Simulation and Design Optimization of Al Alloy Wheels Subjected to Biaxial Fatigue Loading. Ali KARA, Ph.D. Barış KÜSÜLÜ

Transcription

1 Simulation and Design Optimization of Al Alloy Wheels Subjected to Biaxial Fatigue Loading Ali KARA, Ph.D. Barış KÜSÜLÜ

2 Contents Company Profile Introduction Biaxial Testing Simulation with LBF.WHEELSTRENGTH Optimization by Design of Computational Experiments Results and Discussions Conclusions

3 CONTACTS ACROSS EUROPE Frankfurt Paris Munich Turin Barcelona Izmir

4 HISTORY Mr. Tonguç Ösen and his partner set up the first aluminum diecasting company in Turkey. OEM sales began for Fiat Auto-Turkey and Ford-Turkey CMS GMBH was founded for the Aftermarket wheel sales in Europe. The second production site was established in Çiğli. CMS Group R&D Center was established in Izmir. CMS Jant Sanayi A.Ş. started mass production Total CMS Group capacity has reached 9 million wheels. CMS Light Alloy Wheels was founded in İzmir. OEM sales began for Renault-Turkey Export to European automotive factories, Renault-France and Fiat Auto-Italy begins. LDS Lodos Teknik group company was founded. CMS Jant Sanayi A.Ş was founded in İzmir, Gaziemir Aegean Free Zone. CMS Group published its first sustainability report.

5 AWARDS Perfection award by Renault-France. Year 2003 OEM Quality Award by Tofaş / Fiat-Turkey. Regional Contribution 2006 award by Toyota. Quality award by Renault, France and by Toyota. Quality award by Toyota. Project Management & Cost awards by Toyota. Q1 Award" by Ford Motor Company. Qualitas award by Fiat Auto. Qualitas award by Fiat Auto. The Best Supplier award by Oyak- Renault Turkey. Excellence Award by Honda Access Europe. Regional Contribution 2013 award by Toyota. Excellent Overall Performance award by Honda Access.

6 Introduction ZWARP test can be conducted in CMS Mechanical Testing Laboratory since Wheel development considering fatigue requirements and impact reqiurements has used simulation support for more than a decade in CMS. Standard fatigue simulation was not able to demonstrate correct place and correct level for maximum stress due to its monoaxial loading history. Therefore Since 2016 June LBF.WHEELSTRENGTH has been used to develop wheels under biaxial fatigue testing conditions.

7 Biaxial Testing Simulation with LBF.WHEELSTRENGTH In CMS LBF.WHEELSTRENGTH is used with ANSYS Workbench. Finite Element mesh is prepared in ANSYS Workbench, In a previous Project, mesh sensitivity analysis was also conducted and element size was determined accordingly. Boundary condition is applied in ANSYS Workbench *.inp file is exported in order to use in following steps.

8 Biaxial Testing Simulation with LBF.WHEELSTRENGTH GUI of LBF.WHEELSTRENGTH:

9 Biaxial Testing Simulation with LBF.WHEELSTRENGTH Selection of horizontal and vertical load application nodes Selection of loadcases according to wheel requirements

10 Biaxial Testing Simulation with LBF.WHEELSTRENGTH Required Fatigue Strength calculation: By iteration to the allowable damage sum the RFS at the knee point of the S-N curve is calculated for each node. These results can subsequently be evaluated by test results in which cracks had occurred in individual areas

11 Optimization by Design of Computational Experiments A 7" x 16" CMS wheel was used in the study Design of Experiments (DOE) method was used to determine the effects. A full factorial design with one center point was created. Minitab 18 was used in the study.

12 Optimization by Design of Computational Experiments XPOS YPOS LENGHT WIDTH 50,0 105, ,0 110, ,0 105, ,0 105, ,0 105, ,0 110, ,0 110, ,0 105, ,0 110, ,0 110, ,0 105, ,0 110, ,5 107, ,0 110, ,0 105, ,0 110, ,0 105,

13 Results and Discussions Simulation Results Critical points considering the loading in ZWARP were determined as below Simulation results were optimized in terms of stress levels at these points along with the mass of the wheels. Minimize (Normalized Stresses & Mass) Stress levels for these three points were normalized by their maximum values. Spoke Cavity Spoke Cavity Style Window

14 Results and Discussions Simulation Results XPOS YPOS LENGHT WIDTH Mass (kg) Spoke Cavity Window Style ,164 0,653 0,747 0, ,066 0,695 0,769 0, ,988 0,737 0,813 0, ,973 1,000 1,000 1, ,868 1,000 0,989 1, ,863 0,947 0,934 0, ,977 0,705 0,780 0, ,068 0,737 0,824 0, ,120 0,600 0,692 0, ,059 0,726 0,747 0, ,119 0,653 0,758 0, ,170 0,600 0,692 0,926 52,5 107, ,048 0,768 0,835 0, ,124 0,726 0,747 0, ,054 0,811 0,846 0, ,977 0,926 0,912 0, ,112 0,789 0,857 0,963

15 Results and Discussions Statistical Results Pareto charts for objective functions, ANOVA with α=0.1

16 Results and Discussions Statistical Results Pareto charts for objective functions, ANOVA with α=0.1

17 Results and Discussions Statistical Results Factorial Regression Equations Backward elimination method was used with no hierarchy requirement

18 Results and Discussions

19 Results and Discussions Statistical Results ANOVA with α=0.1 and result for curvature calculation thanks to center point showed that: Optimum Mass is out of design domain Of course mass decreases itself by using a deeper and wider spoke cavity Optimum Style Stress for these parameters is in this design domain Optimum Window Stress for these parameters is in this design domain Optimum Spoke Cavity Stress for these parameters is out of this design domain this maybe because of the another factor affecting this stress or we need to look for other factor levels.

20 Results and Discussions Multiobjective Optimization Results Optimum parameters were defined In the style, for all of design points stress results are similar to each other Optimum design is ~195 gr lighter than the heaviest Stresses for spoke cavity and window parts were lowered more than 70%.

21 Conclusions In this investigation, Design of Experiment was used along with ZWARP simulation in order to find the optimum spoke cavity design. The effects of parameters and their interactions were shown. Optimum design was found ~192 gr lighter than the heaviest. For mass and spoke cavity stress optimum parameters are found to be out of the selected design domain. For the selected investigation points lower stress levels were achieved.

22 Thank you. Ali KARA, Ph.D. Barış KÜSÜLÜ