Considering the roll problem of Macpherson suspension, Geely Company has designed a new type front suspension。 The suspension was similar to the structure of Macpherson suspension, but it effectively solved the problem of the roll。 However, the dynamic characteristic and fatigue life of this suspension were poor。 Therefore, these problems of the proposed new type suspension were researched and optimized in this paper in detail。 The paper built a multi-body dynamic model of suspension system。 The vibration displacement of wheel center on running pavement is obtained by experiment and applied it into the multi-body dynamic model to simulate actual working conditions。 Then, we extracted the vibration displacement of a point at the suspension and compared it with the experiment result in order to verify the accuracy of the multi-body dynamic model of suspension。 The dynamic load acting on the suspension jounce bumper was acquired by the multi-body dynamic model and applying it to a finite element model of the jounce bumper。 Then, the stress in the loading position was computed, and it was imported into NASTRAN to conduct fatigue durability analysis of the jounce bumper。 Afterwards, the dynamic characteristic and fatigue life of the jounce bumper were optimized based on genetic algorithm。 At last, vibration isolation ratios of the suspension system before and after optimization were extracted for comparison to verify optimization effect。 The paper innovatively combined dynamic analysis of suspension with assessment of fatigue life。 In this way, the suspension system could possess a better dynamic characteristics and high fatigue life。

2。 The background theory

Fatigue referred to development of local and permanent structural change occurring in cracked or completely fractured materials after sufficient cyclic perturbations due to perturbation stress on a point or some points。 Generally, the cyclic load causing fatigue failure was less than the “safe” load according to static strength analysis。 Traditional static strength analysis methods can’t solve fatigue problems。 Fatigue was one of the main causes for structural failure。 It was the main factor that needs to be considered in structural reliability experiment [5]。

High-cycle fatigue was also deemed as stress fatigue because it was mainly determined by stress amplitude。 Under this condition, the cyclic stress on the material was far less than its yield limit。 The cycle-number was large before damage, in general, it was bigger than 105-106。 When the cyclic stress was small, elastic strain became dominant。 The elastic stress will transform into plastic stress when it gradually increased to a certain extent。 Under this condition, the plastic strain started to become dominant gradually。 Generally, the stress level was low under high-cycle fatigue。 Under this condition, the material deformation was within elastic range, and the stress was in direct proportion to the strain。 Generally, the fatigue properties of the materials can be described by “stress-cycle-number” curve。

As for the high-cycle fatigue problem, the so-called S-N method was used to analyze life cycle based on the stress of the materials or parts。 In this method, the fatigue life was mainly calculated according to the stress or strain distribution of the parts provided by the finite element model。 Such analysis was conducted based on the fatigue characteristic of the materials。 This method didn’t explicitly distinguish the occurrence and propagation of the cracks。 It was mainly used for predicting the entire life of large damage or fracture occurring on the parts。 It can estimate the entire life before abrupt failure of the parts。

Linear cumulative damage theory referred to that the fatigue damage can be linearly superimposed [6, 7]。 The representative theory included Palmgren-Miner as well as modification Miner and relative Miner。 Miner assumed that fatigue failure will occur when the energy absorbed