Web Lines | Unwritten Rules, Now Written

Here are 3 web handling rules Tim Walker keeps forgetting to tell you.

There are a few things that I almost always forget to include in web handling training. Things so obvious (to me) that I forget that it is possible to do it otherwise. Here are a few:

1. Keep Tension

A basic rule of web handling: Tension the web and keep it tensioned. The transition from zero tension to tension will nearly always be prone to defects of lateral control, wrinkle, slip, and scratching.

How a web tracks through a machine changes between high tension, marginal tension, and low tension:

  • High tension webs run straighter, have more traction on rollers, providing more force for spreading devices to prevent wrinkles. The high tension web will run at narrower web width, especially obvious in low modulus and thin webs.
  • Moderately tensioned webs will run at a wider width, have less traction to pull out bagginess, turn idler rollers, and enable spreading devices.
  • Low tension webs will reveal their bagginess, sag from gravity, move easily from interaction with ambient air, and have no tractional force to enable spreading devices.

The transition from zero tension to tension will nearly always create lateral shifting and width changes, both leading to wrinkles. As a web becomes tensioned, it may have cross-roller traction variations. Uneven traction and tension will lead to additional lateral shifting and wrinkling.

For these reasons, avoiding transitions from tensioned to untensioned conditions is a goal of nearly any web process. When equipment is shut down, some web—maybe a sacrificial leader web—is left in the machine. This web may not be tensioned, but it is left undisturbed from the position it stopped, making for a smoother transition to restarting.

2. Avoid rapid speed changes

A second basic rule of web handling: Avoid rapid speed changes. Acceleration and deceleration introduce inertia to web handling. Under rapid speed changes, inertial torques can be many times bearing drag, leading to large swings in tension through a series of idler rollers. Tension will increase in the machine direction during acceleration and decrease during deceleration. High inertial torque can exceed the tractional forces available, leading to slip, loss of lateral or machine direction control, scratching. The change from rapid speed changes to steady speed can change the lateral position and width of the running web, leading to lateral shifting and wrinkles.

Unwinds and winders create a clear problem to the goal of constant speed, tensioned processes. Rolls eventually run out of material or reach their maximum diameter and change must occur. Some processes simply stop, create waste during speed changes and lost time during changeover. Efficient operations use at-speed or zero-speed roll transfer technology to avoid these losses.

Zero-speed splicing is simpler, using a festooning series of rollers to accumulate and dispense the web created during the differential speed of the roll change cycle versus the steady main process speed.

Accumulation length grows with speed and transfer time. Accumulation length equals speed times the sum deceleration time plus splice or roll start time (assuming the acceleration and deceleration times are equal). For example, a process running 200 fpm with a 5 second decal and 25 splice time will need to accumulate or dispense 100 feet of web (200 fpm x 0.5 minutes). This approaches the limit of conventional accumulators (e.g., 10 rollers moving 60 in., for a total of 100 ft). Some processes require long splicing time, such as welding steel web or sewing fabrics, so increasing speeds leads to ever-larger accumulators.

Accumulator size can be reduced by faster acceleration, but the bigger factor is splice or roll change time. Some zero-speed splicers need no more than 1 or 2 seconds to paste two webs and sever the expiring tail. If the same 200 fpm process keeps a 5 second decal, but reduces the splice time to 2 seconds, the accumulator size drops to 23.3 ft.

In most machines, acceleration rates are controlled by the programming. An operator can request 2,000 fpm, but a good control logic will have preset the rate to get to the new speed (such as 100 fpm/s). The programmed ramp is also usually not linear, but an 'S' shape, start out with slow accelerations, reach a peak accel rate in the middle of the accel time, and reducing the accel rate as the target speed is reached. This prevents "ringing" in the speed and tension control loops. Not all systems have this or notice it, but the most aggressive accel cycle will wisely include it.

E-stops are potentially the most rapid speed change. This is a secondary process requirement that I don't get involved with, but the controls team has to. All components need to be considered in E-stop plans. It may be possible for a brake or motor to decel rapidly, but if the roller shaft snaps, core chuck fails, or the internal friction in a wound roll is insufficient, aggressively stopping the motor or brake may not be the safest plan.

3. The web should be narrower than the rollers (and other lateral restrictions)

This isn't 100%, but it is pretty rare to run a process where the web hangs over the edge of the rollers. The unsupported web can scratch or wrinkle outside the roller contact, or worse—break. The corollary to this rule is "the web must be narrower than the distance between (1) the machine side frames or (2) openings it must travel through.

How much narrower? Start with web width, add width variations, add lateral shifting from general web handling, add lateral error of core positioning, add lateral error of unwinding roll side wall, (remove the last two after you go through a good unwind guiding system), and see what the total is. In many cases, 2-3 in. on either side of the web is enough (roller width is web width + 4-6 in.).

I've seen a system where 50-in. wide web was expected to run on 51-in. rollers over a distance of 50–60 ft. It didn't happen. I've seen a 300-ft accumulator where 6 in. of tracking error was unacceptable. We worked it out, but we needed to use a few tricks to do so.

One of my earliest project was to create a table of the process capabilities of three sister plants with polyester film makers (extrusion, orientation, winding, etc). I called the engineers on each of 10 process lines and asked them what their maximum width was. I got answers and published a short report to my boss. I was surprised that this innocent exercised cause a bit of a stink.

There was more to width capability than the simple geometry of the rollers. Die width, stretch ratios, trim widths, acceptable thickness variations, and winder core support all played into the true width capability answer. The stink about this table was that some felt it would be used by management to make decisions on where to run future products, and there might be some important adjustments to the sizing chart that would be overlooked. (And that was partially the intention of the table).

Don't skimp on roller width, and you'll be happier.

Web handling expert Tim Walker, president of TJWalker+Assoc., has 25 years of experience in web processes, education, development, and production problem solving. Contact him at 651-686-5400; This email address is being protected from spambots. You need JavaScript enabled to view it.; www.webhandling.com.


Subscribe to PFFC's EClips Newsletter