Thursday, 7 July 2011

Water Pump Pumps Water!

Despite the original heading of the last WS BFP episode, I've used that heading here and corrected the previous one (owing to the fact that no water had actually been pumped at the time!).

I concluded that the water pump valves did need regrinding. Advised to use genuine valve grinding paste, I 'Googled' it and found that it is available from various manufacturers in a two part tin: coarse paste at one end and fine at the other. Chippenham is not an easy place to buy valve paste but KM Farms Parts had what I wanted.

Typical Coarse/Fine Valve Grinding Paste container
(Spot the product placement in the form of the Worthington Valve handle!)
I hadn't done anything like this before so it was a bit of an adventure working out the best way to use the paste. The fine type seemed to be all that was needed.

The valve seats are basically a pair of flat discs of which the moving one sits loosely around a supporting stem; as such the valve has to seal both at the outer edge and the inner edge for it to work. If the inner edge leaks, then water seeps back through the gap around the valve stem.

Mating parts of an upper chamber (outlet) valve
By inking the inner and outer seat surfaces and rubbing them together, it was clear that there was incomplete contact and scope for improvement. The outer seat needed to be ground down in all eight cases for the inner seat to come into contact. Coarse paste was needed initially as there was a fair amount of metal to be removed.

It seemed from the initial state of the valve seats that the inner seat could never have made contact unless it had eroded considerably during the pump's original working life. If this is the case, then there is much reliance on the boiler inlet 'clack' valve to prevent leakage back through the pump. Alternatively, the valves may rely on being flexed by back pressure to make contact at the inner seat. I had expected the pump piston rings and valves to need to be a tight fit but I'm not so sure now. Some leakage allows the pump to operate without seizing in one position. Time will tell!

Lower chamber (inlet) valve seat
The lower chamber (inlet) valves are of a different design and have indents between inner and outer seats. Again, both seats need to seal for the valve to work properly. As can be seen, the lower valve seats were not easy to get at but fortunately not impossible!

The valve grinding had clearly paid off as the characteristic 'raspberry' produced before was now much better pronounced. Also the inlet could be felt to suck and the outlet to blow.

Moving swiftly on to testing with water. Eat your heart out William Heath-Robinson, here's the test set-up. Recycling is a strong feature!
Complete with spurt!
The pump self primed at the first attempt with the reservoir (watering can) at the same height as the pump but it needed help when having to suck from a lower level. There is also a leaky weld in the inlet pipe which will need repairing.

And the video you've all been waiting for... (also on YouTube).

I'd noticed that the flow did not seem even; half the time it was steady and half a bit slower. At first I thought one pump cylinder was faulty but then realised that this is bound to happen as each pump cylinder has a piston-rod in one half with a consequent reduction in the cylinder capacity at that end! 'I-think-I-can,'I-think-I-can...' comes to mind as the typical two cylinder steam engine rhythm.

So the pump seems to work. I'm still slightly nervous that the pump valves don't seat well but they are probably better than they were when previously in operation.

The user manual (yes, believe it or not, there is a user manual for these pumps) makes no comment about valve seat geometry; however, it does mention the pump piston rings. The four hollow rings are intended to have 1/16 inch of total axial play which I guess is to allow them to fill out when pumping the water to make a seal. What is not mentioned is that the only way to check the play is to remove the piston and rod completely (including the steam end also) and have a look.

No! This is going to remain one of those assumptions that it was OK originally so it's OK now as taking it all apart again belongs in the 'too difficult pile'!


  1. I'm now wondering if you bought me a new watering can, so that you could commandeer the old one...


  2. Do Sentinel locos have automatic control of boiler water level?

    1. Not that I know of. More conventional locos require the fireman to operate an injector when the boiler needs more water. 7109 has the cab mounted pump which can be adjusted to steadily top up the boiler. The engine mounted pumps also operate when moving. It's probably an art to getting the settings right!

  3. It has always surprised me that steam locomotives don't have automatic control of boiler water level because it is an obvious safety measure. I think it would easier to arrange with a reciprocating pump than with an injector.

    1. Not so easy though. Putting water in the boiler is also used to stop blowing off the safety valves so it's not just the level that is important. Of course, in steam days, one was expected to know how to do a job and not to rely on safety devices!


Related Posts Plugin for WordPress, Blogger...