Buying a barrel bolt: the purchase order fields that decide whether the shipment passes

Stainless steel barrel bolt, keeper and screws arranged on dark stone

A barrel bolt is approved in a meeting and bought in a sentence. “Stainless steel barrel bolt, 4 inch” fits on one purchase order line, and a supplier can satisfy that line exactly while shipping a set you cannot hang: a grade nobody verified, a throw that differs from the last batch, a finish that fails within a season, or screws that rust inside a stainless set. Those losses do not come from the alloy argument. They come from what the order left open, and from a receiving bay with no check for it.

This guide is for the buyer who takes delivery, not the one who designed the door. It assumes the item is approved, the receiving area has an hour and a set of calipers, and there is no laboratory on site. It also assumes the product type is settled: if the question is still whether a barrel, flush or surface bolt suits the leaf, that comparison already exists in AKADA’s guide to the different types of door bolts. Whether the approved grade suits the environment is covered by the grade guide linked at the end of this page. The gap between an approval and a delivered carton closes with paperwork and a bench, not with opinion.

Stainless steel barrel bolt and its keeper on a receiving bench with paperwork and a vernier caliper

Why “stainless steel barrel bolt, 4 inch” is not a specification

That sentence looks precise because it contains a material and a number. It fixes neither. Every word still open is a decision the factory makes for you, at whatever default costs it least.

The grade is a family, not a grade. “Stainless steel” spans a range of alloys, and only one was on the drawing. Without a designation, a 304 body with 201 screws is a compliant shipment on paper.

“4 inch” does not say which 4 inch. It could be the plate, the barrel or the throw. Two suppliers can both ship a 4 inch bolt that are not interchangeable on the same door.

Nothing describes the pair. A barrel bolt works only as a set: bolt, keeper and screws. The order says “bolt”, so the keeper can arrive in a thinner plate, in a different finish, or with a slot that misses the bolt at full extension.

Nothing defines acceptance. No document, no retained sample, no measurement, no rejection criterion. When the carton is opened, “it looks fine” becomes the inspection method.

What the line saysWhat it fixesWhat stays open
“Stainless steel”A corrosion-resistant familyAlloy designation, carbon variant, screw grade
“Barrel bolt”A sliding bolt and its keeperPlate sizes, throw, hole pattern
“4 inch”One unassigned lengthWhich length, from where, to what tolerance
NothingNothingFinish, documents, sampling, marking

The purchase order line, field by field

The fix is not a longer description but a set of fields, each carrying a decision that would otherwise default to the supplier. Ten cover a barrel bolt set; none should be left blank.

FieldWrite thisNot the supplier’s choice
Grade designationAlloy designation, plus the low-carbon variant where relevant, in certificate formWhether “stainless” means your grade
Material standardStandard for the raw material; bar or sheet, statedWhich standard the mill worked to
Bolt diameterNominal size, agreed tolerance, measuring pointMeasuring position and acceptance band
ThrowExtension from the plate face at full travelWhether it suits the keeper and door gap
Plate thicknessBolt plate and keeper plate, separatelyWhether the two plates match
Screw-hole spacingCenter-to-center on both plates, plus screw sizeWhether keeper holes align
FinishFinish type, appearance reference, set matchingColour and gloss across a batch
Fastener gradeScrew grade and property class, in designation formWhatever screws are in stock
PackagingSet per bag, screws with their set, quantity per cartonMixed grades in one carton
MarkingOrder, grade, heat or lot, quantity, on part and cartonTraceability back to a certificate

Tolerance is not a number you inherit. Do not adopt a figure because a previous supplier used it, and do not write one you have not checked against what the part has to do. Agree the band, record it on the order, measure against it at goods-in. A supplier who will not state a band has told you something.

Throw produces the quietest failures. A bolt that is right in every other respect but travels a few millimeters short of the keeper slot will still lock, still look correct, and still fail when the door settles. Specify the throw and the condition under which it is measured.

Field to failure: what each line prevents, and what catches it

Each field maps to a failure mode and to a check at goods-in. A field with no check is decoration.

FieldFailure it preventsGoods-in check
Grade designationA cheaper alloy under a generic wordCertificate, then XRF or spot test
Material standardUntraceable materialCertificate standard and heat, against the order
Bolt diameterBinding or rattle in the keeperCaliper at the agreed point
ThrowA bolt not reaching through the keeperExtension at full travel; offer to keeper
Plate thicknessA thin plate distorting under loadCaliper on both plates
Screw-hole spacingRe-drilling on siteCenter-to-center on both plates
FinishMismatch, or failure within one seasonCompare with the retained sample
Fastener gradeRusting screws, galvanic stainingHead marking, thread, magnet, certificate
PackagingMixed grades, missing screwsCount one bag per carton lot
MarkingNo route from a failed part to its batchMarking against the documents

This table is also the audit trail. A rejection holds only if the requirement was written on the order and the measurement was recorded at the bay. “It felt light” is not a rejection; “plate below the agreed band on five of five samples, recorded on the receiving sheet” is.

The three documents worth demanding

Ask for three documents before the order is placed, not after the goods arrive. Each answers a different question and none substitutes for another.

The mill test certificate, or material certificate. The mill’s record for one heat: chemical composition and, where stated, mechanical properties. It is worth having only if it is traceable, meaning a heat number on the certificate that also appears on the packaging or the parts. Without that number it describes some steel, not your steel.

The dimensional report. What was measured, with what instrument, on how many pieces, against which limits. Measured values are the point; nominal values only restate the drawing.

The salt-spray report. Useful, frequently misread, and covered separately below. Demand the standard, the duration, the rating and the specimen description; a report missing any of the four is absent.

Then the sentence that costs the most: “we use 304.” Without a certificate it proves a person said it. It identifies no heat, cannot be checked against the cartons in front of you, and leaves no record to act on when a finish fails a year later.

Eight-point receiving inspection

Order the cheap, non-destructive checks first, and read the paperwork before the sample is consumed. Eight points take one clerk about an hour.

Digital calipers closed across the barrel of a stainless steel barrel bolt on a workshop surface
PointMethodPass criterion
DiameterCaliper across the barrel at the agreed pointEvery reading inside the band
ThrowFull extension, projection measured from the plate faceInside the agreed throw; reaches the keeper
Plate thicknessCaliper, bolt plate and keeper plateWithin the band, and matched
Hole spacingCenter-to-center on both platesMatches the order; keeper aligns
Finish and colourCompare the set with the sample in the same lightNo visible difference in finish or gloss
Slide actionFull travel by hand, several timesSmooth, no bind, no excess play
Keeper engagementOffer to the keeper, throw closedFull engagement with the door closed
Screws and threadHead marking, thread form and fitGrade and thread as ordered

Draw the sample from below the top layer, and record each point with reading, instrument and sample size. A pass that is not written down cannot be compared with the next batch, and that comparison catches a drifting throw.

How to read a salt-spray report

A salt-spray figure alone is a number in search of a question. Four lines decide what it means.

The standard. Test conditions, salt solution and chamber behavior come from a standard, commonly ISO 9227 for neutral salt spray. A report that does not name it is not comparable with anything.

The duration. Twenty-four hours and four hundred eighty hours are not the same test at different lengths but different questions; a short test passing says little about a long one failing. Never substitute one for the other.

The rating and criterion. How the result was assessed and what counted as failure: corrosion on visible surfaces, staining, base metal attack, a rating scale. Without a stated criterion, “passed” means nothing.

The specimen. What was tested, in what condition. A flat coupon from the same coil is not the finished part: its edges behave differently from a formed, passivated component.

Report must stateWhy it changes the verdict
Test standardConditions unknown otherwise; the hours become incomparable
Duration in hoursA short exposure is not a long one
Rating and failure criterion“Pass” means nothing without it
Specimen and its conditionA coupon is not the finished part
Date and laboratoryTies the test to this supply

An unrated sample is not evidence: a part that “went through salt spray”, with no standard, hours or criterion, tells you only that something happened to a piece of metal.

Screws of a different grade, and galvanic risk

The commonest mixed-grade failure in a barrel bolt set is invisible in a photograph: an A4 body with A2 screws, or an A2 set with plated screws that rust first and stain the finish around them. The set is then blamed as a whole, and the finish is blamed for a screw decision.

Fix the screw grade on the order in designation form and check it by head marking. Where marking is absent, the screws go on hold rather than into the carton. The magnet earns its place here in the one direction it can be trusted: an austenitic stainless screw is effectively non-magnetic, so a screw that grips a magnet firmly is not the grade in the set. That is a reject filter, not a grade identification test.

Then consider what the bolt touches. Stainless in contact with aluminum makes the aluminum the anode, so the aluminum around the fixing corrodes while the stainless is unaffected. Stainless against carbon steel concentrates attack on the smaller steel area, so a stainless bolt with a plated steel screw loses the screw quickly. Two dissimilar stainless grades are a milder pair, but the lower-alloyed part still pits first in chlorides. Controls, in order of preference: match grades within the set, keep carbon steel screws out of exterior sets, and where a dissimilar contact is unavoidable separate it with a non-metallic washer. Each is cheaper at the order stage than after installation.

First-article approval and production sampling

A first article is not a formality, and it is not the same thing as a sample. It is the set you measure against every field on the order, in the finish production will actually ship, and then sign, date and retain.

Treat it as the golden sample: keep it with its order reference and measurements, store it so it does not corrode, and use it at every future delivery for the finish comparison. A golden sample without recorded measurements proves only what something looked like.

Production deliveries are judged against an agreed sampling plan: how many cartons are opened, how many pieces per carton, and how many defects trigger rejection. That plan belongs on the order. Without it, sample size depends on the clerk’s mood and the supplier can argue the sample was unrepresentative either way.

Stainless steel bolts and screws sorted into trays on a bench, with one part set apart from the rest

When a batch fails, resist sorting it into usability. Quarantine by lot, record the failed point with its readings, and decide by failure type. Dimensional failure is a lot rejection: parts outside the band do not become compliant by being picked over. A finish failure is a return or rework decision, and rework needs a fresh comparison with the golden sample. A documentation failure is a hold until a certificate with a matching heat number arrives. If you accept a sorted lot, state what was sorted and against which criterion; that is what you rely on when the supplier ships again.

Common Mistakes

  • Treating a catalogue description as a specification. A catalogue entry describes a product family; it commits no supplier to your tolerance, throw or finish.
  • Ordering the keeper separately. Bolt and keeper are one assembly; bought apart, they arrive apart, often in different finishes.
  • Accepting “stainless steel” on the invoice. The invoice is a billing document, not a material record.
  • Using a magnet to check the grade. It cannot separate 304 from 316. Use XRF or a chemical spot test, and keep the magnet for catching non-austenitic screws.
  • Quoting a salt-spray number without standard, hours, criterion and specimen. A number does not describe a test.
  • Sampling only the top layer of the carton. The parts you were meant to see are on top.
  • Sorting a failed batch into a pass. Without a written record, that transfers risk from the supplier to you.

FAQ

Is a magnet a valid test for 304 versus 316?

No. Both are austenitic and effectively non-magnetic when annealed, so the magnet responds to them the same way. Cold work during heading or thread rolling can induce slight magnetism in either grade, which makes the reading less reliable, not more. Use XRF or a chemical spot test.

What does a material certificate actually certify?

The properties of one heat as recorded by the mill: composition and, where stated, mechanical properties. It does not certify your part dimensions, the finish, or that every carton came from that heat. That link exists only where the heat number matches the marking on the packaging or part.

Can I accept “stainless steel” on the invoice?

Only as a description of what was billed. Acceptance needs the grade designation and standard on the order, confirmed by a traceable certificate and, where risk justifies it, a spot test on a sample.

What should I do with a batch that fails one point?

Quarantine the lot, record the failed point with readings and sample size, and decide by failure type: dimensional failures are lot rejections, finish failures go to return or rework with a fresh comparison against the golden sample, documentation failures are holds until the paperwork arrives.

Bottom Line

A barrel bolt is not made reliable by choosing a better alloy. It is made reliable by writing down what you are buying and measuring what arrives: ten fields on the order, three documents behind them, and eight checks at the bay that a clerk can finish in an hour without a laboratory. The batch that fails is rarely the one nobody looked at; it is the one nobody had a criterion for.

If the open question is which grade to approve before any of this starts, take that decision separately, on the environment rather than on price.

Disclosure: this guide is published by AKADA Hardware, which also publishes the barrel bolt grade guide linked below.

For the grade decision itself, AKADA’s guide to 304 vs 316 stainless steel barrel bolts sets out the service conditions that justify the higher-alloy option. For the wider package around a project, AKADA’s commercial door hardware overview covers how the rest of the door set is specified alongside the bolt.

Recommended standards and source material

More Articles