Gas in Thermal Stress

I am trying to characterize a flaw in a structure (a hole filled with air) for a thermal stress simulation. From a structural point of view, the gas has a Young molulus of zero which cause the computation to crash during the calculation of the inverse matrix. So to solve this problem, I've decided to replace the air by a teflon disc. For the geometry, I've used the ''Form Assembly'' option (instead of form union) and excluded the teflon so it's not attached to the rest of the structure (since its supposed to be a gas, there must be no structural contribution from the teflon). The problem is, from a thermal point of view, the teflon is insulated and I cannot add any thermal boundary condition such has a thin thermally resistive layer (say its not applicable) because, as mentionned before, in the geometry, the teflon is not attached to the structure.

What can I do to represent a gas, something that influence the thermal transfer but not the structural resistance?