Sway control devices are becoming ever more useful, and used - Digital Dream Technology support

The principle of anti-sway is, in its basics, simple. A load suspended from a rope will act like a pendulum. If the support-point of the pendulum—that is, the trolley on a hoist or on the jib of a tower crane—starts to move, the pendulum and the load on the end of it will start to sway. Slowing or stopping the trolley will cause the rope and load to sway. Anti-sway systems modify the acceleration and deceleration of the trolley in such a way so as to make the two swayings cancel each other out. The rope stays effectively vertical while the load is being moved.

A very skilled crane operator can get close to such a result manually. Computers modifying the operator’s accelerations and decelerations of the crane trolley can achieve it with near perfection. “With a modern anti-sway system the departure of the lifting rope from vertical is reduced to virtually zero,” says Dieter Feldhinkel, technical director of Mannheim-based SWF Krantechnik.

SmartCrane are another company that have been delivering sway control systems since 2001. They explain the principle they use in more detail: “If an operator moves the control stick to full speed, the SmartCrane Anti-sway Control accelerates initially according to the operator’s demand, inducing an initial load sway. When about half the reference velocity has been reached, the anti-sway ‘coasts’, i.e. maintains constant velocity, for a short time. Then the trolley is accelerated again, this time to the full operator demand velocity.

This second acceleration removes the sway induced by the first acceleration, so the trolley is now travelling at the operator reference velocity with the load hanging directly below the trolley. When the operator releases the stick demanding zero velocity, the same process is repeated in reverse to bring the load to a stop without sway.”

All this requires precise timing. The timing depends on the natural frequency of the pendulum motion, which in turn depends principally on the length of the rope above the load. Anti-sway software takes into account changes in hoist cable length as the load is raised or lowered, feeds its modifications of the operator’s instructions back to the trolley, and does so in real time.

Stefan Elspass is in charge of product management for HMI—Human-Machine Interfaces—and control systems at Demag cranes. Demag has been in the lifting business for many years—it will celebrate its 200th anniversary in 2019—and so can claim first-hand expertise in past and present developments. In 2003 the company launched the first processor controlled rope hoist—a milestone, they say, and a first step towards anti-sway.

“We have a long tradition of serial hoist systems,” says Elspass. “We started sway control in 2004. Since then it has gone through several evolutions.

“Over the past ten years waves of embedded control technology have more and more been entering lifting products, especially for users involved in serial production of high volumes of products.”

There are, says Elspass, two philosophies, or approaches, in anti-sway technology. “The first approach is based on parameterisation. The control systems are set into operation by mathematical parameters or algorithms—a mathematical model of the way in which the load will move when the rope lifts it and the trolley accelerates. The mathematics inside the computer predicts the amount of sway, and sends instructions on trolley movements to reduce this to zero.”

“The key feature is that the SmartCrane Anti-sway Control uses precise timing of accelerations to control the sway, rather than real-time sway measurement and control feedback. It does not require a camera or other sway-sensing device to control sway induced by moving the crane,” says the company.

Ari Lehtinen, manager, industrial cranes automation for Konecranes, describes these as “open-hook” systems. “The open- hook system is controlled by mathematical models,” he says. “The mathematics needs the rope length, and the acceleration and speed of the trolley in order to send appropriate instructions to the trolley.”

Sensors mounted on the drum can detect how much rope has been paid out, and this is the primary data that the algorithm needs. “It is a fairly accurate system,” says Lehtinen. “When it is working in good co-operation with the motor control it gives practically no sway.”

He qualifies that, for accuracy: “It gives practically no observable sway. Because of course if there was no sway at all then the load could not be moved except vertically up or down. It is the angle of the rope that moves it along or sideways. But the sway is so small that operator does not see it at all.”

This first approach uses mathematics to predict the angle that the rope will make with the vertical at any time. The second approach, says Elspass, is less theoretical: it uses sensors to detect the angle of the lifting rope, and feeds back that information to the processor that is controlling the crane. He believes that this approach has advantages: “The parameterisation approach assumes that all the surrounding conditions are ideal: for example that the load really is exactly vertically beneath the hook when lifting begins. But in reality there is often a slight angle, perhaps from the operator being in a hurry, which will induce sway. Or wind conditions can change.

Sensors, he says, give information based on the actual, rather than the theoretical, situation at the crane. As well as detectors mounted on the drum, lasers pointed at detection marks on the rope or at the hook can measure distances to great accuracy.