Incoming inspection is the last gate before electronic screws reach the line. Check everything and every lot needs unpacking, measuring and recording, which strangles warehouse throughput; check too little and problems surface at assembly as cam-out, floating heads or incomplete seating, where diagnosis costs more. An effective IQC plan puts limited resources on the items most tightly linked to assembly failure and writes acceptance rules precisely enough for both sides to reproduce. This article covers item selection, sampling and dispute handling.
Test items should not be copied from a standard index but derived from your own failures. Pull the past year of screw-related defects, rank them by frequency and impact, then work backwards to what incoming inspection must intercept. If cam-out and bit slip dominate, recess depth and head hardness matter; if floating heads and marks are frequent, head flatness and bearing-face burrs deserve attention; if corrosion or appearance complaints dominate, plating thickness and colour consistency belong on the list. Fewer, sharper items also tell inspectors what they are guarding against.
Dimensional items usually cover head diameter and thickness, shank length, major and pitch diameter, and recess size. Threads are checked mainly with go/no-go gauges: the go gauge must run through freely and the no-go gauge must not enter beyond the specified turns; for tapping or thread-cutting screws, crest sharpness and the integrity of the cutting slot also matter. State the datum explicitly - whether length includes the head, from which face thread length is measured, how many points are averaged for head thickness. When two labs disagree on one lot, the datum is usually the difference, not the product.
Sampling plans should be graded by lot size and application. Critical dimensions and functional items generally take a tighter inspection level; appearance can be looser. Two practical points: draw samples from several packing units rather than only the top carton, and record both sample size and defect count so trends can be analysed later. For stable part numbers supplied continuously, skip-lot inspection is reasonable - provided history is long enough and the process is in control.
Plating invites dispute because results vary with measurement location. Agree on three measurement points - thread crest, plain shank and head bearing face - and state the method; judge adhesion qualitatively by bend or tape test; compare colour under a standard light source against the approved sample rather than under workshop lighting. Customers with halogen or RoHS restrictions need periodic third-party reports from the supplier, with incoming inspection confirming that report and batch numbers correspond.
Full material and mechanical testing at goods-in is unrealistic. A workable split is document review plus periodic testing: verify material certificates and heat-treatment records per lot, and run hardness, tensile or embrittlement-related checks quarterly or whenever the supplier changes a process. For thin-wall high-strength parts, baking records must be traceable per batch, because hydrogen embrittlement failures are delayed - by the time a part fractures in assembly, the window for remedy has closed. Writing the test interval into the inspection specification avoids last-minute conflicts with delivery.
Judgement rules should separate critical, major and minor defects: anything affecting function or safety is critical and rejected outright; slight dimensional deviation that can be reworked or used under concession goes through a written concession stating the permitted use; minor cosmetic issues are judged against limit samples. In a dispute, exchange measuring methods and raw data first, then re-measure on the same instrument with the same datum - most disagreements converge at that point. Keeping limit samples and methods as part of the approval file is the cheapest way to avoid arguments.
Incoming inspection of electronic screws is less about strictness than stability: items that map to real failures, sampling that covers real risk, and criteria both sides can reproduce. We ship inspection records and measured values for critical dimensions with each batch and help customers align on datums, so IQC becomes shared control rather than mutual suspicion. Accumulated data can then feed back into process improvement at the supplier - the long-term value of incoming inspection.