laser reprap and toroid focus?

I've seen tons of videos and whatnot on DMLS for plastic and metals but this technique doesn't seem very doable for a home 3d printer.

My idea on solving the metal printing problem was to use a laser to melt the tip of a metal rod through a feeding mechanism.

Obviously there are some design hurdles and I've thought this out quit a bit. Here is a list of problems And solutions i've thought about:

1) High powered lasers can get expensive. I see you can buy 1-20w diodes in cmount style for less than a few hundred dollars. Still trying to figure out a complete price but i have a goal of around $500 or less(possibly more but not by much to make this a worthwhile project) for the complete laser setup; diode, driver, case, optics, etc

2) focusing a laser on a small target is difficult. I've seen the cmount modules on aliexpress for around $15 but it seems the size would still be close to 1mm or more unless i customize it...and the laser cutting heads are around $100 but require mirrors and whatnot.

This leads me to my question on toroids. Since a laser beam is just an electromagnetic wave, couldn't you use an electromagnetic toroid to 'quench'/focus a beam tighter?

3)Feeding mechanism would require a 'purge' when near the end of a rod to prevent slippage...this isn't too difficult if i purge & Cut the new rod to the focal point. The idea would be the angle between the head/feeder is 45 degrees and 22.5 degrees from level for hopefully greatest accuracy....maybe less than 45 degrees?

4) tensile strength of melted metals in layering. Solutions include combination of foward and reverse feeding the rod combined with providing an opposite magnetic charge between the feed rod and the part to promote magnetic attraction from melted tip to the printed part, plus trying to center the beam between the feed tip and metal part in order to get a more consistent bond temperature. For this i would also want to provide time for an area to cool while working on another area, also using as thin of layers as possible.

5) oxidation. I was thinking aluminum would be a good starting point for metal plus its cheap, but oxidizes easily. Vacuum chambers and filling the chamber with a gas seems like overkill...so i thought of a low/medium pressure nozzle to apply nitrogen to the working area to prevent oxidation.

Any helpful input would be much appreciated. I am a software engineer and have a very high mechanical aptitude but don't have a lot of knowledge of lasers so if anyone can point me in the right direction on this would be awesome. I was thinking of adapting one of the open source firmwares to handle my specific application and use a reprap printer design to minimize reinventing most of the mechanics.

1) Yes they can. But to extrude a molten metal will take a nicely focused fiber coupled diode on the order of 100s of watts. A 60W diode can "melt" lead for instance if it is finely focused, but it would take 5-10 seconds to melt a lead bullet. You will need immense power to melt a metal fast enough for an "extrusion" process.

2) No. Light does not directly interact with magnetic fields in open air. Not even another beam of light will interact with it in the ways you might be thinking of. Optics and initial beam quality would be key.

3) Do you have a diagram explaining this?

4) You're essentially talking about welding this metal with a laser. Are you sure there aren't plasma welding solutions for this? I believe fire is still a good way to heat things up, and you cans till do it precisely. Fire is the leading cause of fire you know.

5) Or you could just make everything out of gold. Its soft, and doesn't "do" chemistry with hardly anything. (except maybe Fl)

methinks lasers are not the way to go. The system you've conceptualized is based around a laser that could not be built by the hobbiest and would end up costing 10x your budget to purchase.

By the way, what's wrong with the current 3d metal printing process? Its a bit slow in terms of printing I guess, but results are still faster than other metal shaping methods.

Meatball said:

1) Yes they can. But to extrude a molten metal will take a nicely focused fiber coupled diode on the order of 100s of watts. A 60W diode can "melt" lead for instance if it is finely focused, but it would take 5-10 seconds to melt a lead bullet. You will need immense power to melt a metal fast enough for an "extrusion" process.

2) No. Light does not directly interact with magnetic fields in open air. Not even another beam of light will interact with it in the ways you might be thinking of. Optics and initial beam quality would be key.

3) Do you have a diagram explaining this?

4) You're essentially talking about welding this metal with a laser. Are you sure there aren't plasma welding solutions for this? I believe fire is still a good way to heat things up, and you cans till do it precisely. Fire is the leading cause of fire you know.

5) Or you could just make everything out of gold. Its soft, and doesn't "do" chemistry with hardly anything. (except maybe Fl)

methinks lasers are not the way to go. The system you've conceptualized is based around a laser that could not be built by the hobbiest and would end up costing 10x your budget to purchase.

By the way, what's wrong with the current 3d metal printing process? Its a bit slow in terms of printing I guess, but results are still faster than other metal shaping methods.

Click to expand...

1) idea is To Melt the tip with said laser, not to extrude molten Metal....it would Be Akin to a feed welder But with a laser heater instead...so the 'Extrusion' is only to feed the solid rod to the work surface a fraction of a mm at a time.

2) since the metal is solid, when the Rod gets to a length shorter than the extruder tip to where it would cross the beam, the remaining rod would fall on the project, so you would 'Purge' remaining rod too short to work with without accidently dropping it....then feeding next rod further than needed,melt off excess with laser into a purge tray, then you know exactly where your tip is.

Btw, I'M talking about Melting a fraction of a cubic mm at any one time(with exception of the purge process), maybe like ~100 microns cubed. So this i was hoping would be more accurate and quicker with a laser.

I actually considered the DMLS method, creating a simple vaccum chamber and using a finely machined rake for layering, but i thought this idea might work better considering the danger of powdered aluminum if the vacuum fails...plus being able to use pla or abs spools, ramping down the power and still be able to print plastic with highly available material...not to mention price vs the DMLS printers run in the 10s of 1,000'S

Why not use an arc?

Cyparagon said:

Why not use an arc?

Click to expand...

You know, maybe you are right. I think with the expenses and power loss involved with a laser would put the laser idea to rest, except for non conducting materials but I wonder if arc method could be accurate to a fraction of a mm like I stated above. However a secondary laser tip could be used to maybe use abs or pla with a 1w or so laser, correct?

phillk6751 said:

You know, maybe you are right. I think with the expenses and power loss involved with a laser would put the laser idea to rest, except for non conducting materials but I wonder if arc method could be accurate to a fraction of a mm like I stated above. However a secondary laser tip could be used to maybe use abs or pla with a 1w or so laser, correct?

Click to expand...

Depends on the dimensions of your electrode tip, assuming you would be using a DC TIG arc.

Such plastics tend to melt away from a laser beam in the 1W power range. I think it is due to relatively slow heating. But then again I don't know what its like to melt abs with a CO2 beam - I gave mine up awhile ago. Perhaps powerful pulses can have a "welding" or softening effect? I don't know if its been done.

MIT achieves very high resolution with stereolithography. Give it a read:

FORM 1: An affordable, professional 3D printer by Formlabs ? Kickstarter

Stereolithography - Wikipedia, the free encyclopedia

Yea, i've seen the formlabs project, i think their process is patented, therefore i couldnt commercialize that. It does use a gel instead of spool 'wire' plastic, and prints from the top-down ... very cool design though..

I'm basically trying to commercialize a process for a home printer for various metals, and hoping for dual purposing plastics....maybe i'm just dreaming, but the DMLS videos i saw inspired me, and i thought if you can buy a laser cutter for around 2g, why can't a laser melt the tip of a rod for around the same price and strength....or are the 30-60w co2 lasers not cutting metals? Not to mention alluminum is much easier to melt than say stainless, which would be an ultimate goal in the future.

Oh so you need to sell it. I guess you've got to start from scratch!

Well for DMLS you'd need hundreds of watts in order to process enough material at a time, if you really want to retain the idea of saving time. You can use a 1.2W laser to perhaps melt a tiny bead of solder (lead-tin alloy, maybe silver) at the tiny focal point. Think of 'filament' size in terms of plastic extruder heads. What's the minimum diameter of a bead? 1mm or so? 2mm?

If you want to melt a metal powder surface on the order of mm's at a time with laser, you have to use a spot larger than the high-energy-density focal point's size. There's lots of energy density available in any laser beam's focused point, but the spot can be on the order of um in diameter - which doesn't let you process a lot of material at a time - it would be a slow system. So you need a beam with additional power to it, to where you can have a manageable bead size to process the material with.

Another problem with DMLS is that you cannot rely on the heat transfer properties of the metal since it is in a powdered form to begin with. The laser has to be bright enough to "process" the metal in a swift passing fashion, instead of waiting on the material to heat up before it melds together.

If you can get your hands on a fiber coupled 808nm diode array on the order of ~200W, or perhaps a CO2 system on the order of >150W, I think you would be in a good place to try out and experiment with your tip melting idea. I don't know how you would plan on simultaneously deposit the material as it is heated, (unless you used something more like a length of wire as your filament) but that's what the laser would be for, experimentation.

You could try looking up an engineering firm to consult you with the thermal stuff, the optics, the material processing. You've just got to try.

Meatball said:

Oh so you need to sell it. I guess you've got to start from scratch!

Well for DMLS you'd need hundreds of watts in order to process enough material at a time, if you really want to retain the idea of saving time. You can use a 1.2W laser to perhaps melt a tiny bead of solder (lead-tin alloy, maybe silver) at the tiny focal point. Think of 'filament' size in terms of plastic extruder heads. What's the minimum diameter of a bead? 1mm or so? 2mm?

If you want to melt a metal powder surface on the order of mm's at a time with laser, you have to use a spot larger than the high-energy-density focal point's size. There's lots of energy density available in any laser beam's focused point, but the spot can be on the order of um in diameter - which doesn't let you process a lot of material at a time - it would be a slow system. So you need a beam with additional power to it, to where you can have a manageable bead size to process the material with.

Another problem with DMLS is that you cannot rely on the heat transfer properties of the metal since it is in a powdered form to begin with. The laser has to be bright enough to "process" the metal in a swift passing fashion, instead of waiting on the material to heat up before it melds together.

If you can get your hands on a fiber coupled 808nm diode array on the order of ~200W, or perhaps a CO2 system on the order of >150W, I think you would be in a good place to try out and experiment with your tip melting idea. I don't know how you would plan on simultaneously deposit the material as it is heated, (unless you used something more like a length of wire as your filament) but that's what the laser would be for, experimentation.

You could try looking up an engineering firm to consult you with the thermal stuff, the optics, the material processing. You've just got to try.

Click to expand...

Aha! Thats close to what i'm thinking. Basically melt a 'drop' at a time. And yes, i would use the metal rod like one would use solder. And like in my first post i thought if i for example positively charge the rod, and give the work surface a neg charge, magnetic attraction would draw the molten bead to the project surface by path of least resistance which should in essence be straight down (for ferrous mats only though). I could use the rod feeder in a reverse direction to apply the laser to the project surface if additional melting is required for additional bond strength.

I figured with the accuracy of lasers, this method would create the finest bead size in order to attempt best bonding results.

About bead size, a 1mm diameter might be good for a quick-n-dirty build, but in the world of reprap, 1mm is quite crude. The makerbot claims a layer accuracy of 100 microns, and the filament is 1.75mm dia. I am not exactly sure of the x/y accuracy. However, wouldn't it build faster if i used the smallest bead i could successfully transfer with a consistent size and spacing?
Considering it would take less time for the heat to transfer through a smaller volume vs a larger volume....of course there probably is some sort of sweet spot.

I suppose it would make sense to somehow use a low mw laser to calculate temperature at the tip between duty cycles as to not over/underheat the tip....or maybe i could use a detector to use reflected light from the beam to determine the temperature. I'm not sure the accuracy of laser method of detection though.

Obviously i'd need enough an accuracy that would allow me to determine i haven't heated aluminum to 760c (autoignition temp) while melting point is only 100c below that, but get close enough to bond well.

So I performed a few calculations...and I see what you mean by needing a high powered laser...but I see this as very feasible with 10s of watts.

mass 1 mm^3 of aluminum = .0000027kg

used Specific Heat
to calculate using that mass, specific heat of .9cal/gm*C
from 20C to 660C (melting point from room temp)

it takes 6.514 joules to melt 1 cubic mm.

if i'm trying to melt .2mm^3 then it would take .04971488J to melt it

so here's some calculations on how many of these 'dots' i can melt per second with different Wattage outputs:

1w = 1j/s -> 20 * .2mm^3/sec = 4mm^3 / sec -> .4cm^3/sec
5w = 5j/s -> 100 * .2mm^3/sec = 20mm^3 / sec -> 2cm^3/sec
10w = 10j/s -> 201 * .2mm^3/sec = 40mm^3 / sec -> 4cm^3/sec
30w = 30j/s -> 603 * .2mm^3/sec = 120mm^3 / sec -> 12cm^3/sec
40w = 40j/s -> 804 * .2mm^3/sec = 160mm^4 / sec -> 16cm^3/sec

There are 16.387 cm^3 in 1in^3

So assuming above, to melt enough aluminum to make a 1 inch solid cube this would be 4400 cm^3

1w would take 4400/.4 sec = 11,000 sec -> 183min -> 3 hours
5w would take 4400/2 sec = 2200 sec -> 36.66min -> .611 hours
10w would take 4400/4 sec = 1100 sec -> 18.33min -> .3055 hours
30w would take 4400/12 sec = 366.66 sec -> 6.11min
40w would take 4400/16 sec = 275 sec -> 4.58min

with a 30 or 40w laser, that doesn't seem far off from a plastic reprap machine...lets see what these do:

Makerbot speed is 40 mm/sec (says flow is 24cc/hr..???)

So at that speed we're talking about 110 sec or 1.83 minutes to build the same cube at that speed.

wow...so aluminum appears very feasible with a 40+W at similar speeds to a makerbot. Definately slower, but not bad.

So with a flow rate of 24cc/hr means it can only extrude 6.66 mm^3/sec

So it either leaves x/y gaps, or works very slowly. With a solid at max flow that really means it would take 6606 Sec for the makerbot to make that solid with no gaps...or 110 min, 1hr 50min.

So in reality even a 5w laser would print aluminum faster than the makerbot prints plastic if you choose to leave micro holes....which would be good for nothing more than prototyping really.

Don't forget both that lasers are reflected of shiny surfaces and that the heat would spread around the whole piece of aluminium rather than just your fraction of a mm.