| ROBUST DESIGN OPTIMIZATION OF SUSPENSION SYSTEM BY USING TARGET CASCADING METHOD |
| Dae Oh Kang1, Seung-Jin Heo2, M. S. KIM3, W. C. CHOI4, I. H. KIM4 |
1kookmin
2Kookmin University
3Institue of Design Optimization
4Kookmin Universit |
|
|
|
|
|
| ABSTRACT |
|
This study presents the robust design optimization process of suspension system for improving vehicle dynamic performance (ride comfort, handling stability). The proposed design method is so called target cascading method where the design target of the system is cascaded from a vehicle level to a suspension system level. To formalize the proposed method in the view of design process, the design problem structure of suspension system is defined as a (hierarchical) multilevel design optimization, and the design problem for each level is solved using the robust design optimization technique based on a meta-model. Then, In order to verify the proposed design concept, it designed suspension system. For the vehicle level, 44 random variables with 3% of coefficient of variance (COV) were selected and the proposed design process solved the problem by using only 88 exact analyses that included 49 analyses for the initial meta-model and 39 analyses for SAO. For the suspension level, 54 random variables with 10% of COV were selected and the optimal designs solved the problem by using only 168 exact analyses for the front suspension system. Furthermore, 73 random variables with 10% of COV were selected and optimal designs solved the problem by using only 252 exact analyses for the rear suspension system. In order to compare the vehicle dynamic performance between the optimal design model and the initial design model, the ride comfort and the handling stability was analyzed and found to be improved by 16% and by 37%, respectively. This result proves that the suggested design method of suspension system is effective and systematic. |
| Key Words:
RDO(Robust Design Optimization), RBF(Radial Basis Function), TC(Target Cascading), MDO(Multi-Disciplinary Optimization), COV(Coefficient Of Variance), TEKS(Tabular Elastic-Kinematic Suspension), SAO(Sequential Approximation Optimization) |
|
Preview
Full text via DOI
Download Citation 
Print 
