Since this is a compressive test, I should have said to insert a Minimum Principal Stress result and compare the minimum value (which will be negative) with the Ultimate Compressive Strength, which should have a minus sign on it also.
If the materials have larger magnitude (more negative) values of Compressive Strength in the cold than at room temperature, then a higher force would be needed to create a value of Min Prin Stress that equals that strength. Do you have temperature dependent properties for Young's Modulus? That will change how much stress is created for the same amount of force if it changes with temperature.
The most inaccurate part of this simulation is the Fixed Support and the Force on the top. Real samples will be placed on plates so that the face touching the plate can expand as the stress increases due to Poisson's Ratio. A more accurate model would use frictional contact to press on the sample. One way to reduce the modelling effort is to use a Symmetry Plane and slice the cylinder to half the height. That way, you only need to model one plate. Another simplification is to use two more planes and slice the body into 1/4 of a cylinder.