The optimization result is not symmetric for L-shaped structure

Altair Forum User

Hi guys, 

  I obtained the optimal L-shaped structure using free-shape optimization method. Inner nodes can freely move in thickness direction within a certain range (as shown in figure 1). Minimizing the out-of-plane tensile stress (interlaminar tensile stress) is the objective. The constraints are that maximum load point displacement and total weight are less than or equal to that before optimization.  

  However, in the case of shape change, it is not a symmetric result although the input and loadcase are symmetric (In the attached figure 2). Do you have any idea about the reason? If possible, do you have any idea how to obtain symmetric result?  

  I share the model file and results as follows: 

1) Input deck of the free-shape optimization of L-shaped structure: L-shaped_free-shaped optimization.fem

2) Result of the free-shaped optimization: L-shaped_free-shaped optimization_des.h3d 

 

Best regards,

GUO

 

 

 

Altair Forum User

Hi @Gregory1991

 

Can you try with Pattern grouping with 1 plane symmetry and check if that helps?

Altair Forum User

Hi, Prakash, 

  Thanks for your help. I knew pattern grouping may help.

  However, in my case, the design variables include the left nodes movements, the corner nodes movements and the right nodes movements as shown in the figure. Besides,

the left, corner and right nodes moves along z-axis (normal direction) in the local coordinate 1, 2 and 3, respectively. Therefore, the pattern grouping for all nodes of inner part is not possible since the pattern grouping is just suitable for one design variable. Do you know how to solve this problem?

  Best regards,

  GUO 

 

 

 

 

 

Altair Forum User

Hi,  

  Does anyone have any idea about the above question? 

Altair Forum User

Hi,

 

Sorry, I completely missed it. 

 

Can you enforce symmetric constraint only to the corner and check if that helps?