Key number
Adaptive force is mode-dependent
10-240 N
OnRobot 3FG15: flexible grip caps at 140 N, normal grip reaches 240 N (datasheet v2.0, 2026-05).
Hybrid Page · Tool + Report
Use the tool first to size required force/torque per finger, then use the report layer to validate assumptions, compare options, and reduce RFQ risk.
Published: 2026-05-25 · Last reviewed: 2026-05-25
Tool Layer
Enter payload and contact assumptions to get required force/torque per finger, plus a direct next action.
Recommended range: 0.05 to 25 kg.
Conservative default is μ=0.40 for mixed materials.
2.0 is a typical starting point for industrial handling.
Use 1.0 for slow transfer, 1.3+ for fast starts/stops.
Larger radius increases required actuator torque.
Vertical lift; use this as the default sizing baseline.
Empty state: enter values and click Calculate sizing.
You will get required normal force, per-finger torque, and a decision band.
Report Layer
This section compresses the most decision-relevant numbers, who this page is for, and who should escalate to custom design immediately.
Key number
Adaptive force is mode-dependent
10-240 N
OnRobot 3FG15: flexible grip caps at 140 N, normal grip reaches 240 N (datasheet v2.0, 2026-05).
Key number
Robotiq payload changes by grip physics
2.5 kg vs 10 kg
Robotiq manual lists 2.5 kg fingertip and 10 kg encompassing recommendations at μ=0.6, SF=2.
Key number
Parallel high-force benchmark
650-1,950 N
SCHUNK EGU 70 EI/M/B published range; use as force envelope reference, not topology equivalent.
Key number
Cell-level safety baseline
ISO 10218-2:2025
Robot-cell integration and end-effector hazards must be validated at application level (published 2025-02).
Tool results are only useful if assumptions are explicit. This layer documents model choices and data sources so teams can reproduce decisions, with standards boundaries and unresolved evidence explicitly marked.
Evidence update: 2026-05-25
Review cadence: every 6 months or when core vendor datasheets are revised.
| Model parameter | Setting | Why it exists |
|---|---|---|
| Force model | F_total >= (m * g * orientation * dynamic * SF) / μ | Converts payload risk into required normal force under friction hold. |
| Finger split | Per-finger force = F_total / 3 | Three contacts are assumed to share normal force evenly in first pass. |
| Torque conversion | T_finger = F_finger * r | Maps force to actuator demand by jaw radius from pivot axis. |
| Safety factor default | 2.0 | Practical starting value for early RFQ before full test evidence. |
| Grip-physics boundary | This model targets friction-dominant fingertip grip | Robotiq documents that encompassing stability is no longer friction-driven. |
| Catalog force boundary | Force varies by finger angle, diameter, and grip mode | OnRobot force charts and notes show angle/current/material dependencies. |
| Boundary policy | Out-of-range => bench test + custom sizing path | Prevents false confidence from catalog-only matching. |
| Boundary topic | Verified signal | Required action |
|---|---|---|
| Grip physics | Robotiq manual distinguishes fingertip (friction-dependent) vs encompassing (not primarily friction-dependent) stability. | Use this checker for fingertip-dominant cases; re-run with lower μ and verify if encompassing fixture can be applied. |
| Force window interpretation | OnRobot states force depends on finger angle, diameter, current limits, and material friction. | Do not treat catalog max force as constant across all diameters and motions. |
| Geometry offset sensitivity | OnRobot warns finger offset >0.5 mm at first contact can overload motor and gear. | Add fixture concentricity checks and finger-runout measurement before pilot. |
| Finger length and inertia | SCHUNK publishes max finger length/mass limits; Robotiq and OnRobot provide custom finger constraints. | For long/custom fingers, demand static/dynamic load calculations and proof test records. |
| Source | Signal used | Date | Reference |
|---|---|---|---|
| OnRobot 3FG15 product page | Payload up to 15 kg and force up to 240 N for electric 3-finger class | Reviewed 2026-05-25 | Open source |
| OnRobot 3FG15 datasheet v2.0 | Force 10-240 N (10-140 N flexible), hold on power loss, 5.3 N·m platform torque | Published 2026-05 (v2.0) | Open source |
| Robotiq 3-Finger instruction manual (official PDF) | Fingertip vs encompassing grip boundary, payload assumptions, and W=(2*F*Cf)/Sf logic | Manual revision 2021-06-17, reviewed 2026-05-25 | Open source |
| SCHUNK EGU 70 EI-M-B technical details | 650-1,950 N force, 70 mm stroke per jaw, 80-100% force maintenance, finger-length limits | Reviewed 2026-05-25 | Open source |
| ISO 10218-1:2025 | Part 1 covers robot-as-machine safety and routes integration hazards to ISO 10218-2 | Published 2025-02 | Open source |
| ISO 10218-2:2025 | Part 2 covers integration/commissioning/maintenance for robot cells including end-effector contexts | Published 2025-02 | Open source |
| ISO/TS 15066:2016 | Collaborative operation supplement to ISO 10218; ISO metadata flags revision in progress | Published 2016-02; confirmed 2022, revision status updated 2025 | Open source |
| ISO 12100:2010 | Risk assessment and risk reduction methodology for machinery safety lifecycle | Published 2010-11; last confirmed 2022 | Open source |
| OSHA 29 CFR 1910.212 | Point-of-operation and machine guarding requirements for operator hazard protection | Accessed 2026-05-25 | Open source |
| Framework | What it covers | Decision impact |
|---|---|---|
| ISO 10218-1:2025 | Defines robot-level safety; integration and application hazards are handled in ISO 10218-2. | Catalog matching is only a start. Cell-level safeguards and misuse analysis still required. |
| ISO 10218-2:2025 | Covers design/integration/commissioning/operation/decommissioning of robot cells and integrated components. | If your scenario falls into excluded domains (public access, lifting people, explosive environment), this page output is non-decisive. |
| ISO/TS 15066:2016 | Supplement for collaborative operation; currently marked as to-be-revised by ISO lifecycle status. | For human-robot shared spaces, verify latest revision path before locking force/pressure criteria. |
| OSHA 1910.212(a)(1)-(3) | Requires guarding for point-of-operation and machine hazards in U.S. workplaces. | End-effector selection must include guard/interlock concept, not only force sizing. |
| Open question | Current status | Minimum executable fallback |
|---|---|---|
| Continuous-duty thermal derating curves at max force | No reliable public curve found for all compared models (checked 2026-05-25). | Request vendor thermal derating data by duty cycle, ambient temperature, and supply current before PO. |
| Public fatigue life for custom finger geometries | Vendors provide baseline limits but not universal fatigue life for arbitrary custom fingers. | Define minimum life test protocol (cycles, load profile, failure criteria) in RFQ. |
| Site-specific friction drift across production lots | No transferable public dataset for your exact material/pad contamination pair. | Create incoming QA friction sampling and attach lower-bound μ to the sizing worksheet. |
Compare options using reproducible dimensions: force, stroke, retention behavior, use-case fit, and known limitations.
| Option | Payload | Force range | Stroke | Power-loss / retention | Best fit | Known limits |
|---|---|---|---|---|---|---|
| OnRobot 3FG15 (adaptive 3-finger) | Force-fit 10 kg; form-fit 15 kg | 10-240 N (10-140 N flexible) | External up to 152 mm, internal up to 176 mm | Brake holds grip on power loss | Mixed-diameter handling with self-centering topology | Force and speed depend on finger angle, diameter, and friction/material pair |
| Robotiq 3-Finger (adaptive 3-finger) | 2.5 kg fingertip, 10 kg encompassing | Actuation 40 N pinch equivalent, 100 N break-away | 0-167 mm opening, 155 mm max encompassing diameter | Self-locking raises break-away force over actuation | Best when geometry supports encompassing grip | Fingertip stability is friction-sensitive; object detection can miss thin parts |
| SCHUNK EGU 70 (parallel electric gripper) | N/A (force-driven selection) | 650-1,950 N | 70 mm per jaw | 80-100% gripping-force maintenance version available | High-force tasks with predictable parallel-jaw contact | Not a 3-finger topology; finger length and mass limits strongly constrain setup |
| Custom 3-finger electric actuator module | Defined by test plan | N/A until validated | Custom | Depends on brake/mechanism architecture | Out-of-range scenarios or special geometry requirements | Higher integration, safety, and lifecycle validation effort |
| Risk | Probability | Impact | Mitigation action |
|---|---|---|---|
| Slip during acceleration spikes | Medium-High | High | Increase friction certainty, validate dynamic multiplier with motion trace data. |
| Torque underestimation from long fingers | High | High | Measure real jaw radius and include finger mass/inertia in the model. |
| Catalog force misread as safe payload | Medium | High | Map force to part geometry and friction, not only to payload headline. |
| Unstable pad friction in production | Medium | Medium-High | Run lot-level friction checks and define replacement intervals (Robotiq pads are consumables). |
| Safety scope mismatch with actual deployment | Medium | High | Map application scope against ISO 10218-2 exclusions and run ISO 12100-style hazard review. |
If your force/torque result is near boundary, send assumptions now and lock a validation path before pilot tooling is frozen.
Each scenario maps assumptions to an action path, so the team can decide quickly without skipping validation.
Assumption: 2.5 kg part, μ=0.35, dynamic=1.25, radius=22 mm.
Process: Checker result lands in review band; move to pad upgrade + slip test.
Output: Prefer high-force adaptive actuator or custom finger geometry.
Assumption: 0.8 kg part, μ=0.6, dynamic=1.0, radius=16 mm.
Process: Checker result lands in good-fit band with safety margin.
Output: Catalog adaptive electric 3-finger is usually enough.
Assumption: 4.2 kg with eccentric center of mass and quick repositioning.
Process: Result often jumps to out-of-range due torque demand.
Output: Switch to custom module or redesign fixture/orientation path.
Assumption: Measured μ varies from 0.2 to 0.5 across batches.
Process: Run worst-case input to set lower bound before RFQ.
Output: Use conservative SF and define friction re-test checkpoint in QA plan.
Assumption: Sizing result is good-fit, but shared-workspace force/pressure limits are not validated yet.
Process: Treat checker output as preliminary only and move to ISO 10218-2 + ISO/TS 15066 assessment path.
Output: Decision stays pending until safety validation evidence is complete.
Grouped answers for sizing, integration reliability, and procurement execution.
Send your checker output, part photos, and cycle profile. We reply with actuator options, risk notes, and validation plan.
Inquiry Email
Send target torque/speed, quantity, and destination for faster RFQ response.