If you do a simple tension test on an un-notch specimen, the area under the stress-strain curve obtained shows the modulus of toughness not fracture toughness. In other words, this area is the amount of energy which is necessary to fail a specimen without any pre-existing crack. but to obtain the translaminar fracture toughness it's necessary to use a specimen with crack like the CT or single- or double edge crack specimens.

In my view, using a notched specimen, only the cross sectional area of the specimen subjected to laod will decrease as compared to the unnotched specimen and thus the load required to fracture the notched specimen will be less than that for the unnotched. As the the fracture toughnes is in the energy per unit area (KJ/m2), which should be the same for both the specimens.

Cicumferentially notched cylindrical specimens are sometimes used for fracture toughness measurement. However, the possibility of obtaining such specimens from composite laminates needs to be examined.

Fracture toughness is about unstable crack growth conditions. You need a crack starter. You can try a very small hole in your plate. The load displacement records can be analyzed for any pop-in effects. The crack size could be determined by Scanning electron microscopy.

It is obvious that fracture mechanics needs a pre-crack or notch in a specimen for determining its fracture toughness. Howerve the fracture in un-notched composite specimens is also triggered by the voids and cracks within the specimen (may be at microscale). Then why the simple tension test cannot replace the compact tension test.

Voids are only one of the many features that will lead to composite failure. Fiber waviness, early single fiber break due to local weakness, uneven cure and so on can all lead to failure without voids playing a major role. It can't be assumed that the failure is due to a void acting as an initial crack.

If a unnotched specimen is used, to location of the crack leading to failure is not known and it the energy for its initiation won't be known either. Moreover, you won't really have control over the opening mode.

I recall seeing notched beam bending tests if you want to try a simpler specimen, you may want to investigate that or side notched/open hole tensile specimens.

Fracture toughness is the resistance to crack growth. At the root of a notch/crack, there is stress triaxiality. Therefore, the crack grows from the notch. In a tensile specimen if there is no notch to start with, it develops during tensile deformation, i.e, necking which can be diffuse. This means the damage is developing in a larger volume of the material. Damage mechanics based evaluation of crack resistance is becoming increasingly popular now.

Notched bar tension tests have also been used for fracture toughness characterisation. This has the advantage that the specimen is in a generalised plane strain condition and the size requirements are much more relaxed compared to CT specimens, bringing down the material requirement considerably.

It is evident from the above discussion that in unnotched speciemen the notch location is not known. The crack location can be determined by in-situ tomography of the specimen. However, it will make the test too complex and expensive offseting the benefit of uniaxial tensile tests being simple.

Further, in the stress-strain curve, the global respeonse of an on-axis composite specimen show fluctuations when a crack starts to grow before its ultimate fracture. Can't we determine the energy required to initiate the crack by integrating the area under the curve upto that point. The behavior of on-axis laminates is usually quasi-brittle and no significant necking is observed in the specimen duiring tensile tests. Further, Can the open hole and notched beam bending tesst replace the CT tests?.

Vikram:

In my view, you have read the word Compact test as Impact test that is why you are diverting the discussion. Both the tensile amd compact tests are in static/quasi-static conditions. You are discussing the Charpy and Izot type impact tests for determination of dyanic fracture properties, which is a different scenario and will need a different thread to comment on.

Yes correct i misread the term. Now coming to the topic, IMO this can be checked only by performing experiments. However logically, energy required to move the crack would depend on the location of crack, orientation, type and potentila obstruction offered to crack movement. Needless to say multiple cracks/void present can again put your results in total confusion.