At Andy Marine, we believe that practical manufacturing knowledge is one of the most valuable assets built over time.
Inside our workshop, there are engineers, production specialists, and experienced technicians who have spent years working with stainless steel marine hardware — from material selection and casting processes to finishing and quality control.
Many of these insights are learned through daily production experience: understanding why a component fails, how different materials perform in marine environments, and what details truly matter when developing reliable hardware.
However, this practical knowledge is often shared only within the workshop and rarely reaches customers who are making sourcing and engineering decisions.
ENGINEER TALKS is our effort to organize these conversations and share the manufacturing experience behind marine hardware.
Through this series, we discuss materials, production processes, design considerations, and real-world lessons from the perspective of a marine hardware manufacturer.
Today's ENGINEER TALKS Ep.32 is presented by Engineer Wu, written by me the editor.
Two Checks, Two Unanswered Questions
Picture a hypothetical receiving bench. A box of polished 316 fittings arrives with a supplier report that displays "316L." On the dock, a small permanent magnet is touched to one fitting; it pulls. The report and the fitting seem to point in two directions, and the instinct is to settle it at once — reject the batch or wave the report through. Both moves skip the same step: the magnet has answered one question, the report a different one, and neither says what the buyer still needs to know.
The aim of this episode is not to grade-select a steel or diagnose rust. It is to decide what evidence is worth obtaining when 316 marine hardware is received, and to read each item within its limits.
First, Define What You Actually Ordered
Before any check can be judged, the requirement must be written in a form a supplier can be held to. "316 hardware" is a shorthand, and a shorthand is not a requirement. What was ordered — a specific grade designation, and which product specification governs it? Is the item a casting, a wrought component, or an assembly of parts made differently?
The distinction is not academic. Cast and wrought alloy designations carry different product specifications and composition limits; a cast-equivalent shorthand is not proof of identical chemistry, and does not by itself make one an automatic substitute for the other. Judge a casting against its casting specification and a wrought part against its grade specification, and record the ordered designation, the product form and the governing specification before the goods arrive.
What a Magnet Can and Cannot Tell You
A permanent magnet is a useful, inexpensive field tool, and one of the most misread. A hand-held magnet response is not a conclusive chemical grade test; the magnetic behaviour of austenitic stainless steel depends on its microstructure and processing condition, not simply on the grade name.
Two ordinary mechanisms explain a genuine, correctly specified part responding. Cold working can transform some austenite to ferromagnetic martensite, so a formed, sheared or machined austenitic part may pull toward a magnet; the response depends on composition and on how the part was worked, so not every formed 316 item behaves alike. Separately, austenitic stainless castings can contain ferrite and show magnetic attraction. A casting's structure is not a wrought structure, and cast and wrought conditions have to be distinguished.
Treat a magnetic response as an observation, not a verdict: record whether the magnet pulled, and where on the part. It does not identify a cast grade, does not establish conformity, and cannot substitute for knowing whether the part is 316. Where a magnetic property is itself the requirement, that is a separate specified measurement — not something a pocket magnet settles.
What the XRF Reading Actually Measured
Handheld X-ray fluorescence answers a narrower question than its output suggests. Its reading reports the elements it can detect — typically the heavier alloying elements that distinguish steel families — for the spot it measured. It does not measure carbon. A displayed alloy-library name such as "316L" therefore cannot by itself verify the low-carbon requirement of a 316L order: the instrument read the major elements, and the L designation depends on carbon, which the method did not see.
Follow-up questions matter whenever an XRF result is offered as evidence. Which item and which spot were measured — a polished face, a thread, a weld, a coated or contaminated surface? Was the surface prepared, and how? Surface condition matters: coatings, plating, oxide and shop contamination can affect the measured result. Ask whether the measurement location and preparation were suitable for the material the report is meant to identify. If an assembly has separately specified constituents, ask which component was read; the answer decides how far the result extends.
Then separate what the reading is evidence for. Major elements consistent with the 316 family give a useful alloy-family screen, and a clean screen matters — but it does not, on its own, establish conformity of the whole part. A low-carbon requirement is an additional question beyond identifying the major alloy elements, and the governing ordered product specification, not one generic chemistry table, is what the supplied product form has to satisfy.
Closing the Low-Carbon Gap
If the order genuinely requires a low-carbon grade, the carbon question has to be answered by a method that can measure carbon. Spark atomic emission spectrometry can analyse carbon in austenitic stainless steels, but it needs suitable specimen geometry, preparation and calibration or reference material. Some portable optical-emission or LIBS systems can measure carbon when configured and operated for that task; carbon-capable LIBS may require argon and careful surface preparation.
The point is not to name an instrument but to check three things before accepting a carbon result: that the method is carbon-capable, that the operator and laboratory are competent for it, and that the calibration, reference material and report are appropriate. Do not assume every device sold under the same technology name has equivalent capability, and do not treat an instrument label as a measurement. Surface condition and preparation affect material analysis, so agree the test location and preparation with the competent operator, allow for coatings, contamination and product geometry, and accept that preparation or a destructive test may mark or alter a finished surface; where parts are polished or cosmetic, agree in advance how any sampled item will be handled. None of this implies a buyer must purchase an analyser — it is a specification for the evidence a supplier is asked to provide.
The Incoming-Evidence Table
Each check answers its own question and leaves the rest open; the table below pairs a common check with what it can honestly answer and what remains.
| Check or evidence | What it can answer | What still remains |
|---|---|---|
| Hand magnet | Whether the part responds at all and where; useful context and family-level sorting | Not a chemistry or grade result; a response can come from cold work or from a cast structure |
| Suitable handheld XRF result | The major alloying elements the instrument can detect, for the spot measured | Carbon and other light elements; it cannot close a low-carbon (L-grade) requirement |
| Carbon-capable analysis, properly configured | The carbon question, when the sample, preparation, calibration and report are appropriate | The measurement applies to the sampled location; batch inference needs an agreed basis, and other acceptance requirements remain |
| Linked material certificate (grade, product specification, heat or lot) | Documentary identity and a traceable link to a heat or lot | Whether the tested piece represents this delivery, or the whole lot |
| Part-level dimensional, finish and mechanical inspection | The specified characteristics assessed by the selected inspection methods | Corrosion performance in service; none of these establish service-life suitability |
Tracing the Report to the Goods
A report is only as useful as the link between it and the physical parts on the bench. As a buyer documentation recommendation, define the required material and inspection records in the RFQ or order, then match the grade, product specification and heat or lot identity to the actual delivered material and its traceable markings or packing records. Read the material's identification, not only the label on the box; a certificate that cannot be tied to the goods it arrived with is a document, not yet evidence.
Two distinctions keep the traceability review focused. Keep a sampled result separate from a lot disposition: a test on one item or one sample tells you about that sample, and treating it as a release of the whole batch is an inference that must be supported by an agreed sampling and traceability plan, not assumed. And expect finished multi-component goods to differ from a single heat of stock: an assembly may join separately specified constituents, so ask which component each piece of evidence describes. Where records are missing or the identity chain is broken, agree an inspection plan with the supplier or test provider rather than inventing a marking convention that may not apply.
The Receipt Decision
Putting the checks together gives a plain receiving sequence. Confirm what was ordered, in writing. Use the magnet as an observation and record it as one. Accept an XRF result for the alloy-family screen it is, and do not read a library name as carbon proof. Where a low-carbon requirement exists, ask for carbon-capable evidence that names the method, the sample and the preparation. Match the certificate to the goods through the grade, specification and heat or lot identity.
When that evidence conflicts, the proportionate response is to withhold the intended release and request clarification — not to accuse a supplier or open a dispute. A missing measurement is a question, not a verdict.
Keep the boundary honest. Chemistry identification, corrosion performance, mechanical rating, finish and fit are separate acceptance questions. A chemical result answers a composition question; it does not on its own verify finished-part dimensions, load rating, surface condition or corrosion performance in service. Positive grade identification requires appropriate chemical analysis of the required elements against an applicable specification; magnet screening or a main-element-only screen is not a substitute for the elements a method did not measure.
For rust symptoms noticed later in service, see the separate 316 stainless steel corrosion-diagnosis guide. For incoming material identity, the sequence above is the one to run before a batch is released.
