Every cycle, the same bang. A press feeder cylinder, 63 mm bore, 400 mm stroke, cycling 12 times a minute. At full speed the piston hit the end cap hard enough that people two bays over could hear it. The maintenance supervisor said it had always done that. “It’s just how it is with pneumatics.” It isn’t. The cylinder had adjustable cushions, and the cushions were doing almost nothing.

The fix wasn’t complicated, but it took a while to find because the obvious knob was the wrong one.

What a cushion actually does

Near the end of stroke, a cushioning spear enters a bore in the end cap. The remaining air in the chamber has to escape through a small needle valve instead of the main port. Trapped air compresses, pressure builds, the piston decelerates. The needle valve controls the escape rate. Close it and the cushion is aggressive. Open it and the piston runs on past the cushion and slams.

Whoever set this machine up had opened the needle valve fully, probably because the cylinder crept or stalled when the cushion was engaged. The cushion became a straight-through passage. The piston was doing the last 25 mm of travel at full speed and hammering the cap.

There is a rated cushioning capacity for every cylinder, and it depends on moving mass and speed. The 63 mm cylinder at 6 bar pushing a 40 kg carriage moving at 0.5 m/s has a kinetic energy of about 5 J. A 63 mm bore cylinder’s cushion is rated for roughly 3-4 J at this stroke. We were already over the cushion rating before we touched the valve.

The math that said the cylinder was too small

Kinetic energy is straightforward:

E = ½ m v²

For the carriage, m = 40 kg (including the tooling it pushed), v = 0.5 m/s at the cushion entry point. That is 5 J. The cushion absorbs that energy over its stroke length, say 25 mm, so the average deceleration force is 5 J / 0.025 m = 200 N. That doesn’t sound like much, but it’s the average. The peak is several times higher because deceleration isn’t uniform.

The catalog for the cylinder listed cushioning capacity at 4 J for this bore and stroke. We were at 5 J before adding any margin. The right answer was not a bigger cushion setting; it was a bigger cylinder or a speed reduction. Nobody wanted to redo the mounting, so we cut the speed.

We changed the flow control on the extend port from 0.5 m/s to 0.35 m/s. Kinetic energy dropped to 2.45 J, comfortably inside the cushion rating. Then we set the cushion needle to the factory mid-position and ran it. The bang became a soft hiss. The cycle time went from 2.1 seconds to 2.6 seconds, which the production schedule could absorb.

What we should have checked first

In hindsight the sequence was wrong. We started by adjusting the needle valve and made things worse. The order that works is:

  1. Measure the actual approach speed with the valve fully open (or time the stroke).
  2. Compute the kinetic energy at cushion entry.
  3. Compare against the catalog cushioning capacity.
  4. If it fits, adjust the needle. If it doesn’t, slow the cylinder or upsize it.
  5. Only then touch the needle valve, and only a quarter turn at a time.

The other thing nobody checked was the mounting. The cylinder was bolted to a bracket that flexed. The bracket amplified the shock and made the noise worse. A stiffener plate behind the bracket cut the transmitted vibration noticeably. The bang was a combination of three things: an oversized load for the cushion, a fully open needle, and a springy bracket. Fixing any one of them would have helped. Fixing all three made it quiet.

One more detail. The rod side cushion on this cylinder had never been touched. The retract stroke ended in a hard stop too, but it was quieter because the return stroke ran slower and the mass wasn’t being pushed. We set that needle the same way and the return became smooth as well.

Cushioning capacity is a number in the catalog, not a suggestion. If your cylinder slams, check the energy first and the needle valve second. A 40 kg carriage at 0.5 m/s is already past what a 63 mm cylinder can cushion. Slow it down or upsize it, then tune the cushion. The bang is a calculation error wearing work clothes.