Learning Forum

[mutianzi1]

Hello! 

I was doing a two-way coupled FSI following the example of the cantilever beam. In that example, only one time step is used for Transient analysis.

 

I also used one time step in my simulation. My question is that the the structure has very large displacement at the very beginning. I think it's due to the abrupt application of the self weight (standard earth gravity). I am going to increase the external loads, so the structure might fail due to such large displacement. I am wondering whether I could ramp the gravity load or use multiple time steps in 2 way FSI. Could anyone give me some suggestions about this? Thank you! Any help is appreciated.

 

 

[peteroznewman]

The preferred way to accomplish an initial load with zero initial velocity in a Transient Structural model is to have 3 steps.

• Step 1 has Time Integration Off and ends at 0.001 s. This is where the initial deflection occurs due to the gravity load. The transient load is either Inactive or set to zero.


• Step 2 also has Time Integration Off and ends at 0.002 s. Since there were no changes in load from step 1 to step 2, nothing moved, which means the velocity at the end of step 2 is zero.


• Step 3 has Time Integration On and begins with the initial deflection and stress, but with a zero velocity. The transient load is now set to Active or has non-zero values.

This was all explained in your previous discussion.

[mutianzi1]

Hello Pefer, I understood this for transient analysis only.

 

I was wondering may I still use multiple time steps if I conduct two-way coupled Fluid structure Interaction. will the three time steps repeat each time when data transfer between Fluent and Mechanical happens.

 

Thank you.

[peteroznewman]

When you couple a Fluid and Structural model, they are exchanging data on synchronized time steps. Once the first two time steps have been solved in Structural, they are in the past history. The rest of the steps are a continuation of the same time history. Why do you think earlier time steps would repeat?

[mutianzi1]

Yeah, I think you are right.

 

I defined three steps in Transient analysis. But a question mark appear on the Fluid Solid Interface. When I define one time step only. The question mark is gone. Does this mean only one time step is allowed in two-way coupled simulation?

Thank you.

[peteroznewman]

I don't know. I hope someone who does will answer this question, which belongs in the Multiphysics category, but someone might see it here.

[Anish thapa]

Dear mutianzi1 and peteroznewman:

I am facing similar issue. "the fluid solid interface load is not valid for multi-step analysis".

Have you found any solution to it? Any guidance would be helpful.

Thank you.