Electric Dirt Bike Bolt Torque: The Numbers That Actually Hold the Machine Together
Not one of our four product pages publishes a fastener torque value. We looked while writing this: the number holding your front wheel on is printed nowhere on the page you bought from. So this article derives the loads your fasteners actually see, gives torque ranges that match published industry references, and explains the two mechanisms that make bolts come loose — only one of which is vibration.
A hub motor makes the axle part of the drivetrain
On a hub-drive off-road electric bike there is no chain path between motor and ground. The wheel is the motor, so every newton-metre of drive torque is reacted by the axle, the dropouts and the two nuts clamping them. Nothing in between absorbs or reduces it.
Drive torque is power divided by speed for force at the contact patch, multiplied by wheel radius. Using the power figures and wheel sizes on our own product pages:
| Model | Tire size | Power in the sum | Axle torque at 10 mph | At 20 mph |
|---|---|---|---|---|
| ED-SN01 | 26 × 4 | 4,000 W combined | 318 N·m | 159 N·m |
| STK-01 | 60/100-14 front, 80/100-12 rear | 1,500 W rated | 93 N·m | 47 N·m |
| ED-S01 | 26 × 4 | 1,500 W | 119 N·m | 60 N·m |
| T300 | 14 × 2.5 | 2,000 W peak | 108 N·m | 54 N·m |
Now compare that with the axle-nut figure in the standard reference: 29.4–44 N·m (266–390 in-lb) for a solid axle. The nut is tightened to a fraction of what the axle underneath it is fighting — two times over on the T300, ten on the ED-SN01 — and the low-speed column is the worst case, not the top speed.
Why the axle nut cannot hold that on its own
Torque on a nut is not a measure of grip. It stretches the stud to create clamp load, and clamp load is what generates friction between the dropout faces. A dry M12 nut at 35 N·m produces about 14.6 kN of clamp; two dropout faces at a friction coefficient of 0.10–0.20, over an effective radius of roughly 20–25 mm, resist about 60–120 N·m of rotation.
Friction alone gets you to roughly 120 N·m; the lineup puts 93–318 N·m through the same axle at 10 mph. Which is fine: the nut was never the part doing the work. The hardware that actually stops the axle spinning is the keyed slot, the anti-rotation washer, or a torque arm — a shape, not a fastener. If you lose that washer, no torque wrench will save the dropout, and a dropout that has spun once is already damaged. Never grease the dropout faces to quiet a creak.
The torque table to use when the manual is silent
These come from Park Tool's published torque reference and standard fastener engineering. They are ranges for the joint type, not a substitute for a model-specific spec.
| Joint | Fastener | Torque | Note |
|---|---|---|---|
| Axle nut, solid axle | M10–M14 | 29.4–44 N·m | Re-check at 50 miles |
| Brake rotor to hub | M5 | 2–6.2 N·m | Brand-dependent; start low |
| Caliper body mount | M6 | 5.7–12.4 N·m | Often aluminium threads |
| Stem faceplate, 4-bolt | M5–M6 | 5.4–13.6 N·m | Tighten in a cross pattern |
| Seatpost collar | M5–M6 | 4–6.8 N·m | Enough to stop slip, no more |
| Pedal into crank | 9/16 in | 35 N·m minimum | Left side is reverse thread |
| Generic steel bolt, class 8.8 | M6 / M8 / M10 | 10.5 / 25.5 / 50.5 N·m | 75% of proof load, dry |
With no spec at all, we recommend the bottom of the range and a re-check after the first ride. Under-torquing is recoverable in a minute. Pulling the threads out of an aluminium caliper mount is a helicoil, a workshop visit and a week off the trail.
Torque measures friction, not tightness
Every published torque value is really a clamp-load target converted through a friction coefficient: torque equals the nut factor, times nominal diameter, times the clamp force you want. For a dry zinc-plated fastener the nut factor sits near 0.20; with anti-seize on the threads it collapses toward 0.10. Same wrench setting, roughly double the clamp load you asked for.
Engineering Toolbox quantifies the effect from the other direction: graphite cuts required torque by 50–55%, white grease and SAE 30 oil by 35–45%, and their worked example ends with a blunt warning — apply a dry spec to a lubricated bolt and the fastener may overload and break.
This is why "a dab of copper grease on the threads" is the most expensive habit in home mechanics. Dry means clean and dry, and a bolt that came out oily goes back in clean.
You need two wrenches, not one
A rotor bolt at 2–4 N·m and an axle nut at 44 N·m cannot come from the same tool. A 10–60 N·m wrench asked to deliver 3 N·m sits at 5% of its scale, where the reading means very little; a 2–14 N·m wrench asked for 44 N·m is out of range.
Park Tool's own line splits the same way: 2–14 N·m and 10–60 N·m click wrenches, plus a 4–6 N·m limiter for the small stuff. Their care notes matter more than the model numbers. Calibrate every 5,000 clicks or once a year. Wind a click wrench down to its lowest setting for storage, not to zero. Treat head deflection, not the noise, as the signal — at low settings it may not click audibly at all, which is exactly where rotor bolts live. Never break a fastener loose with it.
The bolt that keeps coming loose is not being shaken loose
Park Tool's reference states plainly that the common cause of threaded fasteners loosening is a lack of tension during initial assembly, and that vibration cannot typically overcome the clamping force of a properly sized and secured fastener. Bolt Science's testing points the same way: axial loading alone does not produce significant self-loosening, and the mechanism that does is transverse slip in the joint. In a Junker test on an M8 nylon-insert nut at ±0.65 mm of transverse movement, the nut still rotated before the prevailing torque caught it and left a residual preload.
So when the same bolt is loose again two rides later, work through causes instead of reaching for the wrench:
- It was never at spec. Hand-tight is a fraction of the clamp load you think you applied.
- It was over-torqued once and yielded. A stretched bolt holds nothing, and it will keep backing out.
- A washer is missing, flattened or split. Common at dropouts and caliper mounts.
- The joint is crushing. Plastic fenders, battery cradles and composite brackets relax under load and take preload with them; that joint needs a schedule, not more torque.
- The thread is damaged or the bolt is too short. Fewer than five engaged threads in aluminium is a stripped hole waiting to happen.
Our position: "check every bolt before every ride" is theatre. A correctly preloaded fastener does not move, and hunting forty bolts every morning trains you to skim. Check the eight that carry load, and investigate any fastener that moves twice.
The eight fasteners worth your time
- Front and rear axle nuts. The only fasteners that react motor torque. Torque to spec, confirm the anti-rotation hardware is seated, and never substitute a plain washer.
- Brake rotor bolts. Six small M5 fasteners per wheel, 2–6.2 N·m. A loose rotor pulses under braking long before it fails; our hydraulic disc brake guide covers the sequence, and replacement rotors are on the brake disc page.
- Caliper mount bolts. 5.7–12.4 N·m into a bracket that is often aluminium. This is the joint most often stripped by enthusiasm.
- Stem faceplate bolts. Four bolts, cross pattern, small increments. Uneven torque here walks the bar in the clamp.
- Battery cradle and latch hardware. A 1,664 Wh pack that shifts under impact is an electrical fault looking for a moment; check the mount, not just the connector.
- Spoke tension, not spoke torque. Spokes are tuned by deflection, not a wrench setting — our wheel and hub guide covers the sequence, and spare spokes are stocked.
- Pedals and crank hardware. 35 N·m minimum at the pedal, and the left side is reverse-threaded — the one fastener people cross-thread at 11 p.m.
- Fender, rack and light mounts. Low torque, high consequence: a fender bolt into the tire at speed is a locked front wheel.
Run the full pass at 50 miles, then every 20–25 riding hours, and after any wheel removal or pressure wash. The pre-ride checklist covers the two-minute version you can do at the trailhead.
Frequently asked questions
Where are EDIKANI dirt bikes designed to be ridden?
EDIKANI off-road electric bikes and electric dirt bikes are designed for trails, off-road parks, and private property. Always check and follow your state and local laws before riding.
What should the axle nuts be torqued to?
29.4–44 N·m (266–390 in-lb) for a solid axle, with the anti-rotation hardware fitted and seated. If the nut was loose enough for the axle to spin, inspect the dropout slot before you re-torque anything.
Do I need threadlocker on every bolt?
No. Use medium-strength removable threadlocker on the small fasteners where the manual calls for it, and never as a substitute for correct torque. Threadlocker resists rotation; it does not create clamp load.
How often should I re-torque?
At 50 miles, then every 20–25 riding hours, and after any wheel removal, pressure wash or crash. A fastener that moves twice on that schedule has a cause worth finding, not a habit worth repeating.
Start with the two that hold the wheels on
If you buy one thing after reading this, buy the 2–14 N·m wrench — it covers the fasteners you are most likely to damage and least likely to own a tool for. Then torque the axle nuts to spec, confirm the anti-rotation hardware is there, and leave the rest alone until the 50-mile check. Want the torque values for your exact model and serial? Ask us; we would rather send a table than a replacement dropout. The full lineup of off-road electric bikes and electric dirt bikes runs from the T300 to the dual-motor ED-SN01 and the STK-01.
Sources: Park Tool, Torque Specifications and Concepts and Torque Wrench Use and Care; Bolt Science, Loosening Characteristics of Prevailing Torque Nuts; Engineering Toolbox, metric bolt tightening torques and lubricated bolts and reduced torque. Drive-torque figures calculated from the power and wheel sizes on each EDIKANI product page.
