Multiphase Flow Simulation Fatal I/O Error!


#21

Hi @cecil_pabbathi,

There are force vales onto the model so I don’t think it is not being initialized. I suspect that the time state isn’t present and what is being seen now is simply flow at 0s. I will re-run with the initialization and more time steps to check if this is really the case.

Cheers.

Regards,
Barry


#22

Hi @cecil_pabbathi,

The simulation mentioned and a new one I’ve run with a -3 outlet has managed to converge but still getting nonsensical results. Try changing the number of correctors in the numerics to 10 and see if it maintains the value below the CFL limit.

Cheers.

Regards,
Barry


#23

Hi @Get_Barried,

I created a new model with a different co-ordinate system so that the outlet velocity is a positive value when defined. I dont know if that made a difference because I found that in common with similar projects here. Look under ‘Ver12.5Test2’ - https://www.simscale.com/workbench/?pid=961853807733457623#tab_1-0 . Ran a case with a time step of 0.005, the residual monitor and force plots were stable for 12 secs free of any cfl issues until there was a sudden spike in the residuals and the sim crashed throwing a fatal i/o error. When I looked under the solution fields there was no backflow which could have caused this. Either way I am running another case with a bigger time step. The residuals and force monitors do look promising as shown below. I will keep you posted. Cheers.

Regards,

Cecil


#24

Hi @cecil_pabbathi,

The force plots do look much better and they are converging as well so that is good news. Lets see how it goes. Most important thing is to be able to see flow progressing throughout the domain and movement in the direction.

Do keep me posted!

Cheers.

Regards,
Barry


#25

Hi @Get_Barried,

The simulation crashes once again due to fatal I/O error :frowning: Cant figure out what is causing the divergence. The solution fields unfortunately dont provide any visual aid as to whats going on. Looks normal to me. Any suggestions ?

Regards,
Cecil


#26

Hi @cecil_pabbathi,

This is very strange, at least it isn’t a instant error. Not sure as well what the divergence is caused by, I’m guessing this is some sort of flow error probably as this usually happens when say the amount of mass into domain is more than that going out or even just not progressing at all out of the system so you get a kind of continuity error. At the moment I’m all out of ideas on what could be causing the issue.

Hopefully the @PowerUsers_CFD have an idea on this.

Do keep posting updates if you find a solution and if you’re really running out of time concurrently using a second solver like ansys would allow you to have some sort of a backup just in case.

Cheers and good luck.

Regards,
Barry


#27

Hi @Get_Barried,

As a follow up to my earlier post, the divergence is caused because of the waves hitting the outlet wall. I ran a turbulent case and wrote the data in time steps to see whats going on. Although I do not see this as a deal breaker because the results converge up until that point, I suppose I should be satisfied with the progress made in the last 2 days. Purely out of academic interest, is there a way to define the outlet to absorb the incoming surface waves ? Cheers.

Regards,
Cecil


#28

Hi @cecil_pabbathi

Looks like my guess earlier is correct. Normally in single medium based simulations we negate this by having a pressure outlet that can compensate for any “overflow” of the fluid out of the domain. In your case, your outlet having been defined as 3m/s out should be enough to cover this issue since your inflow is also 3m/s. From what I can see in the simulation, the flow speed is well below so it is quite perplexing.

Try increasing the outlet speed and see if flow actually manages to progress outward instead of just bunching up like that.

Cheers.

Regards,
Barry


#29

Hi @Get_Barried,

Firstly I would like to thank you for the help and support thus far in getting my free surface simulation up and running. The simulations run and converge well with meaningful results. Moving on to the next phase, which is to model FSI wave interactions. I believe this project ( https://www.simscale.com/projects/Ali_Arafat/free_floating_boat_in_a_wave_tank/) is the only reference available and I made sure to replicate the project set up as much as possible, but my simulation crashes quite early with a CFL error. So far this issue was caused because of a backflow. Looking at the solution fields, the initial wave just remains stationary and does not move with time. Here’s the link. Cheers.

https://www.simscale.com/workbench/?pid=961853807733457623#tab_1-0

Regards,
Cecil


#30

Hi @cecil_pabbathi,

Your setup looks alright. Try increasing the number of correctors to 10 or 20 like the reference 6DOF simulation to keep the CFL within limit.

Do let me know how it goes.

Cheers.

Regards,
Barry


#31

Hi @Get_Barried,

Increasing the correctors does eliminate the CFL issue but I keep running into fatal I/O errors. I dont see anything wrong with my bc’s as well. Appreciate it if you could do some tests on my case. Cheers.

Regards,
Cecil


#32

Hi @cecil_pabbathi, great that you are progressing with this :slight_smile:

I’m not sure if this will solve your problem but it looks like it is constraint to rotating about the x-axis, logically wouldn’t it be rotating around the y-axis, so it kind bobs with the flow direction?

If I have a bit of time later ill try run it.

Let me know if you make progress in the meantime,
Darren


#33

Hi @1318980,

Thanks. Much appreciated. Changing the constraints doesnt help. Will let you know if I make some progress. Cheers.

Regards,
Cecil