Using Invar material in Septum Polarizer and Feed

Hi. For a project i had to design a septum polarizer. But they wanted it to work from +55C to -30C. If i hit these limits my crosspol drops below 30 dB if i use Aluminium.So I found this Invar which is %64 iron and %36 Nickel and has a very low thermal expension coefficient.Here is the Link http://www.nickel-alloys.net/invar_n...ron_alloy.html. Aluminium has around 23*10^-6 expansion coefficient while invar has around 1*10^-6. I just wanted to ask if anyone used in their projects or is it usable in EM applications?

Thank you.

Invar has high resistivity similar to stainless steel and won't be used for waveguides without galvanic silver or gold plating. It's main application is for the design of metal-ceramic or metal-glass vacuum sealings.

Isn't it sufficient to make all waveguide and polarizer parts with matched thermal expansion coeffcient, e.g. from aluminium?

When i ran simulations for both aluminium and invar for septum only, I saw only 0.1 dB loss at S31 and S32(3 is the common port, 1 and 2 are input ports) for changing material from aluminium to invar.For a good measure i'm going to plate it too with silver. The thing is Since im gonna use this septum for a wide bandwidth, at the start and end of the band i get bad crosspol values because of thermal expansion affects the polarizer step sizes.