The finest Prusa ever made (and other bonkers ideas) !

Hi guys,

First of all, apologies for the grammar and such, english is not my native language. Secondly, there's a TL;DR at the bottom.

It's been a while since I was on this forum and at the time I just lurked, didn;t even have a username. I build a prusa about 2 years ago, maybe more, I don't remember well (which is basically the same time as when dinosaurs roamed the earth in terms of tech evolution).

Now, here's the catch: I have access and funding to an insane factory which simply makes brake calipers and pumps for really high performance european luxury vehicles and let's just leave it at that for now. I was recently talking to their general manager and he said Heck! Let's give it a go! If anything, we'll learn a lot from the experiences and set some future benchmarks if nothing else. They're not really keen to spend tens of thousands of currency on large scale commercial printers but
they are willing to give me free roam to their tech and see what I can come up with.

I have access to:

A metrology machine (or a coordinate measuring machine, it compares the final product to the original CAD file, accurate within .0's of a micron)

A Roentgen room, where they study random parts of the production line for inner defects

The most insane CNC mills you'll ever see (and of which I'm scared to operate for obvious costly reasons)

Tensile strenght test machines

500 tons force aluminium injection machines

Any tool and material I could want (but mostly aluminium)

Some old (but perfectly functional) vacuum chambers, large enough to fit a possible DMLS system in there, just not sure how low it actually goes, in terms of vaccum. Could also fill with argon.

Well, as you may imagine, the list goes on but these are just a few.


Now, in exchange for my free roam around the plant I'm supposed to at least try and accomplish the following:

1 Kickbutt Prusa Mendel, for starters, just to prove they work and they work well. We won't be using it too much but at least for an idea of protoypes and to get the more skeptical people behind other future projects, such as the following:

1 Powder printer, for ceramics. Proven to work already so it will be a good start to see what we can accomplish. It needs to be bigger though.

From that point on, it's all experimental but I'm thinking galvos, lasers and of course, aluminium powder.


One big issue is their occasional crack in a mold (and those injection molds aren't cheap, let me tell you). If i can accomplish a way of repairing those using a 3d printer, that's the end of that right there I'm just not sure wether I can manage to 3d print on top of an existing object. I'm sure it can be accomplished though. Doing that in metal is a whole different ball game.

I will begin acquiring parts that I don't have already around Tuesday or Wedensday, but since then, let the ideas flow! Cost isn't really a big issue.

For my Prusa, my goal is to make a super accurate one, preferably one that passes the metrology machine test.

I'm thinking: For starters, make the bog standard one I made some years ago, 0.35 Nozzle, belt drive, etc. It will be a good start point.

Then: Super high finish on the smooth rods (we have what it takes) and chrome them, they won't get smoother than that. Ceramics bearings all around.

Leadscrews all around. Als W thinking of replacing the smooth rods alltogether with leadscrews. Will need some sort of gearbox to turn them both at the same time, what do you guys think ?

Dual extruder of course, 0.35 and 0.1 (and will have to deal with those issues as well)

I'll start with wade's extruder at first but will then seek a solution to improve it. Definetly aircraft grade aluminium for the gears and then try something different insead of the hobbed bolt.

Next step would be to supersize it, hopefully double the build area in size. I know heated build chamber and such would be good ideas, i'll see what I can do, but that's in the future.

What do you guys think about gearing down the steppers ? say to a ratio of 1/20. Would that help much? Thruthfully, i'll try it anyway, if the answer is "not really"

So, to conclude: If you have any idea, no matter how crazy or expensive, let me know and once the initial build phase is complete, I'll try and give those a go as well. Thank you so much for taking the time to read this!

TL;DR I have pretty much and tool and expensive machine to play with in order to accomplish printing in plastic, ceramics and have a go at metal. If you have any crazy ideas I'd be happy to see them put into practice.

Respectfully yours,
elktw

Awesome! Exactly what I have been looking for. Do you have access to a sinker EDM? I definitely haves some ideas you can use.

I haven't seen it upclose and personal while taking the first tour but I do believe what I saw in one of the rooms was such a machine, so my guess is yes. What do you have in mind ?

Well, edm gives you the option of making carbide tools. Also, if you can work out a decent printable electrode material, it would allow you to print complicated electrodes to be burned into injection molds etc. I think that would be more promising than repairing old molds with a 3d printer. I am assuming the molds are some exotic grade of very hard tool steel, which would be extremely difficult to print. However, if you had an easy way to make complicated, high precision graphite electrodes, then you might lower the cost of new tooling.

The molds are something like 60HRC steel and even if a single fin on it breaks, for instance, that's a few days of 3 people trying to repair and hone it back. I was merely thinking of having some sort of MIG welder rigged to a delta bot just to do the hard work. Afterwards, that honing process would still be done by hand, but significantly faster, depening on what I can accomplish with the setup. But that's miles away from the first project, I'm looking into dmls before that

I would like to see you convert an old CMM with air bearings, (almost zero friction).

I think any thing that you build will not be as accurate, precise, and smooth than
reusing an old worn out CMM that has air bearings.



An old CMM with a worn out table, but with working air bearings and air pump,
without software, and controls will cost between $2000 to $6000 USD.

If possible incorporate the built in encoders in a feed back loop to the software, and stepper motors.
With the larger mass of a granite gantry you might be able to use it to your advantage for other 3d manufacturing applications.

I was pricing them a few weeks ago, here's a small one (Size: X=14" Y=16" Z=12")
that would make for a nice 3d printer conversion, and it's cheap only $2000:
[www.ebay.com]

Edited 1 time(s). Last edit at 03/08/2014 07:52AM by A2.

I was thinking the exact same thing when I saw that expensive CMM they had (will post some pictures of some of the toys once I get my camera back from my GF), well why don't we just attach an extruder to that thing? That could work.

I'll stick to the prusa first though, I don't think it's wise to just ask for 3 grand right off the batt, without having anything to show beforehand.

Is it worth sticking air bearings on a conventional prusa ?

Later edit:

Some of the folks there suggested it might be worth to try a hidraulic actuation system rather than stepper motors. At least we would overcome any inertia, momentum and friction issues. Not sure how that would be implemented tough, some electrovalves perhaps ?

Edited 1 time(s). Last edit at 03/08/2014 09:28AM by elktw.

CMMs don't really move fast enough from what I have seen. Also, building a gantry that can hold .0005 probably would not be too difficult compared to extruding with that accuracy. I am thinking about a machine with a surface plate for a base and an epoxy granite filled gantry though, which would be a multi use 3d printer, inkjet printer, router, laser cutter, Etc.

Quote
Svdharma
I am thinking about a machine with a surface plate for a base and an epoxy granite filled gantry

Have you found a source for powdered granite?

Quote
Svdharma
CMMs don't really move fast enough from what I have seen.

Some CMM's typically the larger ones are motorized, and are too slow.

The small gantry (Size: X=14" Y=16" Z=12") on the CMM in the picture that I posted doesn't weigh much.
You can move a 100 lbs with your finger with air bearings, it's almost zero friction.
I used to operate a Mitutoyo B231, and I was always impressed by how easy it was to manually move hundreds of pounds with no effort.
I don't think you'll have a problem moving a small-light gantry quickly.

The Z axis is perfectly counter weighted.

If mold repair is their main reason for letting you do this, I hope they don't find out about Optomec's LENS system:

[www.optomec.com]

[www.youtube.com]

Not the main reason but one of the most important. They do know, i actually tokd them about it but the main issue is the 7 figure cost associated with such technologies. They said something along the lines if "ehy don't you just give it the ol' college try, in the long run it would still be cheaper"

Quote
A2

Quote
Svdharma
I am thinking about a machine with a surface plate for a base and an epoxy granite filled gantry

Have you found a source for powdered granite?

Powdered granite is sand. Practically speaking, you want a wide distribution of grain sizes between a coarse sand and gravel up to 1/5th the maximum thickness of the thinnest feature. Epoxy granite has several times the dampening capability of actual granite and AFAIK just as good or better thermal stability.

Quote
A2

Quote
Svdharma
CMMs don't really move fast enough from what I have seen.

Some CMM's typically the larger ones are motorized, and are too slow.

A converted CMM would need to be motorized in order to be used as a 3d printer, so if it didn't come with motors you would have to jury rig something.

Quote
A2
The small gantry (Size: X=14" Y=16" Z=12") on the CMM in the picture that I posted doesn't weigh much.
You can move a 100 lbs with your finger with air bearings, it's almost zero friction.
I used to operate a Mitutoyo B231, and I was always impressed by how easy it was to manually move hundreds of pounds with no effort.
I don't think you'll have a problem moving a small-light gantry quickly.

The Z axis is perfectly counter weighted.

I agree that air bearings are nice, but think about the inertia of a hundred plus pound Gantry, that you are sending in a zig zag at top speed. Say you have a 200kg gantry and you want a 30 ipm feed rate. 30 Inches is roughly .762 meters so the force required to accelerate your 200kg gantry to .762m/s is roughly 152 newtons assuming zero friction, meaning every movement in a zig zag pattern requires 300 newtons of force (you can't forget about decceleration)

Point being, yes, you absolutely could convert a CMM into a 3d printer, and it would be very accurate, and it might even be a great multi purpose machine, but it would also be very slow compared to machines with only 5 or 10 lbs of inertia in the x/y movements.

I am going to respond to the question of gearing down:

I could be wrong but I am not aware of any printer that can actually print at its full movement speed (unless of course its full movement speed is geared down), so it leaves a lot of scope for gearing down while still retaining the same - or better print speed.

I am of the opinion that you will get *better* print speeds with lower gearing (up to a point where your movement becomes to slow) as with lower gears you get *faster acceleration*, unless you are printing rectangles (at full movement speed - hah) around the extremity of your print area your average print consists of short movements with several directional changes, which are achieved quicker with faster acceleration (assuming of course that you have kept the weight of your moving parts to a minimum to reduce the inertia as much as possible).

With lead screws and a larger print volume I would have thought that you would want to move to MENA 23's or bigger, you also may wish to look at other designs which keep the weight of your moving parts down.

Edited 1 time(s). Last edit at 03/19/2014 05:58AM by Slapparoo.

Dear slapparoo (nice nickname btw),

I understand what you are saying, however my question was in regards of precision rather than speed, geared down would mean more steps/mm hence more precision. I will definetly do that and post here with the result. Thank you so much for you reply (and it bums me out TBH that I just posted a "hey I have pretty much any tech we need and the funding to do something amazing" and get so many replys. Oooh well...

Now, an update on the build:

As I now live far away from HMS Britain it's quite hard to order anything without either waiting 30 days for a part to arrive or buy it locally at 5X the cost. But I just got all of the electronics today and will go shopping for some vitamins tomorrow. I settled on a megatronics board with magnetic endstops in the end. simple and cheerful.

Now, it's settled: The first small Prusa will be bog standard at first then will be converted to a dual nozzle (0.33 & 0.1) leadscrew in the second stage then further experimented with the same dual nozzle setup and leadscrews and geared down steppers.

BUT, AND HERE'S THE ALL CAPS BIG ANNOUNCEMENT:

I just had an idea (and the post serves as further proof that it's my idea or so I hope that noboy thought of it yet)

Instead of supports printed with the second nozzle, why not make a dual extruder with 1: A sort of combo of aluminium and alcohol, just to make it extrudable and 2: A ceramics extruder.

There are enough ceramics extruders out there to provide a good starting point and I could adapt them to some sort of aluminium powder paste.

The ceramics will serve as support everywhere, all around and the aluminium paste on the inside. This way, at the final stage, you just pick up a cube made of ceramics with the actual aluminium part on the inside. You put it in a kiln with the appropriate temperatures and voilla! The ceramic hardens, the aluminium melts on the inside but has nowhere to go (maybe include a sprue in each design to allow for shrinkage and whatnot) and in the end just smash the ceramic away with a hammer and you have an aluminium part.

The drawback I can think of is the ceramics nozzle diameter but with enough metal gears and strenght I think we could go down to .5 mm.

There will be some experimenting, mark my words!

Once more, thank you guys for the interest and the posts and apologies for any grammar mistakes (not my mother tongue)

Edit: Smash with a HAMMER not with a HARDENER

Edited 1 time(s). Last edit at 03/19/2014 03:45PM by elktw.