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.33 is presented by Engineer Li, written by me the editor.
Two Bright Fittings, Two Different Release Decisions
Consider a hypothetical receiving bench rather than any particular order. Two cleats of the same size arrive in one delivery, both mirror-bright, both wrapped the same way. One comes with a treatment record naming the operation, the specification it was worked to and the surfaces covered, traceable to its batch. The other comes with an invoice line that reads "polished". Neither appearance establishes whether the ordered treatment and acceptance requirements have been met; the first record provides evidence to evaluate, not an automatic release.
Brightness is an outcome a buyer can see. Treatment identity, measurement basis and coverage are things a buyer has to ask for, and none of them follow from the shine. A receiving inspection that begins and ends with "does it look right" is not wrong so much as incomplete, and the missing part tends to surface later in a place nobody photographed.
An Appearance Finish Needs a Reference Before It Needs a Number
Where a visible finish carries commercial weight, the useful instrument is a physical reference rather than an adjective. For appearance-critical hardware, agree and identify a representative physical finish reference as a buyer acceptance recommendation, then pair it with any separately required roughness measurement instead of substituting one for the other. A reference settles what "matching" means for the visible face; it says nothing about passivation, and nothing about surfaces it does not cover.
Make that reference part of the order rather than a courtesy sample. Identify the actual piece, record which face and orientation it represents, and note the viewing conditions for comparison, because a finish judged under workshop strip light and the same finish in raking daylight are two different conversations.
Roughness figures have their own place, but one figure carries less than it appears to. Ra is an averaged roughness parameter: different surface profiles can have the same Ra, so an Ra value alone does not fully characterise an appearance finish. A deep, narrow profile and a shallow, wide one can average to the same number while reflecting light quite differently. Direction matters for the same reason: mechanical finishing produces a scratch pattern affected by the abrasive and the working direction, so agree the intended direction or lay where visible matching matters, including which way it should run on a given part.
Cleaning, Descaling, Passivation and Electropolishing Answer Different Questions
A finish specification needs to separate the operation performed from the evidence used to accept it. Cleaning, descaling, passivation and electropolishing answer different questions.
Start with wording. Define the intended surface-treatment operation and the surfaces it covers in the purchase specification; do not treat cleaning, descaling and passivation as interchangeable terms. A buyer and a supplier can both believe they agreed on "passivation" while one means a chemical immersion and the other a line on a routing card.
The tasks separate cleanly. Cleaning, scale removal and passivation perform different tasks, so confirm the preparation and treatment sequence with the treatment provider rather than conflating them — including preparation for a component that arrives with cutting oil or handling soil, since acid pickling and passivation do not remove oil or grease and the provider has to define what is suitable.
Scale removal is its own operation. Pickling removes high-temperature scale and the adjacent chromium-depleted surface layer, involves metal removal and can change or dull appearance — a difference in tone that is a finish question, not automatically a defect. Passivation is not a substitute for removal of scale; scale removal may be necessary before passivation can be effective, so accepting a part as "passivated" does not establish that any scale was removed.
What passivation is aimed at is narrower than its reputation. Chemical passivation is aimed at removing exogenous free iron and assisting formation of the passive surface; it is not intended as a material-removal finishing operation. Because it works that way, it does not usually produce a marked appearance change, so brightness alone does not verify that it was carried out: a part that looks unchanged may have been treated correctly, and one that looks brighter may not have been treated at all.
Electropolishing is a different category, because it removes surface metal — but it is not a cosmetic repair route. It is not a reliable repair for visible scratches or other macroscopic irregularities, and its effect on finished features must be agreed rather than assumed.
Mechanical removal belongs in the same conversation for a different reason: it can leave abrasive or other particles embedded in the surface, and cleaning and contamination control remain separate acceptance questions from the finish itself. Carbon-steel grinding dust and shared tooling can introduce iron contamination, and a bright appearance is not evidence that contamination has been controlled.
Which Surfaces, and What State They Must End In
The hardest part of a finish specification is not choosing an operation; it is saying which surfaces it applies to, and what those surfaces must still do afterwards.
Threads, fits, sealing faces and mating surfaces are where this bites. Metal removal that is harmless on a cosmetic face can matter at a thread root or a sealing land, and the answer is not a universal rule but a question put to the right party: ask the treatment provider and the design owner how critical threads, fits and sealing faces are protected and verified. Whether material removal is permitted depends on the final drawing requirements; agree protection or allowance feature by feature rather than masking every functional surface by default. What is correct is that the decision is written down before parts are treated.
Recesses deserve the same explicitness. Visual inspection can find gross contamination, scale, rust and particles, but may miss thin oil or residual chemical films, and geometry can limit inspection of internal surfaces. If a surface cannot be inspected, the record should say so rather than imply a clean bill of health.
Five Acceptance Questions and What Each One Can Settle
This is a buyer's working list, not a quotation of any standard's reporting requirements. Each row pairs a question with the evidence worth agreeing in advance, and states what that evidence will not establish on its own.
| Acceptance question | Evidence to agree | What it does not establish |
|---|---|---|
| Does the visible finish match what was ordered? | An identified physical reference held by both parties, referenced in the order, with the face, orientation and viewing conditions noted | Does not establish that passivation or any other treatment was carried out |
| Is the roughness requirement met on the agreed basis? | Measurement locations, direction or lay, instrument and settings, evaluation basis and the acceptance rule, stated with the result | Does not establish visual uniformity across the part; different profiles can share one Ra |
| Which treatment was actually performed? | A process record naming the operation, the specification worked to and the surfaces covered, traceable to the component or batch | Does not establish scale removal, treatment of unlisted recessed surfaces, or suitability for the intended exposure |
| Which areas were treated, and which could be inspected? | An agreed list of treated surfaces, agreed inspection access, and a note of any surface that could not be verified | Does not establish that an untreated or unverifiable area is functionally acceptable |
| Is the part acceptable as received for its function? | Final functional checks agreed with the design owner for threads, fits, sealing faces and moving parts | Not a corrosion approval and not a service-life prediction |
Read each record against what it actually covers. A treatment record may answer the process question while leaving appearance or roughness unresolved; a finish reference supports visual comparison, not treatment verification.
Releasing the Batch Without Overreading the Paperwork
At receipt, the useful discipline is traceability rather than volume: tie the reference, the measurement result and the treatment record to the same order line and the same component or batch, so a question about one part can be answered without reconstructing the whole delivery. Where an exclusion exists — a face deliberately left untreated, a surface that could not be inspected — clarify it before release rather than after a complaint.
Be equally careful about what a passivation specification is being asked to prove. ASTM A967/A967M-25 describes qualitative tests intended to confirm passivation effectiveness, particularly removal of contaminant iron and other exogenous matter; its scope makes no recommendations on suitability for a particular application. A result against that specification is not an approval of marine service life, and the publisher's scope statement is worth reading directly: ASTM A967/A967M-25.
A short supplier enquiry covers most of the ground before an order is placed:
- Which operation is quoted — cleaning, descaling, passivation, electropolishing, or a sequence — and against which specification?
- Which surfaces does it cover, and which are excluded by design?
- What preparation is defined for components arriving with oil, grease or handling soil?
- What will the treatment record state, and can it be tied to our component and batch numbers?
- If roughness is specified, on what basis, at which locations and in which direction will it be reported?
- How are threads, fits and sealing faces handled, and who confirms the resulting condition?
None of these questions asks the buyer to specify chemistry, and none should be answered with a promise about service life.
Where This Sits Alongside the Rest of the Series
Finish acceptance is one link in a longer chain: our earlier article on structural design, surface finishing and installation carries the wider durability picture, the 316 stainless diagnosis piece follows a finish question that becomes a corrosion question, and the custom hardware article covers the dimensional side.
For a receiving department, the working summary is short: accept a finished surface by agreeing what the visible face should look like, on what basis roughness will be measured, which treatment was performed on which surfaces, and what the part must still do afterwards. Those are four separate agreements, and treating any one as proof of the others is where the arguments begin.
