Yes, as Dravis says, more detail is required in order for us to diagnose your problem. As it happens, Dravis has already pointed out the most likely cause which is the previous layer having cured too far (past its B stage) in order for the new layer to cross link with it chemically - this means you're reliant on mechanical bond which then means that surface prep (keyed surface) is essential. Provide as much info you can as to your process, other materials involved, work environment and timings and we should be able to put a finger on the issue. When used right the inter-laminar bond between layers of XCR should be excellent but it is a fast curing resin and if you miss the window then the bond strength will be negligible without proper prep.

What you're describing really doesn't sound consistent with the way the resin should perform. I appreciate that you are a painter and so you are confident with your mixing but no matter what your level of experience it's always possible to make mistakes (we make them occasionally and we work with these products every day). If you are applying subsequent layers onto tacky resin then the new layer should crosslink properly with previous layer and the bond between the layers should be better than you can achieve by allowing the resin to cure and then keying it back.

The only two logical explanations are that the resin has either cured too far to crosslink (therefore you need to apply your next layer sooner) or there is some surface contamination on the previous layer which is interfering with the next layer's ability to bond to it - this is likely to be caused by insufficiently thorough mixing or inaccurate mix ratio. I would suggest triple-checking your mix calculations (maybe you could share them with us), making very sure you're mixing thoroughly (I suggest using a double-potting technique) and also reducing the time you wait before applying one coat onto the next. Correctly followed, I would be staggered if this does not resolve the issue.