Bath Spa - High tech modelling: Bringing 1947 into the 21st century. 3d printed scenery, cnc milled track, laser cut baseboards and computer control

[HillsideDepot]

The photo in the book is taken, I think, from platform height, with the turntable just off centre of the frame, so it's more elevated than the first CAD, but not as much as the second.

 

As for other photos, there is a 1936 OS extract, a 1897 photo with a Bath stone signal box partially hidden by a 2-4-0, an 1900 photo of an 0-6-0 shunting (the one with the shunting horse stable), a 1921 general view of the platforms with coaches stabled in the centre roads, the turntable shot, a shot taken from the turntable of a Castle in the middle road, a 1959 forecourt shot and then 6 diesel era shots.

 

The turntable shot is credited PB Whitehouse/Millbrook House which Googling suggests could now be held at the NRM.

 

Not sure if that adds much!  

[richbrummitt]

Here you go:

 

 

Some of the planking is a bit haphazard. It seems that the broad gauge 'rail' is left and the narrow gauge has been filled in at the one side. Note what appear to be locking features.

 

 

It seems that the outer ring was cut on site when the gauge was narrowed if I'm interpreting the evidence in the next photograph correctly

 

 

...and another for good measure.

 

Edited July 1, 2013 by richbrummitt

[queensquare]

I had heard that there was another railway in Bath........something about green engines?..........

 

Seriously, I was given the tip about this thread at the club this evening and have just had a quick read through. It looks to be a fascinating project and I look forward to following the progress. Your work on the bridges is particularly impressive.

 

Jerry

[Rabs]

Hi Jerry,

Thanks, I've been following your queen square layout with interest for a while now.  I love the main station building and I'll be very happy if I can achieve results half as convincing.

[Rabs]

A note for anyone how follows this route in the future:

The photo of the down goods area in the Middleton Press book "Swindon to Bristol via Bath" is not the same image at the one in "Steam around Bath" by Arlett and Peters.  I was concerned that I might be paying twice for the same shot.  As it happens the two books share none of the same images, although several are similar but from slightly different angles.  The Middleton press one has a clearer view of the turntable - so I should be able to model that correctly now.

 

A bit more progress on the bridge:

 

 

Sorry, no coin for scale - just my fingers.

 

[richbrummitt]

Fingers are about 20mm across so I've a pretty good idea. A little bit of unwanted surface texture is visible but not major - it should disappear reasonably well in the overall scene. 

 

I have several books on the particular line/area that I am interested in and although some of them share pictures the reproduction often shows a slightly different area, or the exposure is different such that different details become apparent because of how light or dark a scene is. 

[Rabs]

Bit more progress on the bridge.  All the parts are now printed except for the parapet walls, which I may well make from plasticard as they are pretty simple.

A few minor bits didn't print correctly, so I've added these in in styrene where needed rather than reprint whole sections of the bridge.

 

And no, they aren't bent - that's just my camera lens!

 

 

I'll paint them in this form and then do final assembly in situ on the layout.  

[Rabs]

A few bits to this update.  Firstly is the updated plan. I decided to add one more module to the layout and this allowed me to almost double the minimum visible radius to 670mm (26"), which I believe will be much better looking for mainline trains.  I've also laid out the rest of the buildings that I will model in the Dolemeads and the rest of the stretch East of the station up to the tunnel (that's the top row of modules in this image).

This change also gives be a bit more space to the left of the station.  I can make the skew bridge closer to true scale and it gives me more room to hide the transition off-scene, although I've still not decided how to do this at the West end.

The terraces which are all just drawn in lines so far look like this:

I suspect there might have been fewer satellite dishes in the 40s but I'm lucky that three roads of them still exist, so getting a good model of the prototype should be fairly straightforward.

 

Secondly I've made some more progress on the skew bridge, with the parts now sporting a couple of coats of primer.  Now that the shiny red colour is gone you can see the quality of the print much better.  Please excuse my appalling photography, I was really struggling to get the camera to focus on the nearside of the model and not the table.

I'm very pleased with how the rivet detail has come out.  I never would have tried to model these manually so not only is the 3d printer saving me bags of time, it is also achieving better results than I could have done otherwise.

 

Thirdly and finally is a quick mock up of a short section of the viaduct.  This was made by the following process:

- Screen capture the façade in Google Street view

- Correct the perspective in GIMP2 (or photoshop)

- Choose a suitably grubby brick texture for the inside of the arches from http://www.cgtextures.com/

- Scale both images to the correct size

- Print onto photo paper

- Make the structure in 2mm card

- Overlay the printed photos onto the card with PVA

As this is just a mockup I've not been very careful with the joins on the inside of the arches and the edges of the card have been coloured in black rather than a matching colour.  

 

I wanted to test whether I'm happy with printed card structures or whether I should be using textured plastic sheets for my building construction.  To this end I've been trying some of each.  In the card corner I've built one each of a Metcalfe and a Scalescenes terraced house and in plastic I've built a couple of Ratio huts. My conclusions/opinions were:

- In 2mm/N a card kit is closer to true scale results because you shouldn't be able to see brick relief.

- For complex brick patterns (e.g. the curving arrangements of bricks on the central pier of the mock up above) plasticard is a complete pain in the wotsit.  I could 3D print parts for this but I have over 30 viaduct arches and over 50 terraced houses to build and I'd like to be finished at sometime within my lifespan.

- The additional detail you can get in colour and patterns from using a real photo as a starting point more than outweigh the limitations of using flat source materials in my opinion.

- Printing can be done from photos of the actual prototype building (if it exists), such as I did with the arches above.  I don't need to 'represent' the brick patterns or weathering on the walls with some approximation - the model is inherently an exact replica of the real thing.

- No painting!

Edited August 13, 2013 by Rabs

[queensquare]

This really is making excellent progress, the bridgeworks look superb.

 

Your thoughts on printed card v plastic are also interesting. I am a big fan of plasticard and use it almost exclusively for my buildings although I have to say that your results are very impressive. The main problems I would foresee are that firstly my computer skills are simply not up to what you are doing and secondly, the photos you are working from show the colouring/weathering/details now and not as they were seventy years ago. That said, presumably that sort of thing can be dealt with in photo shop.

 

Keep the updates coming, its always interesting to see what's happening on the other side of town on 'the western'!

 

Jerry

[Rabs]

You are absolutely right.  In this case I just whacked on the arches as they appear now, covered in diesel soot and with lots of modern signs and security lights on.  When I come to do this for real I'll take a lot more time preparing the artwork to make it more plausible for the period and to 'undo' any more modern repairs or additions that I can identify.

[Joseph_Pestell]

Bit more progress on the bridge.  All the parts are now printed except for the parapet walls, which I may well make from plasticard as they are pretty simple.

A few minor bits didn't print correctly, so I've added these in in styrene where needed rather than reprint whole sections of the bridge.

 

And no, they aren't bent - that's just my camera lens!

 

DSCN6566.JPG

DSCN6565.JPG

 

I'll paint them in this form and then do final assembly in situ on the layout.  

 

These bridge parts are lovely and I am sure they would be useful to lots of modellers. Any plans to make them available?

[Joseph_Pestell]

A few bits to this update.  Firstly is the updated plan. I decided to add one more module to the layout and this allowed me to almost double the minimum visible radius to 670mm (26"), which I believe will be much better looking for mainline trains.  I've also laid out the rest of the buildings that I will model in the Dolemeads and the rest of the stretch East of the station up to the tunnel (that's the top row of modules in this image).

Lasercut box with Bath layout templot version v4.png

This change also gives be a bit more space to the left of the station.  I can make the skew bridge closer to true scale and it gives me more room to hide the transition off-scene, although I've still not decided how to do this at the West end.

The terraces which are all just drawn in lines so far look like this:

terraces.PNG

I suspect there might have been fewer satellite dishes in the 40s but I'm lucky that three roads of them still exist, so getting a good model of the prototype should be fairly straightforward.

 

Secondly I've made some more progress on the skew bridge, with the parts now sporting a couple of coats of primer.  Now that the shiny red colour is gone you can see the quality of the print much better.  Please excuse my appalling photography, I was really struggling to get the camera to focus on the nearside of the model and not the table.

2013-08-13 23.01.00.jpg

2013-08-13 23.01.23.jpg

2013-08-13 23.02.46.jpg

I'm very pleased with how the rivet detail has come out.  I never would have tried to model these manually so not only is the 3d printer saving me bags of time, it is also achieving better results than I could have done otherwise.

 

Thirdly and finally is a quick mock up of a short section of the viaduct.  This was made by the following process:

- Screen capture the façade in Google Street view

- Correct the perspective in GIMP2 (or photoshop)

- Choose a suitably grubby brick texture for the inside of the arches from http://www.cgtextures.com/

- Scale both images to the correct size

- Print onto photo paper

- Make the structure in 2mm card

- Overlay the printed photos onto the card with PVA

2013-08-13 22.57.08.jpg

2013-08-13 23.04.20.jpg

As this is just a mockup I've not been very careful with the joins on the inside of the arches and the edges of the card have been coloured in black rather than a matching colour.  

 

I wanted to test whether I'm happy with printed card structures or whether I should be using textured plastic sheets for my building construction.  To this end I've been trying some of each.  In the card corner I've built one each of a Metcalfe and a Scalescenes terraced house and in plastic I've built a couple of Ratio huts. My conclusions/opinions were:

- In 2mm/N a card kit is closer to true scale results because you shouldn't be able to see brick relief.

- For complex brick patterns (e.g. the curving arrangements of bricks on the central pier of the mock up above) plasticard is a complete pain in the wotsit.  I could 3D print parts for this but I have over 30 viaduct arches and over 50 terraced houses to build and I'd like to be finished at sometime within my lifespan.

- The additional detail you can get in colour and patterns from using a real photo as a starting point more than outweigh the limitations of using flat source materials in my opinion.

- Printing can be done from photos of the actual prototype building (if it exists), such as I did with the arches above.  I don't need to 'represent' the brick patterns or weathering on the walls with some approximation - the model is inherently an exact replica of the real thing.

- No painting!

 

I think you are right about this - at least so far as brickwork is concerned. All the textured plastic brickwork that I have seen has mortar courses which are bigger than the bricks.

[Rabs]

These bridge parts are lovely and I am sure they would be useful to lots of modellers. Any plans to make them available?

 

Thanks.  I'm very happy to make the 3D files available, but I'm not sure how much use they would be to people without their own SLA printer.  The parts are too big and have details too small for Shapeways FUD, and even if they would accept it the cost would be somewhere north of £200 for the bridge if ordered from them!  It wouldn't be cost effective for me to produce these to order, as each one of the four trusses ties up my printer for over a day so if wouldn't work as a commercial proposition.  Anyway, if you'd like a set of sketchup or stl files I'm happy to send them to you, just drop me a PM with your email address.

[Rabs]

Fair enough - so long as you're happy!

 

If you are modelling the viaduct by the Old Bridge, there was an arch there used by the Home Guard in the Second war, how do I know - my Grandfather was in the home guard!

 

Hi Tim,

My recent design changes mean that I now will have some of the arches between Skew Bridge and Old Bridge on the boards.  Do you know which arch they used? It would be a nice feature to add.

Thanks

[richbrummitt]

My conclusions about card v plastic for structures were similar to your own. With some more care taken on the 'production' version that bridge will be a nice model and if photographed well should look just like the real one.

[Rabs]

Yes, one of the challenges will be getting it in even lighting because there are lots of houses around it and shadows would cause no end of problems.  I need a nice cloudy day with diffuse light for it so I may make a trip sometime in the autumn.

[Tim V]

Hi Tim,

My recent design changes mean that I now will have some of the arches between Skew Bridge and Old Bridge on the boards.  Do you know which arch they used? It would be a nice feature to add.

Thanks

It was something my mother told me. Can't ask her any more. I believe it was the one immediately by the road bridges, seen in the middle in this picture.

http://www.bathintime.co.uk/image/199079/electric-tram-crossing-the-old-bridge-detail-c-1900

[Rabs]

It was something my mother told me. Can't ask her any more. I believe it was the one immediately by the road bridges, seen in the middle in this picture.

http://www.bathintime.co.uk/image/199079/electric-tram-crossing-the-old-bridge-detail-c-1900

 

Ah, that's a pity.  It's just outside of the area that I'm modelling.  Perhaps if I ever get to the point to extending the layout I'll get to add it then.  There are some interersting features just off both ends of the scenic area and my design makes it quite easy to extend later.  Anyway, I've bitten off quite enough for now.

 

Since the last update I've been playing around with bits of code 40 N gauge track from fiNetrax (http://www.rmweb.co.uk/community/index.php?/topic/65617-finetrax/page-2 and http://www.britishfinescale.com/product-p/finetrax-turnout-b6.htm)

This has recently been launched (just a few weeks ago).  At the moment only plain track and straight B6 turnouts are available, but others will be following soon I've been told.  The plain line uses sleeper sprues, just like 2mm easitrac.  The pointwork uses moulded chairs and a one-piece, cast frog with small pegs which are glued into holes in a milled track base.  The results are fantastic (see the thread above), and a huge step forward in N gauge.  This track, which I've been following on NGaugeForum since its inception is what stopped (saved?) me from going the whole 2FS hog.

 

However, none of my points are straight and every one is different.  So, to make use of the time saving features of this track I need to make some different bases.  My home-built cnc mill to the rescue:

 

I loaded my templot template for a LH B6, set on a curve to the right, that I need into CAD, and drew the 3d shape of the track base from it.  I then calculated the toolpaths with my CAM tool and sent this to the mill:

 

And here is the finished track base with the chairs fitted and some of the rails in place.  On this occasion I didn't get the depths quite right, so the webs between the sleepers are too thick.  On this point it doesn't matter because it's in a yard and the sleepers are completely buried.  I'll make sure that I get it right on the ones where it will show.

 

I will also need some different frogs for my other turnouts, as this one in the shunting yard is the only 1:6 frog that I will need.  The others are a mix of 1:7 to 1:10.  Also, some are semi-curved and some are fully curved.  The owner of fiNetrax has kindly offered to help me go through the process of casting custom frogs for these.

Edited September 2, 2013 by Rabs

[backofanenvelope]

Hello Rabs, just found your thread but have been following your progress on 3D group and really like the 3D visualisations and use of modern techs on your build. I have been armchair modelling all sorts of ideas that might make it into reality one day! Just a small Q. what type of laser are you using for the boards and whats the thickness of the ply?

 

Really looking forward to seeing your layout develop.

 

Regards

Tom

[queensquare]

Hi Rab, an interesting post. I'm surprised you are using those rather unsightly cast frogs, particularly as very little of your track is 'standard' geometry. Filing up and soldering crossings is really not difficult, particularly for a chap with your obvious ability - it is much better looking and allows unlimited variety. 

 

I think I shall stick to my soldered trackwork, making custom bases on a home made CNC mill is certainly not low tech!!

 

Jerry, who always champions the KISS principle, primarily because its the only one I understand:-)

[bmthtrains - David]

Can I have your workshop please?...drooling with envy...

 

David

[Rabs]

Hello Rabs, just found your thread but have been following your progress on 3D group and really like the 3D visualisations and use of modern techs on your build. I have been armchair modelling all sorts of ideas that might make it into reality one day! Just a small Q. what type of laser are you using for the boards and whats the thickness of the ply?

 

I've cut them on an HPC LS6090 (http://hpclaser.co.uk/index.php?main_page=product_info&cPath=1&products_id=11), which I have access to at a maker's club in Cambridge.  They are in good quality 6mm laser ply (http://hpclaser.co.uk/index.php?main_page=product_info&cPath=3&products_id=97)

 

Hi Rab, an interesting post. I'm surprised you are using those rather unsightly cast frogs, particularly as very little of your track is 'standard' geometry. Filing up and soldering crossings is really not difficult, particularly for a chap with your obvious ability - it is much better looking and allows unlimited variety. 

 

I haven't made a final decision yet, the milled bases and push-in chairs approach is amenable to soldered frogs as well.  I'm going to finish, paint and ballast this one to assess the final effect.  What aspects of the cast frogs are unsightly to you?  There is a slight colour mismatch at the moment, but that will be sorted by painting.  There are some gaps and misalignment at the ends in my pictures above but that's just because I've only rested the rails in place and not fettled the fit yet.  Otherwise I'm pretty happy with them: they have the correct bullhead rail profile to match code 40 rail and I think that the centre part between the rails will be invisible when it's painted.

 

 

I think I shall stick to my soldered trackwork, making custom bases on a home made CNC mill is certainly not low tech!!

 

 

You are right, it isn't.  I'm not evangelising any particular method here, just describing what I'm doing.  I went down this route because I have a cnc mill and it can make parts much more quickly and accurately than I can by hand.  I also want to support Wayne Kinney's fiNetrax because it's the first step forward in British trackwork in N since peco launched code 55.   It will be good for modellers in N in general if it succeeds as it deserves to.

 

Jerry, who always champions the KISS principle, primarily because its the only one I understand:-)

 

But what's simple depends on your experiences, no?  For me, using my mill is keeping it simple because I built it, I know all it's ins and outs and it took only about 45 mins preparation and 30 mins cutting (where you can be 90% doing something else and 10% supervising and hovering near the E-stop button) to make that base, starting from templot and ending with the part in my hand.  In future they will take me about 15 mins prep and 30 mins cutting now that I've got the process down.

For you, doing all manner of brass etch kits is simple because you've had lots of practice and get fantastic results.  For some others keeping it simple means taking stuff out of a box and putting it on the track.  'Simple' seems very subjective to me!

Edited September 2, 2013 by Rabs

[Rabs]

Can I have your workshop please?...drooling with envy...

 

David

 

 I am proud of it but it's not come about by throwing cash around, at least not nearly as much as you might expect for a cnc workshop!  I built the mill myself from scrap parts, I designed and built all the mechanics and electronics myself. Total cost was about £200.  The cnc converted myford ML10 lathe was a bit more (~£400), but similar things come up pretty regularly on ebay.  The Boxford TLC series and Denford Orac series are pretty good small cnc lathes.  They were originally designed for schools in the 80s and 90s but modern H+S, coupled with the decline in technically trained teaching staff, means that they almost never get used and hence can be picked up for a fraction of their original cost.  (Cost of a new Boxford 125TCL ~£10k, price on ebay for a 2nd hand one if you are prepared to be patient: ~£300)

 

Elsewhere I've got a small pillar drill and a grinding and polishing wheel that I got for a tenner from a closing down business.  I built my temperature controlled soldering oven from a £15 toaster oven and a £5 PID controller from China.  Other than those and the 3d printer in the house I've just got the usual set of hand tools.  In this I was lucky to inherit a lot of my grandfather's modelling tools.

Edited September 2, 2013 by Rabs

[queensquare]

  'Simple' seems very subjective to me!

 

I have to say I fully agree with much of your response, my post certainly wasn't meant as a criticism and I hope it wasn't taken that way. You are absolutely correct when you say that my fettling bits with comparatively basic tools and raw materials may not be seen by many as simple, the main reason being your skills with machine tools are way beyond my ability.

As for the cast crossings, they look clumpy to my eye although that may well be because they are simply rested in place so I will reserve judgement until they are painted and ballasted. That said I stand by the much greater flexibility of crossings assembled from individual, soldered components.

 

By the way, I love your workshop, particularly the way it has been put together.

 

Jerry, desperately trying to dig himself out of the whole dug by his earlier slightly clumsy post:-)

[Kris]

Very interesting to see the progress that you are making with the pointwork that you are producing.