Wednesday, 4 June 2014

Pipe Reality Come True

My previous main steam pipe article was at the stage of it being persuaded into a shape that should fit. In practice, it wasn't that easy.

The first challenge was that, being such an awkward shape, it would have been easy for Sentinel to make it fit with 7109, the pipe and all tooling collocated in the factory. With the pipe being made five miles from 7109, it could only be made to approximately the right shape using a jig made using the old pipe geometry. Mendip Steam's Andy Melrose eventually gave in that some heat was going to be needed in the cab to make it fit. Thus 7109 had her first fire lit since 1968!
First fire in 46 years
Not much steam was created but it did seemingly allow the pipe to ease into place, or so we thought.

Everything was fine until the second challenge. The nice new end pieces were made to drawing size but that did not mean to say that the original mating joints were made that way.
Nice new end fitting exactly to drawing size
The mating end under the water tank had at one time been repaired and the hole into which the new pipe had to fit was not actually round! The two parts would not fit together.

There was nothing for it but to grovel about underneath with a miniature grinder (it's not made for a human to squeeze under the water tank) and make the mating hole bigger, not a lot of fun!
The underneath end finally in place
The fitting in detail...
...fed by a pipe around the firebox...
...fed from up here in the cab...
...from the regulator all the way up here!
So the pipe is at last in place. The white insulation is made from glass fibre ladder tape wound over partly dry, tacky red oxide paint to keep it in place. Finally the ends and a centre join are secured in place with galvanised wire.
Glass fibre ladder tape
The pipe joints are sealed using 3mm thick annealed copper washers, made from copper sheet, with lashings of SteamSeal. The copper washers were made by cutting the outer perimeter with a hole saw and the inner hole with a screw tightened hole punch. Glad I wasn't in a hurry for these, it took ages.


Saturday, 24 May 2014

Sanding Gear (1)

After a few months of apparent inactivity with Sentinel 7109, things have actually been happening (although much of my time of late has been spent defining safety critical operational roles and training material for Midsomer Norton); the first of these is the fitting of a replacement sandbox pipe at the off-side front.

For ages, there have been three original pipes painted and ready to fit to feed sand from the four sandboxes to the rails; however, the fourth one had disappeared some time ago or so I thought. I'd searched and searched, convinced that the fourth one existed, but finally resigned myself to having to make a new one, flange and all.

Eventually, on close examination of the sand box itself, I found that there was a flange still in place underneath with clear evidence of a pipe having been broken off from it.

So I'd been wrong; the fourth one hadn't existed after all.
New Sand pipe (1)
The new pipe is made from mild steel with a longitudinal weld, i.e. not malleable iron or cold-drawn seamless steel. A 3/4" BSPT thread needed to be cut to fit the flange on one end before bending to shape. (It's difficult to clamp a bent pipe for thread cutting!).
New Sand pipe (2)
After shaping into a mirror image of the near side front pipe using a 12 ton hydraulic pipe bending machine, the lower end was cut diagonally to feed the sand into just the right place.

Finally it was screwed into the threaded flange still attached to 7109. (28 tons of solid metal certainly held the flange tightly in place!).

There's much more to come on sanding gear, particularly working out how the levers need to operate the four boxes.

Tuesday, 18 March 2014

Nut & Bolt - Big Time!

Years ago, my mother recounted the story of her first driving lesson in which, once she had got the car moving, her main concern was how on earth to stop it!

Health and safety aside, the principle has stood the test of time. Sentinel 7109's original handbrake screw thread was clearly beyond use when it first arrived at Midsomer Norton. I never took a photo of it but the square shaped 'Acme' thread bore a strong resemblance to the 60 degree angles of a metric thread profile! As the last resort for stopping, it did not provide much in the way of confidence!

The way the handbrake works is quite simple. It's a big vertical bolt with a handle and a nut that's linked to a crank. When the handle is turned, it pulls on the crank which pulls on the brake. (There's some more about it here: Steam Brake Cylinder).
Handbrake Nut and Linkage (below cab floor)
Handbrake Handle (above cab floor)
Best efforts were tried to repair the damaged thread by machining it away and welding new metal in place to allow a new thread to be cut.

It worked initially as in the video below (Also on YouTube).

However, after a while, the cutting tool decided it wanted to dig a little deeper and pulled off some of the metal causing a complete write-off of the original shaft.

Mendip Steam Restorations were tasked with the job and completed both the finishing of a new nut (with posh phosphor bronze 'Top-Hat' insert) and a new handle shaft.
New Handbrake Shaft and Nut
The original handle assembly has been attached to the top of the new shaft. I wasn't sure about a plain thrust bearing and decided that, for minimal cost, a roller thrust bearing could be used as shown below. Whether it will prove strong enough in practice, I don't know but at least there is the fall-back to the original construction.
Roller Thrust bearing below the handle
New nut & insert from below
New nut & insert from above
So, with any luck, we'll now be able to stop 7109 after we've made it go!

One feature which I do intend to add is a shield of some sort to prevent grit getting into the thread; hopefully this will prevent the damage that happened to the old one.

Saturday, 18 January 2014

Vacuum Braking (6) Testing (1)

I have to admit to being mildly excited! I'd previously been unable to include an article on Sentinel 7109's Jet Pump (Vacuum Exhauster) device itself and only realised today when I'd completed a highly sophisticated (i.e. garage based!) compressed air test of it.

So here is the little sucker:
Penberthy GL-1 Jet Pump
It's not very big as can be seen from the picture (hopefully it will be big enough!). Steam (or air) enters from the right and exits on the left. It draws a vacuum through the sideways port.

For quite some time, I've had to be satisfied with the calculations I'd done in selecting this particular Jet Pump; however, it's not a technology I've been previously familiar with and have also been at variance from others who have fitted vacuum exhausters to Sentinels before. So my R&D has not been without some risk as to its success.

My first attempt at testing the jet pump was to connect it to an air compressor. However, this produced a poor result which I put down to the narrow bore of the air feed pipe being less than the bore of the jet pump itself. Thus it was never going to be able to feed sufficient air at the required pressure (around 60 psi).

By a stroke of luck, I found a scrap air cylinder at Midsomer Norton Station that had a 1/2" bore outlet and a valve. I considered that if I could pump this up with the compressor, I could then use it to feed the jet pump from its 1/2" bore outlet, at least for a short time anyway.

So here's the test rig in all its sophistication:
Jet Pump Test Rig
The air cylinder is on the left with the compressed air inlet at the top via the orange pipe fitting. There is also a non-return valve on the inlet. The valve controlled outlet is by the black air pressure gauge. The jet pump is above with an Ex-BR vacuum gauge on its suction port. The jet pump outlet is skywards.

This time I was much more successful (which led to my excitement that what I'd calculated actually worked!).
Vacuum gauge showing 21" Hg with about 66 psi air pressure
(It's easier to read on its side!)
I'm not concerned at the 66 rather than 60 psi. Neither gauge has been calibrated so the figures cannot be relied upon absolutely; however, I'm satisfied they are good enough to show the principle.

Here it is in video form: (also on YouTube).

Jet Pump Test

Some years ago, I visited the USA when printed 'T' shirts were becoming popular. I bought one with the message: "Engineer - Individual who turns abstractions into malfunctions". Hopefully not always!

Friday, 17 January 2014

Vacuum Braking (5) Design (4)

One difficulty with Sentinels is that the 275 psi boiler pressure is above the 250 psi rating of many off-the-shelf valves. At 275 psi, water boils at 212 DegC so special materials are needed to withstand these conditions. Any valve that has to isolate a steam supply from a boiler has to be rated as such.

Three new ones are required on Sentinel 7109:

  1. A 1/2 inch type to isolate the vacuum braking supply.
  2. A 1/2 inch type to isolate the supply for a steam cleaning lance.
  3. A 1.25 inch type for the boiler's blow-down valve. (See later).
They were made in Switzerland by Valtaco and sourced from Poynton Valves. This is what they look like:
The 1.25 inch and two 1/2 inch ball valves
These are known as ball valves because the moving part of the valve is ball-shaped (but with a hole through). To begin with, investigation had sent me towards using gate valves as these are intended for isolating supplies rather than regulating a flow. However, gate valves are better suited to liquids at relatively low pressures. Ball valves can be built to be much stronger and are very effective at isolation duties.
Full bore when open
Those of you who have followed my blog for some time will recall that a blow-down valve had been obtained some time ago; however, my wariness about fitness for purpose of high temperature and pressure components has been heightened as the project has progressed.

The original blow-down valve was made from gunmetal with flange fixings. Whilst it looked the proper job, when I came to investigate the safety valve mounting flange requirements, I found that I would need a BS10 Table 'F' rated type made from Carbon steel. The Table 'F' flange is 1/2 inch thick which is somewhat more than the flanges of the blow-down valve. Add in that gunmetal is not as strong as Carbon steel and it becomes clear that the original would not be strong enough. Not having any specification to support its construction means that I really can not be sure it is good enough.

The ball valves pictured above are made from stainless steel with Carbon reinforced PTFE seals and are specified for 212 DegC and 275 psi. I thus have complete confidence that these valves will be strong enough. Also, to coin a phrase, they were 'Reassuringly Expensive!".
Related Posts Plugin for WordPress, Blogger...