logo
Latest company news about Hydrogen Embrittlement: The Delayed Failure That a Tensile Test Will Never Catch

November 18, 2026

Hydrogen Embrittlement: The Delayed Failure That a Tensile Test Will Never Catch

The failure that arrives after the inspection report

A class 10.9 stud bolt passes the tensile test, passes the hardness check, ships with a clean certificate — and then, three days after installation, it cracks in service at a load well below its rated capacity. No overload, no impact, no visible corrosion. Just a clean, brittle fracture surface where a ductile steel should have stretched.

This is not a material defect and not a torque error. It is hydrogen embrittlement, and it is the one failure mode that a standard incoming inspection is structurally unable to catch — because the bolt that will fail this week looks identical, on every gauge, to the bolt that will hold for twenty years.

Where the hydrogen comes from

The hydrogen that destroys a high-strength fastener is usually invited in by the very process that protects it from rust:

  • Acid pickling before plating — the first hydrogen source. Pickling solutions attack the steel surface and release atomic hydrogen, which diffuses into the substrate.
  • Electroplating — zinc, cadmium and zinc-alloy plating all generate atomic hydrogen at the cathode. A fraction of that hydrogen is trapped inside the steel when the coating seals the surface.
  • Sacrificial environments in service — fasteners under cathodic protection, or carbon steel bolts electrically coupled to stainless components in wet service, keep absorbing hydrogen long after the plating line.

The cruel arithmetic is that embrittlement risk rises with strength. Up to class 8.8, the risk is low enough that it rarely drives decisions. From class 10.9 upward — and on high-temper alloy grades such as A193 B7 in small diameters — it becomes a design variable that must be managed, not assumed away. The harder the steel, the less plastic capacity it has to accommodate trapped hydrogen at a stress concentration.

Why bolts fail days later, not on the test bench

Hydrogen embrittlement is a delayed, sustained-load failure. The hydrogen atoms migrate toward the highest triaxial stress zones — typically the first load-bearing threads and the thread root — over hours and days. Once a critical local concentration is reached, a micro-crack initiates, propagates, and the bolt fractures below yield.

Three consequences follow for anyone buying or specifying plated fasteners:

  1. A tensile test immediately after plating proves nothing. The bolt is at its most vulnerable hours after it leaves the line, not minutes. A lot that passes the bench test on Tuesday can fail in the field on Friday.
  2. The failure looks like everything else. Fracture surfaces resemble brittle overload, so root-cause investigations routinely mislabel hydrogen embrittlement as "bad steel" or "over-torque" unless someone specifically checks for it.
  3. Sustained load is the trigger, not peak load. A flange stud bolt sitting under constant preload is the classic victim — precisely the application where stud bolts are used most.
The fix is a bake, and the fix must be written down

The remedy is old, cheap and effective: drive the hydrogen back out with a low-temperature bake before the trapped gas can do damage. The standard practice is baking within a few hours of plating — typically 4 hours or more in the 190–230 °C range depending on strength class and coating, with class 10.9 and above treated as mandatory and class 8.8 treated as good practice on safety-critical joints.

ISO 4042, the threading standard for electroplated fasteners, formalises these hydrogen-management requirements, and sustained-load verification is covered by test methods such as ISO 15330 and the ASTM F519 test specimen. Two things matter commercially:

  • HDG is a different risk profile. Hot-dip galvanizing involves a thermal cycle at ~450 °C, which releases much of the absorbed hydrogen as a side effect. That is one reason hot-dip galvanized studs carry a lower embrittlement flag than thin electroplate at equal strength — and one more reason the zinc-versus-HDG choice is a joint design decision, not just a cost line.
  • The bake leaves no visible trace. A bolt that was baked and a bolt that was not are indistinguishable on the bench. The only evidence is a process record.
What to put in the purchase order

Because the evidence is invisible, the requirement must be contractual. For electroplated fasteners of class 10.9 and above — stud bolts for flange service included — the specification should require:

  1. Baking after plating as an explicit line item — temperature, duration and time window between plating and baking stated, with process records retained per batch.
  2. Sustained-load verification on a sampling basis — per ISO 15330 or an ASTM F519 wedge specimen, not just routine tensile testing.
  3. A plating process note, not just a finish name. "Zinc plated" alone does not tell the buyer whether the line runs alkaline or acid, or whether the bake step exists.Suppliers who can answer that question in writing are the ones who run it.
  4. Caution in hydrogen-charging service. Where bolts operate under cathodic protection or couple with stainless in wet chloride environments, strength class selection and coating choice deserve a joint review with the supplier — the exposure continues after installation.
The supply-chain angle

For buyers auditing a fastener supplier, the de-embrittlement step is a surprisingly effective litmus test. It requires nothing exotic — an oven, a thermometer, a logbook — but a factory that bakes on schedule, records every batch and can produce those records with the packing list is also the factory that runs heat treatment, gauging and traceability with the same discipline. Ask to see the baking log. The answer tells you more about the supplier than the certificate does.


Shaanxi Flourish Industrial Co., Ltd. supplies stud bolts, hex bolts and bolt-and-nut assemblies in grades 4.8–12.9 with zinc plating, hot-dip galvanizing and PTFE/Xylan coatings. Electroplated class 10.9 and above is de-embrittled after plating as standard practice, with baking records available on request.