The LNER Encyclopedia

Hi Dave,

I think a two-piece resin roof (or half roof) may be the best approach if future needs arise in sufficient numbers, but I would have to make completely new masters - a task I've been able (and glad) to dodge so far. I have in mind the lower roof parts with a deep channel down the middle, not fully cut away down to the ceiling level but with a bridging layer thin enough to be cut out at a later stage if so desired. The upright sides of that channel would provide faces to which a clerestory glazing sheet could later be stuck, each side. The central roof and its rounded ends, with the whole of each side completely open (under the part with a straight top) would be the other component, the end sections being solid and deep enough to fill the channel ends in the lower roof. A rebate around the edges of the open sides of that upper roof piece would seem sensible, to provide for reasonable firm location and bonding of the glazing. The sides of the upper roof piece could be deliberately inset sufficiently compared to its roof to allow for a thin further overlay for those not wishing to use paint and pen to mark out the windows and panels.

That is a similar idea to what I tried, I think mine could be improved if I were to remake the mould but so far I've manged to get enough that are good enough to use plus some spares.

A little more about the results of those currently abandoned attempts to cast one-piece clerestory roofs with transparent clerestory sides:

Here's the first attempt at casting-in some transparent material, showing that the resin managed to bleed around the insert so completely as to cover the entire outer (and inner) face with a thin film, despite the apparently tight fit in the mould. This view, from a higher angle also shows that the transparent insert had sunk into the inverted mould to such an extent as to almost break-through the surface of the central roof strip, near its edges. This made the lip at the edge of that portion of the roof extremely fragile.... ......and liable to break off during attempts to flake the cured resin off the faces of the transparent insert, thus, as you can see from the stepped edge here: Also, my most careful attempts to flake the rein off the glazing failed to avoid scarring of the surface. The inclusion of a removable space-holder in the initial casting, followed by gluing in of the glazing piece gave better results, especially when combined with the use of some little slips of plasticard cast-in to the central roof, preventing the space-holder pieces from sinking too far down into the mould and thus preserving the thickness of the edges of that part of the roof : Whilst the most convenient way to get the glazing to lie flush with opaque end portions of the clerestory sides, and to remain nice and straight along its length, was to glue it in place while the roof was still in the supporting mould, that approach generated another problem. The necessary cyano, no matter how carefully I applied it, managed in places to bleed around onto the outer face of the glazing. That not only fogged and roughened parts of the surface (which might have been rectified with careful fine scraping / abrasion followed by a wash of darkened gloss varnish), but worse still it bonded the glazing to the mould rubber in one or two spots. Within a couple of attempts, this method was resulting in the tearing free of tiny patches of the surface of the mould. Time to give up!

Deciding to waste no further time on that method has allowed me to make some very long overdue further progress on the second cleretory dining vehicle, the ECJS D30 (or GNR D 73) pantry third, which is seen here with its original multi-part "master" sides now replaced by one-piece, wire-reinforced resin castings, its interior fitted out, and two "plain sided" pieces of clerestory roof cut to length and standing ready for joining and blending on a layer of 30 thou plasticard (whose edges I have temporarily painted light brown) which will act as the gutter / cornice and as a flat ceiling that can be stuck firmly to the tops of the sides, holding them straight at the correct overall width. It additionally raises the roof to just the right height... More updates tomorrow I hope.

As you all probably noticed, I've also completed the headstocks, added buffers, painted same plus the solebars, added the strips that form the stepboards onto the lower ledges of the Z sections, and spray-painted the under-fittings. Here are the trimmed-to-length parts of the roof: A better look at the interior, with basic partitions and tables from plasticard, resin seats from the same mould I used for the D.79 seating. I decided to model the (sliding, I think) doors between the dining saloons "open" to allow a tiny bit more light through the interior, and with the intention of using the solid clerestory roof this time, there was no need to portray the partitions above cantrail level. The open doors meant that I couldn't use any the intermediate partitions to hold the sides of the carriage straight, so in a couple of seating bays I put the tables on full-width stretchers and bonded these to the carriage sides. I've taken care to get the heights of all of the features to match the previous D.79, which itself was carefully matched to the "most correct" versions of Doncaster teak vehicles that I possess. Whilst the modified 247 Developments bogies that I've "used up" on this carriage (well I took the trouble to build them, so why not....) don't enjoy the same accuracy of features and definition of detail as the versions mastered by Mike Trice under the finished D.79, I think it is fair to say that they could have been a lot worse!

When I put the two halves of the cast roof together, the right way up, on flat surface, it looked as if I was going to have some trouble lining up the features at the joint. If I lined up the clerestory portion, then the upper surfaces, the edges, and the line of the destination board brackets on the lower parts of the roof all failed to match. The symmetry and the control of thickness of my home-made castings clearly wasn't perfect! It occurred to me that the hardest thing to put right would be a mismatch at the joint in the clerestory portion, so I decided to support the roof inverted on a couple of lengths of aluminium angle, pushed tight up to the clerestory sides, with the two parts of the roof also pressed firmly down where they met, while I introduced a couple of minimal spots of thin superglue to the joint. I then checked the appearance of what would be the top of the joint before adding more superglue to get a thorough bond. I was pleasantly surprised to see that this method lined up not only the clerestory, but also the top surfaces of the lower roof and the lines of the destination board brackets. The rounded eaves of the roof needed a little trimming with file to get neat alignment, but no so much as to damage the parts of the destination board brackets that I ideally wanted to retain. A bit of filing of the underside of the roof was also needed to create a flat, rather than a stepped joint, but all in all, it didn't hurt a lot.
The cast profile is now bonded to the ceiling / cornice sheet, and awaits further attention. I would have done more today, but it was too nice an afternoon not to go out for officially permitted medically essential exercise. Naturally, I made sure that the lockdown guards checked me out, and checked me back in again afterwards.

I really wish I could report a lot more progress on my EC D.30 pantry-third carriage model. Unfortunately, the real world hasn't yet quite finished making annoying demands on my time, but I have been chipping away at the project. The joined halves of the roof have been fully blended, permanently glued and pinned to the carriage sides, end joints also blended, the destination board racks tidied up, moulded vents in wrong places cleaned off, pilot holes drilled for some of the new torpedo vents, and primer applied. A first attempt to create an imitation "glass-like shine" on the solid sides of the clerestory using car aerosol lacquer was a failure, so further rubbing down and application of more bland grey primer followed. Three coats of thinned, darkened gloss varnish were then brushed onto the clerestory sides as an alternative attempt to produce the illusion of "glazing". I wasn't massively impressed by the result but I've decided to press-on and see how it looks with the "teak" frames and panels painted on. The images below, which fail to show any hint of glossiness (which IS there, I assure you), show just the teak base coat applied by bow-pen and straight edge. I've also taken the jumper cables and sockets off the carriage ends, prepared some connector bellows with end plates, and have painted up several multi-coat teak test pieces on off-cuts of the moulded resin panels that were left-over from making my master sides. As this carriage has sides moulded in a different colour compared to anything that I've previously "teaked", I wanted to be sure of the effects of using various different streaky base coats and coloured varnishes, especially as I wasn't quite satisfied that my previous EC D.79 diner had finished up with sufficient amounts of orange in the final hue of the "teak".

Well that looks jolly good Graeme, I have not forgotten mine it's looking at me on the workbench.
As soo as the little bugger is finished it may get started.

I'll race you Dave.....

The rather anaemic rendition of the resin colour in those photos above isn't quite how it looked to the human eye. The sides and ends have now had the first stage of my chosen teaking process, an uneven "grain oriented" brushed coat of a 2:1 mixture of clear satin varnish and Humbrol 113 (apparently "rust", but not to my eye) slightly thinned. The basic resin had been coloured with Humbrol 133. The whitened 133 and the streaky 113 seem to have given just enough of a suggestion of wood grain, without the 113 beginning to dominate. Fingers crossed.

I should put in a word of thanks to Chris Knowles-Thomas of SRG Phoenix Coaches by the way. He kindly confirmed my belief that some of the "oddments box spares" torpedo vents that I'd used on my D.79 diner had come from SRG and has supplied me with some more to match so that the D.30 will (I hope) have maximum visual compatibility with its next of kin.

Atlantic 3279 wrote: ↑Mon May 04, 2020 2:28 pm....SRG Phoenix Coaches by the way. He kindly confirmed my belief that some of the "oddments box spares" torpedo vents that I'd used on my D.79 diner had come from SRG and has supplied me with some more to match...

Wow, I thought Phoenix Coaches had ceased production many years ago!

I think the official line on the website is that kit production is "suspended" while the question of worn tools is reviewed. It would seem that certain parts may still be had.

With any luck, this now looks less iron-deficient, after phase 1 of the "teak" application: I've also been off at a bit of a tangent this afternoon. An acquaintance wanted to know if I could supply a boiler to suit a Gresley K1. The only thing I had of suitable diameter was the J6 pattern, but with firebox too short, barrel too short, smokebox too short on wrong kind of base, and all of the firebox fittings in the wrong places. My acquaintance was quite happy to have one of those as a basis for work, taking my view that "nothing is impossible". His plan intrigued me though, because at some stage I may need an extended J6 boiler for another project, hence I was keen to see how an extension of such a boiler might work out. I had cast two boilers, just in case, and this afternoon I got out the razor saw, files, glues, plasticard, abrasive paper and filler.
Now, instead of this: I have this: I hope my friend doesn't mind me taking part of the job off him. Sorry the pictures are dismally grainy.

Atlantic 3279 wrote: ↑Mon May 04, 2020 6:55 pm I think the official line on the website is that kit production is "suspended" while the question of worn tools is reviewed. ...

That review has been going on so long that I took it to mean "ceased production"

Horsetan wrote: ↑Mon May 04, 2020 9:41 pm

Atlantic 3279 wrote: ↑Mon May 04, 2020 6:55 pm I think the official line on the website is that kit production is "suspended" while the question of worn tools is reviewed. ...

That review has been going on so long that I took it to mean "ceased production"

You've evidently been looking rather more than I have.


Teaking attempt, phase 2: Tiny amounts of Humbrol 69 yellow, on just the bristle tips of a 3/16" flat brush have been brushed/scrubbed out almost as far as they would go, in the direction of the mock grain. It is little more than dry brushing, the surface certainly being dry to a brief touch as soon as the brushing is finished, but the change towards a mixed, grained, "woody" hue is quite marked. My test-pieces by the way strongly indicated that the vivid yellow of Humbrol 69 gave much better results (when followed by my mixture of Phoenix Precision teak paint and clear satin varnish) than either a paler yellow or and orange-yellow. I felt that the former made the final teak colour too dull and earthy, the latter made it too red. A clearer image of that extended J6 boiler for Gresley K1: The base of the smokebox is a challenge for the end user to address.......

TTFN

Teaking phase 3 completed earlier today. Unlike the almost dry application of the yellow coat, trial soon confirmed that the phase 3 mixture of Precison Teak and satin varnish had to go on wet enough to allow each whole panel or each group of window frames to be fully coated and brushed out to smooth finish while still brush-able, with just fairly subdued flowing grain effect. Drier application produced a rougher surface and too greatly exaggerated grain effect in my opinion - almost the dreaded "first attempt at wiping down after a collision between a yellow coach and a van load of cocoa" appearance. In needlessly cruel close up, this is the current state of affairs: Another nice even coat will be needed.

I then set up the following picture, to reveal how my attempts at a reasonably consistent teak shade have drifted over the years, not helped by loss of availability of the Dulux Brushwood teak varnish that I was originally using. The Gresley FK, top right in this view, was the product of my original teaking recipe, and I tried to emulate that on later vehicles, using different methods. The "cheap and cheerful" clerestory BFK conversion top left had to be done quickly, when we were trying to build up stock quantity and variety for Grantham, so that featured a an attempted "short cut method" of only two coats over the Margate biscuit-coloured plastic, both coats slightly different mixtures of browns with clear varnish. Although "okay" it lacks the golden /orange glow of what I see as proper varnished teak. To my consternation, I actually went even further adrift with the EC D.79 restaurant car bottom right. I believe its dull brown tone stems from the starting point of a relatively dull earthy brown moulded resin base and too pale a shade of yellow used as the dry grainy middle coat. The new D.30 pantry third, bottom left, seems well on the way to a fair match to the original FK,and with any luck it will end up just a touch darker and duller so as not to clash too badly with its D.79 partner. It's a good job I already have variety in my fleet, and Grantham's stock adds to that variety. In a train, the uniform lining helps to make slightly different teak finishes on carriages look more compatible anyway...

A further (phase 4) coat of browned varnish has helped towards the desired shade on the new carriage, although it is not quite as I want it, yet. The clerestory top and sides are also drilled now in readiness for the plethora of torpedo ventilators and gas lamp chimneys. I've departed from the Isinglass diagram in respect of some of the positions, based on what I believe can actually be seen in the photograph of No 314 in Hoole's book, although the leafy tree behind the carriage doesn't help when it comes to confirming the roof fittings!