Jump to content
 
  • entries
    122
  • comments
    870
  • views
    104,176

Diagram U29 Revisited


MikeOxon

1,041 views

I write two blogs, one on pre-Grouping GWR and one on Broad Gauge GWR but, while I am experimenting with 3D printing, I shall be hopping between the two so, if you wish to follow my various trials, it’s necessary to look at both.

 

From 2D to 3D modelling

 

It’s almost six years since my modelling aspirations took a great leap forward, through my acquisition of a ‘Silhouette Portrait’ cutter.  The first prototype that I chose to model with my cutter was the Tri-composite carriage, built in 1886 to diagram U29.  This carriage presented quite a challenge in preparing the 2D drawings!  In fact, I commented at the time that “It turned out that Diagram U29 is somewhat 'the design from hell', since every panel seems to be unique! It certainly gave me plenty of practice in using the drawing tools.”

 

I enjoyed making the Tri-composite model and a number of other models from a similar period but there were some aspects of my constructions that I never felt were really satisfactory, especially my fabricated clerestory roofs.

 

Now that I’m exploring the possibilities of 3D printing as a modelling tool, I thought it would be interesting to re-visit the panelled carriages from the late 19th century and see if I could now produce something better.

 

After creating a few relatively simple 3D models of carriages from earlier Broad Gauge times (described in my other blog), I decided that it was time to move on and tackle something a little more complex.  I felt that I had gained a reasonably firm toe-hold in Fusion 360 drawing techniques, so a panelled coach seemed a reasonable next step.  I also thought that 3D modelling offered the potential to create a clerestory roof rather more easily than the fabrication techniques I had used earlier.

 

Since I already had my 2D drawings of diagram U29 ( prepared in Autosketch and saved as DXF files) I could use these to create the various panels and mouldings by use of the push-pull tools in Fusion 360.  The first stage was to prepare the carriage body and I decided that, rather than add raised mouldings, I preferred to ‘emboss’ the panels into the outside walls.

 

My 3D Modelling Process

 

As described in an earlier post, I started from one end of the carriage, drew the end-wall profile and then extruded this profile to the length of the carriage.  I used my 2D drawing, pasted onto the end wall, to emboss the panels by 0.5mm from the outside framing.  From previous experience, this depth is around the minimum that will show clearly when using my printer.

 

One caveat is that all shapes to be extruded must be ‘closed’  I found that one of my panels had a tiny gap in one side that only became visible on very close inspection.  It had me stuck for some time, as I tried to determine why that particular panel refused to cooperate!  What’s good enough for the Silhouette cutter will not do for the more demanding task of 3D construction.

 

I also realised that there were several different depths to which different features of my model needed to be recessed.  These include the panels, the various window sections, and the window openings themselves.  In each case, the part of the drawing to be extruded must be placed on the appropriate plane of the model for the ‘push-pull’ tool to work.  My method was to make the shallowest panel sections first, then shift the ‘sketch plane’ to this deeper location, to select the next areas (e.g. window frames) for further depression, and so on.

 

To save having to work on the two sides separately, I completed all the detailing on one side then ‘split’ the model lengthways and created a mirror image for the other side, as described in a previous post.

 

Some of the key stages in constructing my model are shown in the following screen shots from the Fusion 360 software.  In fact, having struggled at times to find my way through the various modelling procedures, I went back through the steps, taking these screen-grabs as I went, and found it a lot easier the second time!


U29-1-4.jpg.02270da21b7331f7cd952c46d4f3932b.jpg

Stages in 3D Modelling diagram U29

 

The roof required separate consideration, since I did not want a solid model but had, nevertheless, to provide support for the roof panels.  I decided to split the model horizontally at eaves level and treat the carriage body and roof independently, as regards interior fittings, with the intention to join the parts together again for final assembly.

 

Once the roof had been removed,  I drew a rectangle on the top of the sold carriage body and extruded downwards, to open out the interior of the carriage, leaving a 1mm thickness floor at the base.  I then drew rectangles for the bases of the partitions and seats, and ‘pulled’ these upwards to their appropriate height within the carriage body. (all these procedures are described in more detail in a previous post)

 

For the roof, I left the flat base in place, to support the side sections and used the ‘push-pull’ tool to open out the interior of the clerestory section.  I used my existing 2D drawings to place the panels and window openings on both sides of the clerestory.  To ensure adequate support for the clerestory roof, I allowed the ‘Cura’ slicer software to generate supporting in-fill within the raised section, which I planned to remove after the model was printed.

 

Once I had drawn these two components in Fusion 360, I exported the data to the Cura slicer program, to prepare for printing.  Although the overall dimensions lay within the specified print volume (130x130x130mm) of my Geeetech E180 printer, the Cura software would not accept the model until I rotated it diagonally on the print bed.  The screen display of the two components, when prepared for printing are shown below:

 

U29-6.jpg.218b50e1d84f9626fa53640fd256ad38.jpg
Screen Views of Model Ready for Printing

 

Printing the Clerestory Roof

 

Because of its shorter printing time, I decided to print the clerestory roof first.  I was very pleased with the outcome and felt that the surface finish, after being printed on the diagonal, seemed smoother than when I had printed items aligned along the X-axis.

 

The space inside the clerestory was filled with a ‘grid-style’ supporting structure for the top surface and, at first, I was a little apprehensive about how easy this infill would be to remove.

 

U29-7.jpg.8815ba230120e690f3dd27c21406bc3d.jpg
Clerestory in-fill During Printing


Once printing had completed, the underside of the roof did look rather daunting but, in practice, the in-fill was not really difficult to remove.  The support structure is much weaker than the main structure and I found there was little possibility of damaging the main structure while attacking the in-fill.  

The first stage was to insert a knife into the slots around the edges of the infill.  Then it’s a case of chiselling away at the material until the cavity has been cleared.  After initial trepidation, I found I could proceed quite rapidly, scraping away all the debris!

 

U29-8.jpg.095c5124d9351c3786ca49bc7c160c65.jpg

Stages in Scraping Out the In-Fill


Printing the Carriage Body

 

While I was engaged in all this scraping and cleaning-up, my 3D Printer was busily getting on with the carriage body.  This is an advantage of 3D printing - it does not stop the modeller from getting on with other things!   The carriage body took over 5 hours to print, so I had plenty of time not only for modelling but also to take my cycling exercise.

 

Unlike the roof, the body needed very little by way of cleaning-up, apart from opening out all the window openings.  I was pleased that most of my design objectives were achieved in this ‘first go’.

 

U29-9.jpg.6b52cd6d9fc49d8a7fccd464458762a1.jpg
Printed Carriage Body after Clearing Window Apertures


Assembling the Components

 

There’s quite a lot of ‘fettling’ still to be done but I finish this post with a photo of the assembly after placing the roof over the body.

 

U29-10.jpg.37915198f01f9c8f918de79a0d9dc71b.jpg
3D Printed Tri-Composite Carriage to Diagram U29

 

The result is a much more robust body than I had made from Silhouette-cut plasticard.  With 3D printing, I don’t have the advantage of pre-printed body sides, which I could produce with the Silhouette cutter.  One thought is to investigate combining both techniques, with 3D printing for the main structure and Silhouette-printed laminates to provide elaborate pre-Grouping liveries.

 

More ideas to try :)

 

Mike

  • Like 2
  • Craftsmanship/clever 8

6 Comments


Recommended Comments

  • RMweb Gold

Very nice Mike, looks like you have solved the roof challenge convincingly. 

 

You seem to be moving fast with this. It's strange to think it's only a few years since the Silhouette was cutting edge (no pun intended).

 

Regarding the print quality, has your current printer reached its limit? I ask because you briefly mentioned the Anycubic Photon printer earlier.

  • Like 2
Link to comment

Thank you, Mikkel.  The clerestory was my main interest in this project and I'm pleased with how well it has come out  - as good as an old Tri-Ang roof, anyway :)  

 

My original idea was just to print small odds and ends, and I had never envisaged making complete carriages when I bought the printer.  It just seemed a natural 'training exercise', based on making a box.  All the hard work goes into the computer modelling. Printing only involves inserting the SD card and pressing the 'print' button - then wait for 5 hours or so.

 

My printer is limited to 130x130x130mm size, so this carriage is close to the length limit (for construction in one piece).  Surface finish is reasonable - I've done no surface fettling at all on the carriage I photographed, apart from scraping out the window apertures.

 

This E180 printer's big attraction for me was that it fits well into the domestic environment.   Good looking, very portable, and no unpleasant fumes or chemicals.

 

The Anycubic Photon print volume is even smaller (115 x 65 x 155) and involves the use of hazardous materials.  Also, the smell of the resin is not pleasant.  On the other hand, surface finish and precision are very high so it should produce finely-detailed small parts.  At present, though, the technology does not fit into my modelling environment.

 

I don't think my Silhouette cutter is redundant either.  The ability to print detailed panels and decals has an important role, which I hope to combine with 3D printing in future.

  • Like 2
  • Informative/Useful 1
Link to comment

Mike,

 

Outstanding work.  

 

I think the use of vinyl overlays would be the way to do the livery. - but you'd need multiple layers and have a lot of removing of panels/beading to do on each layer.  I think a compromise of using paint - probably sprayed on for the main body colour(s), with the vinyl used to overlay beading to give the colour and lining on raised mouldings.  That said, if you could match paint to vinyl colours you could do cream panels with the thin chocolate line around them as per Wenlock's recent blog posts on his 7mm bogie Brake Third.  

 

Keep up the excellent work.  

 

Duncan

  • Like 1
Link to comment

Thank you, Duncan.  I have plans to combine Silhouette techniques with 3D-printing in my next experiment, so please keep watching :)

Link to comment
  • RMweb Gold

Hi Mike, I think you're spot on there. 3D printing for a robust shell and then the silhouette etc for your livery application. I once did a test of printing transfers onto waterslide paper and then cutting them out with a Silhouette - it worked pretty well too.

 

A couple of quick ones, have you seen this on Thingiverse? It's a website with free 3D models on it, and 2 of them appear to be GWR coaches - probably not of uch use to you, but you may find them interesting https://www.thingiverse.com/thing:3153287 

 

The other thing is, if you have a machine with Windows 10 on it, there is an app on there called 3D builder (also available form Microsoft's store app). While it doesn't have a lot of functionality, it's stl repair feature is fantastic. There are no options (sometimes this is a good thing), and it's never failed me with regards to meshes that are not quite right.

  • Informative/Useful 2
Link to comment

Hi Jason, thanks for the info. and suggestions.  I'm pleased you agree with my planned way ahead.  I do think that my two machines will prove to be complementary.

  • Like 3
Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...