3D track - fun with laser-cutters

[Phil G]

Hi Martin,
In my experience which I admit is not vast,
I am finding by far the best option for the laser cut holes is to use the clip fit plug option.

I believe the reason is the clip is acting as a micro size scraper and pushing kerf residue around the clip as its pushed home. The residue then seems to behave in exactly the same way the FDM material does. IE holds the chair securely in position. The clip fit also leaves good provision for underside gluing if you want too.
My understanding is James has also found clip fit works fine with laser cut timbers as well.
cheers
Phil.

 

[Martin Wynne]

Hi Martin,
In my experience which I admit is not vast,
I am finding by far the best option for the laser cut holes is to use the clip fit plug option.

I believe the reason is the clip is acting as a micro size scraper and pushing kerf residue around the clip as its pushed home. The residue then seems to behave in exactly the same way the FDM material does. IE holds the chair securely in position. The clip fit also leaves good provision for underside gluing if you want too.
My understanding is James has also found clip fit works fine with laser cut timbers as well.
cheers
Phil.

@Phil G @James Walters

Hi Phil,

That's interesting, thanks. Is that with plywood or MDF or something else?

I can imagine that it all rather depends on the composition and density of the different grades of plywood or other wood or card or cork, etc. Also whether the ply or card or cork or whatever has been sealed with shellac or other sealant.

Also perhaps on the thickness of the outer ply skin, relative to the depth of the side tang.

There is probably a lot of learning and trial-and-error needed before we can make settled recommendations on the best options to use in any given case. I'm desperately trying to prevent folks running away with the idea that everything is done and dusted and ready to go.

cheers,

Martin.

 

[James Walters]

I've had success with clip-fit in plywood, wouldn't like to say if I preferred it over push-fit or not. But the plugs do go in well and the socket size can be slightly more forgiving.

That said, this evening I used push-fit for this little experiment....

 

[Phil G]

To date I am not that impressed with AI, it does everything you don't want well, and very little you do want.
If you ask it to cook your boiled egg, I hope you like very hard boiled eggs
cheers
Phil,

 

[James Walters]

re your question about types of wood used, only plywood in my case that is Birch, Gabon and popular plywood, at 2, 2.5 and 3mm thick nominal. Maybe James can answer the question of MDF.
@James cool looking wagon turntable. I need a lot of these for my LNWR warehouse, 15 from memory including the three inside the ware house. are you driving that with your turntable design?
cheers
Phil,

Hi Phil, yes the little turntable works wirelessly from the controller which was in one of my videos recently. There may be a few of them on my S4 challenge entry, along with some nifty little working capstans.

Also, I wouldn't recommend MDF for track work. I only used some in my video as it was a demonstration piece, and the only material to-hand of a suitable thickness.
What I might recommend however as an alternative to ply is sheet acrylic. I've made a few bits and pieces with it, and it has a lot going for it, dimensionally very accurate, stable, cuts superbly (and consistently) and engraves nicely. How about wood grain, or those little milled chair seats you see on some timbers. Sure, it's a bit shiny and doesn't take paint well, but with a laser the surface can be suitably keyed. Another benefit to Acrylic is that being a hard material (which doesn't creep), it gives a push fit chair something to push against without deforming. In other words a more positive fit.

I'll post some experiments on here soon. I've made some with clear acrylic for demo purposes so that the chairs are visible within the sleepers, but they really don't photograph well so I'll do some in an coloured material.

 

[Phil G]

yes the little turntable works wirelessly from the controller which was in one of my videos recently. There may be a few of them on my S4 challenge entry, along with some nifty little working capstans.

Hi James,
Capstans were also used to pull the wagons into HN warehouse. I have been thinking how to model then working as well, I was thinking about magnets but had not formed any firm plans yet. will be interesting to see what you come up with. I do like the idea of shunting in a group of wagons (main vans) and then sending these individually into the warehouse via the turntable and capstans.

re acrylic, interesting comment. for me the real advantage of ply is all in the look after staining.

I can see what you getting at, but I think I would prefer FDM over acrylic. Each to his own however and there is never any harm in having a play with something different. I do agree MDF does not work on any level for timbering as a solution.
cheers
Phil,

 

[Phil G]

Hi James,
Forgot to ask when will that wireless turntable be on on your web site for sale? Which I believe you mentioned, along with the turnout motors control mechanize you were planning to offer for sale as well?
chees
Phil,

 

[James Walters]

Not sure yet, still lots of development work to do. The system works brilliantly, but there are some quality of life improvements to add.
I'll post more details when I can be accurate about them.
Best,
James

 

[Martin Wynne]

I do agree MDF does not work on any level for timbering as a solution.

@Phil G

Hi Phil,

Not so fast!

MDF may not be too successful for through-cutting, but it works very well for 2.5D CNC milling. It is made from wood after all:

EM. 4mm MDF. timbers 1mm deep. sockets 3mm deep.

There is a surface texture which when stained looks fine from any normal distance. Sawmill marks and splits can be easily scratched into the surface if a more distressed look is wanted for old sidings, etc.

I haven't done anything with the CNC machine for months, but it's not forgotten and I want to get back to it eventually as another option for the plug track bases.

cheers,

Martin.

 

[Phil G]

Hi Martin,
I agree its there as an other option. I even thought it was maybe the go to option for a whilst, I have had a play with it on my router. it's quite good at the cutting stage I agree with that.
Have you tried staining MDF after its been machined though?

I did, admitting most stains here in NZ are water based. It caused the MDF to swell and in a couple of cases the sleepers to come apart! cant speak for your MDF, but on my router using local MDF the cut tests after staining were not very impressive .
I gave up with MDF at that point to be honest.

That's just me thought, I am not ruling it out for all by any means. Maybe with a solvent based stain you get a different result I don't know.
cheers
Phil,

 

[Martin Wynne]

Have you tried staining MDF after its been machined though?

@Phil G

Hi Phil,

Yes, using a water-based stain, without any detectable problems. But I applied it with a damp sponge, rather than sloshing it on with a brush and saturating the MDF.

I also tried artist's acrylic water-colours, which I rather liked. A streaky effect with slightly different colours can be achieved without too much effort. Real timbers are not all the same colour:

cheers,

Martin.

 

[Phil G]

Hi Martin,
I did as you say slosh it on. I will revisit the idea with a damp sponge approach just to see what I get, (depth of penetration is why I slosh by the way)

I 100% agree with you about variation, its the variation of colour in modelling that make it look more real, So I do try to create variation hues, I also think a bit about old oil in the center of the tracks from the locos.
Today that is very evident even on old concrete sleepers. No idea what it was like pre WW2 times to be honest, anybody have a view on that? Its noticeable from photos how well the area owned by the railway was kept at that time though.

My go to method on ply, is to slosh on the base coat to take it to something like a creosoted look, let it dry and the pick out with different stain options, we have over thirty shades at the local paint shop avaible in test pot size, all of which you can mix and match. I do like your idea of artist's acrylic water-colours though, I will give that a try for sure.
cheers
Phil

 

[Steve_Cornford]

Hi Phil,
I have found Indian Ink very effective on lasercut ply sleepers. Available in different colours & can be diluted to taste.
It lets the natural grain show through.
Steve

 

[Phil G]

Thanks Steve, that's interesting do you use black or is there a colour you start with?
will get some and have a play with that.
cheers
Phil,

 

[Phil G]

PS was the attached DFX file the art work you sent for laser cutting?
cheers
Phil

 

[Phil O]

What I might recommend however as an alternative to ply is sheet acrylic. I've made a few bits and pieces with it, and it has a lot going for it, dimensionally very accurate, stable, cuts superbly (and consistently) and engraves nicely. How about wood grain, or those little milled chair seats you see on some timbers. Sure, it's a bit shiny and doesn't take paint well, but with a laser the surface can be suitably keyed. Another benefit to Acrylic is that being a hard material (which doesn't creep), it gives a push fit chair something to push against without deforming. In other words a more positive fit.

Hi James,

How about giving the acrylic a rub with fine wet or dry paper to give the graining effect? And a key for paint. Would it affect the ability to laser cut it, if you abraded the surface before cutting? which would be quicker and easier than each individual sleeper or timber.

 

[Steve_Cornford]

Hi Phil,
Windsor & Newton Inks
Black Indian Ink,
Peat Brown.

I cannot remember who suggested it originally, but their advantage is that the ink just stains the timbers without interfering with the bonding of Exactoscale ABS chairs to the plywood using PlasticWeld (or MEK etc).
Some folks just use tea.....


DFX file
That was a mistake. I was going to comment on getting the grain of the ply oriented in the right direction, but then thought I had better make that comment in the Laser Cutting discussion instead, ersed my text but forgot to delete the attachment!
Steve

 

[Steve_Cornford]

Hi Phil,
Like you, I would quite like to use lasercut plywood timbers, but only if we can get the grain of the ply parallel to the long edges of the timbers.
If I take a LH B7 crossover as an example, I believe I am correct in stating that as long as we have not shoved any of the timbers, then if we laser cut a straight LH B7 crossover, we can use those timbers in a curved LH B7 crossover. Restrict any timber shoving to sleepers adjoining the crossover.
To this end I produced a box file containing a straight P4 LH B7 crossover with equalised incremental timbering 54mm TS track centres, then performed a 2D export and edited the resultant DFX file using Inkscape to rotate individual timbers (complete with their sockets), in order to get them aligned with the grain.
Then bunched the resultant timbers and removed the shortest of the duplicate lines between timbers (to stop the laser cutting twice & more importantly being charged twice).
The idea being to get these cut in 1.5mm or 1.6mm ply.

I then copied the straight B7 crossover (yellow) to produce a curved version (green), then exported a sockets only 2D DFX of the green one in order to get some 3mm cork sheet cut as a trackbed.

I have not managed to get either of these test pieces cut by anybody yet.

Steve

 

[Steve_Cornford]

Hi James,
In your latest video that shows the 3mm timbering base cut, was the laser cutting shown in real time?
Steve

 

[John Walker]

I have just completed my first cut using my Falcon 2 22W laser cutter in 1.5mm birch laserply from Hobarts. I bought some neodymium magnets to hold the corners down flat. I wanted to see exactly what was happening so I didn't use my cover and extractor fan. There was a heck of a lot of smoke. Now I know what happens I will use the cover!

I need to make some adjustments but I think I may be using too much power so that is probably the best thing to get right first. I set the speed to 10mm/sec and the power to 80%. Should I reduce both of them or try one at a time? The cuts a clean and dark brown.

Reards
John Walker

 

[Hayfield]

Hi James,
Capstans were also used to pull the wagons into HN warehouse. I have been thinking how to model then working as well, I was thinking about magnets but had not formed any firm plans yet. will be interesting to see what you come up with. I do like the idea of shunting in a group of wagons (main vans) and then sending these individually into the warehouse via the turntable and capstans.

re acrylic, interesting comment. for me the real advantage of ply is all in the look after staining.

I can see what you getting at, but I think I would prefer FDM over acrylic. Each to his own however and there is never any harm in having a play with something different. I do agree MDF does not work on any level for timbering as a solution.
cheers
Phil,

I think High Level Models do a wagon chassis,

 

[Phil G]

Hi John,
First step for any new material is to do power speed test grid, even if your not going to use lightburn (which I would recommend) get it on the free 1 month trail it has an excellent power to speed, program do to Laser tools>material test.
see this video

you don't need to do a material test by batch but for an new type or new thickness of wood or any other type of material you should.
you will basically end up with a material library of both type and thickness. Again Lightburn is great, as you can store this even by laser head type which is very useful see this video

This is primarily a safety precaution, as once you have the best power to speed ratio, you greatly reduce the likelihood of a fire.

my advise at least to start with, is try a cut though tests in both 1 and 2 pass options. Sometimes two passes is actually better than one but not always. Its very likely your best power speed is not 100% and very fast but a lower power setting and thus a lower and more manageable speed.

Also be aware lightburn can be set as both mm/min (there recommendation for diodes) or mm/sec there recommendation co2 lasers as there more powerful. I note James often uses mm/sec which I presume is simply because he started with a co2 laser. its not an issue as long as you remember diodes run slower so its more common to talk in mm/min
cheers
Phil,

 

[Phil G]

Hi John,
Do you have air assist? if not get one its basically a must have.
cheers
Phil,

 

[Paul Boyd]

I think High Level Models do a wagon chassis,

It was called the Flyshunter. I say “was” because it no longer seems to be listed on the website.

 

[James Walters]

I have just completed my first cut using my Falcon 2 22W laser cutter in 1.5mm birch laserply from Hobarts. I bought some neodymium magnets to hold the corners down flat. I wanted to see exactly what was happening so I didn't use my cover and extractor fan. There was a heck of a lot of smoke. Now I know what happens I will use the cover!

I need to make some adjustments but I think I may be using too much power so that is probably the best thing to get right first. I set the speed to 10mm/sec and the power to 80%. Should I reduce both of them or try one at a time? The cuts a clean and dark brown.

Reards
John Walker

Hi John,
Great to hear you making progress. The speed setting seems a little slow to me. The material tests which Phil has linked to are helpful, and well worth a try. I think you should up the speed to about 17mm/sec and run at 100% power. And see how that does. If it still severs the material jump up from there. I should think that 1.5mm ply should cut at about 25mm/sec with a clean lens, and perfect focus, but it's good to go a little below the theoretical maximum cut speed as it is very annoying to have areas which haven't cut right through.
I think the last lot of 1.5mm ply I cut was at 20mm/sec speed and 100% power.

 

[John Walker]

Thanks Phil and James.

Yes I am using air assist.

I used the Material Test Tool today. I used 4mm squares at 10/15/20/25/30 mm/sec and 64/66/68/.../100 % power. The squares fell out at 10mm/sec and 72% power and above. Nothing else cut right through. Since Lightburn seems to default at 10mm/sec and 80% I tried that.

Everything cut right through.

The nibs are still breaking too easily although I increased them to 0.5mm x 0.5mm. I'll check that and increase them again.

The sleepers measure 2.43 by 25.47 so very slightly under size, should be 2.5 x 25.5.

The slots have come out on size at 0.8mm wide and the chairs produced a few days ago fit them but are not easy to insert. I will use Templot to adjust the plug shape a little bit to be more pointed at the bottom with a slightly longer parallel section at the top.

The gauge is good if I push the rail hard against the outer jaws. The inner jaws are really thin and don't help with that. I had been playing with the length of the slots using resin printed sleepers. Those settings may be affecting this so I will revisit those.

Maybe this birch plywood is rather hard. It will be interesting to compare these results when I move on to the 3mm ply for S7. I have the much cheaper LaserPlyLite in Italian Poplar and Birchface/Italian poplar for that. They seem much softer and are much cheaper than the 1.5mm birch ply.

Regards
John Walker

 

[Phil G]

Hi James,
Just to clarify are you saying on 1.5 mm ply your are running at 1200 mm/min 100% power (20x60) on the Creality falcon 2 laser? and theoretically you could push it up to 1500mm/min if everything is in A1 condition?

Also do you have a view on the YouTube comments "its not advisable to run at 100% all the time with diodes" as it shortens there life quite a bit.

Interestingly with my laser tree 20W x-mass present, they don't advise always being on 100% for the same reason. (which could also be a clever cop out statement as well I guess)

 

[Phil G]

Hi John
re the material test there is a bit more to it than which fell out and which did not,
The way I found to be the best was to look for the ones which only just fell out or in fact has burnt though but not fallen out, that will be a combination of both power and speed, you can also measure the squares and you will find some are more dimensionally accurate than others.
Personally I try to avoid 100% power if I can, even if it means cutting slightly slower, that's because there is quite a bit of noise on the internet about always running at 100% will have a marked effect on laser head life. To be honest I don't know how correct that is.
I would be interested in James view on that one.

The sleepers measure 2.43 by 25.47 so very slightly under size, should be 2.5 x 25.5.

The slots have come out on size at 0.8mm wide and the chairs produced a few days ago fit them but are not easy to insert.

re the difference in size of the sockets and the sleepers, I agree I am finding the same. There are several ways to fix it, but for me its showing there is something different about an internal kerf offset and an external kerf offset. So you need to be carful as Templot only has a single kerf offset setting.
what does matter is to remember if and when you measure your material test holes there an internal kerf.
cheers
Phil

 

[Martin Wynne]

The nibs are still breaking too easily although I increased them to 0.5mm x 0.5mm. I'll check that and increase them again.

The sleepers measure 2.43 by 25.47 so very slightly under size, should be 2.5 x 25.5.

@John Walker

Hi John,

In the next program update I have changed the default nibs to 0.6mm.

If the sleepers are coming out under-size this could be because:

1. the cutter is cutting a kerf wider than you have set in Templot -- try increasing it, or

2. there is backlash on the axis drives -- try tightening the belts, or

3. there is an X/Y resolution error on the cutter, or

4. something else.

For 2D exports there is no shrinkage setting (plywood doesn't shrink when cut), but you can correct resolution errors by changing the output scaling:

cheers,

Martin.

 

[James Walters]

Hi James,
Just to clarify are you saying on 1.5 mm ply your are running at 1200 mm/min 100% power (20x60) on the Creality falcon 2 laser? and theoretically you could push it up to 1500mm/min if everything is in A1 condition?

Also do you have a view on the YouTube comments "its not advisable to run at 100% all the time with diodes" as it shortens there life quite a bit.

Interestingly with my laser tree 20W x-mass present, they don't advise always being on 100% for the same reason. (which could also be a clever cop out statement as well I guess)

Yes. Although I prefer to use units of mm/sec rather than mm/min. I find mm/sec is easy to visualise in my mind, and adjust on the fly.

Sure, everything wears out quicker if pushed to 100%. The manufacturer of the laser I use at work advised the same 10 years ago when the machine was installed. They also said the tube would only last 5 years with careful use. I usually (but not always) cut at 100%, and is still on the original tube.
Personally, I'd always use the fastest speed possible to achieve an accurate clean cut. Which may or may not be 100%. I don't see the logic in doing otherwise.
Bear in mind that for much of the time, (possibly much more than 50%), the laser may well be engraving on a very low power setting.

Should we slow the speed down too, to reduce belt wear?

 

[Martin Wynne]

re the difference in size of the sockets and the sleepers, I agree I am finding the same. There are several ways to fix it, but for me its showing there is something different about an internal kerf offset and an external kerf offset. So you need to be carful as Templot only has a single kerf offset setting.

@Phil G

Hi Phil,

I don't believe it. A laser beam can't know which side is the finished part, and which is the waste.

Templot has only one kerf setting, and is only going to have one kerf setting, because it is taken from the cutter specs.

You need to change the socket size settings for laser-cutting. The defaults are for FDM bases because I don't know what they should be for laser-cutting.

cheers,

Martin.

 

[Phil G]

Hi James, thanks for your thoughts.

Should we slow the speed down too, to reduce belt wear?

well to be honest on my setup belts are ever bit as important as laser head life, I find at too higher speed I do get overrun which I think is belt stretch not sure? and my max speed is only 2000 mm/min anyway.

To be honest unless your cutting a lot of straight lines the acceleration and deceleration values will have as much to do with the true speed as the actual speed setting I guess.
I would say I am being a bit more conservative on my power and speed than you are.
If I get chance I will have a play at slightly faster settings and see if there is any improvement.
cheers
Phil,

 

[Phil G]

@Phil G

Hi Phil,

I don't believe it. A laser beam can't know which side is the finished part, and which is the waste.

Templot has only one kerf setting, and is only going to have one kerf setting, because it is taken from the cutter specs.

You need to change the socket size settings for laser-cutting. The defaults are for FDM bases because I don't know what they should be for laser-cutting.

cheers,

Martin.

Hi Martin,
I am not arguing your logic here, only the simple fact my result's are constantly the same as John is getting. so something is different maybe inside or outside was a bad way of explaining it, maybe it's proximity to previously freshly cut material. I don't know why, but I do know its a measurably different value.
If you read my post it was actually about using the material setting in lightburn which closely resembles the same scenario you get when cuttings sockets.

I was not asking you to change anything, you have already made it very clear your are not going too.
I was simply pointing out to John there is only 1 Kerf off set setting in Templot.

FYI I have already found my own work around which I am happy with. Its outside of Templot, which is why I did not go into any detail and its my choice to do it that way,
cheers
Phil,

 

[Martin Wynne]

I am not arguing your logic here, only the simple fact my result's are constantly the same as John is getting. so something is different maybe inside or outside was a bad way of explaining it, maybe it's proximity to previously freshly cut material. I don't know why, but I do know its a measurably different value.

@Phil G

Hi Phil,

You keep saying this, but you haven't posted what dimensions you are using:

If the sockets are coming out the wrong size, why not change the size settings to get what you want, instead of assuming it is related to the cutter kerf?

cheers,

Martin.

 

[Phil G]

Hi Martin,
The simple answer is, because its not only in Templot am I seeing this issue.
so I am getting used to making offsets every time I use the laser, it has something to do with both the material and the material thickness
and its something I believe I would benefit by having a data table for.
the value for 2,5 and 3 mm plywood seems to be from 0.04 to 0.06 mm difference in kerf when say measuring a socket vis an outside dimensioned value.
cheers
Phil,

 

[Phil G]

Hi James,
Have you ever tried using a camera with lightburn? I note they actually sell some that seem to go inside enclosures and can be viewed on the lightburn screen whist working. There use is stated as observing the workflow! but I though could be useful for watching for hot embers etc. not sure they would pick that up though.
would welcome your thoughts?
cheers
phil,

 

[James Walters]

Not yet, although I have used one on a plasma cutter.
I think they are a bit like a DRO on a lathe or mill, not necessary, but nice to have.

If I invest in one, it would be fixed inside an enclosure, as I think it would be too easy to knock one of the type on a pole over the machine.
Lightburn sell cameras with a fish-eye lens which can be set-up to match the view on screen. This allows you to accurately place a drawing over a workpiece, as well as other benefits such as pass through cutting/engraving. But to be of any use in that regard they need to be properly calibrated, hence I'd prefer to have one which was fixed firmly. The inside of the lid of a CO2 machine being the ideal place. To be honest, I think the technology may have moved on since I considered one for my CO2 machine so my thoughts may be misleading.

Best,
James

 

[Martin Wynne]

Hi Martin,
The simple answer is, because its not only in Templot am I seeing this issue.
so I am getting used to making offsets every time I use the laser, it has something to do with both the material and the material thickness
and its something I believe I would benefit by having a data table for.
the value for 2,5 and 3 mm plywood seems to be from 0.04 to 0.06 mm difference in kerf when say measuring a socket vis an outside dimensioned value.
cheers
Phil,

@Phil G

Hi Phil,

You still haven't said what dimensions you are using in Templot, and what size the sockets actually measure?

If I had some hard figures, I could add some setting options in Templot.

But I know from past experience that the only way to get definitive information about anything is to get one myself and find out.

cheers,

Martin.

 

[Phil G]

Hi James,
Good info my current setup would lend itself to a lightburn fisheye mounted in the roof area see attached photo
@ Martin that's the 10W laser tree unit still attached, by the way the P-DA-01 is actual in the grey box just viable near the front door

cheers
Phil

 

[James Walters]

Agreed, that would seem to be a good place for a camera. I've not looked recently, but the Lightburn site had camera specs, and recommended particular ones for certain sizes of machine, and a fitting height above the bed.
For my own machine, the camera would have to be mounted on the inside front of the lid and would only work when the lid was up, such that the field of view would be enough to fit the whole bed in. Consequently it would be of no use with the machine running. But I don't think that's really the point of a camera system. If I need to see how the cut is going I peer in the window on the lid.