A novel bi-graded honeycomb was proposed by introducing both in-plane and out-of-plane thickness gradients into a regular honeycomb. The graded honeycombs were additively manufactured by fused deposition modeling (FDM) with polylactic acid (PLA) and then tested for axial crushing. Numerical simulation models were constructed through LS-DYNA and validated using experiment results. Based on the super folding element (SFE) method, theoretical models of the proposed bi-graded honeycombs were derived and the accuracy for crushing response was validated against the numerical results. Finally, an active learning based multi-objective optimization algorithm was used to seek the optimal design. The results showed that the bi-graded design for honeycomb structures could improve energy absorption capacity and decrease the peak crushing force in the Pareto frontier manner. The specific energy absorption of the optimal bi-graded honeycomb could be 45.6% higher than that of the regular honeycomb while the peak crushing force was controlled at the same level.

通过将平面内和平面外的厚度梯度引入常规蜂窝中，提出了一种新颖的双梯度蜂窝。通过熔融沉积模型（FDM）与聚乳酸（PLA）叠加制造渐变蜂窝，然后进行轴向破碎测试。通过LS-DYNA建立了数值仿真模型，并利用实验结果进行了验证。基于超折叠元件（SFE）方法，推导了所提出的双梯度蜂窝的理论模型，并根据数值结果验证了压碎响应的准确性。最后，采用基于主动学习的多目标优化算法进行优化设计。结果表明，蜂窝结构的双级设计可以提高能量吸收能力，并以帕累托边界方式降低峰值破碎力。最佳双梯度蜂窝的比能量吸收比常规蜂窝的比能量吸收高45.6％，同时将峰值破碎力控制在相同水平。