In my view, there are three methods to obtain fracture parameters for your simulation:
1. Experimental testing as mentioned it in my first response to your question.
2. Published literature if they are using the same material as that of yours.
3. Numerical calibration: This is a parametric evaluation of the fracture properties through simulation. In this case, you assume some realistic values in the simulation, and then compare the results with experiments say tensile tests. If the global response of your simulation matches with the experiment, then you select those fracture parameters as the final ones. I hope this will clarify your ambiguity.
Usually double cantilever beam (DCB), End Notch Flexure (ENF) and Mixed Mode bending (MMB) tests are used to determine the fracture toughness of resin-rich interfaces in composite materials. For fiber fracture in composites or metallic fracture, Compact Tension tests can be used. A simple tensile test gives the energy absorbed by the materials till their fracture. In Abaqus, fracture toughness of a material can be determined by using Virtual Crack Closure Technique (VCCT) as well as J-Integral.
In the Linear elastic fracture mechanics (LEFM) based simulation for evaluation of J-integrals based on stress intensity factors, you will only need Young's Modulus and Poisson's ratio of your material. So the tensile test data may be enough up to this extent. The crack you will have to define in the model. The detailed procedure for J-Integrals determination with 2D and 3D models is given in Abaqus example problem manual titled as:
1.4.2 Contour integrals for a conical crack in a linear elastic infinite half space.
Look at the example and then formulate your problem. I hope this may solve your problem.
Both the K (stress intensity factor) and G (Energy release rates) are related in LEFM for plane stress and plane strain conditions (see any book on fracture mechanics). The crack will propagate in a material when G is equal to or greater than Gc ( or K equal or grater than Kc) for that specific material under given loading. The FE models based on J-integral or VCCT only determine K or G for the structure under given loading conditions. However, the fracture toughness Kc or Gc are the material properties which can be determined experimentally. For this purpose the ASTM compact tension tests can be employed. In Abaqus, Gc will be required as input parameter for damage/fracture simulations.
I saw the above discussion. To directly answer your question: No. you cannot determine fracture toughness of a material using Abacus or any other FE based tools. You need to perform fracture mechanics tests to get the fracture toughness value.
In my view, there are three methods to obtain fracture parameters for your simulation:
1. Experimental testing as mentioned it in my first response to your question.
2. Published literature if they are using the same material as that of yours.
3. Numerical calibration: This is a parametric evaluation of the fracture properties through simulation. In this case, you assume some realistic values in the simulation, and then compare the results with experiments say tensile tests. If the global response of your simulation matches with the experiment, then you select those fracture parameters as the final ones. I hope this will clarify your ambiguity.
Usually the fracture toughness is an input to Abaqus that you use to determine loads at which cracks will initiate or propagate.
In the case of Fibre Reinforced Composites you will need to perform the following tests:
- DCB for mode I fracture toughness (ASTM D5528 or ISO 15024)
- ENF for mode II ( ASTM D7905 )
- MMB for mixed mode delamination (ASTM D6671)
If you have experimental data for any loading situation where you can observe a crack propagating as a function of the applied load, you could indeed work backwards and the fracture toughness would be the value which would reproduce your physical test.
However, it is very difficult to do that because that would require your experiment to be either pure mode I or II. If not, you would need to know what your mode mixity ratio is.
Therefore, if it is not practical for you to perform one of these standard tests, I recommend taking values from literature.