Fatigue life Nf,50 & too early stops in 4PB tests

Dear colleagues, Some time ago I raised a question about the validity of the traditional fatigue life definition Nf,50. Recently I examined the data of 4PB testing on a mix with a special binder. I could not use the ACP-F model because all the 9 tests stopped before phase III started. I have placed some figures in the attached file. One of the problems is the standard setting for a 4PB test. The test will stop if the Smix has decreased to half its initial value measured at cycle 50 (ASTM) or 100 (EN) or the number of repetitions is above a standard-setting (e.g. 2,000,000 cycles).

My advice will be to carry on the test until the stiffness modulus Smix has decreased to 30% of the initial value. I'm rather sure this will be enough for most of the 4PB testing.

Looking forwards to your reactions,

Best regards

Ad Pronk