By 1918 Heron and Gibson had designed, manufactured, and tested cylinders that consisted of open-ended machined steel barrels with an external thread on a portion of the upper end and a mounting flange on the lower end. The finned cast aluminium head, which was fitted with valve seat inserts and valve guides, was internally threaded. The pitch diameter of the internal thread on the head was slightly smaller than that of the external thread on the cylinder barrel so that the head had to be heated in order to allow assembly. This resulted in a joint that was mechanically secure at the cylinder's operating temperature and provided the best escape path for waste heat. In concept, if not in exact detail, the modem air-cooled cylinder had arrived, but not everyone was ready to accept it, possibly because of the “not invented here” syndrome prevalent in some companies.
Modern technology!
Still interested in finding a link to actual assembly of a modern cylinder. I dare say, little would have changed. From this snippet it would appear the entire barrel enclosed by the casting becomes an interference fit...including the threaded portion.
It may be a long, long time ago for me since I graced the halls of the Mechanical Engineering department of the old QIT but I still get curious about how things work.