Crack Growth Testing Beach Marks

Dear Fellow Machanicians,

I want to make some beach marks in fatigue cracks – what are the best loads?

To verify crack growth modelling we have put some notches in square (10×10mm) section steel tensile fatigue specimens – so there's a reasonably controlled initiation – at least location.  In some the notch is across one side in others it's just on one corner, 0.5mm deep. I have tried one and didn't get the obvious beach marks I was hoping for.

We would like to grow the cracks at fatigue stress that is somewhat like the service levels, about 250MPa with R-ratio 0.05.  So I'll set it to do a few thousand cycles like that then do a few thousand at a lesser load. In the first test I used 1/3 the load (R=0.05) for 1/5 the number of cycles in the full load block.

I've been searching but not found any obvious guidance – does anyone know what works or where to find out?

Regards,

Robert