How to optimize the design of pre-stressed plastic corrugated pipe for impact resistance?
Release time:
2023-04-28
How to optimize the crashworthiness of prestressed plastic bellows?
The energy absorption stability and controllability of the collision process of prestressed plastic bellows are relatively good, so it is very suitable for the design of energy-absorbing structures. However, due to the high nonlinearity of the bellows crashworthiness problem. It is difficult to optimize the design of its crashworthiness. Wang Xiao et al. applied the orthogonal optimization design method to optimize the crashworthiness of the bellows locally rather than globally, and only optimized the maximum acceleration single target.
Its main geometric parameters are: bellows inner diameter DD=2R wave height h wave distance q bellows wall radius r corrugated angle (the angle between the line segment and the axis) α (only 5 of these 4 parameters are independent and the wall thickness of the corrugated segment t It is difficult to directly optimize the crashworthiness of the bellows, and the prestressed plastic bellows is an axisymmetric structure composed of multiple corrugated and straight edge segments in longitudinal section. The ideal approach is to establish an approximate model of the crashworthiness analysis and then perform the optimization calculation. In order to obtain an approximate model, the experimental design is carried out to determine the sample points, then the virtual experiment is performed on each sample point, and finally the desired model is obtained by fitting the test results.
Homogeneous experimental design and Latin hypercube experimental design, etc. The latter two methods evenly disperse the test points in the space of design variables, and the commonly used experimental design methods are orthogonal experimental design. When the range of influencing factors is large and the level is large, the uniform experimental design and Latin hypercubic experimental design can greatly reduce the number of tests, so it is applied in this study. After experimental design, 21 initial sample points were obtained, considering that the second-order response surface model had poor fitting effect near boundary points.
Related News