Weld Heat-Affected Zone Hydrogen Cracking Case Study

This case study examines delayed hydrogen cracking in the heat-affected zone of a welded steel bracket. The weld initially passed visual inspection. Cracks appeared later, after cooling and restraint stresses had acted on a hard heat-affected zone containing diffusible hydrogen.

The case is realistic rather than tied to a named incident. It shows why weld quality is not only bead appearance. A safe welded joint depends on material chemistry, heat input, cooling rate, preheat, consumable handling, joint restraint, residual stress, inspection timing, hardness evidence, repair control, and release criteria.

Case Context

A quenched-and-tempered steel lifting bracket was welded to a base frame during a late manufacturing change. The joint was a multi-pass fillet weld around a thick lug pad. Production wanted to release the assembly after visual inspection and an immediate surface crack check.

Forty-eight hours later, a delayed inspection found several short cracks near the weld toe. Sectioning and hardness mapping showed that the cracks were in the heat-affected zone rather than in the weld metal centerline. The failure mechanism was hydrogen-assisted cold cracking in a hard, restrained HAZ.

Simplified Joint Data

Quantity	Value
base material	quenched-and-tempered structural steel
plate thickness	25\ \text{mm}
weld type	multi-pass fillet weld around lug pad
ambient temperature during welding	12^\circ\text{C}
original preheat used	75^\circ\text{C}
internal target preheat for high-risk condition	150^\circ\text{C}
original inspection timing	immediately after weld cleaning
delayed inspection timing	48 hours after welding
maximum measured HAZ hardness	428\ \text{HV}
internal HAZ hardness action limit	350\ \text{HV}
crack depth from follow-up ultrasonic sizing	a=3.0\ \text{mm}
simplified nominal stress for screening	\sigma=180\ \text{MPa}
assumed geometry factor	Y=1.12
conservative fracture toughness screen	K_c=70\ \text{MPa}\sqrt{\text{m}}

The numbers are used for engineering screening. Real weld procedure qualification requires project-specific material, consumable, thickness, restraint, heat input, environment, inspection method, and acceptance basis.

Failure Mechanism

Hydrogen cracking in steel welds usually needs three conditions at the same time:

1. diffusible hydrogen in the weld region;
2. a susceptible hard microstructure, often in the heat-affected zone;
3. tensile stress from restraint, residual stress, applied load, or thermal contraction.

Removing any one of these can reduce risk. Low-hydrogen consumables and dry storage reduce hydrogen. Preheat, interpass temperature control, heat input, postheat, and slower cooling reduce hard HAZ formation and allow hydrogen to diffuse out. Joint design, sequence, fit-up, and restraint control reduce tensile stress.

The original procedure treated the weld as a geometry and workmanship problem. The investigation showed it was a metallurgy and process-control problem.

Step 1: Carbon Equivalent Screening

A common weldability screen for carbon and low-alloy steels is carbon equivalent. One widely used form is:

Use the material chemistry from the mill certificate:

Element	Mass percent
C	0.16
Mn	1.35
Cr	0.35
Mo	0.18
V	0.04
Ni	0.45
Cu	0.25

Calculate:

So:

Engineering Comment

This is a high weldability-risk signal. Carbon equivalent is not a complete cracking model, but it tells the engineer that the procedure should not be copied from a thinner, lower-carbon-equivalent steel without review. Preheat, hydrogen control, heat input, interpass temperature, and delayed inspection become part of the release basis.

Step 2: Heat Input Check

The recorded welding parameters for the critical pass were:

Parameter	Value
arc voltage	28\ \text{V}
welding current	260\ \text{A}
travel speed	4.0\ \text{mm/s}
thermal efficiency factor	\eta=0.8

Approximate heat input per unit length:

Substitute:

Engineering Comment

The heat input was not abnormally low by itself, but heat input cannot be judged alone. The 25 mm thick restrained plate, high carbon equivalent, cool ambient condition, and insufficient preheat created a fast cooling condition near the weld toe. The procedure needed a combined control: preheat, interpass temperature, hydrogen management, and delayed inspection.

Step 3: Hardness Evidence

The delayed investigation measured a HAZ hardness map. The internal action limit was:

The highest HAZ result was:

The exceedance is:

As a percentage of the action limit:

or about:

Engineering Comment

Hardness does not prove hydrogen cracking by itself, but it supports the mechanism. A hard HAZ is more susceptible to cracking when hydrogen and tensile stress are present. The hardness map also explains why the weld bead surface could look acceptable while the local material state near the toe was not acceptable.

Step 4: Why Immediate Inspection Missed the Defect

Immediate inspection after cleaning found no reportable indication. Delayed inspection after 48 hours found surface-breaking cracks near the weld toe, and ultrasonic follow-up estimated a representative crack depth of:

Hydrogen cracking can be delayed because hydrogen diffusion, cooling, residual stress redistribution, and crack initiation do not necessarily complete during the first inspection window. For susceptible welds, inspection timing is part of the method, not a scheduling detail.

Engineering Comment

An immediate surface inspection can be useful for workmanship defects such as undercut, crater cracking, arc strikes, and some lack-of-fusion indications. It is not enough to release a hydrogen-cracking-sensitive joint unless the procedure has evidence that delayed cracking risk is controlled.

Step 5: Fracture Screening of the Detected Crack

Use a simplified stress-intensity screen:

where:

Calculate:

Compare with the conservative toughness screen:

The immediate static-fracture margin appears positive:

Engineering Comment

This calculation does not justify release. It only says the simplified crack is not predicted to cause immediate unstable fracture under the assumed stress and toughness. The joint is still unacceptable because the crack is a manufacturing defect in a susceptible zone, may have branches, may grow under residual or service stress, and may indicate a process condition that can affect other welds. Fracture mechanics supports disposition, but it does not erase process-control failure.

Step 6: Estimate Critical Crack Size for Context

Rearrange the same relation:

Substitute:

So:

Engineering Comment

The critical size is much larger than the detected crack in this simplified screen, but this is not a release argument. The calculation is useful for risk ranking, repair urgency, and inspection planning. It also depends strongly on stress, geometry factor, toughness, crack shape, residual stress, temperature, and whether the measured crack depth is representative.

Root Cause

The root cause was a weld procedure and release-control failure:

1. The actual material carbon equivalent and thickness were outside the basis used for the original preheat decision.
2. Hydrogen control was weak: consumable exposure time, surface moisture, and storage records were not sufficient for the risk level.
3. The preheat was too low for the combined chemistry, thickness, restraint, and ambient condition.
4. Immediate inspection was treated as final inspection even though delayed cracking was credible.
5. Hardness mapping was not required before release for this high-risk weld class.
6. The design change did not trigger a full weldability and NDE timing review.

The crack was the visible symptom. The deeper issue was that the manufacturing route no longer matched the material and reliability assumptions.

Repair and Disposition

The cracked weld was not accepted as-is. The disposition package required:

1. remove the cracked region by controlled excavation;
2. verify crack removal with surface inspection and, where geometry allowed, ultrasonic inspection;
3. re-weld using a revised procedure with higher preheat, controlled interpass temperature, documented heat input, low-hydrogen consumable control, and dry joint preparation;
4. hold the assembly warm after welding when required by the procedure to support hydrogen diffusion;
5. perform delayed surface inspection after the specified hold period;
6. repeat hardness mapping in the HAZ;
7. review whether similar welds on the same assembly require quarantine or supplemental inspection;
8. update the drawing or manufacturing plan so future design changes trigger weldability review.

Repair was treated as a new process event, not a cosmetic rework. Grinding out a crack without changing the conditions that created it would only repeat the failure.

Release Criteria

The revised release criteria were:

Evidence	Release requirement
Procedure basis	Material grade, thickness, carbon equivalent range, joint restraint, and weld process covered.
Preheat and interpass	Recorded values within the approved window before and during welding.
Consumable control	Low-hydrogen handling, exposure time, and storage documented.
Heat input	Pass records within the qualified range.
Delayed inspection	No relevant cracks after the required delay period.
Hardness map	HAZ results below the action limit or formally dispositioned.
Repair control	Excavation, re-weld, and reinspection documented if defects are removed.
Change control	Future thickness, material, consumable, process, or restraint changes require engineering review.

Engineering Lessons

A weld is a local heat treatment, a joining operation, a geometric detail, and an inspection problem at the same time. Hydrogen cracking risk cannot be controlled by bead appearance alone.

For susceptible steels, the practical engineering questions are:

1. Does the material chemistry and thickness require special preheat or hydrogen control?
2. Does the heat input and interpass temperature produce an acceptable HAZ hardness?
3. Is the joint sufficiently restrained to create high tensile residual stress?
4. Is inspection delayed long enough to reveal delayed cracking?
5. Are repair and reinspection rules explicit before production starts?

The strongest release decision combines chemistry, procedure records, hardness evidence, NDE timing, fracture screening, and change control. Without that evidence, a visually acceptable weld can still contain a high-risk heat-affected zone.