Jump to content
 

3D Printing in 2mm Scale


TomE
 Share

Recommended Posts

  • RMweb Premium

I'm just trying to get into laser cutting acrylic to do windows for tiems such as this. There is a Dutch company who you send a file to and they cut, but also thinking about getting my own machine. Anyone any expereince orf that?

 

Do you have a link for the "Dutch company"?

 

David

Link to post
Share on other sites

Finally had the time to finish off my design for the GER butter wagon (p.209 of Tatlow), which was mainly a learning exercise in Fusion 360. I chose this wagon for being basically quite a simple box. However the inset plank detail is actually very varied, so took some fiddling. The vertical plank lines look virtually invisible in the photos in Tatlow, but I put them in with a 0.05mm groove (pipe) compared to a 0.15mm pipe for the horizontal ones. We'll see how it comes out! Of course, now I notice I forgot to include the securing chain around the door catch, which I had been planning to model.

 

It took a while to get my mind into the way of working with the parametric structure of 360 - basically which bits to mirror and pattern - and why to avoid going back and changing anything too fundamental (as all the later edits disappear because their source sketch is now missing)! I drew the door hinges as a "Component", thinking that would be a way of saving time, but actually found it quite the opposite - components don't seem to mirror or repeat when the body they're attached to does (even with an explicit joint). So I won't bother with that function again - or at least not for cosmetic bodywork features. 

 

post-3740-0-59720400-1544397339_thumb.png

 

I've set the design up on its own in Photon with rotation in both directions. Probably a bit high off the plate in hindsight, as the print time came out as 5hrs on the screen (as opposed to 4 hours estimate in the Slicer software). And of course now I post the picture, I've noticed some slightly errant looking supports. I remember seeing a suggestion to take them off BEFORE curing, to get a better finish - might be worth a try (presuming the whole thing doesn't fail ...)

 

post-3740-0-81494900-1544397351_thumb.png

 

Guess I'll just have to see what is waiting for me in the morning!

 

Justin

  • Like 4
Link to post
Share on other sites

  • RMweb Gold

Probably a bit high off the plate in hindsight, as the print time came out as 5hrs on the screen (as opposed to 4 hours estimate in the Slicer software). And of course now I post the picture, I've noticed some slightly errant looking supports. I remember seeing a suggestion to take them off BEFORE curing, to get a better finish - might be worth a try (presuming the whole thing doesn't fail ...)

 

Guess I'll just have to see what is waiting for me in the morning!

 

Justin

 

Excellent work on the van, looking forward to seeing how it comes out! I've given up with the auto support feature in the slicer software and just add supports myself before running it through a validator to make sure there are no unsupported areas.  I also fell foul of the extra print time vs the slicer print time estimate yesterday evening myself! Still, I only really needed 6 hours sleep the day after a dual Christmas party weekend.... -_- :blink:

 

Fresh from the printer plate at 01:30 this morning: 

 

post-1467-0-87458200-1544434025_thumb.jpeg

 

This is my second attempt using Anycubic grey resin on another test PALVAN print after making some adjustments to the 3D model. After a failed first attempt with the AC grey, I went back and adjusted the exposure time and it's produced a reasonable result this time. 

 

Tom.  

Edited by TomE
  • Like 8
Link to post
Share on other sites

First impressions when I got it out were good!

 

post-3740-0-50783000-1544437546_thumb.jpg

 

There is visible stepping, but on a planked wooden wagon, I was actually hoping to get something like this. 

 

However on the other side there was a bit of a "slump", which is also there on the end that was sloping down toward the bottom (and which also had the stray support cutting through it).

 

post-3740-0-14122300-1544437596_thumb.jpg

 

Thinking logically, this seems to begin where the roof started printing, and is only there on the sides closest to the build plate - so probably pull stresses from the weight/volume of the roof. I hollowed the wagon out, but left the walls quite thick (nearly 2mm) and didn't make any special effort to hollow into the roof (so its basically a "solid" part cylinder on top of the walls). I guess this means there is more "suction" from the FEP when the roof starts printing - and the corner closest to the build plate is taking most of those forces - hence the bulge? 

 

I'll try and do some more tonight, with a few changes: beefed up details (especially the locking bar) and one or two missing details; thinner walls and hollowed roof.

 

I'll probably also try reducing the tilt through the wagon a bit to avoid concentrating the stresses on that corner so much. (and do the supports manually!)

 

Really impressed by Tom's Palvan! I can see some stepping on the end stanchions, but the sides look perfectly smooth :) What were the rotation settings?

 

Justin

  • Like 2
Link to post
Share on other sites

  • RMweb Gold

That looks excellent Justin! I did find that while features appear to be fairly obvious & well defined when looking at the 3D model on the monitor, when you print you quickly realise you really need to beef things up a bit to get them to show properly on a tiny model!

 

The sides on the PALVAN are not totally flat and I have a similar ridge from the corner where the roof starts, so I'm wondering if it's because there is more surface area being pulled off the FEP as the roof starts. Now I'm happy with the rivets I'm going to finish the CAD then print some at different angles and try to find the sweet spot!

 

I printed this one at 0.20 layer thickness and 30 degrees angle off the horizontal and with a 10 degree angle on the print plate. 

 

Tom. 

Link to post
Share on other sites

Having absolutely no experience of orienting a 3D print, can I ask whether anyone has tried printing the model upside down with no tilt?  In my minds eye I visualise that because the roof overhangs the ends and sides that when the ends and sides are printed that they will be supported by the up-turned roof.  I can imagine that some internal supports between sides/ends might be needed to minimise bowing.  Just a thought and please tell me to go back to my own thread and mind my own business if I'm taking complete nonsense :-)

 

Ian

Link to post
Share on other sites

Hi Jon,

 

Many thanks regarding my coach test print - it is a very old file, was never finished and needs completely redoing to correct some errors that I've discovered.

 

20mm is too far apart for the cross bracing as it still allows the part cured resin to warp before it reaches the next brace. Regarding supports, I use my old B9Creator software to generate the supports and then export the file as an STL. It doesn't have an auto support function but does have an x-ray function which allows you to see inside the model to add the supports. I find this much more useful and ChiTu Box offers something similar, although I don't find it as easy to use (but I do do my slicing in ChiTu). I believe that both programmes are available for free (ChiTu certain is).

 

Just out of question, what exposure times are you using?

 

Thanks, Steve.

 

Twisting the model seems to have improved surface finish no end

post-10246-0-83993000-1544456413_thumb.jpg

 

This done at 25 degree twist, 10 wasn't enough fixed only one side of the model, other side tumblehome still had stepping.

 

(Remember all, this is at 4mm scale.)

 

Now just need to work out how to best print this, you can see model failed, this being in the bottom corner I think due to lack of support. (And I guess this is the corner with the most force on it when printing?) Also how to arrange the support on the overhang. 50/50 whether it was needed, most of them had pulled off/snapped. So also trying more beefy supports. Final challenge is prints are starting to come off the build plate with more frequency, first ones I did were all ok. I've increased size of support footings and sanding the plate but may try chemical adhesion improved, Monocure Platebond.

 

This was printed at 100 micron layers using Moncoure Rapid White resin at 9 seconds per layer. I have another resin I've been trying called Druckwege but that takes fair bit longer 14 seconds or so per layer. Even with the faster resin, print takes 10 hours of so. As it seems problems generally happen near start of print, in early pases, I'm now getting into habbit of checking 1/3 to 1/2 way through if timing works to see if all going well and aborting if its gone wrong.

 

Thanks again

 

Jon

 

(Edit - PS - the diagonal line across the print is a minor niggle too, possibly a dead row of pixels in the printer LCD screen?)

Edited by The Great Bear
Link to post
Share on other sites

Having absolutely no experience of orienting a 3D print, can I ask whether anyone has tried printing the model upside down with no tilt? In my minds eye I visualise that because the roof overhangs the ends and sides that when the ends and sides are printed that they will be supported by the up-turned roof. I can imagine that some internal supports between sides/ends might be needed to minimise bowing. Just a thought and please tell me to go back to my own thread and mind my own business if I'm taking complete nonsense :-)

 

Ian

Entirely logical Ian! The problem would be that the roof would have to have a raft of supports under it, which would leave lots of witness marks when they're removed. Doing it "right way up" (to the build plate) as Tom and I have done so far, means the supports anchoring it to the plate can attach on the inside.

 

BUT on a simple roof like this, where frankly even the rainstrips could be easily replaced if lost, it might well be less work to clean that up, compared to dealing with artifacts on the sides ...

 

All food for thought - and experimentation - but it's probably never going to be perfect! Certainly using kit costing as little as the Photon does.

 

Justin

Link to post
Share on other sites

Have you considered drawing and printing body and roof as separate items which might get over the problems of proper support for each?   If the roof had a plug on the underside fitting in the body,  then assembly would be easy.

 

I'm following this with interest.  I've just acquired a Shuffle and my problem at the moment is getting up to speed with a 3D CAD package - my thirty years of 2D CAD skills aren't helping much. :-)

 

Jim.

Edited by flubrush
Link to post
Share on other sites

  • RMweb Premium

These set-ups will certainly give someone a head start in making a model of a lifeboat station!

 

Fascinating, promising stuff. Worth bearing in mind that 2mm detail is always best understated.

 

Tim

Edited by CF MRC
Link to post
Share on other sites

  • RMweb Gold

Sorry for the crap photo but I quickly primed the PALVAN before heading out to work this morning.

 

post-1467-0-32867500-1544530376_thumb.jpeg

 

Again, no work done other than to remove & tidy up the support attachments. It does show I have a line similar to Justin's emanating from the top right corner where the sides meet the roof. This was the area of the print with the most surface area in contact with the FEP at the bottom of the resin tank, so I suspect the increased suction required to pull it off the FEP is to blame. I'm going to try increasing the angle the van is printed at next time to reduce that surface area, see if that makes a difference. 

 

The roof on the other hand has come out perfectly! 

 

Tom.  

Edited by TomE
  • Like 6
Link to post
Share on other sites

So I printed four more as a batch last night. None had such obvious "slump" lines at the junction of side and roof as the first attempt. But also there doesn't seem to have been so much logic as to whether sharper or lower angles worked better. One thing that did seem consistent was aligning the length of the wagon perpendicular to the long edge of the build plate, rather than parallel to it, seemed to produce better results.

 

In fact, with this particular model, the best results seemed to be the copies printed with rotation in one direction only. (Which seems to go against received wisdom). However, the one rotated along its length was victim to another errant support I missed removing when setting it up late last night. And the one rotated up on its end, I didn't clean thoroughly enough before curing, and managed to cure on an ugly drip of liquid resin on one corner!

 

The model that was upside down, with a 30 degree rotation up at one end was the worst result of all, with really obvious diagonal stepping, even without the extra surface area of the roof to pull. I will try another on its roof but no rotation, but then though the supports came off quite well, it looked like there was some evidence of stress with little lumps and bumps showing where the supports had been "pulling" the food. So not really seeing that as a solution.

 

post-3740-0-71628900-1544536458_thumb.jpg

 

Another four are printing now, with tweaks and better supports to the more successful orientations of the last print.

 

I think the separate body and roof is probably a good idea longer term. If no luck with this batch, I'll rework to make the roof plug in. And probably design vans that way in future anyway.

 

And Tim, the detail is pretty understated in reality! I "beefed up" the locking bar from 0.2mm to 0.3mm! Ironwork plates are 0.1mm and rivets 0.2mm!

 

Justin

  • Like 1
Link to post
Share on other sites

Finally, a print I'm happy with in all respects!

After several more prints that were compromised in one way or another, I've settled on a model with sides and ends about 1.5mm thick (roof slightly thinner). Rotation has been about 15 degrees through the length of the wagon, and anything from 30 to 40 degrees up at one end.

I've come to realise just how much care needs to go into designing the support structures, and increasing the gap between the model and the build plate to more than 5mm made far more difference than anything else. I found that ChiTuBox is actually easier to set things up effectively than the provided Photon Slicer app.

 

post-3740-0-77062600-1545095024_thumb.jpg

Edited by justin1985
  • Like 4
Link to post
Share on other sites

One of my earliest prints, a C1, is slowly progressing through the works.

 

post-943-0-06047200-1545159705_thumb.jpg

 

There are several things I would do differently (covered earlier) when I get around to printing another one to get a better result now.

Edited by Atso
  • Like 18
Link to post
Share on other sites

Finally, a print I'm happy with in all respects!

 

After several more prints that were compromised in one way or another, I've settled on a model with sides and ends about 1.5mm thick (roof slightly thinner). Rotation has been about 15 degrees through the length of the wagon, and anything from 30 to 40 degrees up at one end.

 

I've come to realise just how much care needs to go into designing the support structures, and increasing the gap between the model and the build plate to more than 5mm made far more difference than anything else. I found that ChiTuBox is actually easier to set things up effectively than the provided Photon Slicer app.

 

attachicon.gifIMG_20181218_003010.jpg

Interesting, thanks for sharing. What was the benefit of lifting that much off the base?

Edited by The Great Bear
Link to post
Share on other sites

Interesting, thanks for sharing. What was the benefit of lifting that much off the base?

 

It seemed like when a corner was closer to the build raft than about 5mm, that corner started to be distorted. If you look at my first attempt (post no.55) you can see the bottom corner has been pulled down from level if you look along the bottom. Not entirely sure why the forces work differently when it's on longer supports, but seems to have made a big difference. 

 

A bigger gap also makes it easier to get the remains of the excess resin out, which can carry on seeking down from the inside after you've cleaned up. If the body is too close to the raft, the resin tends to creep back up. Several of my previous attempts that were otherwise OK were ruined when excess resin reappeared and crept onto the outside during the UV curving. 

 

Justin

  • Thanks 1
Link to post
Share on other sites

This thread has persuaded me to take the plunge. For your delectation I am happy to present my first project. An L&Y wagon, 3 D model built using Sketchup. (Pro version) I build at a scale of 1 m to 1mm in the model ( 1000x full model size) then shrink on the computer, before slicing. The strapping is a bit on the thick size, but all in all for a first attempt it has all gone very well.

post-14910-0-89351200-1545768921_thumb.jpeg

  • Like 11
Link to post
Share on other sites

While the family fell asleep in front of crap Christmas day TV ...

 

post-3740-0-65131500-1545769996_thumb.png

 

A Great Eastern 10 ton, 7 plank, 17ft long open wagon of 1910, based on a drawing in Tatlow.

 

I'm starting to quite enjoy working in Fusion360! The interactive timeline at the bottom is really handy. I've left the superimposed canvases from the scanned drawings, and the various construction panes in the screenshot. This neatly shows the method of drawing one quarter of the wagon, mirroring, adding the non-symmetrical features to the sides, then mirroring again to make a complete wagon. The only details I'm aware of leaving off are the rope securing loops and the mechanical door stop. The interior is a pretty rough representation.

 

I've drawn several details slightly over-scale to help with printing and durability: strapping is 0.15mm thick, rivets mainly 0.225mm dia and raised 0.15mm, and the T section ironwork on the end is much meatier than it should be, and also tapered in section.

 

Justin

  • Like 7
Link to post
Share on other sites

While the family fell asleep in front of crap Christmas day TV ...

 

attachicon.gifGER10tonOpen.png

 

A Great Eastern 10 ton, 7 plank, 17ft long open wagon of 1910, based on a drawing in Tatlow.

 

I'm starting to quite enjoy working in Fusion360! The interactive timeline at the bottom is really handy. I've left the superimposed canvases from the scanned drawings, and the various construction panes in the screenshot. This neatly shows the method of drawing one quarter of the wagon, mirroring, adding the non-symmetrical features to the sides, then mirroring again to make a complete wagon. The only details I'm aware of leaving off are the rope securing loops and the mechanical door stop. The interior is a pretty rough representation.

 

I've drawn several details slightly over-scale to help with printing and durability: strapping is 0.15mm thick, rivets mainly 0.225mm dia and raised 0.15mm, and the T section ironwork on the end is much meatier than it should be, and also tapered in section.

 

Justin

 

Clever :) I like the superimposed drawing, will have to explore that. Do you use the sketch facility in Fusion? I hardly use it, coming from a proper CAD background it is really limited for me so for complex shapes, like coach body profile, CAD would be my starting point then import in.

 

I find the timeline a mixed blessing. Some of my designs now have such a long history it really slows Fusion down indeed sometimes makes unworkable so towards the end am having to turn it off, despite having a high spec pc. The versioning of the files on save is useful and gives you a way back anyway, if needed.

Link to post
Share on other sites

Clever :) I like the superimposed drawing, will have to explore that. Do you use the sketch facility in Fusion? I hardly use it, coming from a proper CAD background it is really limited for me so for complex shapes, like coach body profile, CAD would be my starting point then import in.

Yup I've been using the Sketch function. I'm only vaguely comfortable with CAD, and probably never really made the most of complex curves etc. For wagons, Sketch seems totally fine - and all the core functions like split, mirror, offset and pattern work fine. For me, it seems like a nice halfway house between SketchUp and "proper" CAD, that suits me quite well.

 

I find the timeline a mixed blessing. Some of my designs now have such a long history it really slows Fusion down indeed sometimes makes unworkable so towards the end am having to turn it off, despite having a high spec pc. The versioning of the files on save is useful and gives you a way back anyway, if needed.

I guess the wagons I've designed so far are much simpler than coaches, but I've only really found this to be a problem if trying to reorder items a long way back (e.g. forgot to cut the plank lines before extruding the strapoing - so dragged it back so the reverse of the strapping wasn't also cut by the pipes being subtracted from the sides to make the plank lines). I guess it's also an encoragement to use the pattern, mirror etc functions efficiently! I find another really useful aspect is being able to scan back and alter the settings of an operation (e.g. increase or decrease an extrude) and then have that change filter through all the subsequent steps. I've no idea how "standard" that function is, but I've certainly taken to it!

 

Justin

  • Like 1
Link to post
Share on other sites

That looks fun Justin,

One question/ comment.

 

Do the mirror lines show in the print? And don’t forget an internal vertical groove on the top two planks, where the cupboard doors open. difficult to see from the image if they are there.

 

R

Link to post
Share on other sites

Good call Richard - I had indeed missed it! I also just noticed it's missing notches in the top face to represent both the middle and the hinge sides of the doors. I'll add these in on one segment and drag to before the mirroring to complete it.

 

The mirror lines have no actual thickness, so presumably wouldn't show on a print in any case. But my final step before exporting as an STL for printing would be to select all four "bodies" of the mirrored quarters and do a Combine operation.

 

Hopefully should be able to do a test print on the 27th or 28th. Anycubic sent me a new firmware to hopefully address the excess resin curing on the FEP issue I had last time, and I ordered some fresh FEP sheets from FEPShop in the Netherlands on the afternoon of Christmas Eve - which they not only dispatched the same day, but upgraded from the postal option I chose, to UPS 24hr! So I should be good to go!

 

Justin

Link to post
Share on other sites

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
 Share

×
×
  • Create New...