Learning Forum

[Jonas S]

Hello,

I am looking for background information on DED simulation in Workbench Mechanical. More specifically on the use of mathematical formulas and the E-Birth / Death method in the APDL script, which is created during the solution (ds.dat).

I have opened the ds.dat file with a text editor, but I cannot find any references to mathematical formulas for calculating the latent heat using the specific heat from the material data. Furthermore, the DED_Simulatin_Guide claims that the DED simulation in Ansys Workbench Mechanical is based on the Element-Birth / -Death method. However, the commands ekill and ealive cannot be found in the APDL script (ds.dat file).

Therefore, my question is whether the ds.dat file actually contains the complete APDL script, which is solved with the MAPDL solver, or whether only an excerpt is saved here?
If so, where can I find the corresponding commands (ekill and ealive) as well as the mathematical consideration of the material data?
If not, which commands in the ds.dat file describe the aspects I am looking for (E-Birth / -Death, latent heat, heat conduction)?

Thanks in advance 
Jonas

[John Doyle]

The ds.dat file is submitted to the MAPDL solver in a batch run.  It contains the complete script for running an AM simulation.  While EKILL and EALIVE are not executed directly in this script, the same technology is being used in an automatic process-sequence macro defined by the  AMSTEP,,,command.  

[Jonas S]

Thank you for your answer.
Unfortunately, I cannot find the command "AMSTEP" in the ds.dat file either. Only the command "AMsolve" and the individual am_loops appear in the file. Does the AMsolve command correspond to AMSTEP?
Furthermore there is still the question, how  latent heat and heat conduction are taken into account in the DED simulation? 

Regards 
Jonas

[John Doyle]

Sorry for the confusion.  My answer would apply to LPBF.  DED is more involved.  It typically requires G-Code to define the element "clusters" to be strategically activated (i.e. EALIVE) to simulated the laying down of material.  From  (Chapter 2: Implementation of DED in Mechanical (ansys.com)

Simulation of the Direct Energy Deposition (DED) manufacturing process requires that the analysis follows the print process itself: weld track-by-weld track solidification of the part. Since the thermal (temperatures) and structural (distortion and stress) physics are largely uncoupled (that is, a weak coupling), we can simulate the thermal phenomena first and use those temperature results in a following structural simulation.  From the DED docs...

In a DED process simulation, the model evolves over time, that is, elements are added. We actually mesh the entire part first with Cartesian or Tetrahedron elements and then use the standard element birth and death technique to activate element clusters to simulate the build progressing, where a cluster is defined as a portion of the weld track. Additionally, the relevant boundary conditions also evolve such as thermal convection surfaces. The build step is complete when all the element clusters have been activated (made "alive").

The analysis times and time stepping are also driven by the process parameters and are not known a priori. These details are all handled internally during the solution.

Even though a DED process and a PBF process are similar in some ways, the DED process needs to have a much more detailed level of simulation. A layer-wise simulation is not accurate enough since the velocity of energy deposition is higher in a DED process and so simulation along the weld track over time is required. As the PBF process simulation uses abstractions like super layers (lumping multiple layers for simulation) and layer-by-layer addition (instantaneous element activation for an entire layer), the DED process simulation uses the real welding seams and an abstraction known as element clustering. Clustering is used to split weld lines into smaller pieces of mesh, called clusters, that are exposed to one temperature in a time step. Thus, each cluster is a portion of a weld track that is instantaneously activated using the element birth and death technique. Element clusters are deposited following the deformed shape beneath them rather than in the original undeformed location. You control the size of the cluster with a Cluster Volume option. The solve time and solution fidelity are directly related to cluster volume. The building order of the part can be defined by either applying the G-Code machine file or by using a manual approach of carefully-defined named selections.

[Jonas S]

This is exactly what my questions are based on.
"...use the standard element
birth and death technique to activate element clusters..."
Where can I find the corresponding commands in the APDL script? (EKILL/EALIVE not present in the das.dat file)

"Phenomena such as phase transition will
be captured with the capabilities of engineering data or imported materials rather than simulating these
phenomena."
How are the material data taken into account? (for example, there should be a formula for calculating the latent heat as a function of the specific heat capacity from temperature dependent materialdata, ????ℎ = ∫ ???????? dT)

Regards