Most factory audits are document reviews with a walk-through attached. For assembly businesses that is sufficient. For composite helmet manufacturing it is not, because the safety performance of a shell is created by process control rather than by incoming inspection. A shell that passes impact testing in the sampling round can fail a year later if the press cycle drifts, the fiber orientation changes, or the resin batch quietly changes source.
The twelve checkpoints below are ordered by information gained per hour on site. They assume the buyer has already received the document pack and now wants to know whether those numbers describe the factory or the brochure.
An audit answers three questions: can this factory make the product repeatably, can it prove that it did, and will it tell you when something changes. Everything else — showroom, sample room, hospitality — is noise.
Bring a scoring sheet, photograph records rather than machines, and interview the person who runs the press, not only the sales manager. A factory that runs motorcycle helmets on one line and bicycle shells on another should be audited line by line, because the process windows are not the same.
Ask who owns the shell mould, where it is stored, and what its maintenance interval is. A factory that cannot produce a mould drawing, a steel certificate or a maintenance log is usually subcontracting the shell itself.
Inspect the cavity for polish quality, parting-line wear and repair welds. A worn parting line produces flash that gets sanded away, thinning the shell edge. That defect never appears in a dimensional report, but it does appear in an edge-impact test.
Look at the quarantine area and the incoming inspection record, not the datasheet binder. You want lot numbers, filament diameter checks, sizing chemistry declared per lot, and moisture readings taken at goods-in rather than copied from the supplier certificate.
For resin, ask for batch viscosity and gel-time records plus the storage temperature log. Resin held above its recommended range for a few weeks will still pass a certificate check and still ruin a cure.
Watch a full layup. Ask the operator how the ply book defines orientation, how many plies sit at the crown, and what happens when a ply is cut short.
Then ask to see a rejected or reworked layup. Factories that lay up from memory rather than from a numbered ply book cannot hold orientation across shifts, and in a thin composite shell orientation is one of the largest levers on impact performance.
Every press or autoclave should log temperature, pressure and time per cycle, tied to a work order. Ask for one complete log from the previous week and check whether the ramp and dwell match the approved recipe.
Then ask what happens when a cycle alarms out mid-dwell. The correct answer is a documented deviation with a disposition decision. The answer you often hear is that the part was checked by eye, which means the factory has no defined cure-critical failure mode.
On a bicycle helmet the liner manages most of the impact energy, and it is frequently the least audited area. Check bead density verification, steam pressure and cycle time records, and how the factory checks fusion at the liner surface.
Then check liner-to-shell fit: gap, adhesive coverage, and whether the shell sits under load or drops in loose. A shell bonded loosely to its liner can pass a flat-anvil test and still behave unpredictably on a kerbstone anvil.
Ask whether the factory owns a drop rig, and if so, when it was last calibrated and by whom. A factory without one depends on external laboratories for every iteration, which lengthens development and hides process drift.
Ask to see a recent internal report against the relevant standard — EN 1078 for pedal cycles or ECE 22.06 for motorcycle use — and check that it states conditioning regime, anvil type, drop height and impact location for each test point.
Strap anchoring is a common field failure and a fast audit item. Inspect the anchor geometry inside the shell, the stitch pattern on the webbing, and the pull-test record for the assembly.
Ask for the dynamic retention record. The roll-off test tells you quickly whether the factory understands that a helmet has to stay on the head, not merely survive an impact.
Coating is where composite shells most often lose performance after demoulding. Check surface preparation, primer compatibility with the resin system, and whether the factory measures film thickness.
Then look at solvent handling and oven temperature records. A water-based system applied in a humid bay without adequate flash-off traps solvent under the clear coat and crazes within a season.
Pick a finished helmet from the warehouse and ask the factory to walk backwards to fiber lot, resin batch, liner batch, press cycle and operator. Do it with two helmets from different months.
If the walk-back stops at production week, the factory is not traceable in any sense a recall would need. Custom Carbon Fiber Motorcycle Helmet programmes with several shell sizes and finishes fail this check most often, because the work order covers a family rather than a unit.
Ask for the written change-notification clause and for the last three notifications the factory actually issued to customers. Then ask which sub-tier suppliers changed in the past year.
Mineral source, sizing package and resin supplier are the three changes most likely to invalidate a certification file. A factory that treats them as purchasing decisions rather than engineering changes will hand you a quiet non-conformance.
Ask for takt time on one line, the number of presses, the shift pattern and current loading. Then compare the quoted lead time against that arithmetic.
Factories quoting from an empty order book quote fast and deliver slowly. The audit question is not whether capacity exists, but whether the factory will reserve it for you in writing.
Ask for the rework log and the scrap trend over the last six months. Then ask to see minutes from the most recent failure review meeting.
A factory with no rework log is not a factory without rework; it is a factory that decided not to record it. Rework on a structural shell — sanding, filling, re-bonding — must be defined, limited and recorded, or it becomes a silent design change.
Score each checkpoint from zero to three and treat anything at one or below as a supplier development action rather than an automatic rejection. A realistic distribution of weight looks like this.
| Checkpoint group | What a strong factory shows | Weight |
|---|---|---|
| Mould and layup (1, 3) | Owned tooling, numbered ply book | High |
| Material (2, 10) | Lot-level control, written notices | High |
| Cure and liner (4, 5) | Cycle logs, fusion checks, bonded fit | High |
| Test and assembly (6, 7) | Calibrated rig, retention records | Medium |
| Finishing and traceability (8, 9) | Film thickness, serial walk-back | Medium |
| Capacity and rework (11, 12) | Written reservation, recorded rework | Medium |
If you have time for only one question, ask this: show me the last helmet you scrapped and tell me why. The answer reveals whether the factory has a defined failure taxonomy, records deviations, and lets anyone with authority look at the data.
An Adjustable Basalt Fiber Bike Helmet with a fit dial and a taillight looks simple from outside. Inside there are perhaps a dozen process steps, and each one is a place where an audit either finds control or finds a story.
Audit the process, sample the records, and insist that answers are written down. A certificate describes a batch that has already been made. An audit is the only instrument that tells you what the next batch will look like.
Bring the twelve checkpoints, score them honestly, and move the weak ones into a supplier development plan before the first purchase order rather than after the first field return.
E-mail
: [email protected]
WhatsApp
:
+86 19556521852
E-mail
: [email protected]
WhatsApp
:
+86 19567201995
E-mail : [email protected]
indirizzo aziendale : Dazhou High-tech Zone, Sichuan Province, China

Rete IPv6 supportata



E-mail : [email protected]
