A robot picks a machined part with a parallel gripper. The catalog force looks ample, the air pressure is set, and the part slips during a fast swing or drops at the first oil-covered surface. The gripper cylinder is not undersized. The friction between the fingers and the part is, and the catalog gripping force is only the start of the calculation.
Gripping force vs holding force
The catalog force is the closing force the jaws exert at a given pressure, measured on clean, flat test surfaces. The force that actually holds the part against gravity and acceleration depends on the friction coefficient between the jaw faces and the part. Holding force equals gripping force times friction coefficient, times the number of contact surfaces. A friction coefficient of 0.1 on oily steel means only a tenth of the closing force resists sliding.
This is why the same gripper holds dry, rough parts securely and slips on smooth, oily ones. The closing force did not change; the friction did.
The friction coefficient is not a constant
Dry machined steel against steel might give 0.15 to 0.25. Oily or polished surfaces fall toward 0.1 or below. Rubber and polyurethane pads raise friction to 0.4 or more and conform to the surface. The actual value depends on oil, coolant, surface finish, and the pad material, so a number from a generic table is only a starting point. Test the real production surface, including the coolant or oil that will be on it.
Friction also changes as pads wear and harden, and as coolant builds up on them. A setup validated on new pads can slip weeks later without any pressure change.
Forces during robot motion
A fast acceleration or a quick rotation adds inertial force beyond the part weight. When the robot swings, the gripper must hold the mass through acceleration and deceleration in changing directions. A part that hangs still at 1 g can see effective loads of 2 g or more during aggressive moves. Calculate holding force for the worst acceleration in the direction that pries the part out, not just for static hanging.
Rotating a gripped part about a vertical axis drives it outward against the jaws; downward moves and sudden stops load it differently. The worst orientation in the cycle sets the required grip.
Two-finger and three-finger contact
A two-finger parallel gripper contacts the part on two sides, and friction at both surfaces holds it. A three-finger concentric gripper distributes force around a round part and centers it, which suits shafts and cylindrical components. Encompassing or profile jaws that wrap the part support it mechanically rather than relying on friction alone; they are the reliable choice for heavy or oily parts where friction cannot be trusted.
If the part geometry allows a cradle or hook under a feature, let the jaw carry the load mechanically. Friction should be the backup, not the only thing holding a heavy part in the air.
Jaw pads and surfaces
Replaceable pads let you choose friction and protect finished surfaces. Serrated or knurled steel bites into rough parts but marks them; rubber and urethane pads grip without marking but can be cut by sharp edges and degrade with oil. Diamond-pattern and coated surfaces balance grip and durability. Match the pad to the surface and the marking tolerance, and treat pads as wear items replaced on condition.
Keep the contact flat and square. A jaw that touches the part on one edge concentrates force and lets it pivot; align the fingers so the full pad area contacts the part.
Sizing the gripper and pressure
Work backward from the required holding force: part mass times worst effective acceleration, divided by friction coefficient and contact surfaces. That gives the gripping force, which sets the gripper size and pressure. Use a safety factor of 2 to 4, higher for slippery or critical parts. Don’t size from the part weight alone at rest. Running higher pressure to compensate for low friction increases wear and can crush delicate parts; fixing friction with pads is better than forcing the jaws.
Stroke and part variation
The jaw stroke must cover the range of part sizes and still close firmly. A gripper near the end of its stroke on a large part may lose force or fail to seat; on a small part it may not open enough to release. Check force across the stroke, since gripping force can vary as the jaws move. Account for part tolerance and any chips or burrs that change the effective size.
Sensing and fail behavior
Confirm the jaws have closed on a part with position or proximity sensors before the robot moves, and verify part presence so the robot does not swing empty or release over the wrong location. On air loss, a gripper should stay closed or fail safely; springs or check valves maintain grip if supply drops. A part dropped on a power or air interruption points to a gripper with no fail-closed design.
A worked example
Take a 1.5 kg steel part swung at 15 m/s2, an effective load near 2.3 g, so the force to restrain is about 34 N. With oily steel friction at 0.1 and two contact surfaces, required gripping force is 34 divided by (0.1 times 2), or 170 N. Apply a safety factor of 3 and the gripper must close near 500 N. The same part on urethane pads with friction 0.4 needs only about 127 N with the same margin. The pad choice changes the required gripper by a factor of four, which is why jumping to a larger gripper before fixing friction is usually the wrong order.
Electric vs pneumatic grippers
Electric grippers control force and position independently and report jaw position, which helps handle fragile or variable parts without compressed air. They generally deliver lower peak force than pneumatic units of similar size and cost more. Pneumatic grippers are simple, fast, and forceful but set force mainly through pressure. Use electric where force control, feedback, or lack of air justifies it; for high force in repetitive handling, pneumatic remains the practical choice, provided the friction and pressure are set correctly.
Maintenance and wear
Worn jaw guides develop play, which lets the fingers tilt and reduces effective contact and force. Air leaks in the gripper or fittings lower the real closing pressure even if the regulator reads correctly. Inspect jaw motion for looseness, clean coolant and chips from the guides, and confirm the fingers meet parallel. Pads that are glazed, torn, or packed with swarf should be replaced; a pad change is cheap compared with a dropped, scrapped part.
Mixed and fragile parts
When one cell handles several part numbers, the jaw design and force must suit the most demanding part without crushing the most delicate. Soft jaws or interchangeable finger sets, selected with the part program, avoid one fixed grip that is too weak for heavy items and too strong for thin ones. Force-limited pressure regulation per part can also vary grip. Don’t validate only the most common part; run the heaviest, oiliest, and thinnest items through the same gripper.
Verifying the process
After setup, deliberately test the grip at the fastest robot moves and with parts prepared to the worst oily condition. A grip that passes slow motion can fail at full speed. Periodically recheck after pad wear and coolant changes, and record pad type, pressure, and jaw settings with the part program. If slips occur, inspect friction and acceleration before raising pressure; the evidence of where and when the part moves identifies whether the cause is a worn pad, an aggressive move, or genuinely undersized grip.
Surface condition in the real cell
Don’t assume parts arrive clean and dry when the gripper follows a machining or washing operation. Residual coolant, quenching fluid, or rust preventive changes friction for hours after processing, and parts from different suppliers may carry different oils. If the grip is marginal, the first failures often appear after a material or coolant change rather than on the original sample. Measure friction on parts taken straight from the upstream process, and recheck whenever that process changes. A short test that deliberately oils a part to the worst expected condition is more honest than a clean bench demo and prevents a grip that only works on ideal parts.
Finally, keep spare pads and fingers at the cell. When a grip starts slipping during a shift, replacing a worn pad immediately is far better than raising pressure or running the cycle slowly for the rest of the batch. A small stock of the correct consumables, marked with the part number, removes the temptation to improvise a surface that has not been validated.
Bottom line
The catalog gripping force becomes holding force only through the friction coefficient, which collapses on oily, smooth surfaces. Size for the worst acceleration in the cycle, use high-friction or encompassing jaws, and let geometry carry heavy parts rather than friction alone. Verify grip with sensors and design for air loss. Most dropped parts trace to friction and acceleration assumptions, not to a gripper that was simply too small.
