Going to assume this is largely meant for looking at delamination with laminates, which is a certainly a big topic by itself:
If your material is a well researched (i.e aerospace certified) autoclaved UD prepreg then academic papers are probably your first point of call for values. However, I would urge catuion on taking experimtnal data relating to composites from the literature at face value and will quote one of Prof Christensen (who with Tsai and Hashin are arguably the fathers of composte failure modelling) on this:
“In all fibre composites assessments, the quality of the data is the matter of the first importance. It is not difficult to find examples of well-regarded data that differ by as much as 50% or 100% or even more. Great differences in failure data values for composites is the rule, not the exception”
If you're using a liquid processing based route (RTM, RIFT, VARTM, vac bagging, etc...) and/or a woven or multixial fabric then its unlikely you'll find suitbale properties given the number of variables that will affect the end properties of the composite. In this case you'll need to resort to experimetnal methods (ASTM methods are a good start for basic tests), e.g. the double cantilever beam (mode 1) or end notched beam (mode 2) for which you obviously need a test machine, sutiable fixtures for the machine and preparing the samples plus a device for recording crack propagation (e.g. travelling microscope with trigger or videocamera...use a videocamera synced to a data logger...trust me on this ...or your back, eyes and sanity will pay the price!). Once you've got your experimental data and backed out your fracture toughness values you'll then have some starting points for your cohesive paramaters. Initiation and propagation fracture toughness props come from your fracture test and the data reduction methods (you can cross check and calibrate FEA models using contour/J integrals and such in ABAQUS too if you want, not sure about ANSYS); interlaminar stress can start from ILSS testing (means of directly obtaining interlaminar tensile strength are challenging and questionable, backing this value out from ILSS and other properties is where we start); penalty stiffness can either simply be a value that works or the thickness of the bond line (e.g. take an average from microscopy images) divided by the relvant modulus, depending on your view as to whether this should have a physical basis or not.
With these values you can then start doing a sensitvitiy analysis with various mesh sizes, traction-speration laws etc... until your FEA model replaciates your experimetnal test results. You can also do a bit of optimisation work to speed up models; e.g. increase mesh size and artififaclly knock down interlamiar strength (apparent strenght will increase significantly with element size for a given strength value) to maintain congruence with experimental results. Once thats done; you can start scaling up your models to component and structural models as the scenario requires.
The above is just for starters, scenarios where rate dependence (which pretty much always requires consideration), sandwich constructions, fatigue etc... come in can greatly complicate both expreimental and numerical work which is why there is so much on going work in academia on suchtopics.
Basically, getting fracture toughness propertiies and then usefully using them within FEA is a lot fo work when compared to simple strength and modulus tests which is why such testing and analysis work is often the preserve of PhD students or Post-Docs. Such testing is a lot more involved than regualr strength/stiffness testing and calibration work but I'd suggest that many, if not most, composite structures that fail are 'killed' by delamination so obtaining good experimental data relating to delamination is of great importance. After all, no matter how adavanced your FEA tools are; rubbish in =rubbish out.
![Autodesk Design & Make Report](https://forums.autodesk.com/t5/image/serverpage/image-id/1350519iCB55845EBFC43016/image-size/large?v=v2&px=999" "Autodesk Design & Make Report")