Influence of tensile gauge length on FS

[puli888]

Hi,
Lets say I have a FE model with approximate element size 15mm.It is non-linear dynamic model use LSDYNA.I do not have material stress-strain relation.So,I experimentally measured eng stress-stress data using 1D tensile test with gauge tensile length of 10mm,25.4mm and 50 mm with 6.5mm,25.4 and 50mm extensometer. I got three different failure strains; say 32% 28% and 26% respectively. Now my confusion is which FS I use in my FE model to get an accurate result. Any suggestion guy?
-Saravanan

 

[prost]

Wouldn't you normally use the mean (29%), but test your
assumptions by using 32% in a separate analysis, to see
how much an effect it has on the answer?

You mentioned the element size--shouldn't your results be independent of element size?

 

[nicola31]

The sensitivity of failure strain to gauge length is clear and the question of what to use as the basis for FE model failure strain must have been addressed at some stage.
However I have found no definitive guidance and in its absence I have tended to use the gauge length closest to my element edge length in the loading direction.

I would be interested to hear any other views.

NB: I do not agree with the approach of using the mean. If I have interpreted the original question correctly, we are not dealing with test variation here, it is essentially the same test but we are measuring different parameters (by assuming different gauge lengths). If we take it to the extreme of very large gauge length our failure strain would approach zero - clearly we would not include this in our mean.

 

[crisb]

The problem with the tensile tests is that the specimens will probably neck before failure, and therefore the maximum strains recorded are unlikely to be meaningful for an FE analysis (unless your calculation of engineering starins takes into account the reduction in area which is very difficult).