.

Sound Transmission

Barrier materials are positioned in the noise path to re-duce the amplitude of the sound waves propagating in acertain direction. In automotive application, sound absorb-ing foams are in contact with the metal shell to reduce theamount of noise entering the passenger compartment. Theeffectiveness of a material as a barrier is measured by theTransmisson Loss (TL). The transmission loss is the differ-ence in sound intensity before (Ii) and after (It) the barrierexpressed in dB:

TL = 10 log Ii/It (4)

The transmission loss is not a foam property as such, but isthe performance of a sandwich structure consisting of ametal sheet and a heavy layer with a foam core. In thisstudy we have used the “Petite Cabine” as acoustic trans-mission screening tool (see Experimental). The foam per-formance is characterized by its Insertion Loss (IL) definedas:

IL = TL(with composite) – TL(empty system) (5)

The acoustic performance of the composite is governed bythe mass law, which determines the basic shape of the ILcurve. On the base curve, structural (mass-spring-mass) andcavity resonances are superimposed, which are inherent tothe construction of the measuring chamber [10]. The inser-tion loss performance increases with increasing septummass and increasing foam thickness [11], which is contraryto the volume and weight saving objectives in automobiledesign. Polyurethane foam cores, by their energy dissipat-ing mechanisms, allow for considerable weight and volumereduction in the vehicle while maintaining good acousticperformance.

EXPERIMENTAL

Foam Preparation

Bench scale molded foams were produced in a 30x30x10cm test block style mold heated at 550C using the standardhand-mixing procedure [1]. High pressure impingementmixing machines were used to pour foams in a 40x40x10cm standard test block style mold and in a variety of pro-

duction seat molds. Foam formulations are indicated withineach section of the paper.

Foam physical property testing

Basic foam physical properties were determined in ac-cordance with ASTM, ISO, DIN, NF and major OEM testprocedures. Tropical dynamic fatigue tests were carried outat 40°C and 80 % RH as described in a previous paper[12].

Crushing force and hot IFD’s

A simple test was carried out to measure the foam tight-ness at demold and the level of foam curing. The foam padwas demolded with precautions and one minute after de-mold was indented at 50 % deflection. This value in New-ton is referred to as the “Crushing Force” (CF) Then thefoam was passed through a roller crusher to open all thefoam cells and the pad was again measured for IFD hard-ness at 50 % deflection. This second value, referred to as“Hot IFD” in Newton, gives an indication of the level offoam curing. Obviously, the higher the CF, the tighter thefoam, and the higher the Hot IFD, the better this foam iscured.

Vibration testing

Vibration tests at low frequency (1-10 Hz) were per-formed according to the Japanese Automobile StandardsOrganization (JASO) B-407-82 method on a Itoh Seiki,Model C-1001-DL vibration table. Displacement amplitudetransmissibility versus input frequency is recorded. Addi-tional vibration studies were performed with a MaterialTest Systems (MTS) Model 831 servo-hydraulic loadframe, equipped with TestStar II electronic controls and aRadicalor environmental chamber (23°C; 50 %RH). Testprocedure: 100x100x50 mm samples; static load 40N, dy-namic load amplitude 10 N; frequency sweep from 1 to 10Hz in steps of 0.25 Hz.

Acoustic testing

Normal incidence sound absorption coefficients weremeasured between 100 and 5000 Hz with a two micro-phones impedance tube according to ASTM E 1050 (Brüel& Kjaer 4206). This method utilizes a hollow tube, whichis equipped with a speaker at one end and a sample holderat the other. Plane waves are generated in the tube by arandom noise source, and the decomposition of the stand-ing wave is achieved with the measurement of acousticpressures at two fixed locations close to the sample usingwall-mounted microphones. Using a digital frequencyanalysis system, the complex acoustic transfer function ofthe two microphone signals is determined and used tocompute the normal incidence absorption of the acousticmaterial. Air Flow Resistance measurements were made

with the AFM 80 Instrument from Norsonic Brechbühlaccording to DIN 52 213. Insertion losses were measuredwith the Petite Cabine. This test was developed by Renaultand is used to expediently measure insertion loss for a vari-ety of insulators, which are usually employed as dash orfloor insulators. The test equipment essentially consists of aconcrete base with nine loudspeakers, the excitation cham-ber, separated by a metal frame with a 0.8mm steel metalsheet to accommodate the 700x700x20mm test samplefrom a semi-anechoic hood with a microphone, the receiv-ing chamber. The semi-anechoic hood prevents reflectionof the transmitted sound energy. The loudspeakers are ex-cited with a filtered white noise to produce a sound pres-sure level of about 80 dB from 100 to 5000 Hz in thereceiving chamber with the sheet metal panel.

The Petite Cabine test is conducted by first measuringthe sound pressure level (from 100 to 5000 Hz) producedin the upper chamber with only the sheet metal panel in-stalled. Next the insulator material consisting of the damp-ing material and a 5 kg/m2 heavy layer is installed, and thetest repeated. By subtracting the levels measured with thetest material installed from those measured with the barepanel, the insertion loss (IL) in dB of the insulator is de-termined.

PVC Staining/Aging

Accelerated aging tests at elevated temperature were car-ried out in closed containers in the presence of a PVC(PolyVinyl Chloride) foil. A foam sample of 50x50x50 mmwas cut from the core of the molded foam pad and placedat the bottom of a one liter glass jar. A piece of gray PVCskin, reference E 6 025 373A0175A, supplied by Benecke-Kaliko, was hung with a Chromium-Nickel alloy-basedstring supported by the rim of the jar, which was thensealed. Aging was carried out at 115°C for 72 hours. Aftercooling, the PVC sheet discoloration was measured using aMinolta Chroma Meter CR 210. The smaller the change incolor, the lower the ∆ E measured in this test.

RESULTS AND DISCUSSION

A) SEATING COMFORT

Polyol developments

In 1994, Dow introduced novel polyols, which were de-veloped for cold cure TDI and TDI/MDI blend High Resil-ience (HR) molding [13, 14]. Based on higher functionalityinitiators than those used at that time, the new generationhigh functionality (HF) polyols were designed to give bet-ter foam cure at demold, higher foam resiliency, and betterdurability.

Figure 2: Effect of the Transmissibility at resonance (ar) onthe Transmissibility at 6 Hz (TDI/MDI based foams, density:50 kg/m3, hardness: 20-22 kgf/314 cm2 at 25% indentation)

in the JASO test.

Table 1 : Polyol Analytical Data

Polyol A BCalculated Functionality* 3.0 3.2EW 1810 2150Viscosity at 250C (cSt) 1250 1530* calculated from the initiator functionality, the hydroxylnumber or EW and the level of unsaturation (monol).

3.0

3.5

4.0

4.5

5.0

5.5

0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4

Transmissibility at 6 Hz

Tra

nsm

issi

bilit

yat

reso

nanc

e

TDI-based foams produced with these HF polyols wereevaluated in the Pressure Distribution Test and the SeatOccupied Vibration Test [15]. In the first test, it appearedthat foams based on HF polyols gave better pressure distri-bution than foams based on traditional materials, which isknown to be essential to the comfort of the seat. In the vi-bration test, however, it was found that HF polyols did notshow any significant difference versus conventional polyolsused in high resilience formulations.

Foam thickness, foam hardness and foam resilience arethe main parameters which determine the general shape ofthe force/displacement transmissibility curve. When thefoam thickness is increased, the mass-spring-mass reso-nance shifts to lower frequency. In the model of the har-monic oscillator, this corresponds to reducing the springstiffness by making the spring longer. However, increasingthe foam pad thickness is limited from a designer’s stand-point, since it reduces the passengers’ space in the car.Foam modulus directly impacts the spring constant. Glob-ally, softer foams (lower k) exhibit a lower resonant fre-

Table 2: Physical Properties (foam

Isocyanate TDIPolyol AHardness. IFD 25% Kg 20.5Hysteresis Loss % 21.1Tear Strength N/m 325Tensile Strength kpa 160Elongation % 125Foam Resilience (% 65Dry C-Set (50%) %CT 6.3Wet C-Set (50%) %CT 10.6Humid Aged C-Set (50%) %CT 6.1

quency. But, considering the range of typical OEMs’ hard-ness specifications from 15-28 kg/314 cm2 at 25% deflec-tion- the achievable resonance shift range is narrow andconsiderably restricts foam hardness as a design tool. Yet,it remains important to tune the foam hardness such that thecushion’s dynamic stiffness is minimized under the actualuse deformation. Foam resilience is associated with thedamping constant (C) and impacts the transmissibilitycurve such that with increased damping the amplification atresonance is reduced but the resonance peak is broadened.As a consequence the vibration damping capacity at highfrequency is reduced and the point at which the systembecomes vibration isolating (Transmissibility <1) is oftenreached later. Thus, the model predicts that high resilientfoams yield a higher amplification at resonance but becomevibration isolating earlier. This effect is illustrated inFigure 2, where data from the JASO test are collected fromprevious tests and the amplification factor at resonance isplotted against the transmissibility at 6 Hz: the higher theamplification factor at resonance, the more damping at 6Hz. All of these foam pads had resonance frequencies be-tween 3.2 and 3.6 Hz measured with a thickness of 10 cm.

Although our first generation polyols for HR moldingdeveloped in the early 90’s were suitable for most OEMs’requirements, foam resilience was still borderline for someof those calling for ball rebound of 70% or higher.

index 100, density 45 kg/m3)

TDI TDI/MDI TDI/MDIB A B

20.0 20.0 19.018.9 19.7 18.2315 260 215145 120 205135 95 13570 68 743.3 5.6 3.19.4 19.3 11.74.7 13.9 11.4

Figure 3: Vibration Curves (MTS test) for TDI/MDI (TM20)HR Foams (Polyols A & B)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

0 1 2 3 4 5 6 7 8 9 10

Frequency (Hz)

Tra

nsm

issi

bilit

y

Polyol A

Polyol B

Therefore, Dow developed second generation polyolsaiming at HR molded foams with very high resiliency andfurther improved physical properties [16]. Polyol A andPolyol B, in Table 1 and Table 2, refer to the first and sec-ond generation polyols, respectively. Both with TDI andTDI/MDI blends, the foams made with polyol B showhigher resilience than the one produced with polyol A. It islikely that the increase in ball rebound is associated withthe higher equivalent weight (EW) of the polyol B yieldinga more elastic foam with lower hysteresis loss. In addition,the optimization of the polyol structure calls for the im-proved physical properties listed in Table 2.

Vibration performance simulation was also performed onthe MTS servo-hydraulic frame with samples taken frommolded foam pads at density 45 kg/m3 that had been madesuccessively with polyols A and B. An example thereofrelating to TM20 foams is shown in Figure 3. It is typicalfor these high resilient foams to exhibit a relatively hightransmissibility at resonance (amplification factor 6-7 at 4.5

Table 3: Comparison of the performance of differe(foam density:

Foam Chemistry Hot cure TDHeight loss % 3.8 4.4Load Loss % 18.6 26.

Table 4: MDI HRM Foam

Isocyanate APart weight g 639Molded density Kg/m3 50.1IFD 30 mm N 240BMW tropical fatigueHeight loss % 2.6Load loss % 7.5Corrected load loss % 16.3Peugeot fatigueHeight loss % 3.3Load loss % 14.3

Hz), but to become rapidly vibration isolating between 6and 7 Hz. The slightly softer second-generation foam B hasthe lowest resonance frequency and a slightly higher ampli-fication, in accordance with the resilience data.

Isocyanate developments

In a previous paper [12], we have shown that in a typicaldynamic fatigue test which ran for 15-20 hours, the foamheight and load loss have already reached their final valuesafter the first few hours of the test. For instance, a foamwhich gives 25% load loss after 15 hours has alreadyreached 20% load loss after a 4-hour test. Although thedynamic fatigue test usually reflects the durability perform-ance of the foam, it can also be considered indirectly as acomfort test, since, during a long journey in a car, it is de-sirable for the driver and for the passengers if the seat(Hip) H-point [17] change remains as low as possible.

In a tropical dynamic fatigue test, it was found that theperformance of hot cure molded foams in terms of heightand load loss is generally better than that of cold curefoams, and among cold cure foams, MDI HR moldedfoams are better than TDI or TDI/MDI foams (Table 3).This is why the MDI technology is commonly used formaking foams required to meet drastic tropical dynamicfatigue tests. On the other hand, the major drawback ofMDI based HR molded foams has been the limitation indensity reduction due to poor flowability in complex moldsand to the negative effect of increased water level on proc-essability and foam durability properties [18]. Recentlynew isocyanates and improved polyol formulations haveallowed a significant improvement in density reductionwhile maintaining foam durability [19].

nt foams in theTropical Dynamic Fatigue test45 kg/m3)

I TDI/MDI MDI (current) MDI (new)6.1 3.4 3.5

0 29.8 21.3 16.9

physical properties

Isocyanate A Isocyanate B Isocyanate B595 641 59846.6 50.2 46.9210 226 200

2.5 3.3 4.17.4 10.8 9.115.7 20.2 20.5

3.0 4.4 4.016.1 11.1 12.0

Table 5: Vibration MDI HR Molding (pads75 mm thickness, JASO test)

Isocyanate APolyol A

Isocyanate APolyol B

Foam density kg/m3 56.0 55.5CFD (40%) kPa 8.4 7.8Foam Resilience % 60 58Resonance frequency Hz 4.3 4.7Transmissibility at resonance 4.3 4.3Transmissibility at 6 Hz 1.2 1.7

0

0 .2

0 .4

0 .6

0 .8

1

0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0

H o t M o ld e d

T D I H R

T M 2 0 H R

M D I H R

Ab

so

rp

tio

nC

oe

ffic

ien

t

F re q u e n c y [H z ]

Table 4 shows the physical properties of 60mm thickfront seat pads made with Isocyanate A and Isocyanate Bmeeting Peugeot fatigue and BMW tropical fatigue specifi-cations at low densities. These foams were made on a highpressure machine in a mold kept at 55 °C and were de-molded 3.30 minutes after pouring. Isocyanate index was85.

The data show that Isocyanate B gives slightly softerfoams than Isocyanate A while Peugeot dynamic fatigueproperties are superior. By contrast, Isocyanate A yieldsvery good BMW tropical fatigue results. It is important topoint out here that all of these foams meet BMW and Peu-geot hardness targets for front seats.

Similar foams were tested for resonance frequency andthe data are collected in Table 5. The advantage of thesefoams is that the transmissibility maximum at resonance islower compared to TDI based high resilient foams, suchthat less energy is transmitted to the driver going throughthe resonance motion. But, because the transmissibilitypeaks are broader, they become vibration damping onlyafter 6 Hz.

Whilst the present paper is focusing on vibration per-formance of foams between 0 and 10 Hz, a study has al-ready been published on seating foam behavior until 80 Hz[12].

In summary, new polyols and isocyanates have been de-veloped by Dow to broaden the range of possibilities avail-able to foamers aiming to adjust foam vibration propertiesat low frequencies whilst maintaining superior durabilityperformance. The performance at high frequencies of seat-ing foams is also an important factor contributing to theacoustical comfort of a vehicle as will be discussed in thenext section.

Table 6 Physical Properties of Foam

Foam: HCFD (40%) kPaFlow resistivity N.s/m4 Core

with Skin

B) ACOUSTICAL COMFORT

Absorption

The four main seating foam technologies were examinedfor acoustic absorption as a function of frequency in thestanding wave apparatus. Some key physical properties ofthe foams are collected in Table 6 and the absorptioncurves of the foam cores are displayed in Figure 4. Theabsorption curves of all the foams show high efficiency.The shape of the sound absorption curve is mainly gov-erned by the airflow resistance: open foams show initiallyless absorption but better performance at high frequency.Foams which are more closed exhibit a higher slope in thelow frequency range, but are less effective absorbers athigh frequency.

Figure 4: Sound Absorption for 60mm thick core samples

s Tested for Sound Absorption

ot Molded TDI HR TM 20 HR MDI HR8.0 4.0 4.0 7.0

34’000 7’600 7’000 9’00086’000 8’500 7’500 11’000

Figure 5: Effect of Crushing on Sound Absorption

Figure 6: Effect of Hardness on Sound Absorption

Figure 7: Sound Absorption for 60mm thick sampleswith skin

0

0 .2

0 .4

0 .6

0 .8

1

0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0

M D I H R /U N C R U S H E D

M D I H R /C R U S H E D

Ab

so

rp

tio

nC

oe

ffic

ien

t

F r e qu e n c y [H z ]

0

0 . 2

0 . 4

0 . 6

0 . 8

1

0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0

H o t M o ld e d S o f tH o t M o ld e d H a rd

Ab

so

rpt

ion

Co

eff

icie

nt

F r e q u e n c y [ H z ]

0

0 . 2

0 . 4

0 . 6

0 . 8

1

0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0

H o t M o ld e d

TD I H R

TM 2 0 H R

M D I H R

Ab

so

rpt

ion

Co

eff

icie

nt

F re q u e n c y [H z ]

This general trend is confirmed by Figure 5. Crushing thefoam decreases its airflow resistance by a factor of 2.5 andcauses only a 10 % reduction in foam hardness, but mark-edly decreases the initial slope of the sound absorptioncurve and shifts the absorption maximum to higher fre-quency. Foam hardness globally has a limited influence onthe sound absorption coefficient as demonstrated by Figure6. The initial part of the sound absorption curve of a 3 kPaCFD 40 % hot molded foam is not so different from that ofa 8 kPa, while the absorption maxima clearly differ in de-tail. The absorption characteristics after the first maximumare governed by the interference of the sound wave re-flected from the surface of the sample with the sound wavereflected from the back of the impedance tube. The curvesfrom the TDI HR and the TM 20 HR foams show the larg-est amplitude fluctuation (interference pattern), indicating ahigher intensity of the two sound waves; such foams exhibitlower sound absorbing capacity. The curve from the hotmolded foam is relatively flat, indicating a higher soundabsorption.

A completely different situation is created when the foamspecimens are tested with skin (Figure 7). Thus, in princi-ple, a two-layer composite is formed, especially for the hotcure foams. The skin behaves like a semi-imperviousmembrane applied to the foam, such that a strong absorp-tion maximum is observed at low frequency with deteriora-tion of the absorption at higher frequency. HR foams withmuch thinner and more permeable skin, as indicated inTable 6 by the very small difference in the airflow resis-tance between the sample with and without the skin, showthis effect to a lesser extent. It should be remembered thatduring seat fabrication, when the fabric is added, similareffects are created [7, 20]. In summary, all foam chemis-tries show comparable acoustic absorption, provided thatthe processing is well-controlled and the right cell opennessis achieved.

Transmission

Mid 90’s, Dow introduced a novel polymeric MDI,SPECFLEX* NS 540. This isocyanate was designed toproduce foams at a reduced density and faster demoldcompared to prepolymers in use at that time. Since the in-troduction of SPECFLEX NS 540, the trend to lower themolded densities continues. Density reduction technologyusing an auxiliary blowing agent such as carbon dioxidehas been developed. This approach is interesting as it doesnot have a significant impact on the Insertion Loss (IL)[21]. But, when density reduction is achieved by increasingthe water level in the formulation, the IL can be affected asshown in Figure 8. Comparing the IL of two foams basedon the same polyol and isocyanate with 3.2 and 4.2 % wa-ter respectively.

Figure 8: Effect of water content on Petite Cabine IL Figure 9: Petite Cabine IL of foams based on SPECFLEXNS 540

-10

0

10

20

30

40

50

100 250 630 1600 4000

Frequency (Hz)

Inse

rtio

nLo

ss(d

B)

Voranol CP 6001

Polyol C

Polyol D

-10

0

10

20

30

40

50

100 250 630 1600 4000

Frequency (Hz)

Inse

rtio

nLo

ss(d

B)

3.2 % water

4.2 % water

-10

0

10

20

30

40

50

100 250 630 1600 4000

Frequency (Hz)

Inse

rtio

nLo

ss(d

B)

Specflex NS 540

Isocyanate C

Isocyanate D

The formulations were adjusted to give similar compres-sion hardness. The foam with the higher water level clearlyshows a degradation of the IL in the 500–1250 Hz rangeand an improvement in the 1600-2000 Hz range.

In order to match the higher passenger comfort demandsand to comply with the new legislation, the OEMs needmore effective sound insulation packages. Weight reduc-tion is an ongoing trend in the automotive industry asweight is directly related to fuel efficiency. For cost rea-sons, part weight reduction by increased water content isstill the most generally adopted solution. As the emphasishas now shifted from the aim of simply reducing the inte-rior noise toward designing a quality interior noise or a carsound signature, Dow has focused on evaluating the impactof the polyol and of the isocyanate in this area of the acous-tic performance of HR foams.

Polyol developments

The main polyols parameters influencing foam mechani-cal performance include equivalent weight, functionalityand primary hydroxyl content. In this study, primary hy-droxyl content was kept constant to maintain optimumprocessing. In a density reduction effort, lower polyol func-tionality and equivalent weight were examined. UsingVORANOL* CP 6001 triol as the reference, reduction ofthe polyol functionality (Polyol C) or equivalent weight(Polyol D) yielded foams with comparable IL as shown byFigure 9.

The choice of the polyol does not seem to have a deci-sive influence on the acoustical properties of the producedfoam. However, given the high complexity of the NVHparts, the choice of the polyol is critical in ensuring theprocessing necessary to achieve the desired low density andthe wide processing latitude to accommodate the differenttools of a molding line. With increased water level,

these key characteristics are even more difficult to obtainand Dow has optimized polyol structures and formulationsto produce low density foams at high water level.

Isocyanate developments

Since the introduction of SPECFLEX NS 540, thePMDI-based technology has been widely accepted in theNVH market as a cost-effective alternative to the prepoly-mer-based technology. As the need for low density andbetter acoustic foam grows, Dow is developing new MDIsto recover the acoustic performance and the processinglatitude lost with the incremental increase of the water con-tent in the formulation. The effect of the PMDI composi-tion on the insertion loss has been investigated.

With the formulation based on VORANOL CP 6001, re-placing SPECFLEX NS 540 by Isocyanate C or IsocyanateD produces a significant increase in the insertion loss in the1000 to 1600 Hz bands. As opposed to SPECFLEX NS540, insertion losses of these experimental isocyanateslevel off above 1600 Hz. With this formulation, the acous-

Figure 10: Petite Cabine IL of foams based on Polyol D

tic performance of the foam produced with the two experi-mental isocyanates is not significantly different. Reducingthe functionality of the polyol formulation also reduces thedifference between the tested isocyanates and no significanttrend can be observed. With the Polyol D based formula-tions, a trend similar to the one observed with VORANOLCP 6001 based formulations can be observed but the do-main with improved IL is extended toward the lower fre-quency range. Figure 10 shows typical results observedwhen replacing SPECFLEX NS 540 by the experimentalisocyanates.

To conclude, polyurethane formulations can be adjustedin many different ways to produce foam having the desiredproperties. Some of the formulation variables have beenpresented and analyzed with the aim of evaluating theireffects on the acoustic performance of the final foam. Thedata presented show that by carefully choosing the polyoland isocyanate composition we can selectively influencethe acoustic performance of the foam. This is a first step indesigning an acoustic package providing the desired caracoustic signature.

C) EMISSION OF VOLATILE CHEMICALS

After the polyurethane foam industry successfully elimi-nated the CFC blowing agents from seating formulationsworldwide during the late 80’s, it embarked in a thoroughprogram to reduce and even eliminate all volatile chemicalscoming from the PU foams. Today new low fogging surfac-tants are widely used [22]. Most additives, flame retar-

Table 7: Machine-made foams with conventional a(foam density: 45 kg/m3, dem

Type of polyolVORANOL CP-6001 + VORANOL CP-1421 phpPolyol EPolyol FWaterCrosslinkerConventional amine catalystsReactive amine catalystsSurfactantIsocyanate ACream time sGel time sRise time sFree rise density kg/m3

Crushing Force N50 % Hot IFD NPVC staining test∆ L (black)∆ a (red)∆ b (yellow)∆ E

dants, antioxydants, are designed not to migrate or leach.The next challenge being addressed at the beginning of thisnew millenium is the elimination of amine catalyst vapors.

The reduction of the emissions coming from the aminecatalysts has been addressed by the catalyst suppliers whohave developed non-fugitive alternatives to conventionalcatalyst systems [23, 24]. These studies have confirmedthat it is possible to produce foams with amine catalystscontaining isocyanate-reactive hydrogens. However severallimitations have been identified: to obtain the desired reac-tivity profile, relatively high levels of such catalysts mustbe used and it has been shown that this is detrimental to thePVC staining resistance of such systems. As they are not asselective as the conventional catalysts, it is more difficult tocontrol the blow/gel reaction balance. It has also beenfound that with some reactive catalyst systems, the formu-lated polyol blends tend to lose their reactivity over time,especially their curing ability. Another way to eliminate theemissions coming from the amine catalysts is to attach thecatalytic site to one of the major components of the foam.This approach ensures that the catalyst is effectively incor-porated in the polymer structure. This has been achieved byreacting the catalyst with a trifunctional polyether polyol[25]. Another solution is to use a molecule containing adialkylamine group separated from the starter unit by a(CH2)n- group where n has to be equal to or greater thanthree [26]. These two approaches stress that to be active thecatalytical site should not be sterically hindered by thepolyether chain.

mine catalysts and new autocatalytic polyolsold time 3.5 minutes)

Conventional Polyol E Polyol F100 50 90

- 50 -- - 10

4.3 4.3 4.30.45 0.45 0.450.4 - -0.3 0.2 0.30.5 0.5 0.5

66.9 66.9 66.99 11 10

73 88 66120 95 8331 38 30

610 245 760180 240 285

- 17.7 + 0.3 - 0.5+ 9.4 - 0.6 - 0.9+ 4.5 + 2.6 + 6.520.5 2.6 6.6

At Dow we found a new route to produce polyols withcatalytic activity having various functionalities, equivalentweight and primary hydroxyl levels. Table 7 illustratestypical foam formulations based on some of these newpolyols and demonstrates that very good foam processing,including foam curing, is obtained with the combination ofthese new polyols and relatively low amounts of reactivecatalysts. Hence low PVC staining results are obtained. Inopposition to polyol formulations based solely on reactiveamine catalysts, these formulations show very good stabil-ity with time in terms of foam curing. No finger markingwas observed at demold on foams made from a three-week-old formulated polyol blend. Under the same conditions, aformulated polyol blend with the same reactivity profile,based only on some specific reactive amine catalysts, al-ready shows marking after one week’s storage. Data fromTable 8 illustrate that foams made with a combination ofthese new polyols and a small amount of new reactiveamine catalysts give no PVC staining while their physicalproperties are comparable to those obtained from foamsprepared with a high level of similar reactive catalysts and

Table 8: Comparison of physical properties of foams made(isocyanate blend TDI/MD

Foam typeReactive catalyst package A AAmine level php 0.95 0.Polyol E -Isocyanate index 90 10Core density kg/m3 58.0 60CFD (40%) kPa 3.0 5Tensile Strength kPa 98 12Elongation % 110 11Tear strength N/m 147 18Foam Resilience % 60 6Airflow cfm 2.2 1Dry C-Set (50%) %CT 5.2 3Heat agingTensile Strength kPa 98 13Elongation % 117 10CFD change % - 0.8 0Humid agingTensile Strength kPa 88 12Elongation % 177 16CFD change 1st % - 23.6 - 1CFD change 2nd % - 35.8 - 3CFD change 3rd % 12.2 16Humid Aged C-Set (50%) %CT 10.8 8PVC staining∆ L (black) - 2.8 - 1∆ a (red) + 1.0 + 0∆ b (yellow) + 8.4 + 6∆ E 8.9 7

conventional polyols. They also indicate that PVC stainingtends to improve with increased isocyanate index. The pre-sent study confirms that the autocatalytic polyols developedby Dow for both seating and NVH foams complement thereactive catalysts recently introduced on the market sincethey address some of their weaknesses. Indeed, by combin-ing these new catalysts with Dow’s autocatalytic polyols itis now possible to get good foam processing with almost noPVC staining.

CONCLUSION

To address the ongoing and newer challenges of automo-tive passenger compartment comfort, new polyurethaneraw materials are needed. As it has been shown in this pa-per, both polyols and isocyanates have to be tuned to meetthe requirements of density reduction, elimination of vola-tiles and improved durability. The Dow Chemical Com-pany is actively working with the industry to respond tothese requirements.

with reactive amine catalysts and autocatalytic polyolI, VW test methods)

A B B B95 0.95 0.5 0.5 0.5- - 54 54 540 105 90 100 105.1 62.6 54.1 57.5 56.1

.2 8.0 4.1 5.7 6.79 146 78 125 1310 105 109 108 947 224 158 212 217

4 61 63 63 60.7 2.1 2.7 2.4 2.3.8 3.8 3.7 3.9 4.1

6 177 77 142 1278 106 115 109 106

.8 - 2.2 - 6.0 - 5.7 -5.0

0 150 84 132 1395 157 184 153 151

4.7 - 14.4 - 30.2 - 24.4 - 23.01.0 - 31.7 - 39.8 - 37.8 - 37.5.4 17.3 9.7 13.4 14.5

.4 8.8 6.4 7.2 7.4

.9 - 0.8 + 0.35 + 0.27 + 0.19.3 + 0.8 - 0.53 - 0.23 + 0.09.9 + 5.6 + 1.9 + 0.92 + 0.07

.1 5.7 2.0 0.98 0.22

ACKNOWLEDGEMENTS

The authors wish to thank their many Dow colleagueswho had a role in generating the data presented in this pa-per, in particular Geert Builtinck, Jean Courtial, Herculesde Freitas, Jacques Labourier and Philippe Osche. Theauthors would also like to acknowledge Ron Herrington forhis help in the preparation of this paper.

* SPECFLEX and VORANOL are Trademarks of TheDow Chemical Company

REFERENCES

1. Herrington, R., De Genova, R., Casati, F. and M.Brown (1997), “Molded Foams” in Flexible Polyure-thane Foams; Herrington, R. and K. Hock, eds., TheDow Chemical Company, Form No. 109-01061, 11.1-11.41.

2. Casati, F.M., Herrington, R.M., Broos, R. and Y. Mi-yazaki (1997), “Tailoring the Performance of MoldedFlexible Polyurethane Foams for Car Seats”, Proceed-ings of the 1997 SPI Polyurethanes World Congress;Technomic: Lancaster, Pa., 402-420.

3. Muster D. and R. Plunkett (1960), “Isolation of Vibra-tions”, Ch. 18 in “Noise Reduction”; L. Beranek Ed.,McGraw-Hill Book Company.

4. Hilyard N.C., Lee W.L. and A. Cunningham (1991),“Energy Dissipation in Polyurethane Cushion Foamsand its Role in Dynamic Ride Comfort”, CellularPolymers (Conference), Forum Hotel, London UK, 20-22 March 1991; RAPRA Technology Ltd, 187-191.

5. Hirabayashi T., McCaa D.J., Rebandt R.G., Rush P.and P. Saha (1999), “Automotive Noise and VibrationControl Treatments”, Sound and Vibration, April, 22-32.

6. Wentzel R.E.and E.R. Green (1997), “Assessing Head-liner and Roof Assembly Acoustics”, 97126, SAE In-ternational Proceedings of the 1997 Noise andVibration Conference, Traverse City, MI.

7. Chen P. and G. Ebbitt (1998), “Noise Absorption ofAutomotive Seats”, 980659, SAE International Con-gress & Exposition, Detroit, Michigan, February 23-26, 117-121.

8. Lauriks W. (1994), “Acoustic Characteristics of LowDensity Foams” Ch 10 in “Low Density Cellular Plas-tics, Physical Basis of Behavior”, Hilyard N.C. and A.Cunningham Eds., Chapman and Hall, London.

9. Voronina N. (1999), “Comparison between Theoreti-cal and Empirical Models for Elastic Porous Materi-als”, Applied Acoustics, 58, 255-260.

10. Tang W.C., Zheng H. and C. F. Ng (1998), “Low Fre-quency Sound Transmission Through Close-fitting Fi-nite Sandwich Panels”, Applied Acoustics, 55 (1), 13-30.

11. Ling M.K., Hilyard N.C. and A. Cuningham (1991),“Fundamental Studies of the Acoustic Behavior ofFoam Backed Floor Covering”, Cellular Polymers(Conference), Forum Hotel, London UK, 20-22 March1991, RAPRA Technology Ltd.

12. Casati F.M., Herrington R.M., Broos R., Miyazaki Y.,Phan Thanh H. and D. Cadolle (1998), “AutomotiveSeating Foams: Improving their Comfort and Durabil-ity Performance”, Proceedings of the PolyurethanesEXPO’98, Technomic: Lancaster, Pa., 417-431.

13. Noakes C., Brown M.J., Duff A.W and S. Pique(1994), “Cost-effective Flexible Molded Foams forCar Interiors”, Conference Book of Papers Utech 94,Automotive, Paper 4.

14. Khameneh K.N., Cosman J.P. and G.D. Hunter (1994),“A New Generation of High Productivity Polyols forHR Molded Foams”, Proceedings of the 1994 SPIPolyurethanes World Congress, Technomic: Lancas-ter, Pa., 253-260.

15. McEvoy J., Thakurta K., Pique S and J. Bladon(1996), “Extra-High Resilience, TDI-based, Polyure-thane Flexible Foams for High-Comfort, High-Durability Automotive Seating”, Proceedings of the1996 SPI Polyurethanes World Congress; Technomic:Lancaster, Pa., 504-511.

16. Berthevas P., Phan Thanh H., Pique S. and Y. Mi-yasaki (1999), “Recent Advances in Flexible MoldedPU Foams for Car Seats”, UTECH Asia ’99 Confer-ence Book of Papers, Automotive, Paper 1.

17. Brown M.J. (1992), “High sag Factor Foams BoostComfort and Performance of Car Seats”, Proceedingsof the SPI-34th Annual Technical/Marketing Confer-ence, Technomic: Lancaster, Pa., 463-467.

18. Thoen J., Elwell R., Sewell R., Pellacani L., PedroniL. and G. Bergianti (1991),“Replacement of Chloro-fluorocarbons in all MDI Automotive Seating Foams”,Proceedings of the Polyurethanes World Congress1991, Technomic: Lancaster, Pa.,100-107

19. Casagrande G. and S. Fregni (2000), “Superior Proc-essing New all-MDI Solutions for Automotive Seat-ing”, UTECH 2000, Conference Book of Papers,Innovations: Automotive Developments, Paper 3.

20. Vicari C. and M. Onorato (1999), “A BE Model forthe Analysis of the Effects of Seats in the PassengerCompartment Acoustic Behaviour”, 1999-01-1790,SAE Proceedings of the 1999 Noise and VibrationConference, Traverse City, MI.

21. Berthevas P. (1997), “Flexible Molded PU Foams areincreasingly being used as the preferred solution forsound insulation in Cars, 971264, Proceedings of theNinth International Pacific Conference on AutomotiveEngineering.

22. Klietsch J. and A. Weier (1999), ‘New Low FoggingHR-surfactants Combining Environmental Concernsand added Performance’, Proceedings of the Polyure-thanes EXPO’99, Technomic: Lancaster, Pa., 95-101.

23. Mercando L.A, Kniss J.G., Tobias J.D., Plana A., Lis-temann M.L. and S. Wendel (1999), “Non-fugitiveCatalysts for Flexible Polyurethane Foams”, Proceed-ings of the Polyurethanes EXPO’99, Technomic: Lan-caster, Pa., 103-134.

24. Sikorski M., Wehman C. and H. Cordelair (1999), “New Additive Solutions for Low VOC in HR MoldedFoams”, Proceedings of the Polyurethanes EXPO’99,Technomic: Lancaster, Pa., 135-143.

25. Parks K.L., Gansen P., Gricar J. R. and U. Ulman(1996), “New Developments in Polyurethane Foamsystems for automotive Instrument Panels, Proceed-ings of the Polyurethanes EXPO’96, Technomic: Lan-caster, Pa., 362-369

26. Larbig H., Hinz W. and P. Horn (1995), “ Catalyticallyactive Polyether Polyols”, Proceedings of the Polyure-thanes Conference 1995, Technomic: Lancaster, Pa.,212-215

BIOGRAPHIES

René Broos

René Broos studied chemistry at theUniversities of Antwerp and Gent,Belgium. He joined Dow inPolyurethane R&D in 1980 and hasbeen involved in research and devel-opment projects for shoe-soleelastomers, construction and seatingfoams. Today he is responsible forMaterial Science in the LeveragedProduct Development Group at Dow

Europe’s Polyurethane Research Laboratories in Terneu-zen, The Netherlands. His research is related to polymermorphology and foam structure-property characterization.Dr. Broos holds a Ph. D. in Organic Chemistry from theUniversity of Gent, Belgium.

Jean-Marie Sonney

Dr Jean-Marie Sonney graduated fromthe Swiss Federal Institute ofTechnology in Lausanne (Switzerland)and received his Ph.D. degree inphysical organic chemistry from thesame institution in 1979. After a post-doctoral research fellowship at theUniversity of California, Santa Cruz,he joined the Geneva-based Researchand Development group of BP

Chemicals in 1981 and was transferred to the Dow Chemi-cal Company in 1989. During these years, he had variousresponsibilities in the field of Analytical Chemistry, Qual-ity Assurance, EH&S and Quality Management. He is cur-rently a project engineer in the Development Group for

Molded Foams and Specialties with responsibilities forNVH applications.

Huy Phan Thanh

Huy Phan Thanh graduated from theUniversity of Geneva (Switzerland)and received his Ph.D. degree inOrganic Chemistry from the sameUniversity in 1981. From 1981 to1989, he worked in PolyurethaneR&D for BP Chemicals then for DowEurope S.A since 1989. He has beenresponsible for TS&D in the field offlexible molded polyurethane foams

and is currently project chemist in the Molded Foams andSpecialties group at the Dow Europe International Devel-opment Center in Meyrin (Geneva).

François M. Casati

François M. Casati graduated fromICPI (F), now CPE-Lyon (EcoleSuperieure de Chimie PhysiqueElectronique de Lyon), in 1967. Hehas over 30 years of experience inPolyurethanes, Industrial Amines andBiocides, with S.N.P.E., Recticel,Abbott Laboratories, BP Chemicalsand The Dow Chemical Company.During that time he has held various

positions in Manufacturing, Marketing and Research &Development. He is currently working as a DevelopmentAssociate for the Flexible Molded Foam business of Dow.