3d printed GWR coaches - something big...a D51

[sem34090]

Mr Great Bear!

 

I have been considering this for myself, as there are now several models that I wish to offer for sale but that i.materialise and shapeways put at too high a cost for my liking.

 

As best as I can tell, the only option is to take payment, then order the prints before boxing them up (perhaps with some of the other bits required for completion, such as handrails, bearings, bogies, etc) and shipping off to the customer. Not ideal, but perhaps has the advantage of being able to sell complete kits.

[The Great Bear]

Mr Great Bear!

 

I have been considering this for myself, as there are now several models that I wish to offer for sale but that i.materialise and shapeways put at too high a cost for my liking.

 

As best as I can tell, the only option is to take payment, then order the prints before boxing them up (perhaps with some of the other bits required for completion, such as handrails, bearings, bogies, etc) and shipping off to the customer. Not ideal, but perhaps has the advantage of being able to sell complete kits.

 

A complete kit like the ones I've built so body, interior, underframe, Stafford Road Model Works bogies, Gibson wheels, handrail wire, grab handles, glazing would I think work out at somewhere in the order of £110-120. Around 70% of that being the body, interior and underframe.

 

The Rolls-Royce of kits, a David Geen brass one, is I think £95 plus another £10 or so if you want a better roof. A well built Geen kit would be appreciably better than my efforts I'd imagine (and given the designers knowledge more accurate I suspect). Admittedly my design requires minimal assembly work and a lot less skill. Also the future availability of the Geen kits is uncertain. The better comparison is the Slaters ones, what the re-introduced ones will sell for I don't know, my guess around £60. 

[The Great Bear]

Latest print of a D45 brake third I've gotten done was by a UK company called Photocentric, a sample of what their LC HR2 printer can do - a machine costing a little under £2k so relatively "affordabe". This was done at 100 micron layer thickness. Not as clean as ones done by Ivan at 3d hubs and less resolution in details like the shell ventilators and door vents, but the small details on the sides, the hinges and door stops have come out, which for me is more important. There is some sign of stepping in the tumblehome but painting will hide this I think well enough. So overall I'm impressed and happy with the print - it's giving me food for thought as to getting a machine like this. That they are built in the UK and so support should be easier than something made on the opposite side of the world is a plus for me.

 

 

 

Unlike machines like the Formlabs Form 2, this printer does not use UV light but uses an LCD screen to expose a resin which is daylight sensitive. Hence both machine and resin are more affordable. This coach was printed by Photocentric using their "firm" resin, it's a bit too soft the bottom of the coach is too flexible, if printing again I'd try their "hard" resin. There are other machines using same technology like the Anycubic Photon but their build size is much smaller.

 

Painting of this coach is ongoing...

 

Thanks for looking

 

Jon

[Siberian Snooper]

It would possibly be better to omit the shell vents add white metal ones from MJT or similar. It looks very promising.

[MGR Hooper!]

Latest print of a D45 brake third I've gotten done was by a UK company called Photocentric, a sample of what their LC HR2 printer can do - a machine costing a little under £2k so relatively "affordabe". This was done at 100 micron layer thickness. Not as clean as ones done by Ivan at 3d hubs and less resolution in details like the shell ventilators and door vents, but the small details on the sides, the hinges and door stops have come out, which for me is more important. There is some sign of stepping in the tumblehome but painting will hide this I think well enough. So overall I'm impressed and happy with the print - it's giving me food for thought as to getting a machine like this. That they are built in the UK and so support should be easier than something made on the opposite side of the world is a plus for me.

 

 

 

Unlike machines like the Formlabs Form 2, this printer does not use UV light but uses an LCD screen to expose a resin which is daylight sensitive. Hence both machine and resin are more affordable. This coach was printed by Photocentric using their "firm" resin, it's a bit too soft the bottom of the coach is too flexible, if printing again I'd try their "hard" resin. There are other machines using same technology like the Anycubic Photon but their build size is much smaller.

 

Painting of this coach is ongoing...

 

Thanks for looking

 

Jon

This also looks fairly good!

[Knuckles]

Could you please detail what machine it is they use.

 

I'm to buy a resin one and looking at the Anycubic but as you say the build size is small so maybe larger is needed. Unsure.

[Guy Rixon]

Latest print of a D45 brake third I've gotten done was by a UK company called Photocentric, a sample of what their LC HR2 printer can do - a machine costing a little under £2k so relatively "affordabe". This was done at 100 micron layer thickness. Not as clean as ones done by Ivan at 3d hubs and less resolution in details like the shell ventilators and door vents, but the small details on the sides, the hinges and door stops have come out, which for me is more important. There is some sign of stepping in the tumblehome but painting will hide this I think well enough. So overall I'm impressed and happy with the print - it's giving me food for thought as to getting a machine like this. That they are built in the UK and so support should be easier than something made on the opposite side of the world is a plus for me.

 

 

 

Unlike machines like the Formlabs Form 2, this printer does not use UV light but uses an LCD screen to expose a resin which is daylight sensitive. Hence both machine and resin are more affordable. This coach was printed by Photocentric using their "firm" resin, it's a bit too soft the bottom of the coach is too flexible, if printing again I'd try their "hard" resin. There are other machines using same technology like the Anycubic Photon but their build size is much smaller.

 

Painting of this coach is ongoing...

 

Thanks for looking

 

Jon

 

This one is interesting. The flat surfaces are relatively smooth, but the resolution looks quite low on the mouldings and ventilator hoods. The effect is very much like the vacuum-formed sides that Roxey used to sell back in the 70s.

[Harlequin]

Could you please detail what machine it is they use.

 

I'm to buy a resin one and looking at the Anycubic but as you say the build size is small so maybe larger is needed. Unsure.

https://photocentricgroup.com/lchr2/

 

Build Volume: 196 x 147 x 250 mm

 

Layer Thickness: 50, 100 microns

 

XY pixel density: 97 microns

 

Price: £1599

[The Great Bear]

 

https://photocentricgroup.com/lchr2/

 

Build Volume: 196 x 147 x 250 mm

 

Layer Thickness: 50, 100 microns

 

XY pixel density: 97 microns

 

Price: £1599

 

Yes that's it. Note the price above is excluding VAT

[The Great Bear]

This one is interesting. The flat surfaces are relatively smooth, but the resolution looks quite low on the mouldings and ventilator hoods. The effect is very much like the vacuum-formed sides that Roxey used to sell back in the 70s.

 

As you can see from the machine specs Harlequin posted, the XY resolution is 97 microns and the layer height either 50 or 100 microns, the one I got done was at 100. Because of the print orientation, vertical I think, I don't think the smaller layer height would improve things, also then takes much longer to print, and as I believe the one shown took 36 hours that's probably unworkable.

 

I was having trouble getting decent focus on the model so that might make it look a tad worse than it is but clearly not as crisp as the one done by 3d hubs. But that was on a machine I guess costing at least 5x more.

 

Tweaking the design of the ventilator hoods with a bit more exaggeration might bring those details back.

 

The other interesting comparison would be how the finish compares with the Slaters injection moulded ones, but I haven't got one of those to do that. My feeling is it is similar?

[Joseph_Pestell]

36 hour print time is OK for a home hobbyist. But definitely not a commercial proposition.

 

Superb result though.

[Harlequin]

36 hour print time is OK for a home hobbyist. But definitely not a commercial proposition.

 

Superb result though.

I agree about the print time but this print was definitely not as good as previous ones from Ivan at 3dHubs.

 

It seems that we need a 3D printer specifically targeted at our hobby:

• Resolution: 25-50 microns in all axes (or better, if possible)
• Build volume suitable for 4mm scale rolling stock: 50*60*300mm
• Desktop format
• Easy materials handling
• Minimal part cleanup needed after printing (because that can damage details)
• Cost: say, £2000-£3000, assuming it was reliable and had a decent working life

That's not beyond the bounds of possibility because there are domestic resin printers that claim to have very fine Z axis control, but their XY resolution is lower simply to have a wider build volume for general use. But we don't need wide build volumes, we only need 50*60mm in XY and so the light projection could be focused into a small area for us giving greater resolution.

Edited July 2, 2018 by Harlequin

[Joseph_Pestell]

 

I agree about the print time but this print was definitely not as good as previous ones from Ivan at 3dHubs.

 

It seems that we need a 3D printer specifically targeted at our hobby:

• Resolution: 25-50 microns in all axes (or better, if possible)
• Build volume suitable for 4mm scale rolling stock: 50*60*300mm
• Desktop format
• Easy materials handling
• Minimal part cleanup needed after printing (because that can damage details)
• Cost: say, £2000-£3000, assuming it was reliable and had a decent working life

 

 

When you look at the modular construction of these cheaper 3-D printers, I guess that there is no real reason why one could not be produced quite easily if they are convinced that enough folk would buy them.

[The Great Bear]

I agree about the print time but this print was definitely not as good as previous ones from Ivan at 3dHubs.

 

It seems that we need a 3D printer specifically targeted at our hobby:

• Resolution: 25-50 microns in all axes (or better, if possible)
• Build volume suitable for 4mm scale rolling stock: 50*60*300mm
• Desktop format
• Easy materials handling
• Minimal part cleanup needed after printing (because that can damage details)
• Cost: say, £2000-£3000, assuming it was reliable and had a decent working life

That's not beyond the bounds of possibility because there are domestic resin printers that claim to have very fine Z axis control, but their XY resolution is lower simply to have a wider build volume for general use. But we don't need wide build volumes, we only need 50*60mm in XY and so the light projection could be focused into a small area for us giving greater resolution.

The Kudo 3d is nearest to the above but at or above upper bound of pricing. Suspect one has to add duty / vat on the price.

 

Combination of the specs of photocentrics LC precision 1.5 and HR2 would fit the bill.

 

A larger build volume in the xy plane would allow multiple bodies and parts to be printed at once, bear in mind each of my designs has three parts.

[Harlequin]

A larger build volume in the xy plane would allow multiple bodies and parts to be printed at once, bear in mind each of my designs has three parts.

 

Out of interest, what sort of minimum XY size would contain the three parts of your designs if they are efficiently arranged?

[The Great Bear]

Out of interest, what sort of minimum XY size would contain the three parts of your designs if they are efficiently arranged?

 

 

 

I don't know how much room around each part would be needed or is advisable but roughly if the body and interiror rounded up with a bit to spare were 40mm x 40mm x 250mm and the underframe 40mm x 20mm x 250mm that would give an 80mm x 60mm footprint but that may well be too tight?

 

Other issue is that to do a 70' coach that would need 300mm height or you angle it across the build envelope, which would have worked with the Photocentric HR2 one.

 

I suppose another approach would a smaller (higher resolution) printer and print in halves! With the daylight resin I think in theory you can join bits together before they are cured. But not convinced this would work accurately or reliably enough!

[Quarryscapes]

You need to factor in that a liquid resin print needs to be angled or it will just be a mess. You can see how the photocentric print is angled by the slipped layers leaving a visible diagonal gouge. This happens when the peel mechanism causes the soft resin to be skewed sideways a little before the next layer is added. (The model has to be peeled off the bottom of the tank after each layer is cured). The same mechanism also mean that smaller layer height does not equal better surface - thinner layers mean more peels per model, more chance of one slipping and causing a line. 

[The Great Bear]

You need to factor in that a liquid resin print needs to be angled or it will just be a mess. You can see how the photocentric print is angled by the slipped layers leaving a visible diagonal gouge. This happens when the peel mechanism causes the soft resin to be skewed sideways a little before the next layer is added. (The model has to be peeled off the bottom of the tank after each layer is cured). The same mechanism also mean that smaller layer height does not equal better surface - thinner layers mean more peels per model, more chance of one slipping and causing a line. 

 

Thanks, so if I'm interpreting the photos correctly based on the above the model was printed at around 30 degrees off vertical?

[Harlequin]

You need to factor in that a liquid resin print needs to be angled or it will just be a mess. 

I've been thinking about this statement and I'm afraid I don't understand it.

By definition whatever angle the print is above the base plate the last layer printed/cured will be exactly parallel to the base plate and will need to be peeled. So I don't think an angled print would help with peeling.

I can see that the Photocentrics print has been angled but I suspect that might have been more to fit it into the print volume, as TGB suggested, than for reasons of quality.

You can see that many other prints in this thread were exactly aligned to the print axes - even the very good ones.

So can you explain why liquid resin prints will be a mess if they are not angled?

Thanks,

[The Great Bear]

I've been thinking about this statement and I'm afraid I don't understand it.

By definition whatever angle the print is above the base plate the last layer printed/cured will be exactly parallel to the base plate and will need to be peeled. So I don't think an angled print would help with peeling.

I can see that the Photocentrics print has been angled but I suspect that might have been more to fit it into the print volume, as TGB suggested, than for reasons of quality.

You can see that many other prints in this thread were exactly aligned to the print axes - even the very good ones.

So can you explain why liquid resin prints will be a mess if they are not angled?

Thanks,

 

https://www.3dhubs.com/knowledge-base/how-design-parts-sla-3d-printing

 

This suggests you want to keep the z axis cross-sectional area to minimise peeling forces, which answers why to not print horizontally (if it would fit); is it support for the model that points against doing it vertically, in that the model has to support it's own weight whereas at an angle it can be shared with supports? A lot to learn!

 

I'm also wondering whether the hardness of the resin has an effect on the detail resolution and the rounding in the edges of the panelling, the stuff from 3d hubs was hard, whereas the one I got done was in a firm resin, so had some flex in it. The ventilator grooves were 0.016m deep full size so .21mm in the model, so not so fine as to be beyond printers' resolution; they are just about visible but a bit more would be better. 

Edited July 2, 2018 by The Great Bear

[Kelly]

That print looks rather good.

[Quarryscapes]

I've been thinking about this statement and I'm afraid I don't understand it.

By definition whatever angle the print is above the base plate the last layer printed/cured will be exactly parallel to the base plate and will need to be peeled. So I don't think an angled print would help with peeling.

I can see that the Photocentrics print has been angled but I suspect that might have been more to fit it into the print volume, as TGB suggested, than for reasons of quality.

You can see that many other prints in this thread were exactly aligned to the print axes - even the very good ones.

So can you explain why liquid resin prints will be a mess if they are not angled?

Thanks,

 

It's quite simple. Imagine a flat surface parallel to the tank base of the printer. It is attached to the print bed by supports, so layer by layer you have supports growing like stalactites. Not much material so not much surface area in contact with the base. Then you get to that one layer where the first bit of the flat surface gets printed, suddenly you have a massive surface area on the tank base attached to all the supports in the same layer, which when peeled is very likely to distort if not rip straight off the supports because they have a relatively small contact area.

 

If you angle it, then the section of flat surface  will be printed initially as a tiny surface no bigger than the first support that contacts it, so peels easily. The next layer only overhangs a tiny amount more and so on, so there is less stress on the peel. As this progresses, more supports meet the overhanging surface and because it is already supported by a series of previous supports it just keeps getting stronger.

 

Summary: As surface area increases so does the adhesive force holding the part to the tank base. If that force is greater than the bonds holding the part together, you get failures, and bits stuck to the bottom of the tank which then block the light from drawing more layers, so you get more failures and basically it's a disaster. 

 

It is possible in Some cases (large chunky parts, like solid 20mm lumps type chunky) to print straight on the bed horizontally without supports, however printers tend to press the first few layers onto the tank base to get better adherence to the base so if you do this you will probably find the dimensional accuracy to be off and a small bulge in the 'squished' layers. It's also harder to remove from the base. 

[The Great Bear]

My latest experiment, door T-handles in brass, again by Shapeways. They had to be rather oversized to get them to print so I am not sure how they'll look on a coach. (I know 247 Developments do turned brass handles so may try those too.)

 

 

Not the easiest things to get a clear photograph!

[Siberian Snooper]

Try contacting Slaters for some lost wax brass T handles. I did a few years back before the CooperCraft debacle. I also got the commode handles. 

[The Great Bear]

Here's the completed D45:

 

Once again the basic model rather let down by my finishing, in particular applying the lining transfers (HMRS Pressfix) - in particular on the paneled coaches where it runs along the ribs   Practise makes perfect I suppose but any tips gratefully received...

 

The commode grab handles in brass by Shapeways look good on the coach I think.