Usually welded joints are something that needs to be usually analysed a lot.
The usual thing to check is the butt welding and how it behaves with the structure.
I have stumbled in my work to similar situations, where unknowingly the welding was too strong and has ripped next to the welding due to heat zone changes.
Currently Prepomax does not show weldment stresses, or am I incorrect?
Is hot-spot analysis possible to implement?
Here are also studies from abaqus for possible implementation:
For now, I’d rather vote for more general features such as transformation of results to local coordinate systems and stress linearization. They could already help a lot with weld analyses. When I performed such studies in the past using Abaqus, I transformed the stresses to local CSYS and evaluated them according to engineering code. I also used fe-safe for more detailed fatigue assessment. It utilizes the structural stress method which is specific for this software but there are also other approaches that could potentially be implemented in the future.
Using fe-safe or Abaqus would require aquiring license at this point.
The theory taught of welding assesment mathematically was too simplified at school and did not take in consideration the assembled structure with different weldments in different places.
The hot-spot method on solidworks simulation was something that I got to try. But the values were quite wild, depending on the mesh and where the welds were positioned. It would always exaggerate the stress value quite high.
Otherwise there is probably the nominal stress method which is usually used on boiler weldings, notch stress method and structural stress method?
And the latest version is “Mesh Insensitive Structural Stress Method for Fatigue Analysis of Welded Joints Using the Finite Element Method”
Of course, it’s out of the question. What I meant is that other approaches (not limited to fe-safe) could be implemented.
This article describes the structural stress method developed by Dong at Battelle institute and implemented in fe-save Verity software. The other common methods are:
nominal stress
effective notch stress
hot spot stress
and those could be considered for the potential implementation.
interesting features, seam weld need to be modeled and generates automatically. it is fairly new implementation in commercial ones (ANSYS).
this model based on Fayard etal (2003) and Turlier etal (2010), however it is only for classical shell element. CalculiX shell is expanded to solid element during solver processing. so, these model may not applicable and lead to erroneous due to knot existence. another approach is required if the problem occurs.
i hope so, but it seems the original developer not interested in. actually i was discussed directly since early version. CalculiX has different goals using expanded shell and beam approach for compatibility reason. this is advanced, unique and challenging for many ordinary FE user such as myself.