Extracting a timbering brick from a track plan

[Martin Wynne]

I'm going to print the two, really just as a test and to see how well the clips work - I'm sure there'll need to be some printer adjustments.

@Paul Boyd

Hi Paul,

It's possible to get a good fit of the clips, but it will take a bit of trial and error -- several of the Cura settings affect it, plus these:

Yes, in due course it will be possible to save everything in the SK4 custom data file, including hopefully the slicer setup. But we are not there yet -- I'm not even sure yet what settings will be needed, for example we now need a new option to have a rail gauge-face mark on the timbers if there is no socket.

This option is new today:

It will make it easier to get the clip angled to match the end timbers.

I'm struggling today to do much computer work with a nose-bleed dripping on the keyboard. They've come later this year, I usually get them in February. Global warming?

cheers,

Martin.

 

[Paul Boyd]

Hi Martin

I've actually printed a couple of bricks! As expected, the claws/bars without any adjustments in either Templot or the slicer didn't quite join well - they weren't far off though which was pleasing. This photo is as printed, just cooling off.

This one shows roughly how it'll merge with some track already built.

These bricks were also the first time I'd printed anything that had been sliced by PrusaSlicer. Estimated time was 3h55m, actual time 3h45m. (3mm thick to match the existing track on its base) What surprised me was how much quieter the printing was compared to being sliced with Cura, not only that but one of the stepper motors went up and down musical scales as it was sweeping across the bed, especially on a diagonal - that's a new one on me, and I have no idea what's actually happening there! I changed virtually no settings in the Slicer from the default, and also kept the Templot settings as they were in your previous post, but I have a good basis to start adjusting things. I much prefer the PrusaSlicer Interface to Cura.

I did find a couple of issues over which I should have taken more care, and might be worth mentioning in any user guide, both highlighted in the screenshot below.

The first one in the blue ring was where I'd simply not noticed that the splint wasn't fully overlapping the web. More care needed on my part!

The other in the red ring I feel that perhaps Templot could have made it clearer that the claws weren't joined to any web properly - maybe by colouring the web in the same way that sleepers are coloured? (I'm still bearing in mind the nature of this whole project!) In this case a short splice was probably needed as the target won't reach across the two webs. Something to spell out in any instructions? (I missed it despite one of your earlier screenshots showing a dislocated tommy bar!) In Templot, it looked like this (the STL file was rotated 180-ish degrees for printing) :-

Now I know, I can see on the screen that the claws don't really join, but it didn't jump out at me. Again, something to draw attention to.

The add clip at notch function looks really useful - I'll definitely use that going forwards to save the manual rotation!

Overall, I'm really pleased with this. I'm going to print a brick with a very short bit of track so that I can play with the settings so the claw/bar mesh nicely. I'll have a bit of a double whammy as I'll also be experimenting with PrusaSlicer settings so I'd better get that out of the way first!

Cheers,
Paul

 

[Martin Wynne]

@Paul Boyd

Hi Paul,

That's looking good!

Generally I found the best approach to the sizing is to concentrate first on getting the sizes and spacings of the timbers spot-on in the slicer settings. Don't forget also to adjust the printer shrinkage settings:

Once you know your printer is producing parts to exact size (within the resolution limits of the printer), you can then adjust the clip clearance settings to get a good fit. It's best to work with trial pieces only 2 or 3 timbers long to save on polymer and wasted time.

To prevent disconnected clips, the idea is to use the clip size mouse action (clip then shows red) until you can see it attached under the hatched infill for the timbering* (automatic switch to hatched infill):

It is best to be attached to the timber itself rather than the side flanges for full strength (the default clip is thicker than the flange, but still thinner than the timber, so that it is still under the ballast).

At the same time it is important to ensure that the tommy bar area is not over any other part of the brick, otherwise that would break into the claw space on the print.

In addition you probably noticed the elephant's-foot protection recess around the underside of the clip -- it's important that is clear of other parts too.

Also of course the clip must not conflict with any of the sockets, otherwise the clip would break into the socket on the print.

p.s. you can also use the clip size mouse action to swap the clip direction -- make it very small and watch.

If you use a brim or skirt in the slicer, don't forget to add the brim protection fences around the clips:

So much still to write about and explain to get a working user interface -- but will anyone ever read it all and take it in?

cheers,

Martin.

*p.s. Sorry I messed up the screenshot by having the same colour for both bricks -- I don't have the energy to make another.

 

[Martin Wynne]

We have now done the very thing we normally try hard to avoid -- created a duplicate partial template.

Which needs some further thought, otherwise saved BOX files are going to get very confusing. Duplicated stored templates often cause problems -- partially duplicated templates can only make matters worse.

For 234a I'm minded to add a new bricklaying mode:

With that box ticked, storing a template would mark it as a brick template, i.e. a member of a timbering brick.

Brick templates can then be grouped, saved in their own BOX files, shown overlaid on the underlying track plan templates, or not, etc.:

But I'm worried this is getting out of hand and Templot is just becoming far too complex -- if you were a newcomer to track planning software, what would you make of a tickbox labelled "bricklaying"?

I can hide it all away somewhere in the menus. But then vast swathes of users (do such things exist? ) will never find it -- and I spend the rest of my life explaining where it is.

cheers,

Martin.

 

[Paul Boyd]

Hi Martin

Generally I found the best approach to the sizing is to concentrate first on getting the sizes and spacings of the timbers spot-on in the slicer settings.

Good point! I've just checked against a print (printer calibrated, of course) and the two bricks are spot on, so that's one less thing to have to fiddle with.

To prevent disconnected clips, the idea is to use the clip size mouse action (clip then shows red) until you can see it attached under the hatched infill for the timbering* (automatic switch to hatched infill):

Somehow, that action passed me by, despite having already used the shift and rotate clip functions either side of it. No excuses there, and from your screenshot it's completely clear - just watch the hatches!

In addition you probably noticed the elephant's-foot protection recess around the underside of the clip

I had to laugh when I read that! I had noticed it and couldn't work out why on earth the first layer of the clips was undersized, when the rest of the brick wasn't. I spent a good few minutes looking at settings and stepping through the slicer preview to try to work out what was happening, then your post arrived. Of course!! I seem to have resolved the elephant's foot issue (bed too close to the nozzle, I think) but I might have known you would already have thought of that and come up with a solution! That's one area where elephant's foot would definitely not be good.

With that box ticked, storing a template would mark it as a brick template, i.e. a member of a timbering brick.

I've been starting to wonder about that. My current solution is to change the project title to add a '3d' suffix and save them in a separate folder, but of course if I make any changes to the main layout (unlikely at this stage), I have to repeat them in the 3d versions. Personally, I think tagging a template as a brick template would work well, especially if it can follow any changes to the underlying template (!!!). Would it be possible to have a global "3D planning" flag that turns on or off everything to do with 2D/3D exports at a stroke? Paper/PDF only unless the user specifically turns on all the extra bits? I get the point about bombarding beginners with too many options. It does feel like we're heading towards needing a Templot Lite - a project for the open source guys?

Later I'll be making a 2-timber brick and start playing with settings to get the clips spot on. And then, of course, with a change of material I'll have to fiddle with the settings again. Having said that, it appears that despite all the information out there, butanone does act as a solvent for PLA, at least for the Anycubic brand I'm using. That means I may be able to stick with my current filament, but I need to do some more trials first.

Cheers,
Paul

 

[Martin Wynne]

I had to laugh when I read that! I had noticed it and couldn't work out why on earth the first layer of the clips was undersized, when the rest of the brick wasn't. I spent a good few minutes looking at settings and stepping through the slicer preview to try to work out what was happening, then your post arrived.

@Paul Boyd

Hi Paul,

If you have eliminated any elephant's foot issue, you might want to reduce or zero the settings:

One thing which is fixed at present is the actual size of the clip parts (tommy bar). Do you think it's about right? It's on the small side (in 4mm scale), but for clips between timbers anything bigger doesn't really fit. Perhaps we need two sizes of clip -- a larger one for fresh air clips on splints?

especially if it can follow any changes to the underlying template (!!!)

Help! -- you just made me 5 years work! What happens if the underlying changes make it longer, and it no longer fits the brick? Or shoving the timbers disrupts the clip position? Or...

Would it be possible to have a global "3D planning" flag that turns on or off everything to do with 2D/3D exports at a stroke?

Ideally yes. But it gets complex, there are dozens of places in the code where you would need to know which mode you are in, and take different actions accordingly. For example we had a 2-D DXF export for roll-paper CAD printing for years before 3D printers were invented, and I don't want to lose it. It was probably a mistake to link the 3D stuff into it, rather than starting again.

A major task still not yet resolved is bunching all the chairs from a brick to fit a small resin printer, allowing for the fact that they are all at odd angles, and printing a paper map showing which is which, cross referenced to the brick templates. For now, bunching will have to be done manually via the auto-arrange functions in the slicer (and pencil and paper to remember which is which).

Not much left to do, really!

cheers,

Martin.

 

[Paul Boyd]

Hi Martin

Just been playing, and for my printer settings I've found setting the end and side clearances to 0.1mm gives a perfect click fit, and setting both to 0.15mm gives a little wiggle room - I'd probably go for the latter, or more likely somewhere in between. I haven't adjusted the corner relief. I'm inclined to leave the bottom offset settings alone, now I understand what they're for, as they don't do any harm. My little test piece, all of two sleepers long, had the clips correctly sized and placed overlapping the sleepers slightly!

One thing which is fixed at present is the actual size of the clip parts (tommy bar). Do you think it's about right?

I do think they're about right. I did some tests with 2mm scale and 7mm scale track and I'm guessing the clip size is based on the scale rather than track gauge. For my 4mm scale, 7.83mm gauge track they're maybe a little big but certainly nothing to worry about.

I'm now happy enough with timber bricks to start printing them for real. I've looked into resin printers a bit more and still can't see myself getting one, not yet anyway. That means that if I do want Templot chairs, it'll be a Shapeways job

Cheers,
Paul

 

[Martin Wynne]

I was looking for an option to turn chairs off, then realised that there would then be no position guide for the rails. What I actually did was to reduce the length of the socket to something very low - when sliced, there would then be a visible indication from the filament path.

@Paul Boyd

Hi Paul,

In addition to S1, L1, etc. chairs, we now have RG chairs:

RG chairs are nominal fictional chairs intended to produce your rail grooves.

The RG chair dimensions are such that the socket for them is a slot matching the rail head width aligned with the rail edges. Small enough to be covered over by whatever rail fixings you are using.

Then you set the blind sockets option, and whatever socket depth you want for the groove.

It would be possible to do FG chairs also, to produce a rail-foot sized groove, for flat-bottom templates. Another tickbox.

cheers,

Martin.

 

[Paul Boyd]

@Paul Boyd

Hi Paul,

In addition to S1, L1, etc. chairs, we now have RG chairs:

View attachment 3475

RG chairs are nominal fictional chairs intended to produce your rail grooves.

The RG chair dimensions are such that the socket for them is a slot matching the rail head width aligned with the rail edges. Small enough to be covered over by whatever rail fixings you are using.

Then you set the blind sockets option, and whatever socket depth you want for the groove.

View attachment 3474

It would be possible to do FG chairs also, to produce a rail-foot sized groove, for flat-bottom templates. Another tickbox.

cheers,

Martin.

That sounds brilliant, thank you! I've just been writing some notes where I've written that as I'm not printing chairs then I don't need to worry about setting a custom rail size - I'd better amend those notes on the next Templot update!

Cheers,
Paul

 

[AndyB]

Hi Martin,

Have you considered "capping" Templot as 2-D only and creating a different prog to create the 3-D models? Of course 2-D Templot would have a convenient and likely proprietary method of exporting templates to the 3-D prog.

Cheers,
Andy

 

[Martin Wynne]

Hi Martin,

Have you considered "capping" Templot as 2-D only and creating a different prog to create the 3-D models? Of course 2-D Templot would have a convenient and likely proprietary method of exporting templates to the 3-D prog.

@AndyB

Hi Andy,

Thanks for that.

Yes, I've often wondered about the ways forward. If we could rewind to 15 years ago when Templot was a commercial product, such a split would be an obvious course of action and similar to many other software products.

But I don't think my ageing brain is up to it now. Nowadays I make so many silly mistakes just keeping one show on the road. The extra logistics of releasing two different programs, keeping them in step, two different downloads, two different setup scripts, but most of all, writing all the ifs and buts into two versions of the docs would be just too much for me.

For the future, now that Templot has been free for over 10 years, I think I shall just plod along doing what I can with Templot as my hobby project, and hope that users can accept it for what it is. I enjoy tinkering with the code, trying new ideas, but converting it into a slick modern software app is beyond what I can manage or have time for -- too much computer time is proving bad for my health, and with sunnier days coming and Covid restrictions behind us, I want to get back into the hills with my camera.

Although the 3D stuff has significantly increased the complexity of Templot, most of it is still hidden from newcomers to Templot at present. Currently, most new users are unlikely to find the menu switch for the experimental chairing, or be interested in exporting CAD files. But I'm hoping that it won't always be experimental, and if Plug Track becomes a successful concept within the hobby, future newcomers might arrive here looking for it. Then what -- "make the control", "shove some timbers", and now do some bricklaying.

But it is free.

And it's all open-source -- if others want to dive in and develop the code in adventurous new directions that would be wonderful. I'm sure Graeme and Alistair would be delighted to have new kids on the T3 block. Or fork your own and take Templot wherever you want it to go: https://github.com/openTemplot/templot3 or https://sourceforge.net/projects/opentemplot/
or see: https://85a.uk/templot/club/index.php?forums/templotmec-nuts-and-bolts.5/

cheers,

Martin.

 

[AndyB]

@AndyB

Hi Andy,

Thanks for that.

Yes, I've often wondered about the ways forward. If we could rewind to 15 years ago when Templot was a commercial product, such a split would be an obvious course of action and similar to many other software products.

But I don't think my ageing brain is up to it now. Nowadays I make so many silly mistakes just keeping one show on the road. The extra logistics of releasing two different programs, keeping them in step, two different downloads, two different setup scripts, but most of all, writing all the ifs and buts into two versions of the docs would be just too much for me.

For the future, now that Templot has been free for over 10 years, I think I shall just plod along doing what I can with Templot as my hobby project, and hope that users can accept it for what it is. I enjoy tinkering with the code, trying new ideas, but converting it into a slick modern software app is beyond what I can manage or have time for -- too much computer time is proving bad for my health, and with sunnier days coming and Covid restrictions behind us, I want to get back into the hills with my camera.

Although the 3D stuff has significantly increased the complexity of Templot, most of it is still hidden from newcomers to Templot at present. Currently, most new users are unlikely to find the menu switch for the experimental chairing, or be interested in exporting CAD files. But I'm hoping that it won't always be experimental, and if Plug Track becomes a successful concept within the hobby, future newcomers might arrive here looking for it. Then what -- "make the control", "shove some timbers", and now do some bricklaying.

But it is free.

And it's all open-source -- if others want to dive in and develop the code in adventurous new directions that would be wonderful. I'm sure Graeme and Alistair would be delighted to have new kids on the T3 block. Or fork your own and take Templot wherever you want it to go: https://github.com/openTemplot/templot3 or https://sourceforge.net/projects/opentemplot/
or see: https://85a.uk/templot/club/index.php?forums/templotmec-nuts-and-bolts.5/

cheers,

Martin.

I was hoping it might mean less work for you. The 2D part of Templot is extremely stable and you could do a lot of experimental stuff in a 3D prog without fear of messing up Templot, but then I have no idea how it's all structured and I'm far too old now to try to find out

 

[Martin Wynne]

I was hoping it might mean less work for you. The 2D part of Templot is extremely stable and you could do a lot of experimental stuff in a 3D prog without fear of messing up Templot, but then I have no idea how it's all structured and I'm far too old now to try to find out

@AndyB

Hi Andy,

I'm not too worried about messing up Templot -- it's an increasing possibility, but we could always jump back to a previous release.

My fear is that the whole thing is beginning to look impossibly complex to newcomers, and turn them away. Which would be a shame, because as you say the core of Templot is very stable, and has been working fine for years.

Paul suggested a global switch to enable the 3D stuff and chairing. It's too much work to put such switches within the actual code, but what I could do is simply hide all access to it. That would hide the old 2-D DXF exports at the same time, but that might be acceptable -- it wouldn't affect the large-format PDF exports for roll-printing.

Thanks again,

Martin.

 

[Martin Wynne]

RG chairs are nominal fictional chairs intended to produce your rail grooves.

The RG chair dimensions are such that the socket for them is a slot matching the rail head width aligned with the rail edges. Small enough to be covered over by whatever rail fixings you are using.

Then you set the blind sockets option, and whatever socket depth you want for the groove.

@Paul Boyd

Hi Paul,

Blind sockets for dummy RG chairs, making rail groove marks:

These are 0.2mm deep (EM gauge), can be set to any depth.

Will be in 234a soon.

cheers,

Martin.

 

[Martin Wynne]

.
These are for dummy FG chairs (groove matches FB rail foot width):

These are 0.1mm deep.

Baseplates should cover them, or a spot of glue in the groove for direct rail fixing.

cheers,

Martin.

 

[Paul Boyd]

That’s great, thanks! I’m holding off generating the STLs for now - no pressure . I’ve still got some work to do adding all the clips and splints so that’s keeping me busy for a bit. I will remember to check and change the colour and gender of each clip for each brick before exporting!

I’ve also changed the process slightly in that rather than using the boundary rectangle as a visual guide to the bed size, I’m drawing rectangular shapes (actually square for my printer) to the correct size, and naming these brick 1, brick 2 etc. They can be shifted just as easily and the names will correspond to the saved STL file name. I just find it easier to have a built-in record of which brick is which.

The trackpad is starting to look quite colourful!

Cheers,
Paul

 

[Martin Wynne]

I just find it easier to have a built-in record of which brick is which.

@Paul Boyd

Hi Paul,

I'm intending to have some means of printing a part number on each brick. Probably on a splint. I haven't yet written any code to draw text, but we could have a system of dots -- 2 dots = brick #2, etc.

Can you see any advantage in having those rail grooves deeper or going all through the timber? I'm thinking of some sort of brass insert or pin which could be soldered to. PLA melts around 175 degs, ABS higher, so it should be possible to use low-temp solder with an RSU? It's easily doable, just don't tick the blind sockets option.

cheers,

Martin.

 

[Paul Boyd]

@Paul Boyd

Hi Paul,

I'm intending to have some means of printing a part number on each brick. Probably on a splint. I haven't yet written any code to draw text, but we could have a system of dots -- 2 dots = brick #2, etc.

Can you see any advantage in having those rail grooves deeper or going all through the timber? I'm thinking of some sort of brass insert or pin which could be soldered to. PLA melts around 175 degs, ABS higher, so it should be possible to use low-temp solder with an RSU? It's easily doable, just don't tick the blind sockets option.

cheers,

Martin.

Hi Martin

I'm not sure about dots (brick 37!) so if you can, I think '37' would be better. I'm only up to brick 5 at the moment though, and I think I'm only going to get to brick 10 or so - this is just for the bullhead sections of the layout. For the time being I was just going to use a marker pen on a splint as I take each brick off the bed, more so for plain track where there'll be a few very similar bricks being printed at once.

For me, the rail grooves at 0.2mm would be fine, and should show up clearly - I'm printing these at 0.2mm layer height which seems adequate. I'm not sure about using brass pins etc because that's not my preferred method of construction, so I don't really feel qualified to answer that!

Cheers,
Paul

 

[Martin Wynne]

I think '37' would be better.

How about Roman numerals? All straight lines!

Martin.

 

[Paul Boyd]

How about Roman numerals? All straight lines!

Martin.

That’ll do! I still understand Roman numbers,