Coating Deformation Tests: Cup Forming, Mandrel Bend & Pendulum

2026-11-14 · Category: Technical Knowledge

🌐 This article was automatically translated from Chinese. Please refer to the original Chinese version if needed. · View original (Chinese)

Coating Deformation Tests: Cup Forming, Mandrel Bend & Pendulum

Coating deformation tests are a family of empirical methods used to assess how well the coatings of paint, clearcoat and related products resist cracking, and resist detaching from the substrate, as they deform together with the substrate under controlled deformation. Under the current coating test standards system, this family comprises the Erichsen cup test (ISO 1520, GB/T 9753), the cylindrical mandrel bend test (ISO 1519, GB/T 6742), the cone mandrel bend test (ISO 6860, GB/T 11185), the T-bend test (ISO 17132, GB/T 30791) and the pendulum damping test (ISO 1522, GB/T 1730); they share one premise — the deformation is applied by a standard apparatus, and the result is valid only under that apparatus and substrate condition, so a conclusion cannot be transferred across conditions.

Real failure scenarios recur throughout manufacturing: a bent part passes 100% inspection in the plant, yet three weeks after delivery a network of fine cracks appears on the outer radius; a coil-coated panel shows bare metal at the edges after press-forming and drawing; a chassis panel, after being clamped during assembly, turns white at the bend with interfacial delamination. In outgoing inspection all three are recorded as "adhesion qualified", yet no deformation-type test covered the strain that actually occurred in service. An in-site page on coating impact-resistance testing already covers the drop-weight fast-deformation case; this article focuses on the other three static-loading routes — cup forming, cylindrical mandrel bend, and pendulum damping — and makes clear "what is measured, how it is judged, how the report is written, and whether one method can replace another".

1. Fixing the concepts first: deformation tests, impact, indentation and cross-cut are four different things

The introduction to ISO 1520:2006 states this family explicitly as six mutually complementary empirical test standards: ISO 1519 (cylindrical mandrel bend), ISO 1520 (cup test), ISO 6272-1 and ISO 6272-2 (drop-weight fast deformation), ISO 6860 (cone mandrel bend), ISO 17132 (T-bend). Their common feature is being "empirical" — the result is not an intrinsic material parameter but the fact that "the coating did not crack on this panel, at this film thickness, with this apparatus". The engineering implication of use is a single one: align the loading mode with the in-service deformation mode, and only then does the criterion have predictive power.

Pendulum damping (ISO 1522) does not belong to these six "crack-resistance / peel-resistance" standards; it measures amplitude decay and is a viscoelastic response. It is listed alongside cup forming and bend in this article because purchasing documents often write "hardness" and "conformability" together in the same technical requirement line, whereas these three measure fundamentally different physical quantities. The note to ISO 6272-1:2011 further points out that the apparatus produces fast deformation rather than true impact, and that the drop-weight method and static bend or static indentation bear no conversion relationship.

2. TL;DR: eight directly citable conclusions

  • The cup test is run per ISO 1520:2006 / GB/T 9753-2007, with a hemispherical punch of 20 mm diameter moving at 0.1 mm/s to 0.3 mm/s; the conclusion is "whether cracking occurs when pressed to a given depth under biaxial tension", and it cannot be inferred from or to uniaxial bend.
  • The cylindrical mandrel bend is run per ISO 1519:2011 / GB/T 6742-2007; the Type I mandrel-diameter series is 2 mm to 32 mm in 12 steps, tolerance ±0.1 mm; Type I applies to panel thickness not exceeding 0.3 mm, Type II to thickness not exceeding 1.0 mm.
  • Pendulum damping is run per ISO 1522:2022 (the current national standard is still GB/T 1730-2007, adopting ISO 1522:1998); König reads 6° to 3° and Persoz reads 12° to 4°, and what is measured is decay time or number of oscillations, not indentation depth.
  • The 25 µm and 60 µm film thicknesses of the same set belong to two criterion points: the ASTM D522 film-thickness correction factor is added per 1 mil (25 µm), being 1.40 at the 3.2 mm mandrel step and only 0.21 at the 25 mm step.
  • Deformation tests judge the conformability of the coating body; interfacial bonding belongs separately to the cross-cut (ISO 2409 / GB/T 9286-2021) and pull-off (ISO 4624 / GB/T 5210) methods; a single cross-cut report cannot release a deformation requirement.
  • Version gaps must be named in the technical agreement: GB/T 6742-2007 adopts ISO 1519:2002 identically, while the current international edition is ISO 1519:2011; GB/T 9753-2007 corresponds to ISO 1520:2006, and the two countries' normative references also differ in edition.
  • Conditioning follows GB 9278 / ISO 3270:1984 standard atmosphere 23 ℃ ± 2 ℃, relative humidity 50 % ± 5 %; pendulum data whose report omits the measurement time point has no comparability during the stage when curing is incomplete.
  • Simple field mandrel bending, and visual-plus-magnifier inspection after folding, are for trend judgement and incoming quick-screening only, not a basis for release; the deviation of these from laboratory data has not been publicly qualified and is empirical reading.

3. What each of the three routes measures: the stress state sets the boundary of the conclusion

Table 1 Deformation-type method family and what it can / cannot judge
MethodStandard number and current editionStress stateResult expressionWhat it cannot replace
Cup test (Erichsen-type pressing)ISO 1520:2006; GB/T 9753-2007 (IDT)Near-biaxial tension, the dome's outer side stretchedPass / fail at a specified depth, or the critical pressing depth, in mmCannot replace the uniaxial-bend judgement of an edge condition
Cylindrical mandrel bendISO 1519:2011; GB/T 6742-2007 (IDT ISO 1519:2002)Uniaxial bend, outer side in tension, inner side in compressionPass / fail at a specified mandrel diameter, or the first diameter at which cracking appearsCannot replace multi-directional forming (drawing, flanging) requirements
Cone mandrel bend / T-bendISO 6860:2006, ISO 17132:2007; GB/T 11185-2009, GB/T 30791-2014Small-curvature uniaxial folding, coil uncoiling and hemming conditionsPass / fail, or expressed as T-count (1T, 2T)Cannot replace thick-plate bending and deep-drawing conditions
Drop-weight fast deformationISO 6272-1:2011 (large area), ISO 6272-2 (small area); GB/T 1732-2020High-rate deformation, includes strain-rate effectsPass / fail at specified mass and drop height, or the cracking thresholdNot convertible to static bend; the standard itself notes it is not true impact
Pendulum dampingISO 1522:2022; GB/T 1730-2007 (adopting ISO 1522:1998)Viscoelastic energy dissipation under small rolling contactDecay time (seconds) or number of oscillations, must state the swing typeCannot characterise crack resistance, nor be used as indentation hardness

3.1 Cup test: the critical pressing depth under biaxial tension

What ISO 1520:2006 specifies is an empirical procedure: after the coating is progressively deformed by pressing under standard conditions, its resistance to cracking or to detaching from the metal substrate is assessed. The method has two branches — perform a single test at an agreed pressing depth and judge "pass / fail", or increase the pressing depth step by step and measure the critical depth at which the coating first shows cracking or first begins to detach from the substrate. A multi-coat system can be measured layer by layer or as a whole, and this choice must be stated in the report, otherwise the same "no cracking at 6 mm" might come from a single topcoat layer or from a complete primer-plus-topcoat system.

3.2 Cylindrical mandrel bend: the conclusion is tightly bound to the mandrel diameter

The wording of ISO 1519 is "assessment of the resistance of a coating to cracking and/or detachment from a metal or plastic substrate when bent around a cylindrical mandrel". It too has two branches: use a mandrel of specified diameter for a go/no-go, or change successively to smaller mandrels to find the first diameter that causes the coating to crack or detach. The mandrel diameter here is not a "smaller is stricter" parameter but the loading condition itself — the same coating differing by more than one step in conclusion on a 2 mm mandrel versus a 20 mm mandrel is a normal physical result, not quality variation.

3.3 Pendulum damping: viscoelastic decay, not indentation depth

The pendulum method places a rod with two steel balls on the coating surface; after release it measures the time or number of oscillations needed for the amplitude to decay from the start angle to the end angle. The softer the coating, the faster energy is dissipated viscoelastically and the sooner the decay ends. It belongs to a different family from the Buchholz indentation (GB/T 9275), pencil hardness (ISO 15184, corresponding to pencil-hardness testing and baking-paint control) and Taber abrasion (wear resistance and hardness), with no conversion between them; when citing data across methods, they can only be listed side by side, not substituted for one another.

4. How to run the cup test: apparatus dimensions, loading rate and parallel determinations

The key instrument points given in GB/T 9753-2007 (identical adoption of ISO 1520:2006): the part of the punch that contacts the panel is a quenched, polished steel hemisphere of 20 mm diameter; during the test the punch should move at a steady speed of 0.1 mm/s to 0.3 mm/s, and the offset of the punch-sphere centre from the die axis should not exceed 0.1 mm, with rotation prevented. Published instrument data further give the mating dimensions: die bore about 27 mm and outer diameter about 55 mm, hold-ring bore about 33 mm, depth read-out resolution 0.01 mm; panels are mostly steel 0.3 mm to 1.25 mm thick with a side length not less than 70 mm. These dimensions are of "published instrument-data status"; when procuring equipment, follow the accompanying technical documentation.

Hand-cranked loading on a manual cup tester can hardly keep a constant rate over the whole stroke, and the thicker the panel the more pronounced; this constraint is exactly the phrase "steady speed" in the standard text. The industry therefore uses automatic constant speed (commonly a nominal 0.2 mm/s) as the reproducibility improvement, for arbitration and batch release; trend judgement and development screening may still use a manual machine.

The number of parallel determinations is also a version difference: one of the main changes of EN ISO 1520:2006 relative to the 2001 edition is to change three determinations to two, and to add precision data and improve the precision of result evaluation (ÖNORM foreword status). The way a criterion is written therefore needs to make explicit whether it is "neither of two determinations cracks" or "the arithmetic mean of two determinations meets the value"; the burden of proof differs entirely between the two in a dispute.

4.1 The five fields a cup-test report must state clearly

An arbitrable cup-test report contains at least: (1) the punch diameter and loading rate (20 mm, the actual value within 0.1 mm/s to 0.3 mm/s); (2) the substrate grade, condition and thickness; (3) the coating system and the measured film thickness of each layer (declared by the ISO 2808 method); (4) the judgement mode (pass / fail at a specified depth, or critical depth in mm) and the observation magnification; (5) the conditioning temperature and humidity and the ageing time after panel preparation. Without any one of these, the data may be used only for internal trend, not written into an external technical agreement.

It must also be distinguished from the sheet-metal forming cup test: the Erichsen cupping value (IE) for assessing bulge forming of thin sheet and strip is run per GB/T 4156 / ISO 8490, the object being the metal itself rather than the coating, with values rounded to 0.1 mm. A purchasing line such as "cup test ≥ 6 mm" that fails to state the object and method standard may well mean two different tests for supplier and buyer.

5. Cylindrical mandrel bend: diameter series, thickness limits and version gaps

The Type I apparatus given in EN ISO 1519 specifies a complete set of hinge mandrels with diameters 2 mm, 3 mm, 4 mm, 5 mm, 6 mm, 8 mm, 10 mm, 12 mm, 16 mm, 20 mm, 25 mm and 32 mm, tolerance ±0.1 mm; the gap between the mandrel and the hinge base plate is (0.55 ± 0.05) mm, the mandrel should turn freely about its own axis, and the apparatus needs a locating block to ensure the two plates are parallel after bending; the mandrels are made of hard, corrosion-resistant material (for example stainless steel). Type I is for panels not thicker than 0.3 mm, Type II for panels not thicker than 1.0 mm; the standard also states that the two apparatuses give similar results for the same coating, and a specific product usually uses only one of them. The text specially warns that when using the 2 mm mandrel you must confirm the mandrel has not deformed during bending; a deformed mandrel must not continue in use.

Table 2 Cylindrical mandrel bend apparatus and criterion elements
ElementType I apparatusType II apparatusHow to write it in the report
Applicable panel thicknessNot exceeding 0.3 mmNot exceeding 1.0 mmWrite the measured value and grade, not "standard panel"
Mandrel diameter2 mm to 32 mm, 12 stepsSelected per the apparatus instructionsWrite the actual pass / fail step, note tolerance ±0.1 mm
Judgement modego/no-go single mandrel, or step down to smaller mandrelsSame as leftWrite "passed D=xx mm" or "first cracking mandrel D=xx mm"
Extrapolation of the conclusionValid only under the same film thickness, substrate and cure conditionSame as leftList these three conditions together in the technical agreement

5.1 Bend duration and observation timing: the difference between ISO and ASTM status

Published operating material generally describes the ISO 1519 bending action as "smoothly bending the panel through 180° within 1 s to 2 s, coating outwards", stressing that no impact may be used. Method B of ASTM D522 is more explicit: bend around the mandrel about 180°, at a uniform speed completed within 1.5 s to 5 s, the specific duration agreed between supplier and purchaser, then inspect immediately by naked eye for cracking; if no cracking, switch to a smaller mandrel diameter and repeat in an untested area, until failure or until the smaller mandrels are exhausted; this standard takes "the mandrel diameter at which cracking stops" as the coating's crack-resistance value. Both statuses are correct, but the duration itself is a loading condition: citing an ASTM 5 s result as an ISO 1 s result amounts to covertly changing the strain rate.

5.2 Locating the cracking source: coating crack, substrate crack first, or interfacial delamination

After bending, the same crack may arise from three mechanisms: coating tensile fracture (cracking within the film, often fine transverse lines), the substrate yielding or cracking first (the crack running through into the metal, persisting when film thickness changes), and interfacial delamination (the coating separating from the substrate, common with a pretreatment or primer mismatch). Judgement means, ordered by rising cost, are naked eye, magnifier (10×) observation, cross-section or fracture-face microscopy, and dye-penetrant assistance. If a substrate contribution is suspected, a bare panel should be bent under the same condition as a control. Published-source status also notes that in a material cup test, if cracking starts at the clamped edge or the specimen slips, that data point is void — this logic equally governs the acceptance or rejection of edge data in coating tests.

6. Converting mandrel diameter into coating strain: how to read the ASTM D522 elongation table

The mandrel diameter is not itself strain, but strain can be inferred from it. The published text of ASTM D522 gives two tools. The first is the geometric formula for outer-surface strain of a flat panel: outer-surface elongation = 100 t / (D + t), where t is the panel thickness and D is the cylindrical mandrel diameter. By this formula, a 0.8 mm cold-rolled panel bent over a 3.2 mm mandrel has an outer surface of about 20 %; a 0.3 mm panel over a 2 mm mandrel is about 13 %. This is the quantitative basis for "smaller mandrel is stricter": each step down in radius of curvature raises the tensile strain borne by the coating by a notch.

The second is the tabular method of Method B: first take, from Table 1, "the elongation corresponding to each mandrel diameter for a 1 mil (25 µm) coating on a 1/32 in (0.8 mm) cold-rolled steel panel", then add the measured film thickness using the per-mil correction factors of Table 2, total elongation E = e1 + t × c1. The published values of Table 1 and Table 2 are in the table below.

Table 3 ASTM D522 Method B mandrel–elongation–film-thickness correction (public standard-text status)
Mandrel dia. in (mm)Table 1 elongation e1, %Table 2 correction c1 (per mil)Geometric formula 100t/(D+t), t=0.8 mm
1 (25)3.30.212.4
3/4 (19)4.40.263.0
1/2 (12.7)6.750.384.5
3/8 (9.5)9.00.506.2
1/4 (6.4)14.00.719.5
1/8 (3.2)28.01.4020

The two tools serve different purposes: the tabular method gives the strain the coating actually bears (including the film thickness's own contribution), the geometric formula gives the outer-surface strain of the substrate, and the two differ markedly at small mandrel diameters and thick panels — the standard text itself offers a comparison discussion that "a difference exists between calculated and true values". In engineering, cite by the method agreed in the contract, and do not list the two sets of figures as one indicator. Substituting 25 µm and 60 µm shows why film thickness is part of the criterion: at the 3.2 mm step c1 = 1.40, each 1 mil (25 µm) increase in film thickness adds about 1.4 % strain; at the 25 mm step c1 = 0.21, the same increase adds only about 0.2 %. That is, the same conclusion of "passing the 3.2 mm mandrel" is harder to achieve at 60 µm film thickness than at 25 µm — increasing film thickness does not automatically buy conformability.

Deformation cracking must be separated from in-service cracking: cracks visible immediately after bending are caused by the loading; craze-cracking and peeling appearing only after years of use mostly result from internal stress, excessive film thickness and UV ageing combined — see the in-site page on internal stress and excessive film thickness in outdoor coating cracking and peeling. The internal-stress–adhesion–film-thickness triangle on the hard-coating side is covered on the page internal stress and film-thickness limits of DLC coatings.

7. Pendulum damping: the difference between König and Persoz lies in the amplitude window

The current edition of ISO 1522 is the 2022 publication (EN ISO 1522:2022, the DIN-adopted edition DIN EN ISO 1522:2023-02). Relative to the 2007 edition, the official change note lists the main edits as: adding a terms clause; removing the "metal or glass plate" qualifier from the substrate recommendation, because the method equally applies to substrates such as plastics; revising the precision of the König pendulum based on 2006 inter-laboratory comparison results; restoring the tolerance for the time required for the amplitude to decay from 12° to 4° to the ISO 1522:1998 second-edition status; adding a check of rod deformation to the apparatus clause and calibration appendix; editorial revisions and updated normative references. China's current method standard is GB/T 1730-2007 "Paints and varnishes — Determination of resistance to indentation by the pendulum method", an identical adoption of ISO 1522:1998 — this means that if a domestic report writes "per GB/T 1730", its apparatus tolerance and precision cite the 1998 edition status and are not fully equivalent to ISO 1522:2022; cross-border delivery must name the edition in the technical agreement.

Table 4 König vs Persoz pendulum apparatus parameters (ISO 1522 / ASTM D4366 apparatus public-data status)
ParameterKönig pendulumPersoz pendulum
Pendulum mass200 g ± 0.2 g500 g ± 0.1 g
Contact-ball diameter5 mm, two balls8 mm, two balls
Oscillation period1.4 s ± 0.02 s1 s
Amplitude read windowDecays 6° to 3°Decays 12° to 4°
Glass reference-plate damping time250 s ± 10 s430 s ± 10 s
Suitability tendencySensitive to early surface cure and baking progressMore robust for high-elasticity, thicker coatings
Results not comparable toIndentation hardness, pencil hardnessCross-cut adhesion grade

7.1 Why damping time cannot be taken as an externally promised "hardness"

The pendulum reading is affected simultaneously by film thickness, substrate stiffness, curing sequence and surface contamination. The reference value 250 s (König) on glass is for calibration checks only and does not represent a coating meeting a specification; the damping time of the same formulation at 30 µm and 80 µm film thickness does not belong to the same distribution, and citing across film thickness must be noted. A sounder external wording is the four elements "method + swing type + read window + time after panel preparation", for example "König, 6° to 3°, 168 h after panel preparation, 23 ℃ ± 2 ℃ / 50 % ± 5 % RH, measured xx s". The relationship between crosslink density and damping time in UV systems is on the in-site page UV-curing crosslink density.

8. Film thickness, substrate and system coupling: why the same judgement cannot be transferred

Deformation tests are sensitive to three variables, and a change in any one voids the conclusion: film thickness (including its distribution across layers), substrate (grade, thickness, yield behaviour, surface roughness), and system (primer-to-topcoat modulus gradient, degree of cure). Tinplate is uniform in texture and transmits bending clearly, hence the traditional first choice; cold-rolled steel is harder and needs greater bending force; aluminium is soft and low in modulus, deforms in a different mode when bent, so "the coating does not crack" and "the substrate yields first" are easily confused.

The practical approach to locating the cracking source uses all three together: use a magnifier to see whether the crack lies only within the coating (not running through to the metallic-lustre face); use a cross-section or fracture to judge whether the substrate cracked first; use dye-penetrant or solvent colouring to judge whether delamination occurs at the coating-substrate interface or between layers. Once judged, the report conclusion must distinguish "coating crack resistance failed" from "substrate or pretreatment failed", because the corrective directions are entirely different (the former changes formulation and cure, the latter changes pretreatment and primer).

Primer-topcoat matching and toughening thinking: a topcoat's high crosslink density brings gains in damping time and scratch resistance while compressing ultimate elongation; the primer provides adhesion and cushioning, and when the modulus gradient is too steep deformation concentrates at the interface. In coil systems the T-bend (1T, 2T, 3T) and cup test are often listed together; the division of labour is on the in-site page coil-coating three-layer system and T-bend; the deformation and stone-chip test items on the vehicle side are listed under car-paint test standards, whose division of labour with this article is: those two pages list the item inventory, while this article gives the method and how to write the criteria.

9. The causal boundary with pretreatment and adhesion

Incomplete degreasing or an abnormal conversion coating amplifies edge delamination after bending, but the cup test and cylindrical mandrel bend judge the ability of the coating body to bear deformation; the cross-cut (ISO 2409 / cross-cut test and surface roughness, GB/T 9286-2021, see also an in-depth analysis of the cross-cut method) and the pull-off method (ISO 4624 / GB/T 5210, criterion wording in pull-off adhesion acceptance) judge the interface. The sequence of the two test families in batch release cannot be reversed: first confirm with the interface test that the bonding basis holds, then evaluate conformability with the deformation test; releasing a bending condition on a single cross-cut report is a contractual risk borne by both procurement and supply sides.

Cross-cut lattice and adhesion testing: the boundary of division of labour between interface-type and deformation-type tests

Table 5 Division of labour between the deformation family and the interface family in the release process
DimensionDeformation family (cup / bend / T-bend)Interface family (cross-cut / pull-off)Report-field differences
What is judgedThe ability of the coating body to deform with the substrateThe bonding strength between coating and substrate or between layersThe deformation family records diameter or depth; the interface family records a grade or MPa
Loading rateStatic or completed within an agreed durationStaticBoth must record rate or duration
Failure locationCracking within the film, top of the domeInterface, notch intersectionThe failure location should be photographed for the record
Release orderPerformed after the interface conclusion holdsFirstState clearly the order of the two families and the linked batch number

10. Conditioning, ageing and measurement time point

The method standards of the deformation-type tests all link conditioning out to the environment standard (GB 9278 / ISO 3270:1984) and to the panel-preparation and thickness standards (ISO 1514, ISO 2808, ISO 1513, ISO 15528). The standard atmosphere is usually 23 ℃ ± 2 ℃, relative humidity 50 % ± 5 %. Note the version gap: GB/T 9753-2007 and GB/T 6742-2007 reference the ISO 2808:1997 and ISO 1513:1992-era texts, whereas the current ISO 2808 was published in 2019 (covering wet film, dry film and uncured powder coatings, plus Appendix B for rough surfaces and Appendix C for wood substrates); on the drop-weight side, ISO 6272-1:2011 relative to its predecessor also removed the specification for relative humidity, changing it to recording humidity and limiting the time to complete the test after conditioning. If one batch of data mixes reference files of different eras, the tolerances of the thickness method and of conditioning will not agree.

A laboratory scene of conditioning and testing coating performance

The effect of ageing on results concentrates in two places. First, residual solvent and insufficient flash-off show as "whitening, blistering, edge delamination" during bending or pressing, which is hard to distinguish from a deformation failure caused by under-cure, so reading must first examine the process window — the mechanism of solvent retention and the flash-off window are on the in-site page pinholes and boil-ups: solvent entrapment and the flash-off window. Second, pendulum damping keeps drifting with the curing sequence, so a single reading without a time-after-panel-preparation can only serve as a trend; for batch release it is advisable to fix one step, "7 d after panel preparation" or "curing-line exit + x h", and write it into the agreement. When field and laboratory conditioning conditions differ, the report must record both sets of temperature and humidity side by side.

At the acceptance level, the film-thickness, pinhole and missed-defect criteria appear alongside the deformation criteria of this article on the in-site page coating project acceptance specification: the former governs field delivery, the latter governs laboratory conformability; when both sets of criteria are cited at once, the batch number, film-thickness datum and conditioning conditions must agree, otherwise the two reports do not describe the same application condition.

11. Key data at a glance (every line with units or a standard number)

  • Cup-test punch: hemispherical, 20 mm diameter; movement speed 0.1 mm/s to 0.3 mm/s; offset of the sphere centre from the die axis not exceeding 0.1 mm (ISO 1520:2006 / GB/T 9753-2007).
  • Cup-test panel: thickness 0.3 mm to 1.25 mm, side length not less than 70 mm; die bore about 27 mm and outer diameter about 55 mm, hold-ring bore about 33 mm (published instrument-data status).
  • Cylindrical mandrel bend Type I diameters: 2, 3, 4, 5, 6, 8, 10, 12, 16, 20, 25, 32 mm, 12 steps in all, tolerance ±0.1 mm; gap between mandrel and base plate (0.55 ± 0.05) mm (EN ISO 1519).
  • Thickness upper limit: Type I not exceeding 0.3 mm, Type II not exceeding 1.0 mm; the two apparatuses give similar results for the same coating, and a specific product usually uses only one (ISO 1519:2011).
  • ASTM D522 Method B: bend duration 1.5 s to 5 s and agreed between supplier and purchaser; the six mandrel diameters are 25, 19, 12.7, 9.5, 6.4, 3.2 mm; Table 1 elongation 3.3 % to 28.0 % (1 mil coating, 0.8 mm cold-rolled panel).
  • Film-thickness correction: Table 2 factors 0.21 to 1.40 per mil, 1 mil = 25 µm; the 3.2 mm step corrects by 1.40, the 25 mm step by 0.21.
  • König pendulum: 200 g ± 0.2 g, ball 5 mm, period 1.4 s ± 0.02 s, reading 6° to 3°, glass reference 250 s ± 10 s; Persoz pendulum: 500 g ± 0.1 g, ball 8 mm, period 1 s, reading 12° to 4°, glass reference 430 s ± 10 s.
  • Pendulum versions: ISO 1522 current 2022 edition; GB/T 1730-2007 adopts ISO 1522:1998; the national adopting editions of the cup and bend tests also lag the current international editions by one step (GB/T 6742-2007 corresponds to ISO 1519:2002).
  • Conditioning: standard atmosphere 23 ℃ ± 2 ℃, relative humidity 50 % ± 5 % (GB 9278 / ISO 3270:1984); from ISO 6272-1:2011 humidity becomes a recorded value with a limited completion time after conditioning.
  • Parallel determinations: EN ISO 1520:2006 relative to the 2001 edition changed from 3 to 2 determinations; differences in normative-reference editions also appear between ISO 2808:2019 (current) and ISO 2808:1997 (national adopted status).

12. Low-cost field reading: alternative routes for trend and incoming screening only

Without a cup tester or standard mandrels, production lines commonly use three substitutes: simple bending with a set of known-diameter rebar or mandrels, folding a panel and inspecting by eye plus magnifier, and cross-cutting first then bending to see whether the cut edge propagates. Their value is quickly separating "obviously brittle" from "obviously tough", for incoming screening, batch trend and anomaly warning. Their limitations are equally clear: the loading condition is uncontrolled (rate, curvature and clamping cannot be written to a standard status), and the observation scale depends on the human eye, so they cannot be a basis for release, nor should they be written into an external technical agreement.

On-site visual and magnifier inspection of the surface coating of a large component

Table 6 Laboratory method vs field quick screen: what can / cannot be judged
Field practiceWhat it can readWhat it cannot readElements that must be recorded
Simple mandrel bendingWhether obviously brittle, whether batches driftA lower mandrel-diameter limit and an arbitrable conclusionMeasured mandrel diameter, bend duration, observation magnification
Fold then eye + magnifierMacro cracking and whiteningMicro-cracks, onset of interfacial delaminationPanel thickness, number of folds, lighting condition
Bend after cross-cutCombined deterioration of interface and deformationAttributing to adhesion or conformability aloneCross-cut grade, bend mandrel diameter (note if non-standard step)
Field non-destructive assistanceFilm thickness, defects and delaminationConformability itselfInstrument model and calibration validity, method in coating non-destructive testing

The magnitude of the deviation between alternative routes and laboratory data, where no public source can be verified, may only be marked as "empirical reading, no quantitative criterion"; when quantification is needed, samples should be sent to a third party for retest per the aforementioned method standards, or the two parties agree a comparison test. This annotation is itself part of the contract language, to keep field quick-screen records from being treated as a certificate of conformity.

13. Frequently asked questions (FAQ)

Q1: The customer requires "no cracking at cup 6 mm" — what must be confirmed before quoting?

Confirm four items: the punch diameter and loading rate (the actual values within 20 mm, 0.1 mm/s to 0.3 mm/s), the substrate grade and thickness (0.3 mm versus 1.25 mm makes a large difference), the judgement mode ("pass / fail at a specified depth" or "step up to find the critical depth"), and the observation magnification and number of parallel determinations (the 2006 edition status is two). If any of these four differs, "6 mm" is not the same requirement; writing them into an annex of the technical agreement is more convenient than a single line in the main text.

Q2: The same system passes on a 2 mm mandrel but cracks on a 1 mm mandrel — does that count as failing?

It depends on whether the agreement specifies "pass at D = 2 mm" or "step down to find the first mandrel diameter at which cracking appears". The two modes of ISO 1519 give different conclusion expressions: the former is go/no-go and yields "pass"; the latter is a step-down test that reports "the first mandrel diameter at which cracking appears", in which case the result should be written as "first cracking at D = 1 mm", not a blanket "qualified / not qualified". Mixing the two expressions is half of the common cause of acceptance disputes; the other half is not stating the substrate thickness and measured film thickness.

Q3: Cup test or cylindrical mandrel bend — is one of them enough?

Look at the in-service deformation mode. For conditions dominated by uniaxial bend — edge folding, roll-forming, chassis hemming — use the cylindrical mandrel bend (adding a T-bend or cone mandrel if needed); for draw moulds, large-radius deep drawing and dome-like multi-directional bulging, use the cup test. The two stress states differ, and the standards' introductions list them as mutually complementary independent methods without conversion. If a part undergoes both folding and bulging, both must be tested, and on the same batch under the same film-thickness condition.

Q4: Can pendulum damping time replace the bend test?

No. The pendulum reads viscoelastic decay (König 6° to 3°, Persoz 12° to 4°), whereas bending reads crack and peel resistance; the physical quantity and failure mode differ. One coating may have a high damping time yet crack on a 2 mm mandrel (brittle and hard), or a low damping time yet not crack (soft and tough). If the customer requires "hardness", give both the method and the read window, and note that it is not mutually convertible with the indentation method (GB/T 9275) or the pencil method (ISO 15184).

Q5: If film thickness rises from 25 µm to 60 µm, must the deformation conclusion be retested?

Retest it. Film thickness is part of the criterion: the Table 2 correction factors of ASTM D522 are added per mil (25 µm), reaching 1.40 at the small-mandrel step (3.2 mm), showing that film thickness markedly affects the coating's actual strain; and a thick film bears greater stretch at the dome top of a cup test, so its cracking tendency usually rises rather than falls. Changing film thickness within the same batch number is equivalent to changing the test condition, and the original conclusion cannot be extrapolated; quotations or technical agreements should write "film-thickness range + corresponding criterion" as a pair.

Q6: The substrate is aluminium, and the result is one step worse than cold-rolled steel — does that mean the paint is bad?

Not necessarily. Aluminium is soft and low in modulus, deforms in a mode different from steel when bent, and "the coating did not crack" versus "the substrate yielded first" must be judged separately; use a 10× magnifier to check whether the crack exists only within the coating, examine the cross-section or do dye-penetrant if needed, and add a bare-panel control (the same batch of uncoated aluminium bent the same way). If the bare panel already shows yield marks, the deformation-test conclusion should fall back to "the loading condition does not hold", and a thicker or harder panel is then needed to redo it.

Q7: Must the report state how long conditioning lasted? What if it is not stated?

State the temperature and humidity and the conditioning time (standard atmosphere 23 ℃ ± 2 ℃, 50 % ± 5 % RH, per GB 9278 / ISO 3270:1984), and also state the ageing time after panel preparation and the measurement time point. Without these fields, the data can only be used for internal trend, because solvent retention and incomplete curing drift both the bend and the damping results; from ISO 6272-1:2011 the drop-weight method also requires recording humidity and completing the test within a limited duration after conditioning. Completing these four items before external submission is a low-cost way to prevent disputes.

Q8: Delivering to the EU or North American market — how to align method editions?

Three practices: First, name the method and edition in the technical agreement (for example "ISO 1519:2011" or "GB/T 6742-2007") and explain the difference in adoption year between the two. Second, if the customer specifies ASTM, run per ASTM D522 Method A or B, writing the agreed 1.5 s to 5 s duration and the naked-eye judgement. Third, retain dual-edition comparison data for the same batch, for cross-plant re-inspection. When Chinese manufacturing export business is transacted on EXW or FOB, the test method and judgement standard are part of the technical annex and the edition year must be fixed in the contract annex, to avoid re-judging to a new edition after delivery.

14. References and status of the figures

The method standards and their current-edition status cited in this article are as follows: ISO 1520:2006 "Paints and varnishes — Erichsen test for elastic properties of coatings" (corresponding to GB/T 9753-2007, identical adoption, issued 2007-09-11, implemented 2008-04-01, replacing GB/T 9753-1988); ISO 1519:2011 "Paints and varnishes — Bend test (cylindrical mandrel)" (GB/T 6742-2007 identically adopts ISO 1519:2002, its 2023-12-28 review concluded continued validity); ISO 1522:2022 "Paints and varnishes — Pendulum method" (GB/T 1730-2007 identically adopts ISO 1522:1998); ISO 6860:2006 (cone mandrel), ISO 17132:2007 (T-bend, corresponding to GB/T 30791-2014, GB/T 11185-2009); ISO 6272-1:2011 and ISO 6272-2 (drop-weight fast deformation, corresponding to GB/T 1732-2020, replacing GB/T 1732-1993).

Supporting normative references: ISO 1514:2016 (standard panels), ISO 2808:2019 (determination of film thickness, covering wet film, dry film, uncured powder, plus appendices for rough surfaces and wood substrates), ISO 1513 (examination and preparation of samples), ISO 15528 (sampling), ISO 3270:1984 (temperature and humidity for conditioning and testing, corresponding to GB 9278-1988), ISO 2409 (cross-cut, corresponding to GB/T 9286-2021), ISO 4624 and GB/T 5210 (pull-off adhesion).

Status of the figures: the elongations and correction factors of ASTM D522 Table 1 and Table 2, and the outer-strain formula 100 t/(D+t) with its worked example (a 1/32 in panel over a 1/8 in mandrel, about 20 %) come from the public standard text; the die and hold-ring dimensions of the cup-test apparatus and the 0.2 mm/s constant-speed description come from instrument makers' public technical data, of "public-source status", with formal arbitration following the accompanying documentation and the original standard; the pendulum mass, ball diameter, period, amplitude window and glass reference damping time of König and Persoz come from the common status of multiple instrument makers' public data, among which the glass reference time appears as both 430 s ± 10 s and 430 s ± 15 s across different makers' documents — this article takes the more frequent ±10 s, and in external documents follows the manual of the equipment actually used.

Content not included in this article: the material-side Erichsen cup test (GB/T 4156 / ISO 8490) is only bounded, not judged in detail; the drop-weight parameters of the ISO 6272 series and the EN 13523 series coil-specific methods are outside this article's coverage; this article makes no deformation-indicator commitment for any specific formulation — all criteria follow the method and edition agreed in the technical agreement between supplier and purchaser.

Formal citation of standard texts must be obtained through legitimate channels and performed to the agreed edition; the "current editions" listed here are the public standard-information-service status as of the retrieval point (2026-11-14), and publication or review-status changes follow official notices.

15. Closing: writing the method and edition into the contract saves more cost than testing once more

The three routes of deformation-type tests each answer one specific question: the cup test answers "under biaxial bulging, to what pressing depth it does not crack", the cylindrical mandrel bend answers "over what mandrel diameter it does not crack under uniaxial bend", and pendulum damping answers "how fast the surface's viscoelastic decay is". Writing all three clearly, naming the adopted editions, and listing film thickness and substrate as part of the criterion are the prerequisites for the purchasing document and the outgoing inspection to reconcile.

Last updated: 2026-11-14. This article was compiled by the technical team of Kexin New Materials (Guangdong) Co., Ltd.; the external byline is Kexin New Materials (Guangdong) Co., Ltd.; the company serves manufacturing customers, its supply status is B2B, and trade terms are agreed as EXW or FOB, with the deformation criteria of specific coating systems following the technical agreement between the two parties and the agreed method edition.

Tags: #IndustrialCoatings #Industrial Coatings #Coatings technology literature #AnticorrosiveCoatings