Composite optimization-phase 3 (shuffle optimization)

Altair Forum User

Hello,

I am working on a composite optimization problem. I follow the skateboard tutorial but I have several components in my problem (not all composite). I am trying to minimize the mass given some static displacement constraints. Phase 1 (free size) and phase 2 (size) optimizations are problem free but when I import the shuffling file produced in phase 2 to hypermesh, I see that the laminate I am trying to shuffle optimize includes plies from other components. Have you ever encountered such a problem? I cannot share the files due to confidentiality agreement.

 

Thanks

Altair Forum User

Hi,

Does the same happened with size and free size optimization?

 

Can you check which laminate is selected as design variable in Shuffle panel?

 

Altair Forum User

No, it doesn`t happen with free-size and size optimization.

 

I see 4 shuffle design variables in the shuffle panel (I have 4 laminates in my model):

First one includes only the laminate (Lam1) I want to optimize.

Second one includes Lam1 and Lam2.

Third one includes Lam1, Lam2 and Lam3.

Fourth one includes all 4 laminates.

 

Also, all 4 laminates have plies from each other.

 

 

 

Altair Forum User

Hi,

 

Thank you for the information. I will check and update to you soon.

Altair Forum User

Hi @ibozsoy1

 

So, I tried with a sample model where I have 4 shuffling variables and each has one laminate. This works fine.

 

I think the fourth design variable which has all four laminates is what making the problem. Can you retain only the design variable which you are interested in. 

 

Not sure if all the four design variables are created automatically after size, but in my case it did not happen.

Altair Forum User

Hi @ibozsoy1

 

So, I tried with a sample model where I have 4 shuffling variables and each has one laminate. This works fine.

 

I think the fourth design variable which has all four laminates is what making the problem. Can you retain only the design variable which you are interested in. 

 

Not sure if all the four design variables are created automatically after size, but in my case it did not happen.

Altair Forum User

Altair Forum User said:

Hi @ibozsoy1

 

So, I tried with a sample model where I have 4 shuffling variables and each has one laminate. This works fine.

 

I think the fourth design variable which has all four laminates is what making the problem. Can you retain only the design variable which you are interested in. 

 

Not sure if all the four design variables are created automatically after size, but in my case it did not happen.

 

I appreciate your time and help very much.

Actually, I haven`t run the shuffling optimization yet because it seems like there are 2 problems:

1) Four design variables created automatically after size optimization:  

      This is fine, I can retain only the one I am interested in. 

2) Mixed plies in each laminate after size optimization:  

      This is my main problem: when I check the laminates after size opt., I see that they consist of plies from different components. See attached screenshots: size.jpg shows the laminate named 'inner' before size optimization and shuffle.jpg shows the one before shuffling optimization. Component 'inner' is the one to be mass minimized.

 

Is it possible that the problem arises since I choose mass response for all 4 laminates? Optistruct does not allow me to choose only one component so I choose its material (AS4/PEI in the pictures) as mass response which includes all 4 components. 

 

Also, I encounter a third problem after free size optimization. After importing the ...sizing.x.fem file created by free sizing optimization, the total weight of these 4 components does not match the one reported in the output file of the free sizing optimization. Do you have any suggestion for that?

 

By the way, I updated all Altair software in my computer to the latest versions 2 days ago but it didn`t help either.

 

Many thanks.

<?xml version="1.0" encoding="UTF-8"?>

<?xml version="1.0" encoding="UTF-8"?>

Altair Forum User

Hi @ibozsoy1

 

Sorry for a late reply. Not sure why this has happened. I will check with experts and will let you know their comments ASAP!!!

Altair Forum User

Thanks Prakash.

Altair Forum User

Altair Forum User said:

Is it possible that the problem arises since I choose mass response for all 4 laminates? Optistruct does not allow me to choose only one component so I choose its material (AS4/PEI in the pictures) as mass response which includes all 4 components. 

Hi @ibozsoy1

 

Can you use Property instead of material when you are creating mass response?

 

Altair Forum User said:

Also, I encounter a third problem after free size optimization. After importing the ...sizing.x.fem file created by free sizing optimization, the total weight of these 4 components does not match the one reported in the output file of the free sizing optimization. Do you have any suggestion for that?

I think this is because, the initial material density used when having a mass response is 0.9. To use 100% material, use OptiCotrol>> MATINIT>> Enter 1

 

BTW, I wish you a very happy New Year /emoticons/default_smile.png' srcset='/emoticons/smile@2x.png 2x' title=':)' width='20'> 

Altair Forum User

Thank you Prakash, happy New Year to you too /emoticons/default_smile.png' srcset='/emoticons/smile@2x.png 2x' title=':)' width='20' />

 

I used property but obtained the same results, no change.

 

Also, I still have inconsistency in mass although I set MATINIT=1

 

Any other thoughts?

Altair Forum User

Hi @ibozsoy1

 

I will share your findings with experts and will let you their suggestions and feedback.

 

Faisal

Altair Forum User said:

 

I appreciate your time and help very much.

Actually, I haven`t run the shuffling optimization yet because it seems like there are 2 problems:

1) Four design variables created automatically after size optimization:  

      This is fine, I can retain only the one I am interested in. 

2) Mixed plies in each laminate after size optimization:  

      This is my main problem: when I check the laminates after size opt., I see that they consist of plies from different components. See attached screenshots: size.jpg shows the laminate named 'inner' before size optimization and shuffle.jpg shows the one before shuffling optimization. Component 'inner' is the one to be mass minimized.

 

Is it possible that the problem arises since I choose mass response for all 4 laminates? Optistruct does not allow me to choose only one component so I choose its material (AS4/PEI in the pictures) as mass response which includes all 4 components. 

 

Also, I encounter a third problem after free size optimization. After importing the ...sizing.x.fem file created by free sizing optimization, the total weight of these 4 components does not match the one reported in the output file of the free sizing optimization. Do you have any suggestion for that?

 

By the way, I updated all Altair software in my computer to the latest versions 2 days ago but it didn`t help either.

 

Many thanks.

<?xml version="1.0" encoding="UTF-8"?>

<?xml version="1.0" encoding="UTF-8"?>

Hello,

I am also facing the same issue, have you find the solution for this? If anyone know a soluton for this, please help.

Thanks.