Simple Contact Analysis Yielding Inaccurate Contact Areas

[entalto]

Hello everyone!

I am trying to model the contact between a shaft and its housing. The shaft is "fixed" and the perpendicular load is applied to the assembly through the housing. The problem converges but not as I expected.

Both parts are made of Titanium so they are quite stiff. When I ran the same case using a softer material for the housing, this deformed and "adapted" to the shaft surface transmitting the pressure homogeneously.

Using Ti for housing, I obtain two lines of contact (high stress concentration areas) at both ends of the housing, which in my opinion is not realistic. The shaft seems to fail in these superficial areas with stress values exceeding the yielding limit of the material by far.

Case = static NLgeom = off
Parts = 3D homogeneous linear
Contact = Surf to Surf, tangential, Coulomb friction mu = 0.1
Contact = Slave adjustament = NO (I have tried with Specify TOl = 0, same result)
Contact = Interference fit = No allowable (I have not undertood it propperly)

Any help or advice willbe more than welcome.
Thank you very much for reading me!
Cheers!