Hi Brett
here are my motor screens.  It`s a long story but i`ve ended up with 3 different stepper motors.
X Axis has a 3.5mm pitch ballscrew, Y is a 5mm pitch and Z is 2mm

jimpinder thanks for your input, what I`m using is a reverse osmosis filtration unit which basically pumps the filtered water
into a pressure tank, on full flow my gauge shows a maximum output of 1.8 bar (around 24 psi) unless I had some sort of
pinch valve I`m not sure how I could control the flow using Mach3. My nozzle valve is a car`s fuel injector which is just a
thru flow open/close solenoid valve controlled by M8/M9 signals in Mach3.

vmax tnx I`ll look into the misters. I`m looking for the finest possible flow to get the best finish resolution.
I`m just a bit worried that any sort of high pressure will cause the layer to splatter as it lands.
I can buy a 0.05mm nozzle to suit my existing setup. The 0.127mm nozzle I have now produces a frozen line width of around
0.8 mm or so, which is too course.  This line width might be a lot finer if the build speed were a lot faster perhaps.

Still a long way off my target 

I should propably not ask this question, but this project of yours is really got me inquisitive, what are you modeling? I saw some info about the Ice Hotel and the patrons were using ice glasses and mugs for beverage consumption. That should keep a frosty mighty cold or what!

Whacko for cold beer

Say, could you adapt an airbrush but instead of paint, use water. I was thinking of using this as a marking device for a flatbed application. There is many ways to control the flow in Mach3. A bit off the topic, check out this link, http://fabathome.org

Whacko

Interesting isn't it Whacko?

Wil, I would guess you would need to be able to pull around 1.5Gs in your motor tuning to acheive a very clean path. I use this number only because on the simulations I mentioned earlier, that is what they were.

Brett

Yep, got to see the end product.
Wil, you could have some more solenoids in parrallel with needle valves that you preset. So you will activate the needed flow solenoid with a macro call or button/vbscript commands.
Just a thought, don't mind me if I'm way off! It's just that when I hear the word Ice, my mind goes wandering. Some Bourbon on ice, or an "ice" cold beer...

Whacko

howdy Whacko, tnx for your input.
I have actually built the fab@home electronics setup and got it working on my machine.
While they have done an excelent job on their project the build speeds are far too slow for what I want,
the max speed I got was 5mm per second. But really the thing that put me off fab@home was the gui
interface,I never felt I had much control, I much prefer using Gcode and Mach3.

A while back I bought a nice fine airbrush and yep of course it`s already been fully stripped down 
I have the needle valve and seat ready and waiting should I decide to go down that route.
At the moment I`m trying to stick with the nozzle arrangement I have, The holder I made
for the Nozzle valve also incorporates 3 power resistors mounted in a triangular array which
provide me with a heated up air gap (between 3 to 5 deg C) as close to the nozzle end as possible
to prevent nozzle freeze.

Being a bit impatient I could`nt wait the 2 weeks or so to get the smaller .05 mm nozzle so I stuck
a bit of epoxy resin to the bottom of my .25mm nozzle I had and ground a very fine point on the
end of a drill bit, I was able to gently push this thru the back of the nozzle till it just poked thru the epoxy
and have managed to make a .035mm dia hole (measured on a shadowgraph).
So armed with this I thought great I have a much smaller dia flow it`s got to help.
Hey man the flow runs well, at full pressure I can create a water jet the thickness of a hair and shoot this
for about 200mm before the jet breaks up. In practice I want the nozzle between 3 and 5mm from the build layer,
and reduce the pressure as much as possible.
In principle the machine is working, I must just match the flow and freeze speeds to the axis speeds, and still try sort
out this problem of flow blobs during axis pauses.

I think that problem is my hardware, I cannot get more G`s out of my motors without them stalling all over the place.
And so I cannot get the table speeds I need. Perhaps the Stepmaster Driver I`m using is`nt up to the job, or maybe
just bigger and faster stepper motors would be the solution.
If I hold the nozzle in my hand and literally sweep it quickly over a frozen plate, I get a lovely thin and blob free frozen line,
just the job.........

So thats where I am ATM.

Regarding my final goal.
As a mold toolmaker I always wondered about making prototypes a lot cheaper and faster. So I came up with this Ice idea.
Our company manufactures big pump assemblies and all too often bad designing causes problems on final assembly, brackets and
castings being in the way of pipework for example.
What I`d like to do is make scaled models of each of the assembly components, these can be put together as a miniature mock-up
of a proposed build and discussed while still in the design stages of a customers order.

So I want to create molds made from Ice, and my intention is to pour/inject pourable plastic into the cavity, let the plastic harden and melt the ice away to leave the 3D component.
I tell you it`s a pleasure working with water, there are no expensive consumables like those found in other Rapid Prototype systems.

here`s a few pics of the nozzle arrangement, the flow and a crappy blobby effort

Hi, Interesting projects. I dont know whether this is relevent, just an idea. If you have your flow output controlled by a soleniod (similar to how some car engine fuel injectors work) could you attach the solenoid to Mach's spindle step/direction output. By using steps you will be pulsing the soleniod with the S command and have full contol over flow.


John

Amazing! How do you stop the water from atomizing when it leaves the nozzle? I had an idea you wanted to build moulds, somehow like the lost wax method. 10 outa 10 for innovation to you Wil ! Is the expansion factor gaugeable?

As far as nozzles, have you thought of trying the nozzles you get in the High pressure waterjet cutting machines?

Whacko

thanks cjmerlin, that is an interesting idea. I have no clue as to step/dir or PWM so I`ll need to read up on that.
On the Ports & Pins>Spindle Setup .. the General Parameters settings, those Spin up/Spin down delays could be useful.

Actually Whacko I`m wondering if atomization would be a good thing, if the spray were to be broken in micro drops it may have better `freeze on contact` ability. I have been looking at Ultrasonics, the trouble I have is that the nozzle outlet has to be a close as poss to the solenoid valve, or you get pressure buildup and release problems and even drops forming at the nozzle tip.
In principle atomization is the same as mist (as mentioned earlier in the thread) but I cant see being able to make a nozzle produce mist within a circular area the size of .035mm dia.

Regarding expansion - the way I see it - what you spray and measure in terms of line thickness and height will determine the slice parameters you`ll end up using to get the best resolution. But yes I cannot wait to see how accurate say a 10 mm dia hole comes out and what size 10mm + - `plug` I`d have to make to fit the hole.
As you know most moulds consist of stationary or sliding cores and Icemoulds will be no different, apart from not having to slide.. hehe I can have any amount of `undercut` features and not worry how to `eject` the part. I just melt the mould away in 2 secs.

@ cjmerlin
yes I think PWM could be a good tool. Imagine a simplified senario as an example, a syringe. If the plunger was controlled by a stepper motor I could use PWM to slow it down on axis direction changes and have less flow at those points. When the plunger reaches the bottom of its stroke I`d have to have a limit switch that would pause the program and automatically retract the plunger which when it reaches the top of the stroke opens a valve that fills the chamber again and off the prog continues. Do you have any clues as to how to implement PWM at certain parts of a gcode program ?