Thermal distortion tests of aluminum and stainless steel plates

An important upgrade to the STAR detector at the Brookhaven National Laboratory RHIC accelerator will be an electromagnetic calorimeter. One design being considered for this calorimeter involves cast lead modules covering Δφ = 6 degree and 0 ≤ |η| ≤ 1. These modules would consist of alternating layers of lead and sheets of plastic scintillator. The gaps for scintillator between the layers of lead would be created by parallel aluminum plates of thickness ≅6.6 mm = 0.260in. in the mold for the modules. These plates would need to be machined or ground to be reasonably flat, perhaps to ±0.003in., and of uniform thickness from plate to plate. These requirements are imposed by the need to remove the plates from the casting after cooling, and to have good uniformity of the lead layer thickness, which gives good performance for the modules as a calorimeter. Aluminum was chosen for the plates because of its high coefficient of thermal expansion. An important cost in this calorimeter design is associated with the machining or grinding of the plates to proper thickness and flatness. In most cost estimates, it has been assumed that the mold parts could be used many times. This note describes a simple test which was conducted to investigate possible distortions in the plates after repeated heating to temperatures at which the lead would be poured into the mold and cooling