Fusion 360: Discordance of thermal simulation results since the 2023 update

[tiphaine_thomas]

To begin with, I use Fusion 360 on Windows 11 under an educational license.

In December 2022, I modeled many assemblies on which I then did several "Thermal" simulations. On the results obtained, I was able to take temperature data by placing multiple control points.

First simulation

<- Results of my first simulation

Lastly, at the beginning of March 2023, while reopening the software after several months without use, this one to carry out an update. After it was correctly installed, I ran the same simulations again: on the same designs with the same study materials, the same loads, the same mesh and keeping the same location as before for the control points. I get very different results from my first simulations. I took the time to recheck my parameters several times, to run more simulations and I get today the same results different from my first ones.

Second simulation

<- Results of my second simulation

The pictures correspond to 1 setup. I have many others that differ in design but whose simulation parameters are all similar. And for all these others, the results were also transformed.

I have read about the new features of the update, and it seems that there are no changes in the "Thermal" simulations.

So I would need to know if it is the software, which since the update, calculates differently, or if it is my settings.

Thank you !

[karol.suchon]

Hello,

I have a quick look at the defect list reported by the users and solved by the developers. There is one which could have an influence on your results, there is one thing which could be essential. There was a software defect where the surface heat condition was applied on the face where we have bonded condition, the heat source is applied only to one surface, and the second face has applied adiabatic condition. From the pictures, it looks like we could have that kind of issue.
Could you send me the model, I will try to compute it in Autodesk CFD to determine which results are proper and where could be a problem.
BR
Karol

[tiphaine_thomas]

Hello,

You have correctly understood my problem.

You will find the model attached.

[karol.suchon]

In a beginning let's focus on getting proper results and later we could determine what exactly happens, we have some BC applied on the contact faces so the source of the issue is rather known. But I am not sure if the conditions are applied properly.

The convection and radiation BC are applied on the outside faces but we have Collier Fourreau bodies which prevent the environment to take part in the heat transfer exchange.

On the inner face, you applied constant temperature BC, we have bonded contact so that BC is also applied to Ensemble injection, where we have different BC applied.

Probably the software one year ago generates results more accurate results, becouse it ignores some of BC, now after fixing the defect it takes them into account, but you get less realistic results, becouse there are improper loads.

Please check the simulation which I provided below. I am still not sure about the 290 degrees BC on the white body, but you need to decide what is on the boundaries.

I also send you the e-mail from the support case, there we could discuss how the BC should look to solve it properly, I just need some additional data on how the process looks.

BR
Karol

[tiphaine_thomas]

The model represents the end of an injection and plasticizing unit of a plastic injection press. It is the part where the nozzle is fixed and which will then make the junction with the mold. The "Sleeve collar" element is therefore a heating collar which is located on the injection sleeve. So the constant temperature charge is necessarily inside. Maybe my convection and radiation loads are wrong.

I do not know if these indications can help you. I remain at your disposal.

[karol.suchon]

Hello Tiphaine, please send me the PM, I want to discuss with you the simulation settings.
BR
Karol

[tiphaine_thomas]

Sorry but I understand what is PM ?

[karol.suchon]

Private message 🙂