I'm working on a design using a planar cam, where the cam follower follows one circular path, them moves outward to another radius, then back in again. I have circular arcs for the inner and outer segments trimmed to where the constant radius sections end, and I added a NURBS blend curve to join them... it curves out very slightly coming off the inner arc, then bends around to be tangent to the outer arc.

The question is, will a NURBS blend give the minimum acceleration to the cam follower or is there a better way to construct the path?

Yes, certainly I can use lines and arcs to get the same effect; my question was more about the underlying math and what will give the result while minimizing acceleration and pressure angle. Going from an arc to a straight line, or one arc to another, is an abrupt change in acceleration, which is never desirable. A spline by its nature gives a varying curvature, which I think would be the lowest energy curve.

On the CDE add-in "tools" there is a cam generation function. It is at the bottom of the "other" fly-out menu.

Typically you would graph the angle to displacement desired. Plotting the points radially would define the cam. A spline through those points is the cam outline. A spline is the lowest tension geometry through any sets of points.

The CDE function "create cam function" appears to be doing just that but more automated.

Thanks, I'll have to look at the cde cam function. At this point I'm mainly exploring the machine concept, not even sure a cam will work in this application.

In most cases (and most places on this machine) we do, too, but in this case I need to move 12 part holders with suction cups on a dial plate radially outward during one segment of the rotation so I can camera inspect the underside, then back in again. The motion is fairly simple, loads light, and I need to keep the rotating mass as low as possible.