Analysis of an Automotive Sealing system By Vinod Jacob*, Abdul Rajak Shaik, Maruthi Phani Ram Kotti and Ramanath K S Engineering Services and Consultancy Practice, Infosys Technologies Ltd. Electronics City, Bangalore-561229, India. (Submitted for CAE 2001 – The Annual users’ conference conducted by CSM Software Pvt. Ltd. held between 7th and 9th November 2001) Abstract The automotive industry extensively uses elastomers as sealing systems in glass runner, shock absorbers, gaskets and several others. The challenges in the simulation of elastomers are non-linearity in material, geometry, boundary conditions and contact definition. The glass runners are the beadings around the door sash of an automobile. They ensure free movement of glass, prevention of leakage and vibration control. A FE simulation of a glass runner has been carried out using MSC.Marc software, taking into consideration the complex geometry of the elastomer and its non-linearities. The evaluation of displacements, contact forces and stresses help in assessing the suitability of the cross-section geometry of the glass runner and improving it thereon. Keywords: elastomer, non-linearity, MSC.Marc, glass runner, contact. Introduction In the market place of the 21st century, automobile industries face the competition for time to market. As the cost of essential design escalates exponentially with time, FEA tools are meant to play a vital role in speeding up the design process. FEA has become more of a necessity than an option. A finite element simulation has been done on the glassrunner to assess the suitability of the geometry of the same. A glassrun is attached to a doorframe of an automobile to provide a seal between an outer peripheral edge of a raised door window glass and the doorframe. The door glass moves through the glassrun. Each section of the glassrun includes a flexible metal insert that has been stamped and rolled into a configuration such that a cross-section of * the insert is a U-shaped section. An outer layer is extruded around the shape of the insert in a manner such that it consists of two adjacent U-shaped sections with a notch in between (Fig. 1). This outer layer defines a glassrun channel that accepts a vehicle window and a sealing configuration. The fins on one of the U-shaped section grips the metal carrier while as the inner and outer lips of the other U-shaped section seat the door glass. In order to allow the glass to slide through the lips without friction and with minimum drag, a layer of less frictionless material known as flock is sprayed on the surfaces where the glass is supposed to be in contact with the rubber part. A notch is provided between the two U-shaped glassrun channels. This notch helps in providing the flexibility of allowing the U–shaped channel to fit tightly within the metal carrier. A Author for Communication: vinod_duraiwin@infy.com portion of the elastomer is projected out at one end of the U–shaped section. This acts as hinge so that the metal carrier can secure the glassrun tightly. Moreover when the glass is inserted into the glassrun the outer lip of the glassrun puts additional pressure on the glass thus securing the glass tightly such that the glass simply doesn’t get released from the glassrun. In order to ensure that the inner lip is holding the glass, the supporting lip is provided with such a curve as to minimise it’s buckling and thus providing additional stiffness to the inner lip. Pivot until the glassrun metal carrier fitment is as expected. The specification of the metal carrier geometry is sufficient in MARC as the software also permits motion of the geometry as analytical curves. In actual practice, however it is the glassrun, which is pushed into metal carrier for fitment. We found this approach resemble from mathematical point of view. 2. Subsequently the window glass is inserted into the glassrun assembly. Since the window glass is considered very stiff compared to the elastomer, it is also defined as a rigid curve. Then a prescribed velocity motion was applied to get it inserted into the glassrun. This approach simulates the actual behaviour of glassrun movement closely. Glass Supporting Lip Fig. 1: 3D Model of the Glassrun Metal carrier Analysis Methodology The FEA software used is MSC. Marc with MSC.Mentat as pre and postprocessor. The Simulations to be achieved were the insertion of the glassrun into the metal carrier and then the fitting of window glass into the glassrun U-section. The Finite Element analysis for this simulation was achieved in two stages. 1. The glassrun is constraint isostatically in the metal insert location to suppress rigid body motion. The prescribed velocity is given to the metal carrier Notch Fig. 2: FE model of the Glassrun FE model Generation The IGES file of the CAD model is imported and meshed. The interference found at the CAD level is removed in FEM. All the three types of non-linearity, namely the material, boundary and geometric and also contact are present in the FE model. Since the length of the glassrun is considered very large compared to the cross section of the same, the problem is treated to be of “Plane Strain” type. Herrmann formulation with element types of 80 for quad and 155 for tria are used. The glassrun is an elastomer, which is a polymer, which shows a nonlinear elastic stress-strain behaviour. Analytical material models do not entirely describe the stress-strain relationship in a material under all loading conditions. Therefore, experimental data needs to be created such that a reasonable material model may be selected to define material behavior pertinent to the application of interest. Fig. 3: Experimental Data Curve Generally Ogden model is used for very large strain levels of 700% and above while as the Arruda-Boyce option has an applicable strain level of up to 300%. The 2-parameter Mooney-Rivlin option has an applicable strain of about 100% in tension and 30% in compression. Since the simulation doesn’t go for very high strain rate, Mooney-2 material model is chosen. The experimental data curve obtained from previously done research as shown in the diagram below (Engineering Stress vs. Engineering Strain) is fitted for uniaxial, planar shear and biaxial cases and the corresponding Mooney Rivlin constants are obtained to describe the behaviour of Mooney-2 (Fig. 3). First, the metal carrier is inserted into the glassrun and then the glass is placed into it. Since the motion of bodies is taking place on different bodies at different instant, the prescribed motion of bodies is defined as a plot showing the variation of velocity with time. Simulation is carried out in pseudo time domain as a quasistatic analysis. Hence for the initial 15sec a velocity is prescribed to the metal carrier with a target position and then for the next 15sec, the glass is inserted into the glassrun. In order to see that all distributed loads are formed on the basis of current geometry, follower force option is used. As said earlier, the metal insert is the one made fixed. Here isostatic boundary condition is applied. One leg of the metal insert is fully fixed; while as the other leg is made free to move in one direction so that it gives way to the metal carrier entering into the U-section. The type of contact between the various bodies is mentioned in the contact table. The table gives whether a contact is present between bodies and if then the type of contact namely, glue and touch. Between each contact body a coulomb friction is provided as shown in the table below. The various friction 1 2 3 1 Rubber T G G 2 Metal Insert G 3 Flock G 4 Glass 5 Metal-Support 4 5 T T T T T parameters are given in the table below. Table. 1: Contact Properties Table Body 1 Glass Metal-Support Body 2 Flock Rubber Friction 0.03 0.1 Table. 2: Contact Friction Table Analysis Since this problem involves large displacement, 500 Load steps are used for a period of 30sec. The number of recycles given were 20 per load step. Since sudden contact nonlinearities occur, it is felt that the gains obtained in assembly in modified Newton-Raphson can be nullified by increased number of iterations or nonconvergence. Hence full Newton– Raphson method is used. Since the analysis deals with elastomer, deviatoric stress option is used. This takes into effect the residual stress if any present. The convergence ratio of 0.1 is used for the force criteria. Results and Discussion All the results were well within the expectations. The Finite Element Simulation carried out brought out the following observations, 1) Contact force vs. Glass position: As shown in the plot from fig. 4, we can observe that as the glass moves into the rubber sealing the resultant contact force for the inner lip on the glass is increasing. This plot along with the contact force variation of the outer lip gives the power required for the motor to push the glass into the U-shaped section. 2) Deformation Pattern: From figs. 59 the deformation patterns at various increments of 100, 200, 300, 400 and 500 are shown. They are found to be in accordance with our expectations. 3) Pivoting action of Glassrun assembly: The pivot provided at bottom of the rubber sealing is to enhance the ability of the sealing system to grip the glass. The pivoting action is observed in fig 9. The contact forces on the glass by the outer lip will indicate the thrust put by the lips on the glass. This will be a measure of the air tightness of the glassrun and also the resistance of the glass to any sideward force. 4) Supporting inner lip: The very purpose of this supporting inner lip is to provide local strength to the inner lip of the sealing system. This sees to it that the inner lip doesn’t give way because of the contact forces on the door glass. This behaviour can be clearly noticed from the plot shown in fig 10. Here the resultant contact force of the inner lip suddenly shoots up when the supporting lip starts pushing the inner lip along with the force acted upon the glass by the outer lip. The contact forces on the supporting lip will give the force acting on the top of it. 5) Compression of Outer lip: From the simulation it is observed that the compression of the outer lip is around 5%. Changing the material of the flock could change this compression, which in turn would play in altering the contact forces on the glass. 6) Stress strain contour: The stress contours are given every 100 increments. It’s seen that the metal insert gets stressed a lot. This is because of the high stiffness of the metal insert when compared to the elastomer. Scope for further work Acknowledgement: The present analysis illustrates the applicability of MSC.Marc to simulate complex structural interaction of elastomers under large strains. With this basis, parametric analysis of glassrun geometry and material configurations can be carried out so that one can arrive at low weight and functionally efficient configurations. Further a 3D model simulation is being attempted to get a feel of the actual issues involved and also to finetune the results. A composite geometry option is available in MARC, which could be used to simplify the problem. In order to juggle with the various geometry options, this option could be used for large scale simulations and hence save the analysis time. The authors wish to acknowledge Mr. Raghavendra and Mr. Vijay Machigad for providing information regarding various glassrunner designs and configurations. The encouragement for in-house R&D from Mr. Ravishankar, Business Manager and Dr. M S S Prabhu, Sr Vice President of Infosys is greatly acknowledged. Conclusions The finite element simulation of a typical glassrun has been carried out. From the analysis, it is observed that the simulation was close to our expectations. The geometry of the pivot, hinging action of the notch and that of the supporting lip plays a vital role in the power requirement of the power window motor. In addition to it, the material of the flock plays a vital role in the contact force requirement of the glass. The analysis of elastomers can be effectively simulated using MSC.Marc software to provide engineering solutions to the industry. The Analysis Driven Design of a typical automotive component such as glassrun is as illustrated in this paper. References. 1. 2. 3. 4. 5. 6. MARC users manual MARC command manual Experimental Loading Conditions used to implement Hyperelastic and Plastic Material Models, Kurt Miller, Axel Products Inc. Testing Brief, Compression or Biaxial Extension, CompressionOrBiax-Rev-1, June 2000.Axel Products Inc. Finite Element Procedures, Klaus-Jurgen Bathe, Prentice-Hall India, 1997 Engineering Plasticity, W Johnson, P B Mellor, Ellis Horwood Limited, 1983. Resultant Contact Force On Glass. Resultant Contact Force vs. Resultant Glass position. 0.8 0.6 0.4 0.2 0 0 5 10 15 20 Glass Position. Fig. 4: Contact force vs. Glass position Plot Fig. 7: Simulation at 300th increment Fig. 5: Simulation at 100th increment Fig. 8: Simulation at 400th increment Fig. 6: Simulation at 200th increment Fig. 9: Simulation at 500th increment Fig. 10: Magnified view of the supporting Lip Fig. 11: Final deformed outline of the glassrun