Effect of composition on thermal expansion of alloys used in power generation

A study based on published data was conducted of the effects of chemical composition on thermal expansion of several groups of alloys: austenitic stainless steels, Ni-base nonmagnetic alloys, ferritic and martensitic irons and steels, duplex stainless steels, and FCC magnetic alloys. Computer regression analyses were performed on the first three of these groups to establish models to predict the mean thermal expansion coefficient ( alpha ) from the composition. The models predict alpha with a standard error of 0.19-0.23 x 10 exp --6 / deg F, which is comparable to the standard error of experimental measurements of alpha . The most influential elements for each group are Ni (for decreasing alpha ) in the austenitic steels group, Mo (decreasing) in the Ni-base nonmagnetic alloys group, and Cr (second-order decreasing effect) in the ferritic and martensitic irons and steels group. The models should be useful in system designs involving combined use of austenitic and ferritic alloys in high-temperature structures, such as for maintaining clearances or interference fits or for minimizing cyclic stresses. The equations could also help a metallurgist develop an alloy with a specified thermal expansion coefficient. 25 ref.--AA