Yeah, I see what you are saying- the geometry is right. But I'm not sure that it will be as rigid as the geometry suggests.

For example, the horizontal displacment caused by flexing in the bent section would be very close to zero if the bends were close to 90 deg., and the fit between the rod and the supports would have to be very tight and very rigid. Even with that, the rod is still free to bend upward (axis out of the page in your diagram) to allow the flexing.

This is probably getting a little too theoretical. I'd be curious to know if anyone has tested a set-up like this to see what the flexibility is.

- George