Web Guiding Systems
Table of Contents
Webs do not always track down through a machine in a consistent location. On a roll-to-roll operation, for example, any offsets in the unwinding roll will cause the web to get started through a machine in a variable CD (cross direction) position. However, even if a web starts out in the position, it may take slightly different paths through the machine depending on several factors.
First, a variably baggy or cambered web will cause the web to track toward the variably tighter side. Second, variabilities in roller traction or drag can cause a steering of the web. Third, changes in tension will cause changes in how straight the web travels downstream. At zero tension, edge position control is essentially lost. However, there are other edge movement factors which include nip roller draw variations, aerodynamics and so on.
Guides are used to bring the edge or center of a web to a specific CD position. The guide location may be at an unwind to get the web started down through a machine in a consistent position, at an intermediate location, or at the winder to improve roll edge quality. The accuracy demands for the guide may vary from merely keeping the web on the rollers, to minimizing trim loss, to registering several print colors to within a couple of mils. Guides may be either active or passive.
Active guides systems are composed of a sensor, an actuator, and a controller. The sensor can be any detector which can reliably pick up the edges of a web. The most common are pneumatic (nonporous webs), photoelectric (opaque webs), or paddies (thick webs). The web must be flat (free of curl) and stable (free of flutter) through the edge sensor. For this and other reasons, the sensor is often placed near a roller. If two sensors are used, the web could be guided to the front edge, back edge, or center.
The output of the sensor goes to a controller which outputs a movement command to the actuator. If the gain of the controller is too low, the response of the guide will be sluggish and slow to correct. If the gain of the controller is too high, the guide will be hot but overshoot, or may even be unstable. The actuator which moves the guide mechanism may be a stepper motor and ballscrew for smaller assemblies or a hydraulic cylinder for larger assembli`es. The actuator and framework must be stiff for responsive operation.
The steering guide, as shown in Figure 2, is a roller that has ends which are mounted on a pair of angled linear bearings. As the actuator pushes the roller sideways, the raceways cause the guide roller to pivot about an imaginary point called the instant center. The principles of operation are via a controlled misalignment in the plane of the incoming web and the Normal Entry law. In other words, if the roller is angled, the web must bend sideways in order to enter the roller’s axis at a right angle. Since the Normal Entry Law is valid only for traction, any slippage of the web and roller will compromise or destroy guiding.
Steering guides are common whenever there is a long entry span, such as following an air flotation oven. Indeed, this type of guide requires the entry span length to be at least 3-5 web widths long to minimize in-plane bending stresses at the exit of the upstream roller. Similarly, the exit span should be at least 1/2 web width long to minimize stresses on the edges of the web due to twisting about it’s centerline. Finally, the length of the pre-entry span must be shorter than the entry span. A common problem to avoid is to place a spreader or other roller between the oven and the steering guide. This shortens the entering span and makes preentering span longer than the entering span.
There are two basic types of guide movement. The easiest is to simply pivot the roller from one end. While this end pivoted is adequate for some endless forming belts, it is not optimum. A better response can be achieved by locating the instant center about 1/2-2/3 of the entry span upstream. This arrangement causes the roller to also move sideways, thus carrying the web with it. The location of the instant center is determined by adjusting the angles of the linear bearings upon which the guide roller ends sit. The location is where the perpendiculars of the two slides intersect.
When a sufficiently long entry span is not available, a displacement guide may be a better choice. The displacement guide, as seen in Figure 3, is a system of four rollers. The first and last are fixed, while the center two rollers pivot on a common frame about the center of the upstream roller. The web sensor must be located very close to the exit of the third roller. While this is the most common depiction found in advertisements, the system can be arranged in at least 16 distinct ways, each of which have four orientations. These numerous equally effective possibilities allow considerable machine design flexibility.
The only real design constraints are that the spans entering and leaving the pivoting table must be at least 1/2 web widths long, but the three legs can be of different length. Note that the length between the pivoting rollers determines the effective sidelay range for a given twist of the web.
There is an important difference between the behavior of the displacement guide and the steering guide. While the steering guide will move the edge of the web upstream and consequently also downstream of the guide, the displacement guide will move only the downstream web path. Thus, the upstream process must be tolerant of uncontrolled web CD position.
Sometimes it is important on roll-to-roll web lines to start the web down an appropriate path despite poor edge quality of the unwinding roll or variations in web width or roll positioning. In these cases, an unwind guide would be used as seen in Figure 4. The unwind guide would be used as seen in Figure 4. The unwind guide consists of an unwind which can be moved sideways on linear bearings, a following roller which (desirably) moves with assembly, and, of course, sensor, controller, and actuator. While the actuator could be any push-pull device which is strong and fast, typically hydraulic cylinders are used.
Sometimes, straight edged rewound rolls are desired despite edge problems with the unwind rolls or the web path between. In this case, often a salvage operation, the rewind may be guided. As shown in Figure 5, winder guiding system closely resembles unwind guiding. In fact, the major difference is that the edge sensor is fixed to ground on the unwind guide and fixed to the winder frame on the winder guide. If the sensor cannot be mounted off the winder, then it must be slaved to follow the movements of the winder. Rewind guides are somewhat less common than unwind guides. In part, this is due to the greater complexity of the winder which does not lend itself to sidelay as easily as an unwind.
Guide Accuracy and Response
We will want our guide to move an edge towards its setpoint with great accuracy and repeatability. Indeed, some are asking for edge position tolerances of less than 5 mils. Also, in general, we would want the guide to be as fast and responsive as possible. (An exception is to avoid following a ragged or scalloped edge.) The combination of accuracy and response will improve product appearance and reduce trim waste.
To achieve accuracy, repeatability and response, it is important to follow guidelines as given here or by the builder. Bending these rules will usually result in sluggish response or hunting of the guide. Breaking these rules can degrade function to the point of inoperability. However, there are other considerations when designing guides to tight tolerance applications. The sum of edge scallop, sensor resolution and actuator hysterisis must be no greater than the total allowable edge variation. However, we must also avoid structural flexibility and play.
Flexibility of the actuator or framework will require detuning the response of the guide in order to avoid oscillation and other instabilities. The element with the greatest flexibility will be the weak and, thus, determining link in the system. In many machines, it is the hydraulic cylinder that is softest. If so, one might consider an electric ballscrew as an alternative if it has sufficient speed and load rating. However, also keep in mind that the mounting of the actuator can needlessly soften the dynamic system if it is not stout and rigid. While decreasing the guides’ mass can also improve response, it is difficult to achieve significant gains there. It should not be surprising, then, that winder and unwind guides might be more sluggish than other styles because of their greater mass than steering or displacement guides.
One of the most troublesome of the edge position excursions is known variously as the acceleration or ramp offset. Here, the web moves sideways when machine speed is increased or decreased. Usually the rate of movement is most severe at the top and bottom of the speed change rather than on the speed ramp itself. Unfortunately, the term “acceleration offset” belies the true nature of the cause of the edge movement. Indeed, the web does not even know how fast it is moving.
The cause of most acceleration offsets is illustrated by the various web tracks through a machine given in Figure 6. Here we show how the web moves through a machine where one of the rollers is misaligned. In the case of absolute traction, the web will conform to the Normal Entry law on all rollers, including the one that is misaligned. Note how the web moves over as the result of this misalignment. In the case of pure flotation, however, the web is not steered by the ‘roller’ and, thus, passes straight through the machine. The case of sliding is intermediate in that there is a small offsetting of the web.
Thus, every roller or element that touches the web also steers the web. However, if the roller is stationary and the state of traction constant, the path of the web will remain constant. That is not to imply straight. Obviously the web will snake through the machine in conformance to web handling laws. What we are saying is that the path will remain, for the most part, consistent.
However, if the web changes from full to partial tracking the path of the web will change slightly in response. This change in traction will be subtle and not easily picked up by conventional observations and measurements. Nonetheless, it will cause the web to move in response.
While there are many ways the state of traction can change, the most common is due to a tension change on a lightly wrapped roller. This means that if our drive allows tension variations, the web might shift slightly on some of the rollers. Furthermore, the condition which is most difficult to hold tensions is during a speed change.
Thus, as we not see, the acceleration offset is not due to the speed or speed change itself, but rather due to tension variations that can and will accompany the speed change. Our first efforts should then be to tune the drive so that tension is held well at sensors (load cells) as well as elsewhere where there are no sensors. Sometimes tension is held well only at the drive points or sensors, but not elsewhere because the web might be pulling against excessive roller inertia or drag.
However, we can also expect to reduce the severity of the offset if we reduce roller misalignment or other geometrical problems. The surest way to do this is through optical alignment of every roller in the line because even the lowly idler is as capable of shifting the web as any of major process rollers. However, sometimes it is not the roller, but rather an airfloat oven, that is steering the web. Here, the first step is to balance the ovens by clearances and airflow measurements. However, the web will be the ultimate judge of balance as its position should not vary objectionably during tension changes or even dryer engagement/disengagement (for those that can open).
Finally, you want to keep your web as uniform, flat, and baggy free as possible. This is because the web with profile troubles will merely exaggerate the difficulties discussed here. Only when material and machine are made true will edges run consistent.
Written for Faustel, Inc. by David R. Roisum, PhD