W. H. Heydt,
Centripetal force, actually.
I was certain someone would pick up on that when I wrote it. They always do. It's some silly thing kids get taught in school.
You know, if you tie a rock to a rope and get it spinning around and around over your head, that rope does not know or care if it's "centrifugal" or "centripetal" that is stretching it out. Cut the rope and tie a spring balance in there and it will measure the force. It cannot tell you which way it is going either.
So, I say, meh, "centripetal", "centrifugal", makes no odds. Just be sure to get your signs right in any calculations.
What one should really do is for a vector in 3D space from the three components of acceleration reported by the accelerometer and check that the magnitude of that vector is small enough to indicate a free fall condition.
Amag = sqrt(Ax^2 + Ay^2 + Az^2)
Memory in C++ is a leaky abstraction .