基于多岛遗传算法的复合材料缠绕壳体封头分区补强优化

Sectionalization-based reinforcement optimization of composite-wound case dome through multi-island genetic algorithm

  • 摘要: 针对Ф150 mm非等极孔复合材料缠绕壳体,开展了水压爆破试验研究,并通过建立基于三维Hashin失效准则的壳体精细化渐进损伤有限元模型,对壳体复合材料失效模式、爆破位置以及爆破压强进行了准确预测,以验证有限元模型的可靠性。以此为基础,根据壳体封头应力分布状态,建立了基于多岛遗传算法的封头分区补强优化模型,对补强层数和补强角度进行优化,揭示了不同分区补强角度及其耦合作用对封头纤维应力的影响机制,继而获得不同分区最优的补强角度及补强层数,并通过分区补强试验进行验证。结果表明,封头赤道圆至金属接头肩部区域补强角度对纤维应力的影响更为显著,宜采用小角度进行补强;而接头肩部至极孔范围内需采用较大角度分别对轴向和环向进行补强。通过分区补强水压爆破试验结果可知,壳体爆压提高了37.5%,壳体特性系数提高16.6%,说明该优化模型是准确且可靠的。

     

    Abstract: The hydrostatic tests for the Ф150 mm composite-wound cases with unequal poler openings were carried out. In order to precisely predict the failure modes, burst position and the burst pressure of the composite cases, the progressive damage model based on the 3D Hashin failure criteria was established and its reliability was evaluated by experimental results. Based on the finite element model, the sectionalization-based reinforcement optimization model through multi-island genetic algorithm was established to optimize the reinforcing layers and angles according to the stress distribution on domes. The influence mechanism of the reinforcing angles in different subareas and its coupling effect on the fiber stress were revealed and then the optimal reinforcing angles and layers were obtained. In addition, the subareaalization-based reinforcement test was implemented to validate the optimization model. The results of the numerical model show that the reinforcing angle from the equator of the dome to the shoulder of the metal boss has a more significant effect on the fiber stress and the relative small angle should be employed to reinforce domes; however, the relative large reinforcing angle in the subarea ranging from the poler openings to the shoulder of the boss should be applied to reinforce the axial and circumferential directions. The results of hydraulic burst test of the reinforcing case show that the burst pressure and the performance factor increase by 37.5% and 16.6%, respectively, compared with that of composite case without reinforcement, which indicates that the optimization model is accurate and reliable.

     

/

返回文章
返回