OEM Bike Parts Sample Signoff: A Timed Evidence Trail

One line in the federal bicycle rules trips up almost every first-time buyer: a bicycle offered for sale in disassembled or partially assembled condition still has to meet the requirements after assembly according to the manufacturer's instructions. Shipping in a box does not lower the bar, and neither does the fact that the buyer, not the factory, turns the last bolt. That one sentence changes what a preproduction sample is for.

Read it as a scheduling instruction rather than a legal footnote. A sample signoff is not a beauty contest, not a color check, and not a one-afternoon photo shoot. It is a timed evidence trail, and the first gate is the assembly state: what condition the sample is offered in, who assembled it, and which instruction sheet they worked from.

Day 0 Agree the Sample Signoff Sequence

Build the sequence into the sample plan before the supplier cuts metal. Freeze the configuration first: part numbers, revision, material, finish, and the exact build the sample is supposed to represent. A signoff against a moving target is worthless, because months later nobody can say which version of the drawing the signature belongs to.

Next, state the offered condition in the plan, whether fully assembled, partially assembled, or knocked down, and name who does the final assembly and which instruction revision they follow. If the dealer does it, the dealer's instruction sheet becomes part of the sample file, not a topic for later.

Then assign the record points while the calendar still has room: assembly state notes and photos, the fastener log, the chain observation, the adjustment access walk-through, and the closing file audit. Each gets a named owner at the supplier and a named owner on your side. What you cannot define yet, including test method, tolerance, and acceptance threshold, goes into the RFQ as an open question. A supplier who answers that they always do it this way has given you a talking point, not a specification.

Day 1 Freeze the Assembly State

The first day the sample is on your bench, capture three things before anyone rides it: the condition it arrived in, the instruction revision used to assemble it, and every step still left to the buyer or the dealer. The requirement follows the bike into that final state, so a sample that arrives half-built and gets finished in your workshop is judged on the finished result, not on the carton.

Where the sample is fully assembled and ships that way, pinning the state down is simple: photos, date, and the name of the person who checked it. Partial assembly is the case that catches teams out, because the finished condition then depends on work done after the carton is opened, and the instructions for that work sit with whoever performs it.

Put the citation on the file header with its edition. The text here is 16 CFR Part 1512 as published in the 2025 edition of the CFR. Annual editions are snapshots, so record which one you read and re-check it before you lean on it for a later shipment. The scope is a bicycle requirement, not a parts certificate: a signoff on a frame, a fork, or a drivetrain component proves nothing about the assembled bike and is not a road-approval document.

Day 2 Log Fastener Condition

Day two is for hardware. Walk the assembled sample and list every screw, bolt, and nut that attaches or secures a component, then record what each one did during assembly and adjustment. The bar the file has to reflect is that screws, bolts, or nuts used to attach or secure components shall not fracture, loosen, or otherwise fail their intended function during the tests required in this part.

Write observations, not verdicts. A note that no loosening was observed across the assembly and adjustment passes is a record. A single word such as passed is a claim that needs a method, a fixture, and a result, and none of those arrive with a sample unless you asked for them. Note the torque tool used, whether it was calibrated, and whether the supplier sent a fastener test report at all.

The log pays off later, when a field complaint lands on your desk and the real question is whether a bolt loosened because the part was wrong or because someone over-tightened it during assembly. Without location-level records the two explanations look identical.

Week 1 Check Chain Path for Binding

By the end of the first week the sample should be assembled well enough to turn the cranks. Put the chain on and watch it run.

The requirement is narrow: the drive chain shall operate over the sprockets without catching or binding.

Record the chain path, the sprocket engagement, and any catch or bind you saw, in the assembled state rather than in a bag of parts.

This is where signoff files usually go soft. A photo of a chain resting on a chainring looks the same whether the drivetrain runs clean or binds later in the rotation; catching and binding appear only in motion, and often at one specific spot in the rotation, from a stiff link, a guide sitting slightly off, or a sprocket pulling the chain to one side. A single hand-driven revolution will not find it either, which is why the useful record names the exact configuration tested: which sprockets, which chain, which guide, how many full revolutions, and what happened at each one. If the supplier sent a test report, attach it and note the method behind it. If your own team turned the drivetrain by hand and saw nothing, write that sentence and nothing stronger. Resist upgrading an observation into a compliance statement: runs clean on the stand is a record, meets the bicycle requirements is a claim this file cannot support.

Week 2 Confirm Adjustment Access

Threaded hardware has a second job beyond holding parts together: all threaded hardware shall be of sufficient quality to allow adjustments and maintenance. That is an access question as much as a materials question, and the assembled sample is the only place it shows up honestly.

Ask a mechanic to perform one routine adjustment on the bike as built, whether cable tension, brake pad position, or seat height, and note where a tool does not fit, where travel is blocked, or where a structural part has to come off first. An adjustment that requires dismantling is an open engineering question for the supplier, not a maintenance step you assume the field will absorb.

The cost is easy to underestimate from a desk. A limit screw you can reach with the wheel in place takes a minute on a stand; the same screw behind a removed wheel turns into a shop visit, and repeated shop visits are what make a cheap part expensive over a season.

One Month Later Audit the Signoff File

One month later, open the file the way a stranger would. The person reading it will not be the one who built the sample, and often will be answering a customer complaint or an internal audit. Can they reconstruct what was checked, on which configuration, on what date, and by whom?

Cross-reference the four blocks against each other, meaning assembly state, fastener log, chain observation, and adjustment access, then check each against the open RFQ list. Add the scope note that keeps the file honest: the signature records what was observed on this sample, on the date shown, and it is not a test report.

Finish with the list of what was not done. No agreed test method, no supplier report received, no field data, no third-party inspection. Naming those holes protects the signature; leaving them out is what turns a working document into a liability later on.

The objection usually comes from the commercial side: the sample looks fine, the supplier is waiting, and a signature is just paperwork. It is the only thing standing between a sample and a production release, which is why the release rule should be written down before the sample lands. If the file cannot show the assembly state, the fastener condition, and the chain path, and cannot list every untested item, sign conditionally and hold the release. A conditional signoff costs a short delay and some back-and-forth. A full release backed by a file that cannot answer a warranty claim costs far more.