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

[Martin Wynne]

Hi Martin,
I am hoping that you will retain the option to print chairs and timbers separately, as my preference is to be able to paint the chairs and timbers with different colours and it makes life easier if that is done before assembly.

As far as bed fltness is concerned is that really an issue? Do we worry about the flatness of the trackbed to those tolerances?
How flat are the components of out trackbed, plywood, layer of glue, cork, then another layer of glue for the timbers?
Just being a devils advocate here!

What you have produced so far has been very practical, and relatively easy to assemble, even the separate jaws. I am really looking forward to being able to print some check chairs the design of which you have appeared to crack.

As for the crossing chairs for a simple turnout, are there any particular designs problems that you are facing?
Appart from the reams of coding required that is!

Resin printing
I have been wondering about the possibility of speeding up the after print draining process.
After rpinting \i might try replacing the build plate knob with a suitable piece of studding, then inserting the end of the studding into an electric drill and then slowly spinning the build plate in a round plastic bowl as a home made centrifuge and let centripetal force do the rest. All this talk of marbles and cheese cutters have made my mid race.

All this written because my eggs were scrambled this morning!
Steve

@Steve_Cornford

Hi Steve,

No option will be abandoned, apart from the ones which don't work!

I still see FDM-printed or laser-cut bases as the standard approach, because those are the only options where the build area is large enough to be practical in affordable machines. Whether you choose to print whole timbers or individual chairs is all down to the tick-boxes. The whole-timbers approach is only feasible with loose jaws, and I can see some users will prefer the original bash-fit slide-on solid chairs option. Also you are not going to get very many whole timbers on the build plate of the little Mars printers, compared with doing a couple of hundred chairs in one go.

The delay to the crossing chairs is just down to me not getting on with it, there is no specific problem that I'm aware of (yet). I had to back-track a bit to do the fictional SC chairs to create the parts needed, but I just need to eliminate distractions now and get on with it. I used up a day of computer time editing the Zoom recording for last week, for example.

The question of how flat does track need to be depends on the wheel standards. In RTR 00 you can even lay track on the carpet! But in P4 a flat baseboard is essential. The overall height of finished track does depend on various factors, glue thickness, etc., but it needs to be consistent and repeatable. I still remember Tony Miles measuring the thickness of Sundeala sheets at all 4 corners with a micrometer, to the surprise of the shop assistant.

I used the power drill in a bucket centrifuge method when degreasing injection-moulded parts to remove the release agent from the moulding process. It works quite well, but does need to be fast. I can see you retrieving resin chairs from the far corners of the workshop! For drying after IPA washing I use my air blaster through an open window:

https://www.amazon.co.uk/Dusters-EG-1000-UK-Electric-Air-Duster/dp/B073F9NY1X/

A hair-dryer should work too.

cheers,

Martin.

 

[Terry Downes]

@Terry Downes

p.s. Terry,

Looking at your STL rendering, did you miss out the mesh fix? STL files from Templot need to be fixed via this free online service before 3D printing:

https://www.formware.co/onlinestlrepair

There is a link on the export dialog:

View attachment 5110

I also have a paid-for local copy of the Formware fixer. If your file(s) are too large and time out on the online tool, I'm happy to fix them for you. Just post your STL in this topic and I will post the fixed version back.

Everything is in a bit of a tangle because some users such as yourself are eager to start using the plug track designs while I am still developing them and they are unfinished. I'm happy for you to do that, but you must understand that some bits are still only stuck on with blue-tack and might drop off at any moment.

cheers,

Martin.

Hi Martin,
Yes you are correct I had not bothered about the stl repair as I was just checking processes and import functions into my CAD etc. I've just had a play around with dxf import into SolidWorks and 'Bingo' I had some template assembly file issues which I've fixed and here is the result. Templot 3d dxf imported into SolidWorks which gives me the ability to alter the 3d model, turn chairs, loose jaws on/off and add custom chairs. Now to try and resin print a sample with random chairs and/or jaws missing (as circled) with a few separate printed jaws etc. Now to see if I can remember how to fire up the Formlabs Form3 printer.

 

[Martin Wynne]

@Terry Downes

Hi Terry,

That's looking good!

You seem to have imported a Templot 3D DXF into SolidWorks, after previously saying that wasn't possible.

Can I tell users that Templot DXFs can be imported into a bog-standard copy of SolidWorks, or did that require some add-on or other extra function or package?

You wouldn't believe how difficult it is to get definitive information about anything at all which I can pass on to Templot users, without actually buying myself every printer, scanner, software and ball of string.

I see that SolidWorks have a cloud-based version for hobbyists ("makers") at the same cost as DSM Creator (£10 per month):

https://www.solidworks.com/solution/3dexperience-solidworks-makers

With the restriction that files are in the cloud and watermarked, and can only be used for private and non-commercial purposes (similar to Fusion360, which is free).

Whereas DSM is an ordinary local Windows exe, files are stored on your own computer, and can be used for any purpose including full commercial use.

---

P.S. JUST TO EXPLAIN for everyone to avoid confusion -- YOU DON'T NEED any of the above for Plug Track, the files from Templot are ready to use for 3D printing. You don't need any CAD programs or CAD skills.

Misinformation will spread across the internet and forums out of control if I don't try to keep a lid on it by repeating this stuff over and over again.

---

The options you mention, to edit individual chairs and use custom chair designs will eventually be in Templot too. There is already an option to omit individual chairs on a specific timber, and change the direction and randomness of the keys:

cheers,

Martin.

 

[Martin Wynne]

.
Bed of nails:

The idea is that you can place a collection of timbers on there, or any flat object, in any position. And they will then be supported for printing. The size of the raft is set by adding a rectangle in the background shapes.

That's enough digressions. I must get back to the chairs.

cheers,

Martin.

 

[Steve_Cornford]

It might help with the chairs if your eggs are beaten and tossed tonight ?

 

[Martin Wynne]

It might help with the chairs if your eggs are beaten and tossed tonight ?

https://www.studioforge.co.uk/shop/products/single-pancake-chair/

 

[Martin Wynne]

.
When conducting an experiment you have to be ready for unwelcome results.

It's now several months since I first started bash-fitting the chairs into FDM-printed timbers. Today I have noticed that some of the first ones I did are now loose enough to be pulled out quite easily from the timbers, or pushed out easily from below. Despite having been originally a tight bash fit needing a tap from a pin hammer.

I have measured the plugs and I can't find any significant shrinkage, nor changes in the dimensions of unused sockets. Chairs which have been removed are just as tight-fitting in unused sockets as they were originally.

The conclusion has to be that the looseness is caused by plastic flow of the PLA bases. Bash-fitting deforms the socket slightly, and as might have been expected with a thermoplastic material, it will flow over time under stress.

One of my main motivations in developing the plug track was to find a means of chaired bullhead track building which didn't involve the dreadful pong of butanone and the other common modelling solvents. Chairs which would press-fit firmly into position without adhesive seemed an ideal solution, much quicker and cleaner to use.

But if they come loose after 6 months, what might they be like after 6 years?

It might be possible to make them a clip-fit rather than a bash-fit, with a ridge and groove recessed locking design of some sort. But I'm not sure FDM printing is doable for that at 0.05mm resolution, although it would be worth some experiments. The disadvantage I see there is that if you accidentally clip a chair into the wrong socket, getting it back out would be very difficult. At least with a bash fit it can be pulled out with pliers, even if it does scrap the chair in the process.

So slowly but inexorably I think I have to accept a glued assembly.

But which glue, and how applied? There doesn't seem to be any common solvent for the UV resin or PLA materials, so it probably needs a solidifying sealant of some kind. It doesn't need to be quick-set, if we retain the bash-fit initial assembly. The easiest and cleanest to apply would be a sealant such as low-viscosity cyano superglue, or nail varnish, or similar?, applied into the sockets around the underside of the plugs after assembly. The lower part of the plug could have a groove around it to assist in creating a locking seal. Maybe ordinary enamel paint would do? Or maybe:

https://howesmodels.co.uk/product/humbrol-precision-poly-30ml/

which wouldn't have any significant solvent affect, it would just set as a solid cement, but is easy to apply cleanly in small amounts and thin enough to penetrate small gaps.

Whereas applying something into the sockets in advance, or dipping the plugs into something, makes the initial assembly process much messier. It might also clog the slots for the loose jaw pins. But in order to apply something to the underside after assembly, the track must be assembled on the bench, rather than in situ on the baseboard. That's fine by me, but I know some folks prefer to build track in situ.

Lots to think about, more experiments to conduct, and Amazon won't mind.

cheers,

Martin.

 

[Phil O]

Martin,

How about a drop of Loctite 6**, I can't remember the last numbers as used to secure Alan Gibson Wheels and crank pins? The joint can be broken, with some effort, or so I'm led to believe.

 

[Paul Boyd]

Martin,

How about a drop of Loctite 6**, I can't remember the last numbers as used to secure Alan Gibson Wheels and crank pins? The joint can be broken, with some effort, or so I'm led to believe.

Hi Phil,

You’re probably thinking of 601 or 603, both of which I believe need a ferrous substrate to cure.

Cheers,
Paul

 

[Phil O]

Hi Paul,

Yes, it was the 603, that I was thinking about and I forgot about the ferrous requirement. Sorry for the bum steer, Martin.

 

[ralphrobertson]

But which glue, and how applied?

Hi Martin,

Have you tried Plastic Magic? In the past I have used this on PLA with some success but never tried it on PLA and resin so I don't know if it would work.

Ralph

 

[Matt M.]

Hi Martin,

If you are using both UV resin and thermo plastic a slow acting two part epoxy would be best.
Finding a poly glue which bonds different types of plastics is very hard to do.

PLA has some issues with deforming over time due to moisture, heat and internal stresses.

Matt M.

 

[Martin Wynne]

Hi Martin,

Have you tried Plastic Magic? In the past I have used this on PLA with some success but never tried it on PLA and resin so I don't know if it would work.

Ralph

@ralphrobertson

Hi Ralph,

That sounds like dichloromethane, or some equally smelly solvent? I'm hoping to avoid such things, at least for my own use, because I find them so unpleasant to have open in front of me for long periods. Isopropanol (IPA), acetone, and butyl acetate (Humbrol Poly), i.e. the pear drops smells, are not too bad, but the chlorinated solvents do affect me badly. I think it's a hangover from my chain-making days 50 years ago, when we used them quite freely to degrease chain before welding and to extinguish small fires around the welding machines. Not much H&S in those days -- one colleague was badly burned when an hydraulic line burst on a welding machine. There was a mad rush to convert all the welders to a water-based hydraulic fluid after that.

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. Maybe I could with more Weetabix and a 2-pound hammer.

This is how the tapered plugs fit the sockets in the timbers:

You can see that there is plenty of void space for a penetrating sealant applied from below to grip around the plug, and it can also help to seal the loose jaw pin in place. I had in mind to put a groove in the lower side of the plug to assist the process, but based on the results from the white enamel it's not needed (other colours are available).

So maybe the problem is solved? I've been hearing that Humbrol enamels have been withdrawn for H&S reasons and replaced with water-based acrylics, but no doubt there is plenty left in stock at the back of model shops and modellers cupboards. Easily applied with a disposable cocktail stick.

cheers,

Martin.

 

[Stephen Freeman]

think I had some Plastic Magic but it went "off". I don't remember any smell in particular. I have a tube of something similar called Perfect Putty by Deluxe, which says it is suitable for all plastics and resins. It does say low odour and I cannot smell anything, mind you, post-covid I continue to be challanged in that respect and it's sometime since I had it.
May be worth a try.

 

[Martin Wynne]

@ralphrobertson @Stephen Freeman

Hi Ralph, Stephen,

Here is the Safety Data Sheet for "Plastic Magic":

https://www.farnell.com/datasheets/1833490.pdf

It's interesting that I had to go to Farnell to find it, I couldn't find the datasheet on Deluxe Materials own site.

cheers,

Martin.

 

[Martin Wynne]

.
I have now tried this stuff as a plug/socket sealant:

As supplied it is useless. It is too thick to penetrate gaps, and the bore of the needle dispenser is too small -- it clogs up within seconds.

However, after diluting it with Humbrol Liquid Poly (ratio about 2 parts thick to 1 part liquid, for which there is space in the bottle), and discarding the metal needle, it works ok. The hole in the cap after removing the needle will deliver just about the right sized drop, which will flash down into the socket around the plug. The needle can act as a bottle stopper to prevent the hole clogging up, and its sleeve seals over it.

After a couple of hours in a warm place, chairs which were very loose require some force to remove, and no doubt if left longer would get stronger. There appears to be a very slight solvent effect on the PLA base, but nothing that I can detect on the resin chair.

But the result doesn't seem to be any better than my 25-year-old tin of Humbrol enamel, which cost nothing (now), can be applied with more precision using a cocktail stick, can be left with the lid off for an hour or two, and most importantly DOESN'T SMELL. I now have a workshop smelling very strongly of butyl acetate (pear drops) from the short time the Humbrol stuff was exposed.

That's definitely enough distractions -- for the rest of this evening I'm working on the chairs! I have had a boiled egg.

cheers,

Martin.

 

[Steve_Cornford]

Some feedback regarding resin bases. Some months ago I stuck some resin bases onto a cork trackbed with Gorilla wood glue. They are still firmly attached.
I also glued some plywood sleepers down with a sparse application of Gorilla wood glue using locator plugs that were removed after about 20 minutes.
The locator plugs were pulled out with the aid of pliers.
The next day I then bash fitted some resin chairs that had been threaded onto rail, and these are still firmly in place.
I suppose if we were doing an accurate experiment we ought to devise a method of measuring the force required to un-plug the chairs!

I am also wondering whether during the course of ballasting some portion of whatever diluted adhesive is used would find its way by capillary action into the interstices between chairs and timbers?

I reported on Wednesday evening that Lidl was out of eggs , but I did manage to purchase a set of the electrostatic tweezers that you recommended

Today we managed to buy half a dozen eggs in the village of Firle along with some sweet pea seedlings

 

[Martin Wynne]

Some feedback regarding resin bases. Some months ago I stuck some resin bases onto a cork trackbed with Gorilla wood glue. They are still firmly attached.
I also glued some plywood sleepers down with a sparse application of Gorilla wood glue using locator plugs that were removed after about 20 minutes.
The locator plugs were pulled out with the aid of pliers.
The next day I then bash fitted some resin chairs that had been threaded onto rail, and these are still firmly in place.
I suppose if we were doing an accurate experiment we ought to devise a method of measuring the force required to un-plug the chairs!

I am also wondering whether during the course of ballasting some portion of whatever diluted adhesive is used would find its way by capillary action into the interstices between chairs and timbers?

I reported on Wednesday evening that Lidl was out of eggs , but I did manage to purchase a set of the electrostatic tweezers that you recommended

Today we managed to buy half a dozen eggs in the village of Firle along with some sweet pea seedlings

@Steve_Cornford

Hi Steve,

Thanks for the info. I would think that if the chairs start off a tight fit in plywood, they would likely stay that way unless the plywood got wet for long periods. It might then warp and distort, but it doesn't "flow" like plastic materials and change dimensions under stress. It's very possible that the ballasting adhesive would help to lock the chairs in place too.

With the milled MDF timbers, I found that I could brush some water-based sanding sealer around the chairs in situ, in a similar manner to using a solvent on plastic parts. The sanding sealer soaked from view into the MDF to harden it and seal the chairs in place. They are still there. If you did find the chairs coming loose in plywood you might be able to do the same, see:

https://85a.uk/templot/club/index.php?threads/cnc-milled-timbers-instead-of-laser-cut.276/post-2720

I've also remembered this stuff, which seems to work well as an adhesive when fitting chairs which are a loose fit to start with, but I think it's too thick to have a penetrating action as a sealant afterwards:

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

I also now suspect it's a very expensive way of buying ordinary waterproof wood glue!

Glad you got some eggs -- I've got mine under lock and key.

cheers,

Martin.

 

[Martin Wynne]

.
Back to the chairs.

I mentioned that in order to fit loose jaws in the switch block chairs they would need to be in one piece, instead of the previous 2 separate half-chairs. This is the new 3P chair:

The blue outers -- A on the switch rail and B on the stock rail, are the loose jaws+keys. Everything else is in one piece.

You will have noticed a gap in the middle where solid cast iron should be.

There's a reason for that. I've tried a few ideas to get round it, but none worked. As the loose jaw A is inserted in its slot, it must flex outwards to allow the key to pass the rail head and clip back into the rail web. Without a gap in the middle, in so doing the jaw would collide with the opposite fixed inner jaw on the stock rail, preventing it from going fully down into position.

You can't see the pin slot in this view, but it is parallel with the switch rail to allow the pin to flex away from the rail without twisting.

The solution is to leave a gap, and print a small filler piece to be inserted in the gap afterwards. It's not critical, the trains will run without it, and hardly anyone will notice. But you and I will know it's missing, and be unable to sleep at night until we have done something about it. It would have the advantage of locking the loose jaw firmly in place, and if a close fit not need any glue.

The gap is 7/8th inch wide (0.3mm at 4mm/ft) -- more tweezers needed!

All the jaws 3P to 11P will be the same, with A getting progressively wider as the opposite jaw and the gap remain constant.

There is no loose jaw A on the 1P and 2P chairs, it is a fixed support without a key. But there is no jaw on the other side of the rail on those chairs*, which allows the model rail to be flexed into position.

The loose jaw B is interchangeable on all other 4-screw chairs, including the fictional SC chair.

The loose jaw A however will be unique to this chair and not usable on any other. Some careful recording of the contents of the 3D raft will need to be kept.

The filler piece will be one fixed size for all the switch chairs.

If you don't like loose jaws and prefer solid chairs, it will be necessary to have both rails in the chair at the same time, for vertical bash-fitting. But for most single turnouts, it would be possible to slide the switch rail into place afterwards. With the loose jaw option switched off, there will be no gap in the middle of the chair.

If I can make a bit more progress in the next day or two, we can perhaps have a chat about this stuff in the Zoom meeting on Wednesday.

*except on the very long E and F switches, which are inside-keyed on those chairs. They can be loose -jawed when I get to them, but some thought will be needed to ensure they clear model wheel flanges.

cheers,

Martin.

 

[Martin Wynne]

.
Just to add to that, as you might have guessed there are some more tick-boxes. I can't remember how many of these were in 236d:

The above combination does this:

Solid jaws (and no middle gap) for the switch rail, loose jaws on the stock rail. Everything above would be a one-piece chair, the colours are just for clarity.

For those who might find inserting loose jaws between the rails too fiddly. But if you do this, for vertical plugging you would need to install the switch rail before the stock rail. Or slide the switch rail in afterwards instead of vertical plugging.

It's getting to be an increasingly complex set of options. I will add some buttons to set up the most frequently required combinations in one go.

There looks to be some tweaking needed to get the middle jaws to blend nicely at the centre. I might leave that until I have got the crossing chairs done. It's likely there will be lots of such final tweaking to do. It will barely notice in 4mm scale, but might do in the larger scales.

cheers,

Martin.