Not to be picky here, but just so everybody is on the same page ... The metal properties can degrade from the heat. This is not disputable. However, fatigue is not the right term. Metal fatigue is definitely a problem in some parts, but is unrelated to heat (unless the cycles are caused by cyclic temperatures -- not the case here). Metal fatigue is exhibited in metal parts that are subjected to cyclic strain cycles. Whether or not the cyclic nature is a problem for failure depends upon the level of strain. At some levels, the number of cycles can be infinite, without failure. At other levels, the failure point can be well defined by S-N curves, that is strain vs. number of cycles. For example, an aircraft wing is designed to remain safe under X number of takeoff-landings. Each t-l is a cycle, a cycle being one extreme strain level to another, then back again. These numbers have great significance on the life of the aircraft. Not to digress here, but sometimes terms are used that are not related to the problem at hand. I wish you well with your project, and can understand your desire to rebuild the machine. Obviously, time isn't critical for you, only parts costs.