Flow goes unstable as simulation time gets increased

I am still working on improving my 3d-printer part cooler design.I have resolved most of the issues with my simulation, but I noticed in my latest free-air test case that the airflow seems to be doing something very odd, as the simulation time is extended. The airflow seems to start accelerating, and is wandering off into a directional stream in unexpected directions. I am thinking there is something I don’t understand about pressure outlets.

At 100 seconds:

At 150 seconds:

At 200s:

Hi @acwest

Firstly i would like to ask that what is your objective for this simulation. Are you trying to find out the performance of your 3d printer cooling fan?? If you are trying to simulate a heat transfer analysis between the printer nozzle and cooling duct then you would have to use CHT model. I have attached an image below and comparing the image and your CAD model, I don’t think that your nozzle area(hexagon) is appropriate to simulate this case. Again it will become much clear if I can get to know your objective(like- heat transfer/ simulate flow pattern around nozzle…etc…etc).

I hope that you have understood the above given information/issues and will tell me your objective as soon as possible. I am always there to help you in any situation.


Also make sure that you know the specification of your cooling fan(usually given in cfm(cubic feet per minute)) as it will be used as your inlet boundary condition.

Hi Ani,
Currently I am trying to model the airflow both inside the duct, and surrounding the hot end of the printer. The hexagonal prism shape represents my outlet, which I have set to a pressure of 1 atmosphere. I suspect my instability is caused by having too high a time granularity in the simulation, I believe it is sampling every second, which on parts this small is too long

Hi @acwest

For this, you have to assign wall boundary conditions to the nozzle and surrounding faces and increase the size of your hexagon as it will provide more resolution to this problem.

As you are running a steady state simulation, the writing time/ total simulation time don’t have anything related to physical time. Rather they are number of iterations for which your simulation will run and store the data. The time will only come into play when you are performing transient simulations.