You are on page 1of 11

Analysis of a Rubber Oil Pan Gasket RTV Pad Using ABAQUS Explicit

Milton A. DeHerrera, Ph.D


Computational Mechanics Group* Freudenberg-NOK General Partnership 50 Ammon Drive Manchester, NH 03103 USA
*

Current affiliation: Edwards Lifesciences LLC One Edwards Way Irvine, CA 92614

Abstract: Automotive Powertrain sealing components like gaskets are required to maintain

sufficient contact pressure to prevent leakage of fluids to the engine exterior and to keep tensile strains small enough to avoid gasket failure. An existing Oil Pan gasket RTV pad design was analyzed to numerically confirm observed failures that resulted in gasket disintegration and oil leakage. Because of the relatively complex 3D T-Joint geometry and subsequent size of the FEA model, ABAQUS/Explicit was used to analyze the model. The results of the analysis, together with observations from engine tear-downs, are presented in this paper. 1. Introduction
In a previous paper [DeHerrera, Heim, 2000] we presented the rationale behind using ABAQUS/Explicit to solve difficult sealing problems involving confined elastomers. Space limitations and the introductory nature of that paper prevented us from discussing the most useful application of ABAQUS/Explicit in our business: The 3D analysis of Powertrain sealing components.

2002 ABAQUS Users Conference

As stated in [DeHerrera, Heim, 2000], the principal objectives of sealing components in Powertrain applications are to (1) prevent the escape of oil and other fluids to the exterior of the engine, (2) maintain internal pressure during operating conditions, (3) provide sufficient resistance to the engine environment so that failure does not occur within a specified amount of time and/or mileage and (4) meet all of the above conditions at a reasonable cost. Contact pressure distribution is the most important parameter in designing sealing components, closely followed by the maximum tensile strain. 1.1 Phenomenological Strain Energy Model

Because of previous FEA work with similar components, we have used the two-term Mooney-Rivlin phenomenological strain energy law to describe the strain energy for both elastomers used in this study: U = C10(I1 - 3) + C01(I2 - 3) + (1/D1)(Jel - 1)2 (See Abaqus Theory Manual for more details). Most of the elastomers we currently use at Freudenberg-NOK have D1 values in the order of 0.01 (MPa)-1. There are occasions when other users wish to know the value of the Poisson ratio in order to gage how near the material lies to the incompressible limit of = . From the formula relating the initial bulk modulus K0 and initial shear modulus 0 to the Poisson ratio : = (3K0 - 2)/(3K0 + 0) and the relationship between K0, 0 and the Mooney-Rivlin parameters C10 and C01 : 0 = 2(C10 + C01) and K0 = 2/D1 Then the Poisson ratio can be written as: = (3 - 2 D1[C10 + C01])/(3 + D1[C10 + C01]) The D1 value for both elastomers used in this study is 0.497. While this value is close to the incompressible limit, it is sufficiently far from 0.5 to allow the use of ABAQUS/Explicit.

2. Problem Statement
The goals of this analysis were threefold: (1) To evaluate the design of a competitors single-beaded carrier oil pan gasket RTV pad that was being used in a current production engine, (2) to see if the FEA would provide supporting evidence for observed leaking at the gasket RTV pad and (3) to compare the results obtained (for the same geometry) between the C3D4 and C3D10M tetrahedral elements. The gasket RTV pad was modeled using both C3D4 and C3D10M elements, and the front cover gasket protrusion was modeled with C3D6s.

2002 ABAQUS Users Conference

The Mooney-Rivlin constants for the HNBR material used to model the Oil Pan gasket RTV pad are: C10 = 0.35 MPa C01 = 0.06 MPa D1 = 0.014 (MPa) -1 And for the 60 durometer material used to model the protrusion they are C10 = 0.42 MPa C01 = 0.08 MPa D1 = 0.011 (MPa) 1 The gasket RTV pad was compressed to the height of the metal carrier, or 2.54mm. The pad was compressed by two rigid surfaces (representing the girdle and front cover setup of the engine) into a supporting rigid surface representing the pan rail. The front cover gasket protrusion was modeled as a rubber block that extends 1.0 mm past the surface of the front cover. A vertical step of 0.35mm was also included in the model. The front cover block with the protrusion was shifted 0.35mm away from the pad. Figures 1 and 2 represent the 3D models used for each analysis. Figure 3 illustrates the deformable elements that represent the T-Joint being analyzed. 2.1 ABAQUS Explicit Model Setup and Run

The gasket geometry was initially generated from a partially defeatured Unigraphics version 16 production CAD model. The model was imported into MSC/PATRAN where its mesh and that of the adjacent rigid bodies was generated and exported. There were two deformable bodies in this model: The gasket RTV pad was modeled with 41,484 C3D4/C3D10M tetrahedral elements with 25,474 and 60,990 nodes respectively, and the Front Cover gasket protrusion by 492 C3D6 wedge elements with 385 nodes. There were a total of 7 contact pairs, out of which two involved deformable-to-deformable contact and the rest deformable-to-rigid contact. The problem was initially run under ABAQUS Explicit version 5.8-18 and subsequently rerun with ABAQUS Explicit version 6.2-1. We shall only discuss the results from version 6.2-1. Both problems were run on an HP J5600 2-CPU computer. The C3D4 problem ran in 4 hours and 59 minutes of wall clock time, and the C3D10M problem ran in 25 hours and 40 minutes of wall clock time.

3. Results
Figures 4-9 show results for field quantities such as maximum true tensile strain and contact pressure for both the C3D4 and C3D10M runs. We have used the mass-scaling feature of ABAQUS Explicit to run the problem quasi-statically and in much less time than would be required for an unscaled problem. Figures 10, 11 show a plot of the internal, kinetic and artificial energies for the C3D4 and C3D10M models, respectively. A cursory inspection of these figures shows that the problems indeed ran in a quasi-static fashion. 2002 ABAQUS Users Conference 3

Figures 12-13 show scanned pictures from an actual warranty return. These pictures show the deformed gasket RTV pad and can be used to correlate FEA predictions of pad deformations to an actual deformed pad. It can be seen that the deformed shape predicted from the C3D10M run is quite similar to that of the gasket from a warranty return. The most important observation is that the FEA predicts a troublesome low contact pressure region on the RTV pad that seems to confirm the leaking of oil that had been noticed in this particular engine.

4. Conclusions
ABAQUS Explicit can be used to analyze complex 3D elastomeric problems, preferably with the use of higher order tetrahedral elements. Whereas contact pressure distribution is not greatly affected by the choice of C3D4 versus C3D10M, that does not seem to be the case for other field quantities.

5. Acknowledgements
I wish to thank Ms. Katie Olsen for providing the gasket pictures from the warranty return and other technical assistance.

6. References
1. DeHerrera, M. A, and D.R. Heim Using ABAQUS Explicit to Model Behavior of Elastomeric Sealing Components, Proceedings of the ABAQUS Users Conference at Newport, RI, May-June 2000. 2. Abaqus Theory Manual, version 6.2, Hibbitt, Karlsson & Sorensen, Inc., 2001.

2002 ABAQUS Users Conference

Figure 1. Isometric view of FEA model (Girdle not shown)

Figure 2. Side view of FEA model.

2002 ABAQUS Users Conference

Figure 3. Undeformed model of RTV Pad/Front Cover gasket protrusion T-Joint

2002 ABAQUS Users Conference

Figure 4. Deformed shape of gasket RTV pad modeled with C3D4 elements after a 2.23mm compression to a 2.54mm carrier height

Figure 5. Deformed shape of gasket RTV pad modeled with C3D10M elements after a 2.23mm compression to a 2.54mm carrier height

2002 ABAQUS Users Conference

Figure 6. Maximum tensile true strain (mm/mm) in gasket RTV pad modeled with C3D4 elements after a 2.23mm compression to a 2.54mm carrier height

Figure 7.

Maximum tensile true strain (mm/mm) in gasket RTV pad modeled with C3D10M elements after a 2.23mm compression to a 2.54mm carrier height

2002 ABAQUS Users Conference

Figure 8. Maximum contact pressure distribution in gasket RTV pad modeled with C3D4 elements after a 2.33mm compression to a 2.54mm carrier height

Figure 9. Maximum contact pressure distribution in gasket RTV pad modeled with C3D10M elements after a 2.33mm compression to a 2.54mm carrier height

2002 ABAQUS Users Conference

Figure 10. Energy curves for gasket RTV pad modeled with C3D4 elements after a 2.33mm compression to a 2.54mm carrier height

Figure 11. Maximum contact pressure distribution in gasket RTV pad modeled with C3D10M elements after a 2.33mm compression to a 2.54mm carrier height 2002 ABAQUS Users Conference

10

Figure 12. Gasket RTV pad taken from a warranty return

Figure 13. Gasket RTV pad taken from a warranty return

2002 ABAQUS Users Conference

11

You might also like