Blog Archives
Scrap Tank Test Build Part 9 – Finishing the Body
I have now had the castings back for the various fittings for the Scrap Tank; the masters being in part my own 3D prints and some turnings that I commissioned from Jeremy Souter. This is what they look like:
I did not seek to do everything for the whole model as some parts are available from other suppliers and I did not want to duplicate their work. Thus, I needed to get the safety valve/safety valve bonnet from Alan Gibson, a smokebox door from Lochgorm, a whistle from Markits and smokebox door handles from Comet.
Once these, along with the remaining handrails, were fitted, the body is complete and it certainly appears to be taking on the character of the real thing so far as I am concerned!
So next up will be the cylinders, crosshead and connecting rods!
Scrap Tank Test Build – Part 7; Boiler Assembly and Finishing the Cab
Next up is the finishing of the detailing of the cab. Common with many tank engines there were grilles over the rear windows. For these, I toyed with the idea of doing these as a single etch, a bit like the Mainly Trains one (and possibly others) but elected instead that the slight roundness of the bars needed to be captured, so this meant that brass rods were going to be required. If I had either etched small holes or soldered these on top of the cab etch, I felt that getting consistency of spacing was unlikely and that this would detract from the finished effect. Thus, it was time for a little jig.
This jig is simply a sheet of brass with holes for the wire at the appropriate spacings along with half etched lines arranged such that when the jig is folded over, the wire is trapped between them. This is what it looks like with the wire in and the jig folded over (along with a dab of solder to hold it all still):
You will note that in the picture above, I have trimmed the wire rods to a gentle curve to reflect the curve of the spectacle plate and in the picture below, this has been soldered on the ring around the window. The jig is then snipped off and the rods can be cut away. I found that by using a scalpel, it was possible to cut a nick in the rods and then the wire could be carefully lifts so that it snapped at the point of the nick. It was necessary to ensure that the rods were soldered well to the sides as if this joint failed it was then pretty difficult to get them soldered back down neatly; I will include a space jig in the production etch of the kit to give the user a second chance!
There is also a beading around the cab side openings, a common feature on pre-grouping locomotives. This was relatively simple to fit, although I did make it a tad too fat deliberately to assist in the process – it can then be filled back to a thinner dimension and in the process any slight irregularities taken away in the filing. In this example the stanchions are probably a bit far away from the cab sheets, so there will be a slight adjustment on the final version.
Also worthy of note is the cabside number plate, which I am dead chuffed with. This is a cruel enlargement as the whole plate is only 6mm across and to clearly be able to read the text which is only 0.7mm high is pretty good I reckon!
After finishing the cab detailing, it was time to add the boiler onto the tanks/running plate and she is beginning to look like the real thing, although perhaps looking a little naked due to the missing dome and chimney at present!
I have fitted a safety valve bonnet and safety valves from those intended for the Strath/Loch and available from Lochgorm Models. I also formed the front splashers, which I had tried to make easier by the use of some tabs and formers. These did assist in the assembly but I then found that they fouled with the wheels, as I had made the splashers true to scale and the tolerances did not allow for the tabs. I will have another think here and might come up with a jig, as splashers are sometimes a bit painful to fit.
And this is what she presently looks like; definitely beginning to look like the real thing (a reminder of which is below). For those of you that are coming to Scalefour North I will bring her along for you to have a look at. As we are now about up to date with her construction (you didn’t think I can build that quickly did you?!?!) and because I am away the whole of this weekend at Scalefour North, there will be a hiatus a bit before the next posting.
Scrap Tank Test Build – Part 6; Boiler and Running Plate
Now that the much of the bulk of the above running plate work has been completed, the running plate valences can be fitted. As these are nearly always long and thin, they are prone to distortion in the kits I have built – so it is time for another jig!! This one holds the valences at numerous places to stop it flexing and to hold it straight.
With this, it is a doddle to fit the valences in their correct place and solder them without distortion. I did find that the running plate flexed significantly at the end of the tanks; so the final version is going to include a pair of temporary stiffeners that fold down and stop this. This would be the moment when they are removed to allow the valancing to take their place.
And onto the boiler. In a departure from normal practise, I am not including a flat etch to be rolled into a boiler – it is relatively difficult to get even a pre-rolled boiler into a neat tube without a visible seam and if you do not have a rolling machine it is effectively impossible to do so. In addition, where boilers have been half etched to create boiler bands I find that the half etched elements that remain are overly delicate. This was something that caught me out a while back when I drilled such and area to take handrail knobs and badly distorted the metal – this kit is still sitting in its box now and I am probably going to have to replace the boiler.
With these problems in mind, I simply used a piece of brass tube from Eileens; easier and much more durable and if I were sratch-building I would not even think of taking a different route. This did still leave the need for some rolled parts, to make the smokebox and I have sought to use another little trick here to make these easier to fit – some tags and eyes. The tags are strips of half etching that pass through the eyes and then tugged back. This can’t impart a curve into the metal but does allow the parts to be pulled tight and makes it easier to solder into place without much of lip. Mind you, they were a tad short and will be lengthened slightly in the production run.
A second additional laminate is then needed to form the outside of the smokebox and down onto the saddle.
I did find another little error when it came to the front of the smokebox. Whilst the diameter for the front that I had drawn had allowed for the thickness of the two laminates, when you fit these there is also a layer of solder between them and whilst this ought not be that thick, it was just enough to make the fronts too small. In the production run, I will deliberately make this a tad too big as it is easy enough to file it back but much more difficult to add the missing metal (I didn’t, I just made a fresh one from sheet metal). The smokebox door is not mine, the door from the Lochgorm Models Loch is the right size judging by the photographs (note the drawing in the old man’s book has it being smaller but this does not match the photos, so I ignored it in this respect – sorry Dad!).
The downside of using tube as a boiler is that boiler bands need to be considered. I have provided these in the kit (again using the strap and eye technique). I chose to fit them on this kit although in practise I think any metal boiler band is too thick and would probably have done it with a transfer sheet if this was not a test build (done prior to painting, the thickness of the transfer is enough to show through the paint on what will be a single colour to the boiler).
Only the top of the boiler is visible after the first ring and a bit, so can be cut away to leave lots of room for the motor, weighting and DCC chip. I may try and fit this with sound, so who can give a view on what it might have sounded like – a jinty is my favoured guess?
Scrap Tank Test Build 5 – Getting on with the Chassis
With the basic chassis made, it is essential to fit the nuts to secure the body to the chassis as both of these will be concealed with later work. So a quick test fit looks like this and we can get onto the next bit, the coupling rods.
As is not uncommon, these are made from a pair of layers of brass laminated together. You can see that the outer layer is half etched for much of its length, with the full depth only being present at the bosses. I have also sought to make it easier to build these by including them in a folding jig – the folding is underway in the bottom portion of the view. The logic of the jig, indeed the whole kit, is to make a really smoothly running chassis much easier to make. Modern CAD and computer operated phototool creation techniques by the etchers means that it is possible to easily draw and then etch such that each dimension is faithfully repeated on the product. Thus, it is possible to be confident that the wheelbase will be repeated exactly on each side of the frames and also on the coupling rods. However, this accuracy is completely lost if the user has to laminate the two parts together by hand; it is not possible to get them superimposed on each other exactly or repetitively so the spacings of the crankpin holes will change. The jig overcomes this as the fold line is so long that there can not be any twist as it folds, so the two parts will meet consistently and accurately.
It is true that there remain two areas of variability. The first is that the degree of etching will not be exact on every occasion so the holes will be slightly bigger or smaller on each occasion. This can be easily overcome by making all critical holes a tiny bit too small and then opening the holes up with a ream (not a file, reams will open up a hole consistently). The second problem is that a fold is not always consistent on a fold line so the jig can protect against twisting but might not necessarily put the two laminates directly on top of each other. However, the important point is is that they will be correct horizontally, any error can only crop up vertically. Thus, when the crankpin whole is opened up, it is possible that it will move vertically slightly but this will not change the dimension between the holes so the critical dimensions should be retained perfectly.
The above is all true in theory but in practise there was an almighty cock up in my artwork; so I was deprived of finding out. A total case of designer error and when this is yourself, there is no one else to blame……………….
………..I made one of the coupling rods no less than 8mm too long – doh! I have no idea how, but it needed chopping; so it was back to the old fashioned way of making coupling rods despite my high ideals! Fortunately, as they were laminated, it is possible to stagger the cut to make the splice – essentially the same technique as Alan Gibson’s variable length coupling rods. Anyway, after the cutting and splicing, I did get a sweetly running chassis and this is what it looks like. The unusually large wheels for a shunting loco are already making their presence felt!
The chassis is created around CSB’s; continuous springy beams. A spring wire is anchored to the chassis at four points per side (for an 0-6-0) and at the centre of each hornblock. Thus each hornblock is supported on either side and can “bounce” on the spring. However, the clever thing about CSBs is that when a hornblock is depressed, not only does the spring wire flex a bit as suspension, but it also rocks on the anchors so the adjacent wheels push downwards a bit to equalise out some of the deflection. It produces a really smooth chassis and, if it is conceived at the design stage, I think is actually rather easier to both design and build than traditional compensation. This is a close up of a pair of hornblocks and a pair of the anchor points (the other is hiding behind the frame spacer on the right). Also worthy of note is the colour coding of the hornblocks; to enable them to be reinstated in the same hornguide each time. This is probably unnecessary with modern (and therefore consistent) hornblocks and the accuracy of the etching I have noted but old habits die hard!
Scrap Tank Test Build – Part 4; Beginning the Chassis
Putting aside the body for a while, to take a look at the chassis because it is necessary to mount the two together and it is not possible to close up some of the element of the body until this is sorted out.
As with the body, I am trying to take a moderately fresh approach to the chassis to make this a little easier to build than certainly most of the kits I am used to. In this regard, most of the kits for the Highland are quite traditional in their design and I readily admit that all but two of my ideas has been either all out pinched from other designers or at least significantly inspired by them. All I am trying to do is use more of these neat ideas in a single kit to make the life of the builder easier. I am, however, finding that it makes my life more difficult, as there are a lot more moving parts to most components, so more places for the tolerances to be catered for; so as John Price has already said, the list of little tweeks and amendments to make is growing! At least, no one can say this particular kit designer has not built their own model.
Anyway, this is what the chassis looks like in the flat; note that it is a fold up design – this is inspired by the Mousa Models chassis, so a pinched idea!
And this is what it looks like with the basic folds made up. What it achieves is really neat, as it is instantly sufficiently stiff to work as a chassis; by the time a couple of further cross braces have been installed the basic chassis is more than robust enough for its life.
My design uses the same slide in hornblocks as utilised by Comet and Brassmasters for their chassis. After a tiny bit of practise, it is possible to size the hole for the hornguides such that these are just too small when etched. This means that with a few strokes of a light cut file on each side, the hornblock becomes a tight sliding fit. Once all of the hornblocks are in, it is then possible to measure the distance between each on both sides of the chassis and also on the corresponding coupling rod. This is done with digital callipers and by the expediency of measuring the distance at its maximum with the callipers facing outwards and then repeating with them facing inwards the average being the actual distance between the centres. I reckon to be able to measure down to 2 or 3 hundredths of a mm, which is rather better than I can build to! Where there are inconsistences, this is dealt with by a few more strokes of the file on the side which needs to be adjusted to change the centre. This needs to be done anyway to turn the tight sliding fit to a snug but smooth fit for the hornblocks to work properly soif the centre does not need to be changed, the file strokes are undertaken equally on both sides of the hornguides.
This does need to be done after the coupling rods have been formed, of which we will see in the next posting. However, the chassis is also designed with a keeper plate to accommodate all of the cosmetic springing to the model and the ashpan sides. This is secured with a series of 12BA screws to enable it to be removed to allow the wheels/axles to be dropped out. A great boon as the model is built and painted.
To make the assembly of this element easier (in fact in this case a lot easier!) I have created a jig that holds the two layers of the laminate in exactly the right position. The jig is chunky enough to avoid distortion as it is folded up and it locates the parts perfectly. In this particular case, the soldering needs to be done with care as there are folds to make after the jig is cut away and it is important not to fill this with solder before hand.
And this is what the keeper plate looks like – it is pretty delicate until it is mounted but fine thereafter.
And the two components assembled look like this. The beginnings of the cylinders are also visible, this is a slide in module that can be removed for assembly and painting (although the scrap tanks were painted fairly simply, so this is not really relevant on this model).
Scrap Tank Test Build – Part 3; Cab Inner & Roof
I have designed the cab roof and much of the cab interior to be a separate assembly, that can be secured by a series of screws. As can be seen below, there are two screws at the rear that locate into a tool box that sits on where the bunker projects into the rear of the cab. As the screws are somewhat lost in the bunker, I have come up with a little dodge where these are retained by an initial nut that traps them in place but still allows them to twist and thus engage in the cab roof assembly. The other screw comes through the top of the boiler, just inside the backhead.
The roof is connected to these fixing points with some inner liners to the cabs which can be seen here; the nuts for the rear piece are hidden in the toolbox and to the front within the false top to the boiler. You can just rebate in the rear spectacle plate that will take the glazing material.
The actual cab roof has a double skin, to aid its strength, include the lamp irons and also to assist with locating it on the cab. The outer skin includes the ribs that appear on the real roof, including a grove to allow brass wire to be used to form the seam to this. To the perimeter of this, there is a valance.
And this is what it looks like on. I find that I just can’t make roofs sufficiently well to sit perfectly on the body and nothing shouts “its a model” more than gaps where there shouldn’t be any – be this under buildings, roofs or between parts that have to be joined to structurally stand up! This is my solution, which I have used on other builds that I have done but it is so much easier when it is designed in.
Scrap Tank Test Build – Part 2; Continuing with the Body
The next stages of the test build were to do the footplate/tank sides/can exterior.
My initial design for the footplate is not particularly radical, but the test build has shown up that until the boiler is put in place (which comes some way into the build process) the front is somewhat delicate, irrespective of whether the footplate valences are fitted or not. Thus, in addition to the temporary stiffener that can be seen to the front of the footplate in the picture below, stiffeners will be provided to the front half of the footplates. The idea of these can be seen in the following view which shows the rear of the cab. By folding these over at 90o during the build, they give strength to the more delicate parts of components. Some will be incorporated into the finished article, others will simply be discarded when their job is done.
The two tanks, along with the sides to the cab/bunker, are conceived as a single piece (if you go back to my previous posting, you can see this in the flat in the etch). The two halves are separated by temporary spacers to both assist in locating them but also to give strength to the assembly prior to the fitting of the boiler which is where it will get its strength from. It was when I tackled this part, I reached the first disaster – the etchers had failed to half etch from behind so I was missing some fold lines. This was pretty frustrating as it entirely negated the intended efficiency of the design and even though I now have a corrected etch, I had to solder on by cutting the parts at the intended line of the half etch and soldering them together in the more traditional manner – exactly what my design was intended to avoid. As a result of this, there are no neat photos of the tanks being folded up and secured in place, we have to jump on a bit to see this.
The cab fronts that were constructed earlier were no slid into place and I was pleased to find that it all fitted very snugly and in exactly the correct location. I did find that I could put in a further pair of fold up tabs on the running plate that meant that it was essentially impossible to put this in the wrong location, so this is another little refinement that will make its way into the production batch.
The rear of the cab was a similar fold up unit to that to the front, which was pretty easy to build but did have one dimensional error at its base that needed cutting away – well that is the purpose of a test build! All of this, has been created from one piece in maybe three minutes!
And this is what it looks like with the cab rear in place. If you look carefully, a couple of 12 BA screws are just visible in the cut out to the rear of the cab – the purpose of these will become apparent in a future posting but it is another one of my little ideas to make this easier to build/better when built.
And this is what the cab bow looks like from above, after the addition of the splasher tops and backs. One of the issues this illustrates is that this kit, as it stands, will only work for EM or P4 modellers. There is insufficient room to get the narrower gauge/wider wheel treads into the splashers.
Next up will be the cab roof………….
Scrap Tank Test Build – Part 1; Getting Started
I took the weekend off the other week and attended the Spring Railway Modeller’s Weekend at Missenden. It is great to spend two full days just modelling away from the distractions of life and amongst people who are all doing exactly the same. I find it a form of therapy and it is well worth going if you have been thinking about it (and even if you haven’t!).
I took with me the etches that I have had delivered by PPD for the Scrap Tank; with a view to doing a test build using them. The origins of this class are some of the earliest locomotives built for the line; the Raigmore class. In an attempt to increase the life of these new enlarged boilers were fitted to them. Unfortunately for the Highland Railway the boilers were too heavy for their frames and consequently these cracked. This left the Highland with a number of new boilers, wheels and many fittings but no locomotives! Ever the frugal, they recycled these parts into a series of three shunting locomotives which were designed by Peter Drummond and these inevitably quickly picked up the name of Scrap Tanks.
These were rather brutish looking locomotives for the time, characterised by surprisingly large wheels for a shunting locomotive – something compelled on the Highland due to them reusing these from the Raigmore class which were mainline passenger locomotives with 5′ 3″ wheels. For those of you who don’t know what these looked like, this is what we are aiming at:
And this is what we are starting with:
Whilst this may (well has!) got me into some trouble, I have sought to design the kit to be easier to build than the average etched brass kit and certainly easier than the Falcon Brass kits that are the staple in 4mm for many of the Highland’s locomotives. I have sought to do this in a number of ways and the first area tackled, the cab front/interior, illustrates one of these; the use of fold up assemblies to assist not only in creating the shapes but also the laminations. Many of the modern etch designers are using these (especially the 2mm boys/girls) but I have sought to do rather more than most (which has made the preciseness of the design rather more challenging, more of which anon).
The bulk of this assembly starts as a single piece, that is folded up to form the cab floor, splasher sides and the bulk of the cab front. To assist the lamination process, jigs either side of the cab front have been used. Wire rods are slipped through the small holes in these to ensure that they are registered on top of each other properly.
The view below shows the laminations now sweated together and illustrates the square cut outs behind the cab front which are to enable glass/Perspex to be slotted in to represent the sceptical glazing. The view also shows the boiler backhead which is made from three layers of etch (not with a folding jig – yet!). I am pretty pleased with this as this is only 13 * 15mm in size, so the wheels on the backhead are only 2mm in diameter.
To be continued…………(soon too!).