Finite element analysis

I'm clinging on to the top of Cangarda's main engine with one hand (the crocheted glove is a style thing) while leaning back to take this picture. The point of this exercise was to get a better 'grip' on the geometry of the cylinder head for a finite element analysis of the pressure forces. Our first cut at determining the stresses in the head was to assume it was a flat plate. A simplification that was ...simple, but clearly incorrect. The stress result was ~9600 psi for the test pressure, which for an old casting of unknown quality, could be too high...it could fail.

So, we made a 3D, virtual solid model of the head (omitting the stud on the top, which is probably not important).And, after applying simplified boundary conditions, subjected it to the virtual test pressure force.

This is a plot of the Von Mises "stress" for the test pressure. The result is that the predicted stress is actually ~3700 psi...a stress which is much less likely to result in failure.

I put "stress" in quotes since it really isn't a stress. It's a numerical way Von Mises, whoever he was, developed to combine the three principal stresses into single number, which he then compared to the yield stress of the material.