Hackworth valve gear etc.

As the build notes for the engine mentioned the inherent inaccuracy of the 'modified Hackworth' valve gear, I thought it would be interesting to simulate the gear - here are the results.

This first clip shows how the motion works going forwards, as seen from the right hand side of the loco.  (I am using my own terms in the description, and I have made some assumptions about the dimensions of various parts).  The 'connecting rod' from the crank to the piston slide is shown in blue.  The video clip starts with the piston at bottom dead centre with the eccentric at top dead centre. As the eccentric goes round, it drives the motion with the 'eccentric rod' linked to the 'motion arm' .  The motion arm is attached to the 'reversing arm', which is as far back as it will go for the engine to go forwards.   The 'valve rod' is attached to the eccentric rod not far from the motion arm, so that the resulting movement at the valve slide is a combination of the vertical position of the eccentric (matching the crank position) and the movement resulting from the swing of the motion arm (out of phase with the crank).  The movement at the valve slide is such that when it is up, the piston will be forced up by steam from the slide valve; when the valve slide is down, the exhaust is open to let the steam out as the piston goes down.  In between, the steam expands to provide force for the engine.  

The next clip shows the engine in reverse, with the reversing arm as far forward as possible.

The graph shows:

- in mid blue, the position of the piston slide;

- in red, the valve slide positions when going forwards;

- in light green, the valve slide positions in 'neutral'; 

- in purple, the valve slide positions when reversing (look at the chart from right to left).

The horizontal lines show possible positions for inlet and exhaust port positions - the exact dimensions of the slide valve will determine if my assumptions are realistic.

(the valve slide positions are not to the same scale as the piston slide, and are adjusted to fit on the same area of the chart)

I think I can now see why Steve at STW was talking about inherent inaccuracies of the valve gear.  I think that designers aim for symmetry to enable the best control and efficiency.  This can be seen to not be the case here.  The red line shows that when going forwards, the valve is being pushed much further than in all the other cases.  This seems to arise from the eccentric rod over-reaching and forcing the motion arm too far up.  This could be relieved by rotating the Crank Bearing Housing by about 15 degrees (anti-clockwise in the clips).  I am not sure if I have the confidence, knowledge or skills to do this successfully.  (If it works don't fix it!).  I have seen images of Hackworth valve gear in which the slide valve is positioned at an angle to the cylinders to align it directly with the centre of the valve gear - to achieve this would mean inserting a wedge across the slide valve mating surface (not my idea of a risk free improvement).
I have been worrying about the effects of condensed steam forming an hydraulic lock in the cylinders, but I have been told that slide valves are able to cope with this situation.  If (incompressible) water forms inside the cylinder, the excess pressure lifts the valve which is normally held down by steam pressure.  The excess water is supposed to be released and pushed to the exhaust by the steam which rushes through.  I would like to run the engine really slowly if this situation occurs.

I was also wondering about putting the cylinders in a wooden casing or box so that heat losses can be reduced and the risk of burns be lessened.