Proof Load vs Yield and Tensile Strength in Bolt Selection: What the Evidence Settles

Proof load is the qualification number, not tensile strength

Proof load, yield strength and tensile strength are all minimum values set by a standard that a bolt must meet to qualify as a given grade or property class ⁴. They are not three views of one property. ASTM A354 grade BD 1/2-13 must reach 17,050 lbf proof load, 18,500 lbf yield and 21,300 lbf tensile — three separate standard-set minima for one bolt ⁴. Proof load and yield strength are similar tests, so a standard may omit the yield requirement in favour of proof load; SAE J429 is cited as an example ⁴. When that happens, proof load is the only strength value the standard gives you, and it becomes the number a quote or datasheet has to be read against. ISO 898-1 class 12.9 is specified the other way, with a minimum tensile strength of 1220 MPa and a minimum yield strength (proof stress) of 1100 MPa ¹⁰. For grade selection, the practical consequence is that a supplier's headline tensile figure does not by itself show the bolt qualifies. Ask which standard and edition the bolt is certified to, and which of the three values that standard actually specifies.

Proof stress by grade: rule out grades below the joint requirement

Published proof stress differs by grade and standard, and the values below are indicative — they must be confirmed for the applicable diameter and standard edition before they go into a specification ¹. The table is a first screen: a grade whose published proof stress sits below what the joint needs is ruled out before any torque or preload work begins.

Equal tensile strength does not make two grades interchangeable

A4-80 stainless and property class 8.8 both carry a nominal tensile strength of 800 MPa, but they are not automatically equivalent ¹. Their minimum proof or yield strength, ductility, hardness, friction, tightening response, galling behaviour and compatible nut and washer requirements can all differ, and a change of material can change preload scatter and fatigue performance even when the bolt dimensions stay the same ¹. The indicative proof stresses make the gap visible: roughly 600 MPa for A4-80 against roughly 640 MPa for 8.8 ¹. The same logic applies to the mating part. A Grade 8 bolt with a Grade 2 nut will fail at the nut, so grades are matched across the joint rather than chosen bolt-first ¹². Where a joint genuinely needs 10.9- or 12.9-level preload, ordinary A2/A4 stainless should not be substituted merely for corrosion resistance; the alternatives named are a qualified coated high-strength fastener, duplex stainless, a corrosion-resistant alloy or a redesigned joint, depending on environment and governing code ¹. Before accepting a substitution, check the standard of manufacture as well as the strength value. Grade 8 bolts have strength levels similar to A490 bolts but are made to a different standard (SAE J429 against ASTM F3125) and are generally not interchangeable in structural steel without engineering approval ¹¹.

Set preload as a bounded fraction, then pick a control method the joint can tolerate

One common method calculates preload as bolt tensile strength divided by a safety factor, typically 2–3; with an 800 MPa bolt and a factor of 2, that gives 400 MPa ². The sources do not establish a preload target expressed as a percentage of proof load, so treat the tensile-strength method as one published approach rather than the rule. Preload has to be bounded on both sides. Too much can yield or break the bolt, damage components or facilitate stress corrosion cracking; too little causes loosening, fatigue, joint separation or slip ¹⁴. The largest possible preload is not the best ¹⁴. Correct preload is what allows smaller or fewer fasteners to be used, so the target is a design input, not a margin to be maximised ¹⁴. How closely the target is hit depends on the control method. Torque control scatters ±25–40% because friction and thread condition intervene; angle control scatters ±10–15% and is better for critical joints ⁷. Bolt elongation, hydraulic tensioning and torque-angle methods are used where higher accuracy is needed ⁷. Choose the method against the scatter the joint can absorb, not against the tightest number available.

Confirm the diameter and standard edition before the size is frozen

Proof load is specified per diameter. The ASTM A354 grade BD figures — 17,050 lbf proof load, 18,500 lbf yield, 21,300 lbf tensile — are stated for the 1/2-13 size specifically ⁴. A value quoted for one diameter cannot be carried to another, and the sources in this set do not reproduce the proof-load tables by diameter and thread pitch. That makes the standard edition a procurement item, not a footnote. Ask the supplier to state the standard and edition the bolt is certified to, and to supply the proof-load value for the exact diameter and pitch being quoted.

Rule out grades and coatings the service environment will not tolerate

Strength class and environment interact. ISO 898-1 explicitly warns that class 12.9 fasteners are susceptible to stress corrosion cracking, including hydrogen-assisted SCC, and should be used with extreme caution in corrosive environments ¹⁰. The hardness band explains the split: class 12.9 sits at 39–44 HRC, in the extreme-risk zone with virtually zero tolerance for atomic hydrogen, while a class 10.9 bolt at 32–39 HRC may tolerate several parts per million of diffusible hydrogen before crack initiation ¹⁰. Where high strength is needed without that risk, zinc flake coatings are used for high-strength fasteners, often grade 10.9–12.9, with zero hydrogen embrittlement risk, and for service up to 300°C continuous ²⁰. If the joint needs 10.9- or 12.9-level preload, ordinary A2/A4 stainless is not the corrosion-resistant substitute — a qualified coated high-strength fastener, duplex stainless, a corrosion-resistant alloy or a joint redesign are the routes named ¹. No source in this set quantifies a proof-load derating factor for coating or temperature, so any derating in the specification has to come from the applicable standard rather than from a supplier's table.

Write the inspection and traceability scope into the RFQ

A quotation that omits proof-load qualification evidence cannot be checked against the standard, so the RFQ has to ask for it explicitly. A reliable technical quotation needs the standard and edition, bolt type, diameter, length and thread pitch, material and property class, coating or lubricant, quantity, operating temperature and corrosion environment, joint description, torque or preload requirement, nut and washer pairing, lot traceability, MTC/CoC and the inspection requirements ⁹. A request reading only 'M16 stainless bolt' is not enough ⁹.

• Standard and edition the bolt is certified to, plus the proof-load value for the quoted diameter and pitch ⁹
• Property class, coating or lubricant, and nut and washer pairing stated as one system ⁹
• Lot traceability and MTC/CoC for the delivered lots ⁹
• Hydrogen embrittlement testing per ISO 15330 at 75% of proof load for 200 hours minimum, no failures permitted, where high-strength fasteners are involved ⁵
• Friction coefficient testing per ISO 16047 where the joint has a specified clamp load ⁵
• Coating adhesion testing on each production batch where hot-dip galvanized hardware is specified ⁵

Decide whether the higher grade is actually needed

Over-specifying wastes money and under-specifying creates safety risk; higher-strength fasteners are not always necessary and may raise project cost unnecessarily ^{16, 22}. The joint is only as strong as its weakest link, so the nut, washer and clamped material have to be compatible with the bolt's strength ¹⁶. Correct preload is the lever that lets a smaller or fewer fasteners do the job, which means the grade decision should follow the preload calculation rather than lead it ¹⁴. Where a structural standard governs, the grade is not a free choice: substituting standard hex bolts where A325 or A490 structural bolts are specified is a code violation ¹².

Where the sources disagree

Two disagreements matter for the specification. The first is the basis for preload: one source derives it from tensile strength with a 2–3 safety factor, while the hydrogen embrittlement protocols express test stress as a percentage of proof load. The second is the status of the hydrogen embrittlement threshold percentages — the source that publishes them also labels them arbitrary, so they should not be written as mandatory limits.

FAQ

Related pages

• Bolts product page
• Fastener Materials for Continuous Service Above 400°C: What the Evidence Settles and What the Code Must Supply
• Pairing a Nut to a Bolt: Property Class, Flange vs Hex, and What the Evidence Actually Settles
• Yield Strength vs Tensile Strength in Bolts: Which Governs Preload, Which Governs Rupture
• Bolt Head Types and Drive Styles: What the Evidence Can Settle Before You Specify