[ametalorb]

Hi All,
I’m running an external flow simulation in Solidworks 2019 which is both transient and looks at CHT. Rouhgly, the setup is an inlet BC that flows a hot gas at around Mach 2.5 over an object which sits in the stream (very similar in setup to this example https://blogs.solidworks.com/solidwo...agon-fire.html but with a different intent). My goal is to learn how long the object can survive this hot gas until it starts to melt. I’ll also note that I don’t currently have the “high mach number” setting turned on because it was really hurting the simulation speed.


The problem I seem to be facing is that I have two very different timescales in this problem. First, the development of the flow from the input BC takes on the order of 1e-5 seconds to occur (i.e., if I set the manual time step any higher than that the solution diverges). Second, the actual heating of the object takes on the order of 10 seconds to reach the melting point. Therefore, leaving the step at 1e-5 leads to multi-day runtimes.


My question is, assuming that my meshing is appropriate, if I wait for the initial flow from the input BC to develop then pause the simulation and up the step size (e.g., to 1e-4 or 1e-3), will I be significantly decreasing the accuracy of the simulation? Also, is there a way to do this automatically, for instance with a tabular timestep that depends on physical time?


I’m a relative CFD novice, so any help would be greatly appreciated. If there is a better way to approach this problem, I’d be happy to hear suggestions as well.


Thanks!
p { margin-bottom: 0.1in; line-height: 115%; background: transparent }a:link { color: #000080; so-language: zxx; text-decoration: underline }

[Boris_M]

In the calculation controls where you specify the manual time step is a small button on the right with ... in it. This allows the definition of a dependency to change the time step size over the iteration or physical time etc.

You could also use a steady-state simulation to get the developed flow and then clone that project and switch it to transient. and use the previous results in the initial conditions of the general settings and only define initial conditions manually for the object for the temperature it should start with. This allows for a faster calculation from then on as the hot jet is developed already and you are only looking at the heating of the object. The flow is so fast, the object should be covered with it quicker than it has any temperature change. So the error should be little.

You can also use Flow Freezing. This freezes the flow calculation but allows for the heat transfer, just the flow field is not calculated each iteration. If you use it periodically it will allow to update the flow and then freeze it again for some time.

This should give you some ideas of what you can try.

Boris

[ametalorb]

Thanks Boris!


I think all of these are good ways of doing what I need. I'll give them a try and post another update if I get stuck on anything else. Thanks again for your help!


Josef