pertaining to this quote - " rather than Rankine (shear failure in whole mass of backfill). Rankine assumes level backfill and frictionless wall."
I agree and did not mean to imply otherwise. If you refer to my post, I said to determine Ka using any method that you chose. The justification of using a 1.5 factor is based on the simple observation of how the pressures vary under a horizontal backfill scenario. I realize this is different from the problem statement, but the use of a 1.5 factor would apply whether it's horizontal or sloping. Maybe I'm not making sense, but I agree that Rankine is for horizontal. You can take the Ka log spiral and multiple it by 1.5 to get an at-rest pressure for your design.
f-d