Pneumatic actuation plays a major role in today's world of computerized automation. It's reliable, economical, and surprisingly easy to use. Understanding pneumatics is a matter of physics. When air inside a container builds up, pressure magnitude is the same at all points within the fluid. The air also pushes out on its vessel uniformly. Pneumatic actuators leverage this fact of fluid dynamics to provide clean, quiet motion, with less waste heat and electromagnetic interference than their electric counterparts. They also excel in applications involving fast repetitive moves, heavy loads, and very smooth motion profiles.

Here's how it works: Compressed air enters an opening in a cylinder and pushes against the interior, including the one wall that can move: the piston. If the difference in force across the piston is larger than the total attached load plus frictional forces, the piston floor drops out. The resulting net force (proportional to the force to mass ratio) accelerates the load, converting pneumatic to linear mechanical power. With air power as the driving force, pneumatic actuators are safe for hazardous environments where electric sparks must be avoided.

Calculating force

Single-rod double-acting pneumatic actuators — those with air ports on both sides of the internal piston — are the most common in industry. We'll now explore the physics behind pneumatic motion, using this type as our example.

All pneumatic force depends on two things: air pressure and piston area. Let's say air pressure is set at 50 psi. If the diameter of the piston is 7 mm, then it has an effective surface area of 38 mm2. Force, then, is the product of pressure and area:

Force = Pressure × Area

Converting from psi to N/mm2, pushing force equals 13.1 N. However, notice that the piston rod reduces the effective area on its side, meaning that pull force is not as great as push force. Specifically, for pull force:

Effective area = Piston area - Rod area

If we assume the rod is 3 mm in diameter, then the rod area is 7.069 mm2. The effective area becomes 30.931 mm2 for a pull force of 10.66 N.

In contrast, double-rod double-acting actuators have equal push and pull forces and can be used for both actions. Rodless variations too have equal push and pull force, but higher than that of double-rod types for the same bore. Since its output linkage isn't contained, the internal piston drives an external slide as it moves. Similarly, cable air actuators have an external slide tied with a cable that's wrapped around a pulley at each piston end.

No matter what cylinder type is used, correct sizing is essential. Bigger cylinders — though they increase system natural frequency and allow faster accelerations — require larger, more expensive compressor pumps to power the system. As with any motion technology, oversizing is something to avoid.

Powering up

Compressors use mechanical power to generate pneumatic force, supplying the pressurized air that moves cylinder pistons. This kind of powering has distinct advantages. Most importantly, one economical air compressor and some readily available air can power several axes and devices. That's because the compressor only needs to supply the average amount of air required by the application per machine cycle. (Shortly, we'll talk more about what makes this true.) Compressors can also be installed where space is not critical. With the powering unit removed from the action, pneumatic actuators at points of motion can be small compared to the power they produce. They're also suitable for environments where contamination is unacceptable.

In heavier lifting applications, a sufficiently sized compressor can move any load put to its cylinders. Where positioning is the main objective — a sophisticated but increasingly common pneumatic application — the proper size ensures controlling pressure can move systems at adequate velocity and acceleration rates. However, another essential element in pneumatic power systems, pressure reservoirs, must also be present and sized correctly. (Note: Oversized compressors do not eliminate the need for reservoirs.)

Pressure reservoirs, also called accumulators, are precharged to store air under pressure when the system isn't moving. Their stored capacity takes care of loading when there is more than one actuator in the system — in effect, acting as pneumatic capacitors by storing air pressure for later use.

Accumulators are helpful for two reasons: They keep system pressure relatively constant, and store energy. The former is important for accurate motion control and smoothness when moving slowly; the latter allows for quick, dramatic power and eliminates the need to size compressors for peak loads. To function effectively, accumulators must be placed near the pneumatic system's inlets/outlets, which we discuss next.

Valves and controls

A more detailed look at pneumatic cylinders reveals valves, there to regulate cylinder speed and power by controlling the speed at which air enters the cylinder. Like accumulators, the physical placement of valves is paramount. Valves should be placed near or on cylinders to minimize the volume of gas between the two. Minimizing latencies and charge/discharge times helps keep system natural frequency high. Short, solid piping rather than hose should connect valves to cylinders, since hoses contract and change shape, and any change of area affects pressure and controllability.

Pneumatic controls meter the air through valves. The oldest employ on/off switching, with only two valve positions during operation: fully-open/released or fully-closed/pressurized. Sometimes called “bang-bang” control, valves in this setup are closed shut to build pressure at a constant rate, then snapped open once the system reaches the required pressure. While initially inexpensive, output motion is discrete and jerky. As a result, the shock and vibrations generated by these bang-bang systems can eventually cause equipment damage.

To smooth motion driven by bang-bang controls, sometimes cylinders are used as dampers. (In short, one flow control is connected to another with tubing.) Even so, bang-bang controls aren't sophisticated enough for precise positioning. One problem is air's compressibility, which can hinder the accuracy of output motion.

To compensate, newer electronic systems include variable valves. Sometimes, these are electronically controlled proportional valves, so named because they adjust the valve setting proportional to a command. Other times, they're servovalves designed to leverage the benefits of pneumatics and precise control. Linear valves (those that use most of their range) can function as servovalves or servo-quality proportional valves with feedback capabilities. The latter are generally easier to control, so response time and flow rates that match the application can be chosen to avoid oversized valves.

For high-performance motion, a desirable attribute is zero overlap, meaning these valves have no dead zone between active control ranges where fluid pressure is increased and decreased. Overlap can be useful in manually controlled systems, but never in precision positioning.

By definition, servovalves alone cannot guarantee precision performance. Accurate control of airflow (and the resulting motion) also relies on transducers and feedback loops. Pneumatic-system precision improves when multiple feedback loops are included; specifically, two proportional/integral/derivative feedback loops — for position and differential pressure — allows coordinated positioning with repeatability to ±0.7 mm or better.

For the position feedback loop, a motion controller gets its input from transducers mounted inside the cylinders. (Increasingly, magnetostrictive linear displacement transducers are used for this because of their consistent output without needing to home.) Then, for accurate pressure feedback, information is taken from pressure transducers at the cylinder's rod and cap ends, mounted on either side of the piston to measure differential pressure. From this information, the resultant differential force is calculated.

Advanced control

:One of the main benefits of servovalves in pneumatic systems that end-of-stroke positions are controlled electronically, eliminating banging and bottoming out. And because the system is automatically adjusted under program control, equipment is less prone to damage and downtime from adjustment errors. When using a motion controller though, the PID gains do not have to adjust for system pressure. Keeping system pressure constant reduces the need to change these controller gains — which underscores the importance of properly sized accumulators. If an accumulator is appropriate, then the controller can quickly and accurately respond to actual external changing conditions by adjusting drive output.

With position and pressure sensor input, controller algorithms can determine the most appropriate cylinder hovering position. Motion from this point reduces the required stroke distance and speed, which allows for more manageable deceleration and reduced impact. Although some simpler systems are capable of positioning from some midpoint, this is generally not precise or variable. Unlike discreet pneumatic systems, advanced servopneumatic control systems are capable of an infinitely variable midpoint cylinder position. Servovalves can be controlled by motion control algorithms that calculate precisely how much to adjust the valve for proper metering of air in and out of the cylinder. This allows for accurate and variable open and closed positions, as well as infinitely variable midpoint positions. The smooth motion that results increases throughput and decreases damage from shock.

Some controllers can also execute high-level commands for smoothly curving motion that carries a pneumatic actuator from one coordinate position to the next. For example, they can linearize a nonlinear motion such as a cylinder rod pushing on a rotating arm.

Connectivity is essential for sophisticated pneumatic operations. Servopneumatic cylinders are best controlled with local, dedicated controllers, but fieldbuses must centralize signals for overall system synchronization. These field or device buses must interface with other system control elements such as PLCs, industrial PCs, and HMIs; suitable for this are standardized, high-performance fieldbus connections such as Ethernet and Profibus.

Alternative spring arrangements

Some single-acting actuators use springs to move the piston in one direction; when pressurized, air pressure overcomes the force of the spring and compresses it. One caveat: The air on the spring side must be allowed to escape, or it will effectively stiffen the spring effect. Alternatively, air alone in single and double-acting cylinders can be used to provide spring force by closing off one side. Generated force is easily calculated. Pressure is known on the open side to be atmospheric; the closed side starts at atmospheric and increases in pressure as the volume is decreased:

P1V1 = P2V2

where volume is the included stroke of the cylinder multiplied by the effective piston area.