Been a while since there has been discussion here but I thought I'd add some info I found as I was researching this.
The original Synova patent (https://patents.google.com/patent/EP1833636B1/de?oq=EP1833636B1) and the Avonisys patent that is derived from it (https://patents.google.com/patent/WO2017093331A1/en, it adds air jet stabilisation if I understand the patent correctly) are quite informative about how the required effects are achieved. There does appear however to be some patent dispute between the above 2 companies and I can't find the exact outcome if there was one:

https://www.patentlitigation.ch/synova-and-avonisys-litigating-over-water-jet-guided-laser-technology/
https://www.patentlitigation.ch/the-dark-side-of-the-novum/
https://www.laserfocusworld.com/industrial-laser-solutions/article/14221945/synova-files-appeal-against-avonisys-with-the-swiss-federal-supreme-court

The conclusion on the patent front thus has to be that the patents for neither party have run out and we should be careful about risking patent infringement litigation if attempting a DIY build. Theoretically it should be fine for hobbyist, home use but sharing the knowledge on how to do it could already land you in hot water, and where that line is drawn is... blurry at best. Tread carefully imho.

Edit: It also looks strongly to me like they settled and Avonisys AG went out of business. The LAM research site only mentions they're no longer able to take new customers. Whether this is awaiting the outcome of the appeal or final is up in the air.

Hi Peeps! Any advancements on this project? Just found it and looks AMAZING machine... I am currently investigating to build my own 1-2kw laser to be able to cut metal sheets up to 10-15mm thick, but this looks more promising, safer, and possibly much less power would be needed as the process retraces several times until the cut is made, so can be much cheaper build..!

From what I've read over the past couple of days, I would say less water pressure gives an unstable water jet.. higher pressure should help the jet stay coherent for more distance... adding the airjet around it also extends the range and reduces puddling

If there is anywhere else we can post comments and discuss advancing this machine, please advise!

Cheers!

This project has gone way down the priority list since my milling machine can do high precision work in steel and titanium now. But I would still love to see a low cost design for one of these. A common pressure washer pump should take care of the water, and aliexpress has 50W diode lasers that will probably work. You'll need to make a mount for the laser that can move in two axes so you can precisely center it over the nozzle, and may need fine angle adjustment as well. Also some way to focus it, which I haven't educated myself on yet. Perhaps the most difficult piece to fabricate will be the window chamber where the laser enters the water and passes through the nozzle. Other than that you've just got the water filtration system and a simple CNC XY table and frame.

As for places to discuss, I think this it. Though if you create a project page, we can discuss in the comments there too :)

Hey man! Please see my post to Lance further down, I feel he has knowledge I need to continue! Been reading a lot, modelling the Avonisys system, and would like to take this further on possible a better platform to be able to contact with participants more regularly and to be able to share docs on the material, etc.. I am quite disconnected with the current options, I am using Notion with my Precious Plastic delegation here, I don't know if this would be enough to be able to put everything in one place, but if is to be an open source product, I feel should be more publicly accessible..? Or maybe later we can post full plans on some open source website...*

*I understand Synova patents have expired, I have no knowledge of Avonisys patents, but as they are a younger company, they may still be vigente - for myself I have no intention of making a machine to sell, only for my own use, so AFAIK shouldn't have any problems..? scratch that, it appears that in the US even replicating a patent is infringement, in the UK it is not as long as there is no comercial gain

Would also have to find Avonisys patents and see really how much they have patented of the tech, given that the base tech was from expired Synova patent 🤔...

https://www.avonisys.com/avonisys-ag-wins-patent-litigation-case-against-synova-sa.html shit, they hold a load of patents! I live in Spain, so not sure if this is like UK or US... maybe will have to go back to the Synova system to be safe ;-)

Hello! I am writing a paper on an application for this technology, and I came across your post via Google. I can contribute to helping understand lasers as I'm the former laser application scientist for Control Laser. 

Thanks! The main question for this project is whether there's any viable route to acquiring a suitable laser. Diode pumped Nd:YAG seems to be the way to go, but I haven't found anything outside of impossibly expensive industrial price range.

There's also the question of how much laser power is needed in this application. If my thinking is correct, you only need to reach a minimum threshold of watts per square mm where some material gets vaporized, and then it's only a matter of time to cut through any thickness, unlike dry cutting where thicker stock needs more power since you have a short time to cut all the way through before it melts into a puddle. But I have no idea what that threshold might be (I'm mostly interested in cutting steel). Or what size of Nd:YAG rod would be needed for it. Is there an equation to calculate the pulse energy from the size of the rod?

I understand what you're saying, and I like the direction you're going. You also understand the concept of how the laser works; it's similar to your microwave: you can cook at 300Watts, and it takes twice as much time to cook than if you did it at 600W. There are nuances to it, but this is the basic theory; e.g., we assume that 300W cooks anything.

You're right in that a diode-pumped, Nd:YAG, laser is the way to go for cutting steel on a budget. This laser can provide you with high per-pulse energy, more than what a fiber laser could achieve. Further, the lower the q-switch frequency, the higher the pulse energy, which isn't necessarily true with fiber lasers because either firmware caps the limit, or it's not possible to go below the optimum pulse repetition frequency (PRFo), which is where we see advantages. 

The diode-pumped lasers I'm familiar with are those from the now-defunct Quantronix Corporation. Specifically, the Hawk model. The Hawk I version was a basic Fabry-Perot cavity using flat mirrors with a 10% Output Coupler (OC) optic (the front optic). The straight-cavity layout was the most basic diode-pumped, q-switched, laser: Highly Reflective (HR) mirror | Q-Switch | (Optional) Brewster polarizer | (Optional) Mode Selector | Pump Chamber | Shutter | OC mirror. 
Adding the mode selector allows to take the multi-mode beam and turn it into a low-order or even TEM00 beam at the sacrifice of power; a better-quality beam leads to better cutting quality. Adding the polarizer results in better-quality cutting, at a small sacrifice of power. You can add these two items later if you wanted to do some tests first.

The most common power output for industrial marking and lite engraving applications is 20-Watts, at a per-pulse energy of 2mJ; i.e., at 10kHz, 20-Watts of power is achieved. In the Hawk I, it means using a 50W pump chamber with a 3mm Nd:YAG rod, and 10% OC. The pulse-width is around 100ns. If I remember correctly, the q-switch was a 75W Gooch-Housego, operating at 27MHz. Frequency doesn't matter too much for what you want to do, just match the frequency to the q-switch resonant frequency. 
To run the system, you need: Diode power supply, water chiller (Termotek P300 is what we used with the Hawk, but you don't need DI water, so maybe there's another chiller with similar cooling specs?), RF generator (27MHz, 100W), and some sort of control card to make it work. The water for the chiller needs to have an additive to keep everything working properly. 

I've done some checking, and the store wavetopsign on Aliexpress appears to have all of the necessary items to create a basic laser at relatively affordable costs. I'm not affiliated at any way; I'm sharing my observation. I note that the diode modules they offer are clones of the first design Northrup Grumman Cutting Edge Optronics (NGCEO) came out with. The NGCEO module is what I put into the Hawk lasers when they needed to be repaired (this was my design, an improvement which Quantronix R&D signed off on). The Quantronix-built modules were notorious to fail within a year. In contrast, I spoke with a former customer recently who indicated the first laser I put the NGCEO module into 11-years ago is still going, so this is a mature design for a module.

As far as an equation for calculating pulse energy from the size of the rod, I frankly don't know. There are multiple factors that come into play aside from the rod size. Rod size can help determine the threshold laser, the output diameter, beam quality, and of course the amplification. However, other factors such as the OC %, cavity lengths, and others come into play when determining the final beam quality, and output power.  The software Zemax is what is often used to assist with such equations, but it's expensive. There may be alternatives that I am not aware of.

Hi Lance,

I respond to this post as the long detailed post below does not have a "reply" option..?

I found this article a couple of months ago after finding out about WJGL, and now have an itch thats needs scratching! I have been investigating and since reading this thru the first time, I now understand (just about) all of the text (thou I am still a self taught noob in lasers, so please understand if my questions should be obvious!), so I write to please pick your brain a little more ;-)

So after seeing the many types of laser, I can understand that a pico/femto second pulse width laser (either a Q-switched or pulsed pumped) is desirable as with a low average power, you can attain Kw pulse power enough to ablate and even plasmerize (?) metal, and the beauty of this method allowing to repass material cleanly many times means also can use lower power.. however, in the literature, studies, webs, and videos I have revised so far, there is an abyss in the quoted laser powers:

Avonisys caption a couple of their earlier videos with notes saying they were using 2800w

Synova state in some of their literature they use 50-400w

Various scientific papers state using laser powers of even down to 11w! (although some studies where more about scribing, drilling and grooving rather than cutting)

Various other papers talk about using multi Kw lasers too!

So I understand that Avonisys using 1064nm lasers probably lose a % of that to water absorption (although it is clear that have enough to still cut thru metal)(had found 1 study that mentioned *simulated* calcs gave 90% transmission to 532nm and only 20% to 1064nm, for a 2cm waterjet🤔) , and 532nm will absorb less, but also that 500bar pressure and 0.1mm or less column width will flush any heated water before 1064nm have much time to affect the transmission % (it certainly would not have time to evaporate, I think!),.. this is still a complex subject for me!

532nm is normal freq doubled from 1064nm, so there am losing power in transition, but the amount lost is possible less than the amount absorbed by 1064nm in water... 

Still haven't found much in the way of products on Aliexpress using 532nm, almost all is 1064nm (or blue lately which is not a choice AFAIK..?)

So after what I have read so far, am still in the dark (pun intended) about which laser and what power to use.. price is the most restricting factor, obviously if can use lower powers then better/cheaper, but haven't been able to find the starter kit you mention for 900$/50w, and am half and half about getting at least a 1kw fibre laser anyway, as I had already been thinking about making myself a dry metal laser cutter for the past few months, and being able to fallback to that option would always be good (however the extra chilling gas feeds, etc involved is why had not yet gone ahead!).... but all is open to new learning! 

Still in modelling stage, have managed to model a first try (no sizes used) of the Avonisys set up, I feel it may be slightly easier to use a separated air jet on stead of a coflow jet as Synove use, as I feel that coflow would have to be perfectly made to guide and to not interfiere and worsen the effect (although there are papers that show this method can have quite a beneficial effect on the length of water jet coherence, as long as I can maintain coherence even a few mm until the workpiece, atm I do not consider it priority to obtain really long waterjet coherence), am currently modelling the XY calibration mech, and once happy with it want to start looking for real components to be able to make a dimensioned model to finally end up with a BoM

I, too wonder what these low power, water guided lasers would be able to cut?  This method seems primarily able to overcome depth of field problems cutting thicker materials.  Not a trivial task but otherwise, you still seem to need the 10k+ metal cutting laser, or am I missing something?

What's interesting to me is that it can produce precise, high quality surfaces with minimal heat damage in super hard material. Wire EDM is the only other process I know of that can do the same, but is very slow and expensive.

But yes, I think the laser cost is also prohibitive. My hopes and dreams of making do with a cheap CO2 laser were dashed by the fact that water is opaque to infrared.

We are looking at this tool. 

I think one of the companies has a build-it-yourself video (it a satire video). There is also a cute video that shows https://www.youtube.com/watch?v=C1axX7hRMog something you can make yourself.

Depending on what you want to cut they allow you to pick your laser. It should be noted that these are Picosecond lasers, which allows more control over the process. (And on eBay if you can find a used one are still pricy). I was mystified they were not just pumping out huge volumes of steam. The big pluses are no heat-affected zone at the cut surface, good surface finish, There appeared to be no smoke or residue to contaminate the part with and they were claiming 1Um tolerance. (.00001 inch) on size and location. The machines Price wise run from $.6M to sky is the limit based on size and number of axis. So, I do not have enough information but since some of the Newport motion systems are in the same price range most of the cost are probably the motion system rather than the laser although it's probably a few bucks more. Because the cutting gap width is less than .002 inch your subsequent cuts either have to follow in that groove or recut the sidewalls so repeatability and location are going to be pretty important. (I believe a .006 inch nozzle diameter 10000 PSI generates about 100 pounds of pressure upwards lift, whatever you use will have to be stiff enough to control that and don't stick body parts in front of the jet. The laser alignment will have to allow you to center a .002 laser beam into a .006 diameter waterjet. So however you make your alignment is going to be another small challenge. (and don't forget to wear your laser safety goggles for the laser you pick). If you pick an infrared laser you will have to tell me how that's done.  Back to the water D-i water systems are available on Amazon and the things that take the air out of the water are also available. For the water nozzles, you might try Wardjet's website or there might be some other types of things out there I have in the past seen diamond and Safire donuts with .006 holes wardjet also has pumps which are generally too powerful for this application, but might guide you to used one that's not at full power.

https://www.plantservices.com/blogs/plant-nexus/diy-make-your-own-water-jet-cutter-for-less-than-200/

https://www.avonisys.com/avonisys-waterjet-guided-laser-systems-and-technology-integration-packages.html

https://www.laserfocusworld.com/software-accessories/positioning-support-accessories/article/16549947/waterguided-lasers-create-clean-cuts

https://core.ac.uk/download/pdf/38924027.pdf

http://ampl.mech.northwestern.edu/research/current-research/waterjet-guided.html

https://www.sciencedirect.com/science/article/pii/S2212827118307017

https://www.synova.ch/

https://www.wazer.com/

Thanks for the links! Looks like I wasn't the only one to think of a pressure washer pump, seeing that Applied Sciences video in the plantservices link :) I don't think we'll need 10000 PSI unless trying to cut really thick material. I'd planned to use the pump as-is, around 2000-3000 PSI.

I'd like to use a machine with an XY table and no Z axis. Judging by the Avonisys videos cutting through thick material without changing Z, it doesn't appear to matter exactly how much distance there is between the nozzle and cut. So the nozzle can be solidly mounted to deal with the recoil, and with the way XY tables move, the area under the jet will always be well supported against downward pressure. But it depends on whether I can protect the table from the laser. Hopefully having a layer of material that scatters the light will prevent damage (like how tattoo remover lasers scatter harmlessly in flesh). Otherwise I'd have to go with a CNC router style design with moving gantry over a stationary bed, which would have more stiffness issues.

You do make a good point about repeatability with such a fine nozzle. 2mm leadscrew with anti-backlash nut, 0.9 degree stepper motor and microstepping should theoretically have 2.5 micrometer (0.0001") resolution, but who knows how accurate it would really be. Hopefully the cutting still works even if the jet half-way overlaps a wall. The final part precision should still be a lot better than you'd get off a mill made from similarly cheap components, plus able to cut super hard materials.

I found those videos last week as well and looked into it a little.

You need to build a Z focus adjust to focus the laser just PAST / outside the nozzle of the water jet or the nozzle gets eaten away quickly. The nozzle needs to be diamond coated for the longest life, but I believe there are some available for regular water jets for pretty cheap.

You basically take a short thick wall metal tube and mount a nozzle in the front (like those used in 3D printers or oil burners) and drill a hole in the side to feed the water. The back of the tube gets threaded, so you can screw in a second tube with the lens mounted in it. As you turn the inner tube the lens moves up and down focusing the laser through the nozzle. You will probably need to grind a ridge for O-rings either side of the threads to keep water from leaking out.

The rest is mostly a xy table to move the work piece and a fiber laser. I think a CO2 laser could work as well so either buy or build one of those. You will need at minimum 40W to get any decent cutting but many of these systems are several hundred watts.

I don't know how you would make sure that no air bubbles form inside the head though or any "whitewater" which would de-focus or obstruct the beam.

That was a fun rabbit hole to dive into.  Avonysis has some nice videos.  Their mechanism seems to be arranged with the laser behind a pressure chamber and nozzle.  The pressure chamber seems to be fed from the side, which keeps the path clear for the laser.  Does the laser then seal its side for the water to go out the other.  Would be very interested if this can be done at low pressure!

Yep, here's a diagram showing that's exactly how it works https://www.synova.ch/technology/laser-microjet.html

Apparently Synova patented it 25 years ago, so I guess it ran out and that's why Avonisys is able to get in on the game. And why we're finally starting to hear about it, much like Stratasys with 3D printing.

And yeah, it is tempting to give it a try with one of those $15 130psi pumps on ebay before buying the bulky pressure washer pump. I'm sure it would limit the cutting depth, but maybe good enough for R&D work.

It's not looking like there will be an easy way to get laser that can burn metal, though. Sounds like pulsed Nd:YAG pumped by laser diodes is the way to go, but will have to be all custom built, and the diodes are fairly expensive.

I've been looking at Aliexpress where some shops are offering a diode-pump laser getting started kit for relatively cheap. I mean, still expensive for a DIY'er, but when purchasing a complete laser in the 10s of thousands of dollars is the alternative, it's cheap! Then, adding the optics and controls for green laser is recommended. Green is a wonderful color to cut with because it's absorbed readily by many materials. 

Ah, I didn't see this message before responding to your other one. Is it the 50W, $900 one? That may indeed work.

Do they show or discuss any applications other than metal cutting? I'm wondering if it would work on any materials that would be suitable for development (wood, delrin) and how to start the cut, given the fluid will perhaps spoil the beam where it hits the uncut surface, initially. 

It is exciting and I'm already wondering if you can use the water as a lens by making the point at which the laser enters the liquid a convex surface. 

They don't say anything about softer materials, but I don't see why it wouldn't work well on plastic. Wood would be too warped to use after it dries out. My primary interest is cutting cycloidal discs from hardened steel, so softer materials would only be for practice before I build a deadly monster. I tore apart an old DVD drive last night, so I now have a laser diode to play with :)

I don't expect starting the cut to be any different than continuing the cut. The Avonisys honeycomb video from November 2019 looks like there's usually quite a bit of splashing, and the stream just pierces through it for the most part.

Interesting thought with the water lens. I was thinking to use a flat glass window in the ceiling of the water chamber, with the laser and lens suspended above it with some kind of adjustment mechanism to tweak the position and angle. But you could put a tube full of water on top of the glass window, filled so the water forms a convex lens by surface tension. I suppose it would evaporate and change curvature over time, though.

I'm expecting great difficulty getting the laser aimed precisely at the nozzle opening, and focus correct to get it confined to the stream. At least with a low power laser I don't have to worry about burning the nozzle from the inside out.

You expect to cut metal with a DVD laser? You realize they are a few mW and not the 20+ W needed to cut metal. You also need a CONVEX not concave lens like meniscus of a tube of water would give you.

The DVD laser is just to practice aiming at a nozzle. My research over the past few days is not looking too promising for acquiring a metal cutting laser, though. Continuous wave CO2 lasers seem to be the only thing easily available, but pulsed high power seems to be what's needed. Wikipedia says it's easy to Q-switch a CO2 laser with a rotating mirror to get high pulsed power, but I can't find anyone actually doing it. Diode-pumped Nd:YAG seems to be what the pros are using, but very expensive.

Understanding the concepts of "near-field" and "far-field" is the secret to laser alignment. Essentially, you want a laser to pass through a near-field reference hole and a far-field reference hole without having the beam clip. I imagine someone has described this process on YouTube, but if you can't find anything, then I can help.
I don't recommend attempting to Q-Switch a CO2 laser; this will be quite a research project for you, and since water absorbs CO2 laser quickly, it's a loss.

Synova posted a few case studies where they have cut through IGBT modules, which have plastic housings. So, I imagine it's possible to machine these other materials, but they do mention "brittle materials" as being the target.

It is possible to use water as a lens. Usually, it's a mixture of oil and water that create the lens. Check out "varioptic lens" and "electrowettability."