Learning Forum

[AndreaPreziuso]

Hi everyone!

I made a traction test of a composite specimen. I plotted the stress-strain curve and i want to identify the point where the failure occurs. I inserted inverse reserve factor but i have some questions about how it works:

The IRF is the inverse of the safety factor, so it's the ratio between max stress and stress limit, but, if the IRF is related to the entire composite which are the stress limits? How can i use IRF to determine the failure stress of the composite?

Another related question: i made two simulations, one with the damage initiation criteria and evolution law, and one without. If the damage occurs only when limits defined by a criteria are rached (I used Hashin), why the slope of the two curves (the one from damage and the one without damage) are different even before reaching the limit?

Maybe these are dumb questions but i hope you'll help me.

Thanks for your help!

[Sean Harvey]

The stress or strain components (normal,shear) are compared with their respective orthotropic limits, and the largest value of IRF is shown. I am assuming you are using composite failure tool in Mechanical, in which case the properties window will report the critical failure mode. If you pick a simple failure criteria for example max stress, then it will tell you the direction and mode, as an example. With Hashin or others failures that have intereaction the failure can be combined so it may not be so clear to know what is the failure stress. A good exercise is to report the ply stress and compute IRF by hand for each direction/mode (using max stress as it is simple). Be sure to request the stress results in the fiber/matrix orientation as these are oriented with the ply orientation and can compare with the stress limits. Once you understand how it works, you can change to Hashin or other.
The slopes you may be seeing could be caused by damage initiating right from the start. My first guess is you should check the time stepping to make sure you are applying the load in small increments. Auto time stepping with large max substeps, and at least a decent amount of initial and min substeps, so the solver is not applying too much load from the onset. If you are doing this already, then please post an image of the two curves. I assume you are reporting the load/deflection curve.
Thanks so much
Regards Sean