Hi thanks for your reply.

The stress ration is 0.1 in my case. I tested 55 samples of different notches at same nominal load. As nothes are different the samples will fail at different levels. My understanding was the resultant SN curve which I get should be considered 50 % Probability of survival / failure.

Now the tensile strength for the specimen (Al 2618 T6) would be 420 MPa which coresponds to 97.5 % probability of survival as per FKM standard. The resultant SN curve I got, should it be treated as 50 % probability of failure or 2.5 % as tensile stregth values corespond to 97.5 % probability of survival for the material. I know if I would have used mean tensile stregth value then 50 % probability of survival/failure for SN curve is justified or there is no need to draw relation between probability of static tensile strength value used and the SN curve generated ?