Cylinder head rebuild: a micrometer with a zero error, out-of-calibration torque wrench, and reused stretch bolts
Your workshop is rebuilding the cylinder head of a four-cylinder petrol engine. The manufacturer's head bolt procedure is: tighten to 40 N·m, then 90 degrees, then a further 90 degrees, using new bolts. The shop has a 20-100 N·m click-type torque wrench last calibrated 14 months ago (calibration interval 12 months), a 5-25 N·m wrench calibrated 3 months ago, and no torque-angle gauge. A 0-25 mm outside micrometer reads 0.03 mm when the jaws are closed and clean. With it, a technician measures a crankshaft main journal as 47.97 mm; the specification is 47.95-48.00 mm. He also measures the old head bolts at 96.2 mm; new bolts are 95.0 mm and the service limit for reuse is 96.5 mm, so he plans to reuse them to save the customer money. He intends to achieve each 90 degrees by "counting flats" on the bolt head. The customer has approved the gasket replacement and asked you to keep the cost down.
1. Correct the micrometer reading, decide whether the journal is within specification, and state the checks you would make before condemning the crankshaft.
Read the worked answer
Correcting the reading The micrometer reads 0.03 mm with clean jaws closed, so it is reading 0.03 mm high. True dimension = indicated − zero error = 47.97 − 0.03 = 47.94 mm.
Against specification Specification 47.95-48.00 mm. 47.94 mm is 0.01 mm below the minimum, so on this measurement the journal is out of tolerance.
Before condemning it A 0.01 mm judgement on an instrument carrying a 0.03 mm error is not a basis for scrapping a crankshaft. Do this first:
- Clean the anvil and spindle faces, re-zero the micrometer with its adjusting spanner, and verify against a gauge block or a known standard.
- Re-measure at three positions around the journal (90 degrees apart) at each of two depths, using the friction thimble and consistent pressure; check for taper and out-of-round as well as size.
- Check the part is at room temperature - a crank straight out of a hot wash or a hot workshop reads small, and aluminium and steel expand differently.
- Cross-check with a second instrument (a calibrated micrometer or a bore gauge set to the same size).
- Measure all journals, not just one: the correct decision may be grind undersize with matching bearings rather than replace.
Only if a zeroed, cross-checked measurement confirms below 47.95 mm do you record the crank as out of service. Whatever the outcome, record the zero error and the correction on the inspection sheet - an unrecorded instrument error will mislead the next job.
2. Evaluate the torque wrench and torque-angle plan and state what you would do before the head goes on.
Read the worked answer
Torque wrench The 20-100 N·m wrench is the only one that covers 40 N·m (the 5-25 N·m wrench is out of range), and it is 14 months past a 12-month calibration interval. It is not fit to be used for a cylinder head. Tag it out, send it for calibration, and borrow or hire a calibrated wrench in the meantime. Using it would leave a 40 N·m preload that could be 10% or more out, which the 90-degree steps cannot correct.
Torque-angle step Counting flats is a recognised field method, but it must be done properly:
- A hex head has six flats, so one flat is 60 degrees. 90 degrees = one and a half flats.
- Mark the bolt head and the head casting with a paint pen before starting, so the mark shows true rotation and not a slip.
- Apply the angle in one continuous sweep with a breaker bar - never stop and re-strike, and never add the two 90-degree steps together into one 180-degree turn.
- A torque-angle gauge or a torque-angle wrench is more accurate and should be the first choice; counting flats is the fallback, and only with clear marks and a second person checking.
Other prerequisites Bolt holes chased and blown out, threads clean and lightly oiled exactly as specified, the gasket fitted dry or as specified, the correct torque sequence from the centre outwards, and the angle steps applied in that same sequence. Note both the final torque and the angle completion on the job card. Explain to the customer that the angle steps, not the torque figure, set the final clamping load - this is where head gasket comebacks come from.
3. Should the old head bolts be reused? Weigh the 96.2 mm measurement against the customer's cost request and state the decision and conditions.
Read the worked answer
The measurement Old bolts measure 96.2 mm; new 95.0 mm; the reuse limit is 96.5 mm. So 96.2 mm is 0.3 mm under the limit and, on the length criterion alone, the bolts pass.
Why length alone is not enough These are torque-to-yield bolts: they are designed to stretch plastically during the angle steps and hold the clamp load by tension. Once stretched, they have less reserve, and length does not detect every failure mode. If reuse is permitted at all, it must be on the manufacturer's terms, with additional checks:
- Shank diameter measured with a micrometer and compared with a new bolt - the necked region must not be measurably thinner.
- Visual check for necking, thread damage, corrosion and fretting under the head.
- Bolts that have been used more than the OEM's stated number of times, or that have been over-torqued or overheated, are scrap regardless of length.
Decision If the OEM procedure says "use new bolts" - as many do for this fastener type - fit new bolts, and say so clearly. If the OEM permits reuse when under the limit, then reuse is legitimate only after the diameter and visual checks pass, and only with clean, lightly oiled threads and the correct sequence.
Handling the cost request Show the customer the price of a set of bolts against the price of a repeat head gasket job - strip-down, gasket, oil, coolant, machining and a second labour charge, typically many times the bolt cost. State plainly that the bolts are cheap insurance on a job where a comeback is expensive. If the customer still refuses new bolts, record their instruction, note the reduced likelihood of a successful repair, and do not offer a warranty on the head gasket.