The Micrometer That Reads Twenty Thousandths Low
A 4-cylinder petrol engine is stripped for a bearing knock and the crankshaft is on the bench for measurement. The manual gives main journal diameter 52.000–52.020 mm, maximum out-of-round 0.020 mm and maximum taper 0.020 mm. The only micrometer in the right range is a 50–75 mm model. Checked against its 50.000 mm setting standard, it reads 49.980 mm. The crankshaft came out of the parts washer ten minutes ago and still feels warm; shop temperature is 32 °C. You measure each journal at 0° and 90° and record: No.1 51.988/51.991; No.2 51.972/51.958; No.3 51.995/51.993; No.4 51.982/51.985. A colleague looks at the figures and says the crank is worn out and must be ground 0.25 mm undersize, which means ordering bearings and a two-day wait. The service counter wants an answer within the hour.
1. State the micrometer's zero error, apply it, and give the true diameter and out-of-round for each journal with a pass or fail against the specification.
Read the worked answer
Zero error The micrometer reads 49.980 mm on a 50.000 mm standard, so it under-reads by 0.020 mm. Correction: true size = reading + 0.020 mm. Before relying on this, confirm the error is the same higher up the range, for example with a gauge block or a second standard near 52 mm; a zero error set at 50 mm is not guaranteed to be linear across 25 mm of travel.
Corrected results (spec 52.000–52.020, out-of-round max 0.020)
| Journal | 0° | 90° | Out-of-round | Result |
|---|---|---|---|---|
| No.1 | 52.008 | 52.011 | 0.003 | Pass |
| No.2 | 51.992 | 51.978 | 0.014 | Fail - undersize 0.022 on one axis |
| No.3 | 52.015 | 52.013 | 0.002 | Pass |
| No.4 | 52.002 | 52.005 | 0.003 | Pass |
Out-of-round is within limits on every journal; the failure on No.2 is size, not ovality. Taper across the length also has to be checked, not only two planes at one station.
2. On the corrected figures, what do you recommend, and what must you check before committing to a two-day regrind?
Read the worked answer
Recommendation The colleague was working from uncorrected numbers, which make every journal look undersize. On corrected figures only No.2 is out of specification, so the hour should be spent confirming that measurement rather than ordering parts on it.
Checks before any regrind
- Temperature. The crank is warm. For 50 mm of steel, expansion is roughly 11.5 × 10⁻⁶ per °C, so 15 °C above reference is about 0.009 mm. Warm parts read large, which means the journals now passing by 0.002–0.008 mm may be borderline or below minimum when cold. Let the crank stabilise, or apply a correction, and re-measure.
- Second instrument and operator. Re-measure No.2 with a calibrated micrometer, cleaning the anvils first and using the ratchet stop.
- Grinding economics. If No.2 truly fails, the practical repair is usually to grind all journals to the next undersize so one bearing size fits and the crank stays balanced. Confirm that size is available, that it stays above the manufacturer's minimum diameter, and check fillet radii, thrust faces and runout.
- History. Check the bearing shell markings; the crank may already be undersize, which changes what 0.25 mm means. Only after those checks should the customer be told the crank must come out of the schedule for two days.
3. Name two other measurement errors that could mislead you here and explain how to control each.
Read the worked answer
1. Technique error - misalignment and clamping force If the micrometer is not square to the journal axis, the reading is a chord rather than a diameter (cosine error). Over-tightening the thimble squeezes the anvils together and under-reads; a slack grip leaves a gap and over-reads. Control: use the ratchet stop with two or three clicks, hold the frame square to the axis, and measure at three stations along the journal in two planes at 90°, recording every value rather than the one that looks right.
2. Contamination and handling An oil film, grit or a burr on the anvil adds thickness; holding a steel frame in a warm hand for minutes expands it against the setting standard. Control: wipe the anvils and the journal with lint-free paper, zero-check immediately before and after the set of readings, and use a micrometer stand or an insulated grip for long measuring sessions.
A third trap worth mentioning is reading the worn band only - journals wear oval and tapered, so a single reading at the top of the journal can pass a crank that fails at the bottom. Recording all values and comparing them to the specification is the control, and it is also what makes the measurement defensible if the customer disputes the repair.