Zcorp machine: How I think the layer system works

The printer makes it pass over the build table. The 2d image is made on the top layer of the material. And arm in front has a plow arm that pushes the material before the printing head to ensure the surface of this layer is smooth and uniformed.

That arm travels past another table which is full of build powder. After the plow arm is behind it. This table raises with more of the build powder.

The other table where the print work was done- drops down one layer.

This arm/plow sweeps the new raised powder area from the other side back over the build/print area table- when it returns to the otherside of the table. Process is reapeated all over again

Excess powder is deposited into a bin in front of the build table area.

What do you guys think?

Edited 1 time(s). Last edit at 07/20/2008 09:54AM by makeit.

*edit* D'oh, of course I've mentioned all this to you before. I'll quickly rewrite all this and hope no-one noticed.

Hurrah for funky animations to prove a point. Maybe I should try that approach

Someone (I forget who, sorry) mentioned that a counter-rotating roller is better than a blade, as it will compact the new layer somewhat as well as smooth it all in to place. Scorch's z-corp-alike uses a forward-rotating roller. I don't think the candy fab uses a roller at all. It seems so eminently sensible that I'd be quite surprised to learn that ploughing is sufficient.

The hopper idea is an interesting one though. I'd have assumed that an overhead hopper is mechanically more simple in almost every way, but a side mounted one would keep the overall high way down, especially for a large build volume.

Edited 1 time(s). Last edit at 07/20/2008 06:20PM by Ru.

Please remember that I'm not mechanical guy. I don't understand the terminology one bit. I have a very small understanding of what you guys fire back at me, and google is my friend. When you said "Hooper" I'm picturing the V shaped something over the top of the printer jet head arm- dispensing the powder somehow.

I have to decide soon on what machine I want to tackle and i have to come back to my guy with concise information that does most of the headwork for him. He's the one who understands stepper motors and programming and all that stuff.

I really and truly hope nothing I'm posting is coming off prickish or know it allish.

I'm plain dumb about all of this, and I'm trying to learn about it as quickly as I can.

heres an example of something I want to say but am afraid to.

I don't like how thick the layers are in the reprap. Think you guys should check wax out. I work with it a lot and it can flow out of an extruder in very micro hair thin layers. It would probably bond better too. Of course I don't know how you handle the dispensing trick. You have to heat and hold the wax at 400 degrees or so to have it flow like water you could go for a temp under this but using my waxer all the time i know that this is the water flow rate. And new layers bond to old ones better at this temp.

But where do I post such a thing, and how can it not come across as pompous. I'm so scared of coming across wrong to you guys. I really want you to know that like you I just want a machine that make great 3d parts from 3d computer data.

So what's the point of figuring out how a Z-corp printer works? It basically makes things out of starch impregnated with glue, iirc. From people I know who have them the objects printed are very fragile and of no use save for modeling solids.

I'm for printing things that I can use to build other things.

Z-corp prints can be impregnated with resin and made very strong: [www.youtube.com]
IE strong enough for skateboard

makeit Wrote:
-------------------------------------------------------
>
> But where do I post such a thing, and how can it
> not come across as pompous. I'm so scared of
> coming across wrong to you guys. I really want you
> to know that like you I just want a machine that
> make great 3d parts from 3d computer data.


The only stupid question is the one not asked mate.

Your experiences with waxes are probably beyond what anyone else here has dealt with. Expand on it, think about some ideas, do some basic experiments, and we'll hit you back with feedback related to other areas of the mechanical design that you might not have a lot of knowledge in.

I myself have not built a reprap (I've made the RP parts), however I run SLS Machines for a living, and work with with SLA and Objet occasionally, so I throw my two cents in whenever possible.

Forrest Higgs Wrote:
-------------------------------------------------------
> So what's the point of figuring out how a Z-corp
> printer works? It basically makes things out of
> starch impregnated with glue, iirc. From people I
> know who have them the objects printed are very
> fragile and of no use save for modeling solids.
>
> I'm for printing things that I can use to build
> other things.

I've pretty much learned that you guys want a final part from the machine.


This is not an issue for me. If it didn't have the physical properties I desired. I'd just make a flexible mold from it.

As is- the Zcorp machine looks like it has a detail issue. It can't reproduce fine lines and wrinkles, and the surface has a granular texture to it.

I pretty much figured if I go this way, I'll have to make a mold off of whatever comes out of the machine in silicone.

I can melt in chavant clay or wax into this mold, and rework that cast to smooth it out and add the detail back in.

If the part didn't need detail, but the material strength was too fragile- the same thing is done. Only now I cast a 2 part resin into the mold.

I totally understand where you guys are coming from and how a machine that doesn't offer that right from the start is a hassle/deterrent.


The prefactory is the dream. The model size on it will be smaller than I'd like but the prints are pretty much perfect for my wants and it's more appealing to you guys as well since there is little to no finish work, and the final part is rigid and strong.


Plotter files seem less complex to me, thats another reason I'm aiming for either one of these machines.

Theres one more machine out there called the REVO from [www.gemvision.com]

It uses 4 axis to cuts the 3d part out of wax. I like this machine, but it looks about impossible to make. Even if you could build it. The next big dilemma is the program


The program has to display, scale and edit the model. Then it has to set the cut paths and functions of the axis motors controls.

My models at the smallest poly count are 6 million polys. I know that several CAD programs don't like large file sizes and polycounts. Thats why the Meshlabs split into 2d' SVG image feature is something I like.

Meshlabs doesn't care how big my poly count is- so long as my PC has the Ram and Processing power to handle it, and once it's converted into slice information. The file size issue is gone- since the set SVG image data for the printer to print is going to have nearly the same file size no matter what kind of black and white information is there in each slice. The file size at this point is more about the images size now.

This is why I want to go with the plotter print system. Less technical mumbojumbo to make a part. You don't have to set up cutting paths in gcode.

At most you have to make/add support structures for your part.

Edited 1 time(s). Last edit at 07/21/2008 10:30AM by makeit.

makeit Wrote:
-------------------------------------------------------

> It uses 4 axis to cuts the 3d part out of wax. I
> like this machine, but it looks about impossible
> to make. Even if you could build it. The next big
> dilemma is the program

building a four axis cnc router isn't that complicated,
basically it's nothing but one rotary table more.

And most commercial milling soft can control these

Building a five and six axis machine however will lead you into some serious issues, a head that can turn in two directions will be hard to position (need to know its dimensions in every possible way no to hit anything)
But, you wanted a four axis machine...

A stepper driven rotary table shouldn't be too hard to make (worm gear transmission )

'sid

Edited 2 time(s). Last edit at 07/21/2008 11:11AM by sid.

Quote

A stepper driven rotary table shouldn't be too hard to make (worm gear transmission)

Easy to make, hard to make without backlash

The more I learn about the construction of machine tools, the more I despair of ever assembling a practical one myself!

Okay, simpler:
direct drive!

the only backlash is from the stepper itself...

'sid

sid Wrote:
-------------------------------------------------------
> Okay, simpler:
> direct drive!
>
Direct drive how?

How?

Build round plate with a flange (is that correct here?) drill a hole for the stepper axis, and attach it firmly.

Okay, some non moving plates would be more than useful to fix the rotary to the cartesian, also a ballbearing between the bed and the stepper could improve the quality.
but all in all you have a very simple rotary table with a low resolution.

jaja, clamping jaws required

'sid

... for direct driving a rotating table with a stepper you need a very powerfull one - i'll recommend more then 3Nm, so this would be a matter of size and weight ...

I'm applying a 4th rotary axis for my CNC-mill, so i could use it as a CNC-mini-lathe and for 4-axes- or XY-A-milling strategies too.

Here you must use microstepping - with 'normal' halfstepping you would receive 400 steps (or 800 with 0,9

Quote

It uses 4 axis to cuts the 3d part out of wax. I like this machine, but it looks about impossible to make. Even if you could build it. The next big dilemma is the program

Generating toolpaths for multiple axis machines is tricky, but not impossible. Certainly not if you're willing to spend money on it... commercial products exist to do 4-axis milling for perhaps only a couple of thousand dollars Not sure if there are any open source possibilities. 5-axis CAM tools exist, but the fact that no-one lists a price for them is probably a bad sign.

Driving a complex multi-axis machine is also tricky, but I do believe that EMC2 is capable of such a feat. People have used it to run hexapods and the like.

Quote

Okay, simpler:
direct drive!

Bleah. You're losing an enormous amount of torque already, so you won't be doing any milling on any particularly resistant materials. You also get a lousy resolution, meaning you'll need more complex toolpath generation to make your mill fill in the gaps between steps and even then you're going to end up with some faceting.

You could microstep as Viktor suggested, but then your torque drops even further as step size decreases.

As usual, there's a good reason why this approach isn't taken in the real world

When you apply torque to a stationary stepper motor it moves a little. It behaves like a spring. A powerful stepper behaves like a strong spring. The last thing you want when milling is any give between the workpiece and the tool. That will be why practical turntables use a worm drive.

---

[www.hydraraptor.blogspot.com]

I've seen those 4th Axis rotary tables in action,
it's not that they move back and forth like a linear stepper should.
They're just rotating (mostly very fast)

That means
1) by rotating in just ONE direction there's no backlash: Say hello to wormdrive
2) It could be done with an ordinary DC Motor and an encoder, just like the one for the extruder

'sid

Hi sid,

all the 3/4-axes-machines with rotary tables i've seen (you switch between Y- and A-axis) works either line-by-line (axial, 1+2-axes-mode) or cylindrical (2+1 or 3-axes simultan) - very expensive ones and 5-axes works 4- or 5-axis-simultan, but for our needs the 3-axes-modes should work fine ...

In the axial mode (1+2 = A+XZ) the rotating axe progress a single step an then the mill cuts the complete 2D-contour in XY along the part.

With the cylindrical mode (2+1 = AZ+Y or AZY-simultan-mode) the axe rotates nearly continuous, but it's synchronized with the Z-axe and the X-axe progress single steps per completed revolutions - so you have in both modes a 3-dimensional stepping, what's easier to realize with steppers, than with DC-motors ...

When you want to mill aluminium or harder metals, then you need a very rigid support - here worm-gaers, gear-boxes or tooth-belts are common, but you have to deal with mechanical backslash and/or friction between the gears.

When working with plastics, wood or hard waxes (best for lost-cast-moulding), then a direct driving strong stepper-motor is the better solution.

With small mill-heads (<2mm diameter) you can use steppers with metalls too, as the forces aren't so big.

But then you should have very high rotational speeds too - i'll mill then with maximum speed at 25000 rpm ...

Viktor

Speaking to the printed skate board posted far above...

You can see a scuff on the boards tail end from where he kicked it up into his hands. Also, he didn't do any tricks on it. I question the strength of the product because of this. A plastic skate board could stand up to that especially if it was that thick. Let's see some strength tests to make us believe in the strength of powder printing form z-corp.

Of course SLS of aluminum is a different story...

demented