Stiffness Change with Temperature while under compression

[Amdub]

I am modeling a gasket that is initially being compressed in a threaded joint at room temperature and then undergoes a thermal change to -40 deg C. I have created an FEA model to simulate this and extract the reaction forces and contact pressures generated by the gasket at the two temps. I am also accounting for some gasket shrinkage during the thermal change with a thermal expansion/contraction coefficient. In the model, the reaction forces increase dramatically after the step change to -40 deg C. The thermal contraction effect seems to be minor and completely dwarfed by the increase in stiffness at -40 C. I understand the model is predicting these results because I am telling it to use stiffer material properties at -40 C, but what I'm struggling with is if this phenomenon is really physical. Does it make sense that the reaction force would increase in a compressed gland 2-3X just due to temperature change?

Thanks

 

[Compositepro]

No, it does not make sense. If you cool a rubber gasket clamped between metal flanges, the contact pressure can only go down, because of the higher CTE of rubber.

 

[Amdub]

Ok, but there is also a stiffness change in the material with a change in temp. How does this factor in? The model is effectively using a higher spring rate at a lower temperature and this increase in stiffness is dwarfing the effect of thermal contraction. Intuitively it does not seem right that the reaction forces will go up, but I understand why I am getting this mathematically. Would compression set be more severe at lower temps? This would be a loss in stiffness+thermal contraction which makes more sense.

 

[Compositepro]

Compressing a cold rubber is not the same process as cooling a compressed rubber. The constraints on the rubber are also important. If the gap is fixed and the rubber cools, contact pressure must go down. Modulus is not an issue is CTE shrinkage.

 

[acspain]

Amdub,

Be careful of the material glass transition temperature. As you are mostlikely aware FEA can not model material phase transitions. Most rubber formulations transistion to a glass prior to -40. Even if the Tg listed for the rubber is below -40, the transition from a rubber (very viscous solid) to a glass occurs over a temperature range. Therefore the modulus is going to start increasing prior to Tg.

I agree with Compositepro, especially when you consider that in reality a rubber gasket is going to leak at cold temperatures not hot temperatures due to the coefficient of thermal expansions involved in the assembly, therefore contact pressure is not increasing.

Good luck.