What the code asks of a guard, what our sections actually deliver against it, and the arithmetic in between — including the places our own numbers are not as strong as we would like. Most railing companies publish a span table. This is the working behind ours.
Start here — what we do not haveCR Fence & Rail holds no ICC-ES evaluation report (ESR) and no Intertek code compliance research report (CCRR) for these systems. Several of our competitors do. Everything on this page is our own engineering against published code requirements and published material minimums — it is not a third-party rating, and we will not dress it up as one.
We are telling you this on our own specifications page, at the top, because the alternative is that you find out from an inspector. If your project requires a stamped evaluation report, say so before you order. We will tell you where we stand and, where we cannot serve you, we will say that too.
A guard has to survive four different demands. Three are rarely the problem. The fourth catches people out.
| Case | Magnitude | Where | Clause |
|---|---|---|---|
| Concentrated load — the governing case | 200 lb | Top of the guard, in any direction | IRC R301.5 · IBC 1607.9.1.1 |
| Infill load | 50 lb | Any one square foot of infill | IBC 1607.9.1.2 |
| Uniform load commercial only | 50 lb/ft | Along the top rail | IBC 1607.9.1 |
| Proof load — what a test rig applies | 500 lb | 2.5 × the concentrated load | ICC-ES AC273 |
It includes straight down — someone sitting on the rail, or leaning across it. For a square tube that is the same as sideways. For a rectangular tube it is much harder, because the section is shallower in the vertical direction than the horizontal one, and bending strength grows with depth cubed.
Our cable top rail is a 2 in × 1 in rectangle, so every rail number on this page is computed against the downward case, which is the one that governs. A supplier quoting only the horizontal case is quoting the easy direction.
We size against the 500 lb proof load, not the 200 lb service load. That is the load an evaluation test actually applies, and it means the safety factors below already sit on top of a 2.5× multiplier.
Every calculation on this page uses the guaranteed minimum yield for the alloy — the number the mill will stand behind — never a typical or average value.
| Line | Material | Min yield | PREN | Where it belongs |
|---|---|---|---|---|
| EdgeLine | Galvanised carbon steel | see note | n/a | Ground-level, 36 in |
| ProGuard | A500 Gr B carbon steel, galvanised | 317 MPa | n/a | General structural, 36 and 42 in |
| CoastLine | 6061-T6 aluminium, AAMA 2605 + e-coat | 240 MPa | n/a | Coastal, marine, pool |
EdgeLine's steel has been specified to us as a galvanised carbon steel in the 317 MPa class. Our factory has separately described it as Q235, which has a minimum yield near 235 MPa. Those are not the same material and we have not yet resolved the contradiction with a mill certificate.
Section 05 shows what each assumption does to the numbers. It is the reason EdgeLine carries a height restriction rather than a caveat — where an input is uncertain, we restrict the application instead of hoping.
A top rail spans between posts and is loaded at midspan. For a simple span the bending moment is M = P × L ÷ 4 , and the stress is σ = M ÷ S , where S is the section modulus about the axis being bent.
| Rail | Span | S (mm³) | Stress | vs 317 (Pro) | vs 240 (Coast) |
|---|---|---|---|---|---|
| Guard rail — 50 × 50 × 1.5 mm square | 6 ft | 4,722 | 215 MPa | 1.47 × | 2.09 × |
| Cable — 2 × 1 in × 2.5 mm | 4 ft | 2,911 | 233 MPa | 1.36 × | 1.93 × |
| Cable, at the 4.5 ft span we rejected | 4.5 ft | 2,911 | 262 MPa | 1.21 × | 1.72 × |
| Cable, if CoastLine were annealed 316 (207 MPa) | 4 ft | 2,911 | 233 MPa | — | 0.89 × — fails |
Annealed 316 stainless has a guaranteed minimum yield of 207 MPa, which is lower than structural carbon steel. The identical rail that carries 1.36× in ProGuard's steel would sit at 0.89× in annealed 316 — under yield at the proof load. 316 is a corrosion alloy, not a strength alloy, and a coastal product built from it has to be thicker or shorter-spanned to compensate.
CoastLine answers this in 6061-T6 aluminium at a 3.0 mm wall rather than in a stainless, which is covered in full in section 07. The point that survives is the one about 316: it is a corrosion alloy, not a strength alloy, and any coastal product built from it has to be thicker or shorter-spanned to compensate.
The same arithmetic, live. Change the inputs and watch the margin move — including into failure.
Section Wall (mm) Span (ft) Load (lb) MaterialWhat this does and does not model. It is a simple-span bending check at midspan against yield — the case that governs our rails. It does not model local buckling (which matters below about a 1.5 mm wall), connection capacity, or fatigue. It uses nominal wall thickness; a metric tube may be delivered up to 10% thinner, so treat anything under about 1.3× as having no real margin.
A post is a cantilever. The 200 lb arrives at the top and the moment is M = P × h , which grows directly with guard height. This is where the three lines separate, and it is not close.
⚠ Corrected 2026-09-17 — and the real lever is the STEEL, not the sectionAn earlier version of this page quoted 45 × 45 × 1.5 mm for EdgeLine. That is a proposed size. What ships today is 40 × 40 × 1.2 mm and it is weaker. Stating a proposal as a shipping fact is exactly the error this page exists to avoid.
Our tube is made in China, so the spec has to name a Chinese grade. Every figure below is against GB steel grades and is computed on the tolerance-minimum wall — nominal × 0.90, because a −10% wall tolerance is what the standard permits the mill to deliver.
| Post section | Q195 195 MPa | Q235 235 MPa | Q355 355 MPa |
|---|---|---|---|
| At 36 in | |||
| 40 × 40 × 1.2 — ships today | 0.51 | 0.61 | 0.93 |
| 40 × 40 × 1.5 — EdgeLine recommendation | 0.62 | 0.75 | 1.14 |
| 45 × 45 × 2.0 | 1.03 | 1.24 | 1.88 |
| 50 × 50 × 3.0 — ProGuard / CoastLine | 1.83 | 2.21 | 3.34 |
| At 42 in | |||
| 40 × 40 × 1.2 | 0.44 | 0.53 | 0.79 |
| 40 × 40 × 1.5 | 0.53 | 0.64 | 0.97 |
| 45 × 45 × 2.0 | 0.88 | 1.07 | 1.61 |
| 50 × 50 × 3.0 — ProGuard / CoastLine | 1.57 | 1.89 | 2.86 |
| To clear 1.0 at 36 in | Cheapest section | Steel used |
|---|---|---|
| On Q235 | 45 × 45 × 2.0 | 2.70 kg/m |
| On Q355 | 40 × 40 × 1.5 | 1.81 kg/m — 33% less |
Q355 (GB/T 1591) carries a 355 MPa minimum yield against Q235's 235 — 51% more, for one line on the purchase order. Every Chinese tube mill runs it. That is why our answer to a post that does not reach is a better steel first, and a bigger tube only if the steel is not enough.
⚠ And it is why the grade must be NAMED on the orderPre-galvanised tube in China is commonly rolled from low-grade strip. If nobody specifies the grade, Q195 — 195 MPa — is what a cost-optimising mill will supply, and the Q195 column above fails at every section and both heights.
So our order names the grade and requires a mill certificate for each heat. That certificate is also the first document an evaluation report would ask us for.
Deflection is not the constraint, and we want to be precise about that. AC273 allows a post to deflect h/12 — a full 3.0 in at a 36 in post. Ours moves 5.2 mm (0.20 in) at ProGuard and 13.8 mm (0.54 in) at EdgeLine. Both pass by a wide margin. Yield governs every post we make, which is why this page talks about stress rather than stiffness.
The two families are limited by completely different things, and conflating them is the most common error in this category.
| Picket guard rail | Cable railing | |
|---|---|---|
| Max post spacing | 6 ft | 4 ft |
| What limits it | Top rail bending | Cable sag against the 4 in sphere |
| The rail could go further? | No — it is the limit | Yes, but the cable cannot |
A tensioned cable is not rigid. Push between two supports and it moves, and the code asks only whether a 4 in sphere then passes. Published industry testing of 316 cable at 3⅛ in spacing and 225 lb tension puts the maximum unsupported span at 48 in. Our rail carries 1.36× at that span, so both limits land together — but sag is the one that binds first, and it is why widening a bay is never the answer.
We considered 4.5 ft and dropped it. Here is the whole argument.Post count is a ceiling function, so half a foot of extra span almost never crosses a boundary. On a 12, 16, 20, 24 or 32 ft run, 4.5 ft spacing needs exactly the same number of posts as 4 ft — you have to reach a 40 ft run before it saves one. Meanwhile 4.5 ft drops the rail's margin from 1.36× to 1.21×.
It cost margin and saved nothing, so we took the shorter span.
Our 4 ft maximum is measured along the slope, nosing to nosing — not horizontally. That is the tighter of the two readings and it is what the rest of the industry publishes. The code is more forgiving on a stair in every other respect: guard height drops to 34 in from the nosing line, the sphere opens to 4⅜ in, and the triangular opening at riser and tread may pass 6 in. The loads do not change.
Most of the railing market answers coastal exposure with powder-coated aluminium. So do we. What follows is the part the market usually leaves out: aluminium only works in salt air if three specific things are true, and we are naming all three.
The section grows, because aluminium is weaker than the steel6061-T6 aluminium has a 240 MPa minimum yield against Q355 steel's 355. So CoastLine's post keeps ProGuard's 1.97 in outer size — same base plate, same caps, same brackets — and takes a 3.0 mm wall against ProGuard's 2.0 to land at the same margin.
| Line | Post | Weight | 36 in residential | 42 in commercial |
|---|---|---|---|---|
| ProGuard | 50 × 50 × 2.0 mm, Q355D steel | 3.01 kg/m | 3.16 × | 1.34 × |
| CoastLine | 50 × 50 × 3.0 mm, 6061-T6 | 1.52 kg/m | 2.26 × | 1.29 × |
Same strength, half the weight. 1.52 kg/m against 3.01 — which a customer feels when they carry it up a flight of stairs, and which shows up in what shipping costs.
⚠ The alloy is not interchangeable, and this is where coastal aluminium usually goes wrong| Alloy | Min yield | At 50 × 50 × 3.0, 42 in commercial |
|---|---|---|
| 6063-T6 the cheapest, commonest extrusion alloy | 170 MPa | 0.91 × — fails |
| 6005A-T6 | 215 MPa | 1.15 × |
| 6061-T6 — what we specify | 240 MPa | 1.29 × |
6063 is what you get if nobody insists. It is the default extrusion alloy, it looks identical, and at our section it does not pass. The substitution is invisible without a mill certificate, which is why we require one per heat.
⚠ The cable top rail changes too — 3.5 mm in aluminium, not 2.5Aluminium's lower yield applies to the rail as much as the post. At 2.5 mm the aluminium cable rail carries only 0.95× — it fails. It needs 3.5 mm to reach 1.21×.
We found this with the calculator on this page after the material decision, not by inspection, which is rather the point of publishing one. And it is still the lighter part: 1.59 kg against the steel rail's 3.41 kg per 4 ft length. 6063-T6 never reaches 1.0 at any wall we would use.
Anodising is its own finish system (AAMA 611), not a pretreatment under powder. Its architectural palette is the champagne-to-bronze-to-black family. Black is achievable; repeatable matte black, batch to batch and against a replacement post bought three years later, is not — and that disqualifies it for a product line on its own. On corrosion, a finisher who runs both systems is blunt: "PVDF is relatively chemically inert and will outlast anodizing in corrosive environments."
Where our finish sits against the market| Class | Salt spray | Florida exposure | Who uses it |
|---|---|---|---|
| AAMA 2603 | 1,500 h | 1 yr | Superior Aluminum |
| AAMA 2604 | 3,000 h | 5 yr | Feeney · Viewrail's entry system, which they state sits outside their coastal warranty |
| AAMA 2605 — ours | 4,000 h | 10 yr | Viewrail for coastal only. Above every maker we checked. |
"The vast majority of powder coat failures can be put down to poor or inconsistent pre-treatment methods." That is QUALICOAT, the European architectural-coating body, and it is why our specification carries a number rather than a name: etch weight loss ≥ 2 g/m², the QUALICOAT "Seaside" class, recorded and reported per batch. "AAMA 2605" describes the powder and says nothing about what is beneath it — and a bad pretreatment under a good powder still fails.
We also permit one alternative: a 4–8 µm pre-anodise, left unsealed, powder coated within the anodiser's hold time. Both words matter. QUALICOAT credits it with "totally remov[ing] all the grain boundaries, so there are no places where filiform corrosion may take hold". A sealed anodic film under powder is the expensive wrong version of this.
Our cable holes cut the coating at the one place salt collects. So every cable, drain and mounting hole is machined, deburred and radiused to a minimum 0.5 mm before pretreatment — never after. A sharp edge or a burr cannot hold film thickness: "powder coating will not fully coat over burrs however slight."
Which is why you must never drill a CoastLine post on siteA hole drilled after coating is bare aluminium in the worst possible place. If site machining is genuinely unavoidable, the exposed metal must be sealed back to the powder coating with the coating manufacturer's approved touch-up before the post is installed.
A stainless cable through an aluminium post is a galvanic couple. ICC-ES ESR-4798 §5.5 is explicit that every metal in contact with aluminium must be an alloy approved for direct contact or isolated by an approved coating. Isolating bushings at every cable pass-through and every stainless fastener — specified, not left to the factory or to the installer.
And the honest part, which the industry's warranties generally do not say out loudRead the exclusions on any railing warranty before buying for a coastal site. It is common for the clause voiding cover for "salt spray, salt air… surface oxidation" to sit in the warranty of a product marketed for marine use, and for the identical clause to appear in that manufacturer's aluminium and stainless warranties alike. One major brand grades cover by distance from the water and drops aluminium to one year within a mile. At least one excludes aluminium posts paired with stainless cable by name.
We are not going to pretend aluminium is indifferent to salt. It is the right material here for weight, cost and corrosion behaviour — provided the finish and the isolation are right. That is why those are written into the specification rather than left to the factory, and why we ask you to rinse the railing with fresh water on a schedule. Salt that gets rinsed off does nothing; salt left to concentrate in a drilled hole is what causes pitting.
The obvious way to strengthen a rectangular rail is to make it deeper. We costed 2 × 1.5 in and withdrew it: at that size the graspable perimeter is 7.00 in, past the 6.25 in ceiling for a Type I handrail, which would force a finger recess along the whole length. Where a guard's top rail is also grasped as a handrail — which on a stair it is — that matters.
So the section stayed 2 × 1 in and the strength went into the wall instead. 2 × 1 in is the largest rectangle that stays comfortably graspable, which is why it is the market standard.
A 3/8 in bolt is 9.525 mm. A 10 mm hole leaves 0.475 mm of clearance — about 30% of what the steel code treats as standard for a bolted connection. These plates drop over anchors an installer drilled by hand, where real position error is several millimetres.
11 mm gives 1.48 mm of clearance, essentially the code standard, at no extra cost — the same punching operation with a marginally larger tool. It is not a strength decision, it is the difference between a plate that fits on site and a callback.
The route for a guard system is ICC-ES AC273, the acceptance criteria an evaluation report is written against. It requires physical testing of the specific assembly — post, bracket, base plate and anchorage together, not the parts on their own.
| Member | Deflection limit | At 36 in | Where we sit |
|---|---|---|---|
| Post | h / 12 | 3.00 in | ProGuard 0.20 in · EdgeLine 0.54 in — both far inside |
| Rail, midspan | h / 24 + l / 96 | — | Inside on every section we sell |
Deflection is the easy half and we clear it comfortably. The hard half is that the report tests the whole assembly with its anchorage, and it names the exact post wall, base plate and fastener schedule — which is why the two open items on this page (EdgeLine's steel grade, and confirming our post wall against a mill certificate) are the real blockers, not the span numbers.
What we HAVE done — one truck, on videoWe pulled a guard rail section apart with a truck and filmed it. It is a real demonstration and it is worth watching — it shows what the assembly takes before anything lets go.
It is not a certified test, and it is not something we do to every section. A test that counts for a code report is run on a calibrated rig, to a written protocol, by an accredited lab, on the specific assembly including its anchorage — and it produces numbers an official can rely on. One truck in a yard produces a video. We would rather show you the video and the arithmetic, and call each of them what it is.
Where we areHonestly: we are not there yet, and the span figures we publish are engineering, not ratings. We publish the arithmetic so you can check it, we name the code clause behind every requirement, and we state the open questions in the same document as the answers. If that matters to you more than a number on a table, that is the whole reason this page exists.
RelatedRailing specifications & documentation — the hub, with every page in this set.
Specifications by line: EdgeLine · ProGuard · CoastLine
Guides: Guard rail installation · Cable railing installation · Engineering & code compliance
Full product specifications · All instruction guides
Questions an installer or an inspector needs answered that these pages do not cover? Contact us — we would rather answer it than have you guess.