Rewind Application question (center wind with air clutch)

Greetings!

I have an application request. I need to add a "Rewind Mode" to an existing machine. I wrote the attached sequence of operation and am interested in your opinion.

I found the following threads when I was searching for Center Wind here:

Winder Controls
http://www.plctalk.net/qanda/showthread.php?t=56639&highlight=center+wind

Web Handling math
http://www.plctalk.net/qanda/showthread.php?t=50619&highlight=winder

Oscillation during Centerwinding
http://www.plctalk.net/qanda/showthread.php?t=52439&highlight=winder

Winder application
http://www.plctalk.net/qanda/showthread.php?t=13660&highlight=center+wind

However none of them seemed to apply directly to what I have been asked to do. They did give me ideas and were informative. But I thought I'd start my own thread.

The hardware is ControlLogix 5555 with a Sercos card. There are ten Kinetix 6000 servo drives. I only need to control two of them. They are all small motors & drives. The machine is a prototype for a lab, it is used for developing new products.

Let me know if you see flaws in this control scheme or know a better way to do this application with the hardware that was provided. Meaning I realize if I want to maintain tension there are sensors for this, but the analog laser and the air clutch is what has been provided. Thanks for taking a look.

Oh, as part of my quest I also joined TAPPI and can search through their database. But when I try to add an article to my shopping cart it wants me to pay. The article says it is $0.00 to members (which it says I ARE one) and $30 to non-members. I have contacted TAPPI about this and will figure it out. Just thought I'd mention TAPPI here.

Paul you said a cutch is that a real air bladder clutch on a drive shaft or are you talking about a air bladder on a mandrel? I have never seen a clutch on a drive shaft for a winding system. I have seen them on a few metal decoilers, unwinders, real stands or what ever you want to call them. They are not close to anything I would ever want to see on my winder.
What material will you be winding?
I am on my cell phone so I can't look at you control scheme. I may have time friday to look at it.

Well, I'm not sure how to answer that. The air clutch is the customer's design, the control concept was specified by the customer. I asked many questions about this design, this is what I can tell you from what I was told.

He said they have other machines that work this way. The air clutch is mounted on the center wind roll, and engages more with more air pressure. They want the center wind motor to run a constant speed and the air pressure to the clutch to be varied with roll radius. More pressure to engage the clutch more as the roll radius increases. This because it takes more torque as the radius increases. So the idea is to slip more (less pressure to clutch) when roll radius is small, then decrease slip as radius increases (more pressure to air clutch = more torque).

My control concept in the sequence I wrote is based on what the customer is telling me how he wants it to operate. I spoke with JRW tonight and he said he'd never seen an air clutch used on a rewind operation. I have two servos, why not use the servo in speed control mode pulling against a torque limit that varies with roll radius? JRW called this "indirect tension control".

Another thing JRW was concerned with was the size of the rewind motor. My web width is about 24" max, may only be 16". My roll diameter starts at 4-5" and can be up to 60". My rewind servo motor is about 10" long by 4" wide, kinda small. I will examine motor/drive specs tomorrow and provide this information, but the concern is it is undersized.

It's a lab application and I need to provide a solution that allows a test to be performed, this provides a window of "good enough" that may pass muster.

Further comments are welcome.

Paul wonders how in the world do you read and answer forum posts from a cellphone.

Paul B said:

Paul wonders how in the world do you read and answer forum posts from a cellphone.

Click to expand...

Smartphone with a web browser.

I think the air clutch is on the mandrel. This is what we had on a rewind system. I think you would need to vary the speed on the web according to the diameter of the roll. The adjust the web tension with the bladder in the mandrel. I ASSuME the laser is to measure the roll diameter? Use the diameter signal to adjust the speed reference. Where will you get the tension refrencr from? Are you going to adjust tension off the drive command torque? I would make sure you have plenty of HP behind that motor. I watched a guy design a system to the letter offthe formulas and it worked only on SLLLLLLLLLLOW accel rates. The motor ran out of torque on raid accels. Changed it to a 5 from a 3 and the system didn't have any more issues.
Shoot me a PM with your contact info I may be able to help you out.
What PLC are you going to use?

Tested it today. The control scheme worked. Unfortunately the motor driving the rewind axis is woefully undersized. Once the radius hit 4 inches the motor was at 150% torque. They are looking into a gearbox.

When the rewind axis is started and the roll diameter is small you can grab the shaft and stop it. As I move a piece of paper closer to the laser the 0-10vdc signal to the proportional valve goes up and the roll becomes more difficult to hold. At around 8 vdc the shaft begins to lock (less slip) and it takes off out of my hands. Next time I visit the plant I will take a picture of it.

They relocated another servo that had a 3:1 gearbox on it. Once it replaced the one with no gearbox it worked. Torque was around 55% with half a roll. We are going to bring in an AB engineer on Monday to study the application and come up with a better solution, perhaps a larger motor and a tension sensor. What we have works but is clearly not optimum.

In case you are interested this is what the air clutch looks like.

Jay

Paul,
As stated by your coustomer you want the tension on the rewind roll to stay the same so you will want to keep the air pressure on the clutch the same so your not changing the tension in the roll and just controll the speed of the rewind motor. If your getting good data from the laser you can do the circumfrence caculation to run the speed of the rewind motor. Keep it about .25% faster than the speed of the first motor.

Forgive me if my comment is out of place, but would not a load cell roll or dancer confoguration be the best measure of the goal of a good roll. Let the torque applied vary to achieve the required roll wind tension.