Templot Club Archive 2007-2020                            < back - search archive  browse archive images

...the home DIY printers deposit plastic onto a plate, on ours it is heated glass.  This deposition process means that any overhangs are unsupported.  Test prints of arch-like components all show a degradation of the under-surface at the top of the arch, amounting to about 30 degrees of arc. 

BTW it would help if the scale/gauge were mentioned in these posts. I'm assuming O gauge, but I could be wrong.

For filament deposition printers (such as the Makerbot I use for the MERG kits) the effect of temperature can help overcome this. The two servo mount kits illustrated were printed at different temperatures. The higher temperature makes the filament (PLA in my case) less viscous and so it does not span the gap.

Howard wrote:

For filament deposition printers (such as the Makerbot I use for the MERG kits) the effect of temperature can help overcome this. The two servo mount kits illustrated were printed at different temperatures. The higher temperature makes the filament (PLA in my case) less viscous and so it does not span the gap.

Howard,

What you describe matches our droopy experience of overhangs with a RepRap, and with a Mendel90. It is not too serious a problem generally, one can make an allowance in the model to allow a sufficient degree of manual fettling; or design the component to avoid it.

 Model railway Scale is not relevant to the work we have been doing, but when I model it is usually in 4mm.

 As a matter of interest, what temperature do you use for your extrusion of PLA ?

Les G

You are correct: Lower viscosity = stiffer; higher viscosity = more free flowing.

Higher temperature makes the plastic more liquid, and more likely to droop. This is why temperature and filament feed speed are controlled.  Hence my query about extruder temperature for successful bridging.

Les G

I see you did upload it to the Gallery. I have edited your message to include it.

The left hand image was done at 200 degrees, the right hand one (with drooping filament) at 220 degrees. Too low a temperature can mean parts do not adhere to the bed plate.
Howard.

We are currently printing PLA at 210 initial then 200 for the main print. Our Mendel90 has a nozzle  at 0.4 mm. We are using a glass plate heated to 60 degrees.  Adhesion is rarely a problem as long as the glass is squeaky clean; we use paper towel and meths for cleaning between prints.

There have been adhesion problems with small diameter prints at around 6 mill diameter  or less; 

Turnout stereo lithography file.

Andrew Barrowman wrote:

Turnout stereo lithography file.

Hi Andy,

Thanks for the file. I have downloaded it.

Now what?

Do you have a link for the software to use it?

regards,

Martin.

Martin Wynne wrote:

Andrew Barrowman wrote:

Turnout stereo lithography file.

Hi Andy,

Thanks for the file. I have downloaded it.

Now what?

Do you have a link for the software to use it?

regards,

Martin.

You can use the free download of the Makerbot Desktop to view the filehttp://www.makerbot.com/desktop
I had to rescale the turnout to fit, undefined The Makerbot bed plate is 28.5 cm wide.
At 0.1mm resolution it would take 1.5 hours to print; I might give it a go later.
Howard Watkins

Martin Wynne wrote:

Andrew Barrowman wrote:

Turnout stereo lithography file.

Hi Andy,

Thanks for the file. I have downloaded it.

Now what?

Do you have a link for the software to use it?

regards,

Martin.

You can use the free download of the Makerbot Desktop to view the file

http://www.makerbot.com/desktop

I had to rescale the turnout to fit, undefined The Makerbot bed plate is 28.5 cm wide.
At 0.1mm resolution it would take 1.5 hours to print; I might give it a go later.

1547_020513_550000000.png

Howard Watkins

I uploaded them to the Gallery but I can't seem to figure out how to Insert them into the message.

This is what I get when I try. (Using Chrome on Windows10.)

.stl file of a 00-BF V-6 turnout attached.
This file is intended for viewing only. It requires further editing before 3D printing.

.stl file of a 00-BF V-6 turnout attached.
This file is intended for viewing only. It requires further editing before 3D printing.

Andrew Barrowman wrote:

.stl file of a 00-BF V-6 turnout attached.
This file is intended for viewing only. It requires further editing before 3D printing.

New and improved!

I wonder how many Templot users do have 3D printers or intend getting one?

This is for the 4ft-1.5in gauge templates which I have been playing with.

I wonder how many Templot users do have 3D printers or intend getting one?

Hi Andy,

Many thanks for that.

It seems that adding and aligning the chairs is a tedious process?

One which Templot might possibly be able to do itself? Then you could create one-off 3D files to print any template in a full track plan, rather than a few fixed turnout sizes.

It would likely need a "shove chairs" function similar to shoving timbers to resolve any conflicts, such as replacing an ordinary chair with a bridge chair where needed. Which means the chairs would also need to be visible on the trackpad.

So the first question is -- are your chairs in a format which will import into Templot using the existing DXF import function (in the background shapes)?

Is it possible you could send me, or attach here, a DXF file containing just one of your chairs? I can then experiment with getting Templot to include it in the DXF export. Preferably the file would be in the DXF text format, rather than binary, so that I can read it. I can probably convert it if not.

What I'm still very hazy about is how we get from a drawing in TurboCAD to something a 3D printer can use? I've been reading that the STL format is many years old, dating from before 3D printers became generally available, and is generally regarded as too low in resolution for them?

Does your printer lay the filaments in horizontal sections, i.e. everything at one Z dimension in one go, and then step up to the next section and repeat? Or does the head move in 3 dimensions simultaneously, laying down flat angled faces for example? I spent years programming for that sort of thing in CNC milling using ball cutters and special-profiled D-cutters, but I'm thinking 3D printing is not quite the same. For example is a round filament the only option? I can see that in some applications a nozzle which extruded a flat strip might be more useful, especially if could by done at an angle from a swivellable nozzle, adding a K dimension to the X,Y,Z.

I can see that I am going to have to get one of these 3D printers and experiment before I can make sensible changes in Templot.

I have found this before, no amount of asking folks or reading docs gets me enough information. For example making Templot dpi-aware for hi-res tablet computers. The only way I could do that was to get one and try it. It was a lot of work, and expensive, but I'm very pleased with the result. But there has been minimal feedback on the subject. I wonder how many Templot users do have 3D printers or intend getting one?

regards,

Martin.

I probably only answered half your questions, so keep asking!

Attached version might be better. I removed the junk that was in a 2-D page tab and set the origin to something more sensible.

Well I thought I had attached it Trying again.....

No clue at all about the junk I'm afraid. I've never even looked in any of the files.

For the sake of drawn appearance you might radius the corners, but for printing at this scale it's a waste of time, it's only going to make the stl file even bigger, and that slows down the pre-printing process, a lot. (I know this first-hand.)

The printer automatically puts a radius on the corners.

Here is a very basic DXF file. It seems to be compatible with TurbocadPro16 and TurbocadDeluxe21 both for reading and writing. The file consists of two "blocks", a 2D rectangle block and a 2D triangle block.

What I hope we might be able to do is get Templot to produce a DXF file that has simple two-dimensional "blocks" positioned on the template at the places where chairs should be located. Each type of chair (3 bolt, bridge, check, etc, etc) would be represented by a unique two-dimensional block in the DXF exported by Templot.

Currently I am working in Templot to create "chair markers" represented by primitive rectangle outlines on Templot's own output.

That would be a neat feature of it was on the template output, it would add chair placement for modern functional chair based turnout construction.

One small thing - I don't want the non-gauge side of the rail on the template. It's a nuisance when I am trying to snap the chairs to the correct intersection. I eventually figured out how to get only the things I want in the DXF, but it took me quite a while, and I've probably forgotten how I did it by now.

If there was some sort of canned profile for the DFX it would help. Maybe there already is and I just don't know how to do it

If you can include the single block marker that is oriented correctly at each rail gauge-line/timber center-line intersection, that will still be a huge step forward.

I'm not entirely sure what to make of this. There is no end of weeping, a not unexpected amount of wailing, and a commensurate amount of gnashing of teeth (not to mention rending of garments) on a site that we know regarding the cost/difficulty of making decent looking turnouts for a reasonable price.

The best thing of all is that 3D printing lets a modeller use any old scale/gauge/standard they care to choose (just like Templot). A cynic might suggest that some are looking for an excuse rather than a solution

As soon as I have got a DXF out of this, and you have proved block substitution in CAD and successful printing, I'm going to be getting one of these printers myself and experimenting. At present the chairs are a bit primitive compared with the injection-moulded ones created by Len Newman, but everything has to start somewhere.

whereas the 1PL, 1PR, etc., block chairs are always square-on to the main road.

Can you point me at a list of the various chair types? The information I have is all from the 1930's.

Andrew Barrowman wrote:

Can you point me at a list of the various chair types? The information I have is all from the 1930's.

Hi Andy,

See: http://www.scalefour.org/downloads/gwrtracknotes/R4290A.pdf

That is Paddington's 1959 re-drawing of the Standard Railway Equipment (REA) chair designs and contains a few differences in the dimensioning compared with the original SRC designs. Not enough to make a a difference in a model. The GWR and BR(W) always had to be different.

AS1 is an ordinary chair. AS1J is a joint chair used on 12" wide sleepers adjacent to rail joints.

I can't find an online source of the original 1923 designs. They are in the NERA reprint book of the LNER 1926 SRC Permanent Way drawings.

regards,

Martin.

Do you have access to the Permanent Way Institute's Handbook? It seems to have some relevant information, but I've only skimmed over it.

For anyone who interested in dabbling in 3D, TurboCAD deluxe V20 is on sale at the moment from Cad and Graphics in the US for $49.99 (35 GBP?)

Obviously a bit more expensive than any of the free CAD programs out there, but it does work very well, and I'll be happy to help anyone who wants to give it a shot.

(I'm not connected in any way with either of these companies.)

Martin Wynne wrote:

whereas the 1PL, 1PR, etc., block chairs are always square-on to the main road.

Well that has always been my understanding, and all the switch drawings I have show this. Including the Exactoscale templates.

Unfortunately it cannot be. There is only one pair of L and R block chairs in each size, which are used for both left-hand and right-hand switches. The chair drawings clearly show the outer jaw for the stock rail square to the rail, with the inner jaw angled to match the switch rail (blade):

2_210026_220000000.png

I have therefore now modified the code for this:

2_210052_560000000.png

Unless and until anyone can shed more light on this?

regards,

Martin.

I think there must have been two varieties of L and R block chairs, one for each hand of turnout, but I could be entirely wrong.

A question: (I suspect the answer is no.)

If I make a 3D model of a straight turnout that consists of sub-models where the sub-models are timbers combined with their chairs into a single block or group, can I then convert it into a curved turnout by re-positioning the timber/chair blocks on to a curved version of the turnout?

In other words, will I be able to re-use any of the timber/chair blocks without changing the position and/or alignment of the chairs on the timbers?

My more limited version of TurboCad 20 allows you to write PDF files, if that's what you want.

Are you any further on with acquiring a printer? I'm hoping to see some turnouts in 7mm.

Andrew Barrowman wrote:

Are you any further on with acquiring a printer? I'm hoping to see some turnouts in 7mm.

You don't know me very well if you think I move that fast. The next job on my list is to decorate the Christmas tree.

Martin.

Andrew Barrowman wrote:

A question: (I suspect the answer is no.)

If I make a 3D model of a straight turnout that consists of sub-models where the sub-models are timbers combined with their chairs into a single block or group, can I then convert it into a curved turnout by re-positioning the timber/chair blocks on to a curved version of the turnout?

In other words, will I be able to re-use any of the timber/chair blocks without changing the position and/or alignment of the chairs on the timbers?

Hi Andy,

The answer is yes.

regards,

Martin.

Am I correct in thinking that the timbers in a turnout are spaced along the center-line of the main road and that the spacings do not change as the turnout is curved? Looking at what Templot does that seems to be the case, but maybe I'm missing something.

If that is the case, I think I ought to be able to print a straight turnout base with a single flexible web that runs along the main road center-line attached to all the timbers. That might allow me to print a common base that could be used for different turnout curvatures.

Andy,

Just one thing to consider in your development - the ability to push rails through the chairs when constructing the turnouts.  I have current experience with building interlaced turnouts in G1 using Cliff Barker's combined sleeper/chair units - similar to what you are designing - and I've found it difficult to thread rail through a lot of chairs before the resistance requires a lot of force.   Cliff Barker's chairs have a good grip on the rail and individual chairs will slide quite easily along the rail,  but I find that once I get up pushing the rail through ten chairs,  then it gets very difficult and I think some sort of damage might occur if I tried to go much further.   I've yet to try using something like wireman's lubricant to see if I can improve matters.  However,  in the turnouts I have constructed so far I can manage with two sections of rail in the longest rails - the stock rails - with the joints close to where the joints between stock and switch rails would be.  However,  I might run into problems if I was modelling a much longer turnout type using "C" and upwards switches,  my current turnouts all using "B" switches up to 1:8.

Jim.

Just one thing to consider in your development - the ability to push rails through the chairs when constructing the turnouts.  I have current experience with building interlaced turnouts in G1 using Cliff Barker's combined sleeper/chair units - similar to what you are designing - and I've found it difficult to thread rail through a lot of chairs before the resistance requires a lot of force.

100_020900_040000000.jpg

Jim Guthrie wrote:

Just one thing to consider in your development - the ability to push rails through the chairs when constructing the turnouts.  I have current experience with building interlaced turnouts in G1 using Cliff Barker's combined sleeper/chair units - similar to what you are designing - and I've found it difficult to thread rail through a lot of chairs before the resistance requires a lot of force.

100_020900_040000000.jpg

If you print the keys separately then the rail will just drop into the chair jaws from above.  Insert the tapered keys (narrow end first!) and then break/cut off the sprue.  

Warning: if used outside then glue the keys in place otherwise the birds will think that the keys are seeds and eat them!

Photo shows 7mm scale GER chairs; sorry about the depth of field.

Regards Dave R
Off The Rails Shop

I've encountered a situation where the X chair isn't fully supported on its timber. When this happens on the prototype, is the timber shoved to fully support the chair?

Andrew Barrowman wrote:

I've encountered a situation where the X chair isn't fully supported on its timber. When this happens on the prototype, is the timber shoved to fully support the chair?

Hi Andy,

Where did you get the dimensions to the chair centre and the timber centre? Normally they would be the same.

But where the model flangeway is overscale, this affects the position of the knuckle. It's not possible to chair a crossing entirely prototypically if the flangeway is not to scale. 

regards,

Martin.

It's easy enough to nudge the timber to accommodate the chair if that's legitimate.

There seems to be a minor incompatibility between Templot generated DXF files TurboCAD. I'm working in metric units in TurboCAD, but I have to set Templot to output the DXF in inches.

Andrew Barrowman wrote:

There seems to be a minor incompatibility between Templot generated DXF files TurboCAD. I'm working in metric units in TurboCAD, but I have to set Templot to output the DXF in inches.

Hi Andy,

A traditional DXF file contains only the numbers, it doesn't contain any indication of what units the numbers represent.

This means that when working with DXF files it is necessary to set the required units in both the exporting and importing programs -- normally the same units for both. That's why Templot puts a note of the units in the default file name for reference.

When importing a DXF file into TurboCAD, click the Setup button on the TurboCAD file dialog, and then set the units to match the file before opening the file:

2_160524_310000000.png

However, looking at TurboCAD DeLuxe 20, that option appears not to be available in the file setup. Instead, it seems that it loads the DXF file in inches regardless, and it is then necessary to change to Metric units at Options > Space Units menu item after opening the file.

Although I said I would do no more on the chairing, I have in fact been doing so, and I now have an algorithm to set 1 or 2 bridge chairs as necessary at the heel of a switch:

2_160542_230000000.png

When only one bridge chair is needed, I can't find any consistency in the REA sources as to which rails get the bridge chair. Different drawings show different arrangements. Anyone?

For Templot purposes I have for the present decided that the bridge chair goes under the turnout-road rails. A possible argument against this is that a skewed bridge chair has the chair screws closer to the edge of the timber than a skewed ordinary chair would have. This may become more of an issue in chairing crossings on square-on timbering. Discuss.

regards,

Martin.

In this shot it looks as if the the bridge chairs are on both stock rails.
Turnout

This location is easily accessible, so I will go and have a closer look.

Martin Wynne wrote:

whereas the 1PL, 1PR, etc., block chairs are always square-on to the main road.

Well that has always been my understanding, and all the switch drawings I have show this. Including the Exactoscale templates.

Unfortunately it cannot be. There is only one pair of L and R block chairs in each size, which are used for both left-hand and right-hand switches. The chair drawings clearly show the outer jaw for the stock rail square to the rail, with the inner jaw angled to match the switch rail (blade):

2_210026_220000000.png

I have therefore now modified the code for this:

2_210052_560000000.png

Unless and until anyone can shed more light on this?

regards,

Martin.

consistent with your statement that they are always "square-on to the main road"

2983_291849_380000000.jpg

This is a sample crossing printed by Shapeways in sintered nylon. There are quite a few things wrong with it, but I rather like the material. The definition isn't quite as good as other types of material, acrylic polymer for example, but the trade-off is that this stuff takes a lot of abuse.

Andrew Barrowman wrote:

consistent with your statement that they are always "square-on to the main road"

Hi Andy,

No, not to the main road, square-on to the stock rail.

I'm sticking to that, and drawings which show otherwise are just plain wrong, because block chairs with seatings square to the switch rail simply aren't available.

(At least for REA switches, pre-grouping switch designs can vary a lot.)

regards,

Martin.

Andrew Barrowman wrote:

2983_291849_380000000.jpg

This is a sample crossing printed by Shapeways in sintered nylon. There are quite a few things wrong with it, but I rather like the material. The definition isn't quite as good as other types of material, acrylic polymer for example, but the trade-off is that this stuff takes a lot of abuse.

Hi Andy,

Looking very good.

Does it need to take a lot of abuse? I think I would prefer the best achievable detail and definition, providing it is not so delicate that it can't be handled to insert the rail.

Once glued in position and ballasted, I don't see the need for great strength. You are surely not going to nail them down, as some folks do with set-track?

re: your experiments on RMweb with plastic sections of the rail -- a couple of points:

1. wheels should run on the wing rail until they reach the vee, so if the wing rail is plastic they would lose pick-up, as in a "dead frog". See the wear marks:

2_202055_480000000.png
thanks to Mick Nicholson for this photo

2. if model wheels rub against the check rail, they can be lifted fractionally off the running rail, again losing pick-up if the check rail is not live.

Experienced trackbuilders make the check rails live for this reason. With functional plastic chairs that means making an additional connection to the check rail.

regards,

Martin.

That's looking very good Andy.

You do realise that my reference to daylight showing through steel keys was in jest?

Do I understand that you are now planning to make these 3D prints available commercially? Do Peco know about this?

Martin.

it gave me the idea that spring steel keys avoid the problem of which way a wooden key should be facing.

I forgot about the wooden check rail keys. Which direction are they supposed to be driven?

That sounds promising Andy.

Can you provide a quick summary of these materials, relative costs, method of printing, etc? Are they all feasible for home printing?

The resin version looks great, and being able to solder close to it is a big plus. Is the resin available with a filler such as chopped glassfibre to strengthen it?

When you say it is fragile, how fragile? Can it be damaged by normal handling? If an occasional chair jaw snaps off, can it be glued back on? What with?

Martin.

What would be a reasonable minimum thickness for the timbers?

The only bit that really needs to be metallic is the rail's head (which is pretty much the same for bullhead and flat-bottom rails {ok, except for the tilt}).

Hi Andy

I love this thread and where you are taking things but a word of caution re the metal rail cap idea, not sure of the resistance of nickel silver but the c.s.a. of the metal will be seriously reduced when compared to a normal rail and will likely compromise the overload protection of a dcc system without a large number of droppers. Attaching the latter may well be difficult too.

Edit: will not only be problematic with dcc but also analogue systems.
Regards

Mike

mikewturner wrote:

Hi Andy

I love this thread and where you are taking things but a word of caution re the metal rail cap idea, not sure of the resistance of nickel silver but the c.s.a. of the metal will be seriously reduced when compared to a normal rail and will likely compromise the overload protection of a dcc system without a large number of droppers. Attaching the latter may well be difficult too.

Edit: will not only be problematic with dcc but also analogue systems.
Regards

Mike

Hi Mike,

You are quite correct. Nickel-silver isn't a very good conductor, and the cap would have a small CSA. It definitely will require droppers.

I have some ideas about how best to attach them.

Regards,
ab

Gosh! Isn't this exciting

Stay tuned for more exciting episodes of "What to do with your time when you are no longer gainfully employed".