opensource cnc edge bender project

For discussions related to designing and making ski/snowboard-building equipment, such as presses, core profilers, edge benders, etc.

Moderators: Head Monkey, kelvin, bigKam, skidesmond, chrismp

User avatar
chrismp
Posts: 1443
Joined: Fri Feb 13, 2009 9:00 am
Location: Vienna, Austria

Post by chrismp »

The new DIWIRE 2.0 seems like it would be capable of bending steel edges with some modifications. http://www.pensalabs.com/

Here you can find the code for their software: https://code.google.com/p/diwire/source/browse/
ben_mtl
Posts: 583
Joined: Mon Mar 31, 2008 1:47 pm
Location: Sherbrooke, Quebec
Contact:

Post by ben_mtl »

looking at it right now :p
http://blog.pensanyc.com/tagged/diwire

I'm still unsure on how they bend long smooth radii (I'm not done with all their videos...) : do they set the bending pin at the right angle and feed the whole arc length, or do they feed 5mm -» bend -» retract -» feed 5mm -» bend -» retract.. and so on until the full arc length is done... for sure the 2nd solution would be easier on the machine (especially for edge material and tight radius) but until you improve the resolution (the "5mm" in my example") you'll end up with a faceted circle, but then better resolution means way more bending time...
The first method, while needing more power to feed the material would produce virtually smooth arcs and would be faster that method #2

Maybe a mix between the 2 methods is possible depending on the radius you want to bend..

I love where this is going ! I'm all pumped out :)


EDIT : just had a look at the 2.0... looks like we have a winner and it took care of most of the points I highlights a couple posts ago ! Great find Chris !
A bad day skiing is always better than a good one at work...
User avatar
chrismp
Posts: 1443
Joined: Fri Feb 13, 2009 9:00 am
Location: Vienna, Austria

Post by chrismp »

I think for the bends we're doing, it's fine if you just turn up the resolution and live with the longer bending times. Even for production runs it wouldn't be that bad if it took a couple minutes to bend one edge.

The DIWIRE 2 will be available commercially...unfortunately there's no release date yet.
User avatar
threeninethree
Posts: 172
Joined: Fri Feb 23, 2007 8:11 pm
Location: Massachusetts

Post by threeninethree »

I thought that sketch looked familiar! :D
~ Matt
User avatar
falls
Posts: 1458
Joined: Tue Oct 27, 2009 5:04 pm
Location: Wangaratta, Australia

Post by falls »

I think using the wire bender idea is better than the 3 dies because your in feed is all in a straight line. When you use the three wheel edge bender and plunge the bending wheel it creates the arc so the edge entering the bender changes angle from coming in on a straight line. With the wire bender it is only on exiting the bender that the angle of exit varies.
If their spring loaded feed rollers using bearings and a knurled wheel work for wire then something similar or using urethane /rubber rollers should work well for edge material.


I take your points about being able to bend parabolic shapes thus using the 5mm resolution method. All my shapes are radii based because I found bending splines too much work. Splines tend to look better in the end though.
Don't wait up, I'm off to kill Summer....
gozaimaas
Posts: 663
Joined: Fri Feb 10, 2012 4:17 am
Location: Nagano Japan

Post by gozaimaas »

Random idea.
Would a mig welder wire feed assembly be any good to feed the edge?
User avatar
MontuckyMadman
Posts: 2395
Joined: Fri Jun 20, 2008 9:41 pm

Post by MontuckyMadman »

that pensa labs thing looks like a grade a POS.
sammer wrote: I'm still a tang on top guy.
User avatar
Head Monkey
Posts: 310
Joined: Sun Jul 17, 2005 4:53 pm
Location: Carnation, WA
Contact:

Post by Head Monkey »

I'm willing to help out, but after I finish MonkeyCAM.

A few thoughts for now. I can't claim to have read the entire thread carefully, sorry if I'm repeating something.

- I don't think there's a big difference in the physics between a 3-roller bending design and the wire benders linked here. For a constant radius they're setting the pin in a fixed location and forcing the wire against it. That's likely more resistance then the 3-roller design, but the same overall principal: you need a robust feed mechanism with no slip, and you need to adjust something gradually as you feed to produce a smooth transition between radii as you go.

- I agree that an advantage of the wire bender design is the straight in-feed.

- For feeding: our edges have teeth evenly spaced. Feels like they need a gear to interface with, not smooth rollers (even rubberized). Imagine a wheel with short pins which match the spacing of the tines on the edges. Instead of placing wheels flat against the sides of the edge, place the wheel with the pins perpendicular and below the edge. Use simple bearings in various orientations to hold the edge in position as it passes over the feed gear.

- Arduinos are great little devices; cheap, easy to program and debug, and lots of pre-made, pluggable modules to interface with various other devices. It should actually be easy and cheap to put together a kit based on Adruino to control this.

- I think a stepper for the feed is ideal since you can accurately control the feed rate. Ditto for the motor to adjust the bending pin/roller.

- Software for this shouldn't be hard. If it were me I'd make the controller accept simple g-code for x & y. Computing the program given a DXF of the overall board shape should be very easy.

- Calibration is going to be interesting, and I think can be made pretty automatic. I suspect there's a zone where attempting to produce a very small bend will produce nothing. I can't remember the exact terms off the top of my head, and it's late, but metal is elastic and will return to its original shape if you don't push it past it's deformation point. We need to determine the minimum deflection needed to deform the metal without it springing back to straight, and there will need to be a step between "straight" and "smallest bend".

- There's a minimum radius to which we an bend edges depending on the tine spacing. I want the software to prevent me from attempting to exceed that.

I hope that helps.
Everything I know about snowboard building, almost: MonkeyWiki, a guide to snowboard construction
Free open source ski and snowboard CADCAM: MonkeyCAM, snoCAD-X
ben_mtl
Posts: 583
Joined: Mon Mar 31, 2008 1:47 pm
Location: Sherbrooke, Quebec
Contact:

Post by ben_mtl »

MontuckyMadman -» POS = ?? sorry again but English's not my first language so all those "acronyms" take me some time (and confusion) to understand... I would not want to miss a very useful contribution of yours. thanks.

HeadMonkey -» All very valid points.
I'm not convinced about the "gear shapes / pins" for the feed rollers for one reason : I bought edges from 3 different sources since I started building : no source had the same teeth pattern...

Also I'm with you for the minimum and maximum bending limits (for too large radii as well as too small).
If terms are the same in French and English, the deformation thing you're talking about is plastic deformation VS elastic deformation. What it means in reality is if we can come up with graphic showing "bent radius = f( bender wheel position)", the curve won't pass at the origin... Test will hopefully provide both the relation between bent radius and rollers configuration, as well as the offset of this curve.

As far as programming goes, g-code would be "easy" to generate, I agree. I'm gonna look into how Pensa uses their data to provide instructions to their machine too.

I'm also convinced thatwith the material we are using we will probably need a "mixed bending process", where the bending is done differently for longer and shorter radii.
For larger radii you can feed the edge through a relatively fixed roller configuration ("relatively" is because it can slightly vary for radius change in this "larger radius range").
However forcing the feed into a tight roller configuration will definitely cause slippage or at least put much pressure on the feed mechanism (if we can use smaller steppers.. the cheaper the better ! + the Arduino might not be able to handle any size of stepper...). So for small radii I think the move-bend-move-bend... method would be the way to go... but it comes with its disadvantages (faceted curves if resolution is poor... but if the resolution is smaller than the "extrusion head -to- bending pin" distance the parameters will change too.
A bad day skiing is always better than a good one at work...
User avatar
falls
Posts: 1458
Joined: Tue Oct 27, 2009 5:04 pm
Location: Wangaratta, Australia

Post by falls »

Piece of ... (POS)
Don't wait up, I'm off to kill Summer....
User avatar
chrismp
Posts: 1443
Joined: Fri Feb 13, 2009 9:00 am
Location: Vienna, Austria

Post by chrismp »

User avatar
chrismp
Posts: 1443
Joined: Fri Feb 13, 2009 9:00 am
Location: Vienna, Austria

Post by chrismp »

And another video of Signal Snowboards' edge bender:

http://snowboarding.transworld.net/1000 ... snowboard/

They're using a design with three roller much like twizz' vise bender. I don't see any special sort of feed mechanism like rubber rollers or something that grabs the edge teeth.

Here are a couple screen grabs from that video:

Image

Image
Richuk
Posts: 1146
Joined: Sun Jan 31, 2010 9:53 am
Location: The Duchy of Grand Fenwick

Post by Richuk »

Lets say for a moment that the arcs required for skis require the drive wheels to be designed to avoid slippage. Is there any reason why the drive wheels cannot be provided from above and below the edge, not at the side? This would allow the tines to act as a tyre tread and not a cog, maximising the number of different types of edge designs to be used. Cutting a 2.5mm square slot in the rubber wheel would maximise the interface between the drive wheels.
Last edited by Richuk on Tue Nov 12, 2013 2:46 am, edited 1 time in total.
User avatar
richie
Posts: 184
Joined: Wed Jun 23, 2010 7:25 pm
Location: New Zealand
Contact:

driving the wheels on edgebender

Post by richie »

My thought was drive the 2 end wheels at the same speed off the same motor via belts/cogs combination to add some traction and one of these end wheels is on the bending "push" motor to deflect the edge into a bend. That will drive without slippage I'm pretty sure. Close the loop and put a tacho on a free wheeling non driven middle wheel that follows the edge movement so its got absolute precision say 2000 pulses per revolution or whatever just accurate..... I think a combination of that and some clever programming (leave that to you smarter guys!) and its a winner eh. I'm keen for one, I can do the electronic side and mechanicals I just have not got the time nor the patience or skills to make a good job of the programming.....
MonkeyCAM and SnoCAD - https://github.com/mikemag
Ski binding mounting https://github.com/splitn2/DrillSki

Richard Harcourt | www.splitn2.com | Christchurch New Zealand
rich@splitn2.com | www.facebook.com/splitn2
User avatar
chrismp
Posts: 1443
Joined: Fri Feb 13, 2009 9:00 am
Location: Vienna, Austria

Post by chrismp »

Post Reply