Structural simulation early in the design process can lead to prospective solutions otherwise overlooked. When developing a welded assembly, managing a viable simulation model presents the engineer with a unique challenge. In the next ten minutes we’ll explore a simple solution that provides design flexibility while at the same time maintains the integrity of the detail in the welded sub-components.
I love the WELD and MERGE feature management. I never thought to use the 'find' functionality in that way. I use merged models to run heavy fabrication simulations regularly and want to offer a word of caution regarding the additional weld geometry that you add to your merged model by use of fillets. Since your merged model already represents full penetration welds adding additional geometry will result in unrealistic stiffness at the welded joints (unless you are truly manufacturing a fillet weld over a full penetration bevel weld). Additionally, the curved geometry that fillets generate is also not very realistic, maybe at the toe of the welds but definitely not through the throat.
I use chamfers to simulate weld geometry along with specific sized gaps that take into account normal weld penetration which prevent fully merged plates. This, in a way, represents a worst-case scenario where you have gaps between plates and the load only transfers through the weld material. I only use fillets at the weld toes to represent post-treatment of the welds (e.g. grinding, tig-toe dressing, etc.) when a smoother load path is required from weld to base plate material. One last comment- any plates that are stacked one on top of the other will again result in unrealistic stiffnesses so increasing their outer diameter to match other features will further increase stiffness reducing the true behavior of the structure. The only time I leave the merged model as-is is when the plates generate a T-joint and the two fillet welds connecting the plates together are in total the same as the thickness of the perpendicular plate. When viewing the stress plots that result one would have to use their best engineering judgment to decide if high regions of stress are due to the merged geometry (i.e. singularities) or if you do have a load management issue resulting in material yielding...
Thanks again for the video.
p.s. great website... your portfolio is quite impressive.
Excellent! You are so right! The effects of local heat and residual stress due to cooling should also be considered.
Where would the world be without us engineers??!? (Who else thinks about this stuff??)