Issue with a Custom Particle Property API (Particle Body Force) in EDEM 2025 during EDEM-AcuSolve co

Question

Hello,

I am running EDEM-AcuSolve coupled simulations that are similar to the tutorial (ACU-T: 6106 AcuSolve - EDEM Bidirectional Coupling with Mass Transfer). Instead of spray coating, the particles are initially given a moisture content and will get dried.

My simulation can run correctly on two older computers, i.e. moisture decreases with time as particles dry. However, on a new computer, the coupled simulation runs, but the custom moisture property remains exactly at its initial value throughout the simulation. There is no crash or DLL loading error. The same source code was recompiled on the new computer using Visual Studio.

I have the same versions for EDEM and AcuSolve, and was able to repeat the tutorial on the new computer.

Has anyone encountered issues where property updates fail specifically after moving to new hardware or a newer Visual Studio compiler? I am wondering if there are specific compiler flags (e.g., floating-point model) or environment variables that could cause the API to fail silently.

Any suggestions on how to debug this further would be greatly appreciated.

Yangyang_Shen · Answer

Hi,

Thank you for the suggestion to test the model in EDEM-only mode and for the advice on using the Visual Studio debugger.

I have followed your recommendation and ran the simulation in EDEM without coupling AcuSolve. In this case, the moisture content remained constant. However, I believe this is expected behavior since the mass transfer relies on the coupling with hot air in AcuSolve to drive the drying process, the particles wouldn't show a moisture decrease without that external trigger. Indeed I observed the same result when running the standalone EDEM portion of the tutorial (link below).

https://2025.help.altair.com/2025/hwcfdsolvers/acusolve/topics/tutorials/acu/acu_6106_intro_cfd_r.htm

As I mentioned before, I could repeat the tutorial successfully. One point of difference I've noted is that the tutorial uses the built-in "Spray Coating Update" model, while my simulation utilizes a custom particle body force model that I modified from CSprayPBF.cpp. Please see attached for the source codes.

CSprayPBF.zip

The most confusing aspect remains the hardware transition. This exact source code and setup produce the expected drying results on my older machines, but the property updates fail to register on this new computer (with the same versions of EDEM and AcuSolve, either using the DLL compiled on the older machine or a fresh compilation in Visual Studio).

I will try the debugging method as you suggested to see if the internal values are being calculated correctly before they are passed back to the EDEM memory space. I'll let you know what I find.

Thanks again for all your help and suggestions.

acupro · Answer

You may also try using the DLL compiled on the older computer - when running on the newer computer (if you haven't already tried that).

Stephen Cole · Answer

Hi,

I'm not aware of any specific issue that would cause this, if there is a memory error in the code then this could manifest in different ways.

Just to confirm if you load and run the model in EDEM only does it run without warnings, if you try running it uncoupled it may show any errors.  However if it loads and runs I would expect it to work in the same way, especially if the EDEM version is the same.

You could create a debug version of the code and attach it to visual studio debugger, this allows you to see the specific values which are been set to the code, this is often the best process for checking an issue.  If you do debug its best to create a simple case which is just two particles interacting, otherwise it can be very confusing on identifying what is happening.

The factory tutorial gives an overview of the debug process

https://community.altair.com/discussion/39622/edem-api-tutorial-6-debugging-snooker-factory