Oasis 3DP printer

Powder and inkjet printing
User avatar
dragonator
Site Admin
Posts: 419
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Oasis 3DP printer

Post by dragonator » Sun May 22, 2016 3:39 pm

I have gotten far enough with the HP45 that I can start with a new printer. I need something to test electronics and firmware with, and I might as well make it the functional printer I was planning to make. But first briefly what I have learned form the previous printers:

Focus: http://ytec3d.com/focus/
Was my first powder printer. It is cheap to build and can handle powder printing. That is basically all it does. It was designed as a cheap powder printing platform to test SLS and 3DP on. SLS never really worked on it, but the backup, 3DP did work good enough on it. (The fact that 3DP on focus was the backup is the reason Plan B is called Plan B).

The piston design jammed after a few hours of use and wobbled. The spreader design was novel, but also dead slow. Even the tiniest of prints took in the hour range (on a printer that jams every few hours). A good thing about Focus was that it was cheap and simple.

Plan B: http://ytec3d.com/plan-b/
Plan B was designed with speed in mind. The double feed piston and independently driven spreader were caused by this. On Plan B I used manufacturing techniques available to me at the time. The laser cut aluminium was easy for me to get, but difficult for others. Also the focus on speed made Plan B rather expensive.

The biggest issues with Plan B are that it only supports one printhead and everything is tied together. It is not tinker friendly at all. It is expensive and requires difficult to get parts. That said, Plan B ran way more reliable and a lot faster. At this point I found out that firmware is a lot bigger bottleneck than hardware on the speed. I never got good software to run for Plan B, something I hope to improve on the next printer.

Oasis:
I hope to make this printer a lot more hackable and easier to make. The laser cut aluminium frame has made way for a simpler wooden frame. The gantry and hoppers are (and should) not be connected to each other and can be changed independently. This is what the current design looks like.
Oasis Y00 WIP1.JPG
Oasis Y00 WIP1.JPG (141.45 KiB) Viewed 11310 times
The frame itself is a simple wooden box. The current itteration is designed to be cut from a single 122x61cm sheet of wood and has a size of roughly: 61x44x35cm. The design can quite easily be scaled to be made from different sheet sizes at this stage. On the top is a polycarbonate cover. This allows for a heated and/or draft free environment. I have 2 hinge designs, one requires more height to open, the other more space in the back.
Oasis Y10 hinge mechanisms.jpg
Oasis Y10 hinge mechanisms.jpg (115.02 KiB) Viewed 11310 times
The gantry is a very basic X-Y gantry with one motor for each axis. There are no complicated belt paths on this one. The gantry in it's current form can easily be scaled simply by using longer guide rods. At this moment the final axis is vertical, but I am still doubting about whether I should mount it vertical or horizontal. The spreader is going to be a Misumi rotary shaft. Misumi is an industrial supplier of everything mechanical, and a shaft is around 25 euros. I cannot design something cheaper and more accurate. The spreader in this design will be a rolling one.
Oasis Y20 basic overview.JPG
Oasis Y20 basic overview.JPG (92.55 KiB) Viewed 11310 times
The hoppers are more interesting on this design. There is no threaded rods and guide rods. I wanted Oasis to be tabletop, and any conventional piston design requires at least twice the stroke in space to work. The design is something I discussed with Ezrec here: http://ytec3d.com/forum/viewtopic.php?f=8&t=15
Oasis Y30 piston overview.JPG
Oasis Y30 piston overview.JPG (53.55 KiB) Viewed 11310 times
The basics are this. There is a piston that gets it's lateral guidance from the walls. The piston is suspended from 4 cables at the top and one cable from the bottom pulls it down. Elsewhere in the printer is a horizontal lead screw that actuates the cables of the piston. This way 1mm of lead screw gives 1mm of piston motion, every time. The cables run through 4mm bowden tubes to get to the pistons. The piston tubes are 120x120x5mm aluminium square tubes. Smaller tubes can be used to test a lot of small batches.

The design is still very fluid and can still be modified to suggestions by others. Want a feature, have a suggestion, have a question, please ask.

Wonko
Posts: 88
Joined: Sat Aug 15, 2015 8:13 pm

Re: Oasis 3DP printer

Post by Wonko » Mon May 23, 2016 8:49 am

This looks really fantastic!

I like the Y axis vertical design as it allows for an arbitrary number and size of ink cartridges. A horizontal design would be more limiting, you have plenty of height available, and the horizontal design could get in the way of the spreader.

I also like the design of the pistons. It must be spring-loaded somehow to keep the tension, I assume. The hopper will have to go down and up on every layer by a millimeter or two.

I am wondering it it is mechanically feasible to put the X rod in the front all the way in the back at the top of the case without compromising stability and precision.From my experience, anything with wet powder will be messy at some point. Powder will start to accumulate on the front rod, and it will clog the bearing, unless the user wipes it clean after every loading and unloading phase.

Last edit: or you could mount the mechanics on a frame that folds up, just like the lid itself. So for loading and unloading, all the linear motion transports would be completely out of harms way.
Attachments
dust_x.jpg
dust_x.jpg (61.91 KiB) Viewed 11299 times

User avatar
dragonator
Site Admin
Posts: 419
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Oasis 3DP printer

Post by dragonator » Mon May 23, 2016 6:15 pm

The reason why I was still doubting was because the horizontal Y-axis has nothing on the back. The circuit of the printhead is pointed that way. I will probably keep it vertical for now, because it is better.

In the pistons there is one wire in the middel pulling down. It is this wire that is spring loaded. The 4 wires coming from the top will be rigid.

Putting both X-rods in the back can mechanically not be done without compromising strength, but the whole gantry can be reinforced to compensate. As a rule of thumb, maximum deviation on a beam quadruples if it is overhanging as opposed to supported on both sides. Also, with a drive on only one side, the gantry needs to be more rigid in the sideways direction as well.

That is not to say that it isn't a good idea. Having the front free will make the design infinitely better with cleanup and accessibility (and looks). The X-axis is quite static, making only slow and constant movements. It is an ideal candidate for a floating axis. The dust cover would be optional (because it is also a dust trap), but I am very willing to give it a try. In the next few days I will design a gantry supported on only one side. It will probably have 10 or 12mm rods for the X-axis and a bit more material for rigidity, but I think I can make it work. My only concern is the rigidity of the spreader.

Putting the gantry in a movable frame sounds less desirable to me. I think I will have a harder time making that idea work. I will focus on a floating gantry for now.

User avatar
dragonator
Site Admin
Posts: 419
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Oasis 3DP printer

Post by dragonator » Wed May 25, 2016 7:16 pm

I did a lot of thinking on how to make the floating Y-axis construction stiff enough for practical use. I now have a concept that might work.

First of I reinforced the X-axis and put both axes in the back of the frame. The rods will be somewhere between 10 and 12mm thick. On the X-axis are 2 thick blocks that hold the bearings and connect to the Y-axis frame. The Y-axis frame gets most of it's strength from a 60x30x2mm aluminium tube. It is not even really the bending force I am trying to stop here. It is torque along the length of the Y-axis that is hardest to stop. The aluminium tube should be more than rigid enough. On the top of the aluminium tube are 2 threaded rods. These keep the tube up and provide a place to adjust the tube horizontal and vertical. mounted on the tube will be the Y-axis rods and the spreader.

I might still move a few dimensions around to get more horizontal stiffness, but all in all I think I can make this work. I like the open front the most about this concept. I have a busy weekend ahead of me, but I will try to get a few more hours of design into this concept to finalize it a bit.
Oasis Y00 WIP2.JPG
Oasis Y00 WIP2.JPG (165.66 KiB) Viewed 11277 times
Oasis Y00 WIP3.JPG
Oasis Y00 WIP3.JPG (166.08 KiB) Viewed 11277 times

Wonko
Posts: 88
Joined: Sat Aug 15, 2015 8:13 pm

Re: Oasis 3DP printer

Post by Wonko » Wed May 25, 2016 8:26 pm

Oh, wow! That does look very sturdy. Looking great!

Philipp
Posts: 34
Joined: Sat Mar 12, 2016 10:55 am

Re: Oasis 3DP printer

Post by Philipp » Sat May 28, 2016 10:03 am

Hello,
I have a proposal for a Z-axis:
You could work with a pulley like that.
You can use a timing belt as a rope.

The belt could run in slots. This could be generate by plates that is mountet square in side the The belt could run in slots. This could be generate by plates that is mounted inside the square profile.
Attachments
Z-Riementrieb1.jpg
Z-Riementrieb1.jpg (5.11 KiB) Viewed 11255 times

Philipp
Posts: 34
Joined: Sat Mar 12, 2016 10:55 am

Re: Oasis 3DP printer

Post by Philipp » Sat May 28, 2016 10:14 am

Hello,
You can do the same thing with two belts, the stiffness of the belt would prevent from tilting of the piston transversely to the belt as a belt on the right and a belt is attached to the left side of the piston. Then both belts must only run synchronously, the piston is forced out.
Best wishes
Philipp
Attachments
Z-Riementrieb2.jpg
Z-Riementrieb2.jpg (4.77 KiB) Viewed 11255 times

Philipp
Posts: 34
Joined: Sat Mar 12, 2016 10:55 am

Re: Oasis 3DP printer

Post by Philipp » Sun May 29, 2016 8:14 am

Hi Guys,
another even if very crazy solution is a screw without thread. Bearings run tangentially on a shaft or in a tube, are disposed inclined and tightened against the pipe or shaft.
In a round pipe which is quite simple in a square tube is slightly more complicated, but not impossible.
http://www.thingiverse.com/thing:310263

I could very well imagine that you can build the fabric feed of a sewing machine, a very good solution. Or something like a Cartridge gun. A discontinuous drive would actually fit very well to the movement.
Presse.jpeg
Presse.jpeg (1.21 KiB) Viewed 11246 times
What is a hydraulic solution. You placed in the space below the piston a balloon, which is filled with water. As balloon could use something like this:
https://www.youtube.com/watch?v=zuR15lxk8Wc
https://www.youtube.com/watch?v=taQAM4i ... tml5=False

These are just so proposals to stimulate the imagination.
Philipp

User avatar
dragonator
Site Admin
Posts: 419
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Oasis 3DP printer

Post by dragonator » Sun May 29, 2016 7:19 pm

The main challenge for Oasis is that I do not want lead screws and guide rods in the pistons. Mechanically speaking, guide rods and lead screws make the piston design a lot simpler, but to get a stroke of n, you need 2xn plus some additional space for mechanisms to make it work. This means that if I want a stroke of 150mm, I need at the very least 300mm of height to the top of the piston, and realistically speaking 350-380mm. This is why Plan B is quite a tall machine. Because I want a tabletop machine, I need something that works without guiderods or leadscrews in the piston itself. For example, the current design I have gives a maximum possible stroke of 180mm in a build height of 240mm.

Your first idea with the pulleys running through the piston would probably work, but it would not guide itself. I probably needs a lot of guidance in the piston to keep it straight. The second idea should work. Having belts on at least 3 corners will mean that the piston can be fairly thin. Keeping the powder out of the belt is tough. Synchronizing is a challenge, but not at all impossible.

I need to think about how a fabric feed like system could be made for a powder handling piston, but I am going to think about it. I do like the DIY bellows. Not for filling with water, but if my linear systems every get in trouble with powder clogging it up. These bellows are quite expensive to buy, but the ones in the link look really good.

Philipp
Posts: 34
Joined: Sat Mar 12, 2016 10:55 am

Re: Oasis 3DP printer

Post by Philipp » Mon May 30, 2016 8:20 am

Hello,
I think with the belt tribes is not so difficult. There is no relative motion between powder and belt, which is good in principle for the process. The belts run in a groove, and are connected to the piston, so they act as a feather key.

Maybe you can install a magnet in the piston, and can mount a spindle drive outside the cylinder, which moves a counter magnet? Then you would have the drive outside of the powder, and not much longer than the effective stroke, with a very simple mechanism.

Best wishes
Philipp

Post Reply