Musings from Bridgnorth pattern-shop

Occasional notes from the team of volunteers beavering away at Bridgnorth's Pattern-Shop.

Brian Oldford has been busy re-inventing the wheel! For 82045 in this case. He writes:

30th June
“Excellent progress over the weekend saw the completion of the 5th pair of GRP mouldings for the spokes (bar trimming) and significant progress on the plywood skeleton frame to which these mouldings will be attached.

What started as two 8' x 4' sheets of 15mm plywood is now 2 discs over 5ft diameter with 17 holes appearing in them which correspond with the inter-spoke gaps along with a large pile of waste wood shavings. The piercings for the inter-spoke gaps are being made using our larger router fitted with a template following collar and a 1/2" router cutter. To guarantee the radii between the spokes at the hub and between the rim and spokes would be consistent I made a template for the collar to follow.

The two 8' x 4' sheets were initially "cut and shut" to make the blanks large enough from which the discs could be made. These have then been temporarily screwed together so that both skeleton frames will be identical. I estimate that less than 30% of the original two sheets of birch plywood will be find its way into the final wheel pattern!"

2nd July
"Here’s what hasn't so far ended up as waste wood from the driving wheel formers. 15 holes to complete. The spoke outlines will be "clad" in the GRP mouldings shown below to progress the 3 dimensional profile. The addition of the crank-pin boss will see the wheel's completion.

The two discs (temporarily joined to be identically machined) are 5' 0 1/2" diameter and thus could not have been made from a 10' x 5' sheet without doing a "cut and shut" job.

I'm confident the router cutter will hold its edge to complete all the piercing despite the abrasive nature of the adhesive between the plywood laminations. Thanks to Titman Tools.”

I'm using a standard 1/2" (12.7mm) router cutter (no bearing) inside the concentrically fitted 30mm collar. The collar simply slides along the template guiding the cutter rotating inside it. The template was made so it was exactly 8.65mm back from the wanted cut line. Standard formula (D/2 - d/2). (Note the 2 thou accuracy!).

During a subsequent stage I will be using a bearing guided chamfer cutter to relieve the plywood to cater for the tumbleholm of the GRP mouldings. I will leave a few mm vertical edge on the ply to which the inside of the mouldings will be fixed. I have yet to see how stiff the 15mm thick skeleton frame will be with only the moulding attached. If it isn't stiff enough I have a number of options:
a. Fit a perpendicular plywood stiffener inside each spoke void.
b. Fill the void with PU foam (very carefully!).
c. A combination of the two above. With that I'd need to be sure I can pour the creamy foam mix inside the remaining reduced void."

3rd July
“Today's efforts proved more productive than I hoped. In addition to finishing the piercing and applying the chamfer I was able to apply the gel coat to the moulds for the next GRP spoke halves and trim another pair made earlier but not trimmed. Just before finishing this afternoon I was able to offer up the first spoke and start the fine trimming to get the fit to the former I want.

The pattern is a straightforward two piece split pattern with each half of the pattern is being built-up from the ply-wood datum.

Although the theoretically perfect split line should include a joggle, the application of a little pattern-makers licence obviates the need for an oddside over 5ft in diameter. This is because inspection of a vertical section through the wheel reveals spokes that are slightly angled to the true vertical, i.e. cart-wheel fashion.

This angularity makes for a more flexible wheel that is presumably less at risk when cast and provides a degree of resilience from road shocks in service."

4th July
"Here's a couple of snaps from today. I've started adding some internal reinforcement to one part of the split pattern and trimming the GRP segments so they do not overlap at the hub. When all the internal reinforcement has been added to both pieces of the pattern I'll seperate them and start permanently fixing the spoke segments.

Interestingly perhaps, is the fact although a wheel pattern is fairly large it isn't that complex. There are no complex cores to place accurately. Much of it is quite repetitive once you are set up. The tooling for the spoke segments took a bit of time to do but once done, production is a fairly simple process that's well documented on a number of web sites e.g. CFS Fibreglass Supplies.

Similarly the addition of the stiffening pieces today is very much a repetitive process of cutting 360/17 degree mitres in strips of plywood made from the off-cuts (no shortage there then!) and glueing and cramping them onto the formers with PVA adhesive."

5th July
"Today saw the start of permanent attachment of GRP spoke segments to the plywood former. The gap between each section of the rim allows me to get the adhesive inside easily."

6th July
"The wet weekend finished with me dashing in and out of the pattern shop between showers. A 5' plus diameter pattern isn't that easy to work on in our shop! I did manage to get all 6 completed spoke-half pairs permanently fixed to the 15mm ply backings and I have laid-up a 7th which I hope to remove from the moulds one evening this week. 7 down; 10 to go!"

8th July
"I managed to swing via Bridgnorth this evening and remove the 7th pair of spoke halves from the moulds. Whilst there I found a defect in one of the moulds where, in a small area, I failed to ensure the white gel-coat was fully supported by the resin soaked CFM. This has lead to cracking of the unsupported gel-coat. A number of repair options can be followed. On this occasion I picked out the crazed gel-coal above the void and poured in some new flo-coat to stand a little proud of the original surface level. This will be rubbed back to the original level using progresive grades of abrasive paper in a similar manner to that used when I made the GW tender number plate.

And, just in case I get bored, I have started studying and mulling over the drawing for the driving wheel horn block. It has a number of insidious undercuts that will require awkward coring or loose pieces.."

82045 now has 2/17ths of the driving wheel spokes!

As you can see from the above photo the second pair of spokes haven't been rough trimmed to remove the surplus CFM (chopped fibre mat). The possibility exists that they may seem a little flimsy when assembled. If so Brian will fill the hollow cavity (very carefully) with expanding foam to reinforce it.

The model/plug of the spoke and its corresponding "inner" half, shown below, had been coated with red flo-coat. This was sanded back with successive grades of abrasive to produce a really fine finish. Subsequently this was coated in PVA mould release agent; and today received two coats of white gel-coat prior to lay-up of 3 - 4 layers of chopped fibre mat. When this moulding is removed from the plug it will form the mould within which each of the 17 spoke halves will be moulded.

For the technically minded; 
a: Flo-coat is nothing more than gel-coat with the addition of 2% wax. The wax rises to the surface and thus inhibits the air inhibitors in the gel-coat from working and thus makes the exposed surface cure to a non-sticky state. 
b: The sole reason for using different colour pigments in the gel-coat is so that is easy to see the coated the underlying material 100% without bald areas. The spokes will be layed-up using red gel-coat (inside a white mould) thus avoiding the need to paint them with Burns Patterncoat. Sneeky!

Work has also progressed with 82045’s buffer beam brackets. The coreboxes for both the left and right handed brackets are complete bar minor fettling and final paint. The patterns are almost fully assembled with only the coreprint portion of three of the four pattern portions to build before final fettling and paint. That little lot has pretty much used the two sheets of ply. The photo above shows the now complete patterns.

This photo from Gordon Woodruff shows the pattern set to simultaneously make four of the ends of the table legs for LMS RFO 7511.
This part sits in contact with the floor and has a screw adjuster to level the table itself. An original can be seen inside the boxed cluster of patterns. The other portion (to hollow out the insides) was made by laying up GRP over a model of the wanted sand "core" portion. Guess why Gordon now knows to get the surface dead smooth of a plug and use plenty of mould release agent?!

The pattern set for GWR water tank fillers made for Collett Saloon 9103.

Meanwhile, between footplate turns Neill Parker is re-making some letters for the early GWR style step-plate for the stairs in Wrangaton Signal Box at Kidderminster for Dave Postle.

GWR 13" combined ejector blower ring. Made back in the 1990s it has been in the pattern shop for additions of omitted features. When Brian created the drawing back in the early 90s from some photocopied scraps he then promptly ignored the relevant part of the drawing. Ho hum; to err is human!

We've started the patterns for 82045's bufferbeam gusset brackets now, which remarkably are steel castings on the Class 3s. (To the best of our knowledge all other BR standards use fabrications for these components).
 

Brian has also just started work on 82045's driving wheel pattern. Basically he's made a model/plug of one spoke (actually two halves since it is a spilt pattern) from which we will produce a pair of GRP moulds. In those moulds we will lay up the inner and outer halves of the spoke 17 times. Cue a remark about 16 1/2 spokes!

Also using GRP we're in the process of producing models of the cores of the 10mm Monitor injector. These will be assembled in exactly the same relative position to each other the sand cores would be and then more material will added to represent the body wall thickness. We're trying this approach to try and get all 9 (yes nine!) core prints to locate as perfectly as possible.
 

Gordon Woodruff has recently completed the blower valve pattern for larger Stanier engines.

Meanwhile the 10mm monitor injector body makes progress (slowly) towards completion.
 
The 82045 web site is a little premature as we haven't actually received the plywood for the patterns and coreboxes for the bufferbeam brackets!

The following photographs show the completed pattern, coreboxes and oddside for the pony truck axlebox keep for 43106. The observant may notice it is to be cast with two filler spouts. In practice one or the other will be removed prior to through drilling thus creating either a left or right hand variant.

This is a pattern set for the LMS/BR water tank filler pipe end fitting alongside an original. Unexciting parts perhaps, but important in helping us to make sure the loos flush!

February-2 2008

These are the two halves of the steam pipe clamps for 43106. The first is the inner portion complete with core (to deal with the flange undercut) and oddside to simplify the moulders task. The second is the outer portion which only required an oddside.

This is an 8mm Monitor injector body mounted on our small horizontal borer which was cast last year from a Bridgnorth produced pattern. The pattern and corebox set was made from a drawing (showing all the internal detail) produced by Charles Lamont who painstakingly reverse engineered an existing cast body. Honour for the machining must go to Graham Bennett using the vertical mill and small horizontal borer to get to this stage. An idea of some of the internal galleries can be seen in the photographs of the coreboxes for the 10mm Monitor injector.

For those that missed the first episode in horn ties these two photographs show the two new castings to replace the cracked ones removed from 43106.

No credit can be attributed to the pattern shop for this photograph which shows two wheels being machined for the "Catch-me-who-can" replica being built at Bridgnorth. These are seen being machined by Dave Reynolds on the twin vertical borers from flame cut blanks.

The pattern and core box for new smaller, 5 1/2" diameter, dynamo pulley. Made for Kidderminster Carriage Works, it suits all carriages with a Stones dynamo, that includes all MK1s, the Venturer kitchen car and the Gresley buffet car.

Pictures of the old snifter valve along side a new casting for 43106. Unlike many of the other patterns we've made for 43106 this one isn't of use to 82045 as BR adopted the western style snifter for the standards.

This group show a few patterns that we haven't pictured before: GW28 is a blower flange to suit 4566 and 7812. GW17 is a valve rod cover for 5532.

A close up of the valve cover.

GW09 is a steam heat cap for 43106.

This pattern is a lamp iron converter that allows GW style lamps and headboards to be fitted onto LMS/BR lamp irons.