Who was Joyce?

Sentinel 7109 was named 'Joyce' after the daughter of Mr Sandeman, the chairman of Croydon Gas works where 7109 operated for 32 years until 1960. This is known from "The Locomotives of Croydon Gas & Electricity Works" by J. B. Latham and published in 1970.

New replica nameplate

But that was the easy bit. Current S&DRHT secretary and long standing member, John Baxter, began searching on the Ancestry.co.uk website and I subsequently continued the challenge.

The search began with 'Joyce Sandeman' and 'Croydon'. This revealed only one person 'Joyce Helen Sandeman' and, according to various records, she was born on 1st July 1902 and died a spinster on 12th December 1974.

'Sandeman' being an uncommon name and located in Croydon would appear to be the only lead. Thus all information here is based on Joyce Helen although it could be completely wrong!

Joyce's parents were William James and Annie Charlotte Sandeman. William James Sandeman was the Gas Works chairman and other members of the family had connections with the gas supply industry.

1911 Census Entry

As Joyce died a spinster, there are no children to seek to ask about her so there is no reliable information to tap.

Joyce had two cousins: Ronald Frederick Sandeman, who married Elsie Louise Docking, and Barbara Clare Sandeman who married Cecil George Hine. All of these relatives died in the 1970s or 1980s.

Although I have no public access to this family via the ancestry website, it would appear that there were children who could possibly still be alive and a source of information. It would be lovely to hear from anyone who really knows something about Joyce herself.

I have, however, been speculating a little. I asked myself: 'If I had a steam locomotive where I was chairman of the company, what would be the likely age of my child for me to honour by using their name for the loco?'. In the range 5-15 maybe?

Joyce, being born in 1902, would have been 26 when 7109 arrived at Croydon Gas works in 1928. Seemingly, Joyce was an only child so maybe this could make sense but I'm not convinced this was the reason.

The 1939 census shows Joyce to be in the Cedar Grange Nursing Home in Caterham and described as 'incapacitated'. In that context, 'incapacitated' suggests that she may have been an invalid either from birth or through the consequences of WW1. The incapacity may have meant that she was cared for by her parents and was so dear to them that 7109 was honoured with her name.

This is of course purely speculation. If anyone can shine a light on Joyce, I would be grateful for any information - please leave me a comment to this posting. And, of course, be my guest and come to visit 7109.

Change in Ownership

Andy Chapman and Nigel Dickinson

On 1st November 2016, Nigel Dickinson and I agreed that for Sentinel 7109's future, it would be best if I continued as sole owner.

As a result, from 1st Nov 2016, Andy Chapman has full ownership.

I'd like to thank Nigel for all the work he has put in over six years or so. Without his original idea to restore Joyce to full working order, it would never have happened.

Sentinel 7109 Re-Joyced!

9th July 2016 was a day for celebration as Sentinel 7109 was officially renamed 'Joyce' by visiting Antiques Roadshow expert and railway author Paul Atterbury.

The day began with plenty of boiler pressure but nameplate obscured.

Sentinel 7109 (unnamed) raring to go (Photo Callum Willcox)

Paul Atterbury meeting with Andy Chapman & Nigel Dickinson (Photo Michelle Chapman)

7109 re-Joyced by Paul Atterbury with Nigel Dickinson and Andy Chapman (Photo Bob Edwardes)

Zomerset Zoider put to good use (Photo Michelle Chapman)

I've been asked a number of times - Why 'Joyce'?

A very long time ago, works number 7109 was originally named 'Joyce' after the daughter of Mr Sandeman, the chairman of Croydon Gas Works. In Sentinel circles, many refer to 7109 as 'Joyce' so we decided that we should rededicate her with her original name (re-Joyce!).

At the same time, we also wanted to pay homage to the two Sentinels that worked at Radstock from 1928 until ~1960. In their LMS period from about 1930 until 1948, they were painted black and numbered 7190 and 7191. 'LMS 7109' kept the same idea whilst also keeping the 7109 works number. So Sentinel 7109 now carries the Croydon and Radstock identities in her new guise.

Paul's cab ride out towards Chilcompton (Photo Bob Edwardes)

Midsomer Norton's wonderful signal box with Joyce posing in front (Photo Callum Willcox)

Despite Joyce's diminutive size, she is a lot more powerful than she looks. Draw-bar capability is 19,600 lbs according to Sentinel's marketing literature.

Joyce eases up the hill dwarfed by 70 tons of carriage & brake van (Photo Callum Willcox)

I'm indebted to Callum Willcox for shooting and composing an excellent video recording the day's events. It can be seen on YouTube by clicking here. Don't miss it.

Sentinel 7109 as LMS 7109 - Joyce

Needless to say, a great day was had by all - particularly Joyce!

Next day, the S&D's 08 diesel shunter was detached and 'Joyce' hauled the four afternoon trains...

Braided Flexible Hose

Leading up to the March 6th 2016 deadline for the 50th anniversary of the closure of the Somerset and Dorset Railway, I was faced with many outstanding jobs to get Sentinel 7109 in steam for the event. I knew I had eleven or so pipe runs to fabricate for carrying superheated or saturated steam as well as water and oil, all at 275psi boiler pressure.

While I could have kitted myself out with copper pipe tooling, it was going to make the job too lengthy to meet the deadline. I’d been recommended stainless steel braided hose as a possibility and after some internet searching found Guyson’s Kaptech hosing. (www.Guyson.co.uk).

Stainless Steel Braided Hose

After some discussion with Guyson’s technical department, I found that they could make to order in a fairly short time the hoses I wanted with the right lengths and connections to suit the existing fittings on 7109. In addition, each pipe run had to suit the pressure and temperature ratings required.

All hoses would have to withstand 275psi. Superheated steam hose would have to withstand the pressure at up to 380DegC, the max superheated temperature measured by Sentinel in their test department. Saturated steam would be up to 230DegC. Cylinder oil from the mechanical lubricator and water from the feed pump would be at relatively cool temperatures.

I chose all the steam hose to be 3/8” Nominal Bore (NB) annularly corrugated 321 stainless steel lined; the water feed hose to be 1” NB PTFE lined and the cylinder oil feed to be ¼” NB PTFE lined. All would need only the single outer braiding layer.

The maximum pressure and other ratings for the annularly corrugated hose are in the following table (from Guyson):

Stainless Steel Hose Ratings

Thus my 3/8” NB hose with single braiding would take 100 bar (~1500psi) at room temperature.

The pressure derating table for temperature is below.

Pressure derating for temperature

At 380DegC, the pressure rating would still be around 74% of 100bar, i.e. ~1100psi. This is a very high specification indeed with plenty in hand for a Sentinel. The PTFE lined hose was not so highly specified for temperature but not a weakling by any means.

PTFE Lining rating

The ¼” NB cylinder oil hose would thus handle 240 bar (~3600psi) and the 1” water feed hose 55 bar (~825psi).

In order to purchase the various hoses, I prepared a spreadsheet table with the following information (two examples shown).

Information to specify a hose

I sent this to Guyson who responded with their own internal manufacturing version for me to check. After a few iterations, we settled on a final spec for manufacture and I placed the order. It all arrived in a few weeks.

The following photos show how the hose was used.

3/8" hose as part of the blower supply

These short sections of hose were later connected to a ‘T’ piece leading back to the blower valve. Despite ordering a 6” length, I found that the flexible part was much shorter than 6” due to the end fittings. Thus the curve radius had to be tighter than I wanted and the final appearance not ideal.

This was not the only time I found the fitting length getting in the way of a tight bend.

3/8” hose from the feed pump regulating valve

The connection to the feed pump valve required a hex nipple to fit a female hose end fitting. These fittings are a hydraulic type but similar in many ways to the original for copper pipe. Hydraulic fittings generally use a cone seat rather than flat face with copper washer to seal so I had to use a lathe to convert the nipple cone into a flat face.

3/8” elbow fitting at feed pump

Elbow fittings can make life simpler but need to be thought out carefully before ordering. Again I found I needed a parallel thread hex nipple and copper washer to fit the pump. Being superheated steam, I was not happy with a tapered thread fitting here.

3/8” in and out feeds for steam brake valve

The pressure capability of the flexible hose is provided by the outer braiding with an annularly corrugated inside tube providing the flexibility. As such, the failure of flexible hose is down to abrasion of the braiding. Where the hose had to pass through a hole and could chafe the sides, I wrapped silicone fire sleeve (silicone coated glass fibre) around as protection. While the hose can move about, it is now well protected from chafing. For hose such as the steam brake supply, pressure is applied and released. The hose hardly moves with the changes in pressure.

3/8” whistle and pressure gauge feeds, the latter incorporates the loop

For the boiler pressure gauge feed, I was able to wrap the hose into a loop to keep the steam out of the gauge.

The whistle has been supplied by flexible hose too; however, this may be the cause of the whistle seeming to be choked with condensation unless used frequently. Despite using various sizes of orifice to regulate the steam flow rate to the whistle, it always has to ‘clear its throat’ before coming on tune. I suspect that condensation accumulates in the inner tube corrugations with the hose not being lagged so further investigation is required. Suggestions?

1” PTFE water feed hose from feed pump to single check valve

With the pulsing of the Worthington Simpson pump, the 1” PTFE hose moves about quite a bit. I’d be better off with an anti-shock bulb at the pump end to smooth out the pulses.

¼” PTFE hose, 11 feet long section in the cylinder oil feed

The cylinder oil feed takes a tortuous route around the pipe work underneath and then circumvents the boiler. There was an eleven feet gap in the original pipe to the regulator assembly. The ¼” PTFE hose filled this gap very elegantly. I used ‘nutserts’ to attach the hose to the boiler cladding.

In summary, I would say that flexible hose is an elegant alternative to copper tubing; however, there are various factors to be borne in mind.

1. It lacks the authentic appeal of copper piping.
2. It can be ordered ready to fit (provided you get your specification right – not trivial and prone to mistakes).
3. The hose is flexible but cannot twist and can thus be awkward to fit at times.
4. The length of the fittings may get in the way. Consider an elbow fitting if a right angle exit is needed.
5. For short hose lengths, take great care about the minimum bend radius specification.
6. For joins in the flexible hose such as a ‘T’ joint, the standard hydraulic cone fitting can be retained. They work very well between themselves but won’t mate with the original copper washer fittings.
7. Any type and size of fitting can be ordered for any diameter of hose. This could give some pretty bizarre assemblies but technically it is possible.
8. Lengths have to be thought out carefully. It is tempting to think that a bit extra will make life easier. In fact the extra can get in the way and it’s ‘return to factory’ if it needs shortening.
9. Chafing must be prevented. An annual inspection of the braiding should be performed – no different to any other pressure tubing.
10. If vibration is likely, make sure the hose is secured in place as with any other tubing.
11. Stainless steel fittings are the norm. There is little saving to be had with none stainless types.
12. Copper washers are needed for flat face parallel thread fittings (same as with copper tubing fittings). I also used Steamseal to make sure.
13. The PTFE inner lining can kink if twisted.

The entire set of flexible hoses was in the region of £1000. Would I recommend this approach? Certainly!

Glandular Replacement(1)

My Glandular Diversion article showed a temporary measure to reduce the water ingress into the crankcases from the engine water pumps. Now the proper job has to be done to stop the water leaking in the first place by renewing the gland packing.

This instrument of torture is used to remove the old packing.

It can probably also pull corks too!

The shaft is flexible and cleverly made so that it does not twist as might be expected. It is screwed into the old packing; the packing can then be pulled out layer by layer, five layers in this case.

The drawing below shows the gland arrangement.

Gland construction
(this is actually an engine piston rod gland but very similar)

First the packing tightening nuts have to be removed. These are conical nuts which are designed to be difficult to over tighten.

Compression fitting released

Then the packing collar is dropped down to show the packing material.

Old packing showing

The torture tool removes a layer at a time.

Two packing layers removed

I was surprised to find that the old packings were made from layers of leather stitched together.

Extracted packing layers

The layers can just about be seen in the enlarged photo below.

Leather goods

Repacking to follow.

Glandular Diversion

My recent article, 'Glandular Fervour', showed the gland packing replacement task ahead. However, prior to that, I had a little idea to temporarily prevent the leaked water ending up in the crankcase.

Grommet

By wiring a grommet around the pump shaft to just the right tightness, hopefully, with the shaft well greased, any leakage should be diverted away from the lower gland and out to the drain.

Fingers crossed!

Post script:
The grommet idea worked quite well particularly when the grommet was well above the lower gland.

Being a Dipstick

In an August 2013 article, I was left with a mystery. Why did the rear engine require so much less crankcase oil than the front engine to reach the dipstick half-full mark?

At the time, I'd assumed there was something lodged in the rear crankcase but didn't investigate further. Some three years or so later, while preparing to clean out the rear engine's oil sump, the actual reason has become apparent.

Take a look at the following photo with the two dipstick heads highlighted.

A new slant on dipsticks

Although the photo is taken from slightly above the level of the dipstick heads, it is clear that the far one is not only at an angle but considerably lower than the nearer one, about 3" in fact. I'd not taken any notice of this before and assumed that the dipsticks told the truth. When I checked the two dipsticks together, I found they were identical in length and markings. Thus the lower one dipped 3" further into its oil bath registering full rather in advance of actual fullness!

This answers the mystery of the differing oil requirements for each engine. What worries me is how long has it been like this and why are the two engines different?

Of course, the rear engine will have been running with two fewer gallons of crankcase oil than the front one. Hmm...

At some time in the future, the rear dipstick will need fixing. In the meantime, yet another mystery has also been solved.

In this photo, there is a drain cock to let water out of the sump and another highlighted in the top right.

Extra drain cock, top right

I'd wondered what this extra drain cock was for but now I know. Examining the front engine's dipstick for where the oil level would actually be in the sump, it is clear that this drain cock would let out any oil above the full dipstick level. Thus it is an alternative way of checking for sufficient oil without a dipstick.

The rear engine had a blanking plug instead of a drain cock so I swapped the two. The rear engine oil level can now be checked using the drain cock while the front engine's dipstick can still be relied upon.

The mystery has been solved but why the two engines are not identical remains an unknown.

Glandular Fervour

In February 2016, when Sentinel 7109 took its first self powered trip in decades, it was clear that there were a few problems. One being that the various glands which seal piston rods and pump plunger shafts have gone hard or have been overtightened in ignorance recently or in the past.

There are 18 of these associated with the two engines and their boiler feed pumps. Some are worse than others, particularly the pump glands which are letting water out and down into the crank cases. Whilst crankcase oil is designed to allow the water to separate out from the oil and be drained, I'm not sure it was designed to cope with gallons of water!

These photos are from April 2009 and show the state as found of the front engine pump glands.

Glands within the pump assembly

There is an upper and lower gland, one should prevent water leaking downwards from the pump, the other to prevent water that did leak from entering the crankcase below.

A bit Rusty?

It's difficult to tell whether the rust had accumulated during use 50 years ago or subsequently.

No better from this angle!

After much restoration, the front engine pump glands are shown operating in this video clip on YouTube. (Sorry, it's too big to include here).

Clearly there is work to do! I took apart the rear engine's upper pump gland and poked in the camera to see how it was.

Rear engine's upper pump gland

The gland material looks pretty hard with a small gap around the plunger. This will have to be dug out and replaced before it will work well. Fortunately the shaft seems to be fairly parallel so it may not need replacing. Anyway, it's easier to try a simple fix before a complicated one so I'll do the packing only to begin with.

I've been recommended greased hemp rather than graphite or PTFE for water pump applications. It needs to be 3/8" square section. Sentinel used to have an off-the-shelf item for this but it doesn't seem to be available any more for some strange reason...!

Front Engine Sludge

Vertical Sentinel locomotive engines were designed to allow for water to accumulate in the crankcase sump, the lowest part of the engine itself.

The water comes from condensation and leakage in the boiler feed pump piston glands. It's dealt with by using special 'Crank case oil' which is formulated to enable water to separate out from the oil quickly and sink to the bottom below the oil itself. A drain valve is provided to enable the water to be let out periodically.

Front engine sump cover in place showing the water drain valve

I'd noticed when draining the water that it seemed to drain out intermittently as if there was something not quite right in the front engine's crank case. Additionally, the dipstick did not seem to register whenever I added some more oil.

The only way to find what was happening was to remove the crank case cover and have a look. It wasn't a difficult job but required the oil to be drained first (about 11 gallons of it!).

There was at least 1/2" of thick sludge in the sump.

Nice! Gungy viscous mess about 1/2" deep or more

I drained the sludge and put the crank case cover through the parts washer to produces something a little more serviceable.

Even nicer!

The crank case cover is sealed by a gasket containing a perforated metal coarse filter.

Coarse filter as found

I tried to remove the filter but it did not want to come off easily. I also didn't want to break the filter in case I couldn't get another. I took the easy way out by blowing compressed air through it to remove the remaining oil and any potential blockages.

Coarse filter after (it's not from a beehive!)

The cover is now back in place awaiting the oil to be replaced, hopefully without any gasket leakage!

Re-Chained Malady

Last time I wrote about fitting the two drive chains, I'd been concerned that one chain seemed much slacker than the other. I counted the number of links on each chain and checked that the corresponding sprockets had the same numbers of teeth. All were present and correct. 

Fellow member of the Sentinel Drivers Club Keith Reynolds visited the Midsomer Norton station open day at the end of February 2016. He explained that chains can stretch and that this was probably the cause. The stretch can be detected by laying the chain on its side and lifting one end. The curvature of the chain indicates the level of wear in the link roller bearings. The greater the curve, the greater the wear.

Although I could not do this while the slack chain was in situ, kicking it showed that it had a lot of sideways play.

So I decided that the best option was to simply replace it with another of the spare chains and take a chance on the result.

I was lucky as can be seen below. The two chains are very similar in tension now and I didn't have to do any axle adjustment.

Nearside chain replaced (the less shiny one)

The two chains are now similarly tensioned.

So a day well spent.

My apologies for the lack of blog material lately. Actually operating 7109 is great fun but without a 30ft selfie stick, it's very difficult for the driver to get any of his own photos!

Testing Times (1)

Tuesday 23rd February 2016 was an important day - the day an inspector called. In this case Peter Hawkins the boiler inspector.

It had been a busy few weeks leading up to this but essential to be ready for action on the S&D 50th anniversary of the line closure on March 6th 1966.

There were various tests: A general inspection of paperwork covering materials used, certificates of conformity and calibration; a hydraulic test to 1.5x working pressure (412.5 psi) to check boiler fittings were attached properly; a maxed-out steam accumulation test to abuse the boiler and check the safety valves could cope; a test of the ability to keep the boiler filled with water under maximum output conditions.

Giving the clouds a shock!

If you are wondering what Sentinel 7109 sounds like when being steam tested, try this video clip: (Higher def. on YouTube)

We also tried a less formal haulage test on the 1 in 53 gradient as in the following video clip (Higher def. on YouTube):

There are interesting times ahead!

(And it passed the tests too!).

Rebirth of 7109!

Blessed with brilliant cool and bright weather, Thursday 18th February 2016 was the beginning of a new life for Sentinel 7109 when she emerged from the shed at Midsomer Norton station 56 years after her retirement at Croydon Gas works in 1960.
This video clip says it all (Better resolution on YouTube).

The 'beeps' in the background are the shed's smoke alarms giving their salute!

Back into life

A picture speaks louder than words.

Chained Malady!

Just to complete my last article about enchaining Sentinel 7109, here is the new view beneath the frames looking from the front.

Drive chains in place

It's not clear from this photo that the chain on the right (of the photo) is slacker than the other. Despite very careful setting up of the axles and even more thoroughly checking that the chains were actually the same length, one chain insists on drooping. At present, I don't know the cause of this although I expect they will equalise with running in.

I also previously referred to items called chain 'slippers' and these are in the photo below.

Chain 'Slippers'

Whether these resemble Sentinel's idea of a chain slipper, I don't know; however, they did the job required of them.

Unchained Malady!

The last time I wrote about chains was in January 2012. That was to do with renovations, now it's time to actually fit them to Sentinel 7109.

After some discussion, I concluded that to fit these heavy items safely in an awkward situation was not going to be easy and required the preparation of some special tools to assist. It was going to be quite a complicated process to be able to get the chains into position mid-way between the sprockets and then get the chains over the sprocket teeth so that the two ends could be joined.

The first tool is a trolley to carry the chain under the loco.

Chain trolley

The centre of the chain is placed in the centre of the trolley with each end piled up to allow it to be pulled over its respective sprocket.

Trolley with chain in position

The trolley and chain are then put on a plank to wheel the chain into position. The upright support piece of wood is to prevent the trolley drifting off the plank and on to my lap whilst underneath!

Ready to go under

The next tool is, as Sentinel called it, a chain 'slipper' to allow the chain to be dragged over the sprocket teeth. One is needed for each sprocket.

Chain 'slippers' in place

I made the slippers from Aluminium strip left over from another job. I drilled a hole at each end and put in a bolt and nut to enable a strap to pull the ends together.

Axle sprocket with slipper

For the second slipper, I used two bolts so that a ratchet strap could be used to tighten the slipper with static straps to hold it once tight (I only have two ratchet straps so limited resources!).

Ratchet & static straps on drive sprocket

The chain was then wheeled into place...

Ready to drag chain over slippers

...and dragged over the drive sprocket.

Up...

...and over

The chain was then secured to prevent it from slipping back over.

Secured from slipping back

Then the slipper was pulled out after removing the bolts from one end to allow it to pass under the chain. (It had to be pulled quite hard!).

The same process was then applied to the axle sprocket and the two ends pulled together with a light duty lorry strap.

Ends pulled together from afar...

...and from anear with both slippers removed

At this stage, I came to the conclusion that the light duty ratchet strap was not strong enough to pull the ends together and that a heavier duty type might do the trick.
This photo shows that even that was not going to solve the problem.

So near yet so far!

I had assumed all along that the axles would be in the same position as when the chains were removed. However, clearly that is not the case and the axles and brakes will all have to be adjusted to get the sprockets the right distance apart for the chains to fit with the correct tension (+/- 2" movement mid way between sprockets).