Residual Stress from welding in FEA NASTRAN

[gdechaine]

I have a contractor of mine performing a FEA on a structure which experiences loading in all axis at several different locations. The structure consists of vertical I-beam support frames and 2" plate welded on to the supports. My question is what effects will the welded joints have on the overal stress analysis of the structure. Right now they are constraining the structure as if it is 1 solid object. Can someone please give me some advice as to what effects welds will have on the overall analysis of the structure? Fatigue will not be an issue. Just residual stress left over fromt welds. They are using NASTRAN for analysis. Does the program have properties which can simulate residual stresses resulting from welds? Advice from any structural and weld engineers would be much appreciated.

 

[GBor]

I typically look at the stresses in the Heat Affected Zone (HAZ) and compare them to knocked-down strength values. In other words, if I have a 36ksi steel that has a strength of 20ksi in the HAZ, I look to make sure that stresses are less than 20ksi in weld areas.

 

[hondaknight1]

I have to agree with GBor analysis of welded structures can be straightforward. But only if you have full penetration welds, good weld prep and dressing after welding. Then you can do as Gbor says and make sure the stresses are lower than some reduced amount in the welded areas and HAZ. Residual stresses are a non issue, when rolled and forged components (such as I beams) are tested for residual stress by hole drilling or electron diffraction it is found that they have stresses in some areas which are typically approaching yield stress. Any welding you do isn't going to make this any worse than it already is...