Valve gear

The plans for SPEEDY include the original layout by LBSC as well as detailed drawings for Don Young's modifications. These modifications were an attempt to improve the valve timings but these fell short of the ideal too.

'''It would appear that the reverser rod on the original plans fouls the underside of the side tank. One solution to this problem might be to make the vertical arm shorter and move the point it's attached to on the pole reverser down to reduce the travel.'''

The geometry was re-worked by Don Ashton who is a recognised expert in the field, and the dimensions for those are shown here. Imperial units are used throughout. The length of the connecting rod changes to 7-1/4", helping with the clearance at the cylinder gland.

K. N. Harris also re-worked the valve gear, and some builders abandoned Walschaerts gear altogether in favour of Baker valve gear. Trying to make a truly accurate scale model that works well is probably impossible because the original requires the crosshead end of the union link to be on the centre line of the small end bearing of the connecting rod. Doing this appears to lead to less than ideal valve events.

The locomotive can be constructed with any of these arrangements. No official detailed drawings for the Don Ashton arrangement are available but the existing drawings can be modified to these dimensions.

Notes for implementing Don Ashton's geometry...

The Union link will need to be modified to accommodate different offset introduced by the way it's attached to the crosshead. The clearance between the expansion link and the radius arm is very small at half gear and care must be taken at this point. The position of the expansion link pivot also needs to clear the gear frame which will have to be moved a little to accommodate the movement.

NB:- The 1.365" and 0.178" measurements on the diagram are square to the Radius Rod and NOT the centre line of the axle. The expansion link is not standing vertically but sits on the arc described by the Radius Rod which is at a slight angle.



Modelling the valve gear
Designing Walschaerts valve gear is far from simple and in the days before computers this was sometimes modelled with card and pins to make a phsical model. The link below is to a computer simulation that has Don Ashton's new geometry for SPEEDY as one of the sets of data.

http://avocetconsulting.com.au/modeleng/

Don Ashton's book (see link below) is available if you want to delve deeper into the subject

http://www.camdenmin.co.uk/products/design-procedures-for-walschaerts-and-stephenson-s-valve-gears Here's a screen shot of the simulator. Don't be fooled by the dated look of this program, the graphics may be rudimentary but the information it provides is excellent.

A very powerful feature of this program is its ability to create families of Indicator Diagrams from the simulated geometry.

Indicator diagrams
The following three families show the consequences of the differences between the various valve geometries, but first, here's an explanation of what these diagrams show.



There are two plots on each Indicator diagram which are shown in mirror form for easy comparison, the one on the left is the diagram for the front of the piston, the other is for the back of the piston. Remember that this is a double acting system with steam driving both ways. The Simulator can generate these families in both forward and reverse gear but here we are only looking at diagrams for Forward gear. The linkages don't provide symmetrical valve events in both directions.



Let's look at the Left Hand plot only for this explanation...

Starting from the top left corner the steam has just been admitted at the Dead Centre and has risen to the maximum. Following the line along to the right, the piston is moving and the steam is still being admitted.

If we go all the way along to the end of the line, somewhere around 78% of the Piston Displacement, the inlet finally closes to live steam and now the pressure drops as the steam expands until at about 93% the exhaust port opens and the steam is released almost instantly.

From the bottom right of the diagram we now go left with the pressure just above atmospheric as the piston performs the exhaust stroke.

At about 9% of piston travel, the exhaust port closes and the pressure begins to rise as the remaining steam is compressed. The piston continues to move left on the diagram with the pressure rising until the inlet port opens and the pressure almost instantly rises to the maximum again.

Each successive plot on the diagram shows the effect of 'notching up' the valve gear by moving the reversing lever closer to the middle position. The lever is moved in equal increments, and the valve gear should ideally provide equally spaced curves.

The original SPEEDY valve gear suffers from having an uneven spread of curves, meaning that 'notching up' isn't smooth and progressive. It also has a cut off in full gear which could be a little higher for better starting.

It's not immediately clear why a high percentage of cut off is necessary until you picture what happens in the opposing cylinder. If you imagine that the cut off is 50%, then the cylinder on the opposite side would be at a dead centre and unable to provide any turning effort. Of course the steam can still expand, but the pressure will rapidly diminish as the piston continues to move. Clearly this arrangement would be likely to leave the locomotive stuck at this position without a little help from the feet of the driver, and that would be embarrassing. So it's clear that 50% cut off is hopeless and that the higher this figure is, the more torque there will be available for starting.

Looking at Don Young's diagram, the spacing of the curves is much better but at the expense of much less cut off in full gear. This means that although this gear will 'notch up' more evenly, the starting torque is less than with the original gear.

So finally to Don Ashton's gear, and you can see that both of these issues have been rectified. Not only are the curves much more evenly spaced than on the original design, but the cut off in full gear is also later. This gives the locomotive more starting torque and more evenly spaced progressive 'notching up'

Solid modelling
3D Modelling of Don Ashton's geometry shows that a practical model can be made with these dimensions. The picture shows a model and dimensioned drawing from it which confirms that the dimensions have been followed. The dimensions are in Metric units.

The model shown has been modified to include an additional 1/8" between the cylinders and the front wheel as described elsewhere.



Practical considerations
Some parts of the valve mechanism are prone to wear and have been fitted with Bronze bushes on the original design. Case hardened pins are sometimes used to extend their life. Some modellers are experimenting with miniature ball bearings which can be accommodated as long as the inside pin diameter is reduced. Such bearings used in reciprocating parts may be difficult to keep lubricated and a light grease may provide the answer to this. The cost of these bearings is only a matter of £2 to £3 so replacing them is not unduly expensive.

The valve gear rod ends can be fitted with thin walled hardened Silver Steel bushes which can have as little as 0.5mm wall thickness. These can be fitted with Loctite and left dead hard, ie just quenched to harden them from red heat. If they are a little tight, you can use a piece of Silver Steel and some grinding paste to ease them.

The axial end float of the Middle drive axle is usually kept to an absolute minimum so that the valve gear doesn't have to accommodate and more than is absolutely necessary. The front and rear axles are allowed more float, typically 0.4mm (1/64" ) each side to allow the locomotive to follow the curves in the track.

There needs to be enough play or flexibility in the valve gear to allow for the middle drive axle to lift approximately 1.3 degrees, causing a twisting motion on the eccentric rod. These figures have been found through practical experience.

The piston valves spool dimensions are different for each of the valve gear designs, the ones for Don Ashtons are on the line drawing above.

The fit of the piston valve spool is critical to the efficient running of the locomotive because there are no seals to prevent the passage of steam.