Hello Carletes,

The piping system is like a spring.

When the pipe supports are moved ("pushed") by the expanding piping, they slide on some surface that resists the sliding movement by friction. When the piping has expanded to its hot position, the force needed to overcome friction is no longer applied to the support - but some of the force has compressed the "spring" (piping system) and that portion of the force of expansion remains. If there were no friction, the support would have moved further and there would be no force stored in the "spring" (piping system).

When the piping system returns to ambient temperature it first releases the stored force of (cold-to-hot) friction, and then again moves ("pulls" on) the supports as it contracts, sliding the support to a new location, somewhere between the installed position and the hot position. If there were no friction, it would return to the point from where it started.

But really it doesn’t slide smoothly, rather the expansion/contraction builds up the force (stores energy in the "spring") against the support until it overcomes the resisting force of friction and then the support "hops" to a new position (exhausting the force to a level less than the opposing friction force). This continues until the temperature of the pipe reaches a steady state (hot or cold). But, in either case there will be some "stored" force in the "spring" that is something less than the force needed to overcome the sliding friction and make the supports move.

The (coefficient of) friction that opposes the movement of the supports changes due to time, temperature and a variety of other factors. No two engineers seem to be able to agree upon what is a reasonable friction coefficient to use in the piping model. That being the case, I am sure you will have several responses to your question.

http://www.engineeringtoolbox.com/friction-coefficients-25_778.html

Regards, John.