Learning Forum

[arnau994]

Hello,

I am running an FSI model, but I get this error just after reading the ICFD boundary conditions (after finishing the initialization, and once the ICFD solver is on).

*** Error 40746 (SOL+746)
      more then 2000 additional nodes in crash front, stop inroutine hhl59s

Does anyone have any information about this error? I have not been able to find anything on the internet or LS-Dyna documentation. Besides, it is quite strange because it is not a crash impact model but a FSI

Any input is very much appreciated! Thank you for your time.
Arnau

[Reno Genest]

Hello Arnaud,

Which version of LS-DYNA are you using? I am checking with a developer.

Also, have you tried to remove keywords in the model to try to identify what triggers the error message?

 

Reno.

[Reno Genest]

Hello Arnaud,

Could you try to run the model separately, that is, the structures part only and then the ICFD part only. Is one of these triggering the error message? 

 

Reno.

[arnau994]

Hello Reno,

Thank you for your message, below I am replying each of your messages.

I am using LS-Dyna 11.2, double precision and in SMP configuration. In theory, I am only running the essential keywords for this model (I have followed the tutorial on the LS-Dyna examples webpage "Intermediate: FSI tutorial flap"), and have only changed the mesh input for both the Fluid and Structure, as well as adding the fixation constraints on the correspondent nodes. However, I have used a laminate material MAT58 for the whole structure. Tomorrow first thing in the morning will try to use a common Aluminum with MAT1 and check if I get the same error.

I also tried to separate the model, but I have only run the ICFD part. The ICFD simulates successfully both explicit and implicit. I will also try to run the structural part alone, but I am not sure how to do it as there are no loads applied to the model.

I will keep you updated regarding the material changes and the strucutre-only simulation. In the meantime, I will be very grateful to hear from the developer.

Thank you very much for your time Reno.
Arnau

[Reno Genest]

Hello Arnaud,

A crashfront is usually associated with the structure, especially with composite materials. Here is more information from the LS-DYNA Manual Vol II:

"Crashfront elements
are elements that are direct neighbors of failed (deleted) elements."

And *MAT_058 has a crashfront algorithm. Here is more information from the LS-DYNA User Manual Vol II:

"Crashfront. To start the crashfront algorithm, input a value for TSIZE. Note
that the time step size, with element elimination after the actual time step, becomes
smaller than TSIZE"

I would expect the error message to go away when using *MAT_001.

 

Let me know how it goes.

 

Reno.

[arnau994]

 

Hello Reno,

I have put the Tsize value for both laminate materials back to 0 and the FSI model now can run without errors aparently.

However, I can’t be sure about it. Although I am running an explicit model (I am using *ICFD_Control_Time and *Control_Timestep for the fluid and solid solvers respectively, and without any implicit control cards), it seems that the structural solver (I guess) ignores the timestep given and is performing an implicit analysis with dt = 2.67e-09. Therefore, the structural solver is still running in its first iteration and I can’t tell if everything is going fine. I am attaching a screenshot of it as well as the *Control_Timestep control card.

The timestep is set to be 0.05s. Also, as I said before, the ICFD part of the model runs as expected with this explicit configuration.

Let me know what you think, please.

Thank you very much.
Arnau

 

[Reno Genest]

Hello Arnaud,

It is good that the error went away. The crashfront algorithm in *MAT_058 seemed to be causing the problem.

You should create a new thread for the new problem. This way, it will be easier for other users to find relevant information about a specific problem.

By the way, you are running explicit. Refer to the LS-DYNA user manual for the lctm parameter on *CONTROL_TIMESTEP. Also, Google "how to perform implicit simulation in ls-dyna?" and you should find the answer.

 

Reno.

[arnau994]

Hello Reno,

Yes of course. Sorry for mixing topics.

Thank you very much for your help!

Arnau