One of the methods that I have seen to be extensively used is to employ Repliset®. Taking Replica is typically used to measure the crack size, especially to understand the crack growth behavior. It is employed in an interrupted test, where the fatigue test stops every n cycle and a replica is taken. The set of sequential replicas are then viewed under a microscope to monitor the crack size during its initiation and growth. You can read about the procedure in the following articles:
Some commonly used techniques for measuring the crack length as a function of fatigue cycles are DCPD, ACPD and digital image correlation technique. Other techniques such astraveling microscope, displacement/CTOD gauges, photographic technique using some grid lines are also been in use.
Dear Mohammad Mahtabi and Sajith ,Thank you for your answers , but in these techniques , are including with out remove specimens from fatigue device and check crack , then return ?
In the Replica technique, one should stop the test periodically, but there is no need to remove the specimen from the machine. I believe in the methods that Sajith recommended, the specimen can be still on the machine. I know that Digital Image Correlation (DIC) is a remote method that does not even need to stop the test. A brief introductory text about DCPD and ACPD can be found in:
I am interested in formulas to estimate the crack length (CT specimen) given the fatigue machine strokes (typical displacements of the pistons). I found this following formula in a literature paper, but it is not clear if the "control displacement" can be the stroke or an additional strain gauge. Unfortunately I have not the ASTM E647 yet. May you help me in understanding this, if you please?
In Figure (Kim et al. 2015):
a: crack length
W: specimen width
E: elastic modulus,
ν: displacement between the measurement points,
B: specimen thickness
P: force.
Ca~Cf: constants determined from the measuring location