The Vacuum That Took Too Long to Grip: Flow Path, Not Cup Size
A vacuum gripper taking 1.5 seconds to build enough vacuum to lift. The cycle allowed 0.5 seconds. The robot was waiting for the vacuum switch. The customer thought the ejector was too small. We measured the plumbing. The ejector was fine. The tubing was the bottleneck.
How long vacuum takes to build
When the ejector turns on, it pumps air out of the cup, the tubing, and the manifold. The vacuum level rises as the air is removed. The time to reach -60 kPa depends on two things: the ejector’s pumping speed (liters/min at vacuum), and the volume to be evacuated (cup + tubing + manifold).
t ≈ V / Q_pump · ln(P_initial / P_target)
Where V is the volume to evacuate (liters), Q_pump is the ejector’s free-air pumping rate (L/min), and P is the pressure ratio. For our setup: V = 0.05 L (cup) + 0.2 L (tubing, 4 m of 6 mm) + 0.1 L (manifold) = 0.35 L. Q_pump = 20 L/min (the ejector’s rating). t = 0.35/20 × ln(101.3/40) = 0.0175 × 0.93 = 0.016 min = 0.96 seconds. We measured 1.5 seconds. The extra time was the tubing’s restriction — the effective pumping speed at the cup was lower than the ejector’s rating because the tubing resisted flow.
The tubing restriction
A long, narrow tube between the ejector and the cup acts as a restriction. The ejector pumps at 20 L/min at its own inlet, but by the time that pumping reaches the cup through 4 m of 6 mm tubing, the effective speed is lower. The conductance of the tube (how much air it passes) depends on the tube’s diameter and length:
C ∝ D⁴ / L
The conductance scales with the 4th power of diameter. Doubling the tube diameter makes it 16× more conductive. Shortening the tube by half doubles the conductance. The 4 m of 6 mm tubing was the bottleneck.
What we changed
1. Move the ejector closer to the cup. We mounted the ejector directly on the tool, 100 mm from the cups, instead of 4 m away on the electrical panel. The tubing between ejector and cup dropped from 4 m to 0.1 m. The volume dropped from 0.35 L to 0.06 L. The vacuum build time dropped to 0.2 seconds.
2. Use larger tubing between the valve and the ejector. The supply air to the ejector was also 6 mm, 4 m long. We upsized to 8 mm. The ejector got full supply pressure. The pumping speed improved from 20 L/min to 28 L/min.
3. Add a small reservoir. A 0.2 L reservoir between the ejector and the cups pre-charges the vacuum. The cups see instant vacuum, even before the ejector builds. The robot doesn’t wait.
| Setup | Volume (L) | Effective Q (L/min) | Time to -60 kPa |
|---|---|---|---|
| Original (4 m, 6 mm) | 0.35 | 12 | 1.5 s |
| Ejector at tool (0.1 m, 6 mm) | 0.06 | 18 | 0.2 s |
| Reservoir added | 0.06 + 0.2 pre-charged | 28 | <0.1 s |
The vacuum switch timing
The PLC shouldn’t wait a fixed time for vacuum. It should wait for the vacuum switch to confirm. The switch (set at -60 kPa) closes when the vacuum is ready. The robot lifts. If the switch doesn’t close within 1 second, the PLC alarms (leak, blocked cup, etc.). This is faster and safer than a timer.
The rule
For fast vacuum cycles, the ejector must be near the cups. Long tubing between ejector and cup adds volume and restriction. The ejector on the panel, 4 m away, means the vacuum has to travel through the tubing. Move it to the tool. Pre-charge with a small reservoir. The vacuum that took 1.5 seconds now takes 0.2.
The vacuum speed rule: Vacuum build time is dominated by volume and tubing restriction, not ejector size. Move the ejector close to the cups (under 200 mm). Shorten and widen the tubing. Pre-charge with a reservoir. A bigger ejector won’t help if the tubing between it and the cup is 4 meters of 6 mm.
