Originally Posted by jur

The original Swift seatpost wall thickness is 3.2mm. That means the moment of inertia for the two cases differs by 30%, I take that to mean the Dahon post would bend 30% more than the Swift for the same applied stress. I think perhaps this was the reason for your frame's failure, rather than the tube diameter. I think it highly unlikely that a seatpost that is marked as 34.0mm would be slightly less, especially in seat posts where this is a critically controlled quantity.

I have ditched my earlier shim design in favour of a heavier, stronger design. The first shim was simply a section of 34mm tube hollowed out to a wall thickness of 1.2mm, to shim up the 31.6mm seatpost to 34.0mm. The (mental) problem I had with this design, after satisfying myself it would never shear off, was one of elastic bending. The manual that came with my frame says that the seat post is an integral part of the strength of the frame, so I thought the thin-walled shim would be unable to provide the required bending moment at the junction of seat tubes.

I therefore made a new adapter, not a shim this time. Starting with the same tube material, I hollowed out just enough to allow the Ritchey seat tube to tightly slide in up to its minimum insertion mark. The rest of the tube is the original 3.2mm wall thickness. The wall thickness at the seat tubes junction is 3.2mm since the Ritchey post is too short to reach this point. The section extending into the bottom section seat tube is longer than the original post, so I have even more strength further down than with the original Swift post.