The objective of this work is to demonstrate the correlation between a simple laboratory test bench case and the predictions of the OOFELIE Multiphysics software in order to deduce modeling guidelines and improvements. For that purpose two optical systems have been analyzed. The first one is a spherical lens fixed in an aluminium barrel, which is the simplest structure found in an opto-mechanical system. In this study, material characteristics are assumed to be well known: BK7 and aluminium have been retained. Temperature variations between 0 and +60°C from ambient have been applied to the samples. The second system is a YAG laser bar heated by means of a dedicated oven. For the two test benches thermo-elastic distortions have been measured using a Fizeau interferometer. This sensor measures wavefront error in the range of 20 nm to 1 μm without physical contact with the opto-mechanical system. For the YAG bar, birefringence and polarization measurements have also been performed using a polarimetric bench. The tests results have been compared to the predictions obtained by OOFELIE Multiphysics which is a simulation software dedicated to multiphysics coupled problems involving optics, mechanics, thermal physics, electricity, electromagnetism, acoustics and hydrodynamics. From this comparison modeling guidelines have been issued with the aim of improving the accuracy of computed thermo-elastic distortions and their impact on the optical performances.