Freeform RPS Issue

[[Ry...]]

Good afternoon, I have a free form geometry with an RPS alignment that has me stuck. Unfortunately I can not share screen shots or specific dimensions, but I will lay out the scenario to the best of my ability.

Imagine taking a pipe, cutting it in half, and then pressing down on either edge to make a wide U shape when looking from the -Y direction. From the +Z it would look like an oval. The perimeter is normal to the Z axis (thankfully!). The RPS has 3 points on the half pipe surface to set Z, 2 points on the -X surface to set X, and a single point on the end to set Y.

I am able to use the "cylinders" on the ends of the oval, the -X wall, and one singular best matched by eye point on the top surface to create a rough X,Y, Z - With X and Y being fairly accurate and true. A CAD model is being used.

I have created my Z1, Z2, Z3 points using Space points and a basket array based on [BSC] X and Y positions. X1 and X2 positions based on space point basket array [BSC] Y positions on the -X edge. Y position based on space point basket array on the tip of the part. To set Free form RPS Z1, Z2, Z3 have Z checked, X1 and X2 have X checked, and Y has Y checked. I have it set to iterate 3 times.

The question is, how do I make sure the Z1, Z2, Z3 iterate to find the Z value based on the [BSC] dimensions of X an Y? It seems to be iterating to find the best Z value [which I dont know what that should be since Z was arbitrarily lined up by eye in rough alignment!] I need it to find best Z value when best fitted to the BSCs for X and Y which are reasonably accurate from rough align.

I apologize for the novel, and hopefully it makes sense to someone since I cant share images or models.

[[Ry...]]

Update: I changed my iterations from 3 to 5 and it aligned the model significantly better! Now I have a number that I believe. Still begs the question above, how do I make sure it finds the Z values based on the best X and Y CAD fit to basics?

[[Ma...]]

Wouldn't start system help with that?