Direct action

When I last put the cases for and against the use of linear motors, it was clear that while linear drive technology had a future in an increasing number of applications, ballscrew technology was not about to go quietly.

That was four years ago. Today, I can justifiably draw the same conclusions. What is also clear, however, is that confidence in linear technology has not risen anything like as sharply as predicted.

This may be due entirely to the general downturn in world markets, but I cannot discount the possibility that many of the initial problems that builders were faced with are still there and have yet to be dealt with satisfactorily.

It seems that linear technology just cannot produce the power required for heavy machining operations without producing excessive heat, due to the size, or number, of motors required. Supporting this is the fact that most of the more recent innovations have focused on the increased speed or compactness linear technology offers over conventional drive systems.

In 2000, business consultant, Frost & Sullivan predicted that the machine tool sector would surge ahead of semiconductors as the largest application market for linear motors, accounting for almost a quarter of Europe’s market by 2008.

This prediction looked accurate, based on the growth of the market at the time, and was supported by global leader Siemens, which forecast that over the next few years 20–30 per cent of all machine tools will be equipped with direct drives. Clearly, this is not the case.

Linear motors are manufactured in two basic forms — ironless and ironcore. Both systems have the advantage of not requiring support bearings, bearing blocks, a motor coupling and a screw, making application designs simpler. The iron-core systems can produce greater thrust than the ironless systems, making them more suitable in heavy-duty applications.

Neither system, however, is capable of producing the mechanical advantage of ballscrews. As a result, they do not deliver the same levels of continuous force. If they are pushed too hard, there is a risk of demagnetisation and failure.

The greatest drawback of using linear drives is caused by the high attractive forces generated between the coils and magnets. These are typically five times greater than the axial thrust provided, and are capable of deforming the machine structure unless accounted for during the machine tool design phase.

This is not straightforward, as any increase in mass will have a direct effect on acceleration. Another problem associated with these high attractive forces is that small particles of ferrous swarf will be attracted into the drives. If things go wrong and the system becomes contaminated, the cost of repair will be high.

There are also the health issues, as the strong magnetic fields are known to interfere with heart pacemakers.

Linear drives do, however, have the advantage of high acceleration — typically 10m/sec2, to a contouring speed of up to 120m/min. They have no backlash or pitch error and do not suffer from the effects of mechanical friction or wear on the drive components, meaning high contouring accuracy can be maintained. They also operate at low noise levels and can be kept stable by using coolant circulation around the system to control temperature.

The big problem with replacing a conventional drive with a linear motor is that it must be designed into the machine tool from the start; motor, control and software must be matched perfectly. It’s only when somebody is looking at a totally new machine that they will be likely to make the leap from existing to newer technologies.

This is the approach taken by DMG which claims to have largely overcome all of these problems by designing its machining centres with linear drives in mind from the onset, supporting my argument that it is impossible to simply adopt linear drives to a machine that was designed for conventional drives.

General DMG design features include twinned motors to prevent torsional twist; a chiller unit to maintain a constant working temperature; and thermal temperature sensors and feedback systems to compensate for local growth. However, the most interesting aspect of DMG designs is that everything expands and contracts in balance (thermo-symmetric), thus stabilising the machine.

Citizen has taken the lateral approach of designing its new R07 Type VI CNC sliding head auto (due to be launched at the EMO exhibition in Hanover this month) based on the fact that linear technology offers compactness plus high-speed one-hit machining cycles capable of holding one micron repeatability.

By combining linear drive technology to both of the machine’s tool posts, high productivity is guaranteed with rapid acceleration to maximum traverse rates of 20m/min.

Sodick claims that because EDM is a non-contact machining process, this makes it a perfect fit for linear motor technology. Unlike machining centres that take advantage of the table speeds of linear motors, EDM uses linear technology's speed for the Z-axis (ram stroke) in order to create its own flushing capability. It also uses the speed of linear motors to react to changes in the spark gap. Linear motor EDMs will excel in difficult-to-flush applications.

These examples demonstrate that linear technology is a viable option — so why is there a problem taking up the technology? Well, most would argue that ballscrews are far cheaper to run, and they do not normally require expensive linear scales to provide accurate positioning data. A servodriven ballscrew is more suitable when high-feed forces are involved, and is more economical to install for shorter axis movements.

More importantly, they do not require the higher levels of complexity for emergency stop arrangements — critical for any machine tool safety.

The ballscrew provides mechanical advantage as it transforms rotary motion into linear motion. The lower the lead of the screw, the higher the mechanical advantage. By the same token, the lower the lead, the lower the axis velocity for a given servo speed.

For example, a screw with a 5mm lead can travel at 10m/min if the servomotor is capable of 2,000 revs/min. A screw with a 10mm lead will travel 20m/min (about 792in/min) given the servo motor revs/min, but will require twice the motor torque to provide the same axis thrust.

For precision applications, a ground low-lead angle is recommended, which inherently limits the speed that a driven table can achieve.

Linear motors have no such restriction.

Two problem areas associated with ballscrews are cumulative/lead variation errors and backlash.

Cumulative error over the length of a ballscrew can be compensated by a CNC control system; however, lead variation error requires a vast amount of computational power to completely correct, hence most algorithms merely ease the problem, and not fix it.

While linear drives continue to gain ground in machine tool applications, these are likely to remain at the high end of the spectrum.