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

[James Walters]

@James Walters

Hi James,

Looking at that again it is fine for single turnouts, but is going to be a minefield for complex formations. Even ordinary double track is going to fall to bits:

View attachment 5164


I think we possibly need a range of options for this function (more tick-boxes!):

View attachment 5165

The question is -- which one should be the default setting for cutting a kerf line? Or something else? Do we need an option to change the setting individually for each end of each timber?

Also, if it is undesirable for loose bits to remain on the laser bed, presumably something similar will be needed for the sockets? Without affecting the bash fit.

Is it reasonable to assume that everyone with a laser-cutter will have some CAD software and an ability to use it, to edit the Templot DXFs? It's my firm intention that it should be possible to make plug track using the files from Templot without needing any CAD software or skills. I'm confident I can achieve that for FDM printing, and hopefully also for CNC milling.

But I'm beginning to wonder for laser-cutting. The output at present can put a kerf-adjusted line around every timber and socket. But if it is laser-cut as-is, it's going to fall apart on the cutter bed and how will you know which timber goes where? Is there any backing sheet which the plywood can be attached to, which is unaffected by the laser beam?

I know some users of plywood bases put a linking web somewhere between the timbers. I put the sprues option on the timber ends so that they could be easily trimmed off after tracklaying without distorting the timbers. But they have similar problems to the above in complex formations. Some users put a web under the rails, where it is very difficult to remove after tracklaying without distorting the track, but if not removed it makes ballasting very difficult and destroys the realism of bullhead track -- no webs between the timbers here:


View attachment 5166

A noticeable detail in that photo is the clean square end on the check rail. One of the advantages of the loose-jaws option is that you can have such neat flat rail ends, because the chairs never get threaded onto the rail.

cheers,

Martin.

Hi Martin,

To answer your questions in reverse order:

3. It is reasonable to assume most (if not all) laser users will have access to CAD software. That is to say the laser driver software (I use Lightburn) has some CAD functionality, and it is not difficult to adjust/delect vectors within that environment. I suspect that most (if not all) will also be using CAD software such as AutoCAD or Fusion360 etc. which they are using to create their own files. In which case they are probably more likely to use that to adjust the .dxf file Templot produces.

2. I hadn't given thought to the bits dropping out of the chair sockets when we spoke before, but these fiddlesome bits are the most likely to drop into honeycomb bed holes. But to attach them with nibs would likely require a modification to the chair plugs, as the remaining 'half-nib' which is left behind after breaking-out the waste will be difficult irksome to sand away.
I hope that makes sense, here's how I suggest the chairs might need adapting:

You will see I've added a clearance groove up the end of the plug.

I think that to modify the chairs to suit perhaps a minority of users with as laser might be to the detriment of the FDM/Milled plug track and if I were you I'd discount chair modification.


However, having just written that, and produced a drawing of a modified chair, I'm now having second thoughts, the half-nib would almost certainly compress - it might even help the bash-fit quality of the plug.
So, after all that I'd say a single "nib" in a chair socket would be a good way to go. Preferably on the 'long side' to avoid compressed nib residue affecting the gauge.

1. Given my answer (3) above, I'd say option 1 of the five you present would be the most useful and the one which I personally would like to see as a default, with the option to switch it off to leave an option 3 situation. If there was the facility to adjust the length of the upper vectors which cross I should think all bases would be ticked. It would be no hassle whatsoever to make further adjustments to these in CAD post Templot if someone chose to do so.
I'd be happy to do some testing, I can recreate what you have proposed above in CAD or can cut something from a test file which you produce.


With regard to the timbering falling apart when lifted from the laser bed, there are a few simple solutions, I'd suggest that either of the following would do the trick.
A. Draw a rectangle around the timbering, so that in-effect becomes the fret, The additional nibs we are discussing here keep the whole thing together, and importantly allow the fret to be snapped away either after the timbers have been glued down or the turnout built. The rectangle could be produced in Templot, or drawn in the laser software mentioned above (3).

B. Stick a line of making tape over the plywood across the timbering after it has been laser cut to keep all the bits together.

As I mentioned before, I think Templot currently has all that is required for laser cutting to be useful. And that most people doing the laser cutting themselves will have the skills to achieve what they required with everything as it is. Personally though, I'd find the nib function really useful and I'm sure others would too.

Finally, I couldn't agree more regarding sprues between the sleepers/timbers.

I hope the above is clear/useful,

James

 

[Martin Wynne]

If there was the facility to adjust the length of the upper vectors which cross I should think all bases would be ticked. It would be no hassle whatsoever to make further adjustments to these in CAD post Templot if someone chose to do so.

@James Walters

Thanks James for the detailed reply.

A setting to adjust the length of those vectors is already there:

For the sockets, I think nibs in the corners might be the best option, because there is already a setting to increase the corner relief on the plugs (for CNC-milled sockets with radiused corners):

Would 4 nibs be too many to push out, or do you just make them smaller and weaker?

I'm getting a bit nervous about talk of most users editing the Templot DXFs. I know folks have been doing that with the DXFs for 20 years, but they haven't been mentioning Templot plug track in the same breath. If lots of technical posts appear on the forums and YouTube about CAD editing of Templot Plug Track files, I just know it is going to put a lot of potential users off the whole idea. I'm getting frequent emails along the lines of "I would love to try building the plug track, but I'm a complete computer numpty". I keep reassuring them that if they are capable of printing paper templates from Templot on a 2D printer, they will be able to make plug track on a 3D printer.

cheers,

Martin.

 

[James Walters]

I think those adjustments are all that is needed, when the time comes I'll have a play.


I understand your concerns Martin re laser cutting, and I wouldn't want to add to any confusion. Hence why I've not previously mentioned my experiments on the web.
When I mentioned most users, I was referring to most using a laser cutter. These are likely to be the people who fiddle with the .dxf's - who on the whole are less likely to be computer numpty's - but more likely to post about about their experiments.

I don't know what to suggest really, as it's my guess that those who feel confident to experiment with the .dxf's will always do so. In the same way that someone with confidence in 'traditional' techniques might have cut-up and curved a paper template in the way things were done pre Templot.

I totally agree, and would defend that the plug track concept can be produced successfully with a 3d printer with no more computer skill than using an office printer. That is the joy and genius of the system which you have created.

Best,

James

 

[Phil O]

Laser cutting is not an area that I have much knowledge of, but I believe that it's possible to partially cut/burn away some of the timber, like a half-etched tab, when etching brass or nickel silver. If that's indeed possible, would that help in creating a gap under the rail, when using under the rail webs.

 

[Martin Wynne]

Laser cutting is not an area that I have much knowledge of, but I believe that it's possible to partially cut/burn away some of the timber, like a half-etched tab, when etching brass or nickel silver. If that's indeed possible, would that help in creating a gap under the rail, when using under the rail webs.

@Phil O

Hi Phil,

I think that's done by changing the colour in the DXF file. I'm not too clear how you burn away an area, rather than the normal process of cutting a thin line. Presumably an area 1mm wide would need 5 lines 0.2mm wide, at half-power. Or maybe the beam-width (focus) can be adjusted while cutting. I doubt the depth can be set with much precision, although it is presumably repeatable on the same material with the same settings. Over to James.

What I do know is that adding such tabs in Templot along the existing rail lines is out of the question, unless you can find me an extra year or two of coding time. Such half-tabs would need to be created separately, although they could be placed under the rails manually if desired.

Anyone who is using under-the-rail tabs at present is editing the DXF file in CAD to achieve that.

cheers,

Martin.

 

[Martin Wynne]

@James Walters

Hi James,

I should perhaps have mentioned earlier that if all you want to do in the CAD or laser software is to add a few extra cut lines, you can do that quite easily in Templot, without needing any CAD.

Then just draw lines with the mouse, in the background shapes:

The result in the DXF is:

A bit tedious to do, but not difficult. It would make it possible to prepare for laser-cutting a complex track formation without needing any other software.

It's much better for me to answer questions with "click this, set that..." rather than "get software X and learn how to use it".

(Note that this works only in the 2-D exports -- in 3-D exports the background shapes are used for other functions.)

cheers,

Martin.

 

[Steve_Cornford]

When I received my experimental laser cut ply sheets from the lasercutting bureau, one side was covered in masking tape and all the "chads" from the sockets were still present, and all the timbers/sleepers were still present.
If the "chads" did fall into the honeycomb bed would it matter?
Could you not just vacuum them up, or use some other method to remove them?
I just sent the .DXF files as output by templot to them, having first set the colours of the lines using the Templot settings.
Sockets were set to a coulour that meant "inner" cut, Timber outlines to the coulour that meant "outer" cut, and the Timber numberings to the coulour that meant engrave.
Next time I would not bother with the Timber numberings as I would just use a printed copy of the same template to identify the timbers.

I used https://modelshop.co.uk/Static/WorkShop/Laser-Cutting-And-Engraving

ps James, I have ordered your ballast vacuum device to help in my experiments

 

[James Walters]

Martin,
Thank you for the instructions above - all very useful.

Steve, the "chads" issue is not a big problem, but they can become wedged-in and get missed by the vacuum cleaner, especially where the bed frame creates blind pockets. Over time they become truly carbonised which whilst not a big problem it does tend to lead to increased staining of the underside of parts.
A honeycomb bed can be tricky to keep clean at the best of times so anything which reduces the need for a thorough wash and brush-up gets an upvote from me.

Thank you for your order. I'm afraid I missed the post this morning but it's ready to go and will be with the post office as they open on Monday.

 

[Martin Wynne]

When I received my experimental laser cut ply sheets from the lasercutting bureau, one side was covered in masking tape and all the "chads" from the sockets were still present, and all the timbers/sleepers were still present.
If the "chads" did fall into the honeycomb bed would it matter?
Could you not just vacuum them up, or use some other method to remove them?
I just sent the .DXF files as output by templot to them, having first set the colours of the lines using the Templot settings.
Sockets were set to a coulour that meant "inner" cut, Timber outlines to the coulour that meant "outer" cut, and the Timber numberings to the coulour that meant engrave.
Next time I would not bother with the Timber numberings as I would just use a printed copy of the same template to identify the timbers.

I used https://modelshop.co.uk/Static/WorkShop/Laser-Cutting-And-Engraving

ps James, I have ordered your ballast vacuum device to help in my experiments

@Steve_Cornford @James Walters

Hi Steve,

You kindly sent me some samples of your commercially laser-cut plywood sheets some while ago. I was unable to do anything with it because you had reduced the socket width to 1.7mm, so my chairs with 2.0mm plugs wouldn't fit.

But I did notice a few things:

1. the underside of the sheets was covered in masking tape, and no loose parts were missing.

2. the masking tape showed no signs of burning, so must have been applied after laser-cutting.

3. removing the masking tape caused the loose parts to fall out freely.

4. the cut lines on the underside were noticeably finer than the cut lines on the top.

5. the result of that was that the timbers were slightly wedge-shaped, being around 4.04mm wide on the underside, and 3.97mm wide on top. (design size is 4.0mm)

There are 2 possible explanations I can think of:

a. the laser lens was focused on the bed of the machine, not the top surface of the plywood; the bed of the machine is a flat heatproof surface; after cutting, the top surface was covered in masking tape, or some other sticky surface, or a vacuum lift plate of some kind to retain all the loose parts; the cut sheet was then turned over; masking tape was then applied over the underside of the sheet; finally whatever was on the top surface was removed.

or b. the sheet was cut twice, once from each side mirrored; the laser lens was focused on the surface of the plywood; after half-cutting the underside it was covered in masking tape; the sheet was turned over; accurately re-registered on the bed, and the top surface was cut again, without adjusting the focus for the thickness of the masking tape.

I have no idea how plausible or likely any of that is, or what other clever tricks or devices are used. What I do know, as with everything else in plug track, is that the only way I can know anything for sure is to get one of these machines myself.

At present I don't have anywhere I can safely use such a machine, so getting one is going to have to wait until I have made a safe space for it. It certainly can't go in the 3D printing den, with gallons of IPA sloshing about!

In the meantime I must get on with the chairs. I'm feeling some urgency to get a finished FDM turnout done and visible to the world, before the whole thing unravels in all directions!

cheers,

Martin.

 

[James Walters]

I think your first explanation is the most likely.
The tapered 'kerf' is typical with laser cutting and can be minimised (but not entirely eliminated) by focussing on the centreline of the material. Although that does depend upon the type of laser, and the speed at which it was cut.

I'm really looking forward to seeing a completed turnout Martin, so please ignore all of this distraction for now.

In the meantime I'll laser a piece of plain track using just the settings in Templot, to prove it can be made to work with the plug track chairs and publish the results and laser settings 'recipe' on here for others to try. Once its been proven, we could come back to nice-to-have laser tick boxes etc when the chairs are all done.

My best to all,

James

 

[Martin Wynne]

I think maybe I was over-thinking the chair fixing problem. The chairs don't need any adhesive to fix them initially, they are a firm bash-fit in the FDM timbering base. We just need a sealant of some sort to penetrate around the tapered plug to prevent them coming loose months later.

Yesterday I found an unopened Humbrol tinlet of white enamel paint which must be at least 25 years old. It had completely settled in the tin, so I stirred it clockwise for a fortnight, and then anticlockwise for a week, until it looked usable again. I found the loosest chairs of those I had been testing -- almost loose enough to fall out of the timber if turned upside down. I added a drop of the white enamel using a cocktail stick into the underside of the sockets around the plugs, and left it for 6 months to see what happened. Well actually I left it overnight in a warm place. The result this morning is that I can't budge them with any normal level of force. This is how the tapered plugs fit the sockets in the timbers:

One week on and I can report that the chairs are now solidly fixed in the timbers. I can't push them out by hand. It needs a hammer and punch to push them out from below.

So it seems Humbrol enamel paint is the sealant of choice for resin chairs in FDM timbers. Apply with a cocktail stick into the underside of the sockets after fitting the chairs. One little tinlet would probably do the whole layout.

cheers,

Martin.

 

[Martin Wynne]

.
Just adding this while I'm working on it -- it's very unlikely I shall ever remember to mention it again.

• If you create plain track with a continuous check rail, the first check-end chair from CTRL-0 (flare in) will be created, and all the middle check rail chairs. However, no final check-end chair (flare out) will ever be created.

(That doesn't actually matter, because continuous check rails on plain track are added by mirroring the template somewhere in the middle. Such templates are created as approach track on shortened turnout templates.)

Martin.

 

[James Walters]

I have done some experiments this morning using the 2d .dxf export from Templot, with the specific aim of testing the 'kerf' settings to attempt to establish a suitable default when laser cutting plywood for the plug chairs. I have (mostly) avoided using any other CAD input as so tried to replicate how someone with no CAD experience might approach the job.
The chairs were straight out of Templot, printed on an Elegoo Saturn with standard Grey resin on default settings.

I'm using 3.2mm plywood as that was to hand.

Here are the results:

The top loose sleepers were cut with the default 0.2mm kerf using the file straight from Templot. Because they weren't fixed into a fret, they tended to move about whilst cutting as so neither the sockets or timber outlines were cut accurately. I tried cutting the sockets first and it helped, but even so the results were not ideal.

Next up, I took the same file, bunched the pieces together and made a small 'nib' between them to stop them moving about.
I did this in the laser software (not CAD per se.). To do this I drew a .5mm radius circle between the bunched sleepers, deleted the line bridging the circle, and then deleted the circle to leave the nib. I also removed the overlapping lines so that the long edges didn't get cut twice once the timbers had been bunched. I added the kerf legend to the sleepers and cut again (with a .2mm kerf). To reiterate, this was done with the laser software not a dedicated CAD program.
The top-left sleeper plate was the result. The chairs were a bit loose and fell out when I turned the piece upside down.

Next I tried again with a .4mm kerf. This time the fit was good but not tight - I'd say ideal with a little glue.
Next up, I tried .5mm kerf (by increasing the kerf I was closing the socket slightly). This time the chairs were a bash fit and it took quite some force to squeeze then in. A little more than I was comfortable with - although I didn't break any chairs.

Next I tried a .45mm kerf - this time the fit was excellent. They pushed in firmly, but could be removed without damage if care was taken. The outer width of the sleeper was 3.39mm, I couldn't measure the length as I only have a 25mm micrometer to hand. A little over width, but not that anyone would know.

So what has been the learning?
Well, theory would seem to suggest that .2mm kerf ought to be ideal. But variables such as cleanliness of the laser, exact focussing, etc. all come into play. What has been proved is that it is completely possible to produce the data directly from Templot, and that the kerf adjustment is all that is required to produce repeatable results. There is no need for CAD.

I'm going to produce a test .dxf with different incremental 'kerf' offsets ranging from 0 to 0.6mm in 0.1mm increments. I'll post it on here.
I'd suggest all that one would need to do is use it as a test cut before each cutting session/material - try fitting the chairs and see which works best. Martin, if such a test piece were available within Templot then it would save others experimenting such as I've done.

Secondly, it has confirmed my suspicion that some form of fretting is essential. Not just from the perspective of retaining the pieces from movement during cutting, but also in terms of accurate alignment when they are glued down. If 0.05mm of adjustment is the difference between good and optimum fitting tolerances, then I'd suggest that sticking down individual timbers to that level of accuracy could be potentially difficult.
Of course the 3d printed track bases avoid that worry and are good to go as-is.

I hope that is of some use.

This system is excellent. Martin you are a Genius.

Best,

James

p.s. Yes I know that the chairs on the 0.4 and .5mm versions above are the wrong way round.

 

[Martin Wynne]

Next up, I tried .5mm kerf (by increasing the kerf I was closing the socket slightly). This time the chairs were a bash fit and it took quite some force to squeeze then in. A little more than I was comfortable with - although I didn't break any chairs.

Next I tried a .45mm kerf - this time the fit was excellent. They pushed in firmly, but could be removed without damage if care was taken.

@James Walters

Hi James,

Many thanks for taking the trouble to do all that. The results look great.

Can you clarify what you mean by "some force to squeeze them in"? For bash-fitting in FDM timbers I use a pin hammer. On the loose-jaw chairs, with the aid of a punch made by snipping the ends off a cocktail stick. For the solid chairs by inserting the rail, and a sharp tap on a hardwood block placed on the rail top. I would imagine similar methods for plywood, but maybe not?

For the plain track sleepers the socket is 2mm wide in sleepers 3.3mm wide. With the result that there is only a weak 0.65mm of ply in the side wall of the socket.

For the 4mm wide timbers there is a stronger 1mm side wall on the sockets (except for L1 bridge chairs).

Do you suspect that the side wall is flexing in the sleepers, making the fit seem easier than it otherwise would be? I'm wondering if the optimum kerf width might differ for sockets in sleepers and in timbers? While remaining the same for the timber outlines? I can see 3 kerf settings being needed instead of just one -- for timber outlines, timber sockets, sleeper sockets (and L1 chairs).

If the result of a firm bash-fit is that the side wall of the socket stretches, this would affect the track gauge. There is a gauge-tweaking adjustment in the exported socket positions, but for consistent gauge results in say P4 we may need to accept a looser fit than a firm bash-fit.

Thanks again,

Martin.

 

[James Walters]

HI Martin,

I too used a pin hammer to bash the chairs in, gently of course with a bit of shaped timber given the lack of cocktail sticks here.
In my experiments I found the .45mm kerf gave a nice feel when inserting the chairs. .5mm felt too tight. Neither has flexed the sockets - I've checked (as best one can) with a micrometer. Either would work fine.

Just tried some timbers in-between jobs here, but made a mess of it. The outsides were fine but the chair sockets too loose. I must have done something wrong in my haste. I will investigate later when I have more time.

I will check the gauge of the examples above this evening when I have access to my gauges.

Best,

James

 

[James Walters]

Gauged Perfectly.

 

[Steve_Cornford]

Hi James,
I am puzzled as to why 0.5mm kerf gave a tighte fit than 0.45mm kerf. Surely a wider kerf wiould give a looser fit?
Steve

 

[Martin Wynne]

@Steve_Cornford

Hi Steve,

The Templot setting doesn't change the width of the cut kerf -- that's a feature of the machine laser.

The Templot setting controls how far Templot shifts the cut line from the true line. Shifting it further into the socket moves the laser into the socket, which reduces the size of the socket.

cheers,

Martin.

 

[James Walters]

Agreed Martin,

Steve, where I have mentioned the 'kerf' above I have been referring to the adjustment of the 2d kerf width in the .dxf output settings of Templot. My apologies if that has caused any confusion. The actual kerf is of course a laser factor.
Perhaps that setting ought to be renamed to 2d Cutter Kerf Off-Set or Kerf Compensation? Just a thought.

James

 

[Steve_Cornford]

Thank you Martin, I understand now.
Come to tnink of it, I produced the .DXF files for the bureau with a vsersion of Templot prior to your adding the laser kerf options.
I Was advised by the bureau to aim for a minimum distance between cuts of 0.8mm, so for 3.3mm sleepers that meant reducing the socket width from 2mm to 1.7mm.
Looks as though James has managed to get down to 0.65mm between cuts, thus maintaining a socket width of 2mm.

James,
Are you able to laser cut 3mm cork sheet on your machine?

Steve

 

[James Walters]

Hi Steve, Yes 3mm cork no problem at all. I've lasered and engraved all of my cork with the timbering. I'm about start another round of trackwork and my thinking is that I'll bond the cork to the plywood trackbed first and then cut and engrave it in one go. Holding rolled cork flat can be a pain and if it lifts even just slightly then it can cause things to go south.
I've just been pondering whether or not to incorporate a sound deadening layer into the mix or not bother.

James

 

[Martin Wynne]

Perhaps that setting ought to be renamed to 2d Cutter Kerf Off-Set or Kerf Compensation?

Hi James,

Well yes and no. In any event I can't fit all that on the button!

The intention is that you know how wide your machine cuts, and enter that dimension. So calling it the cutter kerf width seems reasonable:

Also if it is called the offset instead, it will need to be halved from the current setting.

p.s. this setting isn't used only for laser-cutting, it is also needed for CNC-milling with simple CNC engravers.

cheers,

Martin.

 

[Martin Wynne]

I've just been pondering whether or not to incorporate a sound deadening layer into the mix or not bother.

@James Walters

Hi James,

You might find this topic of interest:

https://85a.uk/templot/club/index.php?threads/tip-of-the-day-trackbed-edges.561/

https://85a.uk/templot/club/index.php?threads/what-thickness-of-trackbed-do-you-use.314/post-5672

cheers,

Martin.

 

[Martin Wynne]

.
I mentioned that the check rail end chairs are going to be a bit of a minefield.

On the prototype it's easy, there is just one size of check-end chair, left and right.

On the model it gets more complicated. We have to allow for differing flangeway gaps for 00/EM etc. In addition the check rail adjustments function allows for changes to the flare length and end gap. Plus the overall length of the check rail can be changed, so the end chair might be on a different timber, or spread across multiple chairs on several timbers.

Here's a scruffy video clip showing what happens to the check rail end chairs as the check rail end is adjusted:

https://flashbackconnect.com/Default.aspx?id=M2dAKtajlbwh5aWMLB8MMg2

As you can see, the check-end chairs will be for a specific check rail and specific timber on a specific template, and not interchangeable with any others. It will be important to make a note of which one goes where.

cheers,

Martin.

 

[James Walters]

@James Walters

Hi James,

You might find this topic of interest:

https://85a.uk/templot/club/index.php?threads/tip-of-the-day-trackbed-edges.561/

https://85a.uk/templot/club/index.php?threads/what-thickness-of-trackbed-do-you-use.314/post-5672

cheers,

Mart

.
I mentioned that the check rail end chairs are going to be a bit of a minefield.

On the prototype it's easy, there is just one size of check-end chair, left and right.

On the model it gets more complicated. We have to allow for differing flangeway gaps for 00/EM etc. In addition the check rail adjustments function allows for changes to the flare length and end gap. Plus the overall length of the check rail can be changed, so the end chair might be on a different timber, or spread across multiple chairs on several timbers.

Here's a scruffy video clip showing what happens to the check rail end chairs as the check rail end is adjusted:

https://flashbackconnect.com/Default.aspx?id=M2dAKtajlbwh5aWMLB8MMg2

As you can see, the check-end chairs will be for a specific check rail and specific timber on a specific template, and not interchangeable with any others. It will be important to make a note of which one goes where.

cheers,

Martin.

This is astonishing. I can't begin to imagine how complex the programming task is, but the result is superb. Very exciting.

 

[Steve_Cornford]

Martin,
Thanks for creating that scruff video.
Easily worth a thousand words.
Steve

 

[Steve_Cornford]

Hi James,
With reference to cork sheet, I was thinking of flat cork sheet rather than the stuff that comes in rolls.
If you bond the cork to plywod first how can you remove sections of it once cut?
I plan to experiment with using a 3mm cork trackbed with just the sockets cut out, then stick 1.6mm ply sleepers/timbers to the cork using "locator" chairs which get removed then slide real chairs onto rail and plug in to the correctly located timbers.
Or in fact I could try using loose jawed chais instead of "locator" chairs and leave them in situ.
Lots of things to experiment with, before I throw in the towel and go the FDM route!
Steve

ps vacuum ballast parcel arrived today , thank you

 

[Steve_Cornford]

Having exported a 2D .DXF file, I use an open source program called Inkscape to view it and check out the various "layers" that are output in the file.
Martin has given each "layer" a useful name.
Inkscape is also useful for creating 2D drawings for sending to Photo Etch suppliers, or for that matter for printing onto paper & sticking to card for making base layers for buildings etc.

 

[Phil O]

Inkscape, is also useful for those with cutting machines, such as the Silhouette Cameo and portrait, also the Cricut, other machines are probably available. They will usually cut Plasticard etc. Up to 10 thou and scribe accurately, for cutting with a scalpel, upto about 40 thou.

 

[Terry Downes]

I have done some experiments this morning using the 2d .dxf export from Templot, with the specific aim of testing the 'kerf' settings to attempt to establish a suitable default when laser cutting plywood for the plug chairs. I have (mostly) avoided using any other CAD input as so tried to replicate how someone with no CAD experience might approach the job.
The chairs were straight out of Templot, printed on an Elegoo Saturn with standard Grey resin on default settings.

I'm using 3.2mm plywood as that was to hand.

Here are the results:

View attachment 5196
The top loose sleepers were cut with the default 0.2mm kerf using the file straight from Templot. Because they weren't fixed into a fret, they tended to move about whilst cutting as so neither the sockets or timber outlines were cut accurately. I tried cutting the sockets first and it helped, but even so the results were not ideal.

Next up, I took the same file, bunched the pieces together and made a small 'nib' between them to stop them moving about.
I did this in the laser software (not CAD per se.). To do this I drew a .5mm radius circle between the bunched sleepers, deleted the line bridging the circle, and then deleted the circle to leave the nib. I also removed the overlapping lines so that the long edges didn't get cut twice once the timbers had been bunched. I added the kerf legend to the sleepers and cut again (with a .2mm kerf). To reiterate, this was done with the laser software not a dedicated CAD program.
The top-left sleeper plate was the result. The chairs were a bit loose and fell out when I turned the piece upside down.

Next I tried again with a .4mm kerf. This time the fit was good but not tight - I'd say ideal with a little glue.
Next up, I tried .5mm kerf (by increasing the kerf I was closing the socket slightly). This time the chairs were a bash fit and it took quite some force to squeeze then in. A little more than I was comfortable with - although I didn't break any chairs.

Next I tried a .45mm kerf - this time the fit was excellent. They pushed in firmly, but could be removed without damage if care was taken. The outer width of the sleeper was 3.39mm, I couldn't measure the length as I only have a 25mm micrometer to hand. A little over width, but not that anyone would know.

So what has been the learning?
Well, theory would seem to suggest that .2mm kerf ought to be ideal. But variables such as cleanliness of the laser, exact focussing, etc. all come into play. What has been proved is that it is completely possible to produce the data directly from Templot, and that the kerf adjustment is all that is required to produce repeatable results. There is no need for CAD.

I'm going to produce a test .dxf with different incremental 'kerf' offsets ranging from 0 to 0.6mm in 0.1mm increments. I'll post it on here.
I'd suggest all that one would need to do is use it as a test cut before each cutting session/material - try fitting the chairs and see which works best. Martin, if such a test piece were available within Templot then it would save others experimenting such as I've done.

Secondly, it has confirmed my suspicion that some form of fretting is essential. Not just from the perspective of retaining the pieces from movement during cutting, but also in terms of accurate alignment when they are glued down. If 0.05mm of adjustment is the difference between good and optimum fitting tolerances, then I'd suggest that sticking down individual timbers to that level of accuracy could be potentially difficult.
Of course the 3d printed track bases avoid that worry and are good to go as-is.

I hope that is of some use.

This system is excellent. Martin you are a Genius.

Best,

James

p.s. Yes I know that the chairs on the 0.4 and .5mm versions above are the wrong way round.

View attachment 5197

Hi James, I previously tried laser cutting timbers using a

Gauged Perfectly. View attachment 5200

Hi James, may I ask what Laser machine You are using? I have a Omtech 40K and my experiments have not been as successful as yours seem to be but, these discussions have urged me to have another go. I assume all of these photos are 4mm scale?

 

[Martin Wynne]

.
While I think of it.

On the prototype the the rail is dropped into the chair onto the rail seat, and then inserting the key pushes it across under the grip on the inner jaw.

If you opt to add the non-prototypical filler chunk below the key,

the chunk is part of the rail seat, not the key:

This means that if you opt for the loose outer jaws, it will not be possible to drop the rail straight in vertically in the prototype manner.

It would be necessary to wiggle the rail into place between the filler chunk and the inner jaw (or slide the chair on the rail, from which it could easily fall off).

Which means the filler chunk option is not really compatible with the loose jaws. It would be better to leave the filler chunk switched off (the default) when using the loose jaws option.

In all my experiments I have not found any reason to use the filler chunk, and I'm wondering whether to retain the option? I added it because I thought the key might need some support in the larger scales for 3D printing. It's not prototypical, although it seems to be present in injection-mould slide-on chairs (Exactoscale, C&L, etc.) and also in the Finetrax kits.

Notice that without the filler chunk, nothing engages against the rail foot on the outer side of the rail. The rail is located only by means of the key. That's the way the prototype works, and it seems to be working well in plug track too.

cheers,

Martin.

 

[Terry Downes]

.
While I think of it.

On the prototype the the rail is dropped into the chair onto the rail seat, and then inserting the key pushes it across under the grip on the inner jaw.

If you opt to add the non-prototypical filler chunk below the key,

View attachment 5206

the chunk is part of the rail seat, not the key:

View attachment 5207

This means that if you opt for the loose outer jaws, it will not be possible to drop the rail straight in vertically in the prototype manner.

It would be necessary to wiggle the rail into place between the filler chunk and the inner jaw (or slide the chair on the rail, from which it could easily fall off).

Which means the filler chunk option is not really compatible with the loose jaws. It would be better to leave the filler chunk switched off (the default) when using the loose jaws option.

In all my experiments I have not found any reason to use the filler chunk, and I'm wondering whether to retain the option? I added it because I thought the key might need some support in the larger scales for 3D printing. It's not prototypical, although it seems to be present in injection-mould slide-on chairs (Exactoscale, C&L, etc.) and also in the Finetrax kits.

Notice that without the filler chunk, nothing engages against the rail foot on the outer side of the rail. The rail is located only by means of the key. That's the way the prototype works, and it seems to be working well in plug track too.

cheers,

Martin.

Hi Martin, this is all looking brilliant, I can not wait for the next update and start producing more of my turnouts via your marvelous system. Regarding the rail seat chunk, personally, I like the chunk because it gives a nice positive position so, I would prefer its left in and if people don't want it they can easily remove with a scalpel. Terry

 

[Martin Wynne]

I would prefer its left in and if people don't want it they can easily remove with a scalpel. Terry

@Terry Downes

Hi Terry,

???

With a scalpel? Why print it, if you intend to remove it? If you don't want it, don't tick the box:

Am I missing something?

cheers,

Martin.

 

[Terry Downes]

@Terry Downes

Hi Terry,

???

With a scalpel? Why print it, if you intend to remove it? If you don't want it, don't tick the box:

Am I missing something?

cheers,

Martin.

Sorry, I miss understood. I just meant to say that I would utilise the chunk option if available. Your comments seemed to imply that you were thinking of removing the 'fill below key' option.

 

[Martin Wynne]

Your comments seemed to imply that you were thinking of removing the 'fill below key' option.

@Terry Downes

Hi Terry,

Well yes, I was wondering about it. The tick-boxes and options are multiplying fast and must be bewildering to anyone seeing them for the first time. I was looking for somewhere to simplify.

In practice I very rarely remove anything, it just gets shunted to a more obscure place. For at least 20 years there has been no reason for anyone to need this function, and I can't remember the last time I used it myself, but it is still there in the menu:

And I wonder how long it is since anyone put the scalebar on, along the bottom of the trackpad (other colours are available):

It serves no purpose at all since the ruler tool was introduced, but it's still there if you want it.

cheers,

Martin.

 

[James Walters]

Hi James,
With reference to cork sheet, I was thinking of flat cork sheet rather than the stuff that comes in rolls.
If you bond the cork to plywod first how can you remove sections of it once cut?
I plan to experiment with using a 3mm cork trackbed with just the sockets cut out, then stick 1.6mm ply sleepers/timbers to the cork using "locator" chairs which get removed then slide real chairs onto rail and plug in to the correctly located timbers.
Or in fact I could try using loose jawed chais instead of "locator" chairs and leave them in situ.
Lots of things to experiment with, before I throw in the towel and go the FDM route!
Steve

ps vacuum ballast parcel arrived today , thank you

Hi Steve, I'm sorry I missed this yesterday.
I may have misled you here. My plan was to cut the plywood track bed/cork underlay to size in one go, and engrave the templot file onto the cork in one operation. I would then stick the 3mm ply timbering onto the cork afterwards. The timbering would be cut separately and so removing the waste sections would not be a problem.

I'm pleased your parcel arrived, I'll be interested to hear how you get on with it.

Best wishes,

James

 

[James Walters]

Hi James, I previously tried laser cutting timbers using a

Hi James, may I ask what Laser machine You are using? I have a Omtech 40K and my experiments have not been as successful as yours seem to be but, these discussions have urged me to have another go. I assume all of these photos are 4mm scale?

Hi Terry,
These were cut on a 30w Techsoft laser machine at work, 4mm scale 18.83mm gauge.

The machine in my own studio is a generic Chinese machine (80w), which does a better job in that it will cut faster which tends to limit the 'dwell' of the laser on the material, this tends to produce a cleaner cut.
I will repeat this experiment on my own machine in due course to add to the pool of knowledge, but am currently waiting for a replacement tube to arrive as mine has failed. It should have been with me 3 weeks ago, this is the downside to having a generic machine from an overseas manufacturer.

I would definitely encourage you to have another go. I am so impressed with the entire concept and am itching to produce a turnout using the system.

All the best,

James

 

[Martin Wynne]

.
The gauge-tweaking adjustment works fine for plain track, but within pointwork it is making the geometry very difficult to keep all the rails straight and in alignment. It is just about feasible for the turnouts, but for diamond-crossings it is going to be a nightmare to implement.

So I have reluctantly decided to remove it. I have to draw a line on the complexity of this project somewhere:

Which means the option marked X will be removed from the next program update.

But if you find the track gauge is coming out wrong, there are still numerous ways to correct it. The most likely reason will be incorrect shrinkage settings on the timbering base. There are 3 different places where the DXF/STL export can be scaled to correct for errors -- at A , B , and C above.

In addition to that there are scaling options in most 3D-printing and laser-cutting software to correct sizing errors.

If all else fails it would be possible to do a small scale/gauge conversion in Templot, and keep separate BOX files for 2D paper printing and 3D exports.

In practice I've been getting very accurate results from my printers, so I doubt any corrections will be needed in 00/EM/0-MF etc. They might be for P4/S7 etc.

Another possible cause of dimensional error is backlash in the axis drives on 3D printers and laser cutters. I wrote a G-code utility program to correct for backlash on my little MINIBO printer, which made a massive difference to the printing accuracy (it's available for use on the buttons at the bottom of the dialog above). If necessary we could do someting similar for other machines.

cheers,

Martin.

 

[Martin Wynne]

.
Another day, another tick-box. And barely any space left for it:

Handily placed with the cryptic label S1J on J1 There is no space for more explanation.

This is a S1J joint chair on timber J1 at the switch front:

There is also a S1J joint chair on J2, adjacent to the switch-front rail-joint in the usual way.

The original REA drawings show an ordinary S1 chair on J1 (and on any other switch front timbers between the switch toe and the rail joint, if there are any), with S1J on the joint timber J2 only.

However, many photographs show S1J chairs also on the J1 timber(s).

The new tick-box lets you choose between S1 or S1J on the J1 timber(s). I have set it ticked as the default, but I'm in two minds about that, I might change it.

Some pre-grouping designs used ordinary 10" wide sleepers for J1 and J2. There is an option for that in the custom switch settings -- it's been there right back to the beginning of Templot. If the switch front is set to sleepered, the above chair settings will be ignored and they will all have ordinary S1 chairs (S1J chairs won't fit on 10" sleepers).

cheers,

Martin.

 

[Martin Wynne]

.
I'm pleased to say that after a week-long battle I have finally got the flared end check-rail chairs mostly where I want them:

That's the normal REA CC parallel chair and above it the CCR angled chair for the flared ends. In EM.

If that was all that was needed it would have been done ages ago. Unfortunately it isn't:

Notice the near end of the near check rail. The timber has been shoved to a skew, changing the position of the chair on the rail. The check rail has been extended in length in front of the crossing. The flare length has been shortened, and the end gap on the check rail has been widened. Such things can be necessary in complex formations such as a tandem turnout, or to match the check rails for a specific prototype. But it still needs a matching chair:

I'm not entirely happy with this distortion of the prototype CCL chair for it -- the check rail seat could be improved (check rail seats are shorter than running rail seats because check rails do not need to support a vertical load).

But it will do for now, so that I can get on with the chairs for the V-crossing.

I knew from the start that the check-end chairs would be the most tricky part of the whole project, so I'm glad to have got this far now.

cheers,

Martin.