What a weight-per-meter control system actually does

The short answer first: what it does is "spot the deviation and correct it automatically"

A weight-per-meter control system (some people call it a weight-per-meter controller, or an inline weight-per-meter control unit) exists to solve one very concrete problem: during extrusion, can the weight of cable per unit length be kept tracking the target value?

Manual sampling works like this. Every so often you cut off a length of cable and weigh it; if it is heavy you speed the haul-off up a little by hand, and if it is light you slow it down. The trouble with this method is not that it is inaccurate, but that it finds out too late — by the time you have weighed that piece, the machine has already run another few dozen or hundred meters, and the material for that length has been consumed.

A weight-per-meter control system compresses the cycle of "weigh, calculate, adjust" down to the order of seconds and lets it run automatically and continuously. The signal flow around the whole loop looks like this:

Step one: inline weighing to obtain the weight signal

The system inserts a weighing stage into the extrusion line and continuously collects the weight of the material passing through it. The key point in this step is that the sampling frequency has to be high enough: if a point is taken only once every ten-odd seconds, the fluctuations that occur in between are completely flattened out, and however accurate the later calculation is, it means nothing.

Haul-off speed is collected at the same time — because weight per meter is "weight divided by length", and if you know the weight but not the length you cannot work out weight per meter at all.

Step two: convert to weight per meter and compare it with the reference value

Dividing the weight by the corresponding length gives the weight per meter, normally expressed in kg/m. The system compares it with the setpoint and obtains the deviation.

Take an actual reading from the shop floor as an example: on one production line the measured weight per meter was 0.1413 kg/m, the reference wall thickness 1.650 mm, the wire diameter 17.50 mm, the material density 1.42 g/cm³, the haul-off speed 60.013 m/min and the extrusion output 508.70 kg/h. These figures have conversion relationships between them and were not simply filled in at random — weight per meter, wall thickness, conductor diameter and material density are locked together, and knowing any three of them determines the fourth.

A deviation has only two directions, but the causes can be completely different

If weight per meter is on the heavy side, it may be that the extrusion output is too high, or that the haul-off is too slow; if weight per meter is on the light side, the output may be too low, or the haul-off too fast. Looking at the weight-per-meter figure alone, there is no way to tell which cause it is. The system therefore has to look at haul-off speed and screw speed at the same time before it can judge which direction to adjust in.

Step three: coordinated adjustment, where the actions have to come in pairs

This is the step in the whole system that most tests your understanding of the process. There are two common ways to adjust:

• Adjust the screw speed: this changes the extrusion output per unit time. The effect is direct, but it affects the state of plastication, and moving the speed too hard can bring new fluctuations with it.
• Adjust the haul-off speed: this changes the amount of material used per unit length. The response is fast, but too large a change in draw ratio will also affect appearance and dimensions.

In practice systems usually make the two coordinated: adjust one by a small amount first, watch the effect, and top it up with the other if necessary. On the system's HMI the two values, screw speed and haul-off speed, are displayed side by side, so the operator can see at a glance which one is moving.

Step four: repeat checking in a loop, forming the closed loop

Once the adjustment has been issued, the system carries on sampling to see whether weight per meter has come back into the target band, and if it has not, it adjusts again. That is what the words "closed loop" mean: it does not assume that a single adjustment will land on target, but holds the deviation down through repeated small corrections.

One easily overlooked detail is the interval between repeat checks. If the interval is too short, the system over-reacts to an adjustment that has not yet shown itself and oscillates back and forth; if the interval is too long, it cannot keep up with the change. This parameter is related both to the length of the line and to the haul-off speed, and it has to be tuned to the actual operating conditions during commissioning.

How exactly it differs from manual sampling

Three things it solves along the way

Because the measurement is continuous, the system can also do a few things along the way that a manual method cannot:

• Material usage statistics: how much each reel of cable actually weighs, how many meters it holds and when the change took place are all recorded automatically, so there is no need to leaf back through a notebook to reconcile the figures.
• Wall thickness compensation: this is what "wall thickness compensation 0.020 mm" in the shop-floor reading refers to — because of factors such as material shrinkage and eccentricity, the actual wall thickness and the value converted from weight per meter show a fixed offset, and the system allows an operator to enter a compensation amount so that the displayed value sits closer to the measured one.
• Process trend logging: the historical curves of weight per meter, screw speed and haul-off speed are all kept. When something goes wrong, looking back through them is far more useful than arguing from memory.

What it cannot solve

Being clear about the boundaries is more useful than listing every function. A weight-per-meter control system governs "the total amount of material used per unit length"; it does not directly govern:

• Eccentricity: the total amount of material is correct, but it is distributed unevenly, so one side is thin. That has to be solved through the die head, the die and concentricity.
• Appearance defects: burnt material, porosity and roughness are related to temperature and screw shear, and they are not a weight-per-meter problem.
• Changes in the raw material itself: if the density of incoming material differs from batch to batch, the system needs the reference value to be set again — it has no way of knowing by itself what that value should be.

Frequently asked questions

Does weight-per-meter control always need a weighing device?

Anything called "weight per meter" must have weight measurement at its core. What differs is where the measurement is taken: weighing the loss of the hopper, weighing the finished product, or a combination of the two. We set out the specific differences in our comparison of loss-in-weight and gravimetric systems.

Can it be retrofitted to an old production line?

In most cases, yes; the amount of modification depends on the control interfaces the existing line has. What really needs to be confirmed is whether the extruder and the haul-off motor can accept an external speed setpoint — if they cannot, the question of the actuation side has to be solved first.

Once it is installed, is manual sampling still needed?

We suggest keeping it, but the frequency can be reduced substantially. Manual sampling then changes role from "the basis for everyday adjustment" to "checking whether the system has drifted", and doing it once a week or at a construction change is enough.