Cart
Your cart is empty

A hub motor drives the wheel forward and pushes back on its own axle with equal force. That reaction torque is what spreads dropouts, spins axles and severs motor cables on high-power builds — and it’s why torque arms exist. Whether you need one arm, two, or a custom plate is decided by measured torque, axle fit, dropout strength and regenerative braking — never by the wattage on the box. This guide covers how torque arms work, when one is enough, when two are needed, and how to mount them safely on 1000W–4000W hub motors.

A torque arm moves the axle’s reaction force away from the small dropout slot and into a larger, stronger part of the frame. Wattage alone cannot tell you how many you need. A 1000W rear motor on thick steel dropouts with no regen, a tight axle-plate fit and an approved single-arm setup may be fine with one engineered arm. A 2000W–4000W build, a front hub, aluminium dropouts, cargo loads or regenerative braking should trigger an engineering review of that specific motor, controller, frame and braking setup. Two thin, loose arms are not automatically safer than one properly fitted heavy-duty arm — and no torque value, washer order or mounting direction should ever be copied from one motor to another.
Main guide: Can Any Bike Be Converted to Electric? Compatibility Checklist. Related in this series:
A torque arm is an anti-rotation component that fits over the flat-sided motor axle and anchors to the frame or fork. It does two jobs: it stops the axle rotating in the dropout, and it transfers reaction torque into a stronger area of the frame. It is not a wheel-retention device — do not rely on it to hold the wheel if axle-nut clamping force is lost.
The physics is simple. The motor turns the wheel forward, so the axle pushes back with equal force. The axle flats press against the dropout sides, and repeated loading can spread, crack or deform them. A torque arm increases the distance between the axle and the point where that force is resisted. The numbers are bigger than most riders expect: a 12mm axle producing 40 N·m exerts a spreading force of just under 1,000 lb on each dropout — which is exactly why motor wattage alone does not describe the mechanical load.
A torque washer is a small tab engaging the dropout — limited local support, heavily dependent on correct seating, and no substitute for an engineered arm on a high-power build. A torque arm extends the reaction point away from the axle and anchors via a clamp, eyelet, bolted plate or custom mount. A torque plate is usually a larger or custom plate fixed around the dropout area, suiting high measured torque, unusual geometry, regen, or dropout faces needing reinforcement. Don’t treat any of these names as proof of strength — construction and fit matter more.
Plenty of guides claim one arm is always enough below some wattage and two are mandatory above it. That is an oversimplification, and it can be wrong in both directions.
Motor input power is voltage × current. Mechanical power is torque × angular velocity. So a low-speed motor can produce substantial axle torque at modest wattage, while a high-speed motor can show a bigger wattage figure without proportionally higher starting torque.
A kit marketed as 1000W may draw well over that under acceleration, and a 3000W motor on a lower-current controller may not load the axle as hard as another 3000W system. Assess these separately: nominal rating, battery current limit, controller phase current, peak electrical input, measured motor torque, and regen torque.
One related myth: axle reaction torque comes from the motor itself, so wheel diameter changes the force at the tyre and the resulting acceleration — but a bigger or smaller wheel never makes a weak dropout safe.
One high-quality torque arm may be sufficient when the complete installation has been technically reviewed and all of the following hold true:
This is why the part matters more than the count. Grin notes one of its thick arms is often sufficient, while thinner third-party arms may need doubling up — evaluate the actual component, not just the motor’s wattage.
Rear dropouts are commonly thicker and better supported than front forks. Even so, a rear motor can still spread shallow or damaged dropouts, and a front-wheel failure causes immediate loss of steering and wheel retention. A front hub needs a fork-specific assessment regardless of wattage.
Steel usually deforms before it fractures, but still needs correct axle and washer fit. Aluminium is more sensitive to local stress and poor washer seating, and may crack rather than visibly spread — it deserves a more conservative assessment. Aluminium and carbon frames should never be cold-set or re-spaced to make a motor fit. Steel does not make a torque arm optional by default.
Two arms can share load and add redundancy — but only when both are installed correctly. Consider dual arms when the build includes any of the following:
| Factor | One engineered heavy-duty arm | Two thin universal arms |
|---|---|---|
| Axle fit | Can be tightly controlled | Often variable |
| Load path | Defined by one mount | Depends on two mounts working |
| Redundancy | Lower | Higher — only if both actually work |
| Installation space | Easier | More difficult |
| Cable-side installation | Usually avoided | May interfere with the cable |
| Suitability | Depends on the verified rating | Not automatically safer |

The motor cable often exits through the axle end. A torque-arm slot must not cut or compress it, and removing connectors to pass an arm over the cable may affect warranty or connector integrity. Cable-side hardware must also leave enough axle-nut thread engagement. Never cut a motor cable during a normal installation.
Acceleration loads the axle one way; regen reverses it. That alternating load causes back-and-forth movement, fastener loosening, dropout fretting and widening axle-plate clearance. Grin’s guidance is to minimise play and use opposed preload or a dedicated clamping design for reversing torque.
Record maximum axle diameter, distance across the flats, flat length available, thread diameter and length, cable-exit position, and shoulder and washer dimensions. “12mm” or “14mm” usually describes the round section — the distance across the flats can differ, and the insert must match the actual flats.
A correct fit gives full contact on the flats, minimal rotational play, no point loading, no hammering the plate on, and no material removed without written approval. Grin’s current designs use hardened splined inserts and publish printable 1:1 profiles so you can test-fit before buying.
Evaluate material grade, heat treatment, axle-slot thickness, resistance to deformation, corrosion protection, fastener grade and any manufacturer test data. Avoid unsupported rules like “every arm must be 6mm” or “aluminium arms are always unsafe” — grade and geometry must be considered together.
| Design | Strengths | Main check |
|---|---|---|
| One-piece plate | Fewer bolted joints; can be frame-specific | Limited adjustability |
| Multi-piece adjustable arm | Fits more frame geometries | More joints that must stay secure; orientation matters |
| Clamping axle block | Reduces axle movement both directions; suits reversing torque | Must match axle dimensions precisely |
The axle insert is only half the system — the frame attachment has to carry the load too. No location is universally strongest; it depends on the frame.
| Mounting area | Potential advantage | Main check |
|---|---|---|
| Chainstay | Often closer to the axle | Tube shape and component clearance |
| Seatstay | Can provide a longer arm | Tube wall and clamp position |
| Dropout eyelet | Clean bolted mount | Eyelet structural capacity |
| Custom dropout plate | Direct load path | Engineering and fabrication quality |
An eyelet may be designed only for mudguards or light racks. Before using one as an anchor, verify thread size, insert material, weld or braze construction, surrounding frame thickness, load direction and manufacturer approval. A clean appearance does not prove structural capacity.
A clamp design can suit some frames, but it must be part of the manufacturer’s approved design, sit on a suitable frame section, avoid brake hoses and cables, avoid crushing thin tubing and stay accessible for inspection. Don’t add rubber under a clamp unless the manufacturer allows it — a soft layer may protect paint but can introduce movement.
Never clamp a torque arm directly to an unsupported carbon tube or dropout. A carbon-frame installation requires motor- and frame-manufacturer approval, structural metal inserts, a designed torque-transfer system and professional installation.

Check the axle reaches the intended dropout depth, both sides sit evenly, the cable is not trapped, anti-rotation washers engage, the wheel is centred and the rotor and gears clear the frame.
The torque arm must mechanically engage the flats. Friction around a round axle surface is not enough.
The mounting must transfer reaction torque into the frame without pulling the axle out of an open dropout, sliding along a tube, rotating through an adjustable joint, loading a weak accessory tab or bending around a sharp corner. There is no universal “always point it forward” rule — orientation changes with front vs rear, left vs right, dropout direction, arm design and regen. Grin documents that reversing one of its earlier designs could encourage the axle to leave the dropout — which is exactly why the product-specific diagram must control orientation.
Free movement lets the axle build momentum and strike the arm before load transfers. Check axle-to-insert movement, adjustable-joint movement, eyelet-bolt clearance, clamp slip and dropout wear.
Seat the wheel and axle hardware, position anti-rotation components, fit the arm without final tightening, align wheel/rotor/gears, tighten axle nuts to the verified spec, secure the frame attachment, then recheck alignment and cable routing — following the sequence approved for your motor variant, not a generic video.
Acceleration applies reaction torque one way; regen reverses it. That alternating load causes axle rocking, slot wear, fastener loosening, dropout fretting, cable movement and washer settlement.
| Feature | Standard sliding arm | Clamping / regen arm |
|---|---|---|
| Main load direction | Primarily one direction | Both directions |
| Dependence on preload | Higher | Lower when correctly clamped |
| Axle play tolerance | Must be carefully controlled | Clamping can reduce movement |
| Regen suitability | Product-dependent | Designed for reversing load |
Grin’s V7 uses a clamping axle interface specifically intended for regenerative-braking applications, while its other models use different anchoring arrangements.
Controller settings change phase current, acceleration ramp, regen strength, reverse torque and peak mechanical load. KirbEbike’s smart controllers and displays are matched by power. Controller current limits vary by kit — check the figure on the product page for your exact variant, and confirm whether it is a battery-side or phase-current rating.
These options are assessed on identical fields — axle size, mounting method, regen suitability, material and main limitation. They are not ranked, and none is a universal winner. Confirm current specifications and stock on each product page before ordering.
| Option | Axle sizes | Mounting | Regen | Material | Main limitation |
|---|---|---|---|---|---|
| Ebike Solution M14 arm (UK) | M14 (14×10mm flats) | Hose clamp to tube/fork | Not specified | 5mm stainless steel | Limited published test data; verify fit |
| Grin V5 | 12 / 14 / 16mm | Bolts to fender/rack eyelet | Not regen-specific | 3mm stainless, hardened 17-4 insert | Depends on eyelet location and strength |
| Grin V6 | 12 / 14 / 16mm | Frame clamp + two hose clamps | Not regen-specific | 3mm stainless, hardened 17-4 insert | Needs an approved tube/clamp arrangement |
| Grin V7 | Separate 12mm / 14mm | Frame clamp + hose clamps | Designed for regen | Hardened 17-4, grade 12.9 bolts | Axle sizing and space must match exactly |
| KirbEbike kit hardware | Per kit variant (135–142mm) | Supplied with high-power kits | Confirm per variant | Confirm per variant | Identify the exact part for your kit |
| Custom engineered plate | Made to the measured axle | Frame-specific mounts | Design-dependent | Specified by the engineer | Needs engineering review and precise fabrication |
KirbEbike supplies installation accessories alongside its motor wheels and conversion kits, and its own guidance flags skipped torque arms and wrong dropout sizing as major risks. Identify the exact arm or plate shipping with your specific 1000W–4000W product — material, thickness, axle compatibility, one-arm/two-arm requirement, installation drawing and replacement route. The high-power kits it ships with include the 52V 2000W MTX, 60V 2500–3000W MTX and 72V 4000W kits.
Designed around the exact frame and axle, using strong existing mounting points, a large contact area, and able to support very high measured torque. Limitations: needs accurate material selection, engineering review and precise manufacturing, must not introduce sharp frame loads, and cannot be judged by appearance.
Torque-arm hardware ships with the high-power kits — matched to the motor, with the axle spec and installation drawing for your exact variant.
Confirm the axle is fully seated, the arm cannot rotate by hand, the mounting point is structurally suitable, axle nuts use verified torque values, rotor and drivetrain run freely, the cable is protected, and fasteners are marked so movement shows up later.
Only after an approved installation: support the driven wheel safely, check rotation at low power, confirm no cable movement, listen for contact and test the brake cutoffs. The widely repeated “hold the brake and blip the throttle” test deliberately applies high load while stationary — not a safe universal check on an installation you are still verifying.
After the first rides, check for changed axle-nut witness marks, fresh metal around dropouts, paint cracking, arm or clamp movement, rotor rub, wheel misalignment and cable twisting. Set ongoing intervals by power level, off-road use, cargo load, regen, frame material and fastener type — not a generic mileage figure.
Do not retighten and keep riding, add washers without diagnosis, fill damaged dropouts with adhesive, fit a second generic arm over the damage, file the dropout further, or exceed the specified axle-nut torque. Have it inspected by the kit supplier, a qualified mechanic or an engineer familiar with hub-motor loads.
Most failures trace back to a short list of avoidable errors:
A torque arm does not increase acceleration or motor torque — it controls the equal and opposite reaction torque acting on the axle. The correct setup depends on actual torque, controller current, axle dimensions, dropout strength, frame material, motor position, regen and the arm’s own design.
One properly engineered arm may be sufficient in some reviewed rear-motor installations. Two arms or a custom plate may be necessary for higher loads, reversing torque or weaker dropouts. Two generic arms are not automatically safer than one tested component, and no torque value, washer order or mounting direction should be copied across unrelated motors. If uncertain, send clear photographs and measurements to your kit supplier before applying motor power — and run your frame through the wheel and dropout fit check first.
A 1000W conversion deserves a torque-arm assessment as a matter of course. Whether it needs one or two depends on actual motor torque, frame material, axle fit, dropout strength, controller current and whether the motor is front or rear.
The axle must be fully seated, the arm must fit tightly over the axle flats, and its other end must anchor to a structural frame point in the manufacturer-approved direction — using the product-specific washer order and torque settings, not a generic sequence.
Not solely because it is labelled 2000W. Many 2000W builds do need dual arms or a custom plate, but base the decision on measured torque, arm strength, dropout design and engineering approval.
There is no single figure for every 3000W motor. Torque depends on winding, wheel speed, controller phase current, voltage and operating conditions — use measured or manufacturer-published torque, not a calculation from wattage.
A 4000W system normally needs a heavy-duty, system-specific solution — dual engineered arms or custom torque plates. Generic thin stamped arms are not adequate without testing.
Not automatically. Two correctly fitted arms share load and add redundancy, but two weak or loose arms may perform worse than one tested heavy-duty arm.
It may. Regen reverses axle torque and can rock the axle in both directions. A tight clamping design, opposed arms or another regen-approved system may be needed.
Stay in the know
Conversion guides, fitting advice and the deals worth knowing about. No spam.
By signing up you agree to our Privacy Policy.
Buyer's GuideApp-Programmable E-Bike Conversion Kits: What You Can Tune and Why It MattersA programmable controller lets you decide how the bike delivers the power it already has. That is a control feature, not a speed feature, and the distinction matters...
Maintenance & RepairHow to Tune a 2000W E-Bike Controller: DC Current, Acceleration Strength and Ride ProfilesTuning a high-power controller is not about finding the maximum. It is about deciding how much current the system can safely supply, then shaping how quickly that current...
Ebike Conversion KitGeared Hub vs Direct-Drive: Which E-Bike Conversion Motor Should You Buy?One mechanical difference explains almost every trade-off between these two motors: a geared hub has an internal clutch, and a direct drive does not. The clutch is why...