Templot Club forums powered for Martin Wynne by XenForo :

TEMPLOT 3D PLUG TRACK - To get up to speed with this experimental project click here.   To watch an introductory video click here.   See the User Guide at Bexhill West.

  • The Plug Track functions are experimental and still being developed. Some of the earlier pages of this topic are now out-of-date.

    For an updated overview of this project see this topic.   For some practical modelling aspects of using Plug Track see Building 3D Track.

    The assumption is that you have your own machines on which to experiment, or helpful friends with machines. Please do not send Templot files to commercial laser cutting or 3D printing firms while this project is still experimental, because the results are unpredictable and possibly wasteful.

    Some pages of this and other topics include contributions from members who are creating and posting their own CAD designs for 3D printing and laser-cutting. Do not confuse them with Templot's own exported CAD files. All files derived from Templot are © Martin Wynne.
  • The Plug Track functions are experimental and still being developed.

    For an updated overview of this project see this topic.   For some practical modelling aspects of using Plug Track see Building 3D Track.

    The assumption is that you have your own machines on which to experiment, or helpful friends with machines. Please do not send Templot files to commercial laser cutting or 3D printing firms while this project is still experimental, because the results are unpredictable and possibly wasteful.

    Some pages of this and other topics include contributions from members who are creating and posting their own CAD designs for 3D printing and laser-cutting. Do not confuse them with Templot's own exported CAD files. All files derived from Templot are © Martin Wynne.

Experimental Plug Track: 3D-printed, CNC-milled, laser-cut

Quick reply >
Hi Martin,
The other even more deliberate mistake is that you have not first threaded the chairs onto rail before "plugging" them into the sleepers!
:):):)
Would the chair distort if it was already threaded on rail?
Perhaps a few boiled eggs are called for?
Seriously though, please keep up the good work on the chairs, as they are looking really good.
Steve
 
_______________
message ref: 4742
Hi Martin,
The other even more deliberate mistake is that you have not first threaded the chairs onto rail before "plugging" them into the sleepers!
:):):)
Would the chair distort if it was already threaded on rail?
Perhaps a few boiled eggs are called for?
Seriously though, please keep up the good work on the chairs, as they are looking really good.
Steve
@Steve_Cornford

Thanks Steve.

There was rail in them when I inserted them! I removed the rail to show more detail of the chairs for the photo.

For the main rails in a turnout, or plain track rails, the rail used as an insertion tool doesn't have to be the finished rail. You can remove it to create a chaired base if you prefer.

That won't work for the wing rails because of the knuckle bend. Once the wing rail and chairs are in place, the rail can't be slid out of them.

Because of resolution effects in the home printers, the chair plug sizes and socket sizes do vary a fraction. You might be lucky and find that some of the chairs can be pressed home using fingers. But mostly the chair jaws risk being damaged if you do that. Using the rail to press them home applies pressure to the base of the chair and works fine.

The idea is to thread the chairs on the rail, roughly position them at the correct spacings (the printed paper template helps), and locate them loosely in the sockets. Then work to and fro along the rail with a suitable implement such as a wooden spoon, gradually increasing the pressure until all the chairs are pressed fully home. You can feel a satisfying click as they get there.

It works fine, but as with most things, a bit of practice helps.

All the settings are entirely under your control in Templot, so if you prefer to make the chairs an easy fit, and apply some glue in the sockets first, that's fine. But don't overdo it, because the intention is that the sockets set the track gauge. If the chairs are too loose you would need to use track gauges in the usual way and need a fast-setting glue.

Or if you prefer a half-way house with a gentle push fit, the finished turnout can be turned upside down after assembly, and an adhesive such as penetrating cyano can be applied to the bottom of the chair plugs to seal them firmly in place.

At some stage I shall have to write all this stuff up and make some videos.

cheers,

Martin.
 
_______________
message ref: 4744
.
I have moved several posts from this topic into a separate topic:

https://85a.uk/templot/club/index.php?threads/building-3d-track.520/

I know a lot of users are getting very confused by all the different options and methods being discussed. I'm hoping that separating actual modelling and track construction topics from the topics about Templot software developments might add some clarity. Maybe. :unsure:

cheers,

Martin.
 
_______________
message ref: 4750
.
It's proving tricky to get the socket sizing right for the L1 bridge chairs where they are close together. Making them a tight push fit means the first one distorts the base a fraction and makes the second one more difficult to fit.
Hi Martin,

If it's possible to reduce the density of the timbers that might help. I've found it's possible to make them very low density.

Andy
 
_______________
message ref: 4758
Hi Martin,

If it's possible to reduce the density of the timbers that might help. I've found it's possible to make them very low density.
@AndyB

Thanks Andy.

I'm currently using a fill density of 40%. The problem is that there isn't any measurable fill between the sockets, to be changed:

L1_socket_walls.png


My thought was to make that socket wall stronger, rather than weaker? So that one plug doesn't affect the fit of the other.

I can see that reducing the fill density in the rest of the timber would allow the chairs to push apart a fraction, but that would affect the gauge accuracy.

I'm currently using the Cura 4.13 slicer above. The new Cura 5 has some new controls for wall thickness and dimensions, but I haven't tried it yet -- it's next on the list. At present that wall thickness is measurably thicker than the design size, even though the overall Y accuracy is good -- sleeper length is typically 33.98 - 34.03 mm (EM).

I've currently got the plugs with significant corner relief so that the ends of plug make contact with the socket wall only in the middle:

L1_plug_outline.png


I'm thinking of offsetting the contact points a little, so that they are not exactly on opposite sides of the wall. This might prevent one plug affecting the fit of the other, but risks twisting the chair in the socket. I may need to tighten up on the plug side clearance to avoid that.

For the L1 chairs the plug is wide enough to have 2 contact points at one end only, which would avoid any twisting and might work well. But the S1 chair plug is not wide enough to do that.

In the end I might have to accept that a gentle push fit is the better option, with some glue in the socket for long-term strength -- there is space for it in the corners. I think I may have found a suitable glue:

https://85a.uk/templot/club/index.php?threads/building-3d-track.520/post-4755

Always something to think about. :)

cheers,

Martin.
 
_______________
message ref: 4759
Hi Martin,

I see what you mean. Of course I'm printing the chairs along with the timbers using a 0.2mm nozzle which probably makes things quite a bit different.

Have you experimented with round pegs and holes at all? I'm wondering if the rail could provide the alignment. Highly non-prototypical of course although I have noticed some FB fixtures don't define the direction of the rail. They only seem to determine the gauge.

Cheers,
Andy
 
_______________
message ref: 4760
Have you experimented with round pegs and holes at all? I'm wondering if the rail could provide the alignment.
@AndyB

Hi Andy,

This has been suggested several times. I haven't actually tried it because I can't see any advantage in discarding information which Templot already knows -- the angle of the chair on the timber. And I can see several problems:

1. Using the rail for chair alignment requires a very close fit on the rail. At present the chairs can slide along the rail quite easily, so there is no stress on the chair jaws. The chairs are aligned from the socket, not the rail, and can't be inserted at the wrong angle.

2. With a round peg, how to recover the situation if a chair is accidentally pressed home at the wrong angle?

3. The edge of the chair base is very thin and fragile. The small overhang from the rectangular plug helps to protect it, and avoids any 3D printing problem from a larger flat horizontal overhang -- which would arise from a round peg.

4. Tolerancing a round object on X and Y for a close-fit is much more difficult than a rectangular object.

There is the "exactopips" alternative idea of having a pip on the timbers and a hole in the base of the chair, matching the original Exactoscale turnout kits:

https://85a.uk/templot/club/index.p...3d-printed-cnc-milled-laser-cut.229/post-2437

But that makes the chair very fragile and difficult to support in the printing process to create a dead flat underside. I may look at that one day -- the pip could be round or rectangular. It would be of no relevance to laser-cut or CNC-milled timbers.

Whereas the present design can cover 3 different timbering methods. It would have been 4 if the Cameo cutter had worked. Even more options if anyone tries other methods such as die-cutting, blanking and punching:


cheers,

Martin.
 
_______________
message ref: 4765
Hi Martin,
With rereference to the pair of L1 chairs in close proximity, I think the thickness of the fillet between the two sockets might be outside the scope of a laser cutter using 1.5mm plywood, so a possible solution would be to implement a solution similar to your proposed check rail chais, where the plug part of the chair(s) is extended sideways so that they touch and in effect replace the need for the fillet between the two chairs.
That way when you "plug" the second chair in in holds both itself and the first chair in the correct position.
Regards Steve
 
_______________
message ref: 4768
Hi Martin,

I have been thinking about this problem and my thoughts would be to combine the two plugs and chairs so that the top of the plug forms the timber fillet between the two chairs. The plug will hold the two chairs in the correct orientation. I have no idea if it's feasible.
 
_______________
message ref: 4769
.
Many thanks for the comments.

My hope was that the standard chairs, S1, S1J, L1, P, and others, would be generic -- meaning that the chairs would all be the same, and any L1 chair say, would fit in any L1 socket on the template, or in any L1 socket on any other template.

Modifying the plug dimensions where two L1 chairs are close together would wreck that, and those L1 chairs would be specific to that location on that template.

Which is obviously doable, but the whole process is getting more and more complex, and one day I'm going to have to explain all this in instructions and videos. Cue another paragraph to explain the above, and answer support questions from someone saying they have got an L1 chair which doesn't fit the socket for it.

Developing Plug Track is one thing -- and very enjoyable. Explaining it is something else and a nightmare facing me. I'm getting fearful that the whole thing will overwhelm me.

My thinking now is to replace the wall between close L1 sockets (or an L1 and S1 socket) with a loose resin-printed spacer component. This would allow the L1 chairs to be the standard generic L1 chairs, and a fiddly little spacer would be inserted between them in a single long socket during assembly. I haven't tried this yet, so I don't know how practical it would be. For thin laser-cut plywood it would be very tiny.

It has to be yet another option tickbox, because the existing design works fine for CNC-milled sockets, and probably also for FDM sockets where a looser glued fit is adopted. So another paragraph in the instructions to explain that. :(

cheers,

Martin.
 
_______________
message ref: 4770
Hi Martin,

I have been thinking about this problem and my thoughts would be to combine the two plugs and chairs so that the top of the plug forms the timber fillet between the two chairs. The plug will hold the two chairs in the correct orientation. I have no idea if it's feasible.


I think this is the best method, that leaves the L1 chairs with the same size plug as the other chairs
 
_______________
message ref: 4771
I am assuming that for the instance where you have a very close pair of L1 chairs, or a close L1 S1 combo, that you would "plug" the stock rail assembly in first, in which case would it help if there was a temporary L1 sizes peg inserted into the switch/closure rail position thus strengthening the fillet whilst the stock rail chair was inserted, then just before inserting the switch/closure rail assembly whip out the temporary peg.
 
_______________
message ref: 4772
.
Every day is a schoolday. :)

When Cura 5 was released I had just spent a lot of time fine-tuning the settings in Cura 4.13 to get the best results for the timbering bricks on my BIBO prnter. So I held off upgrading and starting again, despite the claims from Utlimaker that Cura 5 offered significant slicing improvements in the accuracy of fine detail.

I'm wishing now that I had upgraded straightaway.

Cura 4.13
index.php


Cura 5.1
L1_socket_walls_cura5.png


For the same STL file and slicing settings in Cura. There is now no infill between those problem socket walls, and a few other differences can be seen.

The printed result is very different. The printing is quit a lot cleaner with less stringing and a better top surface.

But the main difference is that the socket walls are much thinner, and the sockets are now much closer to the design size, and more consistent one to the next.

Which means that chairs which were mostly a very tight fit in the brick from from Cura 4.13 (especially the wider plugs on the S1J and L1 chairs) -- are now in the same brick from Cura 5.1 so loose in the sockets that they are falling out, and have to be sealed in place with glue.

So now I must start again a tweak the socket sizes all over again. :(

I'm wondering if I shall have to accept that a fully press-fit design isn't going to work for FDM? I wanted to eliminate gluing because it is messy and time-consuming, and makes it more difficult to make corrections and repairs.

Press-fit works just fine in CNC-milled MDF timbering. But the results from FDM printing seem to be too variable. If I'm struggling to get consistent results from my BIBO printer, I'm wondering how to set defaults in Templot which will work on all other printers?

Perhaps a gentle push fit, with some adhesive to seal the chairs in place long-term, is a more practical approach?

At least this has solved the issue with close-together L1 chairs, and similar problems I was expecting with the special crossing chairs.

There is another test brick printing on the BIBO right now, so I had better go and see how it's doing... :)

cheers,

Martin.
 
_______________
message ref: 4776
.
Every day is a schoolday. :)

When Cura 5 was released I had just spent a lot of time fine-tuning the settings in Cura 4.13 to get the best results for the timbering bricks on my BIBO prnter. So I held off upgrading and starting again, despite the claims from Utlimaker that Cura 5 offered significant slicing improvements in the accuracy of fine detail.

I'm wishing now that I had upgraded straightaway.

Cura 4.13
index.php


Cura 5.1
View attachment 4058

For the same STL file and slicing settings in Cura. There is now no infill between those problem socket walls, and a few other differences can be seen.

The printed result is very different. The printing is quit a lot cleaner with less stringing and a better top surface.

But the main difference is that the socket walls are much thinner, and the sockets are now much closer to the design size, and more consistent one to the next.

Which means that chairs which were mostly a very tight fit in the brick from from Cura 4.13 (especially the wider plugs on the S1J and L1 chairs) -- are now in the same brick from Cura 5.1 so loose in the sockets that they are falling out, and have to be sealed in place with glue.

So now I must start again a tweak the socket sizes all over again. :(

I'm wondering if I shall have to accept that a fully press-fit design isn't going to work for FDM? I wanted to eliminate gluing because it is messy and time-consuming, and makes it more difficult to make corrections and repairs.

Press-fit works just fine in CNC-milled MDF timbering. But the results from FDM printing seem to be too variable. If I'm struggling to get consistent results from my BIBO printer, I'm wondering how to set defaults in Templot which will work on all other printers?

Perhaps a gentle push fit, with some adhesive to seal the chairs in place long-term, is a more practical approach?

At least this has solved the issue with close-together L1 chairs, and similar problems I was expecting with the special crossing chairs.

There is another test brick printing on the BIBO right now, so I had better go and see how it's doing... :)

cheers,

Martin.
Hi Martin

I believe I may have mentioned just how much better Cura 5 was than 4, and that was just Beta at the time 😀

I’ve just bought a new PC (they’re so complex now that I’m no longer confident in building my own). I’ve installed Cura 5.1.something and copied the profiles across, but the key thing is that despite having a much better UI, I haven’t reinstalled PrusaSlicer!

Incidentally, this PC is Windows 11, and Templot runs just fine.

Cheers,
Paul
 
_______________
message ref: 4777
I believe I may have mentioned just how much better Cura 5 was than 4, and that was just Beta at the time 😀
@Paul Boyd

Thanks Paul. :)

Glad Templot is fine on Windows 11. Please let me know if you spot any quirks. Thanks.

Are you saying that you don't like the Prusa Slicer? I've installed it, but not made any use of it.

Another factor in all this is the different reels of polymer. I'm using eSun PLA Plus, but it's not all the same. I have 2 reels of brown, one about 6 months old and one over 2 years old. Both the same part number. The newer one prints noticeably more stringy than the old one. On the other hand the old one has gone a bit brittle with age, and tends to break in the extruder feed if I don't watch it. I assume that after heating the brittleness is removed, I can't see any difference in the strength of the finished bricks.

I also have a reel of grey, which prints much more cleanly than the brown. I've been using the brown for the timbers because it makes a better contrast with the chairs for photos.

Neither of them are particularly realistic for railway sleepers and will need painting (as will the chairs of course), so I may change to using the grey in future.

If anyone reading this would like me to send you a few sample chairs and bits of timbering base, please let me know your postal address and I will see what I can do. Click this link to contact me: https://85a.uk/templot/club/index.php?misc/contact Don't expect it by return of post because FDM printing the timbering bases is quite slow, it might take a few days if everyone asks at once. :)

cheers,

Martin.
 
_______________
message ref: 4778
@Martin Wynne

Hi Martin

Are you saying that you don't like the Prusa Slicer? I've installed it, but not made any use of it.

I do like PrusaSlicer, and I got better results with default profiles than I did with Cura 4, as well as it having (for me) a nicer UI. It also supports my 3D mouse! However, one flaw for me was that there was no hole size expansion option so getting a hole to print at a fairly precise size (e.g., for pressing in threaded inserts) was impossible. At least, I couldn't find a way to do it. Cura 5 still has that option, but has also leapfrogged PrusaSlicer for print quality in my opinion. I don't like having different programs for the same function, and it's also messy having project files and gcode with the same file extensions but produced by different software, so certainly for now I'm going to stick with Cura 5. That may of course change in the future!

I've only ever used Anycubic PLA, and so far the white, black or blue of various ages have all behaved identically, but none are more than a year old. I've not noticed any difference in performance with age, but I know some people go to quite extraordinary lengths to keep their reels in a temperature and humidity controlled environment!

Thinking about the hole size expansion option, that actually might make press fit sockets usable for FDM by adjusting the socket size to suit, with a bit of trial and error, when printing. The "expansion" can also be negative if needed. Essentially, I'm thinking along the lines of Templot producing sockets to a set size, then the user adjusting the printed size at print time. That's exactly what I do when printing my own designs.

Cheers,
Paul
 
_______________
message ref: 4779
Thinking about the hole size expansion option, that actually might make press fit sockets usable for FDM by adjusting the socket size to suit, with a bit of trial and error, when printing. The "expansion" can also be negative if needed. Essentially, I'm thinking along the lines of Templot producing sockets to a set size, then the user adjusting the printed size at print time. That's exactly what I do when printing my own designs.
@Paul Boyd

Hi Paul,

I've wondered about that. My thinking was to keep the trial and error in Templot, so that I can make it easy for users and create videos etc., as I don't know which slicer users will have with any printer they might be using, and what options it might have available. Cura can be a bit scary with so many different adjustments.

Also the hole correction applies to both X and Y dimensions, whereas I'm trying to make an easy fit on the X dimension (along the rail), and a precise press fit on the Y dimension (for track gauging).

The downside of doing it in Templot is that you have to go through the full STL export and mesh fix/repair for every iteration. Doing it in the slicer means you can use the same fixed STL for each try.

Regards storing reels of filament, keeping them dry isn't much of a problem at present! But I wonder if the age-embrittlement is related to daylight exposure, and reels need to be stored in the dark?

cheers,

Martin.
 
_______________
message ref: 4780
@Martin Wynne

Hi Martin

Also the hole correction applies to both X and Y dimensions, whereas I'm trying to make an easy fit on the X dimension (along the rail), and a precise press fit on the Y dimension (for track gauging).

Maybe I've misunderstood something! If Templot outputs a socket dimension with an easy fit on X and press fit on Y, setting a hole size expansion would simply restore that from both X and Y printing (usually) a bit small, wouldn't it? Setting the hole size correction to make Y a press fit would apply the same correction to X, still giving the easy fit. I think...

I think my engineering mind says that the "CAD" output should be to the required size, and it's up to the manufacturing process to ensure those sizes are met. All very theoretical though because I've only printed rail guides on timber bases so far, where the blind socket dimensions are not critical!

Regards storing reels of filament, keeping them dry isn't much of a problem at present! But I wonder if the age-embrittlement is related to daylight exposure, and reels need to be stored in the dark?

I do store mine in the dark, in the original boxes, but that's mainly just where they happen to be stored rather than from any consideration about ageing!

Cheers,
Paul
 
_______________
message ref: 4781
Maybe I've misunderstood something! If Templot outputs a socket dimension with an easy fit on X and press fit on Y, setting a hole size expansion would simply restore that from both X and Y printing (usually) a bit small, wouldn't it? Setting the hole size correction to make Y a press fit would apply the same correction to X, still giving the easy fit. I think...
@Paul Boyd

Hi Paul,

Yes and no, or maybe. :)

I'm not sure conventional engineering comes into it much. This is my current design:

plug_socket_fit.png


It's not feasible to make a close fit all round and expect consistent results. The X-Y resolution on the steppers is 0.05mm steps, so rounding effects mean some sockets will be 0.05mm larger in one or both directions than other sockets. In a few cases there might be a 0.1mm difference.

The sideways position of the chairs along the rail is not too critical, so I'm allowing an easy fit at the sides of the sockets even if the socket is 0.05mm narrower than the design size. Currently the design side clearance is 0.03mm on both sides. It could be a bit more, but on the shorter chairs it risks the chair twisting in the socket.

The endways position of the chairs is more critical, because it sets the track gauge. Since changing to Cura 5.1 I've set a bash fit interference of 0.03mm overlap at each end. A bash fit allows more tolerance in the components than a proper press fit. The resin is quite a bit harder than the FDM polymer which deforms to contain it. The corner relief reduces the force needed to a reasonable level, allows for radiused corners in the sockets, and provides somewhere for the deformed polymer to go. And also a space for any adhesive if desired.

I've drawn the above square-on, but of course many sockets in a timbering brick will be at an angle, so the clearances will be controlled by both X and Y axes on the printer.

My thought was that users would be adjusting the plug end fit to suit their printers, leaving the side clearance unchanged. I had in mind a built-in automated 3D printer calibration function similar to the one for 2D paper templates. With options for those who prefer a gentler push fit, or an easy fit for gluing.

But that's still a long way off, in the short term changing the slicer settings is easier than frequent trial and error trips to the online mesh fixer.

The above clearances apply to 4mm/ft scale. In theory they should stay the same in the larger scales, rather than scaling up. We shall see.

What shrinkage setting do you allow in your CAD designs? Or on the slicer? For PLA Plus on my BIBO printer I currently have 0.28% on X and 0.16% on Y. That's currently set as the defaults in Templot but again others will need an automated calibration function for that. It's difficult to know how much of that is polymer shrinkage, and how much is a resolution correction for the printer. Obviously some of it must be printer correction, otherwise the two values wouldn't differ. The slicer is set to 100% scaling.

For the Z dimension I have the shrinkage set to zero. This makes it easier to match the various thicknesses in the finished part to the layer setting in the slicer (currently 0.12mm layers). The result means that the finished part is shrunk slightly smaller on Z than the design size, but for the timbering bricks that is of no consequence.

For the resin printer I have the shrinkage set to 1.5% all round, to get the plugs to the design size and ensure the proper rail fit in the chairs.

p.s. many thanks for the contribution to the funds. :)

cheers,

Martin.
 
_______________
message ref: 4783
Hi Martin,
I would like to try resin printing the timbers WITH the chairs. I can see that I can create/export timbers only or chairs only which is great. I would like to create/export an stl of the chairs in the timbers 'without' the rails. is this possible within the current version by switching buttons on/off?

Terry
 
_______________
message ref: 4788
I would like to create/export an stl of the chairs in the timbers 'without' the rails. is this possible within the current version by switching buttons on/off?
@Terry Downes

Hi Terry,

Sure. Just export a 3-D file and don't switch off what you don't want switched off. :)

But, but, but.

The only reason to do that is if you want plain track with REA chairs and wooden keys. In which case it works just fine, you can create plain track panels with proper rail lengths and sleeper spacings, wider joint timbers, special joint chairs. Keying direction can be changed, the driven key offset can be randomised, the chair screw head angles will be randomised.

It looks great. If resin-printed. (FDM printing of chair detail would be very poor, and probably won't fit the rails.)

But as I keep saying, for pointwork this whole Plug Track thing is still extremely experimental and unfinished. You are jumping ahead of me. :(

Even if the project was finished, it is a daft idea to print chairs and timbers combined for pointwork because you can't use it to build proper pointwork. You cannot thread bent rails through the chairs without damaging the chairs. The only way to assemble such pointwork is to cut the rails where the prototype isn't cut and/or split the base into smaller pieces, such as a separate switch front section.

That's the whole reason for using separate chairs and threading them onto the rail before plugging in place.

And while that isn't finished, you will currently get a complete mess anyway, through the switch heel, V-crossing and check rails:

3d_export1.png


Because only these chairs are currently done:
sample_chairs.png


If you don't want the rails, just switch them off:

3d_export4.png


But you won't be able to thread the rails through that, because the diverging stock rail has a set bend in it at the switch front (red mark), which would damage the first two chairs as you try to pass the rail through them:


3d_export2.png


Even worse for the knuckle bend in the wing rails, and bent check rail flares -- I can't illustrate that because as I say it is unfinished.

At that stage the STL file will be extremely messy because it will include the sockets through the timbers, the chair plugs, and the chair supports:

3d_export3.png


So you need to switch them off too:

3d_export5.png


Without the sockets you probably don't want the stiffener side flanges on the timbers, and you may prefer trimmable sprues on the end of the timbers instead of webs between them:

3d_export6.png


Also without sockets the timbers don't need to be so thick, so you will probably want to change the timber sizes.

Finally you need to check that you have entered your correct shrinkage setting for your resin, otherwise the track gauge will be wrong. It is not too critical to set the exact shrinkage if you are printing only the chairs (the default in Templot is 1.5%), but if you are resin-printing chaired or socketed timbers you need to get it just right for an accurate track gauge.

Here is all the stuff to untick or set:

3d_export7.png


Clear the drop-down combos by right-clicking on them. Ignore the colours because they don't apply for STL files. The online STL mesh fix will probably fail due to the file size and timeout. In which case you can attach it here and I can fix it for you, see:

https://85a.uk/templot/club/index.php?threads/online-3d-mesh-fix-repair.519/

p.s. There is a bug in 234d which means your entered DXF/STL file name won't be honoured and you get the default file name regardless. Sorry about that, I have fixed it for the next update release. You can of course rename the file afterwards as you wish.

cheers,

Martin.
 
_______________
message ref: 4789
Hi Martin,
That works perfectly. I will play around with print/shrinkage settings and check the resultant gauge etc.

I presume the shrinkage settings for '3D for CAD' is zero so I will start with this as an initial test.

I use Formlabs PreForm to view stl files and slice for my Form3 resin printer.
Terry
templot in preform.PNG
 
_______________
message ref: 4790
I presume the shrinkage settings for '3D for CAD' is zero so I will start with this as an initial test.

Hi Terry,

Yes, that option is exported at actual size.

(I assume I have tested it at some stage, but I don't remember doing so. It would be wise to check. :) Do not regard any part of the Plug Track as finalised and tested software.)

Unless you have set some scaling elsewhere in Templot.

If exporting a DXF, you can choose inch or mm units. STL is mm only:

dxf_dialog1.png


cheers,

Martin.
 
_______________
message ref: 4791
Quick question that has probably got an answer hiding somewhere. I'm looking at at 3D printing my own chairs, possibly the sleepers too, (in small panels), using a resin printer. Are the stl files produced by Templot scaleable to a particular rail profile/code? Initially I'm thinking of code 82 S scale BH but other scales may be used in later projects.
 
_______________
message ref: 4803
Unless I am mistaken. you don't need to scale the STL files, but create your Templot templates in your required scale to start with, then templot outputs the STL files in the correct size, but as Martin has pointed out this is still experimental.
I think Martin is suggesting resin printing for the chairs, but then FDM printing for timbers/sleepers ( or milling them from MDF, or laser cut plywood)
 
_______________
message ref: 4804
Having created an S scale plain track template, I saved it to background, then invoked the export to DXF/STL etc and the following screen was presented:-
1661189909925.png


On this screen there is a panel in top righht which allows you to select a particular rail profile, but unfortuantely(for you) code 82 is not one of the preset options, so you would need to click on the [ set Custom rail] button and enter the dimensions of your code 82 rail, this then sets the (custom code) radio button.
Hopefully you can find out the necessary dimensions.

I would strongly recomend reading through the whole of this topic as it is a mine of information that would be useful for you to know.
Hope this helps,
Steve
 
_______________
message ref: 4805
.
At long last I have made a bit of progress with the chairing:

switch_render.png


switch_render1.png



These are the heel chairs for an REA C-switch:

switch_heel_chair_render.png


After several false starts I think I have got something usable.

In order to allow vertical plugging assembly, it is necessary for the switch block chairs to be in 2 separate parts. As you can see I have allowed a 2-thou gap between them as an assembly clearance, which I think will be easily lost under the paint. It can be made smaller in the light of experience.

It has been tricky to get the proper rail-fit where the jaws are skewed on the chair base, because of rounding effects in the data. But I think I'm close enough now -- there is no point in obsessing about tiny discrepancies which are an order of magnitude smaller than the resolution of the printer. The challenge is not in the printing but in keeping the STL mesh fixer happy!

The middle area of the block chairs differs quite a bit from the REA drawing, but I think it will look the part between the rails. I may come back to it later and have another go. The difficulty is to allow for the split parts, any possible timber shoving, and for overscale 00/EM/0-MF flange depths.

This will be in 236a which I'm hoping to release shortly.

n.b. If you use the option to make fine adjustments to the track gauge, that function shifts the position of the jaws on the chair base. It doesn't change the size or position of the chair base. It is intended for minor gauge corrections only. If you attempt to make a significant gauge change that way, the inner jaws on the 2-part chairs are likely to be unusable. Significant track gauge changes should be done on the template design in the usual way (gauge menu).

E and F switches should be inside keyed on some of the block chairs, but that option will have to wait for another day. It's likely to be a problem anyway in 00/EM/0-MF.

The missing soleplate is very obvious, so I think that will be next on the list. :)

em_c_switch1.jpg


cheers,

Martin.
 
_______________
message ref: 4951
.
More tick-boxes! :)

I'm writing this while I think of it. I shall never remember to mention everything again later. :(

part_chair_tickboxes.png


In 236a there are new options for the 2-part switch block chairs, to permit each half to be included or omitted from a template independently. This is necessary when exporting the chair files for 3D printing, otherwise they would merge in the printing process. Use overlapped partial templates to create a full set of half-chairs at convenient separations for the printer. It all becomes easy and obvious when you do it, it's just something else to explain. I will be making some videos.



extra_opening_chairs_dxf.png


The new extra clearance at open switch blade option needs more explanation:

extra_opening_chairs.png


The REA designs for A to D switches (i.e. the sizes most usually modelled) have 2 block slide chairs (lower picture) beyond the normal bolted slide chairs. These have no inner jaw for the moving switch rail, only a rear support jaw for the switch rail when the switch is closed against the stock rail. The switch rail (blade) is free to flex away from the support when the switch is opened.

Beyond those 2 chairs are a number of block heel chairs (upper picture). In those both the switch rail and the stock rail are keyed conventionally and fixed. It is a single chair (on the prototype) because there is not enough room between the rails for extra chair bases and screws for separate chairs.

All this is fine on the model when the switch is closed. But when the switch blade is open, along comes an over-scale 00, EM, or 0-MF wheel etc., and two things might happen:

1. the back of the wheel might rub against the open switch blade, slowing the vehicle and possibly breaking the chairs.

2. the wheel flange might clonk the inner key jaw on the first block heel chair.

To avoid any such happenings, in those gauges we need to remove the key and jaws on the first block heel chair to clear the wheel flanges, and to allow the switch blade to flex open a bit wider.

Ticking the extra clearance at open switch blade option does that. The first block heel chair is swapped to a block slide chair instead. So there will now be 3 block slide chairs and one fewer block heel chairs.

I said "might happen" because it depends on the actual wheels and back-to-back settings you are using. In P4, S7, etc., the wheels should run clear, but in 00-SF, EM, etc., you will need to do some trial and error with your models to decide what to do.

For this reason this option isn't ticked by default in 00, EM, etc. Templot has no way of knowing what wheels you are using, and it would be a shame to depart from the prototype if it's not needed. So it's up to you to remember to tick this box if necessary.

For 00-BF and "Standard 00" and similar in larger scales (i.e. flangeways wider than 3.5" scale), if the box is ticked 2 chairs will be swapped to block slide chairs instead of 1. But I'm not really expecting many users of those gauges to be interested in plug track.

Notice that the stock-rail part of the block chair is identical for both types, and interchangeable for all templates, like the standard chairs. So they can be printed in bulk and used anywhere.

But the switch-rail part of the chair is not interchangeable on different chair positions or different templates. They are not even interchangeable on the same timber for the main-side or the turnout side of the switch.

(They are so used on the prototype, but there is a small geometrical discrepancy in the REA designs resulting from that, because the switch timbers are square-on instead of equalized through the switch. Consequently the turnout-side is fractionally tight to gauge when prototype REA switches are first assembled. Soon dealt with by 24 hours running-in under traffic.)

All of which means it will be important to identify carefully which chair is which when 3D printing. It would be a good idea to print out a paper template of your timbering brick and mark it up before 3D-printing it.

cheers,

Martin.
 
_______________
message ref: 4954
.

A couple of changes in 236a:

changes_236a.png


To reduce clutter I have moved the brick settings off the trackpad into a separate dialog. Click the top 3D button to show it.



There is a new hide all option on the shove timbers dialog. This hides the timber outline for all the timbers in the template. Individual timbers can then be unhidden by clicking the hide outline button, which acts as a toggle.

Hiding a timber's outline removes the timber from the template, but leaves its chairs in place (to be captured by shoved timbers on some other template).

This differs from the omit all and omit timber buttons, which remove the entire timber from the template, including the chairs.

Do not hide a timber's outline and leave its chairs floating in space, uncaptured by any other timber. This will create a fault in the exported DXF/STL file, and possibly an error message when exporting.

cheers,

Martin.
 
_______________
message ref: 4968
Hi Martin,
Congratulations on achieving your red letter day!

On the subject of 2-part chairs, you mentioned
But the switch-rail part of the chair is not interchangeable on different chair positions or different templates.
By "different templates", would you class two separate B7 templates that had different curvatues as different?
In other words could we print a "standard" set of 2-part chairs for a left-handed B7 template, and use them for any left-handed B7 template of the same persuasion?
Or do we have to print a unique set of the switch-rail chair parts for every left-handed B7 say?

V-Crossing options
If I look at a B7 template (for example), all the chairs on timber X8 are portrayed as S1 chairs.
I have noticed that, whether in OO-SF or P4 say, with Regular crossing set the S1 chairs on timber X8 on the crossing slightly overlap, this overlap disappers if Curviform crossing is set.
I know this is all still experimental, but will it be possible to set the TS side S1 chair (i think it is chair 3) to be an L1 chair if using a Regular crossing?
 
_______________
message ref: 4984
By "different templates", would you class two separate B7 templates that had different curvatues as different?
In other words could we print a "standard" set of 2-part chairs for a left-handed B7 template, and use them for any left-handed B7 template of the same persuasion?
Or do we have to print a unique set of the switch-rail chair parts for every left-handed B7 say?
@Steve_Cornford

Hi Steve,

All the chairs for a Left-Hand B-switch will be identical with any other Left-Hand B-switch, curved or straight, and even on say a B-8 instead of a B-6, provided you haven't done any timber shoving on the switch timbers (unless you have timber-shoved the other B-switch identically).

But they won't be usable on a Right-Hand B-switch. Even though they will appear at first sight to be almost identical.

(And on the prototype they are identical -- therein lies a geometrical conundrum in the original REA data, related to the position of the set in the TS stock rail and the squrare-on timbering. As drawn, REA switches are fractionally under-gauge through the turnout-side of a switch. Templot templates are not.)

It will be possible to do some mixing and matching of the switch chairs if you know what you are doing and don't get in a muddle. This has the obvious advantage that if you do lose a chair in the printing, or damage it on assembly, you can pinch a replacement from a different template, which you might have created as a spare set of chairs just in case of such a need arising. There is a lot of practical learning still to be done on all this, but we can't really get into that until I have made a bit more progress with some more of the special chairs.

V-Crossing options
If I look at a B7 template (for example), all the chairs on timber X8 are portrayed as S1 chairs.
I have noticed that, whether in OO-SF or P4 say, with Regular crossing set the S1 chairs on timber X8 on the crossing slightly overlap, this overlap disappers if Curviform crossing is set.
I know this is all still experimental, but will it be possible to set the TS side S1 chair (i think it is chair 3) to be an L1 chair if using a Regular crossing?

Currently everything beyond the wing rail front joint at X1-Z is a complete mess and you shouldn't try to make any sense of it. I haven't so much as looked at it yet. When I have got it done, Templot will automatically swap S1 chairs to L1 where they conflict (as it already does for the switch heel).

But we do have a bit more progress:
soleplate_fdm.png


A soleplate on the S1 timber. :)

cheers,

Martin.
 
_______________
message ref: 4985
Great progress being made Martin. Fair play to you. I am inching closer to being able to dip my toe in all this experimental wizardry you are producing as I will be getting a laser cutter capable of cutting ply timbers for my birthday next month. Then I just need a second hand resin printer and I can try my hand at a bit of plain 21mm track at last!
 
_______________
message ref: 5003
Hi Martin,
Using 236a I managed to produce a raft containing a B7 switch set of chairs, so thank you for this update.
I first created a background shape (for the raft) that was slightly smaller than an Elegoo Mars 2 Pro build plate split in half lengthwise.
Then created a P4, B7 turnout to sit on this with an approach length of zero and reduced the length so that it fitted within the "raft" shape.
I then used your new option for 2-part chairs to switch off both switch rail parts of the chairs. I then saved to background.
Then on the control template, I used "shove timbers" to omit all the timbers apart from the timbers containing the 2-part chairs.
Then I used your new option for 2-part chairs to tick the crossing rail chairs option & untick the stock rail chairs option.
So far so good. The plan was to then create two more partial templates, one for the Main Line crossing chairs and one for the Turnout Side crossing chairs.
Starting from this position:-
1665309297883.png

I sected real> rails> omit rails & then tried unticking turnot-road crossing rail to remove the turnout road partial chairs and the rail was omitted but the chairs remained:-
1665309544299.png

So then I deduced that the partial chairs actually belonged to the stock rail, so I ticked turnout-road crossing rail, but then unticked main-road stock rail and voila acheived the desired result:-
1665309687090.png

Is this a feature or a bug, and if a bug, then I found the work-around?

I then moved the template up so that when saved there would be separation between the two halves of the partial chairs.
Then repeated but the opposite way round for the main-road crossing rail & saved to achieve my desired 3 partial templates.

I have attcahed the box file, shape file & the fixed .STL
I have not tried printing the result yet as my Mars 2 Pro is still in its box unfortunately, still trying to work out where I can locate it & keep senior management happy!

Regards Steve


I
 

Attachments

  • em2pro_p4_lh_b_switch_chair_raft.box
    29 KB · Views: 72
  • EM2Pro half raft.bgs3
    996 bytes · Views: 68
  • em2pro_p4_lh_b_switch_chair_raft_fixed.stl
    5.3 MB · Views: 62
_______________
message ref: 5006
Back
Top