What is open-loop torque control?

Closed-loop tension control requires tension feedback by either a transducer or dancer roller. If you control tension without a transducer roller (a.k.a. load cell or tension roller) or a dancer roller, then you are using open-loop tension control. Torque control and draw control, without tension feedback, are both open-loop tension control methods.

If your control brake torque applied to the shaft of an unwinding roll, you can increase or decrease the web tension by adjusting the brake torque. Similarly, using a torque-limiting clutch control on a winding shaft will create tension, but is not closed-loop control without tension feedback.

Both unwinding or winding torque control systems need to change torque proportional to roll diameter to maintain constant tension. An advanced torque control system will include a diameter sensor to automatically adjust the torque proportional to roll diameter, coming close to constant tension control, but without tension feedback, they are still open-loop tension control.

The greatest weakness of an open-loop torque system is inertial torque. Web tension is proportional to the resisting torque divided by the roll or roller radius. For steady speeds, a diameter-based torque compensation system will do a reasonable job in creating uniform tension; however, during acceleration and decelerations, the roll or roller mass will add or subtract from the applied clutch or brake torque, creating a significant uncompensated tension variation.

Why use open-loop torque control?

  1. Unwinding – Many tension insensitive processes use a constant (or manually adjusted) torque brake to control the tension in the unwinding web. Many unwinding processes use a diameter sensor to maintain a constant ratio of torque to roll diameter, creating a consistent brake-induced contribution to web tension. (Note:  This approach often ignores the contribution of friction and inertia to web tension.)
  2. Rewinding – Constant torque center winding creates good wound roll structure in many cases, especially if the roll buildup ratio is less than 3 or 4:1. Constant torque center winding has a natural tapering with the tension decreasing inversely with roll diameter. This natural taper creates the desirable roll structure of firm layers near the core and more loosely wound outer layers. Constant torque center winding has less trouble with roll cinching and cinching-related telescoping since there is not increasing need for torque transmission.
  3. Differential rewinding – Differential winds multiple rolls on a common shaft, but unlike locked-bar winding, allows each roll to turn independently. To maintain the desired tension on each independently rotating roll, differential winding systems are designed to control the torque transmitted to each core.
  4. Tendency driven roller – A tendency driven roller is a hybrid device, partially web-driven and partially motor-driven. Tendency driven roller are most commonly used where the traction created by wrap angle, tension, and web-roller traction coefficient is potentially insufficient to overcome a idler roller’s bearing and inertial drag. In most tendency driven rollers, the bearings between the shaft and roller shell are used as a low torque clutch. The shaft is driven at the desired roller rpms and the bearing drag transmits enough torque to bring the roller shell up to speed. With almost no differential between the tendency roller bearing’s inner and outer races, the torque required by the web to make slight changes in speed from the shaft’s baseline rpm is minimal and accomplished with the low traction available.
  5. Local tension change – If you want to change the tension from low to high or high to low on one roller, simply connect a clutch or brake to the roller and ensure there is sufficient web-roller traction to prevent slippage.
  6. Combination winding – I think of combination winding as surface winding with a center torque assist. Since there is no space to measure tension between the nipping surface drive roller and the winding roll and the roll diameter and circumference are constantly changing, open-loop torque control is the obvious way to assist the surface winder in creating roll tightness.
  7. Nipping rollers – In most nipped roller systems, one roller is driven (often steel roller) and one roller is idling, driven by through the web or through contact between the rollers outside the web’s width. In some cases, a thicker web prevents contact outside the web width and the torque transmission through the web causes damage or defects, the logical solution is to add a clutch and motor to the nipping roller to provide the turning torque to overcome drag and rolling resistance independent of the web.

Why use motors vs. brakes or clutches?

(Caveat: As usual, when I’m writing about controls, it’s best to get some confirmation about what I say.  -tjw)

When I think about motors vs. brakes and clutches, I can think of three differences.

  1. Motors have a faster response time.
  2. Motors (the right motors) operate in all four quadrants of control, in either forward and reverse and in either power or regeneration (braking) mode. This is especially helpful for unwinding large, high inertia rolls. Braked unwinding systems rarely apply and compensate for inertial torque. Without inertia compensation, an unwind’s tension will increase during acceleration and decrease, sometimes causing a slack web, during deceleration. A motor-driven unwind can add more torque to help accelerate a roll and act as a brake during decel.
  3. Motors can run in speed mode. Every web line needs at least one pacer.

Early in my web handling days, I tried to set up a torque-to-torque web line. I was working on a torque-controlled winder (at the end of a film line) and wanted to run some winding experiments off a portable braked unwind. So we put in the unwind, threaded the web to the winder, and turned things on. Nothing happened. The brake torque’s tensioning was too high relative to the winder torque tensioning and the unwind won the ‘tug of war.’ If we turned the brake setting down, the winder would win for a while, then stall as the unwind tension went up (due to unwind tension = torque / decreasing roll diameter). If we turned the unwind brake too low or off, the winder would accelerate. This system had no pacer, so it was a battle of torques and F=ma to determine the line speed.

How much torque is needed?

When you hook up a motor to a roller or winder, there are several demands for torque. Add these all up to get the motor torque requirement.

  1. Web torque = Tension x radius (for adhesive webs, the tension must be greater than the adhesion peel force)
  2. Inertial torque = I (moment of inertia) x accelerations (or deceleration)
  3. External load torque = for nipping roller, the nip load will increase the drag in the system, sometimes estimated as tensioning equal to 10% of the nip force.
  4. System torque losses = frictional losses and efficiency losses

Why is diameter feedback used with torque control?

If you want constant tension as a roll diameter increases or decreases, you have to adjust torque. Tension from torque = torque / radius. If the radius is changing, torque needs to change proportionally to keep the torque contribution to tension constant.

By combining torque control with roll diameter, an open-loop torque control system can do a reasonable job-creating constant tension at a constant speed. Certainly, this approach is an upgrade from having an operator manually change torque with a diameter (or forgetting to). Any of the following methods are used for open-loop, diameter-compensating torque control.

How is diameter measured?

There are three common and two less common ways to measure or calculate diameter at unwinding or winding.

  1. Follower arm or nip roller – A pivot arm or sliding carriage is loaded against the unwinding or winding roll. An encoder or LVDT detects the arm or carriage motions as it moves with changes in roll diameter.
  2. Ultrasonic distance measurement – An ultrasonic distance sensor is mounted pointing at the core or roll’s outer diameter, detecting the distance to the roll and converting the signal to roll diameter.
  3. Tachometer ratio calculation – An encoder or tachometer monitors the revolutions per minute (rpms) of the winding or unwinding roll. A second encoder or tachometer monitors the rpms of a roller of known diameter in no-slip contact with the moving web (often the nearest driven roller). The roll diameter is calculated as D(roll) = D(roller) x n(roller) / n(roll) where n is rotational speeds in rpms. The disadvantage of this method is that it requires a moving web and cannot provide any roll diameter information needed to set the starting rpm of a unwinding motor.
  4. Web thickness and revolution calculation – For a given web thickness, diameter is continuously calculated as the core’s outer diameter of the core plus one web thickness for each roll revolution. This type of calculation will be off by some amount due to thickness variation, web compression, and entrained air.
  5. Dancer roller jog calculation – For driven unwinds and winders with dancer roller feedback, the motor-dancer system can be used to sense initial diameter. In the stall mode (line under tension, but at zero line speed), jog the motor 45 or 90 degrees and monitor the dancer motion. The dancer (with 180 degree wrap) will move 1/2 the arc length dispensed or accumulated by the small roll rotation and allow a calculation of roll diameter.

All of these methods are useful for any winding or unwinding system, even if motor driven. Initial roll diameter is used to: 1) set a motor’s initial rpm and 2) calculate inertial torque needs of acceleration.