Marbelup

I print through I.Materialise and they don’t have a problem printing multiple copies of the same part on a sprue. A couple of years ago, they formalised their sprue requirement to ensure small parts don’t get lost, as that was their main concern previously. Since then, they haven’t objected to any of my sprued designs including, for example, 12 identical parts to make 3 pairs of bogies.

Surprising to get a size error of that magnitude. Where has the error crept in? I would assume the STL files are the correct size?

An old venetian blind slat could work for the roof, especially if you can get one from an older house with the wider slats. I have a stock of them which I have used in the past for 4 wheel vans and a small railmotor.

If designing a multi-part model in PLA or similar, you can “weld” parts together using a fine tipped soldering iron. I mostly use Colorfabb PLA/PHA blend for larger models as it is less prone to warping.

It might need additional supports? Did you take a photo prior to removing supports, or a screenshot of the automatically generated supports? Curious that you have included the coupler mount, presumably for Kadees or similar? Body mounted couplers generally perform better, e.g. with much less risk of derailment when pushing a train.

Don't expect to be able to print items in scale thickness, e.g. if representing metal prototype parts. The minimum printable thickness very much depends on the type of printer being used. For example, I work on 0.5 mm minimum thickness for small details and 0.8-1.5 mm for main body parts, as these suit I.Materialise SLA print capabilities. (Some small resin printers can print considerably finer, I know.) Strictly speaking, I.Materialise specify 1 mm wall thickness as absolute minimum and I have to take risk/responsibility for anything less. For something like a hopper or open wagon, I design the body to true scale dimensions outside then remove the inside to achieve the desired wall thickness. Thus the visible top of the sides is thicker than scale, but the same goes for commercial plastic models. Sometimes, I disguise thicker-than-scale parts, e.g. brake gear levers, by bevelling the rear edges so the visible edges appear thinner while the part is still reasonable robust.

I printed some test bogies in i.Materialise grey resin a few years ago. I found that if I made the sideframes thin enough to flex to insert the wheelsets, then they tended to spread over time with the weight of the wagon on the pinpoint bearings. So, I tried thickening up the sideframes so that they are quite rigid, and printing the bogie in 2 pieces, as per attached screenshot, which are held together with screws. This approach has been successful. I do use standard brass bearings press fitted into holes in the sideframes. I have also printed a 1-piece bogie in which the brass bearings are pressed in from below, to fit the wheelset. This works provided the axlebox/bearing is large enough to have at least 0.5 mm material either side of the slot which houses the brass bearing. I haven't tried including the conical bearing hole in the 3D print, so I don't know how it would wear over time. My bogies are in Sn3.5 scale but wagon size/weight is similar to other 16.5 mm gauge models.

You could try asking for an exemption given that you are bringing your business to IM rather than Shapeways, and that the design of the models is proven by previous printing. Just a thought. Also, is it your intention to print your locomotive in many parts to be assembled as a "kit"? I do the opposite, and try to minimise the number of parts, to 1 if possible. I understand that, especially with a steam loco, it might be desirable to print different parts in different orientation to minimise stepping on curved surfaces. IM do allow parts to be "sprued" now, with certain conditions, which I now use for small parts like bogies to minimise the impact of the €5 fixed charge per part.

I haven't used it myself, but I know other 3D modellers using Autodesk Fusion 360 which is free for hobbyists. Would that be another option?

Large curves should not get "planked" if you use decent CAD software and use appropriate settings when exporting STL file. I have done several cylindrical hopper wagons in HO and S scales and the curved sides come out quite smoothly apart from some stepping towards the top due to the printing process itself. The hopper in the attached photo is S scale and printed by i.Materialise in their standard resin. The sides have not been smoothed at all and 2 coats of paint have been applied.

Looking back at the photos from June 8, 2016, I am pretty sure the layering on the sides of the wagons is the residue from the wax support material used for Shapeways FUD. I haven't ordered from Shapeways since I first started out in 3D in 2012, as I wasn't impressed by the residue, although some people seem to get good results with FUD. I have had some success on wagons with similar overhangs by making the underside of the overhang slope upwards, which can elimate the need for supports of any kind while not being too obvious on the finished model. An example is the C-channel side sill in the screenshot. I usually print in i.Materialise Standard Resin and, prior to that, their Grey Resin.

Narrow Planet sell products from Tyneside Models so, if all else fails, they might pass on a message for you. Their web site is https://shop.narrowplanet.co.uk/ and they have a good old- fashioned email address.

I just sent a message to Tyneside Models via Shapeways to let them know Harold has been trying to contact them. I can understand Harold's issues as the message appears as a Forum "conversation" with 2 "members" but there is no confirmation of the recipient's identity. The messaging system is all a bit vague. I am not a regular Shapeways customer, but I do have a login. I do order fairly frequently from i.Materialise, because I prefer their Standard Resin material.

The message form appears when I tried it.

Sven, what software did you use to generate the supports form this model?

I am pretty sure the cleanup process for the Form 2 is pretty similar. A friend who has one for model railway use bought an extra washing tank and UV oven to speed up the finishing process. With regard to estimating the volume of resin required for a model, Netfabb does this based on the STL file.

Shapeways have a facility to send a message to the owner/designer of the model. Failing that, do a Google search for Tyneside Models and contact them direct.

I have found the Colorfabb PHA/PLA filament is much less prone to warping than either ABS or pure PLA. Also, if you use "blue tape" on your print bed, 3M #2090 tape provides strong adhesion, even compared to other 3M tapes.

Tip for future: You should be able to use just about any photo editing software to rotate your photo before uploading it to RMWEB. I find Irfanview (free, Windows only) very good for rotating, cropping or resizing photos.

As far as I know, the commercial printers offering PLA use an extrusion type printer, so the result is not that much different to home printers. Having used a home printer (Makerbot) for a few years now, there are a few serious issue. Where supports are needed, the supports, if done in PLA, are too strong and very difficult to remove without damaging delicate parts of the model. Dual head printers can do supports in other materials, such as PVA which is water soluble, but I found there was poor adhesion between the different materials. I also found that larger models in PLA (say, > 150 mm) are prone to warping and pulling away from the build platform. I have had better result, in terms of warping, with PLA/PHA blend filament, but it can still be a problem. I have printed some functional parts, including a rotary car dumper, in PLA/PHA and have not had any problems with degradation over several years. I mainly use the Makerbot for prototypes to test the fit of parts, where the lack of detail (compared to SLA prints) is not an issue. Where possible, I try to arrange prints so supports are not needed, even if that means joining several parts to complete the model, which can be done quite easily by "welding" with a soldering iron. In some cases, I have designed my own support structure 'which can be easier to remove than the automatically generated supports. I have printed some larger test models, such as large diesels in S scale (around 300 mm long) in 3 sections which are "welded" together. These prints were useful to test the fit of the bogies and drive mechanism, but the finish and detail were not suitable as a finished model.

As far as I know, the commercial printers offering PLA use an extrusion type printer, so the result is not that much different to home printers. Having used a home printer (Makerbot) for a few years now, there are a few serious issue. Where supports are needed, the supports, if done in PLA, are too strong and very difficult to remove without damaging delicate parts of the model. Dual head printers can do supports in other materials, such as PVA which is water soluble, but I found there was poor adhesion between the different materials. I also found that larger models in PLA (say, > 150 mm) are prone to warping and pulling away from the build platform. I have had better result, in terms of warping, with PLA/PHA blend filament, but it can still be a problem. I have printed some functional parts, including a rotary car dumper, in PLA/PHA and have not had any problems with degradation over several years. I mainly use the Makerbot for prototypes to test the fit of parts, where the lack of detail (compared to SLA prints) is not an issue. Where possible, I try to arrange prints so supports are not needed, even if that means joining several parts to complete the model, which can be done quite easily by "welding" with a soldering iron. In some cases, I have designed my own support structure 'which can be easier to remove than the automatically generated supports. I have printed some larger test models, such as large diesels in S scale (around 300 mm long) in 3 sections which are "welded" together. These prints were useful to test the fit of the bogies and drive mechanism, but the finish and detail were not suitable as a finished model.

Andy - Have you posted photos of your 3D printed turnouts and track? I would be interested to see what you have done. I have printed some lengths of sleeper bas to suit flat-bottomed rail for Sn3.5 (16.5 mm), S standard gauge (22.5 mm) and dual gauge. I haven't attempted turnouts yet, although I have made many turnouts previously on timber sleepers using Pliobond glue to attached the rails.

In recent weeks, I have been having major problems with Standard Resin prints. Models with large horizontal surfaces (e.g. 20 x 20 mm) often exhibit a serious bulging effect, such that the printed surface is 0.8 to 1 mm above the desired surface level. i.Materialse are saying this is an inherent characteristic of Standard Resin, despite the fact that some prints are still working well, and the problem has only surfaced in the last month or so. I had a couple of prints in the middle of last year with a slight bulging effect, but it has certainly got much worse recently. I have suggested the problem must be linked to a particular printer or printer(s) and provided examples of some prints from the same order which printed prefectly while others exhibited the bulging effect. However, i.Materalise don't seem to be able to pinpoint the problem, and don't seem interested in tracking it down. The problem is so bad that I am no longer accepting orders from my Marbelup Models customers, as I can't guarantee getting quality prints from i.Materialise, effectly shutting down my "hobby business" after 5 years of operation. I have had some parts printed in Gray Resin, at additional expense, to complete existing customer orders after the Standard Resin prints were unusable. i.Materialise gave me a credit for some unusable prints, but are now refusing to do so claiming that a "slightly different output" is normal, despite the areas of bulging greatly exceeding their specificed dimensional accuracy.

For what it is worth, I have discovered that some blue tapes are better than others. After using 3M 2093 tape for some time, the next roll I bought was 3M 2090 which has significantly better adhesion. I think the difference is due to the slightly textured surface. I use a Makerbot dual head printer, with original aluminium platform. I mainly use ColorFabb PLA/PHA blend filament, which I found has minimal warping for large prints, and sometimes pure PLA.