GWR N6 Loco Coal Wagon in 7mm, a 3D beginner's first attempt

[GWR57xx]

One of the main drivers for me getting in to 3D printing was to be able to produce wagons that otherwise wouldn't be available either as RTR or as plastic kits (I can't solder for toffee so brass kits are a non-starter for me).
As I am (very slowly) building a layout based on a GWR running shed complete with coal stage, I will need several loco coal wagons for the ramp.
The only suitable kit I am aware of is the Slaters N13, of which I will have a couple.
I'm not aware of any RTR, although Minerva are producing the N32 (not a loco coal wagon and anyway too late for my period).
So far my journey into 3D printing has been limited to small bits and pieces for the coal stage that haven't challenged my lack of 3D design skills too much.
Until getting the printer I have never done any 3D design, so I've been on a rather huge learning curve getting to grips with DesignSpark.
I think I am now getting a bit more fluent with it, but sometimes I have been staggered by the results of some commands which bear no relation to what I was trying to achieve!
I like the distinctive look of the N6, with its rounded corners and 9" extension round the top, so decided to bite the bullet and get on with trying to draw up the easy bit - the body.

After rather a lot of hours work I was quite pleased with the result, so tried to print it. I tried doing it flat to the build plate to avoid having to clean off a forest of supports, and also to make the print far quicker. Initially the result looked promising while it was hanging upside down draining off, but after cleaning up it was clear that the suction cup effect had blown holes around the perimeter of the base.

 

I tried again having slowed down the lift and retract speeds to try to reduce the build up of pressure and give the resin time to escape the cup, but to no avail.
These first attempts were with Elegoo Standard Grey, so I thought maybe a different resin would succeed. Unfortunately Elegoo's ABS-like grey behaved almost exactly the same.
So I tried again, this time having removed the 9" side extensions to see if the reduced height would work, but the base still blew out.

I even "blew the bl**dy doors off" thinking that this would surely alleviate the pressure build up but was surprised that even this didn't work.

So I decided I'd have to split the model up into several pieces and make it a kit. I split it into three parts: the floor and two half side pieces, and added alignment tabs and slots to help with assembly later.

Trying to print both halves together on the plate at the same time failed yet again. I guess the pressure of the resin flooding past the end caused that to also fail.

 

Printing one half side by itself was a success, so that's how I'll have to do it I guess.
I now have quite a collection of failed N6 bodies, but they will come in useful for storing small parts away from the carpet monster.

 

 

 

The two halves loosely slotted together:

 

 

 

[GWR57xx]

Having found a way forward with the body, the difficult bit (chassis) was next.

I'm working from a postage stamp sized dimensioned diagram and a few fuzzy photos, so I'm only aiming for a rough approximation rather than an exact replica, as with the body.
This is currently on the printer as a test. It still needs some work, even if it prints ok.

[Miss Prism]

I feel those boxes are set a bit too low.

[Miss Prism]

[brossard]

I'm always impressed by someone who's prepared to try new things and persevere when they don't don't go as desired.

 

Excellent stuff.

 

John

[Miss Prism]

It's got me wondering - did the GWR draw springs uncompressed or under a typical load?

 

[GWR57xx]

2 hours ago, Miss Prism said:

I feel those boxes are set a bit too low.

 

Does it look any better with a bit of context?

This shows it with a 3'1" wheel.

The axle is aligned exactly with the centre of the circular part of the front face.

 

PS: I meant the centre of the circle of which only a semicircle is visible, not the centre of the face, but I'm not sure that's made it any clearer!

The axle is centred on the horizontal line that cuts across the bottom of the semicircle?

I know what I mean!!!

 

Edited August 10, 2023 by GWR57xx
clarification

[GWR57xx]

I know I'm missing a nipple (oo-er missus) above the "W" - on the to-do list.

I just wanted to see how it prints so far.

 

Hot off the printer:

 

 

First impressions, it looks pretty good so far.

 

After washing:

 

 

After my problems with the body I am very pleasantly surprised and encouraged.

 

The box on the left of the photo has a lot more rear support than the one on the right - just experimenting. The springs are extended at the back up to the front of the W-iron, and the axlebox is extended at the back up to the rear face of the W-iron.

 

The one on the right has a gap between the springs and W-iron and the axlebox had space behind it.

 

Both seem to have come out as designed, and at "normal viewing distance" I don't think the differences would really be noticed?

 

[Miss Prism]

17 minutes ago, GWR57xx said:

Does it look any better with a bit of context?

 

Yes.

[GWR57xx]

17 minutes ago, Miss Prism said:

It's got me wondering - did the GWR draw springs uncompressed or under a typical load?

 

 

Maybe I need two versions - one empty, one loaded? The coal stage ramp would have some of each...

 

But then, apart from me and you, who is ever going to notice? 

 

Besides, there will be so many other "approximations" and compromises on the layout that this would probably fade into insignificance.

 

But if it's wrong it's wrong, and I may as well try to get it right at the outset ...

 

[Miss Prism]

I have always held a heretical view on buffer height, and it's all down to the different interaction between model buffers and prototype buffers (the latter being far more tolerant of height difference when it comes to buffing forces), so I aim for a nominal 3' 5 1/2 " (13.8mm, plus or minus a constructional smidge, in 4mm scale) regardless of whether an item of stock is loaded or unloaded. (GWR buffer heights rose gradually from 3'4 1/2" to 3'5 1/2" over the years.)

 

 

[GWR57xx]

4 minutes ago, Miss Prism said:

I have always held a heretical view on buffer height, and it's all down to the different interaction between model buffers and prototype buffers (the latter being far more tolerant of height difference when it comes to buffing forces), so I aim for a nominal 3' 5 1/2 " (13.8mm, plus or minus a constructional smidge, in 4mm scale) regardless of whether an item of stock is loaded or unloaded. (GWR buffer heights rose gradually from 3'4 1/2" to 3'5 1/2" over the years.)

 

 

 

Thanks - I'm aiming for a buffer height of about 24.2mm (7mm scale), which isn't as easy to get right as I naively assumed when I began this journey.

 

In the interests of science (and getting it right) I had a go at superimposing your very helpful diagram (wish I'd started from there) on my design:

 

I'm surprised it's as close as it is to be honest, given how I arrived at it (measuring things off photos on my PC screen mostly!).

It is a tad low, as you say ...

 

 

[stevel]

4 hours ago, GWR57xx said:

One of the main drivers for me getting in to 3D printing was to be able to produce wagons that otherwise wouldn't be available either as RTR or as plastic kits (I can't solder for toffee so brass kits are a non-starter for me).
As I am (very slowly) building a layout based on a GWR running shed complete with coal stage, I will need several loco coal wagons for the ramp.
The only suitable kit I am aware of is the Slaters N13, of which I will have a couple.
I'm not aware of any RTR, although Minerva are producing the N32 (not a loco coal wagon and anyway too late for my period).
So far my journey into 3D printing has been limited to small bits and pieces for the coal stage that haven't challenged my lack of 3D design skills too much.
Until getting the printer I have never done any 3D design, so I've been on a rather huge learning curve getting to grips with DesignSpark.
I think I am now getting a bit more fluent with it, but sometimes I have been staggered by the results of some commands which bear no relation to what I was trying to achieve!
I like the distinctive look of the N6, with its rounded corners and 9" extension round the top, so decided to bite the bullet and get on with trying to draw up the easy bit - the body.

After rather a lot of hours work I was quite pleased with the result, so tried to print it. I tried doing it flat to the build plate to avoid having to clean off a forest of supports, and also to make the print far quicker. Initially the result looked promising while it was hanging upside down draining off, but after cleaning up it was clear that the suction cup effect had blown holes around the perimeter of the base.

 

I tried again having slowed down the lift and retract speeds to try to reduce the build up of pressure and give the resin time to escape the cup, but to no avail.
These first attempts were with Elegoo Standard Grey, so I thought maybe a different resin would succeed. Unfortunately Elegoo's ABS-like grey behaved almost exactly the same.
So I tried again, this time having removed the 9" side extensions to see if the reduced height would work, but the base still blew out.

I even "blew the bl**dy doors off" thinking that this would surely alleviate the pressure build up but was surprised that even this didn't work.

So I decided I'd have to split the model up into several pieces and make it a kit. I split it into three parts: the floor and two half side pieces, and added alignment tabs and slots to help with assembly later.

Trying to print both halves together on the plate at the same time failed yet again. I guess the pressure of the resin flooding past the end caused that to also fail.

 

Printing one half side by itself was a success, so that's how I'll have to do it I guess.
I now have quite a collection of failed N6 bodies, but they will come in useful for storing small parts away from the carpet monster.

 

 

 

The two halves loosely slotted together:

 

 A very nice looking design and prints, I have not had any success printing flat without supports, and would suggest using sacrificial supports, and printing upside down, and with the floor as a separate part.

Here is a rendering of an N3 coal wagon in 4mm showing the support system. After snapping of the support there is minimal sanding to clean it up, and printing at 8-10 degrees from horizontal will help eliminate suction issues. This method has also worked with some 7mm wagons I am currently working on.

 

 

 

[GWR57xx]

12 minutes ago, stevel said:

A very nice looking design and prints, I have not had any success printing flat without supports, and would suggest using sacrificial supports, and printing upside down, and with the floor as a separate part.

Here is a rendering of an N3 coal wagon in 4mm showing the support system. After snapping of the support there is minimal sanding to clean it up, and printing at 8-10 degrees from horizontal will help eliminate suction issues. This method has also worked with some 7mm wagons I am currently working on.

 

 

Thanks - so is that a single print?

 

I have a Mars 3 so a 7mm wagon just barely fits when flat on the build plate. I think trying to angle it would only work with a shallow depth of model, otherwise it will overflow the useable area?

 

If I can get my "kit" of parts to print successfully separately then I'll call that a win!

 

[stevel]

9 hours ago, GWR57xx said:

 

Thanks - so is that a single print?

 

I have a Mars 3 so a 7mm wagon just barely fits when flat on the build plate. I think trying to angle it would only work with a shallow depth of model, otherwise it will overflow the useable area?

 

If I can get my "kit" of parts to print successfully separately then I'll call that a win!

 

For the 4mm it is a single print, just drop in the pin point bearings and it ready to go ish. I have been able to print 7mm wagon body on the Mono SE with a shallow angle but can print a complete kit, about 12 parts, on the Mono X in one go. I think you will find your kit of parts, will benefit from some sacrificial supports, as from experience I use them on nearly everything I print, it makes clean up so easy, and I get less distortion issues.

[GWR57xx]

I've now made a few improvements to the suspension unit thanks to @Miss Prism's helpful diagram.

 

The changes include:

• shortening and widening the W-iron
• moving the spring hangers in slightly
• re-profiling the springs
• shortening the axlebox body slightly
• adding the missing oil filler pot

The superimposed image now looks like this:

 

All in all, very worthwhile. Thanks for alerting me to the issues Miss P.

 

Now that I've got a bit more competent in using DesignSpark I'm really starting to enjoy it.

 

[GWR57xx]

I need a brake ...

 

Simple lever brakes, single side only for the N6, ta very much.

 

Fortunately Miss P's earlier message reminded me (not that I should have needed reminding) of the fantastic resource that is gwr.org.uk, which provided ample information to (hopefully) prevent me getting it wrong while making it easy to draw up:

 

 

I didn't include the flange-way arm as it's on the back so would never be seen.

 

Edited August 13, 2023 by GWR57xx
confusion

[dpgibbons]

From what I've seen you'd get the best results on the wagon body by angling the print.

 

FYI Scorpio do an N6, N20 and N21 and DK Parkin may still have ABS versions of the latter two.

Edited August 12, 2023 by dpgibbons

[GWR57xx]

7 minutes ago, dpgibbons said:

From what I've seen you'd get the best results on the wagon body by angling the print.

I can’t envisage how to do that successfully, as there is always going to be a large suction cup in at least one corner?

The way I’m doing it now is a bit long-winded but does seem to produce good results and also avoids having to remove any supports and the pimples they leave behind.

[dpgibbons]

To quote FormLabs:

 

"Orienting large, flat surfaces to an incline of 10–20º drastically increases the success rate. By printing a model at an angle, you are reducing the surface area of each layer while also decreasing the amount of contact the print has with the tank. Reducing the surface area means that the print is subject to less force as the build platform raises with every layer"

[beejack]

On 10/08/2023 at 13:12, GWR57xx said:

I tried again having slowed down the lift and retract speeds to try to reduce the build up of pressure and give the resin time to escape the cup, but to no avail.
These first attempts were with Elegoo Standard Grey, so I thought maybe a different resin would succeed. Unfortunately Elegoo's ABS-like grey behaved almost exactly the same.
So I tried again, this time having removed the 9" side extensions to see if the reduced height would work, but the base still blew out.

 

Try increasing the exposure times of the initial layers and the subsequent layers.  I use 60 for the intial and 8 for the rest  and have found I have few failures due to layer seperation.

[GWR57xx]

2 minutes ago, beejack said:

 

Try increasing the exposure times of the initial layers and the subsequent layers.  I use 60 for the intial and 8 for the rest  and have found I have few failures due to layer seperation.

What printer and resin are you using?

Those settings seem high - Elegoo recommend 3s for their ABS-like grey v2 resin on a Mars 3.

When I changed over to the ABS-like I ran all the various exposure calibration prints I could find with different exposure times to find the optimum for the kind of printing I’m doing (i.e. not the miniature figures that most of the tests seem to be aimed at). I got up to 5s normal exposure before details started to blur, so I’ve stuck with that.

I’ll try at test at 8s just to see what happens though, thanks for the suggestion.

I hope I’m not tempting fate saying this, but I’ve had no trouble with base plate adhesion using 10s base layer exposures, and don’t get any “elephants foot” either (I use UVTools to optimise base layers).

[GWR57xx]

3 hours ago, GWR57xx said:

I’ll try at test at 8s just to see what happens though, thanks for the suggestion.

 

I printed the calibration matrices (XP2 & Phrozen's XP Finder) at 8s normal exposure.

The prints have been washed but not yet cured (I'm letting them dry), so it remains to be seen if these prints are less resilient than those with lower exposure times.

They don't look as bad as I expected, but there is obvious "bloat" - i.e. small holes and gaps are filled in, raised objects are wider than designed.

I'd previously stopped at 5s on the calibration tests after running 2.5s, 3.0s, 3.5s, 4.0s, 4.5s & 5.0s as this seemed quite a jump up from Elegoo's recommended 3.0s.

I'm now curious as to how much higher than 5s I can go, so I might experiment further.

 

[GWR57xx]

I've done a bit more experimentation.

I looked again at the 8s results from the flat exposure tests and decided it was worth trying the more demanding vertically oriented tests to see how they turn out.

I'd previously settled for 5s as the results seemed pretty good.

Here's a comparison of the 5s and 8s test results:

 

First of all I have to say how impressed I am that this little entry-level printer can achieve anything like this level of detail at all.

 

Here's the Phrozen test piece:

 

A) the holes on the 5s piece are quite a bit more open than those on the 8s.

B) not easy to tell from the photo but the very finest hair-like columns on the 8s piece are much firmer. More significant is the row of hex columns on the far right, where the 0.3mm to 0.6mm didn't print at all on the 5s but are perfect on the 8s.

C) the patterns on the 5s piece are more defined.

D) the hole in the "nut" is slightly narrower on the 8s. The 5s hex measures 3.06mm across the flats compared to 3.09mm for the 8s, so it's grown a bit.

 

Here's the Starship test:

 

A) the biggest difference is the strength of the finest hair-like columns on the wings.

B) the trenches on the 8s are very slightly less well defined.

c) the gaps between the walls are equally slightly less wide.

 

All in all though, both exposure settings have produced good results, it's just a trade-off between different features.

 

[GWR57xx]

Exposure tests are all well and good, but they don't really tell me how well my wagon will print, so I ran a test print of the solebar at 8s:

 

This was printed flat to the build plate with no supports, similar to the photo above.

I am super happy that the brake handle and ratchet came out so well - I was really expecting them to fail.

This means I won't have to make them as separate parts and glue them on.

 

They weren't quite perfect though:

 

Totally unsurprisingly they both have a bit of delamination at the ends.

I have now added supports at these points, so hopefully they'll be perfect next time.