Brainstorming: Compact piston drives

Powder and inkjet printing
ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Brainstorming: Compact piston drives

Post by ezrec »

For my next printer (a 'large desktop' size, currently in the concept phase) I'm trying to come up with ways to implement a more compact piston system.

I have a few ideas:

* A Spirllift style system http://pacospiralift.com/
- Pro: Off-the shelf device, in the scale needed for 'large desktop' powderbed systems
- Con: Looks quite expensive.
* Scissor jacks
- Pro: Should be quite compact
- Con: May be difficult to translate rotary motor action into lift movement
* A drag chain lift http://www.google.com/patents/US20120261629
- Pro: Might be doable with bicycle parts
- Con: Patented recently!

Any other ideas I should consider?
User avatar
dragonator
Site Admin
Posts: 605
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Brainstorming: Compact piston drives

Post by dragonator »

I am not probably not capable of making a full sized answer until at least Wednesday, so I'll throw my idea here now and add images of it later (publishing stuff right now, and finishing other stuff).

There is a set of systems I say on some ancient 3DP printer (think 80s) made by a guy I know who was one of the earliest inventors of the technique.

The idea is to have a set (3 or 4) of cables/wires running from the top, supporting the piston. The cables run along the wall, move under the piston, and keep it up. How the cable is moved depends on a lot of factors, but the cable can be winched up or down from a motor under the piston (which was the system he used) or from the side of the printer. Some other fancy wires are also needed to pull the piston down, but the basics do work quite well.

The biggest problem I found with any technique is not that it is hard to make the linear drive in a way that 1x drive height is 1x parts height, it is guidance. Guiding the piston along the whole way without taking more than 1x the space of the travel is more difficult. The only place to guide against is the walls, and powder also accumulates there.

Food for thought, I will think about it for the coming days.
ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Brainstorming: Compact piston drives

Post by ezrec »

A solution to the guidance problem is to use a wide coil spring under the powder plate.

Assume powder plate is attached to a coil spring, sufficient to fully eject a loaded part/feed bin when fully compressed. The powder plate's edges are rollers in bearings in the corners.

Further assume the powder plate's center is connected to a steel wire rope, which goes through the bottom of the spring, to a winch under the base plate. (The winch can be relocated anywhere you can mount pulleys to)

Terrible ASCII art:

Code: Select all

  ||......................................||
  ||......................................||
  ||======================================||
  ||O                                    O||
  ||==================X===================||
  ||            SSSS  :  SSSS             ||
  ||            SSSS  :  SSSS             ||
  ||            SSSS  :  SSSS             ||
  ||            SSSS  :  SSSS             ||
  ||            SSSS  :  SSSS             ||
  ||----------------| : |-----------------||
  ||                  :                   ||
                    WW:WW MMMMMMMM
                    WWWWW=MMMMMMMM
                    WW:WW MMMMMMMM

Key:
  || - Bin walls
  .. - Powder
  == - Piston plate
    O  - Piston plate rollers
    SS - Coil spring
    :  - Steel wire
    WW - Winch drum
    MM - Stepper motor
ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Brainstorming: Compact piston drives

Post by ezrec »

.. of course, the holding torque needed for this design is absurd, but it may help poke our brains into other directions.
ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Brainstorming: Compact piston drives

Post by ezrec »

.. which could be completely mitigated with the following winch gearbox.

Assume the steel wire is 1mm in diameter.

The wire will wind onto a winch spool, where the entire wire will wind onto a single layer on the drum.

The ends of the winch spool are 1mm higher than the wire, and threaded so that one complete turn of the winch will advance the winch 1mm. The rear of the spool has a high-torque gear (steel).

A threaded cylinder contains the winch spool. The cylinder has an opening for the wire to feed (at the top middle of the cylinder) and the sides are slotted with a linear bearing. The linear bearing contains a worm gear (engaging the cylinder's high-torque gear) and the work gear is attached to a stepper motor, and the end of the worm gear are engaged via a bearing to the spool's thread lip through the slot.

Theory of Operation:

The stepper motor rotates the worm gear, which rotates the spool, which either picks up or releases the steel wire, which compresses or releases the spring, which raises or lowers the piston.

The spool is threaded into the cylinder, so that the feed hole of the wire is always positioned such that a single layer of wire will be wound, with no overlap. This makes calculation of stepper rotations to piston position trivial, and prevents tangling in the spool.

The slots and linear bearings for the worm gear is to allow the worm gear to move with the winch spool as it moves along the cylinder.


... Probably easier to explain if I can make an OpenSCAD diagram with some animation...
User avatar
dragonator
Site Admin
Posts: 605
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Brainstorming: Compact piston drives

Post by dragonator »

The constant diameter winch is a good idea. The mechanism that properly fills the winch should actually not be necessary if the distance between the winch and the first guide wheel of the cable is kept somewhat large. I would also add a system to the winch that allows it to be adjusted. This would work by adding a stationary block to the winch shaft, and with an adjustment screw on the winch drum. Turning the winch drum screw would alter the position of the drum relative to the stationary block, allowing for fine adjustments (think a few millimeters).

A system like this is needed for one reason. I do not think that one cable and spring is capable of both keeping the platform stable and allowing adjustments to be made. The system itself should work just fine, but I would do it with 3 cables and springs instead of 1. The 3 cables can be mounted on the same winch, but adjustment needs to be possible.

For the spring I would look into conical springs. They can become as flat as the spring wire thickness when compressed. Getting them in the sizes required is quite hard, but you could wind them yourself. Worm gears should do the job of locking the piston into place, but will also reduce the speed and efficiency of the system (though efficiency is irrelevant due to the amount of travel). Getting a decent worm gear for a decent price it tough, but other than that it is a good idea.

The only real problem I have with the spring system in general (because I had it as a concept in my head for a time as well) is that it is not a mechanically locked system like a lead screw or cables from above AND below. The springs determine the entire system. They need to be strong enough to push up a full buildbox of any possible material, but every gram of force also goes straight into the printer. Due to the constant load on the printer, the usage of most plastics (especially 3D printed parts) is impossible around the buildbox. The parts would warp too much over time.

These problems could be stepped over if the rest of the system works fine though, it is just something to keep in mind when designing. Plastic warps over time with constant strain, metal almost not.
ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Brainstorming: Compact piston drives

Post by ezrec »

Yes - the "traveling winch spool" could easily be replaced with a traveling cable guide and pulley. In that case we would need some software adjustment to account for the moving guide, but it's a Small Matter Of Programming.

As for gears - metal all the way. The real trick would be to find "stock" worm and gear arrangements, an work with those, instead of a custom geartrain.

Ideally the winch could be purchased as a BOM from McMaster & Carr (or their Metric equivalent) with a small amount of machining required. Think hacksaw and drill press.
User avatar
dragonator
Site Admin
Posts: 605
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Brainstorming: Compact piston drives

Post by dragonator »

Another option would be to not wind a cable on a winch, but mount cables to a horizontal linear slide. This linear slide can then be moved with a lead screw. This way you don't have to worry about the constant diameter of the drum and the system becomes self locking. It also is easier to adjust if you have more than 1 cable (which I think you will need).

With the available materials, this may be an even easier system to build than finding fairly cheap worm gears.
ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Brainstorming: Compact piston drives

Post by ezrec »

dragonator wrote:Another option would be to not wind a cable on a winch, but mount cables to a horizontal linear slide.
And we have a winner! I like this one a lot, especially as with proper cable routing, this can do both the 'push' and 'pull' operations on the piston plates, and can drive both the part and feed bins at the same time with just one stepper.

I'm thinking 'very small' for this design - about the same total volume as the M3D FFM printer, so Spectra line should be sufficient for the < 1Kg of powder total.

Know any good software for designing laser-cut acrylic? Or OpenSCAD macro sets?
User avatar
dragonator
Site Admin
Posts: 605
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Brainstorming: Compact piston drives

Post by dragonator »

The Push pull on one axis is indeed a lot easier, if you want to mount the piston on cables coming from the top, and the bottom (to keep it tensioned). Alternatively, you can also mount springs on the cables if the only purpose of the cable is to keep tension on a system.

While the driving 2 pistons with one stepper may seem like a good place to save money, be careful. When preparing the powder bed it is incredibly handy to have independent control over the pistons. Also in some cases you might need to over-feed the feed piston to get a full layer on the build piston (to account for losses and compression). The later can be solved my making the feed piston slightly (<5%) bigger than the build piston. Other than that, it is a good place to save 50-100 Euro.
Post Reply