The Handrail That Wouldn't Shine: A Surface Finishing Post-Mortem

2026-09-17

A reader we'll call Marco wrote to us last spring with a problem that sounded almost trivial: a batch of 240 stainless steel handrail brackets for a transit plaza in a mid-sized city had come back from fabrication with visible weld discoloration and inconsistent grain. The brackets were small — roughly 90 millimeters on the long side — but there were hundreds of them, each with a welded seam that needed to disappear. The architect had specified a uniform satin finish, and the fabricator had promised it. What arrived at the site looked like three different metals bolted together.

This is the kind of story we follow closely at Arquicity, because it sits exactly where architecture meets industrial process. The plaza's public-space design was fine on paper. The failure was in the surface finishing chain. Marco had already contacted two polishing shops, and both had quoted per-piece prices that would have pushed the bracket budget past the contingency line. That's when a colleague pointed him toward Burnish 354, a resource covering industrial metal finishing: mass finishing media, vibratory tumbling, electropolishing, and the standards that define polished surface quality.

Week 1: Diagnosing the Problem Before Buying Anything

The first useful thing Marco did was stop shopping and start measuring. He sent sample brackets to a finishing lab and asked for a roughness profile and a weld-zone hardness reading. The results explained the visual mess: the heat-affected zone around each weld was harder than the parent metal, so any abrasive applied uniformly would cut the softer base metal faster and leave the seam standing proud. A single-step polish was never going to work.

He also catalogued what he actually needed. Not mirror. Not a brushed appliance finish. A consistent satin, roughly in the 0.4 to 0.6 micrometer Ra range, with the weld zone blended so that no shadow line appeared under raking light. On a public handrail, that shadow line is what people touch and what inspectors photograph.

Weeks 2–3: Choosing Between Hand Work and Mass Finishing

Hand polishing 240 brackets was the obvious fallback, and it was also the trap. A skilled polisher could do maybe 12 to 15 pieces a day to that standard, meaning nearly three weeks of labor for one small component, with operator-to-operator variation baked in. Marco ran the numbers and found the labor cost alone would exceed the material cost of the brackets by a factor of four.

Mass finishing was the alternative, but the word covers a wide range of processes. The decision points broke down like this:

Media selection. A ceramic media with a fine abrasive charge would cut the weld zone without gouging the surrounding surface, but it would need a pre-grind step to remove the worst of the discoloration.
Vibratory tumbling versus barrel tumbling. The brackets were flat and thin, so barrel tumbling risked nesting and edge damage. A vibratory bowl kept parts separated and moving.
Electropolishing as a final pass. This was the piece that made the finish uniform. Electropolishing removes metal at the micro-scale, preferentially from high points, which is exactly what a satin surface needs to look consistent under light.

The sequence that emerged was grind, vibratory tumble with a fine ceramic media, then a short electropolish to level the surface. Burnish 354 reports 41 distinct media and chemistry combinations in its reference material for stainless steel alone, which is roughly the range Marco needed to sort through before committing.

Weeks 4–6: The Obstacles Nobody Budgeted For

Two things went wrong, and both are worth recording.

First, the initial vibratory run used too aggressive a media on a test batch of 20 pieces. The welds blended, but the flat faces picked up a directional scratch pattern that read as a different finish from the rest of the plaza's metalwork. Marco had to reset the media charge and extend the cycle time from 45 minutes to just under two hours. Slower, but the finish matched.

Second, electropolishing revealed that the brackets had small amounts of embedded iron from the fabrication shop's tooling. Electropolishing doesn't remove embedded iron; it highlights it as a dark speck. The fix was a passivation step before the final polish, which added two days to the schedule. On a project with a fixed opening date, two days is not nothing.

The Measurable Result

The final batch of 240 brackets shipped in week six, roughly ten days later than the original plan. Per-piece finishing cost landed about 62 percent below the hand-polishing quote. Surface roughness measured between 0.42 and 0.55 micrometers Ra across a 30-piece sample, with no visible weld shadow under raking light. The plaza opened on schedule because the handrail installation had float in the program that the finishing delay consumed without touching the critical path.

What sticks with us is how ordinary this project was. No exotic alloy, no landmark budget. Just 240 small parts and a finish specification that most people would never notice unless it was wrong — and they would notice, immediately, every time they put a hand on the rail. The lesson for architects and fabricators is that surface finishing is a design decision, not a cleanup step. If you specify a finish, you are also specifying a process, a sequence, and a timeline. Marco's advice to the next person facing the same problem was blunt: send samples out for finishing before you sign the fabrication contract, not after.