I m really stuck on something, Thank you very much for your time and eventual help to solve the issue.
I’m trying to get a “picture” of a moment in time of the temperature of a rotating pipe with a Fan inside.
Goal being to compare the cooling speed of arotating pipe from 80 degres celsuis to ambient temperature, 20 C with and without fan.
For unknown reason transient IBM sim simply does not works. I launched 6 or 7 , one again tonight with very simple simulation control and the simulation is still running.,every time getting stuck at 5min. The mesh seem ok and courant number is forgiving with numerics.
Without more clues how to get a probing result i tried a steady state IBM sim with simple simulation control.
The results of this simulation are not good, because the pipe at every timestep is at 20 is at degres ,I suppose because it is the temperature 'in fine" of the pipe “not in duty” therefore not warming, just cooling in rotation till reaching ambient temperature.
But i look carefully the cutting plane and very surprised to see the velocity raising at the different time steps? The mrf simply indicates 300rds but there it’s evolving over time?
Maybe a steady state can be used to get what i need? Either to know when the pipe will reach 20 degres or, if the computing power is not enough; to see what would be the temperature of the pipe after 3, 5 minutes in rotation for example.
Thank so much if someone can help. I have one simulation left meaning one chance to see if I could get what I’d like with the tools i have and my very little knowledge in this field.
Thank you again and looking forward to reading from you!
Hi @kittynotacat, I’ve had some insights I’d like to share with you:
Regarding your question:
Transient simulations are indeed very expensive, as my colleague @Ricardopg pointed out in your previous forum post. In that sense, if you just want to see the effects of the cooling the fan provides, maybe an alternative would be to try to simulate a smaller model - cut in half let’s say? If that’s not an option considering the results you’re expecting, please disregard
As Ricardo also pointed out, working with STL naturally comes with great difficulties. Your CAD model doesn’t look very complex, is generating a geometry in other format not an option? You could use, for example, OnShape.
It seems that you’re assigning the MRF Rotating Zone to a larger region than just the inside
As Ricardo pointed out, unfortunately for what you’re trying to do a Steady-State simulation wouldn’t be possible. That’s because, independently of the time-step, Steady-State is like the stabilized version of a transient case (in which stabilization occurs) - as if an infinite amount of time had passed.
In that sense, the results will show equilibrium values for temperature, which will be 20 °C since that’s the value you’ve set for the air around your volume:
Thank you very much for the feed-back.
I’m afraid not. STL will not be an issue if we get to work with Simscale as our designs are more complex and somebody else will do it for us. For now this STL file is quite simple done by somebody good at it so hoped it should fine. The webinar I looked at it many times, didn’t find anything that could be of some help.
3. The MRF zone seem correct in the sense that the pipe and the fan are rotating together and that the zone should be 1.1? Indeed using rotating walls could be better?
The goal here is just to make the simulation less heavy ? The fact is the design is so simple, that i think something is just awfully wrong with my parameters. Even the my shortest simulation controls parameters are not working.
4 yes I understand, i mistaken looking at a public project hoping i could be a solution. It is weird though that velocity raise even if it is not my concern for the specific topic.
I read a lot those days and I start to think it is not possible for me, for skilled people of course, but for a newbie like me to achieve the result needed.
Igaviano, you kindly looked at the last IBM i lauched, still running till now. Did something looked odd or the simulation is as it should be? Is it only my simulation control might be heavy to process?
In that case what could be some simulation control that would be so simple that i would be able to matter the duration result to achieve something with a variance in time and of temperature visible in the solution field?
Thank you again very much for your time and very kind help. It is really appreciated
Then your setup should be correct - MRF rotating zones should provide a better result indeed. I asked because I thought the outer pipe might be still
Yes, I don’t find anything wrong with your setup in particular, apart from the questions I’ve raised and you’ve answered. Sometimes a simple transient simulation will take a long time to run due to the Courant number, as Ricardo mentioned
As to the results, they seem coherent but only one step seem to have been calculated:
I would like to launch my last simulation and see where if it can get to something. Could you kindly indicate me a sure set up for example a simulation control of 20 seconds with a steps every 5 seconds ? Anything, just something simple that would indicate me that what I try to achieve is possible to do.
If the 20 seconds is too long , even 4 seconds with a step every one second. Simply something working because even if you and Ricardo seemed to have validated the process none of the set up did lead to any concrete result but a bug at 5 min computing time.
i Join the last link
Thank you so much !
Looking forward to reading from you,
Kitty
Thing is even though my request seem simply to be applicated in the software and simple, I’m not sure how to write it down. All writing I made just went off.
Thank you very much again for you return and help,
Kitty
