Torque Wrench, Crowfoot Adaptor and a Dropped Tool
An AMM task specifies 45 N·m for a bolt on a landing gear actuator mount. The shop torque wrench has an effective length of 380 mm from the centre of the square drive to the hand-grip index mark. Its calibration certificate, issued 10 months ago against a 12-month interval, states an accuracy of ±4 % of the indicated value. Access requires a 65 mm crowfoot adaptor fitted in line with the handle, extending the effective length. The AMM tolerance on the fastener is 45 N·m ±3 %. Earlier in the shift the same wrench was used to undo two stiff bolts, and it was knocked off the bench onto the hangar floor between tasks.
1. Calculate the wrench setting for the in-line adaptor and the fastener torque range that could result, given the stated accuracy. Does the setup demonstrate compliance with the AMM? What changes if the adaptor is turned through 90°?
Read the worked answer
Setting with an in-line extension
M_set = M_required × L / (L + E) = 45 × 380 / (380 + 65) = 45 × 380/445 = 38.4 N·m
The adaptor increases the lever arm, so the wrench must be set lower to deliver 45 N·m at the fastener.
Resulting torque at the fastener
Actual = 38.4 × 445/380 = 45 N·m nominal, but the wrench reads within ±4 %, so:
- Low: 45 × 0.96 = 43.2 N·m
- High: 45 × 1.04 = 46.8 N·m
Compare with the AMM
Permitted = 45 ± 3 % = 43.65 to 46.35 N·m
The permitted window is ±1.35 N·m; the instrument uncertainty alone is ±1.80 N·m. The uncertainty is larger than the tolerance, so you cannot demonstrate compliance with this wrench, even with perfect technique. Correct action: obtain a wrench of sufficient accuracy (for example ±2 % or better) with a current certificate, or use a torque analyser/torque transducer to verify the setting, and target the mid-tolerance value.
90° orientation
With the crowfoot at 90° to the handle the offset acts perpendicular to the applied force, so the lever arm is unchanged: set 45 N·m directly, no correction. Getting this wrong is the classic error - correcting a 90° adaptor under-torques the fastener by about 15 %.
2. What must be done about the wrench that was dropped, and what other practices in this scenario are unacceptable?
Read the worked answer
The dropped wrench
- Remove it from service immediately: red tag it, record it, quarantine it. A drop can bend the beam, shift the mechanism, or leave the pointer/scale reading plausibly but wrongly - the biggest risk is a tool that looks fine.
- Return it for recalibration/verification before any further use. The certificate interval being valid is irrelevant; an impact event invalidates the assumption behind it.
- Do not 'check it by feel' against another wrench.
Unacceptable practices in the scenario
- Using a torque wrench to undo fasteners. It is a tightening instrument, calibrated in the direction and range of tightening; breakaway torque is higher and outside its design use. Use a breaker bar or suitable wrench.
- Applying a ±4 % instrument to a ±3 % tolerance without a metrological check.
- Working without the adaptor correction documented on the task card - the calculation should be traceable, not done from memory.
- Not recording the tool's serviceability status after the drop, which would leave the next technician to assume it was good.
Good practice: pre-task check of the tool's certificate, range and condition; the adaptor correction written on the task card and cross-checked; calibration recall visible at the point of use.