Drive Rollers on Long Part

[musicmkr]

I am interested in any comments on a concept that I've been developing for a bridge type 3 axis mill/drill machine.

My company machines tons of aluminum extrusions for use in an architectural application. I've been developing a 3 axis NC machine to do all of the drilling and notching required for one side of an extrusion in one setup. These extrusions are often over 25' long, and in order to use a linear motion system rigid enough for some light milling I was looking at several thousand dollars of rails and bearings, along with several more thousand dollars machining acceptable surfaces to mount the linear system. Plus all of that money would just be a waste since we only need to have an accuracy of +/- 1/32" (0.0312").

I would like to build this machine using a stationary Bridge (Y-Axis) and Ram (Z-Axis) configuration. However instead of having a traveling bed for the X-Axis I intend to drive the aluminum extrusion directly with drive rollers linked to a 1KW motor. A friction roller system obviously has the possibility to slip so to make sure the part has moved the distance requested my design would read the motion of the part with a rotary encoder. (See Sketch) By attaching two friction rollers to the same encoder I limit the possibility of slippage, and allow the end of the part to travel past the tool. This encoder will close the loop for the servo driver.

In order to home the machine in the X direction I plan to use a photoelectric through beam sensor to detect the end of the extrusion.

I would appreciate any comments or suggestions that you may have, and I would be extremely interested in anyone's experience with a similar application.

[Al_The_Man]

I have in the past been involved in a couple of projects that are very similar, one just recently, a tube cut-off machine, and the other angle punch involved in the production of HydroElect-transmission Towers.
Both incorporated some of the ideas you have mentioned.
One thing I would comment on is if intend doing any kind of Mill/drilling as you suggest, Is to have some kind of clamping vise arangement when stopping to perform the operation. At this stage the servo action of the motor should be turned off but the encoder still monitoring the position. There are commands in some controllers that allow the turning back on of the servo and assuming the current position as valid, so that the motor does not jump to cover the assumed error.
The first project I mentioned, used a servo operated carriage to push the part through the operation stage, The start end of the material was detected by photo-cell detection.
The second was a air/hydraulic pusher that had an encoder on the pusher just for measure reference. Both operated on 25'~30' material.
Al.

[Geof]

An alternative way you could consider is what I have seen on automatic saws. One fixed vise to hold the material while it is being processed and a moving vise with a servo and position encoder for advancing the bar. The moving vise only needs to travel slightly more than your largest advance distance; it loosens, moves back, grips and the fixed vise loosens then the moving one advances and the fixed one closes.

You could do a variation on this and have your moving vise travel the distance of several holes/notches. It would move back, both vises grip, perform process, release fixed and advance one hole, gip fixed and perform process with moving still gripping, release fixed and advance one more hole, regrip on fixed, etc, etc. This could possibly give a faster time hole to hole.

You also mention light milling so maybe you would make it so the fixed vise could be unlocked to move along short guides. For milling you would grip in both vises and use the moving vise to drive the part and the, now unlocked, fixed vise. If your milling was carried out between the two vise you would have good rigidity.

[mxtras]

I think what you have concepted looks fine.

I think there are numerous ways to do this, so in order to stay focused - the only issue I see is the hold-down for drilling/milling as mentioned by Big Al.

Other than that, it looks do-able to me. Neat project!

Scott

[ty1295]

What about have a carrier for the bar that has a gear tooth or something on it to provide non slip movement.

carrier be reusable obviously.

[musicmkr]

Thanks for the input. I am still going back and forth over if this will even work for the application.

Al, I am interested in what kind of accuracy you were able to get with your previous applications of this concept. Did those machines have acceptable repeatability?

Geof, I like your idea and that is in fact the way that many of the other extrusion machines work. I was just looking for a way around buying 50' of linear rail so that's where this idea came from.

I have noticed one error in my concept sketch, and I figured it'd point it out so no one else makes the same mistake in future applications. The Servo Encoder friction wheels must be in line with the Servo drive friction wheels in order to prevent servo lock, or to keep the servo from shooting the part out of the machine after the part is no longer in contact with the encoder.

[Geof]

Originally Posted by musicmkr Geof, I like your idea and that is in fact the way that many of the other extrusion machines work. I was just looking for a way around buying 50' of linear rail so that's where this idea came from.

What I was trying to get at was you don't need more than a 2 or 3 feet of linear rail for the system I described. The maximum travel for the moving vise only needs to be greater than the distance between your holes or notches; it just shuttles back and forth gripping, moving and releasing.

[Al_The_Man]

Originally Posted by musicmkr Al, I am interested in what kind of accuracy you were able to get with your previous applications of this concept. Did those machines have acceptable repeatability?

The repeatability was excellent and in the case of the Tower angle punch, the accuracy was within .001"
On the question of the encoder being in line with the motor, if a system is used that supports master and slave encoder, an master encoder could be placed anywhere on the system, such as the infeed-outfeed point to detect when the material entered and left the process.
I have seen encoder used this way in flying shears, where the encoder is on a spring arm operated by a rubber wheel on the material.
Al.