IBAMR results

[tangjia zhang]

I am currently working on a numerical analysis involving periodic motion and have encountered some issues that I would appreciate your guidance on.

In my current setup, when the inflow velocity is set to 0, the thrust results do not exhibit a stable periodic pattern. However, when I adjust the inflow velocity to 0.2 m/s, the periodic behavior becomes well-defined and stable. I would like to ask if there are any modifications or parameters I could adjust to ensure that the calculation remains stable and yields a periodic result even when the inflow velocity is set to 0.

Thank you very much for your time and assistance. I look forward to hearing from you.

[Boyce Griffith]

Is this with a standard IBAMR example or with your own model? I think we need a little more information to be able to help.

In the meantime, these are just some guesses:

I think you are doing something like a swimming simulation — is that right? If there is no inflow velocity, does the swimmer get close to the boundary of the computational domain? You might be seeing the impact of boundary conditions, which might be reduced if you match the swimming velocity with the inflow velocity, keeping the swimmer close to the middle of the computational domain.

<X.JPG><y.JPG><z.JPG>

--
You received this message because you are subscribed to the Google Groups "IBAMR Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibamr-dev+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/fed90609-7f2a-4b1b-8e44-deb784f37d03n%40googlegroups.com.
<X.JPG><z.JPG><y.JPG>

[tangjia zhang]

Dear Professor Griffith,  

Thank you for your response.  

As you mentioned, we are working on a 3D undulating fin motion model with a tethered system. This is our own model, inspired by the black ghost knifefish.  

The issue we are facing arises when the inflow velocity is set to 0. The computational results become unstable. Theoretically, since the undulating fin performs periodic motion, I would expect the results to also be periodic. However, in practice, the forces in the X, Y, and Z directions are not periodic, as shown in the attached figures.  

Interestingly, when we set the flow velocity to v = 0.2m/s, the results become stable. I’m wondering if this instability might be due to a mismatch between the model nodes and the computational grid, or perhaps an imbalance between vortex generation and dissipation.  

I’ve been trying to adjust the parameters to achieve stable thrust generation, but so far, I haven’t been successful. I would greatly appreciate any insights or suggestions you might have regarding this issue.  

Thank you in advance for your time and guidance!  

Best regards,  
Tangjia Zhang

Boyce Griffith <boy...@gmail.com> 于2024年11月11日周一 21:33写道：

To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/8EF7CB54-1909-4908-8607-B71E598DB666%40gmail.com.

[Boyce Griffith]

On Nov 17, 2024, at 2:44 AM, tangjia zhang <zhangtan...@gmail.com> wrote:

Dear Professor Griffith,  

Thank you for your response.  

As you mentioned, we are working on a 3D undulating fin motion model with a tethered system. This is our own model, inspired by the black ghost knifefish.  

The issue we are facing arises when the inflow velocity is set to 0. The computational results become unstable. Theoretically, since the undulating fin performs periodic motion, I would expect the results to also be periodic. However, in practice, the forces in the X, Y, and Z directions are not periodic, as shown in the attached figures.  

Is the computation itself unstable, or is it physically unstable? I would be very surprised if the computation itself is unstable (e.g., solvers diverge) depending on inflow conditions.

Interestingly, when we set the flow velocity to v = 0.2m/s, the results become stable. I’m wondering if this instability might be due to a mismatch between the model nodes and the computational grid, or perhaps an imbalance between vortex generation and dissipation.  

My guess is that, without inflow, the swimmer is strongly interacting with its own wake — it might require longer simulations to reach periodic steady state. If the Reynolds number is high enough, I am not sure that you would expect to realize periodic dynamics in this case. (Back-of-the-envelope suggests that Re is 1e3—1e4 or higher.) This seems consistent with what you’ve been seeing and doesn’t seem to indicate a problem with the code or model.

I’ve been trying to adjust the parameters to achieve stable thrust generation, but so far, I haven’t been successful. I would greatly appreciate any insights or suggestions you might have regarding this issue.  

Perhaps others who have more experience with swimming simulations can chime in, but if you want to look at thrust generation from rest, I would suggest that you do not tether the swimmer, but instead let it swim. If you want to tether the swimmer, I think that I would suggest looking at thrust generation at a steady inflow that corresponds to the (expected) swimming velocity.

Thank you in advance for your time and guidance!  

Best regards,  
Tangjia Zhang

Boyce Griffith <boy...@gmail.com> 于2024年11月11日周一 21:33写道：

Is this with a standard IBAMR example or with your own model? I think we need a little more information to be able to help.

In the meantime, these are just some guesses:

I think you are doing something like a swimming simulation — is that right? If there is no inflow velocity, does the swimmer get close to the boundary of the computational domain? You might be seeing the impact of boundary conditions, which might be reduced if you match the swimming velocity with the inflow velocity, keeping the swimmer close to the middle of the computational domain.

On Nov 10, 2024, at 10:15 AM, tangjia zhang <zhangtan...@gmail.com> wrote:

I am currently working on a numerical analysis involving periodic motion and have encountered some issues that I would appreciate your guidance on.

In my current setup, when the inflow velocity is set to 0, the thrust results do not exhibit a stable periodic pattern. However, when I adjust the inflow velocity to 0.2 m/s, the periodic behavior becomes well-defined and stable. I would like to ask if there are any modifications or parameters I could adjust to ensure that the calculation remains stable and yields a periodic result even when the inflow velocity is set to 0.

Thank you very much for your time and assistance. I look forward to hearing from you.
<X.JPG><y.JPG><z.JPG>

--
You received this message because you are subscribed to the Google Groups "IBAMR Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibamr-dev+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/fed90609-7f2a-4b1b-8e44-deb784f37d03n%40googlegroups.com.
<X.JPG><z.JPG><y.JPG>

--
You received this message because you are subscribed to the Google Groups "IBAMR Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibamr-dev+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/8EF7CB54-1909-4908-8607-B71E598DB666%40gmail.com.

--
You received this message because you are subscribed to the Google Groups "IBAMR Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibamr-dev+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/CABrhyRT-cxyMco7Fz3qLn86P4fYagjpjUtELWV57BmqEy9vDuA%40mail.gmail.com.

[Amneet Bhalla]

I think there is some confusion about using the terminology “unstable results”. Do you mean “unsteady” or “non-periodic” results instead? We would infer unstable as simulation blowing up or stopping suddenly.  

--Amneet 

To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/D84C6B45-BD61-497A-A46D-392151E49DEB%40gmail.com.

[tangjia zhang]

 Dear Professor Griffith,  

Thank you for your response.  

Actually, there are no issues with the model or the code itself. The problem I am facing is that the computational results are not periodic, which I interpreted as instability. I apologize for any confusion caused.  

Perhaps the non-periodic results arise because the inflow velocity does not match the swimmer’s motion, requiring a longer time to reach a balance between vortex generation and dissipation. Moving forward, I plan to extend the simulation time and allow the swimmer to move freely to see if these adjustments improve the results.  

Thank you again for your valuable insights and support!  

Best regards,  
Tangjia Zhang  

Boyce Griffith <boy...@gmail.com> 于2024年11月18日周一 22:23写道：

To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/D84C6B45-BD61-497A-A46D-392151E49DEB%40gmail.com.

[tangjia zhang]

Dear Bhalla,  

Thank you for your reply, and I apologize for not clearly explaining the issue regarding the “unstable results.”  

What I meant to convey is that the computational results are non-periodic when V = 0m/s, not that the simulation blows up or stops suddenly. However, when V = 0.2 m/s, the results become periodic, and as shown in the figures, the thrust frequency matches the motion frequency.  

I don’t understand why periodic results cannot be obtained when V = 0m/s. I would greatly appreciate any insights you might have on this matter.  

Best regards,  
Tangjia Zhang  

Amneet Bhalla <mail2...@gmail.com> 于2024年11月18日周一 22:32写道：

To view this discussion visit https://groups.google.com/d/msgid/ibamr-dev/CAMETWJ2bR8ELS8REu-h%3DxJn-FD24E6h3ZiHYtdbDApAGWKipPg%40mail.gmail.com.

[Amneet Bhalla]

Hi Tangjia, 

I think the reason could be the one provided by Boyce earlier in his reply: the fin is interacting with its own wake when the inflow velocity is V = 0 m/s. To test this hypothesis, first let the fin swim freely and obtain its steady state swimming velocity. Next set the inflow velocity as the steady state swimming velocity for the tethered case. In the latter case, the net hydrodynamic force over a steady cycle (and not the instantaneous) would be zero. 

Thanks, 

--Amneet 

--

--Amneet 

[tangjia zhang]

Thank you for your explanation. 

Amneet Bhalla <mail2...@gmail.com> 于2024年11月18日周一 23:32写道：