Hi..
I have winder application using controllogix and I want to calculate diameter of material. I used following formula: diameter = (Intg Line_Spd_MPM_Fdbk) * Winder_Gear_Ratio /(Intg Winder_Spd_Fdbk). I found that diameter is not increase (stay constant), how to solve the problem? How to get correct diameter update rate?
Thank you.
James

What are Intg Line_Spd_MPM_Fdbk and Intg Winder_Spd_Fdbk?

I can already tell you I don't think your calculation is going to work. As more material is wound on the roll and the diameter of the roll increases, the length of material in one wrap of the roll also increases

If you can accurately time the duration of one rotation of the rewound roll, then the circumference of the roll is equal to the line speed multiplied by the time for one rotation. Divide the circumference by PI to get diameter.

way don't use an analog sensor (ultrasonic,lasero or similar) and use the core as you refereance or "0" diametrer as you increase the diameter the distance betuine the core and the sensor decrease because the material rolls and this dist X 2 =diameter

[size=3][/size]

Well thats simple enough. I was getting ready to really complicate the solution

well I use this metod to control tencion in relation to diameter and is very simple and acuret
I well look for the codes and posted leater

If you're trying to control a centre-driven rewind start with the formula in the attached doc.

I recently had to do this in a PLC5-15 (without CPT). What I did was use speed feedback and calculate the distance travelled every 1/2 econd, accumulating that as the length.

Steve Bailey on another forum gave me this outstanding method:


I don't know if it's okay to post a link to it here, so I won't, but if you search plctalk you can read the whole thread.

OkiePC

Attached is the sample code of diameter calculation ( RSLogix 5K, ver 13.00). I try this code and it works to control winder motor speed, but I can't get actual diameter related to material wound.
The detail data of our application is 25-100 microns thickness,30-300 MPM speed, empty core 3-1/2" and full roll is 18". How to get diameter update rate (20 count/sec)? Any idea? or Any code for diameter calculation?
For diameter calculation, we have winder encoder feedback to clx and we can use this encoder to measure time for one revolution.
Thank you for reply and ideas.
James

Attached is diameter calculation that I got from rockwell site. Any comment or suggestion or any experience using this calculation?
James

The rockwell verion looks overcomplicated and designed to be a "one size fits all program". It has provisions for many things you may or may not need. Looks like it might depend on an analog diameter input?

Steve Bailey's method worked beautifully for me. I had the follwoing to work with:

Line Speed (Analog from drive scaled 0-400FPM)

Core Diameter (20")

Stock Thickness (0.013" to .055")

Web Thickness (Calculated using Steve's math at average of 0.021")

I added a 1/2 second timer that constantly reset itself.
(AB EXAMPLE: XIO T4:0/DN TON T4:0 .01 50 0)

Each time the timer finished I accumulated the Length of the stock that had passed in inches by dividing the ft./min. signal by 10. {since ft/min = 10*(in/0.5sec)}

L = L + FPM / 10

I also recalculated the diameters of the letoff roll, the rewind roll, and the web winding roll using the 1/2 sec timer done bit to trigger the math using L.

Once you have L for the formula. All you need is t (thickness) and ID (core diameter) to do the math below and solve for OD.

L * t = (PI/4) * (OD^2 - ID^2) or
OD=SQRT( ID^2 + L * t / (pi/4) ) 'SQRT indicates Square root function

In a PLC-5/15 it took about six rungs of several branches each to do all the math step by step for three rolls. It would have been much easier with the CPT instruction.

The result was near perfection. I installed the Length accumulator logic first, and then had pretty accurate length measurements to go with hand measurements of the finished rolls. That allowed me to reverse calculate to find the thickness of the web material that was important for the letoff roll calculation in my case since the web and product were rolled up together. The stuff is hard to measure with a gauge, but calculating it gives an average that works beautifully.

Try it!

You'll LIKE IT!

I did.

All credit goes to Steve Bailey for supplying me with the original formula.
And double check my transformations above before using this please.


PeC

I agree, use Steve Bailey's method. It is much easier

It looks like the Rockwell calculation is using the same method I suggested in Post #3.

((Line_Speed_FPM * 12) / Roll_Speed_RPM) / 3.14

Dividing by RPM (Revolutions per unit time) is the same as multiplying by the time per revolution.

That method works if you know or can measure the roll RPM or time per revolution.

The method that OkiePC cites works if you know the material thickness. It's as accurate as the thickness measurement. If you're winding a material with a high modulus of elasticity, it will work fine. If you're winding a material that will get thinner as tension increases, then you need to measure the thickness with the material under the same tension as it will be when you're winding it. Plus you need to wind it at constant tension.

You still haven't identified what Intg Line_Spd_MPM_Fdbk and Intg Winder_Spd_Fdbk are. What type of winder is this? If it's a two-drum winder, and Winder_Spd_Fdbk is the drum speed, then the ratio between Line_Speed_MPM_Fdbk and Winder_Spd_Fdbk is going to stay constant. You need to use a variable that's related to the speed of the wound roll or the speed of the chuck.

Thank you for replies.

The winder type is center winder, that use to film line application with thickness 25 - 100 microns, line speed is 200 MPM, using dancer roll for tension control with tapper tension (linear tapper with option negative or positive taper). Today I just try OkiePC suggestion and it works but I need to calibrate the length measurement and diameter calculation, but I don't know how it accurate if we run taper tension.

I try rockwell formula and diameter calculation doesn't works. I have spindle speed feedback to clx using encoder and line speed feedback also.

Clx have Integral (INTG) function to count line speed or spindle speed with amount of time. INTG Line_Spd_MPM_Fdbk use to determine material length on, the result of INTG is material length in amount of time. Same thing with spindle speed. I use this calculation (INTG (Line_Spd_Mpm_Fdbk) / INTG (Winder_Spd_Fdbk)) on alumunium slitting machine to calculate diameter ratio that use to scale spindle speed. This instruction also using in Reliance Automax processor for winder application.

OK, now I see a source of confusion.

I interpreted Intg Line_Spd_MPM_Fdbk as an integer variable. You interpret it as the integral of the variable Line_Spd_MPM_Fdbck. If Line_Spd_MPM_Fdbck is indeed the line speed in meters per minute, then integrating it over time gives length.

Similarly, integrating spindle RPM over time gives angular displacement and the ratio of length to angular displacement is diameter.

So, does Intg mean Integer or Integral?

All of the code and examples provided by Steve, Okie and TW are excellent. (I plan to use them in the future)

You say the following about the material.

Wow the material you are measuring can vary by 4X.
What type of accuracy are you looking for ?
If you were after high accuracy because of the varying product width i would lean towards using two laser sensors facing the core that go into a summation amplifer which outputs the result as an analogue signal to your PLC, but hey thats just another suggestion.

Basic questions:

• are you winding one product whose caliper varies between 25 & 100 microns, or are you winding several different products with calipers in that range?
• are you trying to wind a specific length of product and control the end diameter or are you only interested in the end diameter?
• is there a driven roll between the dancer controlling tension and the centre-driven rewind mandrel?

On the tissue rewinders that I am familiar with, we wind a specified length to a specified finished diameter - caliper does not enter into the calculation of the winding velocity profile. If product bulk (caliper) is slightly off resulting in incorrect diameter, we alter the profile to correct +/- 5mm. (finished diameters are 106 - 140 mm).

I should point out that my feet/min signal was coming from a high quality DC drive that controlled a fixed diameter set of rollers in a slitter with about 135 degrees of contact with the product. There was no slippage or changing diameter. Therefore, the signal was very accurate which is another key to using the simpler formula in addition to knowing the thickness.

The INTG mean Integral not integer.

Our line is Embossed Film Line that produce plastic with thickness range 25 - 100 microns, depend on which thickness we want to produce, say we want to produce 29 microns. We run different product thickness. We have thickness control using laser scaner to measure the material thickness and send the signal to extruder section. This signal will use to control material thickness. The material have constant thickness with +/-1% variation. This is known thickness that I want to used it to calculate winder diameter. I used Panelview screen to entry thickness, empty core diameter and full roll diameter. From OkiePC formula, I need to calculate material length first before calculate diameter. I have line speed feedback (in RPM) from 1500 PPR encoder that connected to 1756-HSC module. From line speed feedback then calculate line speed in MPM (base on roll diameter that material contact to it). I used formula Line_Speed_MPM = Line_Speed_RPM * Roll_Dia * PI. With this Line_Speed_MPM then I calculate material length. I used this length to calculate winder diameter base on formula

L * t = (PI/4) * (OD^2 - ID^2) or
OD=SQRT( ID^2 + L * t / (pi/4) ) 'SQRT indicates Square root function, L is material length, ID is core diameter, t is thickness, OD is Diameter.

The winder type is center winder with dancer roll and taper tension, how accurate of diameter calculation? I need suggestion. I want to use diameter to control winder speed (Winder_Speed_Ref = Line_Speed_Ref / Diameter).

[size="2"][/size]For Garry question,

- We are winding several different product in thickness range

- I want to calculate final diameter, the output product is from extruder line

- Before dancer roll we have pull roll and after dancer roll material will passing one roll then go to spindle winder. We have center winder type.

I hope this information will help. Thank you for any suggestion or ideas.

James

Do you have analog position feedback from the dancer?

If so, I have a system that works great:

Just use your "estimated" diameters to calculate the winder speed, and trim it with a trim factor. (say plus or minus 20% to start out). Set up a PID block to regulate this trim signal based on dancer position. ONLY RUN THE PID in auto when the line is running. FREEZE the PID action when the line speed is zero. Use the dancer position as the PV, (You don't want the PID making adjustments to the trim factor in any situation where the adjustment cannot affect the PV or it will saturate). Put the setpoint value where you want to the dancer to run, and then scale the PID output (CV) to a value from 0.8 to 1.2. Then you can just multply it by your estimated winder speed to get the plus or minus 20 percent trim. Once you watch it run, you may be able to reduce the trim scaling or increase it as necessary. I don't like to limit the PID output. I use the full range and do a separate calculation to get the trim factor. This maximizes resolution and keeps the PID generic so it can match others in your program for consistency.

WIth this method, you can set the PID gains very low, and the speed control will be very smooth. I have lines that run up to 300 feet per minute using this method. Theses machines stop and start frequently with 1/2 second accel and decel ramps. The loop (dancer) positions stays within about 2 inches of the setpoint at all times.

Good luck!

Why not add a sensor to detect each revolution of the roll?
then roll diameter = core diameter + material thickness.
Perform this calculation to take place every revolution.

Sure. That's a good enough way to estimate the required line speed, but unless you're working with servo positioners and absolutely perfect product, it won't be good enough without trimming it with the dancer position.

I must be missing something here. Isn't he just trying to keep track of the roll diameter?
Seems as if you may be referring to backtension control.

From the desription he already has line speed from an encoder and thickness from the lasers. Seems as if he then trying to go through all the math to get diameter. Just seemed simpler to add a sensor.

this is the metod I use for diameter
the sensor is ulrasonic whit analog out put and the type of sensor I personaly preferd are the ones that you can program whit a coputer and software
plase the sensor pointed to the shaft and program this distance as you min.distance or "0" volts output from the sensor
if you want a final diameter of 20" program you max. distance or "10" volts output from the sensor at 10 inches and not forgot the core diameter.
this way if the material rolls increase in you core the out put voltaje on the sensor increase in the same proportion

Yes, keeson, if it's possible to mount a sensor, that would be fine for tension control.

Somewhere along the line I thought we are also dealing with speed control of the winder. I may have gotten my threads mixed up, because I thought there was a complaint about speed control with a dancer input to a winder. I just reread tis thread, and didn't see that mentioned. That must have been my misunderstanding.

I used the calculated method in favor of ultrasonic sensors for several reasons:
1. On my slit stock rewinders it would have been impossible to mount a sensor.
2. There is no safe and reliable position to place one at the letoff. On my machine, the letoff is unloaded by dropping the core on the floor!. The full roll can weigh 4000 lbs, and is loaded by an overhead crane. Occasionally they drop one and shake the whole machine. Tough environment.
3. Ultrasonic sensors cost money and require maintenance ... software is virtually free and 0 maintenance.

Thank you for replies and good suggestion to solve my problem. I try to run the line and I get 100 MPM with good stability. Using your formula, the diameter calculation close to actual diameter, just difference about 7-8% because the line speed I get from another casting roll (the section is casting roll - cooling - pull roll #1 - treater - pull roll #2 - winder)but this is enough to scale the diameter from core to full roll. I used following formula:
Winder_Ref = (Dancer_Trim_Pct * Line_Spd_ref) / Build_Up
Build_Up = Diameter_Calc / Core_Dia (core = 94 mm, full roll = 850 mm)
The Dancer control are using PI block with Kp=0.7, Wld=0.03. The input is from dancer_pos_fdbk. I use Scurve (SCRV) to ramp dancer pos fdbk.
LIne speed accel and decel is 100 MPM / sec.
I run the line with 100 MPM and build up ration increase to arround 2.3 and I try to increase line speed to 120 MPM, the dancer start to oscilate from minimum to maximum position. I try to adjust the PI Kp and Wld but not help, I try to decrease the line speed to 80 MPM and the problem is still exist.
How to solve this problem?
Thank you for any ideas or solution.
James

James, you're going to have to "take me there" to be able to fine tune it further. I need to know more detail about the process.

How responsive is your drive to changes in the speed reference? Since you're regulating speed in the PLC, the lower you set the accel and decel rates in the drive, the better it will perform overall, BUT you have to consider all operations of the drive, you don't want to casue faults if a jog mode or some other control will be affected by changing it. With only 8% adjustment range, you should be able to set the P gain very low, and the I gain much lower. I don't use the d term at all for this application. Your symptoms appear to indicate that your PID gains exceed the responsiveness of the drive.

I'm not a PID expert, but to manually tune this application I'll set the P gain very low like 0.01 (PLC5) and the I gain to 0. I adjust the P gain very slowly, and watch the process stop and start at various winder diameters. I'll adjust the setpoint to create a disturbance and watch the trend chart as the PV moves toward the SP. I want minimal overshoot, and no more than 2 oscillations. I repeat this process, graudually increasing the P gain until overshoot is minimal. I then back it off a touch until it's a little overdamped.

Hope that generalization helps.

Thank you for the good forum.

I try to increase the speed until 170MPM, the dancer pos is still in mid-pos and I adjust the dancer trim to +/- 30%, P gain set to 0.7 and I gain to 0.03. The process is stable, until we the material reached the full roll diameter. During this process, we set the taper to 0%. We need to have + taper, that will deacrese as diameter increase. I use the formula Tension_Demand = Tension_Ref - (Taper_PCT * Diameter_ratio). I use linear taper and 9 point of diameter (From core to full roll). I generate this base on function generator of controllogix. Any idea or comment?

Thank you for the suggestion

James