Water Jet Cutting

The posts about Big Blue Saw's free parts day (http://forums.reprap.org/read.php?1,14888 and [forums.reprap.org]) led me to Big Blue Saw's site, which has information on their laser cutting and waterjet cutting [www.bigbluesaw.com] methods.

Laser cutting seems to be getting some attention here already, but the waterjet cutting looked really cool. I'm thinking that it may be cheaper to implement some simple water jet toolheads than to use high powered lasers. The industrial costs of water jet systems seem to disagree, but I think we can come up with something.

A cheap (<$500) water jet toolhead would be great for RepRap. As far as I know, you don't need a sturdy cartbot as you would with mechanical cutting (milling). I imagine that it could find various implementations, parallel to a laser-cutting head's uses. That is, it could be used for cutting individual layers which are then stacked and bonded together (LOM), cutting entire objects (perhaps necessitating further mobility of the head or build object for cutting into sides), or perhaps even for smoothing out rough edges created by additive fabrication (like for gears and stuff). It should be noted that cutting depth can be controlled.

Any comments? I'll probably take a deeper look into this later.

One of the members of my robotics team happens to be an engineer for one of the largest waterjet-making companies in the world. I've toured the factory, and designed dozens of parts that he has cut on these machines.

Waterjet cutters require a pump that can produce incredible pressures, and components all along the way that can withstand that kind of pressure as well as a mixed abrasive. That's before you even get to trying to control and direct the stream. These can easily be multi-million-dollar machines, and they are dirty and take up whole rooms.

I know someone is going to say that everyone thought 3D-printing took millions of dollars in resources too, and look what we've done! However, I think that waterjets belong in the industrial-machinery class, unlike laser cutters and 3D printers that have been found in offices for decades now.

If you've got more questions I can put you in contact with my friend, he knows a ton more about this than I do.

EDIT: "It should be noted that cutting depth can be controlled." - wait, where did that come from? I've never heard of a waterjet that could control cutting depth... maybe you can differentiate between a light scoring and cutting all the way through, but trying to do anything between those is pretty much impossible, even for top-of-the-line machines.

Edited 1 time(s). Last edit at 07/28/2008 03:10PM by Kyle Corbitt.

Yeah, the infrastructure associated with waterjets is significant.

There are always plasma cutters of course... they're not as neat, but they are the sort of things that are affordable by mortal man. Not that they'd be hugely useful in such a small build area, but if you wanted to do sheet metal cutting they're certainly convenient compared to water or laser alternatives.

Waterjet is wet, noisy, slow and abrasive particles get everywhere. I used it for cutting brake lining material (copper loaded etc and not laser friendly).

Its great for cutting granite

Hm. Isn't traditional waterjet cutting equipment designed for cutting metals? If we were only going to cut thermoplastics, then couldn't we scale down the equipment? Wouldn't they require much lower pressures?

Oh, and remove abrasive materials and just use water (to not make such a mess as described by R.B.)?

As to the depth control... there is a max depth at which a waterjet can cut. What does this max depth depend on? Is it pressure? Could you not set the pressure for a desired depth?
Or rather, the speed at which the waterjet erodes (analogous to natural weathering) the target substance is limited by finite variables. Could you not adjust those variable(s) to adjust the cutting speed, and then cut for a certain duration at a certain speed to achieve a certain depth? Similarly, could you not pulse the jet such that it stops when the desired depth is achieved?

My understanding of the process is that in nature, water particles collide with solid particles on the surfaces of objects, like rocks. Given sufficient collision energy, the particles break away from the other particles around them (the rest of the rock). Given sufficient time, enough such collisions occur that a noticeable difference in shape and size of the object is achieved.
Waterjet cutting is this process, accelerated by having a continuous stream of highly energized (high speed/pressure) water particles focused on a small area of the surface.

It does not make sense that we cannot produce a thin, steady stream of water particles with adjustable power (energy/time). Reduce the pressure created by the pump and you reduce the speed of the stream and thus the erosion power of the waterjet.

Is it that there is a cutoff point, below which waterjet power is not sufficient to erode the target object, and above which the erosion is uncontrollably fast?
Or is it just that the erosion is varied and unpredictable (like at a certain waterjet power the speed of erosion varies significantly)?

I could summarize, but this is pretty good on its own, so I'll copy in the whole email:

"It all depends on what you want to cut, how accurately and how fast.
The highest pressure pumps you can buy in a store are pressure
washers, which rarely break 3ksi of pressure (as opposed to our
smallest pump, which makes 40ksi)

Abrasive is a huge benefit for speed for harder materials. You can
certainly cut thin plastic foam, and rubber with a water only head.

Controlling depth is pretty much out of the question. You basically
need to move at a fast uniform speed in a filling pattern over the
area, so reprap"s layer filling algorithms would be good there, but
the hardware just isn't fast enough. The closer you machine is to
infinite acceleration, the better.

My friend Al has seen some forums on-line where people are trying to
design their own waterjets. It's not very safe, and you have to use
good materials. It's best if you have access to FEA tools.

There are free tools online that will calculate you cutting speed
based on material, thickness, pressure, etc.

What was he planning to catch the jet with after it gets through the material?"

Basically, sounds like depth control is out, and water-only cutting is pretty difficult. Also a fair question about catching the stream - the machines I've seen use a several-foot-deep tank of water or large can of ball bearings to slow down the stream.

Hm... based on the limitations you just listed, I think I'll put the idea on the shelf for now.

I'll probably come back to it after I get a few things working on my own setups, but for now it's too theoretical, too challenging, and not necessary enough.

Oh well.

Let me dispel a few myths about waterjets / abrasivejets, and confirm some of the above information.

ETCHING / DEPTH CUTTING
=======================

Yes - waterjets have trouble with depth cutting. You can "mill" with a waterjet, and you can do things like etch part numbers into parts, etc. However, it is awkward to do, and only really useful in certain circumstances, and not something generally done. There are two basic methods used for milling. The first is to simply fire the jet and move it quickly - so quickly that it does not fully penetrate the material. Ideally, this is done at the lowest pressure the nozzle can handle before it has trouble pulling in abrasive (around 15,000 PSI or so.)

Here are some examples of etching/milling with a waterjet (abrasivejet), using the above mentioned technique:

[www.omax.com]

[www.omax.com]

[www.omax.com]

[www.omax.com]

A second method of milling involves cutting out a "mask" on the waterjet, and then either using another waterjet to scan over the mask quickly, or a sand blaster.

These methods can mill materials that are otherwise very difficult to mill, and they do have their place in obscure applications. However, generally speaking, part marking is better done with a laser, and milling is better done with a mill.

MESS AND NOISE
==============

Waterjets are all over the board here. Some machines are very quite and clean, and others are hell-messy beyond belief. Part of this is how they are run, and part of this is how they are designed. For the cleanest and quietest solution look for a fully enclosed system that cuts under water, and uses a direct drive pump. Under-water cutting helps prevent dust and splashing, and makes the cutting process very quiet. The enclosure is a backup in the event of something going wrong, to prevent the occasional spray. The direct drive pump is much quieter than the alternate "Intensifier" design.

MAKING YOUR OWN MACHINE
=======================

People do this, and for some people that

I wonder if a set of 1/4" steel waterjet RepStrap parts would be much different in price than the plywood ones from Ponoko.

[www.reprap.org]

Quote

I wonder if a set of 1/4" steel waterjet RepStrap parts would be much different in price than the plywood ones from Ponoko

You might not want to use these for the X axis/Y axis/Extruder, as they'd make a pretty heavy head to move around, and you might need new motors.

SnailRacer Wrote:
-------------------------------------------------------
> I wonder if a set of 1/4" steel waterjet RepStrap
> parts would be much different in price than the
> plywood ones from Ponoko.
>
> [www.reprap.org]


Any particular reason for steel? Waterjets can do just about any material known to man (except for tempered glass or something like that, I think). Of course, laser cutting is much cheaper for wood/plastic, but that's probably why you're asking about steel.

Ru,
because neither the extruder not the x- and the y-axis are lifted by the motor in any way, it shouldn't be a problem if they are heavier (physically)
Only thing that makes heavier parts difficult to move is higher friction,
so by adding some nice bushbearings that's not a problem

I'm unsure if making the parts from steel does make sense regarding the original size of darwin,
but it may, if the size is increased.

'sid

sid Wrote:
-------------------------------------------------------
> Ru,
> because neither the extruder not the x- and the
> y-axis are lifted by the motor in any way, it
> shouldn't be a problem if they are heavier
> (physically)
> Only thing that makes heavier parts difficult to
> move is higher friction,
> so by adding some nice bushbearings that's not a
> problem


Well, except that steel's greater mass/inertia means you have a lower maximum acceleration, and RepRap needs all the acceleration it can get if I read Nophead's blog correctly.

I don't think that one should care about the lower acceleration if the printingspeed is still under 20mm/s
I don't think that the y-Axis with the extruderhead (that's all that moves) will actually be that much slower just because the "plastic parts" are made of steel.
that's what? maybe 30-40% more weight?
keeping in mind that it is not lifted but moved horizontally,
I do think the original stepper motors are powerful enough to handle that little extra

Still, I don't like the thought of steelcut parts instead of reprapped ones,
that's so "not RepRap-able" that it hurts.
But: that is not my business but Tim's

If he likes to have steelparts.. that's alright with me, and I won't care about the extra weight that much (I would care about friction though )

'sid

>>>"I wonder if a set of 1/4" steel waterjet RepStrap parts would be much different in price than the plywood ones from Ponoko."<<<

If you have a DXF or DWG file of the part, or a photograph I can go from, I could give you ball-park what a job shop would charge for the part. (Visit [listings.waterjets.org] to find people who can do waterjet work for you.)

email me at carlolsen@waterjets.org

Let me know the part dimensions, material, and thickness.

Just about any material can be cut. Aluminum might be a good choice if you want metal but don't want weight. Teflon or similar plastic (UHMW?) might be good if you want low friction.

- Carl

Steel might survive the baggage handling gorillas.
[www.youtube.com]
Although aluminum is probably more practical due to weight concerns.

This would be for a RepStrap, the more options people have to get started the better. The idea is good because it follows the Darwin design, so parts might be interchangeable with Darwin once the kinks are worked out.

SnailRacer Wrote:
-------------------------------------------------------
> Steel might survive the baggage handling
> gorillas.
> [www.youtube.com]
> Although aluminum is probably more practical due
> to weight concerns.

Good point. If anyone happens to pursue this route, be sure to blog about it. One intermediate step I'm considering in getting to Darwin is to build a machine like this one, and my waterjet-running friend referenced above has offered to cut free parts. I was planning on using acrylic, and just a super over-engineered manufacturing technology >< but if aluminum works just as well but is more robust, perhaps I'll transfer to that.

...or I could just get my mill RepStrap in working order, and print the parts for myself! We'll see.

Indeed! If your friend has access to some free waterjet time you could cut the corner brackets out of something strong like aluminum at least.

According to Carl there is 0.030" lost to the cut width, and
smaller parts probably should be attached with small tabs to
keep them from falling into the catch tank if there isn't a
waterjet stone to be used.

He also says that wood cuts 10x faster than steel, and that
aluminum cuts 2x faster than steel, and plastic is somewhat
slower than wood.

Waterjet wood parts would probably be a better deal than the Ponoko
laser-cut parts (about $130/hr + setup time).

SnailRacer Wrote:
-------------------------------------------------------

> According to Carl there is 0.030" lost to the cut
> width, and
> smaller parts probably should be attached with
> small tabs to
> keep them from falling into the catch tank if
> there isn't a
> waterjet stone to be used.

Right, but I'm pretty sure that all water jet CAM software deals with the offset intelligently, and I know that in Flow's, at least, there's an option to auto-add the tabs. It's kind of finnicky though, and I've lost small parts in the tank before. But either way, that's something that the contract company should deal with for you, so I wouldn't worry about it.

> He also says that wood cuts 10x faster than steel,
> and that
> aluminum cuts 2x faster than steel, and plastic is
> somewhat
> slower than wood.
>
> Waterjet wood parts would probably be a better
> deal than the Ponoko
> laser-cut parts (about $130/hr + setup time).

I'd be careful. In the waterjets I've seen (which is not many) not only do they use supersonic water to cut the material, but it's pretty much sitting on top of a tank of water (to catch the stream) as well, and the spray gets everywhere. I'm pretty sure that MDF warps/swells if exposed to water. There may be other varieties of wood better suited to this treatment though.

Yeah, MDF wouldn't be a good choice for material.
I suspect the belt gear portions wouldn't hold up in MDF anyway.

Some sort of engineered hardwood would be better, or aluminum

I'm probably not going to follow this path for a RepStrap,
it's quite a bit more money than a McWire, or Tommelise type frame,
but it would be doable in an apartment, without excessive tools.

---

[olytim.blogspot.com]

in what situations is waterjet cutting best vs. laser cutting? what about crystal/solid state vs. C0

How about setting up a wire EDM? Use it on metals. Or an abrasive wire?

Would 1600bar in a highly focused jet be enough to cut thin sheet steel? (i.e. 1-2mm?

I opened a discussion at the RepLab group discussing this a bit further.

[groups.google.com]

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

What is the better option for cutting MDF 18mm board! Water Jet Machine or CNC Router Machine?

Some body told me that MDF Board can not cut on water jet machine because when it goes in the contact of water it will spoil. Is it true?

Reply soon to my mail id.

Regards,

Atul

Quote

Some body told me that MDF Board can not cut on water jet machine because when it goes in the contact of water it will spoil. Is it true?

Yes, at least for the cheap qualities. You don't need a water jet cutter to demonstrate this.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I would know what is the best machine to buy (waterjet/laser/plasma) i'm cutting a lot of differant materials?
I want to cut anything? What is the cost implication on laser and water?

atul Wrote:
-------------------------------------------------------
> Some body told me that MDF Board can not cut on
> water jet machine because when it goes in the
> contact of water it will spoil. Is it true?

The adhesive used in MDF is water soluble, so any part of the object that comes into contact with water. Better to use a CNC router, CNC jigsaw, or laser cutter.

Or, use plywood instead of MDF.

Has anyone considered using a car fuel injection pump as these go to incredible pressures (1000 Bar /15000psi). They are easily available and could be driven from an electric motor. They do tend to use the fuel as internal lubricant / cooling and I'm not sure of the flow rate.

What sort of pressures and flow rates are we looking for for cutting say a 1mm steel sheet?