Control consistent speed for feeding machine - aluminium foil

ckchew666

Member
Join Date
Aug 2003
Location
Malaysia
Posts
591
Hi,

We've come across a customer plant having problem with their aluminium feeding machine.

Referring to the attachment, the roll of aluminium foil is feed into the machine. In between there is an analog displacement/distance sensor to control the speed of the feeder motor. If it's a new roll, the motor will turn slower, if the roll almost finish the motor will turn faster so that the feed-in speed into the machine is constant. When the roll almost finish, the distance from the analog sensor to the aluminium foil will be greater. The analog signal from the sensor will feed into a PLC 5/30 and the output to control the speed of the motor is from a PLC analog output module (0 - 10V).

The problem is, now the motor feeder is running fast and slow (not consistent), because the aluminium foil will fluctuate during running, and when the sensors sense the fluctuation, it'll cause the motor to speed to fluctuate also. When this happen the aluminium will fluctuate even greater because they are like a PID loop!

We tried to put a timer in the PLC program, so that the PLC 5/30 will capture the sensor signal only every 10th second. And update the output every 10th second.

This does not help much, is there any better solution out there?

Hope to hear from you guys. Thanks.
 
Seems to me the simplest is let the infeed rolls on the machine unwind the feed stock roll.
ASSUMING the machine is constant speed thus infeed rate of the foil is a constant speed.
AND the aluminum will not tear

Dan Bentler
 
Well, every machine I've seen with a feed setup similar to that uses dancer rollers of some type to control the feed, and the machine itself pulls the film into the machine....

Mechanical modifications aside, instead of taking your readings every 10 seconds, maybe the way to go would be to take your readings, feed them into a table, and take the average of a certain number of readings.... Just by using the average instead of the instantaneous signal it may be enough to "clean" up the signal to get rid of some of the fluctuations.
You may instead want to go ahead and use a pid loop of some kind to control it, but I'm not sure how in depth you want to get with the programming.
 
Who worked this bucket last?

My first question is whether or not this setup WAS working in the past. If it was, then something has changed. Have they recently replaced the sensor, drive, motor, etc., etc.? Was it an exact replacement? Was the new sensor, drive, etc. adjusted properly? Are they running a different type of foil? Are they running faster/slower than before?

IOW, don't overlook the obvious.

🍻

-Eric
 
I came across this problem many time.

The option of average is not so bad but sometime it would not help because the fluctuate go too high or low then the resoult is like the timer you uesed.
I assume you used ultrasonice sensor.
When you use ultrasonic sensor you must to be aware that the sensor must to be 90 degree from the web.
also you need to find the right sensor.
I spend weeks to try to solve problem like that. the fluctuate was amazing and when I changed the sensor to Sick it solved the problems in a secound.
I geusse there is different between the sensor filter and sensitivity.
On a paper machine we put the sensor on the outside surface of the roll not like you put it on the loop, the loop is not stable.and might cause this problems too.
 
HI..

I did it on the coil slitting line I made two years ago...

I didnt use a sensor to measure the coil's diameter...I mounted a rotery encoder on the slitter which is inbetween the uncoiler and the recoiler...I measured the speed of the coil(linear speed)...and controled the VFD...the control system was very smooth and the loop was almost constant..check if you can mount a tacho meter on the foil while its moving..and keep the speed constantly at your setpoint speed.. do you have a VFD in you system?? if so then I think this is the best solution.


hope this help.
 
khalil Hi

I think you miss something.
The foil (web) roll changing his diameter when you unwind.
you need to increace the speed when the diameter go down.
Puting encoder to mesure the line speed is not helping at all.Line speed can not tell you what the speed for the motor feeder.
It work only when you drive the roll from out surface
( perimeter drive) not from the core.
Old machine use what we call "dancer" it a kind of arm who lead the roll diameter and have potentiometer on the shaft.
When the roll diameter go down it trun the pot up and the speed increase.
Ultrasonic sensor do it without all the mechanical involved.
 
Web Handling

I have had experience with these types of systems before. It was not aluminum foil, but a foil that was used in shielding comm wire. I have never seen a setup that lets the material run slack and look at the loop. Are you sure this was the way it was running before? The way I have seen it done is to look at PAYOFF roll diameter with an ultrasonic sensor. Yes, I did have to be picky about which sensor. It was substrate related. The Sick seemed to be a good choice. I have also done this with a dancer, and with a dancer - payoff sensor combination. Any of these will work, it just takes some PID setup. I have also used the above on severasl paper and plastics handling systems with superb results!
 
I have worked with unwinds on many different materials, the ultrasonic is fine for determining roll size but it depends on how you want to splice how you design the rest of the system. A splice "on the fly" just about has to incorporate some form of "excess" material that can be taken up while the rolls are being switched, a dancer works better for determining how much excess is available.

If you are just converting from a single big roll to a small roll then a dancer may not be necessary, just a sensor to determine roll size.

Its another one of them "IT DEPENDS".
 
have to agree with Ron. If there were more info available, then a solution would be easier. Adapting control for one instance, or for one part of the machine, without knowing the rest of the set-up could result in more erratic operation.
 
leitmotif: The system is designed in such a way that the feeder motor is used to turn the aluminium web (coil) & the machine job is to do stamping on the aluminium foil. It's an aluminium can making machine.

tom stalcup: We taught of using the average value to control the analog output (feeder motor), but the fluctuation is quite great. When I go online with the PLC-5 I noticed the difference of the analog input can be around 6000 to 8500. Difference of 2500 (I'm not sure what's the max & min value, the system is running 24/7)

Eric: From day 1 the system has this problem, just that the time those engineers does not bother to do anything until now :)

ArikBY: Glad to hear that you have solve this problem before. But can you explain a bit what's the difference between the Sick sensors tha you used compare to other brands? What do you mean outside surface of the roll? Seems like it because of the mounting of the current sensor?

khalil: We want to get a variable consistent speed, when the aluminium web is new, the motor feeder speed will be slower, when the web is almost finished, the motor feeder should run faster so that the machine can have constant in feed of web.

mjisaacs: Can we put another sensor to sense the drum diameter instead of a dancer? Less mechanical would it be better choice?

rsdoran: No, we are not converting from a big roll to some smaller rolls, not like the plastic slitting machine :)

stasis: Appreciate if you can let me know what other infomation that is useful, I have no experience in this kind of machine. Thanks for guiding.

My another question:

1) Is that useful/helpful if I move the current sensor (bet. the machine and the motor feeder) to a higher location, so that the sensitivity/resolution is greater and a flutuacting of the aluminium will not affect much of the motor feeder speed? Possible?
 
I am having a very stupid feeling that we all are confusing a constant speed control with a simple on-off loop control here. Aren't we?

If we are, then a 1-2 sec delay for the sensor, as well as hystheresis adjustement, would be a solution. Just to maintain enough loop for several press strokes, that's all.
 
Enclosed is a drawing to show you where to place the sensor. You can place it anywhere on the unwind end, this is only a suggextion based on the designs I've dealt with in the past. I think you can do away with the middle sensor all together, or use it for a safety (if film gets too tight). By adding an excoder to the payoff and running PID, you can have pretty good control over the payoff speed. You could also (thru an HMI) enter the material thickness, and by using this number and the sensor, make a good guesstimate of the amount of material on the payoff roll.

A better answer for your dilemma may be to only use the sensor detecting the roll diamater and use pneumatic brakes on the payoff, instead of a driven roll. This method works very well, as long as the payoff roll is not so large that you risk tearing the product. You can allow the operator to set an average needed brake pressure, and use the analog feedback to vary the analog output to an E/P or an I/P unit. In this manner, the operator sets the average brake desired by recipe, and the PLC varies the pressure according to the size of the roll.
 
could help

Take an analog sensor and scale it so you will know the diameter of your aluminum roll to be uncoiled. Once you have the diameter use this to find the circumference of your roll in feet. Now take the feet per minute the machine runs and divide it by the roll circumference. This gives yoiu DESIRED UNCOILER RPMS. Now take the DESIRED UNCOILER RPMS and work backwards through your gear ration this will give you DESIRED MOTOR RPMS. Now take the DESIRED MOTOR RPMS times motor poles divided by 120 and this will give you DESIRED MOTOR HZ. Now you will have to scale the MAX and MIN motor HZ your going to run. After you get this answer you will have a RATE and OFFSET. Now DESIRED MOTOR HZ times RATE plus OFFSET be your analog out signal. In the drive you will set your MIN HZ and MAX HZ to what ever you used to scale with. If your machine is variable speed you will have a little more Math. It might beat guessing at what speed to run? And i may not understand your application.
 
as stated by Ron (and what I was agreeing with), how you intend to splice the web has alot with how you need to control the speed.

In an aluminum printing plate plant, the web goes through a series of rollers designed to take up slack in the web when splicing, without affecting the speed of the machine.

For softer material, it is common to use z-scrays. The material piles into the scray & splices occur without regard to machine speed.

In one plant around here that does metalforming, their entire machine slows down during splicing.

How are you going to splice, and how will splicing affect the rest of the machine? If not continuious, does the machine need to stop to be resupplied? What about the section after the 'feed', does it have nip/pull rollers? I'm guessing not, but...

Like stated before, the more information we have, the better the solution can be...
 

Similar Topics

I have to start out by saying I am not a PLC programmer and I have basic programming skills but mainly use software as a troubleshooting tool. I...
Replies
0
Views
47
Hi, I have a 1500 that controls a station with diferents warehouses, but i also have a 1200 that controls one of those warehouses, i have been...
Replies
9
Views
225
I have to provide remote access and control to a touch screen. I was thinking about using Weintek and the Weincloud. Does anyone know if this is...
Replies
3
Views
138
Hello everyone, I'm reaching out for assistance regarding a project I'm currently working on. Specifically, it involves a hydraulic system with...
Replies
7
Views
325
Back
Top Bottom