The Vacuum Cup That Dropped the Carton Every Other Cycle
We picked up cardboard cartons with a single Ø50 mm vacuum cup. The holding force was calculated at about 12 N (at -60 kPa). The carton weighed 0.5 kg (5 N). The math said it would hold. On the floor, it dropped every other cycle. The cartons had a textured surface (corrugated) that leaked air. The vacuum level dropped from -60 kPa to -20 kPa because the ejector couldn’t keep up with the leak. At -20 kPa, the holding force was only 4 N — less than the carton’s weight. The cup was sized for a perfect seal, not for a corrugated surface. We switched to a larger cup (Ø80 mm, more suction area) and a bigger ejector (more flow). The vacuum held at -50 kPa even with the leak. The drops stopped. The mistake was calculating holding force at the ultimate vacuum, not at the working vacuum under leak.
Vacuum gripper sizing has two calculations: holding force (will it lift the part?) and flow capacity (can it maintain vacuum against leaks?). Most people only calculate the first. This article runs both.
Holding Force: The Basic Calculation
The vacuum creates a pressure difference between the atmosphere and the cup. The holding force is the pressure difference times the effective area.
F = ΔP × A_cup
Where ΔP is the pressure difference (atmospheric minus vacuum, in Pa) and A_cup is the effective cup area (m²).
Example: Ø50 mm Cup
A Ø50 mm cup has an effective area of A = π × (0.025)² = 0.00196 m². At -60 kPa (ΔP = 60,000 Pa): F = 60,000 × 0.00196 = 118 N. That holds about 12 kg. That sounds great for a 0.5 kg carton.
But the working vacuum isn’t -60 kPa. On a leaking surface, it’s lower. At -30 kPa (ΔP = 30,000 Pa): F = 30,000 × 0.00196 = 59 N. Still 6 kg. At -20 kPa: F = 39 N (4 kg). Still above 5 N. So why did it drop?
Because the cup’s effective area is less than the geometric area. A bellows cup doesn’t seal across its full diameter — the sealing ring is smaller. The effective diameter might be Ø40 mm, not Ø50. A = π × 0.02² = 0.00126 m². At -20 kPa: F = 25 N. Still above 5 N. The real issue was the acceleration: the robot picked and accelerated at 2 g. The effective load during acceleration is m × (g + a) = 0.5 × (9.81 + 19.6) = 14.7 N. At -20 kPa with the smaller effective area, F = 25 N — still enough, but barely. And with a leak, the vacuum kept dropping. The cup didn’t have margin.
Safety Factor: Don’t Trust the Static Calculation
The holding force must exceed the actual lifting force (weight + acceleration + margin).
F_required = m × (g + a) × safety_factor
- m = part mass (kg)
- g = 9.81 m/s² (gravity)
- a = acceleration of the robot (m/s²). For a 2 g pick, a = 19.6 m/s².
- safety_factor = 2–4 (for vertical lifting, 2; for horizontal or high-speed, 4).
For the 0.5 kg carton at 2 g, vertical: F_required = 0.5 × (9.81 + 19.6) × 2 = 0.5 × 29.4 × 2 = 29.4 N. The cup must deliver at least 30 N. At -30 kPa with the Ø50 mm cup (0.00196 m²), F = 59 N — enough. But at -20 kPa (leak), F = 39 N — still enough but with less margin. The problem was that the leak dropped the vacuum to -15 kPa: F = 29 N — right at the limit. Any variation and it drops.
| Cup Diameter | Effective Area (m²) | Force at -60 kPa | Force at -30 kPa | Force at -15 kPa |
|---|---|---|---|---|
| Ø20 mm | 0.000314 | 19 N | 9 N | 4.7 N |
| Ø40 mm | 0.00126 | 75 N | 38 N | 19 N |
| Ø50 mm | 0.00196 | 118 N | 59 N | 29 N |
| Ø80 mm | 0.00503 | 302 N | 151 N | 75 N |
| Ø100 mm | 0.00785 | 471 N | 236 N | 118 N |
Flow Capacity: The Leak Problem
The holding force calculation assumes a perfect seal. Real parts leak. The vacuum ejector (or pump) must replace the leaked air to maintain the vacuum level.
Estimate the Leak Rate
The leak depends on the surface. Estimate it:
- Perfect seal (glass, machined metal): Near zero. A small ejector is fine.
- Good seal (molded plastic, smooth surface): Small leak, maybe 1–2 L/min (standard liters per minute).
- Poor seal (corrugated cardboard, textured plastic): Significant leak, 5–15 L/min.
- Very porous (foam, wood): Heavy leak, 20+ L/min. Need a reservoir or a bigger pump.
Pick the Ejector for the Leak
The ejector’s suction flow (L/min free air) must exceed the leak rate. If the leak is 10 L/min and the ejector pulls 5 L/min, the vacuum drops (the leak wins). If the ejector pulls 20 L/min, it maintains the vacuum.
For the carton example: leak is about 8 L/min. We used a small ejector (10 L/min free air). That’s barely enough — any extra leak (a worn cup, a rough spot on the carton) drops the vacuum. We upsized to a 25 L/min ejector. It maintained -50 kPa even with the leak.
Vacuum Switch Setpoint
A vacuum switch confirms the part is picked. Set it above the leak-induced vacuum level. If the working vacuum is -50 kPa, set the switch at -30 kPa (alarm if it doesn’t reach -30). Don’t set it at -60 kPa (the ultimate vacuum) — the leak prevents reaching it, and the switch never confirms.
Multiple Cups: The Total Area
When using multiple cups (4, 6, or more), the total holding force is the sum of all cups. But each cup leaks independently. The total flow requirement is the sum of all leaks.
For a 4-cup gripper on cardboard: each cup leaks 5 L/min, total leak = 20 L/min. The ejector (or pump) must pull 20+ L/min. The holding force is 4 × 59 N (at -30 kPa) = 236 N — plenty for a 5 kg box (49 N weight, 98 N at 2 g).
Add check valves at each cup. If one cup loses the seal, its check valve closes, and the other cups maintain vacuum. Without check valves, one leaking cup drops the whole gripper.
The vacuum sizing workflow: 1) Calculate the required holding force: F_req = m(g+a) × safety. 2) Pick cup size(s) where F_hold (at working vacuum, not ultimate) exceeds F_req. 3) Estimate the leak rate based on surface quality. 4) Size the ejector/pump for flow exceeding the leak. 5) Set the vacuum switch above the leak level. 6) Add check valves for multi-cup grippers.
Vacuum Reservoir: For Fast Cycles
For high-cycle applications (pick every 1–2 seconds), the ejector can’t build vacuum fast enough. A reservoir (a small tank) helps.
The reservoir is pre-evacuated to vacuum. When the cup hits the part, the reservoir dumps its vacuum instantly. The ejector then re-evacuates the reservoir during the robot’s move to the drop-off. This gives a fast pick without a huge ejector.
Size the reservoir volume: it should hold 2–3× the cup volume plus the tubing volume. A 0.5 L reservoir for a few small cups works. The ejector just needs to recharge it between cycles.
Tube Size and Volume
The tubing between the ejector and the cup adds volume. A long, large tube takes longer to evacuate (more volume to pull a vacuum on). For fast picks, use short, small tubing (4 mm ID, under 1 m). Mount the ejector on the robot (right next to the cup) to minimize the tube length.
If the tubing is long (over 2 m), the vacuum takes longer to reach the cup. The cup doesn’t hold instantly — there’s a delay. This delay adds cycle time. Mount the ejector close.
When to Use a Vacuum Pump Instead of an Ejector
Venturi ejectors (compressed air) are simple but waste air. For high-cycle or continuous use, an electric vacuum pump is more efficient.
- Ejector (Venturi): Cheap, light, fast to respond. Uses compressed air continuously. Good for intermittent picks (a few per minute).
- Electric pump (diaphragm / rotary vane): More efficient (no compressed air). Heavier, more expensive. Good for many cups, long tube runs, or continuous vacuum. For 24/7 operation, the pump pays back in air savings.
A Vacuum Sizing Checklist
- What is the part mass? (kg)
- What is the robot acceleration? (g or m/s²)
- F_required = m(g+a) × safety (2–4).
- What is the surface? (Determines leak rate.)
- Pick cup size(s): F_hold at working vacuum > F_required.
- Estimate total leak rate (sum of all cups).
- Ejector/pump flow > total leak rate (with margin).
- Vacuum switch setpoint above leak level.
- Check valves at each cup (for multi-cup).
- Short tubing (mount ejector close to cup)?
- Reservoir for fast cycles?
- Ejector vs pump (based on cycle rate and air cost)?
The Bottom Line
Vacuum gripper sizing isn’t picking a cup from a catalog. The holding force calculation is the easy part (F = ΔP × A). The hard part is estimating the leak rate and sizing the flow to maintain vacuum. The cup that dropped the carton wasn’t too small in area — it was fed by an ejector that couldn’t overcome the corrugated leak. Calculate the required holding force with acceleration and safety factor, estimate the surface leak, and size the flow for the worst case. The gripper that holds every part every cycle isn’t lucky — it’s sized for the real leak, not the perfect seal on a sample.
