Templot Club Archive 2007-2020                             

topic: 3307progress report - chairs in the output
author remove search highlighting
 
posted: 16 Jul 2018 20:34

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
As I mentioned, the reports of folks using 3D printing to create bullhead chairs has spurred me to move chairing in Templot from the NOD* list to the to-do list. :)

I started some experimental work a couple of years ago, and I have now returned to it.

It seemed that if the output was to be of any use in 3D printing, the chair outlines would need to have the proper corner radii. It's a lot of extra calculation with 4 chairs on most crossing timbers, but it hasn't impacted too much on the screen response for full track plans. Obviously you would switch the chairs on only for the final printed templates, they are not needed while track planning.

It's fiendishly complex. For example the REA bridge chairs (L1) have a larger corner radius (1.3/4") than ordinary S1 chairs (which have 1" corner radius): 

2_161458_330000001.png2_161458_330000001.png

And GWR slide chairs have a full semi-circular end on the inside, for example.

Templot does all the positioning and rotating of the chairs, and calculates where bridge chairs are needed to fit the space.

For 3D printing it needs to be in the DXF file. So this is where I've got to:

2_161458_330000000.png2_161458_330000000.png

Just the chair bases so far. The next job is to change the rails to a proper bullhead section. Then add the keys, and the chair jaws. And the chair screws.

There is a lot still to do. I may be gone for some time. :)

Just to be clear, the above wasn't drawn in a CAD program, it is the DXF export from Templot. The CAD program is used simply to view it, and then save it as an STL file for 3D printing. All the dimensions and fits will be entered in Templot. No CAD knowledge will be needed -- at least, that's the plan.

This is in Templot2. When I have got it finished and working I will make the code open-source, but I will probably leave it to others to incorporate it into OpenTemplot.

*Nice One Day

cheers,

Martin.

posted: 16 Jul 2018 20:48

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Nice work with the chairing :)

If it isn't too much to ask can you add some code to aggregate all the chairs used on templates in the storage box ( like the timber/sleeper length one which is great ). It would need to be broken down into quantities of S1, L1 types etc. This would make ordering/stock checking a breeze. I am looking at the 'old fashioned' route of using injection moulded chairs and actually glueing them on.......

Rob


posted: 16 Jul 2018 23:59

from:

d827kelly
 
Coventry - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
A great addition to templot!

posted: 17 Jul 2018 08:12

from:

richard_t
 
Nr. Spalding, South Holland - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I've been doing similar, in 2D, using my cad program - I think I'd go slightly crazy doing a full plan, even though the chairs are blocks so they just need putting in the correct place.



Attached is a simple A6 turnout, without the crossing chairs - it's a PDF, as the PNG I generated looked awful.



Attachment: attach_2700_3307_a6-chairs-test-Model.pdf     289

posted: 17 Jul 2018 18:31

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
richard_t wrote:
I've been doing similar, in 2D, using my cad program - I think I'd go slightly crazy doing a full plan, even though the chairs are blocks so they just need putting in the correct place.



Attached is a simple A6 turnout, without the crossing chairs - it's a PDF, as the PNG I generated looked awful.



Hi Richard,

Did you draw it all in CAD or did you import a DXF from Templot?

Andy

posted: 17 Jul 2018 18:38

from:

richard_t
 
Nr. Spalding, South Holland - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Everything bar the chair outlines is from Templot. I'm not that loopy (yet) :-)

posted: 18 Jul 2018 05:35

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
richard_t wrote:
Everything bar the chair outlines is from Templot. I'm not that loopy (yet) :-)
Correct answer! You get ten points (points, not turnouts :) )

posted: 30 Jul 2018 17:14

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Nice work with the chairing :)

If it isn't too much to ask can you add some code to aggregate all the chairs used on templates in the storage box ( like the timber/sleeper length one which is great ). It would need to be broken down into quantities of S1, L1 types etc. This would make ordering/stock checking a breeze.
Thanks Rob.

Yes, that's on my list. Also splitting the timbering into separate counts for sleepers and turnout timbers.

I'm making some progress, as discussed on the 3D printing topic. Chairing works for equalized timbers or square-on, and the chairs remain on the rails when timbers are shoved*:

2_301200_590000000.png2_301200_590000000.png


For 3D printing, the chairs go into the exported DXF files:

2_301202_320000000.png2_301202_320000000.png


As you can see, there is a long way still to go. The chair jaws and keys to add, and then all the special switch and crossing chairs.

2_301203_170000000.png2_301203_170000000.png


Of course this level of detail won't survive 3D printing in 4mm scale, but it might in 7mm, and should do in the larger scales. The screw heads are all at random angles.

The bright colours in the screenshots are simply to make the details more visible.

*not when crabbed.

cheers,

Martin.

posted: 30 Jul 2018 17:34

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:


I'm making some progress, as discussed on the 3D printing topic. Chairing works for equalized timbers or square-on, and the chairs remain on the rails when timbers are shoved:

cheers,

Martin.
Hi Martin,

Thanks.

Yes you are making great progress, interesting to watch the the development of this even if much is way over my head :(

Rob

PS...Ha Ha, Surely you need to take the chair screws out of the chairs before you can shove the timbers otherwise the rail will get slewed over? :D

posted: 30 Jul 2018 18:01

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks Rob.

The 3D printing stuff is going over my head too. I'm still writing DXF code the same way I did in the 1980s. :)

That's why I started this separate topic to cover the track design aspects.

Here is something that will be possible. If you omit a timber when shoving, its chairs go too, as you can see. However, if instead you set its length to zero, making it invisible, its chairs remain unsupported on the rails. Adjacent long timbers can then be extended under them:

2_301256_010000000.png2_301256_010000000.png

With the timbers invisible, shoving them along will be a way to slide the chairs along the rails to any desired position.


2_301320_510000000.png2_301320_510000000.png

UPDATE:

We now have the keys, and outer jaws on the S1 chairs. Gold-plated: :)

2_061704_070000000.png2_061704_070000000.png

Notice the random key positions (and screw heads).

There should be a corner radius around the top of the jaw, but I think I'm going to draw a line at this. I've used as much of the REA drawing as seems reasonable. I'm hoping that the 3D printer will smooth out some of the sharp angles:

2_061707_200000000.png2_061707_200000000.png

Now for the inner jaws, L1 chairs, slide chairs, slab & brackets ...

Don't hold your breath.

cheers,

Martin.

posted: 19 Nov 2018 20:29

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
After a break I'm now working on this project again. :)

Dave asked about the chairing parameters. Here is a list of the settings for the S1 ordinary chairs.

All the settings are in full-size prototype inches (as most else in Templot):

    seat_thick:single;           // full-size inches chair seating thickness   1.75"
    chair_plinth_thick:single;   // chair plinth is a nominal area inside the
                                 // outline perimeter (rectangle on corner rad centres)
    chair_edge_thick:single;

    outer_jaw_face:single;   // outer jaw from gauge-face  (3.15/16" for vertical rail)

    boss_height:single;         // inches full-size up from base of chair...
    ferrule_height:single;
    bolt_head_height:single;

    boss_dia:single;
    hole_dia:single;            // ferrule dia

    key_length:single;
    key_pad_length:single;
    key_pad_taper:single;

    key_max_offset:single;


       // rail section (overall height, head width, foot width as .box file) ...

    rail_corner_rad:single;    // inches full-size  top corner rad on rail  221a
    rail_head_depth:single;    // inches full-size, from rail top down to ...
    rail_web_top:single;
    rail_web_bottom:single;
    rail_foot_depth:single;
    rail_web_thick:single;

      // S1 ordinary chairs   Standard Railway Equipment 1926  S1 ordinary chairs ...

    S1_chair_outlong:single;        // from gauge-face
    S1_chair_inlong:single;         // chair total length 14.1/2"
    S1_chair_halfwide:single;       // chair width 8" /2

    S1_seat_top_halfwide:single;    // almost full width  7.7/8"  /2

    S1_chair_crad_inner:single;     // corner rad 1"
    S1_chair_crad_outer:single;

      // bolt centres ...

    S1_inbolt_1x:single;      // y always 1.75" from end of chair
    S1_inbolt_2x:single;

    S1_outbolt_1x:single;
    S1_outbolt_2x:single;

    //-------------------

      // S1 outer jaw (2 ribs) ...

    S1_outjaw_height_top:single;       // from base


      // top x-section...

    S1_outjaw_half_rib_space_top:single;    // half-width between ribs

    S1_outjaw_rib_depth_top:single;
    S1_outjaw_rib_width_top:single;
    S1_outjaw_rib_rad_top:single;
    S1_outjaw_fillet_rad_top:single;
    S1_outjaw_depth_top:single;        // behind ribs


      // middle x-section ...

    S1_outjaw_height_mid:single;       // from base

    S1_outjaw_half_rib_space_mid:single;    // half-width between ribs

    S1_outjaw_rib_depth_mid:single;
    S1_outjaw_rib_width_mid:single;
    S1_outjaw_rib_rad_mid:single;
    S1_outjaw_fillet_rad_mid:single;

    S1_outjaw_depth_mid:single;     // behind ribs


      // seat x-section...

    S1_outjaw_height_seat:single;            // from base

    S1_outjaw_half_rib_space_seat:single;    // half-width between ribs

    S1_outjaw_rib_depth_seat:single;
    S1_outjaw_rib_width_seat:single;
    S1_outjaw_rib_rad_seat:single;
    S1_outjaw_fillet_rad_seat:single;

    S1_outjaw_depth_seat:single;     // behind ribs


      // plinth x-section ...

    S1_outjaw_depth_plinth:single;     // behind ribs

    S1_outjaw_half_rib_space_plinth:single; // half-width between ribs
    S1_outjaw_rib_depth_plinth:single;
    S1_outjaw_rib_width_plinth:single;
    S1_outjaw_rib_rad_plinth:single;
    S1_outjaw_fillet_rad_plinth:single;

    //-------------------

      // S1 inner jaw (1 rib) ...

    S1_injaw_height_top:single;       // from base


      // top x-section...

    S1_injaw_rib_depth_top:single;
    S1_injaw_rib_width_top:single;
    S1_injaw_rib_rad_top:single;
    S1_injaw_fillet_rad_top:single;
    S1_injaw_depth_top:single;        // behind rib


      // flange clearance x-section...

    S1_injaw_rib_depth_flange:single;
    S1_injaw_rib_width_flange:single;
    S1_injaw_rib_rad_flange:single;
    S1_injaw_fillet_rad_flange:single;
    S1_injaw_depth_flange:single;        // behind rib


      // upper middle x-section (top of rib) ...

    S1_injaw_height_upmid:single;       // from base
    S1_injaw_rib_depth_upmid:single;
    S1_injaw_rib_width_upmid:single;
    S1_injaw_rib_rad_upmid:single;
    S1_injaw_fillet_rad_upmid:single;
    S1_injaw_depth_upmid:single;     // behind rib


      // middle x-section ...

    S1_injaw_height_mid:single;       // from base
    S1_injaw_rib_depth_mid:single;
    S1_injaw_rib_width_mid:single;
    S1_injaw_rib_rad_mid:single;
    S1_injaw_fillet_rad_mid:single;
    S1_injaw_depth_mid:single;     // behind rib


      // seat x-section ...

    S1_injaw_height_seat:single;            // from base
    S1_injaw_rib_depth_seat:single;
    S1_injaw_rib_width_seat:single;
    S1_injaw_rib_rad_seat:single;
    S1_injaw_fillet_rad_seat:single;
    S1_injaw_depth_seat:single;     // behind rib


      // plinth x-section ...

    S1_injaw_depth_plinth:single;     // behind rib
    S1_injaw_rib_depth_plinth:single;
    S1_injaw_rib_width_plinth:single;
    S1_injaw_rib_rad_plinth:single;
    S1_injaw_fillet_rad_plinth:single;


Many of these can be left on the defaults. For customising I'm intending to use easily editable text files -- it would be a massive task to create a conventional user interface to enter so many values.

Ditto and likewise for all the other types of chair ...

cheers,

Martin.

posted: 20 Nov 2018 01:13

from:

DaveJ61
 
Northampton - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Thanks for this. This is absolutely jaw-dropping. What have I let myself in for? :-)
I'm working some strange shift patterns this week, covering for other people but I hope to look at all this in closer detail at the weekend.

Dave

posted: 24 Nov 2018 18:00

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Some progress today on the S1 inner jaw.

For dimensioning purposes I have split the inner jaw into two components. This part I have called the "stand":

2_241233_460000000.png2_241233_460000000.png

It sets the gauge-face of the rail by locating the foot of the rail*. This assumes of course that the rail foot is the same width as the head. If not there will be a gauge_tweak setting to make final adjustments to the chair position.

*vertical rail. I'm not getting into the madness of inclined rail, which in my view is unbuildable in small scale models.

Behind the stand there will be an "web insert" component in the rail web, which will be dimensioned to match the rail section, in the same way as the key on the other side. The two parts of the jaw will blend together in the final render.

The central rib on the REA drawing is more difficult to replicate simply than the ribs on the outer jaw. I think I have got somewhere near without getting too involved in 3D blends which would require dozens of additional facets. An important consideration for models is to keep it below the wheel flanges.

I'm still undecided about how to blend the side of the jaw into the chair base and rail seat. I have removed the silly angled socket in the rail seat which I had previously. On the prototype it is a spherical fillet radius, but that would involve a silly amount of code. On the outer jaw I used a few simple facets instead, although I'm not convinced it looks right. It is easier to see without the rail in place:

2_241247_000000000.png2_241247_000000000.png

This view also shows the tapered ends of the key, to ease rail threading and reduce stress on the chair.

Time for a boiled egg. :)

Martin.

posted: 24 Nov 2018 19:17

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

I'm assuming we'll be able to randomize the rotation of the screw heads :D.

Seriously, looks good. Before you go too much further, if you can generate a complete model we can easily find out file size etc. Obviously a lot of the detail you are adding won't be visible at small scales but it will be visible in the larger scales. Striking a reasonable compromise might take a bit of testing.

I think it might be possible to use Meshlab to simplify/reduce facets on a model. That might allow you to put lots of detail into your models and use Meshlab to post process the file according to the scale, but I'm probably talking rubbish. Every time I try to use Meshlab it gives me a sore head :)

Cheers,
Andy

posted: 24 Nov 2018 19:31

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
Hi Martin,

I'm assuming we'll be able to randomize the rotation of the screw heads :D.

Seriously, looks good. Before you go too much further, if you can generate a complete model we can easily find out file size etc.

I think it might be possible to use Meshlab to simplify/reduce facets on a model. That might allow you to put lots of detail into your models and use Meshlab to post process the file according to the scale, but I'm probably talking rubbish. Every time I try to use Meshlab it gives me a sore head :)
Hi Andy,

Yes, the screw heads are randomized in the file. You don't have to do anything. :)

Likewise the driven position of the keys.

Yes, I'm hoping to post a file for a plain track panel in a day or two -- can you remind me of the section dimensions of your rail?

Meshlab gave me a sore head too. In fact the whole 3D modelling/printing business gives me a sore head. That's why I'm still using the methods and geometry which I used back in the 1980s.

Any joy with your printer?

cheers,

Martin.

posted: 24 Nov 2018 19:50

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Andrew Barrowman wrote:
Hi Martin,

I'm assuming we'll be able to randomize the rotation of the screw heads :D.

Seriously, looks good. Before you go too much further, if you can generate a complete model we can easily find out file size etc.

I think it might be possible to use Meshlab to simplify/reduce facets on a model. That might allow you to put lots of detail into your models and use Meshlab to post process the file according to the scale, but I'm probably talking rubbish. Every time I try to use Meshlab it gives me a sore head :)
Hi Andy,

Yes, the screw heads are randomized in the file. You don't have to do anything. :)

Likewise the driven position of the keys.

Yes, I'm hoping to post a file for a plain track panel in a day or two -- can you remind me of the section dimensions of your rail?

Meshlab gave me a sore head too. In fact the whole 3D modelling/printing business gives me a sore head. That's why I'm still using the methods and geometry which I used back in the 1980s.

Any joy with your printer?

cheers,

Martin.
Hi Martin,

I was actually joking about the screws, but if you can do that, why not?

SMP rail nominal dimensions :

Height 1.92
Width (head and foot) 0.80
Web thickness 0.32
Foot depth 0.32
Head depth 0.60

The last two are a bit sketchy because of the radii where the head and foot meet the web but I think they are pretty close.

The replacement microcontroller board for the printer is supposed to arrive here tomorrow.

Cheers,
Andy

posted: 24 Nov 2018 20:26

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
Height 1.92
Width (head and foot) 0.80
Web thickness 0.32
Foot depth 0.32
Head depth 0.60

The last two are a bit sketchy because of the radii where the head and foot meet the web but I think they are pretty close.
Hi Andy,

Thanks. I will use those for a test file.

I've included the prototype 1:2.75 (20deg) fishing angle (as you can see on the keys) which should hopefully allow for the internal rads on the model rail. It could be increased if nec.

At present there is no location against the outer edge of the rail foot -- as on the prototype, otherwise it wouldn't be possible to drop the rail into the chair. That means location relies entirely on the key, I'm not sure how well that will work for a model. Plenty of trialling and errorring awaits. :)

cheers,

Martin.

posted: 25 Nov 2018 01:14

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Andrew Barrowman wrote:
Hi Andy,

Thanks. I will use those for a test file.

I've included the prototype 1:2.75 (20deg) fishing angle (as you can see on the keys) which should hopefully allow for the internal rads on the model rail. It could be increased if nec.

At present there is no location against the outer edge of the rail foot -- as on the prototype, otherwise it wouldn't be possible to drop the rail into the chair. That means location relies entirely on the key, I'm not sure how well that will work for a model. Plenty of trialling and errorring awaits. :)

cheers,

Martin.
Hi Martin,

I think the keys should print OK in Shapeways hi-res resin but I suspect you'll need to include a locator on the outside of the foot. I tried to do it without that, but it didn't work (for me).

Things are a bit different with a filament printer like the one I have. Unless you add some sort of support columns that will be removed later, it's not possible to have overhangs that exceed about 40% of the extruder nozzle's diameter. The slicer will build the overhang out gradually with each layer which means that the keys won't be too much like the real thing.

For home printing I've completely abandoned any suggestion of keys. The outside jaw just presents a vertical surface that contacts both the foot and head of the rail. The inside jaw presses on the rail's web to jamb the rail against the outside "jaw" while providing enough vertical force to keep the rail in the jaws. None of this resembles prototype practice in the slightest, but seems to work quite well :)

Cheers,
Andy

posted: 25 Nov 2018 01:49

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
For home printing I've completely abandoned any suggestion of keys. The outside jaw just presents a vertical surface that contacts both the foot and head of the rail. The inside jaw presses on the rail's web to jamb the rail against the outside "jaw" while providing enough vertical force to keep the rail in the jaws.
Hi Andy,

Yes, I realised you are doing that. I'm trying to create something that will scale up to the larger scales. If it produces a reasonable looking chair in 7mm scale resin I shall be well chuffed. I know it may have to be drastically downgraded for 4mm scale and/or filament printing, but I will cross that bridge when I get there. It would be easy enough to fill in under the outer half of the key.

What is the nozzle diameter on your printer? I'm just measuring the chimney to see if Father Christmas could get a 3D printer down it. :)

cheers,

Martin.

posted: 25 Nov 2018 02:08

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:

What is the nozzle diameter on your printer? I'm just measuring the chimney to see if Father Christmas could get a 3D printer down it. :)
According to my information he's pretty good at extruding himself and his kit into tight places. 0.2 mm should not be a problem.

posted: 26 Nov 2018 05:32

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
According to my information he's pretty good at extruding himself and his kit into tight places. 0.2 mm should not be a problem.
Hi Andy,

From what I'm reading a 0.2mm nozzle is needed for fine detail, but significantly slows the printing process. Do you find that the case? By how much?

I'm assuming that with a double-extruder printer, if not using the other nozzle for water sols, you could have 0.4mm or 0.6mm on one nozzle to print the timbers fairly quickly, and then use a 0.2mm nozzle on the other to print the chairs onto them? Is that a sensible idea, or have you just fallen off your chair laughing? :)

Some progress with the coding, we now have the inner jaw finished with the insert in the rail web:

1. prototype rail dimensions:

2_260016_310000003.png2_260016_310000003.png


2. SMP rail to your dims:

2_260016_310000002.png2_260016_310000002.png

Notice that the inner jaw insert is much less robust because of the thicker rail web, and the key is significantly deformed because of the excessive head depth.

You can see that better on the end views:

2_260016_310000001.png2_260016_310000001.png

2_260016_300000000.png2_260016_300000000.png

The SMP rail is significantly under scale width, the head is too deep, and the web is too thick. The deformed key which results and the weaker inner jaw is evident. It can't be beefed up much because of the need to clear wheel flanges.

I assumed a fishing angle of 1:1.5 for the SMP rail, that's just a guess, but I doubt it matches the prototype 1:2.75 fishing angle.

I suspect other model rail will be similarly compromised.

What all this means is that I think I'm ready to create a test file. I just need to add the webbing between the timbers.

cheers,

Martin.

posted: 26 Nov 2018 07:34

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Andrew Barrowman wrote:
According to my information he's pretty good at extruding himself and his kit into tight places. 0.2 mm should not be a problem.
Hi Andy,

From what I'm reading a 0.2mm nozzle is needed for fine detail, but significantly slows the printing process. Do you find that the case? By how much?

I'm assuming that with a double-extruder printer, if not using the other nozzle for water sols, you could have 0.4mm or 0.6mm on one nozzle to print the timbers fairly quickly, and then use a 0.2mm nozzle on the other to print the chairs onto them? Is that a sensible idea, or have you just fallen off your chair laughing? :)

Some progress with the coding, we now have the inner jaw finished with the insert in the rail web:

1. prototype rail dimensions:

2_260016_310000003.png2_260016_310000003.png


2. SMP rail to your dims:

2_260016_310000002.png2_260016_310000002.png

Notice that the inner jaw insert is much less robust because of the thicker rail web, and the key is significantly deformed because of the excessive head depth.

You can see that better on the end views:

2_260016_310000001.png2_260016_310000001.png

2_260016_300000000.png2_260016_300000000.png

The SMP rail is significantly under scale width, the head is too deep, and the web is too thick. The deformed key which results and the weaker inner jaw is evident. It can't be beefed up much because of the need to clear wheel flanges.

I assumed a fishing angle of 1:1.5 for the SMP rail, that's just a guess, but I doubt it matches the prototype 1:2.75 fishing angle.

I suspect other model rail will be similarly compromised.

What all this means is that I think I'm ready to create a test file. I just need to add the webbing between the timbers.

cheers,

Martin.
Hi Martin,

I needed some light relief. I spent most of the day dorking around with my printer but despite making some progress, it's still not working. There is at least a theory now, but that will have to wait until the cold light of dawn.

No, I don't think there is anything wrong with your two nozzle idea. I was thinking along similar lines but I noticed recently that, even with 0.2 mm nozzles, most of the time (I have not quantified the amount precisely) goes into printing the chairs rather than the timbers. I concluded two nozzles were not worth the effort and decided it's better to focus on things that would make the printing process as repeatable as possible.

With filament printing it's always going to take a fair bit of time and there really isn't a way to get around that. It's a serial process that requires the acceleration of massive objects. If the process is very reliable/repeatable the best way to increase productivity is probably to add another printer.

Cheers,
Andy

(And now to bed, perchance to dream.)

posted: 27 Nov 2018 02:52

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Here's a 30ft track panel in EM with 12" joint sleepers:

2_262129_120000000.png2_262129_120000000.png

I have tried a diagonal webbing between the sleepers. Several reasons for that, probably all invalid: :)

1. when cutting the web rib from the sleeper, the thickness of the knife blade will bend the rib instead of forcing the sleepers apart (as happens with the usual under-rail webs).

2. the panel is flexible for curving, and will hold the sleepers square to the curve.

3. the intersection of the web ribs makes it easy to follow a drawn centre-line when tracklaying,

4. it is easier to ballast over the webbing if desired than when it is under the rail.

This scheme is unlikely to work for more complex track formations having partial templates and shoved timbers. For those I think I shall implement a means to draw the required webbing on the screen manually.

2_262143_460000000.png2_262143_460000000.png


Here it is without rails, ready for printing: :)

2_262159_130000000.png2_262159_130000000.png

Some companies which used the wider 12" joint sleepers also used wider joint chairs on them. But not all, the ordinary S1 chairs look ok on them (as above). So I think I will leave joint chairs for another day. The L1 bridge chairs are calling me.

cheers,

Martin.

posted: 27 Nov 2018 19:33

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Excellent design. All 4 reasons for the 'odd' webbing sound valid to me but not sure about the curving bit. Surely it depends on the material having some flexibility as the angles between the webs and the sleepers ( and the 'cross' in the middle ) will need to change as the panel is curved. Having tried a few products from shapeways there doesn't seem to be much flexibility to tweak things. It needs a test I think ( and somebody with SMP rail to fit in it :D ). Gauge widening needs to be considered too at some point.

Keep up the good work.

Rob

P.S. If the material flexibility is as issue it may be good to have a jig for rail threading that engages the ends of the sleepers to prevent the them being forced apart.

P.P.S If Santa is bringing you a printer you better get your order in quick as I am having a new workbench and that will take up a lot of room on the sleigh :)


posted: 27 Nov 2018 21:46

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks Rob.

Printer ordered. I'm just hoping the reindeer aren't too exhausted after "black friday", especially if you are expecting them to bring a workbench at the same time. :)

I'm trying to get my head round the conversion to G-code for the printer -- it's hurting.

Andy was asking about some file sizes:

DXF file from Templot: 476 KB

Slic3r set to 0.05mm layers, 0.2mm nozzle, 50% fill.
______________

Using TurboCAD Deluxe I get:

STL file: 2033 KB

Slic3r GCODE file: 3109 KB -- 106,000 lines, filament usage (1.7mm dia): 1277mm. File usable?
____________

Using free software -- 3D Crafter + MeshLab, see: topic 2734 - message 25268

I get:

3DS file : 907 KB
STL file: 2031 KB, so same as TurboCAD, but

Slic3r GCODE file: failed (8KB empty file, 290 lines, filament usage 3.4mm), I need to investigate. :(

Here's the STL file from 3D Crafter/MeshLab viewed in TurboCAD Deluxe:

2_271623_570000000.png2_271623_570000000.png

Here's some info from Slic3r:

                       TurboCad    3D Crafter + MeshLaB 


facets total            49,896        49,844

errors, auto-repaired   14,365        78,081

comprising:
       facets removed       52            78
       facets added      8,320         8,346
       facets reversed   5,993        69,657

I don't know what to make of all that. Over to Andy. :)

cheers,

Martin.

posted: 27 Nov 2018 23:02

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hello Martin,
As a matter of interest what printer are you getting?
Regards.
Trevor.:)

posted: 27 Nov 2018 23:11

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin,

Santa has out-sourced some deliveries this year - he is using Parcelforce now :) Last year he had too many mince pies and couldn't manage some of the chimneys.....

Are you keeping the 'brand' of your new printer a secret ? I have thought once or twice about getting one but get the feeling it could take over my leisure time and divert me away from other modelling activities. Once you and Andy get into production it may be different though......I will just watch the two of you for now.

Rob


posted: 28 Nov 2018 00:34

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Are you keeping the 'brand' of your new printer a secret ? I have thought once or twice about getting one but get the feeling it could take over my leisure time and divert me away from other modelling activities. Once you and Andy get into production it may be different though......I will just watch the two of you for now.
Hi Trevor, Rob,

I'm reluctant to say any more about the printer just yet because I don't want anyone here to regard my choice as any sort of recommendation. My knowledge of 3D printers is absolutely nil at present.

When it has arrived and I can report actual facts about what I find in the box, I will be able to say more about it.

It's news to me that I'm going into production. :)  This will just be a new toy to tinker with, and also to confirm that the DXF files from Templot can be used successfully to create 3D-printed track. I don't like to claim a new feature for Templot until I have actually proved it myself.

cheers,

Martin.

posted: 28 Nov 2018 01:21

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin,

Comments in green, changes in Orange...



Martin Wynne wrote:
Rob Manchester wrote:
Are you keeping the 'brand' of your new printer a secret ? I have thought once or twice about getting one but get the feeling it could take over my leisure time and divert me away from other modelling activities. Once you and Andy get into testing your theories it may be different though......I will just watch the two of you for now.
Hi Trevor, Rob,

I'm reluctant to say any more about the printer just yet because I don't want anyone here to regard my choice as any sort of recommendation. My knowledge of 3D printers is absolutely nil at present. But I bet your knowledge is 100 times what is was not long ago and with your work ethic and attention to detail I am sure you made a good choice..

When it has arrived and I can report actual facts about what I find in the box, I will be able to say more about it. Hopefully all the bits will be there. I gather there may be some initial setting up and such like with most of them - assuming you haven't gone for one made of Meccano or such like.

It's news to me that I'm going into production. :)  This will just be a new toy to tinker with, and also to confirm that the DXF files from Templot can be used successfully to create 3D-printed track. I don't like to claim a new feature for Templot until I have actually proved it myself. I was trying to suggest you would want to test your new features.....rather than starting up in production ( again )...think you have done quite enough of that in the past :)

Just keep us all up to date with your progress as usual  :thumb:

Rob


posted: 28 Nov 2018 01:30

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

If you can create Gcode, that's great.

I typically use Netfabb Basic to repair the STL for Slic3r. Slic3r does have a repair option but it doesn't always work for me. BTW, there's a good chance your printer will run under Repetier-Host. You can download it now and see how the printer will lay down the filament, like this:


2983_271947_000000000.gif2983_271947_000000000.gif

The file sizes seem about right. For my 7 inch track panel:

STL from TCpro 1.4 MB
STL After Nefabb repair 1.4 MB
Gcode from Slic3r 1.37 MB
Filament required 621 mm

I believe your STL files are larger because of the number of facets, but that's not really a problem if the STL compiles to Gcode. The major difference in filament length is probably because my sleepers are only 0.9 mm thick and they not very dense.

Repetier-Host tells me my panels will print in 23 minutes but the estimation is way off. They actually take about twice that, and the major chunk of time is printing the chairs.

I noticed your layer thickness is 0.05 mm. That will increase the print time, a lot. There is an option in Slic3r that lets you specify different thickness layers, so you might use greater values for the sleepers and reduce it when you print the chairs. If your nozzle is 0.2 mm you should be able to use up to 0.15 mm layers (see pic above). I have not used the variable layer option for a long time and I can't remember exactly how to turn it on :)

Meanwhile I await ANOTHER micro-controller for my printer. The replacement bricked exactly the same way as the original. It must be the driver card. I will replace it too before my next attempt.

Cheers, from a very wet Pacific Northwet. At the moment we are on the wrong end of what they refer to as a "Pineapple Express".

Andy

posted: 28 Nov 2018 04:27

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Another thought:

The triangulation webs between the sleepers will add a lot of lateral stiffness. That's really good for turnouts and the like but probably not so good for flexible plain track.

Cheers,

Andy

posted: 28 Nov 2018 13:16

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Andy,

This is what Repetier-Host is displaying. This is using the highest precision settings which the printer is claimed to support.

Chairs from Templot in 4mm/ft and 7mm/ft scales.

I don't yet know how closely these screen previews match the finished result -- obviously a circular nozzle can never create such sharp edges. :?

2_280801_060000000.png2_280801_060000000.png

2_280802_400000000.png2_280802_400000000.png

There are some strange artifacts in the rendering there. Presumably the choice of slicing software can make a big difference to the final results at this level of detail?

Also noticeable is the dominating effect of the key in the overall design. This perhaps explains why the plastic flexi-tracks (which because of the practicalities of injection moulding can have only a minimal key) never look quite "knobbly" enough to properly represent bullhead track having wooden keys.

cheers,

Martin.

posted: 28 Nov 2018 17:06

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hello Andy and Martin,
This is beginning to look really exciting.I hope newer printers can manage to reduce the layers effect on the surface which is an unfortunate effect on earlier types.
Some of the results I have seen on Youtube lately look really smooth on surfaces of items being produced.I think 0.2mm nozzles could be a factor and if they are readily available a further improvement.
Regards.:)
Trevor

posted: 28 Nov 2018 18:24

from:

Paul Boyd
 
Loughborough - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I’m sort of half following this thread, and wondering why you don’t use someone like Shapeways to get stuff printed using the SLS (Selective Laser Sintering) process? At work we have lots of low volume plastics made which suits us wonderfully, and the finish is a lot better than anything I’ve seen using extruded plastic. Although I use a “professional” printing company for most stuff, I’ll also use Shapeways - they’re mainly a hobbyist company. It’s got to be cheaper than buying a printer if you just want track bases for a layout or two!

posted: 28 Nov 2018 18:31

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Hi Andy,

This is what Repetier-Host is displaying. This is using the highest precision settings which the printer is claimed to support.

Chairs from Templot in 4mm/ft and 7mm/ft scales.

I don't yet know how closely these screen previews match the finished result -- obviously a circular nozzle can never create such sharp edges. :?

2_280801_060000000.png2_280801_060000000.png

2_280802_400000000.png2_280802_400000000.png

There are some strange artifacts in the rendering there. Presumably the choice of slicing software can make a big difference to the final results at this level of detail?

Also noticeable is the dominating effect of the key in the overall design. This perhaps explains why the plastic flexi-tracks (which because of the practicalities of injection moulding can have only a minimal key) never look quite "knobbly" enough to properly represent bullhead track having wooden keys.

cheers,

Martin.
Hi Martin,

I don't know what's going on there. I've seen some similar effects and it usually means there will be a problem with the print. I seem to remember Meshlab had something to do with it, but that could be nonsense. You should be able to see if those protrusions are real if you examine the individual layers in plan view.

BTW1, forget what I said about the webs. I was having a Señor moment.

BTW2, have you got your can of hairspray at the ready?

Cheers,
Andy


Last edited on 28 Nov 2018 18:49 by Andrew Barrowman
posted: 28 Nov 2018 18:48

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Paul Boyd wrote:
I’m sort of half following this thread, and wondering why you don’t use someone like Shapeways to get stuff printed using the SLS (Selective Laser Sintering) process? At work we have lots of low volume plastics made which suits us wonderfully, and the finish is a lot better than anything I’ve seen using extruded plastic. Although I use a “professional” printing company for most stuff, I’ll also use Shapeways - they’re mainly a hobbyist company. It’s got to be cheaper than buying a printer if you just want track bases for a layout or two!
Hi Paul,

I have used Shapeways. The SLS samples I have are not detailed enough. I can do better with my $230 filament printer. To get better resolution you have to use Shapeways resin but it is fragile and quite expensive. There are examples in the "3D Printed Track and Turnouts" thread.

Then there's the turnaround time. If you have to make some adjustments to a model it can take several weeks to get to the final article.

Cheers,
Andy

posted: 28 Nov 2018 19:01

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Christmas has come early at Templot Towers. 3D Printer safely arrived. :)

The box arrived covered in "Fragile" stickers and no sign of any rough shunts on its journey from Lapland, so full marks to Santa. Everything was well packed inside.

It's the BIBO dual-extruder model which also has a laser engraving option, which could be very handy for modelling.

First impressions very favourable -- apart from the docs which are all on an SD card found in the bottom of the box after removing everything else, clearly labelled "Please read the user manual on the SD card BEFORE UNPACKING". :roll:

(I can't help feeling a USB stick would have made more sense. My Windows10 system doesn't have an SD card slot, so the first job was to fire up my old Windows7 system and copy the contents of the SD card to the Windows10 system.)

I will write some more when I have got it assembled.

cheers,

Martin.

posted: 29 Nov 2018 09:09

from:

Bernard Haste
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hello Paul,

Can you advise what level of resolution your professional printer achieves? I`m getting 16 microns from my one.

Many thanks,

Bernard

PS Is it possible to get even better than 16 microns presently?

posted: 9 Dec 2018 18:34

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Paul Boyd wrote:
I’m sort of half following this thread, and wondering why you don’t use someone like Shapeways to get stuff printed using the SLS (Selective Laser Sintering) process? ...

... they’re mainly a hobbyist company. It’s got to be cheaper than buying a printer if you just want track bases for a layout or two!
Hi Paul,

Well the object is not for me to get anything. I'm looking to add functions to Templot which will be helpful to Templot users in creating 3D-printed track bases. Quite a few modellers now have filament printers, and those who don't may want to send Templot-derived files to friends or to Shapeways, etc.

I long ago discovered that the only way to create a function in Templot is to try it working myself. It's just not possible to get enough sensible information from data sheets, web pages, or by asking other users on web forums, etc.

So I've obtained a 3D filament printer to tinker with. :)

cheers,

Martin.

posted: 9 Dec 2018 20:26

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Are you keeping the 'brand' of your new printer a secret ? I have thought once or twice about getting one but get the feeling it could take over my leisure time and divert me away from other modelling activities. Once you and Andy get into production it may be different though......I will just watch the two of you for now.
Hi Rob,

Andy warned me that I had a lot to learn about 3D printing, and he wasn't wrong. :(

I went for the BIBO dual-extruder printer:

http://ourbibo.com/

It has turned out to be a bit of a curate's egg. The case and axis drives seem to be well made, but the extruder head is a very agricultural thing. Likewise the work bed. I haven't tried the laser cutter/engraver attachment yet.

On the whole I feel a bit disappointed, especially with the dual extruder option. I was expecting more for my money.

In my ignorance I assumed that a dual-extruder machine would cause the not-in-use nozzle to be retracted clear of the work while the other one is in use. It doesn't.

Failing that, it is obviously essential that if the nozzles are fixed they must be on exactly the same level to within a thou or so. Otherwise one or other of them will bump into the work being laid down by the other.

As supplied one nozzle was about 12 thou lower than the other.

So there needs to be a screw adjustment to get them level. There isn't. The extruder shafts are held in the head block with allen grub screws, but to get access to them requires the extruder head to be dismantled, with no easy means to set them level, and little chance that they will still be so after re-assembly. I see one user has printed a support block to assist:

 http://www.thingiverse.com/thing:2305856

The only sensible method I can see will be to clock over them (or use feeler gauges off the bed), and then skim the required amount off the back of the longer nozzle in the lathe. The two nozzles then to be kept as a matched pair. Which is doable, but a lot more faff than I had in mind. Changing a nozzle is no quick task -- it has to be done hot, and the heater block needs to be held with one spanner while unscrewing the nozzle with another, otherwise there is a danger of breaking the heater connections.

All that having been done, the fixed dual-nozzle design still precludes any Z-moves while laying down a model. The entire model has to built up simultaneously one layer at a time.

In our case that means hopping from one timber or chair to the next, laying a single layer on it, and then hopping to the next timber, and then the next. And finally back to the first one for the next layer. Very time inefficient, and requiring the extrusion to be frequently stopped and started, and risking leaving whiskers of polymer between the timbers. It would be so much better to finish one timber or chair at a time, before moving on to the next (assuming sufficient clearance is available under the head, which is just about the case up to 7mm track bases). 

So for the present I have removed one nozzle and will regard this as a single-nozzle printer for now while I develop some code.

All this meant it took me a couple of evenings to get everything assembled and ready to work as I wanted. If anyone is looking for a plug-and-play 3D printer, this is not it. For that you need to look at models such as the Dremel Digilab range (and pay more).

All that having been said, I'm confident that I will eventually get some successful results from the BIBO, using code from Templot.

cheers,

Martin.

posted: 9 Dec 2018 21:03

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,
Thanks for your feedback on the printer. I can't say I am surprised it didn't work straight out of the box as you hoped. Feedback from others on most brands seems to imply that is normal.

Printing in layers with movement of the plate ( z-axis) between each layer isn't exactly great for track bases as you mention. Is the software not clever enough to be able to specify which chunks to print and in which order whilst avoiding banging the nozzles into previously printed stuff ?

I will be impressed when the first Templot 3D print emerges from your printer.

Rob

P.S. Just taken delivery of a large order of Exactoscale chairs....I will stick to traditional methods while you and Andy B sort out the 3D method :thumb:


posted: 9 Dec 2018 21:54

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Is the software not clever enough to be able to specify which chunks to print and in which order whilst avoiding banging the nozzles into previously printed stuff ?
Hi Rob,

The supplied software isn't. In fact it is not even able to do the standard z-slicing at the resolution I'm looking for, without creating some strange artifacts.

Which is not to say I might not be able to write something which will do that. It's doable if you just want to clear a fixed obstruction in the head, but being able to use the other nozzle subsequently on the same model is a whole new game of chess. And if you don't do that, you might just as well remove the other nozzle entirely. Or buy a single-nozzle printer in the first place!

There are several things I'm trying. For example you might want to use your Exactoscale chairs on the Exactoscale "NewTrack" bases -- I see the bases and a few sizes of the kits are available again:

 http://exactoscale.com/track-components/track-pricing/

But of course that doesn't include a B-7.5 curved crossover, or whatever other custom formation you might want. So the Templot "exactopips" function might come in handy: :)


2_091641_110000000.png2_091641_110000000.png

I have also found that printing reasonably thick timbers takes far too long, so I want to try printing chairs (or exactopips) directly onto ply or limewood or plastic timbers, placed into some sort of jig or locator printed first.

Lots of ideas to try, but I fear the as-supplied software isn't going to be much use for them and I shall have to write something myself. I have already been editing the g-code manually to get anything to work at all. The software would be fine if I needed a Toby Jug.

cheers,

Martin.

posted: 9 Dec 2018 23:10

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Yes using the new technology with Exactoscale or other available products could be a good plan too.

I am a fan of thin timbering as they generally suit the kinds of layout I am likely to model where well tended ballast isn't the norm. Think goods depots, loco sheds, industrial sites and such like. I tend to favour plastikard timbers with thickness to suit - the 0.020"(0.5mm) is good but it precludes printing the bases with webbing between the timbers. It would be good to come up with a method for building trackwork formations at the workbench that could be moved into position on the baseboard but at this thickness the track isn't too robust until finally glued down. I am not a fan of leaving the templates in place on the layout.

Rob




posted: 9 Dec 2018 23:30

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
I tend to favour plastikard timbers with thickness to suit - the 0.020"(0.5mm) is good but it precludes printing the bases with webbing between the timbers.
Hi Rob,

How about printing thin sleepers and chairs directly onto a sheet of plastikard, or a sheet of plywood, or some other sheet material? Assemble on the bench and transfer the whole thing to the baseboard.

The BIBO printer bed is limited to 8" x 6", so it would need to be done in sections and joined up in some way. A strip of plastikard or ply glued under the joint could drop into a slot in the cork underlay.

cheers,

Martin.

posted: 10 Dec 2018 00:36

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Yes, that could work. Would the printed bits bond to the plastikard or ply sheet ?

I quite like the 'jigsaw' method of layout construction that Vincent de Bode described in RailMODEL Digest ( and other places ) a good while back. Basically a flat topped or open frame supporting structure with trackwork and scenic modules interlocking together to form the overall visible scene. You can work on each scenic or track 'module' in isolation at the workbench and it is easy to disguise the baseboard joints ( which spoil many a good layout in my view ). Iain Rice was also involved in the promotion of such ideas. It also gives the possibility of having different modules for some of the scenic areas that can be interchanged to reflect a different time period which I quite like.

Rob


posted: 10 Dec 2018 02:56

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Would the printed bits bond to the plastikard or ply sheet ?
Hi Rob,

I've no idea, but as we are talking about hot molten plastic, I would think they have a good chance of bonding. There are several variables to try, including type of polymer, temperature of first layer, and feed rates. If plastikard I would think molten ABS polymer would bond strongly.

If printing onto such materials the fragile glass bed plate could be discarded, and the nozzle could be driven hard against the sheet for the first layer, against the bed levelling springs. Molten plastic by such means would be injected into the grain of plywood, or the hot nozzle would melt into the surface of plastikard (he said, without trying it or having the faintest idea if any of this would work).

If it turned out that the bond is not too strong, solvent could perhaps be brushed around the printed parts afterwards (butanone for ABS, acetone for PLA).

If printing ABS onto plywood, it might be possible to prime the plywood with Pipeweld, as in the original Exactoscale docs.

Lots of stuff to try. :)

cheers,

Martin.

posted: 10 Dec 2018 16:53

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thank you Martin,

Rob


posted: 10 Dec 2018 18:20

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Rob Manchester wrote:
Are you keeping the 'brand' of your new printer a secret ? I have thought once or twice about getting one but get the feeling it could take over my leisure time and divert me away from other modelling activities. Once you and Andy get into production it may be different though......I will just watch the two of you for now.
Hi Rob,

Andy warned me that I had a lot to learn about 3D printing, and he wasn't wrong. :(

I went for the BIBO dual-extruder printer:

http://ourbibo.com/

It has turned out to be a bit of a curate's egg. The case and axis drives seem to be well made, but the extruder head is a very agricultural thing. Likewise the work bed. I haven't tried the laser cutter/engraver attachment yet.

On the whole I feel a bit disappointed, especially with the dual extruder option. I was expecting more for my money.

In my ignorance I assumed that a dual-extruder machine would cause the not-in-use nozzle to be retracted clear of the work while the other one is in use. It doesn't.

Failing that, it is obviously essential that if the nozzles are fixed they must be on exactly the same level to within a thou or so. Otherwise one or other of them will bump into the work being laid down by the other.

As supplied one nozzle was about 12 thou lower than the other.

So there needs to be a screw adjustment to get them level. There isn't. The extruder shafts are held in the head block with allen grub screws, but to get access to them requires the extruder head to be dismantled, with no easy means to set them level, and little chance that they will still be so after re-assembly. I see one user has printed a support block to assist:

 http://www.thingiverse.com/thing:2305856

The only sensible method I can see will be to clock over them (or use feeler gauges off the bed), and then skim the required amount off the back of the longer nozzle in the lathe. The two nozzles then to be kept as a matched pair. Which is doable, but a lot more faff than I had in mind. Changing a nozzle is no quick task -- it has to be done hot, and the heater block needs to be held with one spanner while unscrewing the nozzle with another, otherwise there is a danger of breaking the heater connections.

All that having been done, the fixed dual-nozzle design still precludes any Z-moves while laying down a model. The entire model has to built up simultaneously one layer at a time.

In our case that means hopping from one timber or chair to the next, laying a single layer on it, and then hopping to the next timber, and then the next. And finally back to the first one for the next layer. Very time inefficient, and requiring the extrusion to be frequently stopped and started, and risking leaving whiskers of polymer between the timbers. It would be so much better to finish one timber or chair at a time, before moving on to the next (assuming sufficient clearance is available under the head, which is just about the case up to 7mm track bases). 

So for the present I have removed one nozzle and will regard this as a single-nozzle printer for now while I develop some code.

All this meant it took me a couple of evenings to get everything assembled and ready to work as I wanted. If anyone is looking for a plug-and-play 3D printer, this is not it. For that you need to look at models such as the Dremel Digilab range (and pay more).

All that having been said, I'm confident that I will eventually get some successful results from the BIBO, using code from Templot.

cheers,

Martin.
Hi Martin,

That's not good. Sounds like you are paying a lot of money for the supplier to build the printer in a nice case. I suspect the double extruder arrangement would only work if you were printing something with a small X dimension. I suppose you could try orienting a piece of track along the Y axis to see if that might work.

I've considered using two printers with different nozzles. One to print the timbers and the other to print the chairs, but that would require some reliable indexing fixture. I think it might be possible to transfer the glass plate with the timbers still attached to the "fine" nozzle printer, but it would have to be done quickly before the glass cooled down too much.

Cheers,
Andy

posted: 10 Dec 2018 20:38

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
I suspect the double extruder arrangement would only work if you were printing something with a small X dimension. I suppose you could try orienting a piece of track along the Y axis to see if that might work.
Hi Andy,

The nozzles are at 33mm centres. So that would work in 4mm scale only for plain track, and only for 00 gauge (32mm sleepers) and not for EM or P4 (34mm sleepers). It also limits the length of the track panel to 6" long in the Y dimension. It would however mean that two identical track panels could be printed at the same time, if the whole thing is to be printed with one nozzle size.

I have come to the conclusion that the nozzle size is not as important as it might seem, if the g-code is calculated accordingly. For example a narrow slot can be printed with any nozzle size:

2_101524_090000000.png2_101524_090000000.png

The nozzle size obviously affects external corners and the width of ribs, but I have found that a 0.4mm nozzle can create surprisingly sharp edges (0.2mm radius).

This of course is the opposite of my experience with CNC 3D milling, where sharp corners are easy and narrow slots are difficult.

However, if we try extruding in 3 directions rather than in 2D layers, a different dimension comes into play -- the diameter of the tip of the nozzle itself rather than the size of the hole in it, and the taper angle on the side of the nozzle. I can see some lathe work on the horizon. :)

cheers,

Martin.

posted: 11 Dec 2018 08:01

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Andy,

A thought occurs (just the one?). :)

If the bed is tipped to match the angle across the nozzle tips, and the g-code is edited to add a Z-component to every X move, it would be possible to print from both nozzles even when they are not level:

2_110051_380000000.png2_110051_380000000.png

Angle exaggerated above. If the actual angle is say 0.5mm across 33mm, that's only 3mm across the width of the bed and possibly within range of the bed levelling springs. That would be a lot easier to set up with feeler gauges than fiddling about trying to get the nozzles exactly level.


Changing the subject, I think (if I'm not missing something blindingly obvious) that I was wrong and it would be possible to build the timbers and chairs one at a time, rather than in layers across the full model, even using fixed dual nozzles, provided:

a. all the timbers are first laid down unchaired one at a time working from left to right in the diagram above, using the left nozzle (say 0.6mm), with the right nozzle always in fresh air, and

b. the chairs are then built on top of the timbers one at a time working back in the opposite direction from right to left, using the right nozzle (say 0.2mm), with the left nozzle then above the empty timbers. If it remains hot it wouldn't do too much damage as it passes back over them while the first layer is being laid on each chair.

cheers,

Martin.

posted: 11 Dec 2018 17:42

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Reflecting while chomping my breakfast cereal it struck me that your initial idea of tuning the nozzles to lie on the same plane (shim washers might work too) is probably the best place to start. If they are on the same plane within a thou or two it's not really any different from what happens with a single nozzle when it traverses a previously printed layer.

With Slic3r there is the option of lifting the extruder when the filament is retracted but I don't even bother to do that. I've tried it but other than extending the print time it didn't seem to make any difference, so you might just try tweaking the nozzles and see what happens.

Regarding printing timbers in one go, to prevent the model turning into a molten blob I think each layer has to cool enough before you can add another layer. I've no idea exactly how much but I do know I had to add a fan to cool the model on my printer to get it to print reasonable chairs. Your printer probably has one already. (It will definitely have extruder fans but that's different.)

Cheers,
Andy

posted: 11 Dec 2018 18:41

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
Regarding printing timbers in one go, to prevent the model turning into a molten blob I think each layer has to cool enough before you can add another layer. I've no idea exactly how much but I do know I had to add a fan to cool the model on my printer to get it to print reasonable chairs. Your printer probably has one already. (It will definitely have extruder fans but that's different.)
Thanks Andy.

There are 4 fans on the extruder head. One on each extruder gear block, and one front and back blowing across the nozzles. Are they the ones you mean? I assumed they are there to stabilise the nozzle temperature, but they would also have the effect of cooling the model.

The extruder fans run continuously, regardless of whether the extruder is in use. The nozzle fans stop and start, or run at variable speed, during the print -- GCODE:  M106 [S{Fan speed (0..255)}]

Fine tuning the nozzles is tricky because they are difficult to remove and replace in exactly the same position -- I think there must be traces of polymer in the screw thread in the heater block. On the whole I think I might prefer to make any compensation after they are bolted tight.

If the intention is to use only flat layers, a simple method would be to attach some abrasive paper to the bed (levelled) and push the head to and fro across it with the (brass) nozzles in firm contact. The size of the nozzle tip diameter (as opposed to the size of the hole in it) should have no effect, or very little effect, if there is never any part of the model above the nozzle level.

cheers,

Martin.   

posted: 11 Dec 2018 21:47

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Yes, that's the fans. They are for cooling the filament after it has been applied. The amount of cooling depends on the feature size. Sometimes they are not used at all.

Mind how you go with shaving the nozzles. The small diameter part is very short. I made one once but I don't remember how short.

Andy
Last edited on 11 Dec 2018 21:48 by Andrew Barrowman
posted: 12 Dec 2018 15:39

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi,
I have taken the plunge and ordered a cheap DIY one (spec looks good enough to me, wide range of materials and decent nozzle size and temperature range etc. We shall see what it is capable of in a few days time.

posted: 12 Dec 2018 17:32

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

You might not need to modify the nozzles. If your printer is like mine you should be able to rotate the threaded tube to adjust the height. I could be wrong but I seem to remember the tube is prevented from rotating by a grub screw.

Andy

posted: 12 Dec 2018 18:09

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
I could be wrong but I seem to remember the tube is prevented from rotating by a grub screw.
Hi Andy,

It is. But the grub screws are inaccessible without dismantling the extruder head. After re-assembly, the alignment is not guaranteed.

cheers,

Martin.

posted: 12 Dec 2018 18:25

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Andrew Barrowman wrote:
I could be wrong but I seem to remember the tube is prevented from rotating by a grub screw.
Hi Andy,

It is. But the grub screws are inaccessible without dismantling the extruder head. After re-assembly, the alignment is not guaranteed.

cheers,

Martin.
Hi Martin,

You could ditch the grub screws and run a lock-nut up the tube? (That would be possible on my printer.)

Cheers,
Andy



posted: 12 Dec 2018 18:44

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Should be a M6-1.0

posted: 12 Dec 2018 20:57

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks Andy.

Correction - the extruder tube is a plain shaft. Only the business end is threaded (M6) to carry the through-tapped heater block. The nozzle screws into the heater block and locks against the end of the tube.

However, first the good news -- I have now modified the extruder head so that (after removing the front fan assembly) I can access the grub screw by passing an Allen key through the blades of the extruder fan. This means that I can adjust the relative nozzle levels with the extruder head in situ.

The result is that I have now installed and levelled two new nozzles, 0.6mm and 0.2mm (instead of both original 0.4mm), and both are feeding polymer successfully.

Now celebrating with a boiled egg. :)

The bad news -- two burnt fingers. :(

cheers,

Martin.

posted: 12 Dec 2018 21:01

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:

The bad news -- two burnt fingers. :(

cheers,

Martin.
From testing the new nozzles or picking the egg out of the pan ? :D

Bet you are glad you got a printer that needed some tinkering.....

Rob

posted: 12 Dec 2018 21:24

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
From testing the new nozzles or picking the egg out of the pan ? :D

Bet you are glad you got a printer that needed some tinkering.....
Hi Rob,

Boiling an egg I can do. 3D printers, don't ask. :)

I'll leave the being glad until I have actually printed something with the new nozzles.

How long I wonder before we can buy a home 3D printer with Epson or Canon on the label instead of Shaoxing Bibo Automatic Equipment Co. Ltd.? To be fair, they do promise to send a spare nozzle and thermistor as compensation if they don't reply to an email within 12 hours. I don't recall such a response from Epson or Canon. But there is an obvious clue there -- they expect you to have the thing in bits most days before breakfast.

cheers,

Martin.

posted: 12 Dec 2018 22:54

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Boiling an egg I can do. 3D printers, don't ask. :)

I'll leave the being glad until I have actually printed something with the new nozzles.

How long I wonder before we can buy a home 3D printer with Epson or Canon on the label instead of Shaoxing Bibo Automatic Equipment Co. Ltd.? To be fair, they do promise to send a spare nozzle and thermistor as compensation if they don't reply to an email within 12 hours. I don't recall such a response from Epson or Canon. But there is an obvious clue there -- they expect you to have the thing in bits most days before breakfast.

cheers,

Martin.
Hi Martin,
 3D printers won't become mainstream tech and available from Argos and John Lewis with Epson or Canon badges on them until the user knowledge level is considerably reduced. Mainstream products for a consumer audience have to be virtually plug&play and have ink that can be just slotted in when required. They need to be understandable by anyone ( although Apple have managed for years :?)

It seems difficult to analyze the actual cost of 3D printed items. Andrew(IIRC) has explained in another topic the cost model used by Shapeways for pricing parts but there doesn't seem to be consistency to me. Home printing cost calculations need to factor in the purchase price of the printer ( it will unlikely be worth selling 3 years down the line ) and even maybe take account of the time spent leveling the hotplate and making adjustments to nozzles and the like - as you are finding out. But hey, you didn't really think you would have track bases with chairs ready to thread with rail in the first week ?

Hope you are taking pics of all the work you are doing on this. you may even be able to teach the guys at Peco how to make 3D models even if you are a bit too late for a wider 4mm scale track gauge:D

Rob


posted: 12 Dec 2018 23:36

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
3D printers won't become mainstream tech and available from Argos and John Lewis with Epson or Canon badges on them until the user knowledge level is considerably reduced. Mainstream products for a consumer audience have to be virtually plug&play and have ink that can be just slotted in when required.
Hi Rob,

The Dremel range seems to be aiming at that level -- the filament spools have an RFID chip in them which automatically sets all the relevant printer parameters for the type of polymer loaded:

 http://digilab.dremel.com/products/3d20

 http://digilab.dremel.com/products/3d45

But hey, you didn't really think you would have track bases with chairs ready to thread with rail in the first week?

No, I was mindful that Andy's topic on his experiments has been running for more than 3 years. :?

To be clear, unlike Andy I'm not actually wanting track bases for myself. My railway interest is to add functions to Templot which can be used to do that.

My own hobby interest, and the reason I decided to spend the money, is to create 3D versions of OS 25K Explorer maps. The BIBO workbed can accommodate 4x3 km grid squares, so a reasonable map area could be built up with a dozen or so such panels:

 http://www.ordnancesurvey.co.uk/blog/2015/06/building-a-3d-map-of-london/

cheers,

Martin.

posted: 13 Dec 2018 00:10

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Three years! My goodness. How time flies.

Martin and Stephen,

If not already obtained, get a can of hair spray (the cheaper the better). A light dusting on the glass bed really helps to keep the model stuck down.

Initially you might try printing on blue masking tape with the bed heater off. Just give the surface of the tape a light rub with very fine sandpaper between prints.

At least that's what I've been doing, mainly with PLA filament. I've never tried printing with ABS - I run my printer in the house and I understand the fumes from ABS are a bit unpleasant.

Happy printing!

Andy

posted: 13 Dec 2018 00:11

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Hi Rob,

The Dremel range seems to be aiming at that level -- the filament spools have an RFID chip in them which automatically sets all the relevant printer parameters for the type of polymer loaded:

My own hobby interest, and the reason I decided to spend the money, is to create 3D versions of OS 25K Explorer maps. The BIBO workbed can accommodate 4x3 km grid squares, so a reasonable map area could be built up with a dozen or so such panels:

 http://www.ordnancesurvey.co.uk/blog/2015/06/building-a-3d-map-of-london/

cheers,

Martin.
Hi Martin,

The Dremel filament RFID chip idea reminds me of inkjet printers that have them. The non-OEM ink cartridge sellers either reverse engineered their own or had you lever the old chip off and attach it to the new cartridge. The Dremel website kindly explains that the warranty is invalid if you use any other kind of filament.

The 3D mapping looks good. I was going to ask what you do about low lying areas but as we only go as low as Holme Fen near Cambridge at 9 foot below sea level it isn't likely to be a problem. I presume hilly areas will just have a printed surface rather than being solid down to sea level ?

Rob

posted: 13 Dec 2018 00:58

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
If not already obtained, get a can of hair spray (the cheaper the better). A light dusting on the glass bed really helps to keep the model stuck down.
Hi Andy,

The BIBO came with a "lipstick" of Pritt-type (actually chinese equivalent) adhesive stick:

 http://www.amazon.co.uk/Pritt-Stick-Original-Glue-Childproof/dp/B004QFI99Y

It seems to work fine. Just a light rub over the work area and the PLA sticks firm (a bit too firm).

The great advantage is that it is water soluble. Run the glass plate under the tap and it's clean ready for next time. I'm not sure if hairspray cleans up that easily?

cheers,

Martin.

posted: 13 Dec 2018 01:52

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
The 3D mapping looks good. I was going to ask what you do about low lying areas but as we only go as low as Holme Fen near Cambridge at 9 foot below sea level it isn't likely to be a problem. I presume hilly areas will just have a printed surface rather than being solid down to sea level ?
Hi Rob,

I don't "do" anything yet. As you pointed out, I've had the printer for only a week. :)

I'm planning that the fill under the hills will be the usual 10% or whatever 3D grid fill structure. Or with dual nozzles it could be a water-sol support, to be washed out and replaced with something more long-term solid (Polyfilla?).

50K terrain contours are available free from OS:

 http://www.ordnancesurvey.co.uk/business-and-government/products/terrain-50.html

So 25K is probably the largest I can go without the result looking "steppy". I would like to go larger, but the much preferable 5K terrain data is a paid-for product.

It would be great to go right up to the 25" maps, or at least the 6", (as on NLS) but it would be a lot of work. The 25" maps don't normally have contours. Some, but not all, of the 6" maps have contours, but shown chain-dotted so would need to be manually traced. The osmaps "Standard" 10K map has solid contours, but shown very faint. Decisions, decisions.

But at present this is all NOD stuff. For the present I need to get the printer going, and get on with the task of coding all the special switch and crossing chairs for Templot.

cheers,

Martin.

posted: 13 Dec 2018 02:25

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Andrew Barrowman wrote:
If not already obtained, get a can of hair spray (the cheaper the better). A light dusting on the glass bed really helps to keep the model stuck down.
Hi Andy,

The BIBO came with a "lipstick" of Pritt-type (actually chinese equivalent) adhesive stick:

 http://www.amazon.co.uk/Pritt-Stick-Original-Glue-Childproof/dp/B004QFI99Y

It seems to work fine. Just a light rub over the work area and the PLA sticks firm (a bit too firm).

The great advantage is that it is water soluble. Run the glass plate under the tap and it's clean ready for next time. I'm not sure if hairspray cleans up that easily?

cheers,

Martin.
Thanks Martin,

I'll need to give that a shot.

(I use meths to remove the hairspray. Sometimes I don't add anything to the glass bed. I just get it squeaky clean with glass cleaner.)

Cheers,
Andy

posted: 13 Dec 2018 12:28

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
The one I have ordered is a DIY job hence the low price tag but I shall report back in due course.

posted: 13 Dec 2018 13:10

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Stephen Freeman wrote:
The one I have ordered is a DIY job hence the low price tag but I shall report back in due course.
Thanks Stephen.

Is this for track bases/chairs or general modelling? I'm thinking that for chaired track construction (no soldering nearby) it would be possible to have 3D printed track gauges. The PLA polymer seems to be immune to most track building solvents.

p.s. don't forget the nozzle is hot. :)

cheers,

Martin.

posted: 13 Dec 2018 15:50

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi,

No my interest is only chairs at the moment, I have a different solution for gauges.

posted: 14 Dec 2018 20:24

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Hope you are taking pics of all the work you are doing on this.
Hi Rob,

I will try to report the good bits and hide the bad bits. :)

Here are the first prints with the 0.6mm nozzle for the timbers (this is 0 gauge). I want to get the timbers looking reasonable before I try printing chairs on them.

I set the timbers 2.75" thick (1.6mm), which is about half most 0 gauge flexi-track, but anything thicker will take a fortnight to print.

2_141435_020000000.jpg2_141435_020000000.jpg

On the left was printed using 0.15mm layers. The ribs are too flimsy and distorted in removing it from the workbed.

On the right I changed to 0.2mm layers for faster printing, and I beefed up the ribs (which slowed it back down).

The results are very accurate for size (59.5 x 5.8 x 1.6 mm), and with surprisingly sharp edges (at least to me). The biggest problem is the knobbly top surface:

2_141446_140000000.jpg2_141446_140000000.jpg

It's not a bad as that close-up, and would probably be mostly hidden under paint. But I would like to get it better if I can. Two ideas to try:

1. drop the nozzle temperature a fraction, and then run it back over the surface in a random zig-zag pattern (without extruding) to smooth out the surface and maybe add a bit of (overscale) wood-grain.

2. use the 0.2mm nozzle to add a thin 0.05mm skin over the top. That would be a bit more time consuming, but it would be only the one layer.

These samples each took about 20 minutes to print, i.e. 5 minutes per timber. The volume of each timber is 552mm^3. The printer spec claims to have a max capacity of 12mm^3 per second. So flat-out a timber would take 552/12 = 46 seconds. A realistic target is probably about 2 or 3 times that, say 2 minutes per timber. So I should be able to double the speed from current 5 minutes when I know what I'm doing.

These were done as-sliced, i.e. hopping from timber to timber for each layer. It's painful to watch. Next I want to try printing each timber one at a time. A problem with that is what to do about the ribs, I may need to change the diagonal design.

The printer came with spool of translucent white PLA filament, which made it impossible to see the details. I have obtained some coloured filament, and at the same time I'm trying the PLA PLUS polymer. This has some (unspecified) additive to make it tougher and more flexible. It certainly seems stronger than the white PLA, and I thought the extra flexibility might be a benefit when threading rail through the chairs.

Andy -- it also has the advantage that it sticks quite firmly to the glass plate (at 50degC) without needing any adhesive. The ordinary PLA didn't do that. The extrusion temperature is a bit higher, I used 215degC for these samples.

When I have got it all working reasonably well with PLA I want to try ABS, so that moulded plastic chairs could be attached with solvent if needed for special formations.

So many different options and settings to try. :)

cheers,

Martin.

posted: 14 Dec 2018 23:52

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks Martin,

Good to see some printed samples. The idea of a thinner top layer could be interesting to try. I have often looked at 3D printed items for sale and wondered about the rough surface finish they exhibit. Presumably the flow rate through the nozzle, the steps between adjacent passes and the temperature of the filament are all variable and maybe it is just a case of finding an optimum set of values for better surfaces ?

Maybe I have missed this somewhere along the line but what are the reasons not to print in ABS ? Compatibility with C&L/Exactoscale products would certainly be an advantage although not of course required if you were printing plain track panels with chairs already in place.

Rob


posted: 14 Dec 2018 23:58

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Looks good! Did you generate the G-code or was it Slic3r? If it's Slic3r you can alter the texture by changing the direction of the top layers. I usually align the "with the grain". Also, how dense are the sleepers?

Your filament might be the same stuff I'm using. It's called PLA/PHA.

According to Colorfabb - "We've developed a new type of filament, it's a blend of PLA and PHA. Polyhydroxyalkanoate (PHA) is an biodegradable polymer produced by bacterial fermentation of sugar or lipids."

Anyway, it is a bit better than straight PLA.

Cheers,
Andy

posted: 15 Dec 2018 00:08

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Rob,

Printing ABS requires higher extruder temperatures and it gives off unpleasant fumes. Not the best thing to have in your dining room, although I suppose you could use the extractor fan above the cooker when domestic authorities are out.

Cheers,
Andy

posted: 15 Dec 2018 00:55

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Andy,

Thanks for your input. "Biodegradable polymer" ? Does that imply printed items have a limited life-span or it is so long in the future we don't need to worry.

I remember now reading of the vapours with ABS printing. I was thinking it may give more flexibility to items such as chairs to allow easier threading and not exhibiting the brittleness you had commented on with other materials. I get the feeling that Martin isn't using the dining room or the kitchen for his track experiments :D although the smells and possibly harmful vapours still need treating with respect.

Rob


posted: 15 Dec 2018 01:02

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
Did you generate the G-code or was it Slic3r?

Your filament might be the same stuff I'm using. It's called PLA/PHA.
Hi Andy,

It was sliced with Cura, which is the default in the software supplied with the printer. But I've been comparing the Cura-generated code with Slice3r -- it's completely different, and Slic3r has far more options. So I have to try them all to find the best for our purpose.

What's obvious however is that none of them are optimized for railway track -- they assume you want to print a toby jug or an alien invader. I've been editing the G-code manually to get what I want, and my thoughts now are that for the timber base the best option may be to generate the G-code directly in Templot instead of using the DXF/STL/slicing route. I definitely want to print one timber at a time, not sliced across the entire model.

The filament I'm using is:   

 http://esunchina.net/products/142.html

Whether that's the same as yours I have no idea. All the suppliers seem to have a PLA Plus option, without much explanation of what's in it. :?

cheers,

Martin.

posted: 15 Dec 2018 02:21

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks Martin.

In Slic3r the top layer direction is determined under Print Settings - Infill, "Fill Angle". Try it either at zero or 90 degrees (depending on the number of layers).

Yes, printing track can be a painfully slow process. I don't mind so much with turnouts because they are a lot of work to make by hand. Despite that, I have many yards of printed plain track bases. Once you get everything dialed in properly you can launch a print and go and do something else.

I must admit when I started trying to print turnout bases I had little confidence that it was going to work at all. I bought my first printer on a whim and the track idea came later. Now I'm convinced it really is quite practical, even with the long print times. The one thing that surprises me slightly is that very few people have jumped on it, but each to his/her own.

Cheers,
Andy

posted: 15 Dec 2018 02:34

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Biodegradable polymer" ? Does that imply printed items have a limited life-span or it is so long in the future we don't need to worry.
Hi Rob,

As Martin pointed out the other day I've been dorking around with this for several years. I haven't noticed any degradation in the things I printed several years ago. They are certainly not falling apart, but I should probably take a more serious look and do a bit of testing.

I've mainly been using PLA and PLA/PHA. ABS would be different and probably last longer, but it probably degrades over time too.

Cheers,
Andy

posted: 15 Dec 2018 03:23

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
I get the feeling that Martin isn't using the dining room or the kitchen for his track experiments :D
Hi Rob,

Not the kitchen, but it needs to be indoors in the warm and dry. I managed to clear a bit of space:


2_142144_210000000.jpg2_142144_210000000.jpg

The PLA polymer is hygroscopic (as are many others). As with injection moulding, if it gets damp steam will be created in the machine -- with very detrimental effects on quality (and even danger, in some cases).

As supplied, the filament is vacuum-sealed with silica-gel. But once hanging on the back of the machine, it is difficult to reseal. You don't want to cut individual lengths for a specific job, because starting a fresh end through the extruder is a lot of hassle (at least it is on this machine). But a 1kg spool could last for several months, so keeping the machine in the warm and dry is important. At present I'm removing the spool after each job and sealing it as best I can in a plastic bag, while the filament remains attached to the printer.

Making a proper sealed spool holder/dispenser has been added to my NOD list:

 http://www.amazon.com/Polymaker-PolyBox-Filament-Filaments-Printing/dp/B075DBPY6F

There is another reason for that with the BIBO -- hanging 2Kg on the back of the printer case is quite a heavy weight, and might long-term distort it (it's made of aluminium/plastic laminate). Bearing in mind that it carries the drive shafts and slidebars directly, a distorted case would be bad news.

PLA appears not to create any offensive smells or fumes, and it hasn't triggered the smoke alarm. ABS when hot has the usual smell of burning plastic, and would need some form of extractor if used indoors.

cheers,

Martin.

posted: 15 Dec 2018 05:44

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I won't be posting any pics of my printing "workspace" :)

posted: 15 Dec 2018 19:57

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
I won't be posting any pics of my printing "workspace" :)
My printing workspace won't look like that photo for long. :)

I can report some success. These sleepers have been given a top skin to hide the knobbly surface - I stopped after one and a half sleepers to see the difference:

2_151426_330000000.jpg2_151426_330000000.jpg

2_151429_060000000.jpg2_151429_060000000.jpg

It was done using the 0.6mm nozzle and added only about one minute to the print time per sleeper.

That's a cruel enlargement (5.8mm wide sleepers), but it's very much smoother -- the difference is quite noticeable at normal layout-viewing distance.

I created an extra 0.05mm (2 thou) layer of polymer, 100% filled, and extruded it over the top of the previous layer without actually changing the Z position of the nozzle.

The result is that it has filled the voids in the surface, but also spilled over the edges a fraction and created a rough top edge on the sleeper. I can probably fix that by removing the outer shell wall from the layer. There is no change in the measured thickness of the sleeper.

It's time these sleepers gained some chairs. :)

cheers,

Martin.

posted: 15 Dec 2018 21:48

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
You wouldn't believe the large number of parts, could take me at least a week to put it all together if I 'm lucky!
Stephen Freeman wrote:
Hi,

No my interest is only chairs at the moment, I have a different solution for gauges.


posted: 15 Dec 2018 22:07

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Stephen,

A before and after picture of your printer build would be good :)

Have fun

Rob


posted: 15 Dec 2018 23:30

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin,

I count at least 12 books on your little shelf that are also in my collection and that includes the Ilford Manual of Photography of course :) Oh, messing about in the darkroom, that takes me back.

Rob


posted: 15 Dec 2018 23:49

from:

madscientist
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I must admit when I started trying to print turnout bases I had little confidence that it was going to work at all. I bought my first printer on a whim and the track idea came later. Now I'm convinced it really is quite practical, even with the long print times. The one thing that surprises me slightly is that very few people have jumped on it, but each to his/her own.


I think there are two reasons , one, is that 3D printing is a minority sport amongst railway modellers and two, track building is even a more minority sport and is already fairly well catered for by specialist suppliers and long stablished processes

PS I have an Ender 3 , ( an i3 clone )

posted: 16 Dec 2018 01:07

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
I count at least 12 books on your little shelf that are also in my collection and that includes the Ilford Manual of Photography of course :) Oh, messing about in the darkroom, that takes me back.
Hi Rob,

Yes, I can smell the fixer now. Remember changing bags? :)

The house is full of books. I don't know how the Ilford manual got on that shelf, it should be on the photography shelf. I've just checked inside the cover -- purchased Aug 1969, which by coincidence is the exact same date as those slate quarry pics.

Your eyes must be better/younger than mine if you can read the book titles, or are you just recognising the covers? Next time I can't find a book, I'll ask you if you know where it is. :)

I did have to do quite a lot of clearing to make space for the BIBO -- it won't look like that for long.

cheers,

Martin.

posted: 16 Dec 2018 01:38

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
madscientist wrote:
I think there are two reasons , one, is that 3D printing is a minority sport amongst railway modellers and two, track building is even a more minority sport and is already fairly well catered for by specialist suppliers and long stablished processes

PS I have an Ender 3 , ( an i3 clone )
That sounds about right. If I ever get around to completing a decent part of a layout with printed track it might get more attention, but I don't really care either way. Meanwhile printing operations are suspended here. Our daughters are arriving on the 27th with three young gents who will be extremely disappointed if there are no trains running on grampa's railway. I have at least managed to lay a gigantic dumbbell loop on the baseboard. Trouble is I keep playing with all the locos and rolling stock that's been sitting around in boxes for years :)

"Oooo! Look at this one. Wonder when I bought that?"
Last edited on 16 Dec 2018 01:43 by Andrew Barrowman
posted: 16 Dec 2018 18:25

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi
Photo of finished/nearly finished not a problem but I'm not about to take it all apart for a photo! I can however show you the empty box :D

I do feel that it will be best to make the control box connectable rather than fully permanently connected as it will be a bit unwieldy otherwise. I do have more nuts and bolts left over than I appear to need but of course am one short of a specific size :(.  I will have to see if I have any spare in the garage.

It will probably take me the rest of the coming week at least to get it fully usable.

The fun will start when I try to actually use it!

Rob Manchester wrote:
Hi Stephen
A before and after picture of your printer build would be good :)

Have fun

Rob


posted: 17 Dec 2018 02:28

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
The first chairs! These are 7mm scale. They don't look too bad at normal viewing distance:

2_162041_280000000.jpg2_162041_280000000.jpg

But not so good when zoomed in:

2_162043_010000000.jpg2_162043_010000000.jpg

Perhaps I should have started with Gauge 3. :)

With Peco gauge 0 rail:

2_162045_030000000.jpg2_162045_030000000.jpg

Don't look so close! This is what they would look like on a layout:

2_162049_200000000.jpg2_162049_200000000.jpg

And this is what they were intended to look like: :)

2_162113_370000000.png2_162113_370000000.png

It's clear how the underscale Peco rail has prevented a proper key shape. That's code 124 -- it should be code 131 for BS-95R in 7mm scale.

Andy, you were right. Attempting to do one chair at a time got too hot, and turned into a melted blob. The software knew it was happening, and moved away for a few seconds pause between each layer, but to not much effect. So it's back to the normal sliced whole method. Whoever invented it obviously knew what they were doing.

On the other hand, in 0 gauge 2 timbers-worth of chairs at a time is the most than can be done without risking the other nozzle damaging the tops of the chairs.

This is the PLA+ polymer. It is indeed tough, it needed some brute force to thread the rail. I will adjust the dimensions for the next try. I'm also going to try ordinary PLA, I think it's probably a bit less chewy and might give better detail.

One thing I do know is that the polymer is deposited so slowly to create these chairs, that it is spending too long hot in the nozzle. I recognize the same effects from injection moulding when travel through the heater barrel is too slow. It may need to be purged from the nozzle into the cess after every few layers. It would be wasteful, but the actual polymer volume is very low.

cheers,

Martin.  

posted: 17 Dec 2018 14:32

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi,
I think I have finished putting it together now and have taken a photo, which I will upload as soon as I can, bit of a delay as I am doing a much needed backup at the moment. I did end up with some surplus hardware (nuts and bolts) but was short of some non-essential bolts. Now to order to some filament.
Stephen Freeman wrote:
Hi
Photo of finished/nearly finished not a problem but I'm not about to take it all apart for a photo! I can however show you the empty box :D

I do feel that it will be best to make the control box connectable rather than fully permanently connected as it will be a bit unwieldy otherwise. I do have more nuts and bolts left over than I appear to need but of course am one short of a specific size :(.  I will have to see if I have any spare in the garage.

It will probably take me the rest of the coming week at least to get it fully usable.

The fun will start when I try to actually use it!

Rob Manchester wrote:
Hi Stephen
A before and after picture of your printer build would be good :)

Have fun

Rob



posted: 17 Dec 2018 20:08

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hello Martin,
                 Are the chairs actually printed onto the sleepers or are they glued on after printing?
Here are some images of prints done by MERG member Bob Gledhill in 4mm.He the did the sleeper bases and chairs separate then glues them together. The print files are available for download by members along with quite a lot of information. Perhaps it might be helpful for you to contact him?
Regards.
Trevor.

Attachment: attach_2786_3307_ballasted_track.jpg     480
Last edited on 18 Dec 2018 08:06 by Trevor Walling
posted: 17 Dec 2018 20:09

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hello,
 Here is another image.

Attachment: attach_2787_3307_junction.jpg     482

posted: 17 Dec 2018 20:11

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Here is an image of various chairs
Attachment: attach_2788_3307_chairs6.jpg     495

posted: 17 Dec 2018 22:24

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Trevor Walling wrote:
Hello Martin,
                 Are the chairs actually printed onto the sleepers or are the glued on after printing?
Hi Trevor,

Sorry I haven't properly explained what I'm doing.

The chairs are 3D printed onto the timbers. The machine first prints a timber and then prints the chairs onto it, all in one go.

At least that is the theory. At present I'm still experimenting with bits of the programs which control the printer, called G-CODE. I'm experimenting with plain track because it is simpler, but in the end most users will be mainly interested in pointwork (flexi-track is readily available for plain track in most gauges).

I don't actually want any track myself. What I want to do is to create functions in Templot which anyone can use to create the files for 3D printing. Either for their own printer, or to send to commercial companies for printing.

It's a complex subject, and the only way I can develop those files is to play about with a printer and try things.

The intention is to create 3D printed track bases to help with track building. That might be any or all of:

1. fully-chaired track bases into which rail is threaded.

2. mostly-chaired track bases which have some chairs missing, to be glued on during construction. That makes construction a lot easier, especially for slips and tandems, etc.

3. track bases which have only the inner half of each chair on them, to align and gauge the rails. The outer half to be glued on after the rail is in place. That avoids threading rail into chairs and makes construction very much easier. My hunch is that could be basis of a workable track-building system.

4. track bases without chairs which have instead a slot or pip or some other means to locate separate chairs. The base could be home-printed (the easier bit to print), with the matching chairs (the difficult bit) sent for commercial printing, or using injection-moulded chairs (C&L, Exactoscale, etc.).

5. chairs printed directly onto some other timbers, such as plywood or limewood.

6. a system in which the rail is also 3D printed, for battery operation, or which can have a metal conductive cap fitted onto the rails. Andy has been experimenting with such ideas.

Thanks for the pics of 3D printed track, there are some more pics in Andy's long topic:

  topic 2734

All this stuff would be vastly better if injection-moulded rather than 3D printed. But that is a very expensive process, and can only ever provide a limited range of fixed trackwork items which can be produced in quantity, such as the range of P4 straight turnout kits from Exactoscale.

3D printing can instead be used at home for one-off bespoke trackwork, such as the track layouts created in Templot. Everything from a basic curved turnout on any radius, to complex junctions, in any gauge or scale.

Home printing can also be exactly matched to the rail section in use, avoiding some of the reported problems in threading the currently available rail into commercial chairs.

There is not much sense in printing plain bases without any chairs at all, because that can be achieved much better and very much quicker and cheaper by the traditional method of sticking timbers on a 2D printed paper template.



To create this test piece

2_162041_280000000.jpg2_162041_280000000.jpg

I used a printer with two nozzles, currently installed as large (0.6mm dia extrusion) and small (0.2mm dia extrusion). The large one prints the timbers at reasonable speed. It is currently loaded with green polymer (showing blue in the pics). The small nozzle is loaded with brown polymer and then prints the chairs onto them (very slowly). In the end I would probably use brown for both, but using two colours makes it easier to see what is happening.

It is necessary to use a small nozzle in order to create finer detail. The difference in nozzle size is greater than it first appears, because it is the area of the nozzle opening which counts. So the ratio between the two nozzles is 2^2 : 6^2 = 4:36 = 1:9. In other words the large nozzle is effectively 9 times larger than the small nozzle. I may have overdone the difference there. I may find that using 0.3mm and 0.5mm nozzles is more sensible (9:25 or 1:2.8). Most home printers have a single 0.4mm nozzle.

I have a long way to go. :)

cheers,

Martin.

posted: 18 Dec 2018 03:51

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
When I took the first test print off the printer yesterday, I confess to being quite disappointed. I was well aware that filament-printed detail could never be a patch on injection moulding, and that home filament printing can't match other 3D printing methods such as cured resin. But I was hoping for something a bit better than I was seeing.

So I started thinking about the many printing settings which I might need to change. Looking at the chairs again today, I'm not so sure.

I don't really believe that the chairs could have improved with keeping for 24 hours, but it is not impossible. I do know that injection moulded parts shrink after injection, and continue to shrink for days or weeks afterwards, unless moulded at very high pressures. These chairs were moulded at zero pressure, so it's quite likely that they would shrink and change quite a lot with time.

Close-up photography can be so cruel to models, so I thought I would set up something more akin to normal layout viewing distance and photos. This is 7mm/ft scale:

2_172227_190000000.jpg2_172227_190000000.jpg

I changed to grey-scale to hide the coloured sleepers. The track in the background is ordinary Peco 0 gauge flexible. It seems to me that when painted and ballasted, the 3D printed version would not be too obviously inferior. This is the same test piece as before, I haven't done anything to it.

Even in what would be regarded as a close-up shot, it doesn't look too bad:

2_172235_490000000.jpg2_172235_490000000.jpg

So today I'm feeling more hopeful. This is of course 7mm, and many will be looking for 4mm scale, a much greater challenge. But this at least looks promising.

cheers,

Martin.

posted: 18 Dec 2018 07:18

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Don't give up Martin. You just need to tweak some of the variables to clean up the chairs a bit. Without knowing exactly what your code is doing I can't really advise, but I'm sure the print could be a lot cleaner than that. If you get stuck send me the G-code and I'll try to help.

The viewing distance thing is interesting. When I post high magnification shots of my chairs people tend to be put off by the visible layers in the print. In practice no one will even be able to discern the layers at a normal viewing distance, even less so after ballasting and painting.

That's what I was saying about putting printed track into a real layout. It will look just as good as injection molded track and if you don't spill the beans and reveal that it's printed, nobody will be any the wiser.

Model railways owe just as much to theater (probably a lot more) than they do to engineering.

Keep plugging away!
Andy

posted: 18 Dec 2018 08:38

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hello Martin,
              I agree with Andrew that it takes quite a bit of experimentation to achieve the optimum finish on prints. From what I have seen from others no two printers are the same and each user takes quite a few tries to optimize what they print. The quality of finish achieved has improved immensely since the earlier days of 3D printing such as this.

 http://en.wikipedia.org/wiki/RepRap_project

Regards.:)


posted: 18 Dec 2018 20:19

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Well done for getting to the printing stage with the timbers and chairs. I wasn't happy that your efforts looked as they did just like yourself. I agree with your opinion that in a layout situation the track may not be subject to the same scrutiny as in your close up photos but, from my modelling perspective, I choose 7mm scale for some subjects to allow greater detail to be included in trackwork, buildings and rolling stock. Due to the compact dimensions of the layouts and the proximity of the operating position the chairs have got to look rather better than the examples you produced or I would stick with current injection moulded components. A club building a 30x12 layout for exhibition with wide sweeping curves would be a different kettle of fish and with a little more tweaking of the settings you are on the right track(!) :)

I have a Shapeways order arriving tomorrow. Shall I spray some of the chairs and post the pics ?

Rob


posted: 18 Dec 2018 22:58

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
I have a Shapeways order arriving tomorrow. Shall I spray some of the chairs and post the pics ?
Hi Rob,

Yes please. Or just a permanent-ink felt-tip marker over them.

I've come to the conclusion that I have made a wrong turn in going for the PLA+ polymer. The web site does actually say that it is suitable for larger models rather than fine detail.

I went for it because I thought I would need the extra strength and flexibility for threading rail. I think I was too hasty.

I'm now convinced that the chairs improve with keeping. I've been looking back at a few test print chairs I made a fortnight ago when setting up the machine, using the original translucent PLA (I inked over them to see the detail). They were made using the default settings using the standard 0.4mm nozzle. They are now looking much better than I remember them at the time, and much better than the ones made in brown PLA+ this week.

But more importantly they are now much stronger. I tried threading rail in them straight off the printer, and they simply fell apart. The chair jaws would crumble away, and could be broken off easily with fingers. A fortnight later, I have only a couple left to try, but I now can't budge them with fingers. To remove the chair jaws would require a cutting tool of some sort. Clearly some form of slow chemical process is taking place after printing -- over to the polymer chemists to explain. Could the solvent in the ink be a factor?

Whatever, I have now ordered some brown ordinary PLA, and will try again when it arrives (hopefully before Christmas).

Looking at the web sites, there seems to be a minefield of different filament suppliers and prices, all claiming to be superior to any other. I'm hoping that I don't have to try every one to find the best for our purpose. :?

cheers,

Martin.

posted: 19 Dec 2018 02:40

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
"Biodegradable polymer" ? Does that imply printed items have a limited life-span or it is so long in the future we don't need to worry.
Hi Rob,

See:

 http://www.scientificamerican.com/article/environmental-impact-of-corn-based-plastics/

"Although it does biodegrade, it does so very slowly. According to Elizabeth Royte, writing in Smithsonian, PLA may well break down into its constituent parts (carbon dioxide and water) within three months in a "controlled composting environment", that is, an industrial composting facility heated to 140 degrees Fahrenheit and fed a steady diet of digestive microbes. But it will take far longer in a compost bin or in a landfill packed so tightly that no light and little oxygen are available to assist in the process. Indeed, analysts estimate that a PLA bottle could take anywhere from 100 to 1,000 years to decompose in a landfill."

cheers,

Martin.

posted: 19 Dec 2018 16:12

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I am hoping to be experimenting myself before too long, just need to get up to date with a few projects that need finishing off. I now have some ABS and PLA to try out.

posted: 19 Dec 2018 16:28

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Thanks for the info on the biodegradeable polymer.

The chair pics will have to wait. UPS messed up on the delivery by claiming there was nobody in at the delivery address despite the fact I have been 10 yards from the door all day :( I will post them when I retrieve the package.

Rob


posted: 20 Dec 2018 20:55

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I sympathise, you could try leaving a note on the door to knock harder. I presume you now have a trip to their drop-off shop to collect.

posted: 20 Dec 2018 21:03

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks Stephen. Yes package collected from the drop-off shop this morning. The delivery driver will be shot for being lazy...... :D
Rob


posted: 20 Dec 2018 23:24

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Rob Manchester wrote:
I have a Shapeways order arriving tomorrow Thursday. Shall I spray some of the chairs and post the pics ?
Hi Rob,

Yes please. Or just a permanent-ink felt-tip marker over them.

cheers,

Martin.
Hi Martin,

A quick pic of some painted 7mm scale chairs. In the middle we have 4 3D crossing chairs and 4 3D check rail chairs. Round the outside are Exactoscale and C&L samples.

2001_201822_260000000.jpg2001_201822_260000000.jpg

Sorry about the depth of field being a bit shallow - the camera is not too well at the moment.

Rob

P.S. The Shapeways chairs are printed in what they call 'Smooth Fine Detail Plastic'



posted: 21 Dec 2018 00:10

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
A quick pic of some painted 7mm scale chairs. In the middle we have 4 3D crossing chairs and 4 3D check rail chairs. Round the outside are Exactoscale and C&L samples.

P.S. The Shapeways chairs are printed in what they call 'Smooth Fine Detail Plastic'
Hi Rob,

Many thanks for those. They look great. When Shapeways say "printed plastic" do they mean from filament, or some other process? Or cured from resin?

Here's the latest from me (7mm, Peco rail):

2_201859_050000000.jpg2_201859_050000000.jpg

Getting a bit better, but not comparable with Shapeways. This is PLA rather than PLA+, printed at a lower temperature* with the fans full blast. And afterwards popped in the freezer for an hour.

*180C on the dial, but I have no idea how well calibrated the machine's thermistors might be, or even how well the two extruders are thermally matched.

This was exactly the same GCODE as before (ex Cura). My next step is to try other slicing software, and I also want to try generating the GCODE directly from Templot. There's no way it will work first time, but the great thing with doing it directly is that everything is adjustable, varying layer thicknesses, speeds, temperatures and nozzle pressures for different parts of the model if it helps.

cheers,

Martin. 

posted: 21 Dec 2018 00:47

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Fine Detail plastic from Shapeways. There seems to be several out of dates pages on the site made more confusing by the fact they changed the names a while back.

One thing I hadn't noticed before with the two makes of moulded chairs was the difference in the keys. The six 3-bolt running chairs in my pic are by both C&L and Exactoscale. The keys in the C&L ones stick out of the chair jaws a long way and are quite skinny in cross section - the Exactoscale ones are shorter and fatter and actually fill the web of the rail meaning they hold it more firmly. The 4-bolt chairs are all Exactoscale ones as they were all I had to hand. The base of the C&L chairs tends to bend as Peco rail is threaded sometimes even fracturing. The Exactoscale ones don't but are still easy to thread onto the rail. The Shapeways ones fit the Peco rail fine although it is obvious that they are more likely to break if care isn't taken to smooth the rail ends and offer it up squarely to the chair.

Your chairs are looking better. We will have to hope you get snowed in for Christmas to give you some more time on the project :D

Rob


posted: 21 Dec 2018 09:15

from:

richard_t
 
Nr. Spalding, South Holland - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Martin Wynne wrote:
Rob Manchester wrote:
I have a Shapeways order arriving tomorrow Thursday. Shall I spray some of the chairs and post the pics ?
Hi Rob,

Yes please. Or just a permanent-ink felt-tip marker over them.

cheers,

Martin.
Hi Martin,

A quick pic of some painted 7mm scale chairs. In the middle we have 4 3D crossing chairs and 4 3D check rail chairs. Round the outside are Exactoscale and C&L samples.

2001_201822_260000000.jpg2001_201822_260000000.jpg

Sorry about the depth of field being a bit shallow - the camera is not too well at the moment.

Rob

P.S. The Shapeways chairs are printed in what they call 'Smooth Fine Detail Plastic'


Nice chairs :-D And you seem to have a lighter touch with the spray can than I do (which isn't hard, I suspect). The trouble with designing your own chairs is that you can become unsatisfied with the existing commercial offerings, and end up designing all of the types of chairs …. or is that just me?

I spent most of yesterday, whilst IT was updating my work laptop to Windows 10, looking into home SLA/DLP printers... Hopefully by the time I'm finished with the fiddle yard, the prices might have come down, and that's the route I'm going to take...

Oh and our UPS man does exactly the same when he is too busy, pretends to have been, and comes back another day when he is less busy. The manager at the depot admitted this, and apologised, but it still happens.

Richard
Last edited on 21 Dec 2018 09:18 by richard_t
posted: 21 Dec 2018 11:04

from:

Jim Guthrie
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
richard_t wrote:

I spent most of yesterday, whilst IT was updating my work laptop to Windows 10, looking into home SLA/DLP printers... Hopefully by the time I'm finished with the fiddle yard, the prices might have come down, and that's the route I'm going to take...


Richard,

I've just got my hands on a DLP printier - a Phrozen Shuffle - and I've been following the progress on this thread with interest now that I am,  like Martin,  a recent entrant to the 3D printing world.

I'm getting on quite well with it but I am learning a lot of the ins and outs of the process - like what supports,  how many, and where and the advantages of adjusting the axes to get the best results.  So far I've only done parts like wagon springs,  axleboxes and buffer stocks,  the main hold-up being me getting my brain around Fusion 360. :D

I think you would get very good prints of chairs in DLP and you might have to experiment with different resins to get the best results.  At the moment I'm working with Phrozen grey ABS and that does seem quite strong but there is quite a wide range of choice.   However,  with the size of the diode screen in the Shuffle,  I don't think I would attempt what Martin is doing in 7mm scale.  I might be able to do similar in 4mm but the envelope of the Shuffle might mean that I would have to do it vertically to get a decent length of track base.  The Shuffle's sizes are X = 65mm, Y = 120mm and Z = 200mm.    There is a larger size - the Shuffle XL where the X and Y axes are almost double and there is talk of yet another larger version next year with the X/Y axes being about A4 size.

25_210603_040000000.jpg25_210603_040000000.jpg

25_210603_270000000.jpg25_210603_270000000.jpg

Here are a couple of pictures of S scale wagon axleboxes and springs printed from files given to me by Justin Newitt of Rumney models.  The quality is excellent.

Jim.

[EDIT]   I meant to add that these prints are not quick.  Typically I am printing files for betwee one and three hours for a height of about 10mm.  The main criteria for print duration is the thickness of the layers and the height of the print.    The area of the print is not a factor - I could print one chair or one hundred in the same time if the height of both prints was the same.

Jim.
Attachment: attach_2791_3307_3DPrinting-007.jpg     259
Last edited on 21 Dec 2018 11:13 by Jim Guthrie
posted: 22 Dec 2018 00:38

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Richard,
The 'lighter touch' you refer to ( thanks ) is due to the use of an Iwata airbrush with MIG or Vallejo acrylics rather than a spray can :). It means I can spray inside too without coating every flat surface of the room with overspray......
Rob


posted: 22 Dec 2018 14:45

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Latest effort. Changed to Slic3r instead of Cura to do the slicing -- a bit better I think:

2_220920_110000000.jpg2_220920_110000000.jpg


Normal viewing:

2_220920_450000000.jpg2_220920_450000000.jpg


Abnormal viewing:

2_220930_260000000.jpg2_220930_260000000.jpg


Next I want to try 6.61mm/ft scale, for a better match for the 95lb REA chair and key to the underscale Peco rail. (It's code 124 -- should be code 131 for BS-95R bullhead at 7mm/ft.)

cheers.

Martin.

posted: 22 Dec 2018 20:43

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hello Martin,
         It may be a good idea to try a different filament for the chairs just to make sure the results are not due to a bad reel of filament.
Regards.
Trevor :)

posted: 23 Dec 2018 03:02

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Latest effort. Changed to Slic3r instead of Cura to do the slicing -- a bit better I think:

2_220920_110000000.jpg2_220920_110000000.jpg


Normal viewing:

2_220920_450000000.jpg2_220920_450000000.jpg


Abnormal viewing:

2_220930_260000000.jpg2_220930_260000000.jpg


Next I want to try 6.61mm/ft scale, for a better match for the 95lb REA chair and key to the underscale Peco rail. (It's code 124 -- should be code 131 for BS-95R bullhead at 7mm/ft.)

cheers.

Martin.
There you go!

Slap some paint on them and nobody will be able to tell they were not injection molded.

posted: 23 Dec 2018 03:30

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Trevor Walling wrote:
It may be a good idea to try a different filament for the chairs just to make sure the results are not due to a bad reel of filament.
Thanks Trevor,

I have in fact already tried two different polymers from different suppliers, ordinary PLA and toughened flexible PLA+.

At some stage I also want to try ABS and maybe others.

But I also want to retain some flexibility which is why I went for PLA+. For example, if only the inner half-chair and the rail seat is printed on the timber, a separate outer half-chair could be fixed in place after the rail is in place against and under the inner jaw. This avoids any tricky threading of bent rails through the chairs.

If the outer half-chair is printed (or maybe obtained from Shapeways) in a flexible polymer, it could locate down into a recess in the timber and snap into place as the key snaps past and under the rail head. There might also be some undercut snap in the timber location. The join in the chair base would be close to the rail foot and go unnoticed after painting. Some penetrating cyano would lock the whole thing solid if necessary.

I have lots of ideas to try, once I have the confidence to set up all the parameters on the printer to obtain precise flash-free results.

cheers,

Martin. 

posted: 23 Dec 2018 03:39

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
There you go!

Slap some paint on them and nobody will be able to tell they were not injection moulded.
Thanks Andy.

Yes, I was quite pleased with them. What I don't know yet, is whether this is near the limit of what can be expected from a home filament printer, or whether I still have a long way to go to get the best out of it. Have you tried anything in 7mm scale for comparison? I shall be interested to see what Stephen comes up with when he gets his printer going.

cheers,

Martin.

posted: 23 Dec 2018 04:50

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Andrew Barrowman wrote:
There you go!

Slap some paint on them and nobody will be able to tell they were not injection moulded.
Thanks Andy.

Yes, I was quite pleased with them. What I don't know yet, is whether this is near the limit of what can be expected from a home filament printer, or whether I still have a long way to go to get the best out of it. Have you tried anything in 7mm scale for comparison? I shall be interested to see what Stephen comes up with when he gets his printer going.

cheers,

Martin.
Hi Martin,

I have not tried to print 7mm chairs but based on my efforts I would say you are approaching the limits. Further adjustments might be more to do with optimizing the process for strength, speed, time etc.

When things are a bit less hectic here I'll take a shot at some 7mm chairs for comparison, but that will not happen for a couple of weeks.

It probably doesn't make much sense at 00/H0 scales, but at larger scales you could just print 2.5D turnout templates. The chairs would be 3D but the timbers would be very thin, almost like paper. The templates would be glued on to plywood timbers. It's a bit like putting timbers on a template only the other way round.

The advantage is that all the correct chairs would fall into place at the correct gauge. No mucking about with gauges. Just slide the rails in and go (OK, that might be a slight over-simplification :) )

Looks great! Keep going.

Andy



posted: 23 Dec 2018 12:15

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
It probably doesn't make much sense at 00/H0 scales, but at larger scales you could just print 2.5D turnout templates. The chairs would be 3D but the timbers would be very thin, almost like paper. The templates would be glued on to plywood timbers. It's a bit like putting timbers on a template only the other way round.
Hi Andy,

Yes, I have already been thinking about printing the chairs directly onto ply timbers.

Since I found that the PLA+ polymer (but not ordinary PLA) will print directly onto the glass without any adhesive, I had another idea. If a suitable bulldozer attachment was fitted onto the extruder head, the timbers could be printed one at a time and be bulldozed off the bed after printing. That way the printer could run continuously to print all the timbers for a large pointwork complex in one go, the timbers to be attached to a paper template in the usual way. I noticed that the diagonal infill lines are visible on the underside of the timber. So with a bit of G-code an identifying mark could be printed into them.

The bulldozer flap thingy needs to be sprung or weighted down into position, and be pushed back up against the bed while printing.

cheers,

Martin.   

posted: 23 Dec 2018 13:00

from:

richard_t
 
Nr. Spalding, South Holland - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Jim Guthrie wrote:
richard_t wrote:

I spent most of yesterday, whilst IT was updating my work laptop to Windows 10, looking into home SLA/DLP printers... Hopefully by the time I'm finished with the fiddle yard, the prices might have come down, and that's the route I'm going to take...


Richard,

I've just got my hands on a DLP printier - a Phrozen Shuffle - and I've been following the progress on this thread with interest now that I am,  like Martin,  a recent entrant to the 3D printing world.

I'm getting on quite well with it but I am learning a lot of the ins and outs of the process - like what supports,  how many, and where and the advantages of adjusting the axes to get the best results.  So far I've only done parts like wagon springs,  axleboxes and buffer stocks,  the main hold-up being me getting my brain around Fusion 360. :D

I think you would get very good prints of chairs in DLP and you might have to experiment with different resins to get the best results.  At the moment I'm working with Phrozen grey ABS and that does seem quite strong but there is quite a wide range of choice.   However,  with the size of the diode screen in the Shuffle,  I don't think I would attempt what Martin is doing in 7mm scale.  I might be able to do similar in 4mm but the envelope of the Shuffle might mean that I would have to do it vertically to get a decent length of track base.  The Shuffle's sizes are X = 65mm, Y = 120mm and Z = 200mm.    There is a larger size - the Shuffle XL where the X and Y axes are almost double and there is talk of yet another larger version next year with the X/Y axes being about A4 size.

Jim.

Hi Jim 
I've been following your work on WT (indeed it was your work in the garden that encouraged me to join...) But I didn't twig that it was a DLP printer.

I agree about Fusion 360 and it may be worse as I'm used to Autodesk Inventor which is very much like Fusion except when it isn't... Unfortunately I've lost access to Inventor when my subscription expired.

Keep up the good work - I'm definitely following with interest.

Richard
Last edited on 23 Dec 2018 13:00 by richard_t
posted: 23 Dec 2018 14:18

from:

Jim Guthrie
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
richard_t wrote:
I've been following your work on WT (indeed it was your work in the garden that encouraged me to join...)
Richard,

Unfortunately,  no work has been done on the garden layout this year.  A bad encounter at an MOT test with a modern,  low mileage diesel cost me a fortune to get it through the emission tests,  so no disposable income available for the next move which would have been the tracklaying.  I've now changed to a petrol car - about a month after that MOT - so there should be a reasonable amount of disposable income for next year. :D

Jim.


posted: 23 Dec 2018 17:14

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:


The bulldozer flap thingy needs to be sprung or weighted down into position, and be pushed back up against the bed while printing.

Hi Martin,

I don't know if it would work on your printer, but on mine it might be possible to mound a fixed beam just above the bed. To knock the print off just drive the bed to run the print into the beam? Not sure if that work or not.

Cheers,
Andy

posted: 23 Dec 2018 18:47

from:

John Durbetaki
 
Gaston - Oregon USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Would the shearing forces be such that the driving forces of the XY motors or their supports might be whacked out of calibration or worse?

Since you thought about printing the timbers one at a time, could you print with a compressed spacing (it looks from the image you could get a timber within the space between timbers, and maybe tighten it up more. Since they are singles anyway, cut them apart and lay them on the paper template. That way, you can fill the printer bed with prints and then get them all off at the end. It would seem markings on the timbers would be needed to identify what goes where...

Just a thought...

John
Andrew Barrowman wrote:
Martin Wynne wrote:


The bulldozer flap thingy needs to be sprung or weighted down into position, and be pushed back up against the bed while printing.

Hi Martin,

I don't know if it would work on your printer, but on mine it might be possible to mound a fixed beam just above the bed. To knock the print off just drive the bed to run the print into the beam? Not sure if that work or not.

Cheers,
Andy


posted: 23 Dec 2018 20:22

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
John Durbetaki wrote:
Would the shearing forces be such that the driving forces of the XY motors or their supports might be whacked out of calibration or worse? Just a thought...

Andrew Barrowman wrote:
I don't know if it would work on your printer, but on mine it might be possible to mound a fixed beam just above the bed. To knock the print off just drive the bed to run the print into the beam? Not sure if that work or not.

Hi Andy, John,

Thanks.

A fixed beam won't work on the BIBO printer because the only bed movement available is the Z drive. The X and Y drives are on the extruder head in the top of the case. The bed drops as the work proceeds.

Here is what I had in mind, if you can call it a mind: :)

2_231443_030000000.png2_231443_030000000.png

A hinged weighted or sprung flap attached to the head and running against the bed. During printing of a single timber, it would be off to the left of the work. Then by driving the extruder head to the right, the timber would be bulldozed off the bed into the bottom of the case. There is plenty of space there. 

It's true that the stepper motors see a design load of near zero. The Z stepper sees only the geared-down weight of the bed assembly, and the X,Y steppers see only the flexibility restriction of the cables and filament feed.

But the construction seems robust enough to withstand a smallish additional load at lowish speed. In the BIBO the X Home position is conveniently on the right, with the nozzles then clear of the bed. So sending the X drive home will not only bulldoze the timber off the bed, any lost calibration of the X drive will be reset when the head hits the Home limit switch. Ready to move back left for the next timber. Even if some X calibration is lost, it wouldn't affect the printing of a single timber, it would simply take place at a different location on the bed.   

Since you thought about printing the timbers one at a time, could you print with a compressed spacing (it looks from the image you could get a timber within the space between timbers, and maybe tighten it up more. Since they are singles anyway, cut them apart and lay them on the paper template. That way, you can fill the printer bed with prints and then get them all off at the end.

Yes, I've been thinking along such lines. Once you decide to print individual timbers to be used on a paper template there are several options. Andy suggested printing only a thin timber to be attached to ply or card or whatever full timbers. That would reduce printing time considerably. I did some tests, and I can effectively print a "timber" only 0.3mm (12 thou) thick, although 0.5mm (20 thou) might be more sensible. (My current tests are 1.9mm thick. Peco 0 gauge track has timbers 3.3mm thick.)

Lots of ideas to try. :)

cheers,

Martin.

posted: 23 Dec 2018 21:14

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
p.s. I have just tested the force available on the X and Y drives when going Home. It is considerable, I can't stop them with reasonable hand force. I wouldn't want to get a finger trapped in there. :(

Martin.

posted: 24 Dec 2018 07:12

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi all,

It might be a week or two before I can get around to actually setting up the machine, with Christmas and current projects that need completion first.

posted: 26 Dec 2018 02:34

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Getting a bit closer. The central rib on the inner jaw is now distinct, and there is now no flashing between the screw heads. I also modified the outer jaw and key to better match the underscale Peco pretend-bullhead section, and made it a closer fit in the chair (7mm scale / 0 gauge):

2_252112_540000000.jpg2_252112_540000000.jpg

2_252113_110000000.jpg2_252113_110000000.jpg

2_252113_340000000.jpg2_252113_340000000.jpg

In theory you make one change at a time. So I made four. :)

1. retraction increased from 0.25mm to 1mm.

2. Z-hopping over perimeters switched on, and set at 0.2mm lift (for 0.05mm layers).

3. infill overlap on perimeters reduced from 15% to 5%.

4. printing order changed to print infill before the perimeter around it.

I don't know which of these was the more important, but I suspect 4. was the most significant in improving the definition. Whatever, I am now quite pleased with the results, which compare well alongside injection-moulded flexi-track (Peco 0 gauge).

It's now time to see what can be done in 4mm scale. :?

cheers,

Martin.

posted: 26 Dec 2018 15:28

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,
Happy Boxing Day :) Looking quite good I must say, all your work is starting to bear fruit.

Do you have any C&L O gauge rail ? Happy to post you a couple of short lengths tomorrow if you want to try a different profile.

Rob


posted: 26 Dec 2018 22:00

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Do you have any C&L O gauge rail ? Happy to post you a couple of short lengths tomorrow if you want to try a different profile.
Hi Rob,

Many thanks for the offer, but I do have some C&L code 125 bullhead rail. I just need to move some boxes to get at it. The Peco rail was easier to find to hand. :)

Somewhere I also have some original Slaters code 131 bullhead, which is the correct scale size for BS-95R rail in 7mm scale. Slaters nowadays are denying they ever supplied such a thing, despite the fact that it is specifically mentioned in their 7mm trackbuilding handbook. :? Presumably admitting it would mean acknowledging that their current code 125 rail is underscale. Even S7 modellers seem to be using it, and Karlgarin are supplying it too.

If you have some 0 gauge bullhead it would be worth measuring it.

cheers,

Martin.

posted: 26 Dec 2018 22:44

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:

If you have some 0 gauge bullhead it would be worth measuring it.

cheers,

Martin.
Hi Martin,
I just measured my stocks of O scale bullhead, not sure it was worth it though :?

All I have is C&L and some rail supplied by Marcway years ago. Both measure 0.125" height and a scale 2.75" head width as expected. The profile is very similar except the Marcway rail appears to have a slightly wider radius between the foot and web.

I did however also find some Peco 'ready made' switch blade sets. The height is the same as above but the head is a bit skiny, scaling out at around 2.6" wide. As I think you pointed out the foot is very under fed.

I remember the Slaters track system and I had the book too once upon a time. The late Bill Hudson used to run the shop at the Matlock museum. I hadn't realised that Slaters still supplied the track parts range. The rail they list is a lot cheaper than the C&L price although Slaters may get a shock if they have to have a new batch done.

Rob


posted: 26 Dec 2018 22:47

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Getting a bit closer. The central rib on the inner jaw is now distinct, and there is now no flashing between the screw heads. I also modified the outer jaw and key to better match the underscale Peco pretend-bullhead section, and made it a closer fit in the chair (7mm scale / 0 gauge):

2_252112_540000000.jpg2_252112_540000000.jpg

2_252113_110000000.jpg2_252113_110000000.jpg

2_252113_340000000.jpg2_252113_340000000.jpg

In theory you make one change at a time. So I made four. :)

1. retraction increased from 0.25mm to 1mm.

2. Z-hopping over perimeters switched on, and set at 0.2mm lift (for 0.05mm layers).

3. infill overlap on perimeters reduced from 15% to 5%.

4. printing order changed to print infill before the perimeter around it.

I don't know which of these was the more important, but I suspect 4. was the most significant in improving the definition. Whatever, I am now quite pleased with the results, which compare well alongside injection-moulded flexi-track (Peco 0 gauge).

It's now time to see what can be done in 4mm scale. :?

cheers,

Martin.
Hi Martin,

That's really excellent. You might try eliminating the Z hopping to reduce the print time. I've tried Z hopping but other than slowing things down I didn't see any difference.

Did you use Slic3r?

How long does it take to print the four chairs?

Cheers,
Andy

posted: 26 Dec 2018 23:49

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
That's really excellent. You might try eliminating the Z hopping to reduce the print time. I've tried Z hopping but other than slowing things down I didn't see any difference.

Did you use Slic3r?

How long does it take to print the four chairs?
Thanks Andy,

Yes, Slic3r. It seems to be more capable than Cura, which was the default supplied with the printer. I'm reading that Simplify3D is even better, but at $150 it's a lot more than free. :(

Printing the 4 chairs (after printing the timbers) takes about 16 minutes at the speeds I'm currently using. So 4 minutes per chair. A typical turnout has around 100 chairs, so that's around 7 hours to print the chairs on a typical 0 gauge turnout. I might be able to speed it up a bit, but I didn't notice adding the Z-hopping making much difference (it might do in 4mm with smaller areas). Changing from 0.05mm layers to 0.1mm would obviously make a big difference, but may lose some of the definition. There is still a lot of experimenting to do.

But that's quite a bit faster than printing the 2 timbers first, which take about 20 minutes to print up to 1.9mm thickness but then another 10 minutes to add the top skin to give smooth tops. I suspect there is more scope for speeding up there, everything really depends on the required timber thickness (and the infill density).

The times are a bit misleading because of course once set running, the printer needs minimal attention for the rest of the day. The BIBO (and other printers?) also has an option to stop the print at any stage, and restart it later, or next day. Likewise if it suffers a power cut or runs out of filament.

I'm now looking at 4mm scale, which should obviously be quite a bit quicker than 7mm. The volume ratio is 343/64, say 5:1, so a fifth of the 7mm time for 4mm? We shall see.

cheers,

Martin.

posted: 27 Dec 2018 00:44

from:

Nigel Brown
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin

You could try 3mm/ft :D  Could be 12 times faster, say 20 seconds/chair.

Nigel

posted: 27 Dec 2018 16:11

from:

Charles Orr
 
Leicester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
8_271110_120000000.png8_271110_120000000.pngHi Martin,
I use Simplify3D for slicing.  It has a very wide and comprehensive range of settings.

Charles

posted: 29 Dec 2018 21:47

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Charles Orr wrote:
I use Simplify3D for slicing. It has a very wide and comprehensive range of settings.
Thanks Charles.

It also costs $150 for much the same range of settings as are available in Slic3r and Cura, which are both free.

However it has brilliant previewing and extrusion animation functions, so I have had a rush of blood to the head and purchased a copy. I'm telling myself that it will be worth it when I get started on the 3D maps. :)

Here are the first trials in 4mm scale. I found that ordinary PLA was too fragile, so I have changed back to the tougher PLA+ which doesn't work so well for detail. Even so I needed to beef up the design of the chair jaws to give them sufficient strength, as Andy warned me I would.

At present the rail fit is a bit too tight, and I managed to thread only 4 of the 6 on the test piece without breaking them:

2_291620_530000001.jpg2_291620_530000001.jpg

2_291620_530000000.jpg2_291620_530000000.jpg

2_291620_530000002.jpg2_291620_530000002.jpg

Above are EM gauge, with C&L code75 rail.

I think I have maybe over-beefed the outer jaw, there is still a lot of experimenting to do:

2_291641_260000000.png2_291641_260000000.png

cheers,

Martin.

posted: 30 Dec 2018 00:53

from:

madscientist
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Have you considered PETG , it’s strong but has some flexibility , or perhaps FPE , of which I have no experience but the specs look good 

Dave
Last edited on 30 Dec 2018 01:13 by madscientist
posted: 30 Dec 2018 01:38

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
madscientist wrote:
Have you considered PETG , it’s strong but has some flexibility , or perhaps FPE , of which I have no experience but the specs look good
Hi Dave,

Yes, I want to try ABS and PETG. I've started with PLA because that seems to be the most popular and easiest to print. Being strong and flexible isn't the only requirement, the most important thing is the ability to print fine detail.

But bear in mind that I don't actually want any track for myself. The object of the exercise is to export files from Templot which everyone can use for 3D printing or to send to Shapeways, etc. So I need to find settings for the files which will be usable by most Templot users. The only way I can do that is by trial and error on an actual printer.

I also have to bear in mind the passing of time. In the time Templot has been available (let alone the time I have been working on it), computing capability has leapt forward massively beyond what was available when it was first released. If it now takes me say 5 years to get all the chair designs and export options done, who knows what new 3D printing capability will have appeared by then? Or how many households will have a 3D printer? So I need to think ahead and put detail stuff in the files, even if there is no practical way at present to print them on a home machine.

cheers,

Martin.

posted: 30 Dec 2018 08:07

from:

Martin Carew
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Great to see all the development for Chairs going ahead! I am guessing this request might be a lesser used type but will flat bottom modern chairs be added in the near future? Specifically the Pandrol types etc?

posted: 30 Dec 2018 10:57

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Carew wrote:
Great to see all the development for Chairs going ahead! I am guessing this request might be a lesser used type but will flat bottom modern chairs be added in the near future? Specifically the Pandrol types etc?
Hi Martin,

Welcome to Templot Club. :)

Flat-bottom designs will come eventually, but I don't think you should hold your breath. Home 3D printing has a long way to go before functional Pandrol clips are feasible in the smaller scales.

UK-style bullhead track has remained largely unchanged for getting on for a century now. It's familiar to many and easy to model. In contrast, flat-bottom track has gone through massive changes over the same period, and is much more difficult to model.

Templot includes the inclined FB-109, BS-110A and BS-113A designs from the 1950s and 1960s, but that's it as far as native flat-bottom support is concerned. Anything later requires the user to create custom templates (and know what they are doing). Even the 1432mm double-curved vertical designs from the 1970s are not yet implemented and need to be improvised. It's all doable, but not necessarily straightforward. A modern transitioned high-speed turnout would need to be constructed from many custom partial templates.

I'm now 70 years old and I've been working on Templot one way or another for 40 years, see:

 http://templot.com/martweb/templot_history.htm

6 months ago I made Templot open-source, in the hope that someone a bit younger might come along to move Templot forward, including perhaps all the modern flat-bottom designs. But so far interest has been minimal -- it's still all down to me. As I said, I don't think you should hold your breath.

Unless of course someone somewhere is beavering away on a "son of Templot" program unknown to me? We can only hope. :)

cheers,

Martin.

posted: 30 Dec 2018 14:07

from:

DerekStuart
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin,
Just a quick OT if I may.
1. Happy Birthday for your 70th (you said you were 70 this year but not when).

2. Please don't mistake lack of ability with lack of interest, re open source Templot. I've done a bit of programming before, but what you've done is well above most of us.

3. Best wishes for 2019.

Derek
Last edited on 30 Dec 2018 14:08 by DerekStuart
posted: 30 Dec 2018 20:01

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
madscientist wrote:
Have you considered PETG , it’s strong but has some flexibility , or perhaps FPE , of which I have no experience but the specs look good 

Dave
Hi Dave,

I tried PETG but the result, though it was strong and flexible, was a stringy mess. It doesn't seem to "pinch-off" during retraction as well as PLA and its derivatives. It would, I think, be a good filament for objects that have large contiguous surfaces that don't require lots of retractions.

Cheers,
Andy
Last edited on 30 Dec 2018 21:21 by Andrew Barrowman
posted: 30 Dec 2018 21:39

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks Derek. Now 70 and a half. :)

Getting a bit better I think in 4mm scale. The rail is held vertical as intended, and by some fluke exactly to 18.2mm gauge EM (I was expecting to need some tweaking there).

Sensible close-up:

2_301608_470000004.jpg2_301608_470000004.jpg

Silly close-ups:
 
2_301608_460000000.jpg2_301608_460000000.jpg

2_301608_470000001.jpg2_301608_470000001.jpg

2_301608_470000002.jpg2_301608_470000002.jpg

2_301608_470000003.jpg2_301608_470000003.jpg

I shall try again with ordinary PLA for cleaner detail, but the PLA+ is very much tougher and can withstand some reasonable rough handling when threading rail.

cheers,

Martin.

posted: 30 Dec 2018 22:48

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Personally, I blame this 'ere modern close-up photography. If you'd published your efforts fifty years ago in the Railway Modeller everyone would say it's brilliant :)

Seventy eh? Gosh, you are old. I won't hit that one till next year (that sentence expires in less than 48 hours)

Cheers,
Andy

posted: 30 Dec 2018 22:53

from:

Nigel Brown
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
Hi Martin,

Personally, I blame this 'ere modern close-up photography. If you'd published your efforts fifty years ago in the Railway Modeller everyone would say it's brilliant :)

Seventy eh? Gosh, you are old. I won't hit that one till next year (that sentence expires in less than 48 hours)

Cheers

Old??? I'm 5 years older than that and still young :)

Nigel

posted: 30 Dec 2018 23:25

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Nigel Brown wrote:
Old??? I'm 5 years older than that and still young :)
Nigel
Indeed you are!

If you are fortunate enough to have the mental or physical capacity, age is not at all absolute. I started skiing late in life. It keeps me young (and it makes me sleep really well :) )

If I don't over-do it this season I'll post some pics from Utah, Colorado, Idaho and Montana.

Cheers,
Andy

posted: 31 Dec 2018 00:01

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Andrew Barrowman wrote:
Seventy eh? Gosh, you are old. I won't hit that one till next year
Hi Andy,

I've no immediate plans to be old. :)

But I've become only too aware that my brain is not what it was. You wouldn't believe how many silly mistakes I make nowadays when working on Templot code. Everything has to be checked, double-checked, and checked again. I'm permanently surrounded by bits of paper with odd numbers scribbled on the corner, because I know I will forget them within seconds if not written down. I find whole chunks of code which I have entirely forgotten about, and half-finished things which I don't recall starting.

Which is why I thought 70 was the time to make Templot open-source and hopefully let A. N. Other take it forward. 40 years is a long time to work on something single-handed. I still enjoy coding, but I wouldn't mind a break from it. :)

cheers,

Martin.

posted: 31 Dec 2018 00:40

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Here is another project that has been going a while although some of the info stinks :(

Toilet Map

Lots of missing info for us all to edit and update. A good excuse for drinking lots of tea too :)

Rob

P.S. Good work on the 4mm chair printing, not bad at all. I have never used Exactoscale rail - is it skinny in the foot area or just an illusion in your shots.




posted: 31 Dec 2018 01:04

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
P.S. Good work on the 4mm chair printing, not bad at all. I have never used Exactoscale rail - is it skinny in the foot area or just an illusion in your shots.
Thanks Rob.

I believe the Exactoscale rail is from the same source as C&L and the scale societies.

(Winterbottom Wire Mills? see:

 http://www.wintwire.co.uk/profile-wire-manufacturers.htm

watch the animation.)

The overall rail dimensions are good, but the foot does look a bit low in those 10-year-old samples. I don't know how well they compare with current stocks. An undersize foot does of course help a bit in threading chairs.

cheers,

Martin.

posted: 31 Dec 2018 13:29

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Hi Martin,

Thanks for the wire link :) I guess the raw material is made abroad and imported and they produce the profiles from it. Rumour has it that Sheffield used to actaully produce the odd ton of steel and other materials a while back.

I suspect my stocks of code 75 BH that I class as being of C&L origin do probably include rail from at least one other source ( EMGS Stores ) but I can't see any difference in profile.

Having been critical of the Scalefour stores at their northern show last year I think I am probably banned from joining :( They just didn't seem to understand that code 82 FB rail should have a head width the same as code 75 BH. Now we don't have Peco IL-115 Code 82 with a correct width 4mm scale head it is an issue.

Happy New Year for tomorrow.

Rob



posted: 1 Jan 2019 16:06

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Nigel Brown wrote:
Martin

You could try 3mm/ft :D  Could be 12 times faster, say 20 seconds/chair.
3mm? Sure. Do you want separate keys with that? :)

Happy New Year everyone.

2019 marks 40 years since I first started plotting templates from computed co-ordinates in 1979.

Here again are some memories from that time which I posted last year:

 topic 3197 - message 23709

cheers,

Martin.

posted: 3 Jan 2019 01:27

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
It's time to start thinking about some of the practicalities.

1. At present, almost every setting in Templot is template-specific. That means each template in a .box file can have its own scale (unlikely), its own rail section, etc. And they can all be output at the same time for the printed templates, PDF files and 2D DXF files, etc.

That's going to be difficult to continue for the chairing options, not least because of the use of BLOCK entities in the 3D DXF files.

It will be necessary to set up the required chair/baseplate dimension settings separately from the templates (or load the chair settings from a file), and then output an entire group of templates all with those same chair settings. It would get far too complex to make everything template-specific.

For most users this will have little effect. But if for example your track plan includes some GWR templates and some LSWR templates, with different chairs, if you want chair detail in the output it will not be possible to print both at the same time in the same print run or the same 3D panel. They will have to be done separately (unless you want to start editing the G-code manually).


2. Which brings me on to the question of panel joins in the 3D track bases. For the paper template prints it doesn't matter if a timber is split across separate pages, and the rectangular trim lines can be easily printed around each page. That's obviously not possible for the 3D track bases.

My printer can handle panels up to 8" x 7" (200mm x 175mm) and that is probably typical of many 3D printers. In any event you wouldn't want to print much larger because of the long print times. So a track plan is going to be made up from several track bases, and I need to devise some means to define the join positions running between the timbers. And then some sort of hook and socket fitting to align them accurately together for construction, unless they are going to be stuck down on paper templates?


3. So far I have fully defined only the REA S1 (3-screw) chairs. But it would now be possible to create plain track bases using them, to any length, curved or straight, in any scale or gauge.

Should I be looking to release a program update in the shorter term allowing only that much? So that folks can experiment with 3D printing from Templot files? It might be some time before I can get all the special chairs defined for turnouts. In the meantime it would be useful to have some feedback from users about the files and any tweaks needed.

The downside of that is that word will go round the web forums that Templot "can only do plain track", an idea which will then be difficult to shift. It is remarkable the way that so many folks ALWAYS get the wrong end of a stick. It's no wonder that "fake news" is so effective.

Thoughts welcome.

cheers,

Martin.

posted: 3 Jan 2019 19:48

from:

Trevor Walling
 
United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
2. Which brings me on to the question of panel joins in the 3D track bases. For the paper template prints it doesn't matter if a timber is split across separate pages, and the rectangular trim lines can be easily printed around each page. That's obviously not possible for the 3D track bases.
Hello Martin,
        Would it be possible to print track panels vertically on top of each other in a stack?
It would make more efficient use of the build area.
Regards
Trevor.:)

posted: 7 Jan 2019 05:48

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Dave,

Here's a sample of PETG.

Some of you might recognize this as a Tri-ang two-start worm and worm wheel. I just replicated the worm wheel on my printer. Obviously the hole in the center is a mess. I'll need to fiddle with temperatures and flow-rates to get it right, but the tooth definition is remarkably good. The surface is also very hard and should withstand a fair bit of use.

Cheers,
Andy


2983_070036_150000000.jpg2983_070036_150000000.jpg
madscientist wrote:
Have you considered PETG , it’s strong but has some flexibility , or perhaps FPE , of which I have no experience but the specs look good 

Dave


posted: 8 Jan 2019 06:35

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
The reason I'm messing about with printed gears is slightly obscure. I'm converting a Hornby tender drive Duchess to four-wheel-drive. The leading and trailing drivers will be gear driven from a rather large motor in the tender and the Duchess will be filled with lead :D

However, all this messing around isn't completely irrelevant to track construction. PETG is very sturdy stuff, but it is a bit quirky when it comes to printing. Those gears are pretty good (later ones even better) but they really need to be printed one at a time (Slic3r allows that option) otherwise there will be lots of "spider webs" between the gears.

So, if chairs were to be printed in PETG I think it would be best if each chair was printed completely before moving on to the next chair. That seems to be the opposite of what works well with PLA, and it would probably require some G-code that I do not think Slic3r will generate.

Cheers,
Andy

posted: 21 Feb 2019 11:48

from:

Stephen Freeman
 
Sandbach - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Everything quiet here at the moment? In the absence of any other developments in the matter I might actually have to make a start on designing some slide chairs (7mm scale GWR) before too long. Don't have any idea  how things will go. The only trouble is I have so much other pressing stuff to do, which is getting in the way.
Last edited on 21 Feb 2019 11:49 by Stephen Freeman
posted: 21 Feb 2019 11:58

from:

madscientist
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I’m just configuring my ender 3 with a 0.2mm nozzle to see how I get on

I also have on the bench an upgrade to a MKS Gen-L board that supports microstepping and I have some 0.9 degree nema17 steppers on order , it will be interesting to see how the actual resolution improvements translate to actual print quality
My ultimate goal is fit a two head system , that carrys a 0.4 and a 0.2 head 
Last edited on 21 Feb 2019 12:00 by madscientist
posted: 21 Feb 2019 18:25

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
madscientist wrote:
My ultimate goal is fit a two head system, that carries a 0.4 and a 0.2 head
Hi Stephen, Dave,

I have set mine up with a 0.6mm nozzle for the timbers, 0.2mm nozzle for the chairs.

Bear in mind that the extrusion volume is determined by the area of the nozzle, i.e. the square of the diameter. Which means a 0.4mm nozzle will extrude 4 times faster than a 0.2mm nozzle (16:4). And a 0.6mm nozzle 9 times faster (36:4).

Which is another way of saying that a 0.2mm nozzle is extremely slow if you want to use it for anything of significant volume.

I haven't done much more on this project recently, I've been busy with other things. I must get back to it soon. Currently it needs some work on the gauging, for rail sections which don't match the prototype dimensions.

Looking back at the prints I made a couple of months ago, I'm more than ever convinced that the prints improve with keeping. It's difficult to be sure of a mechanism for this, but it is likely related to differential shrinkage. Certainly I know from injection moulding that mouldings age-shrink if not injected at a high enough pressure. And 3D filament is extruded at a tiny fraction of injection-moulding pressures.

Judging the results with a fresh eye, I think the 7mm results are really quite good -- on a par with 3D resin prints, and not far behind injection-moulded chairs.

The 4mm results are more iffy. At least so far. Fine at normal layout viewing distance, but not so good in cruel digital close-ups.

The most time-consuming part of the print is adding a smooth top skin to the timbers, to avoid the usual ridged 3D printing effect -- which is more noticeable with the 0.6mm nozzle of course.

cheers,

Martin.

posted: 21 Feb 2019 18:55

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:

Looking back at the prints I made a couple of months ago, I'm more than ever convinced that the prints improve with keeping. It's difficult to be sure of a mechanism for this, but it is likely related to differential shrinkage. Certainly I know from injection moulding that mouldings age-shrink if not injected at a high enough pressure. And 3D filament is extruded at a tiny fraction of injection-moulding pressures.

cheers,

Martin.
Hi Martin,
Off topic a bit as usual but is that the reason that injection moulded RTR open wagon bodies have sides that bow inwards especially when you have had them a year or two ?? Kits aren't much of a problem as when building I make sure the sides bow out ( like the real ones ) so the solvent can at worst make them parallel.

Rob


posted: 21 Feb 2019 19:20

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Rob Manchester wrote:
Off topic a bit as usual but is that the reason that injection moulded RTR open wagon bodies have sides that bow inwards especially when you have had them a year or two ??
Hi Rob,

I don't have any such bodies to examine, but let me guess that this happens when the inside is smooth undetailed? And the feed gate is in the centre of the underside of the floor?

The outer surface of the side is seeing a lower injection pressure because of faster solidification in the fine detail, and also a greater distance from the gate.

It therefore age-shrinks more than the smooth inner surface, and the result is a side bowed in.

But it is poor-quality moulding, the machine pressure should be increased to prevent this happening. Possibly the moulding machine is too small for this job, lacking sufficient ton-lock to allow higher injection pressures.

But that's just me muttering in the corner, don't take any notice. :)

cheers,

Martin.   

posted: 21 Feb 2019 21:06

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Rob Manchester wrote:
Off topic a bit as usual but is that the reason that injection moulded RTR open wagon bodies have sides that bow inwards especially when you have had them a year or two ??
Hi Rob,

I don't have any such bodies to examine, but let me guess that this happens when the inside is smooth undetailed? And the feed gate is in the centre of the underside of the floor?

The outer surface of the side is seeing a lower injection pressure because of faster solidification in the fine detail, and also a greater distance from the gate.

It therefore age-shrinks more than the smooth inner surface, and the result is a side bowed in.

But it is poor-quality moulding, the machine pressure should be increased to prevent this happening. Possibly the moulding machine is too small for this job, lacking sufficient ton-lock to allow higher injection pressures.

But that's just me muttering in the corner, don't take any notice. :)

cheers,

Martin.
Hi Martin,
Thank you. We do take notice of you as your comments usually make logical sense :)

Yes, I think many, if not all, are moulded with a single injection point in the middle of the bottom. You don't tend to get any signs of other entry points. The bottom surface of the wagon is usually hidden by the chassis moulding, until that is you strip the wagon down to add compensation units or such like. The insides of wagons are often pretty plain in terms of detail but then again there wasn't that much on the prototype. Details such as end door hinge bars are usually added to the basic moulding.

Rectifying the issue isn't that easy. On a loaded wagon you can make an insert to keep the sides straight, or even bowed slightly. For empty ones I have done some experiments with hot water to see if forcing the sides apart while sitting the body in the hot water and then letting the body cool. Problem is the materials used seem to be variable, water that is too hot seems to eat away at the plastic, too cold and the body resumes it's concave shape.

Rob


posted: 21 Feb 2019 23:59

from:

Nigel Brown
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
This Cambrian Railways 15T coal wagon was constructed from a 3D printed body from Coast Line Models, who obligingly reduced his 4mm one to 3mm for me, along with buffers and springs / axle-boxes.

528_211857_220000000.jpg528_211857_220000000.jpg

When it arrived it there was a noticeable bow to the sides. Came back to it 3 weeks later and the sides were straight. The sides are sufficiently thin that I wouldn't have been surprised if they had remained warped.

Nigel

posted: 22 Feb 2019 00:21

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Nigel Brown wrote:
This Cambrian Railways 15T coal wagon was constructed from a 3D printed body from Coast Line Models, who obligingly reduced his 4mm one to 3mm for me, along with buffers and springs / axle-boxes.

528_211857_220000000.jpg528_211857_220000000.jpg

When it arrived it there was a noticeable bow to the sides. Came back to it 3 weeks later and the sides were straight. The sides are sufficiently thin that I wouldn't have been surprised if they had remained warped.

Nigel

Hi Nigel,
Thanks for the pic. Were the sides bowed in or out ? What material is it printed in ?

That is quite a loading ( 15T ) for a wagon of the Cambrian period. Presumably it was for internal coal use to loco sheds ?

Rob


posted: 22 Feb 2019 01:02

from:

Nigel Brown
 
 

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Honestly can't remember whether it was bowed in or out. Seem to remember when first painted it was bowed the other way, but again straightened itself more or less. Here it is painted:

528_211955_150000000.jpg528_211955_150000000.jpg

Just checked it again. There's a very slight bow out on the sides and bow in on the ends, but difficult to see unless you actually look for it.

I think the material was what used to be called FED, slightly higher quality than FUD. The buffer guides etc were the same.

They were loco coal wagons built around 1904, Pickerings built 50 and the Cambrian the remaining 20.

Nigel

posted: 22 Feb 2019 14:29

from:

Rob Manchester
 
Manchester - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Thanks for the background info, very interesting.

Rob

Edit -> Sorry Nigel should have made it clear I was saying thank you to you.



Last edited on 26 Feb 2019 00:08 by Rob Manchester
posted: 25 Feb 2019 03:50

from:

Andrew Barrowman
 
USA

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
MatterHackers are now selling a filament called Ryno. It's supposed to have the strength and flex characteristics of PETG without the tendency to 'string'. It might be a good option for chairs but I have not tried it.

posted: 6 Apr 2019 14:40

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
I haven't done much more on this project recently, I've been busy with other things. I must get back to it soon. Currently it needs some work on the gauging, for rail sections which don't match the prototype dimensions.

Looking back at the prints I made a couple of months ago, I'm more than ever convinced that the prints improve with keeping. It's difficult to be sure of a mechanism for this, but it is likely related to differential shrinkage. Certainly I know from injection moulding that mouldings age-shrink if not injected at a high enough pressure. And 3D filament is extruded at a tiny fraction of injection-moulding pressures.

Judging the results with a fresh eye, I think the 7mm results are really quite good -- on a par with 3D resin prints, and not far behind injection-moulded chairs.

The 4mm results are more iffy. At least so far. Fine at normal layout viewing distance, but not so good in cruel digital close-ups

Nothing more to report, I'm afraid. I've been trying to find some greater consistency and easier threading for the 4mm prints, without much success. I'm reluctant to say that this process is suitable for 7mm and larger only, but in 4mm we do seem to be near the limit of what is possible with a home-level filament printer.

Also I have spent some time on the file definitions for the switch and crossing chairs -- with not much yet to show for it, except the realisation that it's going to be an even longer job than I thought. :( Certainly there is nothing yet which I would be happy to release for general use.

Which is another way of saying that it is now over 12 months since the last Templot update 219b was released, and there are several other changes in progress which are waiting to be made available.

So I think I will park this chairing project for the time being, and get back to making the next program update available soon, but without this chairing stuff. Sorry if that is disappointing for anyone. But I will get back to the chairs in due course, if only because I now want them for myself.

cheers,

Martin.

posted: 17 Apr 2019 21:21

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
Martin Wynne wrote:
Martin Wynne wrote:

and get back to making the next program update available soon, but without this chairing stuff. Sorry if that is disappointing for anyone.
Partial rewind on that. I have decided to include some experimental chairing in 223a, so that folks can experiment with 3D print settings if they wish. But only for plain track at this stage, and only for REA S1 ordinary chairs (3-hole):

2_171616_200000002.png2_171616_200000002.png


2_171616_200000001.png2_171616_200000001.png


223a should be available soon.

cheers,

Martin.

posted: 19 Apr 2019 13:08

from:

Martin Wynne
 
West Of The Severn - United Kingdom

click the date to link to this post
click member name to view archived images
view images in gallery view images as slides
I have added a function to make fine adjustments to the track gauge after doing trial 3D prints:

2_190744_590000000.png2_190744_590000000.png


This is for 3D printing only, not for template design. The chair jaws and rail seat are shifted a fraction, but not the chair base (because that is fixed and shown on the 2D printed templates). What this means is that if you enter more than a tiny gauge adjustment, you will either spoil the chairs, or get a very silly result like this: :)

2_190748_530000000.png2_190748_530000000.png

If the gauge is more than a fraction out after 3D printing, it would be better to scale the entire output to suit your 3D printer -- output > enlarge/reduce size > menu items, and tick the scaled box at the top of the DXF dialog above. Or alternatively do the scaling in your printer's slicer software.

p.s. I'm posting these notes before 223a is released mainly as an aide-memoire for myself. Otherwise I shall almost certainly forget to mention some of the details later. Or all of them. :(

cheers,

Martin.



Templot Club > Forums > Templot talk > progress report - chairs in the output
about Templot Club

Templot Companion - User Guide - A-Z Index Templot Explained for beginners Please click: important information for new members and first-time visitors.
indexing link for search engines

back to top of page


Please read this important note about copyright: Unless stated otherwise, all the files submitted to this web site are copyright and the property of the respective contributor. You are welcome to use them for your own personal non-commercial purposes, and in your messages on this web site. If you want to publish any of this material elsewhere or use it commercially, you must first obtain the owner's permission to do so.
The small print: All material submitted to this web site is the responsibility of the respective contributor. By submitting material to this web site you acknowledge that you accept full responsibility for the material submitted. The owner of this web site is not responsible for any content displayed here other than his own contributions. The owner of this web site may edit, modify or remove any content at any time without giving notice or reason. Problems with this web site? Contact webmaster@templot.com.   This web site uses cookies: click for information.  
© 2020  

Powered by UltraBB - © 2009 Data 1 Systems