For 3.0 mm architectural PVDF aluminum veneer, GB/T 23443-2009 sets a minimum pencil scratch hardness of ≥1H, with the second-coat liquid film averaging ≥30 µm and the third coat averaging ≥40 µm [S2]. For coil and sheet product, suppliers typically publish a working range of 1H–2H for the same film system [S1]. The single most common failure is buying a panel that passes 1H on one test rig and fails 2H on another, because the underlying method, not the grade, was never pinned.
This is a method-bound property, not an absolute. Pencil hardness numbers quoted on a datasheet are only as defensible as the test method printed next to them, so the rest of this guide lines up the standards, the equipment, and the pitfalls a spec writer has to lock down before signing a PO.
Applicable Standards: ASTM D3363, ISO 15184, GB/T 6739, GB/T 23443
The Wolff-Wilborn pencil hardness test is the dominant method for coil and architectural PVDF coatings, formalised in 1974 as ASTM D3363 [S7]. The ISO equivalent is ISO 15184:2020, "Paints and varnishes, Determination of film hardness by pencil test" [S8]. In China, GB/T 6739-2006 carries the same method body, and GB/T 23443-2009 "Aluminum Veneer for Architectural Decoration" is the umbrella product standard that sets the acceptance threshold [S2]. For North American high-end architectural work, AAMA 2605 calls out no rupture of the film at the "F" pencil level under ASTM D3363 [S4].
Methodology alignment matters: GB/T 6739-2006 specifies a push speed of 0.5–1 mm/s over a minimum 7 mm stroke, while ISO 15184:2020 calls for a slow, constant speed over a sufficiently long period without a fixed stroke [S5]. Both are nominally "the same test", but a lab running GB/T 6739 and a lab running ISO 15184 can disagree by half a pencil grade on the same panel, which is the difference between passing and failing a 1H callout.
Hardness Range, Equipment, and Test Geometry
The standard pencil set runs 14 grades from 6B (softest) to 6H (hardest), and Elcometer's 3080 free-hand kit is a typical field set bundled with the test method [S3]. A scratch is recorded when a visible defect ≥3 mm long appears under the stroke [S5]. The reported hardness is the highest grade that does NOT produce a 3 mm scratch, which is why data sheets read like "≥1H" or "1H–2H" rather than a single number [S1][S2].
Test load is part of the result. A manual Wolff-Wilborn rig from BEVS (model 1301) and the equivalent electric tester (BEVS 1309) both accept 500 g, 750 g, and 1,000 g counterweights, with the electric version holding speed between 0.5 and 8.5 mm/s to remove operator drift [S5]. A 750 g load at 1 mm/s is the working point most coil-coat labs default to; quoting "2H pencil hardness" without naming the load and speed is not a complete spec.
What a PVDF System Has to Deliver at 1H–2H
A 1H–2H rating is a system result, not a topcoat result. The Worthwill coil product reaches that range on a 25 µm two-coat or 34 µm three-coat PVDF system, with the topcoat formulated at ≥70% fluorocarbon resin by weight [S1]. GB/T 23443-2009 layers the requirement further: a 2-coat PVDF liquid system averages ≥30 µm with local film ≥25 µm, and a 3-coat system averages ≥40 µm with local film ≥34 µm [S2]. For comparison, the automotive multi-coat panels used in the BEVS round-robin test were 60–100 µm total (primer 15 µm, base 10–25 µm, topcoat 10–15 µm, clear coat 25–45 µm), and they routinely grade out at the soft end of the pencil scale [S5].
That thickness difference is the spec writer's friend: a properly built 3-coat PVDF coil at ≥34 µm local film has the resin volume to absorb a 1H scratch without exposing the primer, while a 2-coat 25 µm film is living closer to its hardness margin. AAMA 2605's "no rupture at F" sits one grade softer than the typical 1H Chinese-architectural callout, reflecting the difference between high-build extrusion coating and architectural spray [S4].
Comparison: Coil PVDF vs Spray PVDF vs Powder PVDF on Hardness
Three commercial routes hit the 1H bar with different margin. Coil-coated PVDF (Worthwill-type continuous roll, 25–34 µm, ≥70% PVDF resin) lands at 1H–2H on Wolff-Wilborn and is the workhorse for aluminum veneer panel cladding [S1]. Spray-applied liquid PVDF to GB/T 23443-2009 (2-coat ≥30 µm, 3-coat ≥40 µm average) must hold ≥1H and is the default for bespoke aluminum window door trims and shaped facade elements [S2]. Powder PVDF skips the solvent route and clears the same 1H callout with a minimum local film of 30 µm, which is the option to specify when VOCs or factory re-coat logistics drive the decision [S2].
On a four-criterion comparison, the three systems line up roughly as: coil PVDF offers the fastest lead time and the broadest colour range but the thinnest film; spray PVDF offers the deepest custom colour and field repair but the highest film-to-film variation; powder PVDF offers the cleanest environmental profile and a single-layer thickness reading but a narrower gloss and effect range. Pencil hardness is the one criterion where all three meet the same 1H floor; the differentiation lives in adjacent tests, especially impact (coil routinely survives a 9 J reverse impact with no cracking or peeling) and salt spray (1000 h acetic acid with no creep beyond 2 mm from a scribe) [S1][S2].
Common Pitfalls: Cure State, Substrate, and Operator Drift
Under-cured PVDF is the most common reason a panel fails a 1H callout in the field, and pencil hardness is widely used as a quick proxy for cure state on the topcoat [S4]. A panel that reads 2H in the lab and B–HB on site has almost always been exposed to wet shipment, cold storage, or premature stacking, all of which starve the fluorocarbon crosslink. The AAMA 2605 "F hardness no rupture" line and the GB/T 23443 ≥1H line are both end-state targets, not "as-shipped" targets, so site verification has to allow a cure window before measuring [S2][S4].
Substrate choice quietly moves the result. Hardness tests on aluminum alloy substrates in 3003, 3004, 3005, 3105, 5005, and 5052 tempers H12–H44 are routine, but a 1060-O substrate is soft enough to flex under the pencil, which can mask a marginal topcoat as a "pass" [S1]. Operator drift is the third pitfall: a manual test run by different technicians on the same panel can swing half a grade, which is why electric testers with fixed 500 g / 750 g / 1,000 g loads and controlled traverse speed are increasingly used for incoming-inspection work on hardness tester benches [S3][S5].
What to Write into the Spec and How to Verify on Site
A defensible clause reads: "PVDF topcoat pencil hardness ≥1H per ASTM D3363 (Wolff-Wilborn), 750 g load, 1 mm/s traverse speed, 14-pencil set 6B–6H, with no surface rupture at the reported grade; acceptance per AAMA 2605 for extrusion-coated aluminium or GB/T 23443-2009 for spray-coated architectural veneer." That sentence locks the method, the load, the speed, and the acceptance document in one go, and it survives both a Chinese and a North American submittal review [S1][S2][S4].
For on-site verification, the Elcometer 3080 free-hand kit plus a portable hardness tester rig is the practical toolset: prep the surface with an alcohol wipe per GB/T 9271-2008, run three strokes per grade starting from the soft end, and record the highest grade with no ≥3 mm scratch [S3][S5].
For related coverage, see Laser Screed vs Vibratory Screed: Decision Map for Large Slab Flatness.