When a gear shipment arrives, the inspection report that rides with the packing list is more than paperwork. It is the supplier's sworn measurement of the parts you paid for, and it is the document your quality team will use to decide whether the gears go straight to the gearbox assembly line or back onto the truck. A buyer who can read that report with confidence can accept good parts quickly, reject bad parts before they cause a line stoppage, and hold the supplier to the accuracy grade that was agreed on the drawing. A buyer who treats the report as a box-ticking exercise inherits every measurement dispute as a production problem. This guide walks through what each field on a typical gear inspection report actually means, how to read the CMM data, where the acceptance line sits, and which values should make you pause before signing the release.
The first block of values on any gear report is the tooth geometry itself, measured on a gear measuring center. Tooth profile — the involute curve across the tooth flank — is reported as a deviation in micrometres, usually split into total profile deviation and form deviation. For an ISO grade 7 commercial gear, total profile deviation might sit around 11 to 16 micrometres; grade 6 halves that band. A profile that runs outside the allowed tolerance means the involute is worn or the grinding wheel was dressed off, and the consequence is noise and pitting that shows up within the first hundred hours of operation. Tooth lead — the straightness or helix angle along the face width — is reported the same way, and a lead that deviates beyond tolerance means contact is concentrated at one end of the face, causing edge loading and uneven wear. Radial runout, measured as the total indicator reading when the gear is rotated on its bore, tells you how concentric the teeth are with the datum; a runout that exceeds the class tolerance introduces pitch error every revolution and drives vibration at the mesh. Backlash is reported either as a measured range or as a derived tooth-thickness reduction, and it is checked against the range agreed on the drawing — too little and the gears bind when the oil warms, too much and the drive hammers on every reversal. Tooth thickness, measured over pins or by span, is the controlling dimension that sets backlash, and it must land inside the specified band rather than merely near the nominal value.
The second block is the coordinate measuring machine data, and this is where buyers most often feel lost. A CMM report lists a coordinate system, usually tied to the bore axis and one locating face, and then reports each controlled feature as a measured value next to its nominal and its tolerance limits. Bore diameter, perpendicularity of the locating face to the bore axis, flatness of the datum, keyway width and position, and any geometric callout on the drawing are all listed this way. The rule for reading CMM data is simple: look at the measured value, not the pass stamp, and compare it to both the nominal and the two tolerance limits. A bore that measures at the high end of its tolerance on a warm day may still be inside the band on paper, but your assembly engineer will feel the difference when the gear is pressed onto a shaft sized to the nominal. True position and perpendicularity values are reported as a diameter zone, and a reading that is near the outer limit is a warning, not a pass. The CMM report should also state the probing method and the temperature at measurement, because a coordinate value taken on an unregulated floor drifts with the room.
Acceptance is decided against the accuracy grade written into the purchase order, not against a generic standard. If your drawing calls out ISO 1328 grade 7, then the profile, lead, pitch and runout tolerances from that grade are the acceptance band, and every measured value on the report must fall inside it. Values that land inside the band are accepted as a matter of routine; values that fall outside are rejected, and there is no middle ground on a precision gear. Where the drawing specifies a backlash range or a tooth-thickness band, the measured values must sit inside that range, not merely at one edge. Heat-treatment data — surface hardness, case depth and core hardness — are accepted against the purchase specification the same way, and a mill certificate that shows a chemistry slightly off the nominal but within the material standard is accepted only if the order allowed that standard. The practical test is this: if you cannot trace every controlled characteristic on the report to a clause on the drawing or a line in the purchase agreement, ask the supplier to explain it before you sign.
Several values deserve a closer look because they are early-warning signals rather than simple pass or fail. A tooth-profile deviation that consistently shows a pressure-angle error — the whole involute is tilted rather than locally dented — usually means the hobbing machine was set with the wrong hob pressure angle, and this is a process problem that will repeat on the next batch. A lead that twists, with contact shifted from one end of the face to the other along the width, points to a grinding setup that was not dressed straight, or to heat-treatment distortion that the finish pass did not fully correct. Radial runout that is only slightly over tolerance on every gear in a lot suggests the bore datum is not being controlled the way the assembly uses it, which is a fixture problem at the supplier and a sign that future lots will drift the same way. Hardness readings that vary by more than a few HRC across a single gear are a heat-treatment red flag, and a CMM report where every measured value sits at the same edge of the tolerance band is usually a sign the operator stopped the machine as soon as the part cleared, rather than a sign of a stable process.
Reading the report is the first half of the job; acting on it is the second. Build a short acceptance routine: confirm the lot number on the report matches the lot on the parts, scan the profile, lead, runout, backlash and tooth-thickness rows for any value outside the tolerance band, check the CMM datum features against the drawing, and verify that hardness and material certificates are attached and signed. If a value is borderline, flag it for the assembly engineer rather than accepting it on sight. When a value fails, open the conversation with the supplier by quoting the row on the report and the clause on the drawing, not by sending the whole shipment back. A supplier that can explain the out-of-tolerance value, show you the measuring equipment calibration, and propose a correction for the next lot is a supplier worth keeping; one that answers only with a revised report is not. Send us your drawing and the accuracy grade you need, and we will quote an inspection plan and a report format your quality team can read on the first pass. Our quality system, the measuring room behind it, and the gear-inspection and tolerance references on this site give you the background to check our work, and our engineers are glad to walk through an existing report with you before the next order.