Email: info@bemachine.cn Tel: +86 135 8442 7912

News

Multi-axis coordinated control method of extrusion blow molding machine

Multi-Axis Coordinated Control Methods for Extrusion Blow Molding Machines

Running an extrusion blow molding machine is a lot like conducting an orchestra. The extruder screw, the parison take-off rollers, the clamp actuators, the blow air valves, the mold cooling fans, and the conveyor or winder all need to move in lockstep. If one section drifts, the parison sags, bursts, or ends up with uneven walls. Coordinating all these axes is not just about making them spin at the same RPM — it is about managing timing, force, and position relationships simultaneously under constantly changing load conditions.

What Multi-Axis Coordination Actually Means in This Context

Most people hear "multi-axis" and think of CNC machines cutting metal along X, Y, and Z. In extrusion blow molding, the axes are different. You typically have the extruder screw (volumetric output), one or two traction pin belts or rollers (linear parison pull), the mold clamping cylinder (closing and opening), sometimes a parison programming device for preform shaping, and a finisher or winder. Each axis has its own motor, drive, and feedback sensor.

Coordination means these axes do not operate in isolation. The traction speed must track the extruder output in real time. The mold must close at the exact moment the parison reaches the correct length. The air pin must inflate the parison before the material cools too much. All of this happens within a cycle that may last 8 to 45 seconds depending on part size.

The core challenge is that these axes have wildly different dynamics. The extruder screw responds slowly — mechanical inertia plus melt compressibility means it takes seconds to ramp. The traction servo responds in milliseconds. The clamp cylinder bangs open and shut in under half a second. Trying to synchronize a sluggish axis with a snappy one using a single controller loop is where most systems fall apart.

Master-Slave Architecture and Its Limitations

How the Classic Approach Works

The oldest and still most widely deployed method is master-slave control. One axis — almost always the extruder — is designated as the master. Every other axis receives its speed or position command as a scaled multiple of the master signal.

For example, if the extruder delivers 12 kg/h and the target parison weight is 40 grams, the controller calculates that a new parison segment should be ready every 12 seconds. The traction motor then runs at whatever speed pulls exactly 40 grams of material in that window. The clamp receives a delayed command: open now, close in 11.2 seconds. Simple on paper.

This architecture is popular because it requires minimal communication between drives. Each slave just reads an analog signal or a simple fieldbus word from the master. Commissioning takes an afternoon. Machine builders love it because it keeps wiring costs down and the PLC program stays small.

Where It Breaks Down

The problem shows up during transients. Imagine the mold opens, the operator reaches in to adjust something, and the clamp cylinder holds the mold open 3 seconds longer than planned. In a pure master-slave setup, the extruder keeps pushing melt. The parison droops. The traction belt keeps pulling, stretching the parison thin. By the time the mold finally closes, you have a long, wispy parison that will blow into a thin-walled mess or rupture entirely.

Master-slave also struggles when you have multiple slaves competing for the same resource. Two traction belts on a large parison, each trying to follow the master independently, can end up fighting each other if one encounters slightly more friction. The result is parison twist or uneven wall distribution around the circumference.

The fundamental issue is that master-slave is open-loop from the slave's perspective. The traction motor does not know what the extruder is actually doing — it only knows what it was told. There is no feedback path confirming that the material actually moved as expected.

Cross-Coupling and Deviation Coupling Strategies

The Core Idea Behind Cross-Coupling

Cross-coupling control, originally developed for dual-axis gantry systems in the 1980s, takes a different philosophy. Instead of one axis telling all the others what to do, every axis talks to its neighbor. Each controller monitors two things: how well it is tracking its own setpoint, and how far it has drifted from the adjacent axis.

In an extrusion blow molding context, the extruder and the traction motor form a pair. The extruder controller sees its own speed error (how far actual screw RPM is from command) and also the speed difference between itself and the traction motor. That difference — the coupling error — gets fed back into both controllers as a correction term.

If the extruder slows down because melt pressure spiked, the coupling term immediately tells the traction motor to slow down too. Not because a master said so, but because the two axes detected a mismatch and corrected it together. The synchronization error between them collapses much faster than any single PID loop could achieve.

Deviation Coupling for Three or More Axes

When you add a third axis — say, a parison programmer that squeezes the parison at the bottom to reduce waste — simple pairwise cross-coupling is not enough. Deviation coupling extends the concept by comparing every axis against a virtual reference trajectory that all axes should follow.

Each controller computes its own tracking error and also the average deviation of all axes from that common reference. The average deviation is distributed back to each axis as a corrective signal. This way, no single axis can drift far from the group without feeling a pull back toward the center.

In practice, deviation coupling works remarkably well for the extruder-traction-programmer triad. The parison programmer often introduces sudden load changes when it clamps and releases. Without coupling, the traction motor stumbles every time the programmer activates. With deviation coupling, the traction and extruder both absorb the disturbance cooperatively, keeping parison weight variation under 2% even during aggressive programming cycles.

Electronic Gearing and Cam-Based Synchronization

Virtual Cam Profiles for Cyclic Motion

Blow molding is inherently cyclic. Every cycle repeats the same sequence: extrude, traction pull, mold close, blow, cool, mold open, eject. Electronic gearing maps each axis to a virtual cam profile that defines its position, speed, and acceleration throughout the cycle.

The cam profile is stored in the motion controller as a lookup table — typically 256 to 1024 points per cycle. At runtime, a cyclic interpolator reads the table at a rate synchronized to the master clock, generating smooth trajectory commands for every axis simultaneously. The extruder might follow a trapezoidal speed profile: ramp up, hold, ramp down. The traction follows a profile that starts slightly after the extruder ramps up, peaks in the middle, and decelerates as the parison reaches target length. The clamp follows a step function: fully open, then snap shut, hold, snap open.

The beauty of cam-based coordination is that all timing relationships are baked into the profile. You do not need real-time communication between axes to keep them in sync — the interpolator enforces it by construction. Jitter between axes is limited only by the servo update rate, typically 250 microseconds to 1 millisecond.

PREVIOUS:Energy-saving frequency control system for extrusion blow molding machine NEXT:Method for Synchronous Speed Control of Extrusion Blow Molding Machines