Learning Forum

[jfrench]

I have a customer that asked me about this very issue with Discovery Live. In the attached, they ran the analysis prior, then ran it again and didn't get the same results. When they ran it in front of me, with roughly the same fidelity and with the same configuration and time step, they didn't achieve their original results. Does anybody have some insight into this?

🛈 This post originally contained file attachments which have been removed in compliance with the updated Ansys Learning Forum Terms & Conditions

[Brian Bueno]

JFrench

It's unclear to me what the request is here.  The PDF image appears to show 2 sets of results where the maximum temperatures differ by about 1 degree F.  The bottom image has a probe defined, but the top does not.

Fluid simulations are transient, so the results will fluctuate and may differ slightly at any given time.

The fidelity setting in both these cases is also relatively low.  Lastly, if you want to have more uniform results in a fluid analysis, you can enable the Time Averaged Result Display option.

[jfrench]

Brian Bueno 

 

I completely agree with them being transient and varying slightly from run to run. But this result set is vastly different from run to run. The temperature paths don't even come close to resembling each other, even with the 1 degree variation in scale.

Basically what I was asking is if it's possible to achieve different result sets based on the available (not total) RAM on the graphics card. 

[Brian Bueno]

JFrench 

Are you able to submit that model for further investigation?  I've seen this behavior once in the past.  In that case I had to make some modifications to the geometry to clean it up a bit.