Learning Forum

[Tim T]

I defined "*ALE_STRUCTURED_MESH  # 1" and "*ALE_STRUCTURED_MESH # 2" with *ALE_STRUCTURED_MESH and coupled them to the same Lagrange unit as ALE_STRUCTURED_FSI # 1 and ALE_STRUCTURED_FSI # 2 respectively，I set TDEATH=20 for *ALE_STRUCTURED_MESH # 1, *ALE_STRUCTURED_MESH # 2 TDEATH=default, when the calculation time reaches 20 the program will *ALE_STRUCTURED_MESH # 1, ALE_STRUCTURED_FSI # 1 and ALE_STRUCTURED_FSI # 2 were deleted at the same time, which caused my calculation failure. The specific error is shown in the figure. How should I set the keywords so that the program only removes *ALE_STRUCTURED_MESH # 1 and ALE_STRUCTURED_FSI # 1 and let the program continue? What I mean is that the program automatically removes ALE_STRUCTURED_MESH # 1 and ALE_STRUCTURED_FSI # 1 and then ALE_STRUCTURED_FSI # 2 and the Lagrange cell continue to function until I reach the computation termination time I set.

[Ian Do]

Hi Tim,

Pls describe the details of the physics being modeled. ALE domain is not treated simply as a LAG part, and therefore it is dangerous to just "delete" one. I need to know what you are modeling and why you need to do this to maybe suggest an approach.

Regards,

Ian Do

[Tim T]

The model I established uses two different fluids to impact a concrete block. The first fluid is in *ALE_STRUCTURED_MESH # 1, and the second fluid is in *ALE_STRUCTURED_MESH # 2. I set TDEATH = 20 for *ALE_STRUCTURED_MESH # 1, hoping that the impact of the first fluid on the concrete would end at 20 μs and the second fluid would start to impact the concrete at the same time. However, an error as shown in the figure occurred at 20 μs.

[ilker.kurtoglu]

I assume that you are using two different part ids for *ALE_STRUCTURED_MESH cards. If not, that may be the problem. Moreover, I again assume that you use two different *ALE_STRUCTURED_FSI if you are modelin your concrete block as lagrangian. Then you should also kill the related *ALE_STRUCTURED_FSI card. Another way is that I suggest you to use classical ALE for using two different ALE domains. If you want to perform successive simulation, you can use a restart approach to your problemm.

Regards,

İlker Kurtoğlu

[Tim T]

I using two different part ids for *ALE_STRUCTURED_MESH cards.I also kill the related *ALE_STRUCTURED_FSI card，but the calculation is still incorrect.

[Ian Do]

Hi Tim,

Interesting! Which exe version did you use? 

Have you tried the latest beta version? If you tried the latest beta and this still happens, pls send a clean (remove all KWs not "absolutely required") and well organized input file. I can review it. This may take some time. 

Curious, even if you have 2 fluids impacting the same LAG block at 2 different times, i am not sure you really need to use 2 separate ALE domains(?). In fact, physically that would not be faithful to the physics, as 1 fluid may displace the other during the transition, no? Can you send a schematic of the physics? 

[Ian Do]

Hi Tim,

This may have been due to a bug that has been recently fixed. You can try the latest beta exe to see.

Ian