Coupling spring properties for piles in liquefiable sand

I’m having a tough time matching p_ult (in flac it is Fsmax/L) using only the frictional component with c_s = 0. If I only use the frictional term with c_s = 0 then the calculated p_ult is an order of magnitude smaller than what I get from API curves. I can make up for that by using c_n, but in some cases for sand I was hoping to keep p_ult proportional to sigma_m so that excess pwp generation would scale down the p_ult value. I am not able to match API p_ult for some reason. I there any reason for this?

I am not sure I have a FLAC-related answer here but I am confused with what you are targeting and using. p_ult and API are for p-y response yes? So the priority should be your normal components not your shear ones. You are mentioning p_ult with Fsmax… did you mean t_ult with Fsmax or p_ult with Fn_max? Did you mean phi_n and c_n = 0? Now, regardless of terminology, I see no reason why a link with a frictional component (regardless of cohesion) would not pick up the dependency on the confining stress of the neighboring zone. Also you mentioned F/L which is the force per segment length but this force is per circumference so are you tracking the sigma_m * perimeter in your x-axis?