Poppet valve sizing for tender axle pump

[Bill Shields]

it is a calculation problem.

at maximum rpm of the wheels, what is the volume of the water going through the valve (calculated from the known piston dimensions and rotation).

Knowing the volume, and the open area of the valve, what is the velocity of the water going through the valve?

I will have to go back and dig it up, but Parker has a number of 'do not exceed' velocity for common size O-rings (by material) and liquid as a function of temperature (water is most certainly in the chart).

I am all for accessible pumps.....

FWIW:

I have fixed porosity in castings by subjecting them in a vacuum and wiping them all over with superglue and sucking the stuff into the cavities...and letting it set up...

[Fender]

So with a pump piston 3/8” diameter the area of each piston is about 0.110 in**2 and the flow area around the poppet valve (you said) is about 0.030 in**2. This seems reasonable to me, but depends on the track speed. The pump will be producing about 0.110 in**3 of water from each revolution of the tender axle (both pistons having a 0.5 inch stroke), or about a gallon (231 in**3 per gallon) for every 2100 revolutions, by my calculation.

[Bill Shields]

Concern is flow velocity through each valve at max speed.

What does that work out to?

Look at the open area of each scallop x 3, since that is the number of scallops per valve.

Then look at the lift of the the poppet and the resultant annular area...

Highest velocity is in the smallest area.

Worst case is high velocity (momentum) pulling the o-rings off the seat as the valve closes, causing a pinch.

I have always been a fan of semi captive rings so that flow in s less likely to get under them and cause them to move.

[DaveD]

I believe I've worked this out. I found an old rough sketch of the axle pump, probably done by Bruce, and drawn somewhat to scale. It shows what are almost certainly 1/4" balls closing against what are most likely 3/16" passage bores and with the balls in what appears to be 5/16 holes for clearance. Doing all the calculations, it happens that the cross-sectional clearance of the 3/16" passage/ball seat is 0.028 sq. in. and the clearance in the ball bore minus the ball cross section is 0.027 sq. in., very close to the 0.030 sq. in. that I have total in my three poppet valve flutes.

It is apparent that all this works OK, as I haven't heard of any problems due to check ball bore or passage sizing with either of the two Hamilton axle pumps. The only concern that I might have using my poppets instead of balls is what Bill pointed out above, that being the potential for the higher velocity at the O-ring area causing problems with them, and I don't believe that will happen because the poppet valve lift is pretty high. If that does happen, I believe it would be easy enough to fabricate new poppet valves with a slight concave where the O-ring fits, holding them more securely.

Thanks for the comments. If, down the road, this design works out well, I would be happy to put it on chaski.

[Bill Shields]

High lifts can have problems with floating.

With balls, maybe 10% diameter is a good lift number.

All of these concerns can be handled, but should be considered.
.it is amazing that I have been building steamers for 50 years with ball valve axle pumps, and am caring for one that is 50+ years old...and not had a failure with ball valves.

Guess I am just lucky...or the old UK tried and true designs are valid.