Bill Schwab wrote:
Chris,

If you can visualise the type of ideal

pin joint I'm talking about, the horizontal component of the tension
is not applied to the column at all. It is transmitted around the
column through the joint.


That is not at all what you drew. If you intend to transfer only a
vertical force to the top of the beam, you are going to need a slot or
horizontal roller on top of the beam, with the likely outcome being a
statically indeterminate problem or something that leads to a
contradiction because it is not adequately supported.

Imagine the front forks of a bicycle. Now imagine two sets of forks, one
slightly smaller than the other so that it fits snugly inside. Now make
the column slightly narrower than the smaller set of forks and drill a
hole through it. Attach one set of forks to the tension member, the
other set to the roller support I described, then pass a pin through
both pairs of forks and the column. This joint would behave as I
explained, and does not in itself make the structure statically
indeterminate. Instead it makes it easier to analyse.

Chris