Jump to content
 

Printing Turnouts on a 3-D printer


AndyID

Recommended Posts

Not much difference, except the wheel flanges don't clonk into the chairs with this one, and the surface of the sleepers is a bit smoother. (The sleepers are really the width of point timbers rather than sleeper width.)

 

post-25691-0-09823200-1439014472_thumb.jpg

 

 

Link to post
Share on other sites

  • 2 months later...

Andy

 

Is this the end of the story?

 

Hi Ken,

 

Definitely not.

 

I now have a better printer and I've been optimizing the chair design toensure I produce the correct gauge, clearances, etc.

 

Hoping to post some new turnout pix in the next few days.

 

Cheers!

Andy.

Link to post
Share on other sites

First piece of turnout base off the new printer. I still need to make some adjustments to the CAD drawing, but it's not too bad for a first article. Should look a bit better with some rail installed.

 

This printed in just under 22 minutes. It's printed in brown plastic which has quite a sheen. The top side has been painted with flat brown but looks black in the photo. The timbers are not solid. They are printed with a honeycomb infill.

 

post-25691-0-47302300-1446361230_thumb.jpg

 

post-25691-0-35092600-1446362360_thumb.jpg

Link to post
Share on other sites

What about using conductive glue to fix wires to the bottom of the rails.

Not cheap, but probably goes a long way.

 

http://uk.rs-online.com/web/p/conductive-adhesives/1863616/

Or what looks like similar, but much cheaper, stuff that repeatedly turns up on my eBay searches to irritate me, even though I'm not looking for it!

http://www.ebay.co.uk/itm/Electric-Glue-Model-Railway-Train-Wiring-Track-Motor-Freight-Car-HO-ON30-F-N-DCC-/331696373369?hash=item4d3aa4fe79:g:9EQAAOxyuGFRz-wk

Link to post
Share on other sites

If it is as sturdy as normal plastic sleepers and chairs a quick in and out with a soldering iron is fine, or use 70 degree solder (low melt)

 

Or pre-solder wire to the rail ends, or drill a hole in the end push in a wire dropper and solder before fitting

Link to post
Share on other sites

If it is as sturdy as normal plastic sleepers and chairs a quick in and out with a soldering iron is fine, or use 70 degree solder (low melt)

 

Or pre-solder wire to the rail ends, or drill a hole in the end push in a wire dropper and solder before fitting

 

 

Hi John,

 

Yes - I'll try hot a quick. Think it will be OK. That's why some of the webs are missing.

 

Andy

Link to post
Share on other sites

Looking very nice Andy. Are the sleepers square to the main road or equalized?

 

And any chance of a P4 1:10 one next?  :jester:

 

Quentin

 

Thanks Quentin. They are equalized on this one.

 

P4 is always a possibility, but I thought I might try some 5'3" gauge at 1:76.2 first :)

Link to post
Share on other sites

Talk about "pushing the envelope"  (BTW, why does "envelope" not rhyme with "Penelope"? - try it the next time you are at the post office.)

 

I discovered a couple of things - TurboCAD blows up if I add all the solid 3-D elements in a turnout base to make a single 3-D solid object. The model seems to get just too big for it to handle. The solution seems to be to convert solid models into surface models which is a bit of a pain as it involves another step in CAD.

 

It turns out there is a feature in the slicer program I use that I only just discovered (it was disabled in the version that came with the printer.) The slicer takes the model and slices it into individual layers for the printer. There is a trade-off between vertical resolution and speed - a lot of thin layers take a lot longer to print. Sleepers and timbers don't need high resolution, but the chairs do. It turns out that I can use different layer depths for the sleepers and the chairs, and that will help to improve the rail holding characteristics of the chairs without increasing the time to print timbers.

 

I'm not overly impressed by the horizontal (X and Y) resolution of the printer. It's just about OK, but I'm concerned that the track gauge and clearances might be a bit inconsistent, so I'm going to try using higher resolution stepper motors. It should make a difference, but there are a lot of factors at play. We'll see.

Link to post
Share on other sites

Without wishing derail to your thread, here's a video of a diamond crossing I had 3D printed by Shapeways

https://onedrive.live.com/redir?resid=DEB2591C47894A10!4615&authkey=!AJpNz_vWhQFGJbo&ithint=video%2cmov

Note that this was before I added the wing/closure rails

It did work in anger as part of a large Freemo modular setup (I finished building it night before) but I am intending to get it printed again, and build a new one, with some adjustments and some jigs for filing the short rail sections, there was just too much filing and tweaking and melting bits into place with a soldering iron required for my liking and it has ended up a bit of a bodge. When I built it again I will write it up in a separate thread.

Link to post
Share on other sites

The higher resolution stepper motors (0.9 degrees per full step instead of 1.8 degrees) have been installed and they seem to be doing the trick. So far, the gauge of the samples I've printed has been a lot more consistent but I'll not know for sure that it's completely satisfactory until I've printed a lot more pieces.

Link to post
Share on other sites

  • 3 weeks later...

I think this one might actually be a "keeper".

 

post-25691-0-22507200-1448355768_thumb.jpg

 

Just doing a bit of gauge checking and it seems that the higher resolution stepper motors are behaving as I had hoped.

 

It took a large number of chair iterations before I managed to get the rail vertical. Any tilt has a serious effect on the gauge. Then I had to learn some stuff about CAD that was new to me.

 

The various elements (all the different types of chairs) are objects in a library. Theoretically, I can now simply drag them onto a new turnout template, orient them correctly with the rails and send a file to the 3-D printer - there are a couple of other steps involved so it's not quite that simple, but that's the general idea. But I'm not sure if it will really pan out that way.

 

 

Link to post
Share on other sites

Talk about "pushing the envelope"  (BTW, why does "envelope" not rhyme with "Penelope"? - try it the next time you are at the post office.)

 

I discovered a couple of things - TurboCAD blows up if I add all the solid 3-D elements in a turnout base to make a single 3-D solid object. The model seems to get just too big for it to handle. The solution seems to be to convert solid models into surface models which is a bit of a pain as it involves another step in CAD.

 

It turns out there is a feature in the slicer program I use that I only just discovered (it was disabled in the version that came with the printer.) The slicer takes the model and slices it into individual layers for the printer. There is a trade-off between vertical resolution and speed - a lot of thin layers take a lot longer to print. Sleepers and timbers don't need high resolution, but the chairs do. It turns out that I can use different layer depths for the sleepers and the chairs, and that will help to improve the rail holding characteristics of the chairs without increasing the time to print timbers.

 

I'm not overly impressed by the horizontal (X and Y) resolution of the printer. It's just about OK, but I'm concerned that the track gauge and clearances might be a bit inconsistent, so I'm going to try using higher resolution stepper motors. It should make a difference, but there are a lot of factors at play. We'll see.

I'm looking at buying a 3D printer to do mall work. Having owned the i3 would you still recommend it.

 

Have you a link to the steeper motor upgrade , did the printer software need changing

 

In your estimate what sort of x-y-z resolution are you achieving

 

Thanks

 

Dave

Link to post
Share on other sites

I'm looking at buying a 3D printer to do mall work. Having owned the i3 would you still recommend it.

 

Have you a link to the steeper motor upgrade , did the printer software need changing

 

In your estimate what sort of x-y-z resolution are you achieving

 

Thanks

 

Dave

 

Hi Dave,

 

Yes, the i3 is pretty good for the money. I also have a Printrbot Simple but it uses friction drive which was tending to slip. I like the gantry design on the i3 as it seems to maintain more consistent vertical positioning.

 

I'm not entirely sure what the resolution really is. I think the firmware uses 16 microsteps so it should be possible to figure it out from there.

 

The first thing I did was to swap the extruder nozzle for a .25 mm version. Then I swapped the X and Y steppers for 400 steps per revolution instead of the supplied 200 steps per rev. I also swapped the X and Y drive sprockets for 16 tooth versions instead of the 20 tooth supplied. I bought all the bits on Ebay - not very expensive either.

 

No changes to the firmware. I only had to change the steps per mm in the flash.

 

Cheers!

Andy

 

EDIT: I forgot to mention I also added a fan to cool the filament affter it's extruded. The i3 has a fan, but it's only used to cool the "cold end".

Link to post
Share on other sites

OK. Finally getting to the point.

 

post-25691-0-30240500-1448694848_thumb.jpg

 

Of course my stinking annoying CAD software fought me all the way. Sometimes, for no particular reason, it fails to properly "add" various bits of the chairs into a single solid model. They look OK, but the printer software complains so I have to go back and resort to some tedious editing.

 

But I think I've got the overall process nailed down. Still need to make some tweaks to the models. Bolt heads need enlarged a bit etc. but I think these prints are good enough to finish the complete turnout.

 

 

Link to post
Share on other sites

Andy

 

Looks excellent and perhaps something we all thought not achievable so quickly a few months back. Just to overcome fitting the common crossing together is impressive. How about a slip next  :jester:

 

Would be interesting seeing a standard RTR offering next to it just to compare sleeper size/spacing and to see the difference in the geometry of the turnout

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...