HOW TPV OUT-PERFORMS EPDM IN ACOUSTIC PROPERTIES FOR AUTOMOTIVE SEALING

James T. Browell, Jyco Sealing Technologies, Dexter, MI Shawn Jyawook, Jyco Sealing Technologies, Dexter, MI

Abstract

A comparative analysis of TPV vs. EPDM with respect to noise reduction is examined for automotive weatherstrip body sealing applications. Specifically, EPDM sponge is compared with JyFlexTM, a TPV compound of equivalent stiffness. The study is performed using multiple acoustic tests (road and component), supported by FEA analysis as a diagnostic tool [1].

Introduction

Primary and Auxiliary Body Seals are used to prevent water, air and noise impingement into the passenger compartments of automobiles. While the seal design is critical to prevent leaks and aspiration, the material itself can influence noise transmission. The automotive industry currently uses EPDM sponge for body sealing because of its pliability, thereby being able to accommodate build variations [2]. JyFlexTM TPV’s can replace EPDM sponge and increase acoustic performance, while maintaining equivalent sealing characteristics at a reduced cost.

It has been shown, using a combination of testing and finite element analysis, that primary seal acoustic performance is stiffness controlled at low frequencies (< 1000 Hz) and mass controlled at higher frequencies (>2000 Hz) [2]. JyFlexTM TPV’s are typically two to three times denser than EPDM sponges. Thus for similar sections, the TPV bulbs should have better acoustic performance than an EPDM bulb of equivalent cross-section.

Past studies indicate that primary seal acoustic performance is controlled by mass at higher frequencies [2] and the frequencies typically associated with wind noise [3] – 1000 to 5000 Hz. Because JyFlexTM TPV is a denser material than typical production EPDM sponges, it should follow from mass law [4] that insertion loss characteristics are better for similar cross-sections in similar environments. Subsequently, a series of studies show the extent of the acoustic advantage using higher density material for weatherstrip primary sealing applications in automobiles. Noise Transmission Through Primary Seals

There are many noise sources that can penetrate into the cabin of an automobile, but only a few that penetrate through the primary body seals. Sources of noise transmission through primary body seals are:

1. Wind noise, 2. Far field noise, and 3. Aeroacoustic resonances.

Other noise sources, such as the engine, tires,

structural borne vibrations, and interior cabin resonances can dominate. However, at higher speeds wind noise tends to become the major contributor. Noise impinges through the body seals around the sealing patch (in the worst case, aspiration), between the carrier and flange (for flange mounted seals), and by way of direct noise transmission [Figure 1]. Direct noise transmission can occur through the seal walls and be accentuated by cavity size and vibration characteristics of the seal [2].

Most current production automobiles use triple sealed doors, which comprise of a continuous body mounted primary seal, a continuous door mounted primary seal and a series of segmented auxiliary seals around each door. Observations at the 2006 Detroit Auto Show indicate that double sealed door apertures are only currently being used by U.S. manufactured trucks. The body and door mounted primary seals act as a double barrier to noise transmission and aspiration, while the auxiliary seals are for air and water management.

The relative influence of primary sealing on vehicle interior noise varies from vehicle platform to vehicle platform. Noise through primary seals may not be significant if other sources such as engine noise, window or structural borne vibrations dominate. In fact, even if wind noise through the primary seals is dominant at high speeds, when the noise is insulated, other noises may become audible [see Figure 2]. Thus, care must be taken when choosing a vehicle to compare sealing designs and materials.

Initial Indicators of Increased Acoustic Performance

Page 1 of 7

During the development cycle of several EPDM to TPV conversion seal designs, various acoustic tests are performed in an attempt to verify the advantage of TPV. To date, all indicate that JyFlexTM TPV seals are at least as good at insulating sound as EPDM sponge, while most indicate TPV is better.

To compare the relative acoustic performance of the TPV and EPDM bulb design seals, tests measuring insertion loss (IL) [3] and road tests measuring sound pressure level (SPL) are performed. The IL test set-up is shown in Figure 3 [5] and performed at the Kolano and Saha Engineers, Inc. test chamber in Waterford, MI. If the installed bulb shapes are essentially the same, the results can be considered a direct comparison of the materials.

Initial studies measure insertion loss (IL) on the component level for TPV and EPDM decklid (trunk) seals of the same design [Figure 4] [5]. Results indicate the TPV seal perform slightly better. Although the part designs are intended to be identical, there is a manufacturing variation between the two. This variation results in the TPV part deforming differently than the EPDM part. When the TPV part is forced to deform the same as the EPDM, the results indicate the TPV seal perform even better than the EPDM section.

The initial study shows a comparison of the bulb seals with seal deflections of 5 mm. Both deflected shapes result in a sealing width (“wet-out”) of 5 mm. The difference in IL is significant between the EPDM and TPV seals as shown in Figure 4. Even when the EPDM seal is crushed to a “wet-out” of 10 mm, the IL of the TPV seal is still better overall. Furthermore, the results of the IL test are consistent with those of previous findings [2] – indicating that seal stiffness controls the acoustic performance at low frequencies (<1,000 Hz), while mass controls the frequencies at high frequencies (>2,000 Hz).

Another component study of Upper Auxiliary weather seals indicates that the insertion loss characteristics of TPV outperform EPDM [Figure 5] [6]. At one frequency, the EPDM shows better performance. However, this frequency is coincident with the length top cavity, indicating the TPV part did not close out completely. This is a function of the geometry, not the material. The seal shape was corrected later on during development.

Recently, TPV replaced EPDM sponge in Dodge

Ram truck production vehicles. Internal road and far field noise tests were performed comparing EPDM and TPV seals. The results of these tests are only able to indicate that the TPV seal performs at least as well as the EPDM seal. In the case of the Dodge Ram, other noise paths are dominant and override any audible advantages of the TPV Primary Seal.

Further internal studies are shown here of the 2006

Cadillac Seville primary body seal for road and far field noise [7]. In this case, the TPV performs better [Figure 6]. However, the test results are at only one speed (70 mph).

Component and limited road noise studies show a consistent trend, but carrier and geometry effects cloud the results. A more comprehensive study of a simple primary seal design is required for quantitative results.

A Comprehensive Acoustic Study

A comprehensive study of TPV and EPDM body mounted primary weatherseals with identical cross-sections was completed to show the acoustic performance advantage of low stiffness JyFlexTM TPV. The intent was to perform component and road tests of EPDM and TPV door mounted primary seals with identical installed cross-sections an automobile at high steady state speeds. Unfortunately, the only TPV sections available at the time of the road tests were undersized prototypes. Therefore, the road tests were performed with the intent of validating the component testing and using that validation as a relative measure of improvement with a second set of accurately dimensioned seals.

A 2006 Volkswagon Passat was used as the test

vehicle. The car was a four-door sedan with a 3.6 L, six-cylinder engine, front-wheel drive with automatic transmission and radial tires. Tests were conducted under four different door seal/body seal conditions as follows: • EPDM Door Mounted Seal / EPDM Body Mounted

Seal • TPV Door Mounted Seal/EPDM Body Mounted Seal • TPV Door Mounted Seal/No Body Mounted Seal • No Door Mounted Seal/No Body Mounted Seal

Each set of seals were installed and compressed at least four times each and left overnight with the doors closed before each test to allow for the effects of material set. The vehicle was tested on the smooth asphalt surface of Milan Dragway in Milan, Michigan at constant speeds of 80, 120 and 160 kmph (50, 75, and 100 mph, respectively). For each test run, sound measurements were made at four microphones placed inside the vehicle at the front right passenger center-of head, rear right door top of B-Pillar, rear right door beltline, and rear right bottom of door. Four each operating condition four runs were made each along the same stretched of road for 8 seconds. All data was recorded using a computer-based data acquisition system located in the trunk of the vehicle. 1/3 octave band spectrum analyses and various single number metric were computed [3].

Page 2 of 7

The raw road data indicates the TPV sections are around 2-3 dB worse within the wind noise frequency range (1 – 5 kHz) [Figure 7]. From the road test measurements, in addition to the 1/3-octave band spectrum analysis, various single number metrics are computed to get a better understanding of the test results. When the road test results are broken down into single number metrics, both seals have essentially the same acoustic response [Tables 1, 2, and 3] [3] - less than 1 dB difference is considered subjectively indistinguishable. Thus, even for an undersized TPV section the overall noise perception remains unchanged. The metrics used are defined as follows [3]:

A-weighted SPL is a better representative of how we hear than linear SPL. We do not hear low and high frequencies well, so weighting networks are often used to better represent the linear SPL that is measured to the SPL that humans actually hear. Differences of 2 -3 dB are perceptible by most people. Lower is better.

Loudness is a quantitative measure of the subjectiveresponse to sound. Sones are exponential – 2 sones are twice as loud as 1 sone; 4 sones are twice as loud as 2 sones. Phons are expressed as decibels – 20 phons are twice as loud as 10 phons; 30 phons are twice as loud as 20 phons. A difference of 2 – 3 phons is perceptible by most people. Lower is better.

Articulation Index is a measure of a person’s ability to distinguish speech in the presence of “noise”. A 4 percentage point difference is perceptible by most people, i.e. most people can distinguish the difference between 92% and 96%. Higher is better.

A FEA study of the digitized TPV section indicates

that the wall thickness of the TPV seal has minimum wall thickness directly in the noise transfer path. This results in less noise reduction based on mass, exposed area, and deformed geometry [Figure 8] [8].

The component IL test indicates the same relative

difference between the EPDM and TPV sections [Figure 9]. A fixture mimicking the seal environment was used in the Kolano and Saha Engineers, Inc. test chamber [Figure 3] to compare the insertion loss (IL) characteristics of the seals on a component level. Seal lengths of 400 mm were used with the ends securely sealed with clay. The effects of holes were included consistently between specimens. Outputs were measured from 25 to 10,000 Hz and presented in 1/3 octaves.

A second set of TPV seals were produced to be

identical to the EPDM when installed. Additional finite element analysis (FEA) indicate the second series of

digitized TPV seals are essentially identical in the assembled position as the EPDM production seals [Figure 10], having the same wall thickness in the direct noise transmission path. The deformed shapes themselves agree with external observations of the IL test fixture with the seals installed.

Component IL testing on second set of TPV seals [Figure 11] indicates better noise reduction performance for the TPV seal. The results indicate the improvement will be 4 to 6 dB, and even correlates well with mass law: The density of JyFlexTM TPV is ρTPV = 9.2e-10 Mg/mm3,while the density of the 2006 VW Passat production door mounted primary EPDM sponge is measured to be ρEPDM = 5.6e-10 Mg/mm3. Thus, ∆IL = 20 log (ρTPV/ ρEPDM) = 20 log {(9.2e-10 Mg/mm3)/ (5.6e-10 Mg/mm3)} = 4.3 dB [4]

Conclusions

The results of the road testing was consistent for all three speeds, indicating the production EPDM seal has an approximately 2 to 6 dB better un-weighted acoustic performance than the thin TPV seal from 1000 to 10000 Hz. Test data that is typical of that taken [Figure 7] shows similar trends to the component insertion loss test data [Figure 11]. It is expected that road tests performed with the corrected sized TPV seals would show the same relative improvement to the IL test data as well. However, the magnitude of this difference, especially in the 1000 to 2,000 Hz range, is likely to result in a noticeably subjective improvement in interior vehicle noise at high speeds for the VW Passat, barring other noise sources that emerge.

References

1. Browell, James, “The Ninth International Conference

on Thermoplastic Elastomers”, TPE 2006, Smithers Rapra Ltd., Munich, Germany, November 8-9, 2006.

2. Park, Junhong, Thomas Siegmund and Luc G. Mongeau (Purdue University), Society of Automotive Engineers, Paper No. 2001-01-1411, “Sound Transmission Through Elastomeric Sealing Systems”.

3. Saha, Pranob and Richard A. Kolano (Kolano and Saha Engineers, Inc.), Project No. 2006-170, “In-Vehicle On-Road Door Seal Noise Study on a Volkswagon Passat”.

4. Beranek, Leo L., Noise and Vibration Control,McGraw Hill, Inc. 1971, p.282.

5. Saha, Pranob and Richard A. Kolano (Kolano and Saha Engineers, Inc.), Society of Automotive

Page 3 of 7

Engineers, submitted paper, “Acoustical Performance Testing of Automotive Weatherseals.”

6. Saha, Pranob and Richard A. Kolano (Kolano and Saha Engineers, Inc.), Project No. 2003-084, “Results of Acoustical Performance Evaluation of Two Different Weatherseal Materials”.

7. Saha, Pranob and Richard A. Kolano (Kolano and Saha Engineers, Inc.), Project No. 2003-083, “Results of Acoustical Performance Evaluation of Two Different Weatherseal Materials”.

8. Browell, James T., CAE Project No. 2006.027-2, Jyco Sealing Technologies, Dexter, MI

Acknowledgements Special thanks goes to Ralph Richardson and Mark Steward of Jyco Sealing Technologies; also to John Kopko and Pranab Saha of K&S Engineers, Inc.

Figure 1. Noise Transmission Through Automotive Body Seals

Figure 2. Example of Noise Source Relative Influence

Page 4 of 7

Figure 3. Insertion-Loss Measurement Test Set-up

Figure 4. Insertion-Loss Comparison of a Decklid (Trunk) Seal Design

Figure 5. Insertion-Loss Comparison of an Upper Auxiliary Seal Design

Figure 6. Insertion-Loss Comparison of a Body Mounted Primary Seal Design

Figure 7. Typical Un-weighted Sound Pressure Level Data During the Road Test of the VW Passat

Page 5 of 7

Figure 8. Undersized 1st Run TPV Seal Compared to Production EPDM in Insertion Loss Test Set-up

Figure 9. Initial Insertion-Loss Comparison of the Door Mounted Primary Seals – EPDM and TPV

Figure 10. Overlay of 2nd Run TPV Seal on Production EPDM

Figure 11. Insertion-Loss Comparison of Door Mounted Primary Seals – TPV 1st Run, TPV 2nd Run and EPDM

Page 6 of 7

Table 1. SPL Summary at 80 kmph (50 mph)

Table 2. SPL Summary at 120 kmph (70 mph)

Table 3. SPL Summary at 160 kmph (100 mph)

Key Words: TPV, Insertion Loss.

Page 7 of 7