Perimeter times length, and nothing else
For anything that comes in lengths, the coated area is the perimeter of the section multiplied by the length. That is the entire method, and it is exact. Measure the perimeter in inches, divide by 12 to get feet, multiply by the length in feet, and you have square feet.
The reason this beats every rule of thumb is that it does not care what the section is called. A tape wrapped around the section gives you the perimeter directly, which is why the last shape on the list exists: channel, I-beam, tee, extrusion or a shape nobody has a name for, all handled the same way. The named shapes are conveniences for the cases where you would rather type two dimensions than fetch a tape.
The angle and the tube are the same number
Here is the result that catches people. A 2 by 2 angle and a 2 by 2 square tube have exactly the same coatable perimeter: 8 inches. Not approximately, exactly, and it does not depend on the thickness.
Trace the angle. Down the outside of one leg is 2 inches, across the outside of the other is 2. The two toe ends are one thickness each. Back up the inside faces is 2 minus one thickness, twice. Add them: 2 plus 2 plus 2t plus 4 minus 2t, which is 8. The thickness cancels. Wrap a tube instead and you get 2 plus 2 plus 2 plus 2, which is also 8. The inside faces of the L happen to make up exactly the two sides the tube adds.
Now compare the weight. A 2 by 2 by 1/8 angle has a cross-section of 4 minus 1.875 squared, which is 0.484 square inches. A 2 by 2 by 1/8 tube has 4 minus 1.75 squared, which is 0.938. The tube is nearly twice the steel for identical coating. Anyone estimating paint or powder from tonnage is off by a factor that depends entirely on which sections the fabricator happened to use, and it is not a small factor.
What the adjustments are actually doing
| Adjustment | What it covers | When it is large |
|---|---|---|
| Welds and gussets | Weld caps, tabs, brackets, plates that were never on the cut list | Heavily welded frames and anything with a lot of small attachments |
| Surface roughness | A blasted profile has more true surface than a smooth one, which coating fills | Deep profiles on the first coat over bare steel |
| Masked faces | Machined surfaces, bearing bores, threads and mating faces kept bare | Machined weldments, which can be a fifth of the area |
| Cut ends | Not counted at all by the linear rows | Short pieces, where the ends stop being negligible |
The roughness allowance is worth a word because it is often confused with a loss allowance. Material lost in the pot, on the roller or as overspray is not area, and it belongs in the transfer efficiency or loss figure of whichever material calculator you feed this into. The number here is genuine extra surface that genuinely has to be covered.
Where this sits
The obvious next steps are the powder usage calculator for pounds of powder, the coating spread rate calculator for gallons of a liquid coating, the abrasive blast media calculator for preparation, and the plating time and current calculator if the finish is a deposit. For the weight of the same bill rather than its area, use the metal weight calculator, and for flat blank sizes before anything is bent, the bend allowance calculator.
Questions people ask
Why does the angle option not exist separately?
Because it would compute exactly the same number as the rectangular outline, and having two entries that do identical arithmetic is a good way to make people distrust the page. An angle with legs A and B has a section perimeter of 2 times A plus B, which is the same expression as a rectangle A by B, and the thickness cancels out of the angle exactly. Trace the outline and you can see why: the two inside faces of the L, plus the two toe ends, add up to precisely the two sides that a closed rectangle has and the angle does not.
Should I include the inside of tubes?
It depends on what the finish is and whether anything reaches. For a sealed frame the answer is usually no: the bore is closed off and nothing gets in. For open tube that will be dipped or plated, the electrolyte does reach the bore even if it plates badly there, so the area is real. For spray application into a tube, the honest answer is that a gun reaches perhaps a diameter or two in and then stops, so counting the whole bore over-estimates. The outside-and-bore option is there for the cases where the bore genuinely gets coated, and the measured-perimeter option lets you enter a partial figure if you want to count part of it.
How much should I add for welds and roughness?
On a bolted or lightly welded frame, a few percent covers it and the default of 3 is a reasonable start. On a heavily fabricated weldment with gussets, tabs and stitch welds everywhere, ten percent or more is easy to justify, and the reason is that most of those features never appear on a cut list at all. Surface roughness is separate and only matters on the first coat over freshly blasted steel, where a deep profile genuinely has more area than the nominal surface. Your own comparison of estimated against actual material use, on a job you have already done, beats any figure here.
Can I use this for something that is not steel?
Yes, because nothing in it is a property of steel. Perimeter times length is geometry and it applies to aluminium extrusion, stainless tube, plastic profile or anything else that comes in a constant section. The only place material enters the discussion is the weight comparison in the guide, and the point being made there survives any material: two sections with the same outline and different wall thickness have the same area to coat and different masses, so estimating coating from tonnage is unreliable regardless of what the tonnage is made of.
What if my parts are not lengths at all?
Use the sheet row for anything flat, which counts both faces and all four edges, and use the measured perimeter row for anything that has a constant cross-section however odd. For genuinely three-dimensional parts, such as a casting or a pressing, neither works and you are into either measuring a sample by wrapping and counting, or estimating from a comparable part whose area you already know. A practical shortcut for a small awkward part is to coat one, weigh it before and after, and work backwards through the coverage arithmetic to get the area, which then applies to every other part of that shape.