Wheel-hub bearing fit measured in a hot workshop
A main landing-gear wheel hub is being rebuilt. The aluminium-alloy hub bearing bore is called out on the drawing as Ø120.000–120.035 mm and the bearing outer diameter as Ø120.045–120.060 mm; the required interference fit is 0.010–0.045 mm, all at 20 °C. The hub has stood in the heated workshop all morning and is at 38 °C when the bore is measured with a calibrated internal micrometer: 120.028 mm. The bearing, kept in the temperature-controlled store, is at 22 °C with an outer diameter of 120.052 mm. Coefficients of linear expansion are 23 × 10⁻⁶/°C for aluminium and 11 × 10⁻⁶/°C for bearing steel. The fitter proposes to press the bearing in as measured, arguing the figures are 'within a couple of hundredths of a millimetre' of the drawing.
1. Correct both measurements to 20 °C, calculate the true interference, and state whether the assembly may be built.
Read the worked answer
Bore (aluminium, 38 °C → 20 °C, cooling):
Δ = 120 mm × 23 × 10⁻⁶/°C × (38 − 20) = 120 × 23 × 10⁻⁶ × 18 = 0.0497 mm
Bore at 20 °C = 120.028 − 0.0497 = 119.978 mm
Drawing limits are 120.000–120.035 mm, so the bore is 0.022 mm undersize - it is already out of tolerance on its own.
Bearing (steel, 22 °C → 20 °C):
Δ = 120 × 11 × 10⁻⁶ × 2 = 0.0026 mm
OD at 20 °C = 120.052 − 0.0026 = 120.049 mm - inside 120.045–120.060 mm, acceptable.
Interference = 120.049 − 119.978 = 0.071 mm against a permitted 0.010–0.045 mm.
Decision: reject. The fit is 0.026 mm tighter than the maximum and the housing bore does not conform to the drawing. The 'couple of hundredths' the fitter is dismissing is the whole tolerance band - thermal correction is not a refinement here, it decides the answer.
2. The fitter says it will 'press in fine'. Explain what a 0.071 mm interference does to this assembly, and what corrective actions are and are not permitted.
Read the worked answer
Consequences of the over-tight fit:
- Hoop stress in the aluminium hub rises roughly in proportion to interference - about 60 % above the maximum designed value, with a real risk of cracking the hub at the bore or at the flange radius.
- The bearing is squeezed: radial internal clearance is reduced by roughly 80 % of the interference, i.e. about 0.057 mm. A bearing with, say, 0.030 mm of internal clearance is taken into preload, giving raceway brinelling, high running temperature and early spalling.
- Raceway distortion also destroys the geometry the bearing was ground to, so the eventual failure will look like a bearing fault when the true cause is the housing.
- If it must come out later, a 0.071 mm fit usually means cutting the bearing off, risking the hub.
Permitted actions:
- Quarantine the hub, re-measure in a 20 °C environment with a setting ring, and confirm the reading at several depths and two planes 90° apart.
- Check the bearing part number - an oversize-OD bearing may have been drawn from stores by mistake.
- Refer to the approved repair scheme / SRM for an oversize bearing or a replacement hub.
Not permitted: reaming or scraping the bore 'to suit', since the bore is already below the minimum limit and any material removal needs approved data; heating the hub or freezing the bearing beyond the manual's stated limits to force the fit.
3. How should the bore measurement itself have been taken, and why is a calculated correction a check rather than a production method?
Read the worked answer
- Parts and instrument must be at the same, stable temperature, and the geometry must be allowed to soak (as a guide, roughly an hour per 25 mm of section, or as the manual states) - a hub that has stood in a hot workshop is not at one uniform temperature throughout its section.
- Calibrate the micrometer against a setting ring at the same temperature, and re-check zero after the measurement.
- Use a three-point internal micrometer or bore gauge set from the ring; take readings at 90° in two planes and at three depths, and record the maximum and minimum.
- Measure in a controlled 20 °C environment wherever possible. The arithmetic correction using the expansion coefficient assumes a known, uniform temperature and a known coefficient for the exact alloy; it validates a reading, it does not licence building to a hot measurement.
- Record temperatures and instrument identification on the paperwork so the next person can see how the conforming figure was obtained.