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

[Phil O]

@Phil O

Hi Phil,

Fishplates with bolt holes at 5" centres are intended for flat-bottom rail. Do they fit ok on bullhead? They are 20" long (as opposed to 18" for bullhead) so wouldn't easily fit between the chairs at a bullhead rail joint. Flat-bottom rail fixings do not obstruct the fishplates to the same extent, and the fishplates can be longer. If an 18" fishplate is drilled at 5" centres, the end holes would be very close to the end of the fishplate.

We need the full info on this because Plug Track will need it's own 3D printed fishplates. The standard H-section locking fishplates from the trade won't work because the rail needs to slide into them. Whereas Plug Track rails are pressed home vertically. The fishplates will need to be attached to the side of the rail rather than inserted from the end. The insulator lug on the back of the fishplate will occupy only half the rail height (and can be cut off where the fishplate is on a dummy joint).

cheers,

Martin.

Hi Martin,

My experience of the 2 sizes of fishplates on the ESR is that they fit ok against the fishing surfaces of 95lbs bullhead rail, I don't recall any problems with a direct replacement of broken fishplates with the wrong pitch of holes with the 2 centre bolts, only the inability to fit one or other of the outers.

I don't think that a lot of time or resources were expended by BR on the branch in its final goods only days and the Extension from Merrifield Lane to Mendip Vale was done with recovered materials from various sites and was just cobbled together, sufficiently for the inspecting officer from whatever department it was at the time.

There's an ongoing replacement with bullhead rail and concrete sleepers from the loco shed access crossover towards Merrifield Lane, which includes replacing some very old 85lbs rail, the only 60ft lengths on the original part of the line.

I realise that this may have opened a can of worms, but can heritage be termed as being prototypical? We only used what materials we had available to us.

 

[Phil O]

As a further thought to the above, the 85lbs rail has/had two bolt fishplates.

 

[Hayfield]

I am amazed at the level of detail acheivable by the Sonic mini 4K. I think it is helpful to model the detail as it helps to get the big stuff right. Martin thankyou for the chair pictures. Mine arn't as good as the old track near me is still part of network rail. I have the drawings for the chair plan views but I am having to work out the cross section.

I have refined my ordinary char and made a stab at the SS chair:-
View attachment 2087
View attachment 2088
printed on a phrozen sonic mini 4K. Aqua 4K resin, XY 0.037mm Z 0.030mm.

Excellent piece of work, very impressive

Just out of interest what are the printing costs please

 

[Martin Wynne]

As a further thought to the above, the 85lbs rail has/had two bolt fishplates.

@Phil O

Hi Phil,

Short 2-bolt fishplates were quite common at one time, for both BS-95R and BS-85R rails. They were 9" long (or 11.1/2" long on the Southern). It was thought that by allowing the joint sleepers to be moved closer to the rail ends, the performance of rail joints would be improved. In practice the improved support was undone by the failure of the short fishplates to hold the rail ends properly in line. So the practice of using them died out in favour of the traditional 4-bolt fishplate.

They are worth bearing in mind when Templot timber shoving, because it means the usual 24" timber spacing at rail joints can be reduced to 15" centres if it helps with the timbering layout. But bear in mind that to use them on plain track the standard sleeper spacings need to be replaced with a custom setting. 9" needs to be added to the total spacings, usually divided among the middle spacings. So for example on the 25 sleepers/60ft spacings there are 18 middle spacings of 2ft-5.1/2in centres, which would each be increased to 2ft-6in when used with short fishplates.

The 85lb BS-85R rail was intended for branch lines and sidings. In practice very little was used because such lines were usually relaid with reclaimed serviceable 95lb rail from main-line renewals. I have (so far) completely ignored it in Templot, along with the model trade -- in 4mm where is the code 72 rail or S2 and L2 chairs?

The BS-85R section is 5.15/32" high (1/4" lower than BS-95R). In 7mm/ft that scales to code 125. I'm wondering if that explains the mystery of the frequent use of code 125 bullhead by 7mm modellers, even including the S7 Society? Someone measured a bit of 85lb rail by mistake? To their credit C&L have just replaced their code 125 supplies with a proper code 131 95lb rail, but there have been complaints that it's "non-standard". Modellers are funny sometimes.

cheers,

Martin.

 

[timbersgalore]

Martin

I did mean undersize compared to the nominal values for the C&L track. I estimate that my rail is between 10 to 20% undersize. I think I could have fixed the problem using the presets you have already build in. I don't think the taper on the key was causing the sloppy fit.

After adjusting my CAD to describe the rail profile it terms of fish, I have done a lot of experiments to get the rail grip right. I started of with daylight under the key and adjust the prints to get a good grip of the web. The rail was still sloppy in the sense it was easy to roll the rail from vertical. Rather than fill completely below the key I have tried only filling below the key at the centre of the key, shown purple below.

The is still a bit of roll. I printed this with FootWidth set to 1mm, My rail is 0.9mm.

I think this demonstrates the benefit of customising the 3d print to the actual profile of the rail to hand. The other option is to deliberately make the rail a tight fit but I that is not an option if using a brittle resin.

I think it will take a least another iteration to get something I am happy with.

Steve: my sub layer already has the sockets cut in it.

 

[Stephen Freeman]

I have saved the Phrozen details on my wish list, just an aside , I can't remember if have mentioned here that in 7mm scale I do cosmetic etched (by PPD) brass fishplates, mainly because C&L stopped doing them mainly as a result of Phil not being able to get any more because the artwork had gone AWOL. I know he does alternatives now but there are times when etched is best (and cheaper).

EMGS and Brassmasters do them in 4mm scale.

 

[Steve_Cornford]

Hi Timbers,
Can you confirm that the concept of sockets in the sub layer works ok for you, and that it is practical to do away with any timber sprues?
If it does we can bunch the timbers up and get the most economical cuts of timbers.
Steve

 

[Martin Wynne]

I did mean undersize compared to the nominal values for the C&L track.

@timbersgalore

Hi Timbers,

Those are not "nominal" values for C&L rail. They are just random test data. Please ignore any dimensions you find in the current program release. It is all just dummy data used to test the programming and the process. I grabbed just any data that I had to hand. It is almost certainly wrong. My big fear in releasing unfinished stuff is always that someone will regard it as definitive info and start using it for real. Or worse still post it elsewhere on the web.

It is all utterly experimental at this stage. I will let everyone know when Plug Track has reached a stage where it can be used for actual models, but it won't be for some time yet.

I think you have your key too small and your outer jaw too low, which probably explains why the rail is not being firmly held upright. The key should completely fill the space between the rail head and foot, and the top of the jaw should be level or fractionally below the outer face of the key:

Also the rail seat/platform should extend across to the jaw, so that the rail can be dropped into the chair before being keyed across into place.

Making the rail vertical does affect the dimensioning to some extent. But check rails are always vertical, so the jaw dimensions from the inside of check rail chairs can be a reference. For REA chairs all outer jaws are 5.11/16" above the underside of the chair (except 2-level chairs and thick-base chairs).

Here is an actual GWR / BR(W) 2-bolt S1 chair. The top of the outer jaw is 5.5/8" above the underside of the chair. The jaws are 4" wide at the top:

cheers,

Martin.

 

[timbersgalore]

Martin
Don't worry I am taking very little form the templot chairs. My height comes from measurements of the plastic chairs from the 4mm Soc. The only reason I looked at the stl from templot was to provide you with feedback.

On the other hand the pictures of the fullsize chairs are gold dust to me. Plus your written feedback has been most helpful. Are the close up of the chair casting from your own private collection or are the available on the internet?

 

[Martin Wynne]

Are the close up of the chair casting from your own private collection or are the available on the internet?

@timbersgalore

Hi Timbers,

That chair has been in my garden for about 40 years.

I dragged it into the sunshine and took the pictures this afternoon.

cheers,

Martin.

 

[Martin Wynne]

They recommend cutting the sockets first then the timbers, in particular this service cuts blue lines first, then green, so it would be really useful as you suggested to have a SOCKOUTL layer as well as a TIMBOUTL layer, with colour choice.

@Steve_Cornford

Hi Steve,

Here you go, green and blue outlines adjusted for the width of the cutter kerf. No need to remove the timber extension marks:

Apologies if I have posted all this before, I'm getting in a bit of a muddle again.

I'm going to try to post an update soon, so that you can play with this stuff and report any problems.

cheers,

Martin.

 

[Martin Wynne]

Just out of interest what are the printing costs please

@Hayfield

Hi John,

Do you mean the cost of getting them printed commercially? Sorry, no idea.

If you mean the cost of making the chairs yourself, the actual resin cost is about 3 chairs for 1p (4mm scale).

But at a bare minimum you will also need:

A well-ventilated workshop area away from your living space. A bathroom + extractor fan or openable window might be suitable.

3D resin printer, around £200. The one I'm using (Elegoo Mars 2P) is currently on Amazon at £230, but there are cheaper models, offers and bundles.

A bottle of "ABS-like" liquid resin, around £35 per kg bottle. (Bundles will likely include the cheaper more brittle resin.)

A bottle of 99% isopropanol (IPA) solvent for washing the chairs. Price varies, the stuff I'm using costs £3.80 per litre.

A sunny day.

The amount of resin needed to make a chair is negligible, more resin will be wasted in the support sprues and washing than ends up in the actual chair. For the last batch of 144 chairs which I made, the software says the total cost of resin including supports was 26p. Say about 48p including washing, IPA and clean-up, which makes the chairs cost 0.33p each or 3 for a penny.

How you factor in the cost of the printer is up to you, but once you have one you will be able to make all sorts of other models and parts, not just chairs. But note that a resin printer is no good for making the filing jigs, for that you need a FDM filament printer (which is no good for making chairs below about Gauge 3).

If you don't have any sunny days in stock you can use a UV lamp instead. Cost is the same as a length of string. Or a dedicated UV curing and washing machine to match the printer costs around £100.

cheers,

Martin.

 

[Hayfield]

Martin

Thank you, I think the biggest problem is not having am outside workshop, secondly my inability of being able to use said machine. like injection moulding it seems the costs of materials are quite inexpensive. So I assume it will not be long until someone like Wayne comes along and offers a service at a reasonable cost

I may have misunderstood a previous post but the longevity of the resin seems questionable if unpainted and left in sunlight

 

[timbersgalore]

Martin

Laser cutters are capable of understanding shapes. What it means is that when presented with a nested set of rectangles they work out what is inside and outside. The result is that the cut planner can works out which bits to cut first. It also means that the laser can provide the Kerf offset. It would be worth checking whether a cutting service is automatically applying it own kerf offset to closed figures. That said it would be usefully to have the timber outline and chair socket layers on there own layers.


Here is a screen shot showing the cut layer from my recent test:-

A little shape manipulation is required to join the spruces to the timbers.


Timbers

 

[Martin Wynne]

secondly my inability of being able to use said machine

I may have misunderstood a previous post but the longevity of the resin seems questionable if unpainted and left in sunlight

@Hayfield

Hi John,

For the Templot chairs you don't need any ability to use the printer -- the files from Templot are ready to use. Just follow the instructions to make a few clicks on the computer and switch the printer on.

For other models, lots of modellers are making their model files available if you don't feel able to use CAD design software yourself, for example on RMweb. And there are huge libraries of other stuff online if you need a Toby jug or a new part for your vacuum cleaner.

Yes, resin-printed items are susceptible to UV light (as are many other plastics). They will harden and crumble over time if left in sunlight. They need to be painted to offer some protection, and preferably kept away from strong sunshine. Wayne probably mentions the need to paint his 3D-printed track bases in the instructions.

cheers,

Martin.

 

[timbersgalore]

Hi Timbers,
Can you confirm that the concept of sockets in the sub layer works ok for you, and that it is practical to do away with any timber sprues?
If it does we can bunch the timbers up and get the most economical cuts of timbers.
Steve

Steve
I haven't explicitly tried, though it is an idea that appeals to me. Even though I made the spruces only 1mm wide and took precaution to get no glue under them they were remarkable resistant to being removed(but achievable). The sockets cut by my laser have a taper. I set may laser up (kerf offset) so that I get my nominal dimensions at the bottom. I put at taper on my plugs but slightly less (room for glue). I would need to re jig this so that the sub layer provides the positive location for the chair (provide clearance on the sleeper sockets). That done it stands a very good chance of working out OK. I would recommend using birch ply for this layer.

I would still want to attack a spruce to the timbers so that I did not end up with a bag of match sticks.

I would still want to atach a spruce to the timbers so that I could lift them from the cutter.

Timbers

 

[Martin Wynne]

Laser cutters are capable of understanding shapes. What it means is that when presented with a nested set of rectangles they work out what is inside and outside. The result is that the cut planner can works out which bits to cut first. It also means that the laser can provide the Kerf offset. It would be worth checking whether a cutting service is automatically applying it own kerf offset to closed figures. That said it would be usefully to have the timber outline and chair socket layers on there own layers. Here is a screen shot showing the cut layer from my recent test:-

View attachment 2154

A little shape manipulation is required to join the spruces to the timbers.

@timbersgalore

Hi Timbers,

Thanks for the screenshot. Is that from a Templot 2-D file or 3-D file?

For 2-D the timber outlines and sockets are on separate layers:

In the 3-D files the timbers and sockets are all in one layer because they form a single solid object.

Laser cutters are capable of understanding shapes.

There are no rectangles or other area shapes in the Templot 2-D DXF files. Everything is a simple Line entity. The files are, or were originally, intended primarily for import into a CAD program for further track and layout design purposes.

For use on cutting machines there are other types of machine in addition to Laser cutters -- blade cutters, photo-etching, CNC milling machines, presses, stamping and blanking machines, etc.

I want to provide as many options and settings as possible to make the Templot DXF useful, but ultimately it is down to the user to know their own machine and how best to use the Templot DXF on it. If the kerf adjustment is not wanted, the layers can be switched off or the kerf width set to zero.

At present home laser cutters are much more expensive than an FDM printer, so the latter is more likely to be available to many Templot users. At least for the present my main focus is on the FDM timbering bricks for Plug Track. There has been a lot of recent discussion in this topic about laser cutting, but I suspect the majority interest will be in FDM printing for the timbering. I may be proved wrong as machine prices change, but I still feel that FDM printing will be the most user-friendly option for home pointwork construction, and is capable of excellent results comparable with injection-moulded track:

Also, an FDM machine can be used to make the rail filing and bending jigs needed.

cheers,

Martin.

 

[timbersgalore]

Very much the 2D dxf. I am still basicly using the same work flow as for 227. I have descovered it is very straighforward to place my own socket outlines using the track edge and timber centre line. Is everything you mentioned in the live version of 228 or is it unrealeased?

You track sample looks good, is that 7mm scale?

I have just completed another iteration of the GWR 95 chair. I used no fill below key but with the key heights fixed. I have also switched to ABS like resin. The Phrozen stuf as very little smell. Even able to use it in the house! I still have a small tweek to make the jaw height the same. The good new is that with a close fitting key the rail is firmly held.

Timbers

 

[Martin Wynne]

I have descovered it is very straighforward to place my own socket outlines using the track edge and timber centre line. Is everything you mentioned in the live version of 228 or is it unrealeased?

You track sample looks good, is that 7mm scale?

@timbersgalore

Hi Timbers,

Sorry, I can't remember what is in 228a. I think I put the bare minimum for folks to experiment. I'm planning to release an utterly experimental 228b shortly.

Which track sample? This one is 4mm/ft scale (EM):

I have held off playing with 7mm for now because I just know I will get distracted. I expect the results to look good, but the limited FDM build area will be a challenge.

Your GWR chair is now looking much better. Your key might be a bit short -- prototype keys are 6" long, i.e. 2" longer than the chair jaw.

cheers,

Martin.

 

[Steve_Cornford]

Hi Timbers,
Thanks for sharing all that useful information.

Laser Cutting
For my experiments the laser cutting service I used (4D model making materials) did not mention that their laser cutter understood shapes nor did they tell me what software they used to control the laser cutter.
When I asked them about the difference between their meaning of an inner cut(blue) and and outline cut(green) they just stated that the laser cut all the blue lines first then the green lines.
The principle being that if the timbers were cut before the sockets they could not guarantee that the sockets would be in the right place.
They required artwork with 0.076mm line width.
They also stated that if i wanted timbers 4mm wide and 32mm long I should add half the kerf (in their case 0.2mm) to the dimensions all round, and as far as the sockets were concerened (inner cuts) that I should subtract half the kerf all round.
Using Templot 228a, Martin suggested a work around to achieve this as far as the timbers were concerned, and he had already suggested a method to Ralph for adjusting socket size which I used.
I then used Inksape to open the resultant .DFX file so that I could copy the sockets to a separate layer and change their colour.
Martin has kindly added this experimental ability in to his next Templot release, together with the enhancement of putting the sockets in a separate layer thus making it easy to also produce a cutting file for a trackbed layer that the acts as a jig for the timbers layer, thus obviating the need for sprues holding the timbers in the right position.

The chairs I produce ( using an Elegoo Mars 2 pro, arriving in next 3 hours) will have longer pegs so that they go through the timbers and into the trackbed layer.
Also in my experiment I output both a straight B7 turnout and a curved B7 turnout, as I wanted to see whether there was any difference in the socket poitions on the timbers between a curved and a straight turnout.
Martin has since confirmed that there is no difference.
This means that I can get all the timbers cut as straight turnouts, and get the grain oriented parallel to the long edge of the timbers, also bunch them together to minimise cutting costs.

Rail Profile
I have just measued the web thickness (without filing head or foot) of some of the rails that Wayne Kinney has provided in 4 OO-SF B7 turnout kits with my cheap(powerfix) digital caliper and they appear to be 0.39mm. I checked the accuracy of the calipers using a 0.40mm thick feeler gauge (draper).
I believe Wayne sources this rail from the E.M.G.S.
I then tried measuring some rail lengths purchased from Scalefour Society a couple of years ago and these had 0.39mm thick webs.

Steve