The Short Answer
Three constraints decide the coating, in fixed order: property class, thread, environment. Get the first two wrong and corrosion never matters.
- 8.8 or lower, C2–C3, coarse thread, joint closed once: hot-dip galvanizing at 40–50 μm on the threaded part, with an oversized nut or a thread cut after coating.
- 10.9 or 12.9, C4–C5, fine thread, joint opened for maintenance: zinc flake, with a torque coefficient measured on the delivered finish.
- C5–CX, splash zone, or a joint that cannot be recoated: duplex (hot-dip plus paint) or a stainless grade. Neither coating alone is a service-life guarantee.
C-grades follow ISO 9223.
Two Different Coatings, Not Two Grades
Hot-dip galvanizing is metallurgical: GB/T 18226-2025 §3 defines the coating as the layer formed by immersing steel in molten zinc, with dissolution, chemical reaction and diffusion; the result is a zinc-iron alloy bonded to the substrate.
Zinc flake, covered in the fastener coating systems, is a different family: the same clause defines zinc-chromate flake as a water-based paint applied by dip, brush or spray and baked, with lamellar zinc and chromate. A binder holds overlapping plates; the bond is not metallurgical.
Thickness follows. GB/T 13912-2020 Table 3 gives 35 μm local and 45 μm mean below 1.5 mm, rising to 70 and 85 μm above 6 mm. Threaded parts are in Table 4 because they are centrifuged: 40 μm local and 50 μm mean above 6 mm diameter, 20 and 25 μm at 6 mm and below. GB/T 18226-2025 Table 2 puts Class I highway fasteners at d ≥ 8 mm at 49 μm.
Zinc flake is several times thinner. GB/T 18226-2025 §16 sets a minimum of 8.6 μm for zinc-chromate flake on bolts and nuts and refers performance to GB/T 18684-2002. Zinc-aluminium data sheets commonly specify 8–15 μm total film; that is a product convention, not a standard value.
| Property | Hot-dip galvanizing | Zinc flake |
|---|---|---|
| Bond | Metallurgical zinc-iron alloy (GB/T 18226-2025 §3) | Baked lamellar film (GB/T 18226-2025 §3) |
| Thickness, threaded part | 40 μm local, 50 μm mean, d > 6 mm (GB/T 13912-2020 Table 4) | 8.6 μm min, zinc-chromate (GB/T 18226-2025 §16); 8–15 μm typical, zinc-aluminium (supplier) |
| Thread allowance | 4t = 160–200 μm; needs 6AZ/6AX or post-tap | 4t = 32–60 μm; may fit f or e, not always 6g |
| Thermal cycle | Above the 425 °C or 380 °C tempering floor (GB/T 3098.1-2010 Table 2) | Cured below the tempering floor |
| Damage | Chips; zinc-rich paint repair | Thin; zinc-rich paint repair |
| Appearance | Grey, matte, colour variation allowed (GB/T 13912-2020 §6) | More uniform; silver, grey or black |
A thickness without a location is not a specification: GB/T 13912-2020 §6 requires a 10 cm² area, five readings, and no readings within 10 mm of an edge; weighing prevails.
Thickness and the Thread Allowance
A coating consumes clearance. GB/T 5267.1-2023 Annex D states that a coating thickness t adds 4t to the pitch diameter d₂ of an external thread. Table D.1 gives 3 μm → 12 μm, 8 → 32, 12 → 48. At 40–50 μm, hot-dip means 4t = 160–200 μm.
Table D.2 gives the maximum coating thickness by pitch and position: at P = 1 (M6, M7), 6 μm for g, 10 for f, 15 for e; at P = 2 (M14, M16), 9, 13 and 17 μm; at P = 6 (M64), 20, 23 and 29 μm.
A 40–50 μm hot-dip layer exceeds every position, including 29 μm at P = 6 e, so a standard 6g/6H pair cannot be assembled. The answer is an oversized internal thread: GB/T 3098.2-2015 §5 names 6AZ and 6AX, with a basic deviation above zero, and warns of lower stripping strength; alternatively, ream and tap after coating or cut the thread after galvanizing.
Zinc flake is thin but not zero. At P = 1, a 6g thread accepts 6 μm; an 8 μm coating needs f, and 15 μm sits at the e limit. At P = 2, 8 μm fits g, while 12–15 μm needs f or e. Fine pitches bind first, so the question is whether the drawing’s position covers 4t.
The drawing must carry the pre-coating size. GB/T 5267.1-2023 §4.3 states that external-thread coating shall not cross the zero line (h) and internal-thread coating shall not go below H; threads are gauged after coating with h and H go gauges. Chapter 10(e) makes the pre-plating position purchaser-supplied.
| Coating t | 4t added | Fits 6g at P=1 / P=2 | Fits e at P=1 / P=2 | Result |
|---|---|---|---|---|
| 8 μm | 32 μm | No / Yes (9 μm) | Yes / Yes | M14/M16 in 6g; M6–M10 needs f or e |
| 12 μm | 48 μm | No / No | Yes (15 μm) / Yes (17 μm) | e, or a pre-coating reduction |
| 15 μm | 60 μm | No / No | At the limit / Yes | e with no margin at P=1 |
| 40 μm | 160 μm | No / No | No / No | Hot-dip: 6AZ/6AX or post-tap |
| 50 μm | 200 μm | No / No | No / No | Hot-dip: 6AZ/6AX or post-tap |
Above 5d, Annex D.2 records mid-length thickness one-third to one-half lower than the reference area, so a bolt thickened for the middle can end with an over-thick thread.
Hydrogen and the Tempering Window
Internal hydrogen embrittlement needs three conditions at once: high tensile strength or hardness, a tensile stress, and atomic hydrogen (GB/T 5267.1-2023 §4.4). The mechanism is covered in the hydrogen embrittlement guide; here the question is why hot-dip fits the top property classes badly.
Hot-dip has a hydrogen step and a thermal step. Acid pickling before fluxing is a hydrogen source, the same acid-cleaning route named in the electroplating standard. The thermal step is unique: the part enters molten zinc. Conventional lines run at 445–465 °C (shop practice, not a standard value), above the minimum tempering temperatures in GB/T 3098.1-2010 Table 2: 425 °C for the listed 8.8–10.9 and alloy-steel 12.9 rows, 380 °C for 12.9 carbon steel. A 10.9 bolt tempered at 425 °C has no margin for a 450 °C dip. The 20 HV re-tempering limit in §7.4 is a sample test, not a production guarantee.
A 10.9 bolt runs 320–380 HV, a 12.9 runs 385–435 HV (GB/T 3098.1-2010 Table 3). GB/T 5267.1-2023 Table 2 sorts electroplated parts at 360 and 390 HV, with 10.9 in class B and 12.9 in class C; it does not classify hot-dip, but it marks where the risk starts.
Zinc flake avoids the molten-metal step: applied as a liquid and cured well below the tempering floor, a 10.9 or 12.9 part does not go through a second tempering cycle. That is why flake is common on high-strength joints, not a claim that it is hydrogen-free; cleaning still needs a process sequence.
For a 10.9 or 12.9 joint, the purchasing file should carry tempering temperature and time, post-coating hardness, and a sustained-load test to GB/T 3098.17 on safety-critical joints.
Corrosion Testing vs Service Life
GB/T 10125-2021 defines the salt-spray family: NSS at 35 ± 2 °C with 50 g/L sodium chloride at pH 6.5–7.2, AASS at pH 3.1–3.3, CASS at 50 ± 2 °C. Durations run from 2 h to 1,008 h. §1 states the boundary: the methods serve quality control and batch comparison, not the ranking of materials or the prediction of long-term life; hours are an acceptance index, not years.
A 720 h and a 1,000 h salt-spray result do not mean 20 and 30 years. GB/T 5267.1-2023 Table 8 footnote a says the same. GB/T 18226-2025 uses 168 h NSS with no red rust as a batch floor for hot-dip galvanized highway parts.
The tests answer different questions. Salt spray is continuous fog: it finds porosity and batch variation, not the wet-dry cycle. A cyclic immersion test alternates wet and dry; oxygen reaches the surface in the dry phase, closer to a splash or tidal zone. Atmospheric exposure includes UV, rain, pollutants and time, and is measured in years. Results do not transfer between the three.
The test request needs method, duration, end point (white or red rust), coating thickness, and the acceptance standard; without them, “passed salt spray” is not comparable.
Torque Coefficient Changes the Preload
Torque and preload are linked by K:
GB/T 1231-2024 Annex A gives Ki = T / (F × d). §6.3 requires high-strength large hexagon assemblies to be supplied with a mean K of 0.110–0.150 and a coefficient of variation no greater than 0.077. That band is the delivered friction condition, coating and lubricant included, not a property of the grade.
A coating change invalidates a torque table. For an M20 10.9S assembly, Table 14 sets the preload window at 161–179 kN. At K = 0.110 the torque for 161 kN is 354 N·m; at K = 0.150 it is 483 N·m. Apply 354 N·m to a set whose actual K is 0.150 and preload lands at 118 kN, 43 kN low, a 27% shortfall. Apply 483 N·m at K = 0.110 and preload reaches 219.5 kN, 36% high and above the 179 kN maximum. The yield load of an M20 10.9 bolt is about 220.5 kN (900 MPa × 245 mm²), so the second error puts the bolt at yield.
K must be measured on the finish that will be installed. GB/T 16823.3 and MIL-STD-1312-15 are torque-tension methods; record mean K and variation, not one reading. A zinc-flake system with an integral lubricant usually sits lower and tighter than dry hot-dip, but that is a test assumption, not a value to copy.
Damage, Repair and Reassembly
Hot-dip takes site handling but chips at edges and on impact. Field repair is zinc-rich paint: a barrier with sacrificial zinc, not a rebuilt alloy layer. GB/T 50046-2018 §4.3.9 goes further for building structures: hot-dip bolts, nuts and washers, and after installation the joint receives the same protection as the main structure.
Flake is thin; a knock exposes steel sooner. A small bare spot is protected by the surrounding sacrificial zinc; zinc sits on the active side of steel in the galvanic series (MIL-STD-889C Table II). A large torn area cannot be rebuilt on site; the rework is a replacement.
Reassembly changes friction. A hot-dip thread with a 6AZ/6AX nut runs with clearance and a rough surface; repeated assembly wears the coating and K drifts. A flake thread is smoother and usually lubricated; repeated assembly can remove the lubricant and raise K. Re-torque after reassembly, and re-check K on a critical joint.
Selection Tree
The tree is a filter, not the calculation: a C3 site with a 12.9 bolt still goes to the first branch.
Acceptance Checklist
What to Put in the PO
| PO line | What to write |
|---|---|
| Class and heat treatment | Class, tempering data, post-coating hardness for 10.9/12.9 |
| Coating | Process, standard and thickness at the measurement area |
| Thread | Pre-coating size, tolerance position, 6AZ/6AX or post-tap, go-gauge |
| Hydrogen | Process sequence; baking or sustained-load test where required |
| Friction | K mean, variation, test method and the resulting torque table |
| Corrosion | Method, duration, end point and the acceptance standard |
These are standard values and arithmetic. A specific joint still needs the project engineer to confirm preload, thread allowance and corrosion allowance. Send the joint schedule with class, thread size and pitch, environment, assembly sequence and coating callout for a torque-tension test plan.
References
- GB/T 13912-2020 — Hot dip galvanized coatings on iron and steel articles, Tables 3 and 4, §3 and §6
- GB/T 5267.1-2023 — Fasteners: electroplated coatings, §4.3, §4.4, Annex D, Tables 8 and 10, Chapter 10
- GB/T 3098.1-2010 — Mechanical properties of bolts, screws and studs, Table 2 and Table 3
- GB/T 3098.2-2015 — Mechanical properties of nuts, §5
- GB/T 10125-2021 — Corrosion tests in artificial atmospheres: salt spray tests, §1, §7 and §11.1
- GB/T 18226-2025 — Anticorrosion conditions for highway steel components, §3, §16 and Table 2
- GB/T 1231-2024 — High-strength large hex bolted connections for steel structures, §6.3, Annex A and Table 14
- MIL-STD-889C — Dissimilar metals, §5.2 and Table II
Key Standards
Standards and clauses referenced by this article.
- Metallic coatings - Hot dip galvanized coatings on fabricated iron and steel articles§3; §6; Table 3; Table 4
Thickness minima by substrate thickness (Table 3) and for centrifuged threaded parts (Table 4); local versus mean thickness definitions; measurement area rules; allowance for colour variation; note that hot-dip nuts are often reamed and tapped oversize.
- Fasteners - Electroplated coatings§4.3; §4.4; §10(e); Annex D (D.1, D.2); Table 8 note a; Table 9
The 4t pitch-diameter rule and Table D.1 values; maximum coating thickness by pitch and tolerance position (Table D.2); pre-plating thread tolerance and gauging rules; the three IHE conditions and hardness classes; the warning that NSS hours are not a direct basis for corrosion resistance.
- Mechanical properties of bolts, screws and studsTable 2; Table 3; §7.4
Minimum tempering temperature: 425 °C for the listed 8.8-10.9 and alloy-steel 12.9 rows, 380 °C for 12.9 added-element carbon steel; strength and hardness bands (10.9: 320-380 HV, 12.9: 385-435 HV); 20 HV limit on hardness drop after re-tempering.
- Mechanical properties of nuts§5
6AZ and 6AX hot-dip nut tolerances with a basic deviation above zero, and the warning that a positive basic deviation may reduce stripping strength.
- Fasteners - Torque/clamp force testingMethod reference
Torque-tension test named for determining K or the total coefficient of friction; method name only, no values quoted.
- Fasteners - Preloading test for the detection of hydrogen embrittlementMethod reference
Sustained-load test named for safety-critical 10.9/12.9 joints; method name only, no values quoted.
- Zinc-chromate coatings - Technical conditionsReferenced by GB/T 18226-2025 §16
Performance standard for zinc-chromate flake coatings; named only. Not independently consulted, so no salt-spray hours are quoted.
- Code for anticorrosion design of industrial constructions§4.3.9
Hot-dip galvanized bolts, nuts and washers for steel structures, with the same anti-corrosion measure as the main structure applied after installation.
- Fastener test methods - Method 15: Torque-tensionMethod reference
Torque-tension test method referenced for K measurement; no values quoted.
- Corrosion tests in artificial atmospheres - Salt spray tests§1; §7; §11.1; Table 3
NSS, AASS and CASS conditions; recommended durations; the scope boundary that salt-spray hours are not a material ranking or a service-life prediction.
- Anticorrosion technical conditions for steel components of highway traffic engineering§3; §16; Table 2
Process definitions for hot-dip and zinc-chromate flake coatings; 8.6 μm minimum for zinc-chromate flake on fasteners; hot-dip fastener thickness and 168 h NSS acceptance criterion.
- High-strength large hexagon head bolted connections for steel structures§6.3; Annex A; Table 14
Torque coefficient formula Ki = T/(F*d); mean K 0.110-0.150 with VK not above 0.077; M20 10.9S preload window 161-179 kN.
- Dissimilar metals§5.2; Table II
Galvanic series and design measures for dissimilar-metal couples; zinc listed on the active side of steel.
References
Sources used for fact checking and background context.
- GB/T 13912-2020 - Hot dip galvanized coatings on iron and steel articlesSAMR / SAC
- GB/T 5267.1-2023 - Fasteners: electroplated coatingsSAMR / SAC
- GB/T 3098.1-2010 - Mechanical properties of bolts, screws and studsSAMR / SAC
- GB/T 3098.2-2015 - Mechanical properties of nutsSAMR / SAC
- GB/T 10125-2021 - Corrosion tests in artificial atmospheres: salt spray testsSAMR / SAC
- GB/T 18226-2025 - Anticorrosion conditions for highway steel componentsSAMR / SAC
- GB/T 1231-2024 - High-strength large hex bolted connections for steel structuresSAMR / SAC
- MIL-STD-889C - Dissimilar metalsUS Department of Defense
Deep Reading
More systematic selection, procurement, or inspection guides.
- Fastener Coatings and Platings
The broader coating reference: designation, thickness baselines, salt-spray data and galvanic couples.
This page is one entry in the fastener engineering reference. The library covers standards, dimensions, materials, and failure analysis; every figure is traced back to its source.
Continue reading in the library →