MikeTrice

Having produced the end drawing time to start adding some detail. As part of my guesswork I am assuming that the end steps should be 9" wide by 6" deep, totally unfounded of course . For strength my preference is to use a moulded angle as the basis for making them. Take a piece of Evergreen 2mm x 2mm angle (Ref 292). Unfortunately they do not do an 'L' section so we are going to create one: Now take two steel rulers and clamp one of the angles of the styrene then using a knife run it along against the edge and trim off the unwanted bit. Sorry if the photo is not clear. I staggered the rulers to try and show up what I mean, probably unsuccessfully: Now take a piece of prepared angle and cut off sections 3mm wide: You now end up with a small step, here rested on its shallow side: Carefully pare down the corner of the shallow side forming a rough curve: Repeat the process until you have enough steps: The two lower steps are solvented in place, making sure you are using the right end panel: In this instance I used d-Limonene which gave me a little more time to get them positioned and aligned correctly. A pair of dividers were then set to measure the position of the top step from the base of the coach: This was then transferred to the coach and the point on the beading scribed: Using a knife a section of the panelling can now be removed: The top step can now be fixed in place: Repeat the process for the remaining steps, sit back and pat yourself on the back that it went OK:

If not you can save it as DXF and import it into Studio.

Any chance of a photo?

I mentioned earlier a lack of detailed information relating to various details, which are sort of holding up further progress. Last night I spent time going through various references, pulling out what little hair I have left in the process. In the end after a lot of guesswork I have produced the following arrangement of the steps and handrails, drawn in Inkscape of course: If anyone knows of any glaring errors please speak now or forever hold your peace. I have also decided to add a continuous rain strip to the roof so have updated my roof template accordingly: As far as Inkscape and the Silhouette is concerned there is little left to do, just the axleboxes which will have to wait until the Brassmasters Cleminson chassis arrives. After that it is plain scratchbuilding and soldering. So the question is should I continue in this step by step fashion, or should I just show the finished item (assuming I finish it)?

It sounds as if you are using it to laminate the two pieces together to hold a shape. I have not done that with any solvent. Geoff Kent's approach is to just glue at the roof edges.

Thanks Mike. Try the UHU solvent free, I was surprised just how strong a joint it produced.

51L London and North Western Railway Carriage Torpedo Vents (LNWC042) and 51L Midland Railway Clayton Carriage Gas Lamps - for arc roofed vehicles (MRC041) added. Tiny drops of Uhu solvent free were applied with a cocktail stick and the castings inserted. Will cleam them up a bit more when the glue has dried: Have ordered a Brassmasters Cleminson chassis (RCH type ref MC002) to provide the underframe working parts.

Nick's drawing suggests that you are correct. One each end for the toilet and one in the middle. Looking good by the way.

Question: should the roof have rainstrips?

In order to work out the vent and lamp positions on the roof, rather than measure and mark up directly on the roof, I created a template in Inkscale and printed it off: The printout was cut out and fixed to the roof with a Pritt Stick: Pilot holes were then drilled: The paper template was removed and the pilot holes opened out. They need finishing off:

In hindsight I decided to increase the false roof thickness from 30thou to 40thou. This involved reducing the partion heights once again, this time by rubbing upsade down on a piece of 600 grit wet and dry. Here is the new false roof in place and marked to reflect the partion positions: The roof profile pieces produced previously have been tidied up and fixed in place: The roof itself cut from 30thou styrene being curved the same way the turnunder was produced: False roof back in place: Checking that all the profile pieces are level with the ends. IF there are any high ones, turn upside down and rub on wet and dry until level: The roof is now stuck to the false roof profiles and put under weight to cure: A trial fitting. Not bad.

10thou straight from the cutter. The styrene has a shiny surface so this was mounted on the mat, gloss side up: Some of the components. Bolections bottom right. The side panelling is laid on the sides and clips used to keep it aligned. Once everything is in place, d-Limonene was applied to fix in place: The close up here shows how the 20thou layer needs its window and door openings filed to size following the snapping process especially the 3rd quarterlight from the left, but not yet: The bottom reinforcing strip is cut off and the interior has been refitted to show how the sides a reasonably straight: In reality, not all the parts have cut right through so a little work with a blade is required to complete: The end panelling is located using the clips and solvented in place: Once everything has cured for a few hours the side panelling can be rubber smooth on the 800 grit wet and dry and to remove the gloss finish. The interior is left in place to add strength: The ends are also treated the same way: Door vents and bolections added. Now they need to cure fully before I can finally finish the window apertures.

I am cutting direct from Inkscape, however I could do as you do and use colour separation for cutting the various elements, as the Silhouette driver asks you to select which colour line to cut. To be honest, I have not thought of it. Panelling cutting out in 10thou as I write this.

Nearly there. First job is to find that bolection created a long time ago and change it to black. Now the various components are collected together, duplicated and arranged ready for cutting. I actually need 20 bolections but always include some spare, like the spare droplights. Note, that unlike the earlier 20thou components I find it best to leave space between the various parts, it seems to give cleaner cut. The bounding rectangles are also unnecessary as generally 10thou will cut through. This is the 10thou scribing version: The whole arrangement is duplicated and moved to a new layer and the door outlines deleted. This is the 10thou cut version: We are now ready for cutting (and more modelling).

Once again a quick image to show which parts we want from the side for the panelling layer, as shown in blue, most of it really: Taking out the unwanted parts. Oops, missed one droplight: Got it! Past experience tells me the lower beading is quite fragile, so I have added a reinforcing rectangle below it and various rectangles that will form attachment points: Selecting the side's outer rectangle and new new rectangles the whole lot are combined. That is the side done: Duplicate one of the doorways from the original side so I can start producing the droplights: Here I have drawn a rectangle round the droplight which will define the extend of the droplight: Here the unwanted parts are deleted leaving a finished droplight. Just for variety I also like to create some droplights in lowered position so another clone has been created and the bounding rectangle reduced in height to where I imagine the prototype would be when in the up position: The two droplight elements are moved into a lower position: The lower edge of the outer rectangle was then moved up: The inner shape is converted to a path and node edit selected: The lowest two nodes are deleted and the bottom edge converted to a straight line on the toolbar: Now the lower two nodes can be moved up to just below the opening: Having created one open droplight it was the work of minute to create another even lower version:

I should add that working from carriage diagrams is not an accurate way of producing cutting patterns and as will be seen earlier I doubted some of the dimensions shown. Lacking any reference material other than the preserved 818 clerestory a lot of this is down to guesswork. One of the previously prepared panels is duplicated and positioned to the left and the top adjusted as before to match the arc: Zooming in on the bottom edge and switching to node edit mode, the two lower nodes are selected and the "Add node" toolbar button clicked creating a new central node: The left hand node is moved vertically up (while holding down the Ctrl key) and the middle node moved left horizontally: The two left hand nodes are converted to a curve and the curvature introduced to match the profile guide: The three left hand panels are duplicated, mirrored horizontally and moved to the right hand edge. Just to play safe I selected all 6 and aligned them to give equal spacing then deleted unwanted objects. We are left with this finish end: One end needs a modification to the outer panel to allow for the alarm gear to be fitted, so the produce end is duplicated and the original moved aside for later use. On the other one the left hand corner is zoomed in and I have drawn a rounded rectangle to act as a guide line for the modification. I could at this point combine the two objects but decided to do it a slightly different way: Instead what I did was break one of the nodes at the corner and then move curve to side side of my guiding rectangle, then draw in the missing parts by hand. Node editing allowed me to fine tune this: The modified panel was duplicated, flipped and moved to replace the one on the right, then the old right hand one deleted. I now have the two panels finished:

As you will see, I sort of made up the next bit as I went along. I duplicated the roof arc and moved the copy down. I have drawn a lower guideline for the bottom of the panelling and duplicated the 20thou curve at the side and set its width to 0.75mm then moved it to align with the new edge of the end profile: The new side was duplicated and mirrored and moved to the left to align with that edge. Vertical lines were added in 0.5mm to represent the beading. Additionl beading (shown in red) was added and aligned to the inner edge of the 0.75mm line. All 7 lines can now be selected and using the object align tool, evenly spaced: I thought I would use the inner edge of the 0.75mm line so converted it to black and started breaking it so I just had the inner surface. In the end I did not need to break it as will be seen in the next post: The first panel outline is added using the straight line tool and the various sides and lower edge adjusted to match the guidelines: Zoom in, switch to node edit, select the left two nodes and then change to a curve using the toolbar button: As it has been converted to a curve I now two handles to manipulate and can start changing the corner to a curve: Repeat for the right hand side: I then did the same for the top two nodes so I could curve the line slightly. The original panel outline is duplicated and moved to the next panel position then the top elements moved in node edit mode to suit their new height: So this is what it looks like now:

Thanks SS. It is early days yet but the use of d-Limonene looks very promising. After 4 days a test peice using 20 thou is showing no sign of warping. A similar test piece using 10thou is showing very slight warping but nowhere near as with the traditional solvents. Think I will try it when laminating the panelling and the rest of the construction. Of course to be able to complete the body I do need to do those ends. I cannot put them off any longer. You might also notice that as this topic progresses I have used less basic instructions as by now there is a lot of the same and additional explanation should be unnecessary, however if anyone want me to cover anything in more depth, let me know. First job is to draw the required outline. Here is the end profile I produced earlier in the thread. I have changed some of the colours in an attempt to highlight what I am trying to achieve. In this instance I am trying to create the red line. After duplicating the original profile, the first job is to start deleting some of the unwanted lines: The outer curved lines are converted from a stroke to a path and the fill turned off and the stroke width set to 0.05mm: Switching to node edit mode the outer nodes are broken apart, the resulting path broken apart and the inner unwanted bits deleted. Note that the roof arc stops short of the side. This can be extended by switching to node edit mode, selecting the end node and dragging it down and to the right until it meets the side. The same needs to be done on the other side. Here the roof arc has been extended: Finally a line needs to be drawn to join the two lower edges of the sides. It was easier to delete the floor from the previous images and just draw a line from scratch. All lines selected, changed to uniform thickness, coloured black and bitmap switched off:

The main body joints have now had time to harden sufficiently for the various ends and sides to be sanded to remove any burrs and joints at the ends using a piece of 800 grid wet and dry on a flat surface: Two partitions glued back to back with seat squabs crudely represented with Evergreen 4.8 x 7.9mm rectangular tube. As both sides are symmetrical solvent has been used to fix these: A false floor of 30thou has been fitted and some of the seat/partitions fixed to the floor. To avoid warping Uhu Solvent Free All Purpose glue was used. Remember the unit needs to be removeable so do not glue the partitions to the sides and try not to get glue squeezing out from the floor to the coach body (guess who speaks from experience here!): . The two end partitions having seats on one side only have been glued with the Uhu non solvent glue NOT solvent: Seat backs from 30thou have been cut and fitted. Again for the central partitions solvent was used for this, but for the ends Uhu: Not a warp in site. Success at last: I mentioned earlier managing to get some stray Uhu glue attempting to fix the interior in place permanently. I did manage to separate them but the Uhu is surprisingly strong. An alternative to using the Uhu would be to use Cyano however I found I preferred the working time of the Uhu.

Before continuing an important diversion: styrene warping. Nothing to do with Inkscape, or the Silhouette cutter, but very important none the less. Have a look at this earlier false floor unit based on a 20thou false floor. Nice isn't it? The problem occurs where solvent is used on one side of the styrene only. This test piece was made and photographed one day out of 30thou styrene: Just one day later and it is starting to warp: What I believe happens is that as the styrene/solvent cures it shrinks resulting in a pull on the side applied. Going back to the original floor this partition is made out of 40thou styrene, but more importantly the components attached to it are symmetrically arranged and as a result does not warp. On the end panel though components are only attached to one side and as a result it warps (it is not helped by the thinner material here: So going back to the original photo of the interior the worst curvature is where the seating units are solvented to one side only of the false floor. Result, it warps. This was a problem with my original coach design. Here individual laminations of 10thou just did not have enough strength to resist the warping: Even my latest attempt is not immune. The corner joints are also subject to warping resulting in the sides pulling in: This is not a problem for this build as I will be including an interior with partitions and a false floor. An obvious solution is to limit when solvent is used, which I will cover in a later post. Another option I have seen advocated is the use D-Limonene which is supposed to be less agressive. This huge bottle was bought from Magnacol for the princely sum of £7.50 + £3 postage, a test sample has been glued together and I will monitor it over the next few days. Obviously styrene thickness will also play a part however to date problems have been experienced below 40thou. You have been warned.

In best Geoff Kent fashion, the side is layed face down alongside a steel ruler and a craft embossing tool rubbed along beside the ruler to start forming a curve. A knife handle could be used but remove the blade first! The side is then moved further up the ruler and the embossing tool used again rubbed along its length: Keep going and you will end up with a nicely formed turnunder which can be checked against an end: The turnunder reinforcement was also curved and offered up in place remembering to leave a 40th ledge at the bottom to take the floor. In reality I felt the reinforcement strip was too close to the windows so decided to remove 1mm from it. The spring clips alow fine adjustment to be made until solvent is applied to weld the two part together: Here are the ends which like the partition are doubled up from 20thou. Burrs are cleaned off using a suitable abrasive. I find nail emery boards useful especially where they have a cource grit one side and a medium on the other: I find marking out styrene and getting good accurate cuts rather frought of late. This is down to poor lighting and using a pencil that does not have a fine tip that does not give a very visible line. Instead I now use a 0.1mm UNI PIN Fine Line pen which creates a highly visible, thin mark. Here I am measuring the width of the floor in 40thou using an end as the measuring guide: Before measuring the length, the end is squared up with a small engineer's square: Again following Geoff's guidance the first end is fixed to a side using the square to ensure it is at right angles. The other side is also fixed in the same manor ensuring the end is fixed in the same orientation: The two sides are fixed together and the floor is fixed in place and the whole put aside to cure. I should have follwoed Geoff's advice and measured the floor length against the assembled sides, but thought I knew better. Wrong, as you can see by the slight gap nearest the camera! Before leaving to cure I should add that I fit some of the interior postions but not fix them in place. This is just to help stop the sides being pulled in by the solvent (more in the next post): Additional partitions in place. I did notice at this point that they were slightly taller than I expected so ended up shaving 0.5mm off from the top straight edge:

A piece of 20thou styrene cut to the previously determined size and mounted to the cutting mat. I tend to use a roller to press it onto the mat: The carrier sheet fed into the machine: To provide extra support whilst cutting I use a suitable book: Not the easiest shot to take but you can just make out the cut lines on the sheet after removal from the Silhouette: The basic rectangles snapped out: Now individual snapping out to profile can take place, sometimes aided by a knife: The inner partitions are doubled up to form a 40thou component. Here it has been snapped and is then glued without being separated. Be awarethat the ones that go either end of the coach are only single layer: Snapping out the windows. I have used the knife to cut through more of the diagonal lines and then push out the remaining bits using a blunt tool. This is one I picked up at an exhibition somewhere:

I have been trying to drum up the enthusiasm to post an update to this topic. Having got so far, it has taken longer to write it up than it has to actually do the work, and I have been getting withdrawal symptoms from actually doing some modelling. As a result I decided to change priority and finish the cutting pattern for the 20thou parts, which I could then start assembling and postpose finishing the 10thou panelling until later. It also means I can give you a break and do some posts about construction. Anyway, back to Inkscape. Here is the side view we completed earlier. It is necessary to visualise those elements that relate to the 20thou side, so to illustrate it I have changed the line colour to cyan. Other lines will be deleted. Here I have started deleting some of the panels but stopped from deleting the lower panels: This is so I can zoom in on the door and trace the scribed edges of the doors: The scribe lines are duplicated and moved into position for the other doors, then the remaining unwanted panels deleted: Like a number of other people I do not have any joy cutting 20thou right the way through and have to effectively score and snap. To simplify snapping out the window apertures I have added diagonal cross scores: This is repeated for the other apertures. I have also drawn a rectangle to represent the lower turnunder reinforcement. The length is adjusted to be the side length less 2 times the end thickness, e.g. 2.4mm. Why 2.4mm? Well I often find that 20thou styrene is often 0.6mm rather than the stated 0.5mm and it is easier to sand the sides to length to match the ends than sand the ends down to match the sides: Here is one of the door vents showing again in cyan which line we want to keep: 3 lines are drawn and aligned to give an equal spacing: These are then adjusted to be central to the vent: Once again to aid snapping a bounding rectangle has been drawn: The finished vent is duplicated and the rectangles butted up. Typically with the smaller parts I add a few spare: Here the components so far have been duplicated , the previous parts likewise duplicated and everything roughly positioned: When cutting 20thou I tend to butt the various rectangles up to each other: Again to improve snapping an additional rectangle has been added to the end ofthe reinforcing strips: Everything positioned to my satisfaction. I have added another rectangle to the top right to square everything up. Snapping out should now be straightforward: The existing objects are duplicated, making sure they are not moved, and transferred to a new layer which will represent the scoring layer. The non scoring layer needs the score lines deleted: Here you can see the various layers. Rather than just send a single drawing to the cutter multiple time, I find it easier to just duplicate the template and move to a duplicates layer: As my pattern does not fill an A4 sheet I have drawn a rectangle in order to determine what size to cut the styrene for mounting in the machine: We are now ready to cut. Hide the sizing rectangle, set up the machine and load the styrene on the cutting mat than Print and select the Silhouette printer. On my machine I set the option to indicate the use of the carrier sheet. Clicking on the Cutline Settings tab, then Modify Color to change the line colour to Black Click on OK, then repeat when the Ready to cut dialogue appears, then cross your fingers that is cuts correctly.

It would be nice if it all worked as expected. Unfortunately I have found some interesting 'features'. Here I have drawn a line. The measurement system at the top states it is 24mm high. The measure extension on the other hand gives a different dimension: Zooming in on the end of the line shows that the handles are outside of the node points of the line. Selecting each node individually and noting its y value suggests that the line is nearer the measured length between nodes: My suspicion is that the length on the toolbar is based over handles. There is a setting in the Stroke settings to allow the line to be extended to match the handles and it does now print, as far as I can tell, to 24mm length, so use the measure extension with care!!!!!

I don't want to put you off, but the restored coaches have got the lettering wrong so don't assume they are prototypical. Will be sending you a separate email.