And who says that home 3d printing can't do detail?

[-missy-]

Hello.

 

But I don't think it can easily do complex 3D curves and surfaces that curve in every direction.

 

I can vouch for that. The curves on that Morris van Rabs has just printed were an absolute pain to model. It was one of the most complex thing I have had to model to date. One the plus side at least I know how to do it next time!

 

Missy

[Kelly]

The question is do you polish your Autocad or learn something new.  Student editions of autocad 2014 with a 3 year license appear on ebay for £10 so its not too expensive.  But I don't think it can easily do complex 3D curves and surfaces that curve in every direction.

 

 

For me as it's been so long I'd be better learning something new and preferably cross-platform as  use a mixture of Windows/Mac/Linux across laptops/desktop computers.

 

The other thing to remember is the student version licence has limitations in what you're allowed to do with it. So creating a few items to learn from would be fine. But then later to use it to produce for others/sell might get you in hot water with AutoDesk's lawyers perhaps as the student editions apparently store data in the models/files that allow it to be identified by AutoDesk as which version was used.

[Rabs]

 

Having seen Missy's completed Foden today at the Didcot show I was mightily impressed both with the construction but also the detail & finish of the printing

 

Some impressive work in his thread. Great to see.

 

Thank you both.

 

The question is do you polish your Autocad or learn something new.  Student editions of autocad 2014 with a 3 year license appear on ebay for £10 so its not too expensive.  But I don't think it can easily do complex 3D curves and surfaces that curve in every direction.

 

True, but recreating a precise compound curve is tricky in just about every package

 

Hi Dave PM me your address and I'll send you some FUD. About time I joined in on this thread.

As it happens I've already got a few  bits, but thanks anyway.  (BTW, I'm not Dave but I assume this was in response to my post earlier!)

[Tricky-CRS]

 

 

uote>

 

 

Sorry Rabs auto correct I think, I certainly didn't intend saying that.

 

Glad you got some anyway.

[Rabs]

First an update on the Beetle and Mink prints for Richard, who has been waiting very paitnetly for months for me to work out a way of printing them effectively.  I've spent a lot of time playing around with different ways of joining the beetle and mink wagons printed in two halves.  One of the (many) problems is that it is difficult to get a good mating surface on the joining faces (for exactly the same reason that it's hard to get a good finish on the downward facing side if it's printed in one part).  I've tried all sorts of alignment features and now have a method which is moderately successful but still doesn't look as good as the best of the all-in-one prints for the beetle (which was actually pretty good). 

So, not much progress to report but I've not forgotten about it and still intend to send you some prints as soon as the results are good enough.

 

In other news I have had some success with some other areas of printing.  Firstly is this little test wagon which is a (very incomplete) 4 plank O21 design from my GWR wagon generator (http://www.rmweb.co.uk/community/index.php?/topic/77285-gwr-wagon-cad-generator/). 

 

As you can see, it didn't print properly  because I missed some supports so one corner got messed up and the brakegear didn't quite come out.  However, it's a nice little proof of concept for an all-in-on print with integral pinpoint bearings can work in this material.  The wheels run nice and freely ( as well as Peco N gauge wagons, at least) - despite me miscalculating the length of the axle in the design.  All the features are scale thickness or very close to it and the detail blows the socks off the peco wagon I put it next to.  Not sure how this compares to 2FS wagon standards, but I suspect that it is similar.  For practical, robust models I may need to beef some bits up a little bit - at least until someone produces a stronger resin for me to print with.

 

Finally, and a bit that I'm particularly excited about:

These are a test pair of driving wheels for a GWR 72xx and the pictures really don't do them justice because there's a lot of detail which doesn't show up properly in this yellow until it has paint on it (not to mention a better camera).

I'm experimenting with different methods of making these.  One approach is for me to turn the tyres (I can knock these out very quickly on my CNC lathe in brass, steel or nickel silver) and print the centres, as I have done here.  This is quick, easy and gives a very good quality result.  The disadvantage is that pickups are made a bit more difficult if the centre isn't metal.  I may be able to get around this problem by painting the plastic part with a conductive paint or ink before putting a final paint coat over that.

 

I've also tried printing complete wheels, including the tyre with and without stub axles.  I'm working with a couple of RMWebbers to look into getting these cast in nickel silver

[corax67]

This is such a brilliantly insightful thread - it is fantastic to see just what can be achieved in this burgeoning field - keep up the good work and the pics

[richbrummitt]

These are a test pair of driving wheels for a GWR 72xx and the pictures really don't do them justice because there's a lot of detail which doesn't show up properly in this yellow until it has paint on it (not to mention a better camera).

I'm experimenting with different methods of making these.  One approach is for me to turn the tyres (I can knock these out very quickly on my CNC lathe in brass, steel or nickel silver) and print the centres, as I have done here.  This is quick, easy and gives a very good quality result.  The disadvantage is that pickups are made a bit more difficult if the centre isn't metal.  I may be able to get around this problem by painting the plastic part with a conductive paint or ink before putting a final paint coat over that.

 

I've also tried printing complete wheels, including the tyre with and without stub axles.  I'm working with a couple of RMWebbers to look into getting these cast in nickel silver

 

I think that traction will be better if the tyre is steel.

 

For pick ups there is no problem if you use the flange: a method that has been proven on at least one widely exhibited 2mm layout.

[Marbelup]

Hello Rabs

 

Would it work for your covered vehicles (Mink and Beetle - I had to look them up to see what they looked like!) to print them in one piece but without a solid floor?

 

I did this recently for a bogie covered van (in S scale) and I included a skeleton underframe but no floor as such.  I had the van printed on end (in i.Materialise Prime Gray), to get a good finish on the curved roof, but this meant that one end had support attachments which had to be cleaned up.  Your printer would probably be Ok for the curved roof as the resolution seems very good, so you could print it right way up.

 

The lack of floor is not obvious when on the track, although I have allowed for 1 mm styrene floor panels to be added or, optionally, some done in sheet lead to add some weight.  Photo of finished van below, and attached screenshot shows view from underneath: 

 

 

[Rabs]

I think that traction will be better if the tyre is steel.

 

For pick ups there is no problem if you use the flange: a method that has been proven on at least one widely exhibited 2mm layout.

 

Interesting point - I can't find any data on CoF for steel on nickel silver and nickel silver on nickel silver.  I guess the trade off is more grip (steel) vs. more reliable conduction (NS).  I think that, because my layout has no gradients and prototypical accelerations I might be better off with NS.  I'll do a test and try both.

 

What is the flange method of pickup?  I assume that it's similar to RTR pickups but if you could point me at a link to see how others have done it I'd be grateful.

 

Hello Rabs

 

Would it work for your covered vehicles (Mink and Beetle - I had to look them up to see what they looked like!) to print them in one piece but without a solid floor?

 

I did this recently for a bogie covered van (in S scale) and I included a skeleton underframe but no floor as such.  I had the van printed on end (in i.Materialise Prime Gray), to get a good finish on the curved roof, but this meant that one end had support attachments which had to be cleaned up.  Your printer would probably be Ok for the curved roof as the resolution seems very good, so you could print it right way up.

 

That's the way I'm doing it at the moment, the problem is that the surface that faces down (with the supports on it) doesn't come out all that nicely.

[richbrummitt]

Maybe you could try a near whole body with a smooth(ish) recess in one end and a little detailed end as an insert to fit in to it . The supports would be sub flush to the mating surfaces such that perfection in cleaning the supports off is not a requirement? Probably easier to explain with pictures - let me know if you haven't a clue what I mean and I'll sort some data out.

 

For the Mink I think I should get them resin cast and therefore would only need to make one good one for a master.

[Rabs]

I think I understand but a sub-flush support on a downward facing surface by definition isn't supporting the lowest part, so the bits below it won't print.  I am getting there with them, and the joining process is giving pretty good results (to my eye anyway, you may be more exacting).  I'll do the best I can and send them to you for you to decide if they are good enough.

[richbrummitt]

I think I understand but a sub-flush support on a downward facing surface by definition isn't supporting the lowest part, so the bits below it won't print. 

 

 

[Pugsley]

I'm loving the Spot HT - just a little bit of playing this afternoon has lead to results that far surpass what the B9-R resin is capable of.  Please excuse the quality of some of these pics, they're cropped from full size, but the detail that is possible with this machine is just incredible.

 

 

This is a calibration model that I knocked up so I could determine the capabilities of the machine/resin combinations against a constant reference.  The surface features start from 0.1mm, upwards, and I also included very thin unsupported walls, that I suspected wouldn't work.  The top line of embossed text, just visible on the box to the bottom left of the image, is only 0.06mm wide!

 

 

The bolt detail has also come out extremely well - the circled nut/bolt feature is only 0.4mm across flats and is quite clearly hex shaped under magnification.  With the smaller ones, it's not so obvious, but you can tell that they're not round.

 

 

The 7mm axleboxes have come out very nicely in this new resin, far better than anything else I've had so far.  Shrinkage seems to be pretty low as well, only 1.2% and seems pretty consistent in all directions.

 

 

The extreme closeup shows some minor defects, but this is very extreme magnification, the actual parts are only 8mm across at the widest part.

 

I'm confident enough to say that the results that I've got this afternoon are better than Shapeway's FUD, the level of detail is the same, but there is none of the frosting problems that can sometimes occur with that material.  It's even more impressive given the difference in cost between the B9 and the machine that creates the FUD parts.

[Rabs]

I'm glad the Spot-HT is working well for you, I saw a similar improvement when I switched over to it.

[Etched Pixels]

GWR wagon looks nice. How do the integral bearings wear though. FUD ones look very nice until the vehicle has done a few real world miles...

[Rabs]

It's a good question and exactly what I printed the test wagon to investigate.  I'm also preparing versions which use brass pinpoint bearings and a version where a normal chassis is used with a printed body.

[Rabs]

So, after an absolute age I've got some results to share on the rolling stock front.  These are two models designed by richbrummitt, a GWR Beetle C and a Mink F - both in 2mm scale, so the beetle is about 55mm long and the mink is about 70mm long.

 

I've printed both of these models quite a few times (the beetle is probably over 30 now and the mink about 5!) as I gradually home in on the best way to achieve the end result.

You'll also notice that I've added a couple of GWR plate bogies, of my own design, to this batch.  These should go nicely under the minks.

 

Those following this thread will have seen a few of my attempts, which were pretty good, and may wonder what has changed between then and now.  Well, the answer is that I've gone from 'pretty good' to 'good enough for my layout'  Previously I had slight niggles like small holes in the model, or some details not coming out, or one end being slightly warped.  I'm fairly sure that I've now eliminated these problems, most of the time.  The final question that I was puzzling over was whether to print in one part or two.  One part has the advantage that you don't have to join and then conceal a join in the model.  The disadvantage is that you have to put supports somewhere on the visible faces of the model - which can then leave a mess or destroy detail when they are removed.  Compare the ends of this beetle, the first image is the end which is printed facing up, the second is the end facing down before I've made any attempt to tidy it up:

 

I can make the supported end look much better than this with careful sanding but you always lose the fine strapping and rivet detail at the edges.  It's ok but this is a model which is better printed in two parts.  Here is a side view of the wagon printed in one (above) and two (below) parts:

Sorry about the poor photo here but hopefully the result is clear.  Because there are lots of vertical features it's easy to conceal a join without any real compromise.  In fact, if I'd managed to get the solebar straight it would be hard to tell where the join was.  This is worth the effort to get two good ends.

 

The mink is a different story - because it is mostly smooth sided it is both easier to remove the supports without losing detail and harder to make an invisible join.

(Mink supported end after a first pass clean up of the end where the supports were)

 

(one and two piece minks after filling the join but before second priming and sanding)

 

All in all, the mink could be made into a perfectly acceptable model with either method, but it's easier to do this one in one piece.

 

Here they all are after selective joint filling and sanding:

 

One issue I have had is that the roofs in these models are very delicate.  They print fine but the exposed bits tend to break off when the model is handled:

For my purposes I'm happy to thicken them up to 0.5mm (currently 0.2mm) because I don't think it's that visible a difference.  Others might prefer to print them without roofs and then overlay an etched roof.  Richard, which would you rather?

 

As a final point, my photography isn't very good - mostly because I'm lazy and I use my mobile phone.  If you can't see rivets in any of the photos above it's because of the photo and not the printing.  All of the printing is as crisp as the best of these photos, all over the models (except where there are supports).

[Southernboy]

I've been following progress here for some while Rabs, and have to say I admire your perseverance.

 

This is impressive, pioneering work, and I take my hat off to you.

[PhilH]

I've been following progress here for some while Rabs, and have to say I admire your perseverance.

 

This is impressive, pioneering work, and I take my hat off to you.

 

Absolutely second that.....thanks to all for sharing the journey on here.

[Etched Pixels]

The roof one is a problem with the commercial processes as well. Best approaches I've found so far for anything near scale thickness are to either use etched roofs or cut a vinyl for the roof if the roof is smooth. Both might also help you with the printing as you'd perhaps also then be able to sprue and support in the area concealed by an etched roof ?

 

Alan

[richbrummitt]

For my purposes I'm happy to thicken them up to 0.5mm (currently 0.2mm) because I don't think it's that visible a difference.  Others might prefer to print them without roofs and then overlay an etched roof.  Richard, which would you rather?

 

I believe I amended my own CAD such that the roof is no longer part of the model. I had hoped that the Mink might be possible to use as a master for resin casting (most likely without the buffer housings) and that the resulting models could be finished with an etched roof, brass buffers and bogies from either the 2mm scale association, Bachmann or the NGS - who all have suitable bogie mouldings/castings. Whether this can be achieved at a viable cost is yet to be established but that was/is my hope. The beetle would be more difficult to do because of the windows. I like things to be the size they should be full size at scale size. They must also be robust, which is why I continue to toil making under frames from metal.

 

As a final point, my photography isn't very good - mostly because I'm lazy and I use my mobile phone.  If you can't see rivets in any of the photos above it's because of the photo and not the printing.  All of the printing is as crisp as the best of these photos, all over the models (except where there are supports).

 

The detail on the Beetle appears gorgeous and better than shapeways comparing to pictures - there is a topic here that was linked back on page 2 with the shapeways version (different details on ends, same sides/roof). Incidentally the shapeways prints had fragile roofs and one arrived in a similar condition to the damaged picture. Another issue that I have with the shapeways parts is that there is a diagonal striping in the finished items, which is very visible from some viewing angles, and I detect what might be evidence of similar artifacts due to the layering in your own prints. This is particularly noticeable on the one piece Mink in the side by side photo. It should clean up easily on these flat surfaces. It would have been a real pain to do on the Beetle, so I didn't bother trying.

 

Thank you for all the work and effort that has gone into these two (although I read 35!) models. 

[Rabs]

Some more pictures of the cleaned up ends of the models:

As you can see, the mink really doesn't look too bad and I will definitely print this in one part.

 

The one-piece beetle came out better than I expected.

It does lose some of the rivet detail on the edge straps where I've had to sand down the support points.  This one ends up about even on the two methods.  With practice and patience the joining method and two good ends will give a better model but it's a lot of work just to keep two rows of rivets! For my purposes I will probably do one piece prints of this type of model too.  Most of the stuff that looks like run paint near the roofline is where I wasn't quite rigorous enough in removing the excess resin - it's not typical.

 

Some side views in better light to show the details:

(bit grubby this one - I need to pick the cat hairs off with tweezers before priming next time!  Unfortunately they are ginger cats and the resin is yellow - so it's a bit hard to see them!)

 

(The stanchions here are straight, it's just the rubbish lens that makes it look like they bend to the left at the bottom)

[chris p bacon]

- I need to pick the cat hairs off with tweezers before priming next time!  Unfortunately they are ginger cats and the resin is yellow - so it's a bit hard to see them!)

 

 

Give them to the next door neighbours they wont mind..........

[Rabs]

Well, considering the detail on these models (and current commercial N gauge) I'm thinking that some sort of cleanroom/clean cabinet might be needed!

 

A couple of three quarters views of a beetle printed in one piece.  The roof even survived on this one.

 

And a close up of the plate bogie I drew to go under the mink.  I made the holes for the brass pinpoint bearings too small on this one, so I'll revise the model and try again.  These will take 2mmSA pinpoint bearings and 12.25mm axles, so are suitable for 2mm or N with suitable axles (not standard farish or Dapol axles).  I did look at making one wide enough for standard N gauge axles but it looked silly, sticking far outside the width of the wagon above it.  Unfortunately the 2mm shop seems to be discontinuing their N gauge wheels, and the axles (2-047) to convert farish wheels are quite expensive.  So, time for some lathe work I think.

[Etched Pixels]

Conversion axles are available to put the spoked Farish wheels onto 12.25mm axles. They are not perfect but they work. It would limit the period you can pedantically accurately model to the early 1930s and before when they received disc wheels. Mind you pedantically speaking they should have 3'1" wheels not 3' wheels

 

Alan