Non contact length measurement

[Electrically_Bound]

Couldn't you just integrate the speed of the cable to get its length?

 

[robertmee]

Couldn't you just integrate the speed of the cable to get its length?

And how do you measure speed? With a contact roller which is what he's trying to get away from. Can't rely on the speed of the unwind or winder as that changes due to diameter and he indicated the cable diameter itself may change.

However, EB is onto something. Don't they make non-contact laser sensors to measure speed realtime? I found this link which has tons of stuff for the steel industry. I have to imagine they have the means to measure steel cables.

http://www.steellinks.com/pages/Sensors__Gages_and_Measuring_Devices/Diameter/index.html

 

[Manglemender]

I've used the "integrate Line Speed" method a few times in web handling applications. I found that it's as accurate as any capstan/encoder method due to the vagaries of stretch/shrinkage and slip. I suppose that it really depends on how accurate you need to be?

Nick

 

[robertmee]

I've used the "integrate Line Speed" method a few times in web handling applications. I found that it's as accurate as any capstan/encoder method due to the vagaries of stretch/shrinkage and slip. I suppose that it really depends on how accurate you need to be?

Nick

It's surprising that's your experience as I've been doing web handling for many years and I have exactly the opposite experience....With web handling, you get all the slippage, as your source of speed integration is typically from large diameter rolls that are driven by a drive. So, you're integrating the speed of the drive, not the web. The web's natural inclination is to resist the movement of the roll and that's where slippage losses are introduced. Especially due to speed changes as you likely have sectional drives controlling speed/torque to tension.

In a capstan, the web/cable drives it, it is much smaller diameter (not as much inertia to overcome), the frictional losses are minimal and controlled and under most circumstances is a very accurate means of measuring speed/distance. However, they can be a headache to maintain as noted by the OP, and hence the reason for a non-contact version.

To be fair, I guess you qualified it by "how accurate you need to be". In the case of winding a spool of cable, I'd expect it would need to be very accurate.

 

[Manglemender]

It's surprising that's your experience as I've been doing web handling for many years and I have exactly the opposite experience....With web handling, you get all the slippage, as your source of speed integration is typically from large diameter rolls that are driven by a drive.

I too have been in web handling for many years and it is true that this method does not suit all applications - it doesn't suit: slow speeds, short run lenghts, slipping speed masters, "high" degrees of accuracy. The integration method works best where a machine runs for some time at a steady speed.

Other things to considder: Many materials will shrink after processing (3-4% is not uncommon), what is the effective diameter of your capstan and does that change with product thickness? will contact with the material damage it?

I've used this method: on film winders in Canada and the UK, on various coating lines (including bank note paper) where contact isn't desirable, on an entire range of vacuum metalising machines distributed around the whole world.

My most recent implementation of this method was earlier this year on a film winder making medical grade plastic mesh. The machine is equiped with two counter methods; count pulses and integrate length which they can switch between. Accuracy of calibration is key to both systems (line speed or wheel diameter) but once setup they are as accurate as each other. Because of the strecthy nature of the product, it's difficult to say whether either method was truly accurate.

Nick

 

[jewely]

Did you ever find an affordable solution? I have a need for something similar with the same budgetary constraints?

 

[TConnolly]

Allscot, I realize its now several months after the fact, but yesterday I saw a peperl-fuchs sensor that uses the same basic technology an optical mouse uses. It senses material moving past it and outputs 1 pulse per millimeter.

 

[OkiePC]

Years ago we used a product from these guys:
http://www.betalasermike.com/

It was expensive, but very accurate for our web length measurement.

 

[rdrast]

Allscot, I realize its now several months after the fact, but yesterday I saw a peperl-fuchs sensor that uses the same basic technology an optical mouse uses. It senses material moving past it and outputs 1 pulse per millimeter.

Alaric, I'd be interested in any information you can find on that... I've just been searching, but haven't come up with any device by them that matches your description.

And - Allscot, sorry, I missed your question back from a long while ago... Anyway -

Point and Shoot - Nope. On cable, they require pretty precise positioning, and we had problems with inaccuracies anytime the cable moved more then a couple millimeters. There is absolutely no way we can keep the laser focused on the cable with the consistency required. And our cable is probably significantly larger then yours. We do things up to 1250 MCM cable here, so we don't even run at near the speeds you do when processing #6.

Calibration - (of ANY counter, mechanical or otherwise) Two methods. First, is we have a jacketed stainless steel (wire-rope) cable that has been certified and marked in 100' increments, long enough (more than 1500 feet) to go through most of our equipment, or at least the counter section. That is the preferred 'Standard' calibration, but time consuming. The second method is to run that length through a dedicated steel-on-steel wheel laboratory counter a couple times a year, then when we need to calibrate or check calibration, we move the lab counter out to the line, put it inline, and synchronize the resets on that and the line counter.

 

[OkiePC]

. . . yesterday I saw a peperl-fuchs sensor that uses the same basic technology an optical mouse uses. It senses material moving past it and outputs 1 pulse per millimeter.

It's about time...

 

[JesperMP]

Okie, what does that betalasermike device cost, approximately ?
What was the resolution ?

Alaric, I am also interested in the P&F device.

 

[OkiePC]

It has been so long ago...I want to say it was around $6500. It was used to monitor a wide strip of material used to make radial tires steel belts. Not exactly the same as measuring the length of a wire.

We also used a laser velocimiter from Canon in conjuntion with a PC based controller that would measure the length of splices by detecting the change in thickness (laser doppler). I am not sure that the Canon unit would be able to work on a strand of wire, unless it could be contained exactly in the laser target focal area. The good thing about the Canon velocimiter was its quadrature encoder output. The bad thing was that it only tolerated a very short movement in reverse before stopping its output. In our application, it was common to stop the machine and have an inch or two of backward travel. I could never get any help from Canon to resolve that problem, so we just made mechanical workarounds.

And, we used some thickness gauging lasers from LMI that were really good and under $5000, but I don't know if they had anything for length/velocity.

We demo'ed a system for tire tread profiling that was based on industrial cameras. It would do all sorts of quality control including length but I can't remember the name of the company. They were mainly focussed on realtime monitoring of extrusion profiles and their stuff was really slick. What they did was shine a line projecting laser at 90 degrees to the product and mount a camera at about 45 degrees to monitor the resulting line on the product. Then with software they could do all sorts of magic to tell us the exact profile of the whole width of the strips up to ~12 samples per second. That was much better than the other system we had which could only scan a single point and had to traverse the width constantly.

I could not find anything googling for what Alaric mentioned either. It must be a brand new development, perhaps in a trade rag?

Paul

EDIT: Here is a link to another vendor doing the same thing.
http://www.bytewise.com/products/default.html

 

[JesperMP]

6500 USD. Yikes !
Probably worth it for some applications, but is too much for mine.
An industrial grade "laser mouse" sounds as it could be made for a lot less.

I need to measure the transported length of a mold string on a molding machine. I am worried about the 1 mm per impulse mentioned for the P&F device. However, if the resolution is accurate over a long distance it is OK. I would like to get an accuracy of +/- 10 mm over a length of up to 50 meters.
For the Laserspeed 4000 they quote +/- 0.05% accuracy. That would be +/- 25 mm over a length of 50 meters. It would be just acceptable for me.

 

[TConnolly]

I could not find anything googling for what Alaric mentioned either. It must be a brand new development, perhaps in a trade rag?

It was in their product demo trailer that they drag all over the country. We're working with them on an RFID project and I guess they thought they saw other sales opportunities so they hauled it over to show us what they had. I don't see it on their website, that doesn't mean its not there though because their web site is a PITA to find things on.

 

[jewely]

I'm pretty much in the same boat as JesperMP. I recently got a ballpark figure on the lasermike system that was $8K. Even though that is the cheapest I've found, it's still out of my price range. Surely these "industrial grade optical mouse" devices would be less expensive. Please let me know if you come up with anything. THansk a lot.