The Hidden Defect That Arrives Hours After Welding
A pressure vessel passes visual inspection. The welder signs off. The assembly moves to the next station. Fourteen hours later, a technician discovers a through-thickness crack in the heat-affected zone (HAZ) — initiated at a hydrogen atom trapped in the crystalline lattice of the steel.
Hydrogen-induced cracking (HIC), also called cold cracking or delayed cracking, is one of the most dangerous and economically damaging defect modes in structural and pressure-retaining welded fabrications. Unlike hot cracking or porosity, HIC:
- Initiates below 200 °C, often hours or days after welding is complete
- Is invisible to in-process visual inspection
- Requires simultaneous presence of three factors: susceptible microstructure, hydrogen content, and tensile residual stress
- Can cause catastrophic brittle fracture in service under loads well below the material yield strength
For manufacturers in oil & gas, pressure vessel fabrication, structural steel, shipbuilding, and heavy lifting equipment, preventing HIC is not optional — it is a contractual, regulatory, and safety requirement. This post explains the physics of HIC, what monitoring data actually prevents it, and how real-time thermal monitoring systems enforce the preheat and thermal compliance regimes that eliminate the conditions for cracking.
The HIC Triangle: Three Conditions That Must All Be Present
Hydrogen-induced cracking requires three simultaneous conditions. Eliminate any one, and cracking cannot occur:
1. Susceptible Microstructure
High-strength steels with carbon equivalent (CE) values above 0.40–0.45 are susceptible to HIC. The susceptible microstructure is martensite, which forms when the HAZ cools too rapidly through the austenite-to-martensite transformation range. Martensite is hard, brittle, and traps hydrogen efficiently in its distorted lattice.
Carbon equivalent is typically calculated using the International Institute of Welding (IIW) formula:
CE = C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/15
Steels with CE > 0.45 are considered highly susceptible and typically require mandatory preheat in any credible welding procedure specification (WPS).
2. Diffusible Hydrogen
Hydrogen enters the weld pool primarily from moisture — in the arc atmosphere, on base metal surfaces, or in the flux coating of consumables. Cellulosic electrodes, moisture-contaminated low-hydrogen electrodes, surface condensation, and humid workshop conditions are the main sources. Diffusible hydrogen levels are classified in standards such as ISO 3690, which defines the test method for measuring hydrogen in ferritic steel arc welds.
3. Tensile Residual Stress
Welding always creates residual stress. The weld metal shrinks as it cools and solidifies; the surrounding base material constrains this shrinkage, leaving tensile stress in the HAZ — typically equal to or exceeding the material yield strength in highly constrained joints.
Thermal monitoring directly attacks conditions 1 and 2 by enforcing the preheat and interpass temperature regimes that slow HAZ cooling rates (preventing martensite formation) and promote hydrogen diffusion out of the joint before it becomes trapped.
How Preheat Prevents HIC: The Physics
Preheat works through two distinct mechanisms:
Mechanism 1 — Microstructure control: Elevating the base metal temperature before welding slows the cooling rate through the HAZ. A slower cooling rate through the 800–500 °C range (t8/5 time) shifts the transformation product away from martensite toward softer, tougher microstructures: bainite, ferrite-pearlite, or tempered martensite. The result is a HAZ that does not trap hydrogen efficiently and does not crack under residual stress.
Mechanism 2 — Hydrogen diffusion: Hydrogen diffuses through steel much faster at elevated temperature. Maintaining preheat temperature (and minimum interpass temperature) during and after welding keeps the joint warm long enough for diffusible hydrogen to escape to the atmosphere rather than becoming trapped at critical sites. Post-weld hydrogen bake-out (typically 250–350 °C for 2–4 hours) exploits the same mechanism.
The AWS D1.1/D1.1M:2025 Structural Welding Code — Steel and ISO 13916:2017 both specify minimum preheat temperatures as a function of steel grade, carbon equivalent, joint thickness, and hydrogen content level of the consumable. These are hard requirements in any qualifying WPS for susceptible steels.
Where Manual Compliance Fails
Despite being well-understood, preheat compliance is frequently violated in production:
Temperature gradients are large. A thermocouple or contact pyrometer measures one point. The actual temperature distribution across a thick joint or large HAZ can vary by 50–100 °C from point to point. A welder checks one location, finds 150 °C, and starts welding — while 30 cm away, the metal is at 80 °C, below the minimum required 120 °C.
Preheat dissipates during setup delays. The welder preheats to the correct temperature, then spends five minutes repositioning the workpiece, adjusting shielding gas, or waiting for a fitter. By the time welding begins, the temperature has dropped below minimum at the joint start.
Interpass temperature is not enforced. Multi-pass welds on thick sections require controlling both the minimum (to maintain hydrogen diffusion) and maximum (to prevent overheating and HAZ grain coarsening) interpass temperatures. A contact pyrometer is almost never used between passes in production — it is too slow and disruptive.
Records are absent or falsified. When preheat compliance is only verified by operator attestation, the paper trail provides no real traceability. An auditor reviewing a quality record cannot know whether the preheat was actually achieved before welding commenced.
The consequence: HIC occurs in welds that had, on paper, a compliant preheat procedure. The failure is in the gap between procedure and practice.
Real-Time Thermal Monitoring for HIC Prevention
A thermal camera mounted at the weld station changes the compliance model from attestation to evidence. The camera continuously captures the temperature distribution across the entire joint area — not a single point — at frame rates of 25–100 Hz.
Preheat Verification Before Arc Strike
The thermal monitoring system can be configured to enforce a preheat gate: the welding power source is enabled (or the operator receives a go/no-go signal) only when the joint area exceeds the minimum preheat temperature across a defined inspection region of interest (ROI). This eliminates the risk of starting too cold anywhere along the joint.
With HeatCore, the ROI is defined in the system configuration against the WPS requirements. If any pixel in the joint zone falls below the minimum preheat threshold, an alert fires and the timestamp is logged to the quality record.
Interpass Temperature Surveillance
Between passes, the thermal camera continues recording. The system monitors:
- Minimum interpass temperature: Triggers an alert and withholds welding permission if the joint drops below the WPS minimum before the next pass begins.
- Maximum interpass temperature: Triggers an alert if the joint exceeds the WPS maximum (typically 250–300 °C for many structural steels), preventing overheating-related HAZ degradation.
Because the camera images the entire joint area continuously, both conditions are checked across the full length of the previous pass — not just at a single probe point.
Cooling Rate Estimation and t8/5 Calculation
Advanced thermal analysis can extract the t8/5 cooling time (time for the HAZ to cool from 800 °C to 500 °C) directly from the thermal time-series data. This is the physically meaningful parameter that determines the HAZ microstructure — not preheat temperature alone.
Knowing t8/5 in real time allows correlation with hardness predictions from known CE and cooling models, flagging passes where the actual cooling rate deviated from the WPS-qualified envelope — even if the preheat temperature itself appeared compliant.
Automated Quality Records
Every temperature event — preheat achieved, arc start time, interpass temperature at each pass, maximum temperature reached, cooling curve — is automatically timestamped and stored in the digital quality record. This provides the evidence chain required by ISO 3834-2 for full weld traceability: not just that a procedure existed, but that it was actually followed.
Monitoring PWHT for Post-Weld Hydrogen Bake-Out
Post-weld heat treatment (PWHT) and hydrogen bake-out are frequently required for susceptible steels when preheat alone is insufficient to ensure hydrogen escape — particularly for thick sections, highly restrained joints, or applications with very low hydrogen tolerance (e.g., sour service oil and gas, nuclear).
Typical hydrogen bake-out requirements:
| Steel Category | Temperature | Hold Time |
|---|---|---|
| High-strength structural (CE 0.45–0.55) | 200–250 °C | 1–2 hours |
| Pressure vessel (P91, 2.25Cr-1Mo) | 300–350 °C | 2–4 hours |
| Sour service / NACE MR0175 | 200–250 °C | 4+ hours |
Thermal imaging during PWHT (including bake-out) maps the actual temperature distribution across the component, verifying that every region of the weld and HAZ achieved the required temperature for the required hold time — not just the thermocouple attachment points. This is directly equivalent to the function described in our post on post-weld heat treatment thermal monitoring.
Integration with ISO 3834 and WPS Requirements
ISO 3834-2 (comprehensive quality requirements for fusion welding) requires that welding is performed in accordance with an approved WPS and that the critical parameters — including preheat temperature and interpass temperature — are monitored and recorded.
Thermal monitoring transforms this requirement from a checkbox exercise into a live control system:
- WPS parameters are loaded into the monitoring system for each weld joint type.
- Camera ROIs are configured to cover the joint area.
- Monitoring is active from preheat through to the final pass and cool-down.
- Deviations are logged automatically, with timestamps, temperature values, and operator ID.
- Quality records are generated in a format directly usable for ISO 3834 audits, EN 1090 execution class documentation, and customer PPAP submissions.
This is the compliance architecture described in detail in our digital welding quality records guide and the ISO 3834 audit checklist post.
Carbon and Low-Alloy Steels: Where HIC Risk Is Highest
HIC risk is highest for:
- S355, S420, S460, S500, S550, S620, S690 high-strength structural steels (EN 10025-6 / TMCP grades)
- P265GH, P355GH, P355NH, 16Mo3 pressure vessel steels at thickness > 40 mm
- P91, P22, P11 creep-resisting alloys (also require PWHT for metallurgical reasons)
- API 5L X65, X70, X80 line pipe steels for oil & gas
- S960 and above ultra-high-strength steels (CE often 0.55–0.65)
For all of these, the combination of preheat and interpass temperature monitoring with digital record-keeping is the difference between a weld that passes destructive testing and one that fails in the field.
What a Preheat Non-Compliance Event Looks Like in Practice
Here is a concrete example of what thermal monitoring catches that manual inspection misses:
Scenario: 50 mm thick S460 structural steel, CE = 0.51, WPS requires minimum preheat 150 °C. Oxy-acetylene flame preheat applied, thermocouple reading 165 °C at centre of joint. Operator starts welding.
What the thermal camera reveals:
- At the joint start (left end), temperature is 162 °C — compliant.
- At the joint end (right end, 400 mm away), temperature is 112 °C — 38 °C below minimum.
- The operator welded from left to right. By the time the arc reached the cold end, the HAZ at that location experienced a cooling rate consistent with martensite formation.
Result without monitoring: Weld passes VT and PT. HIC initiates at the cold end 18 hours later. Part is scrapped — or worse, makes it into service.
Result with monitoring: Pre-weld gate fires at the right end of the joint. Operator notified. Additional preheat applied. Gate passes when whole joint reaches 150 °C. Welding starts. HIC risk eliminated. Event is logged in the quality record.
Quantifying the Business Case
HIC-related failures are among the most expensive in welded fabrication:
- Repair cost: Grinding out and re-welding a HIC crack in a thick-section pressure vessel joint can cost 5–15× the original welding cost, including NDE re-inspection.
- Delivery impact: A HIC failure discovered at final inspection delays delivery by days to weeks.
- Liability exposure: HIC in a pressure-retaining or structural application carries regulatory and liability consequences far beyond the direct repair cost.
- Insurance and certification impact: Repeated HIC events can trigger increased third-party inspection requirements or loss of EN 1090 / ISO 3834 certification.
The weld defect cost analysis we published previously showed that inline monitoring typically pays back in 3–6 months from defect prevention alone. For HIC-susceptible steels, the payback window is often shorter because the individual failure events are so costly.
Implementation Recommendations
For manufacturers working with HIC-susceptible steels who want to implement thermal-based preheat compliance monitoring:
Step 1: Classify your joints by HIC risk. Calculate CE for each material specification in your shop. Identify joints with CE > 0.40 and thickness > 25 mm as priority candidates for monitoring.
Step 2: Define the thermal compliance parameters from your WPS. Minimum preheat, minimum and maximum interpass, PWHT temperature and hold time if applicable.
Step 3: Configure the monitoring ROI. The camera field of view should cover at least 50 mm on each side of the joint centreline to capture the full HAZ thermal footprint.
Step 4: Set up automated quality record generation. Every preheat verification, interpass measurement, and arc start/stop event should be automatically logged with timestamp and operator ID.
Step 5: Integrate with your QMS. Non-conformance events (preheat gate failures, interpass exceedances) should flow directly into your NCR workflow for root cause and corrective action, as described in the welding NCR management guide.
Prevent Hydrogen-Induced Cracking with Real-Time Thermal Monitoring
HeatCore provides continuous preheat verification, interpass temperature surveillance, and automated quality records for HIC-susceptible steels — turning your welding procedure into an enforced control system, not just a paper requirement.
See HeatCore in actionSummary
Hydrogen-induced cracking is preventable — but only if the three conditions that enable it are actively controlled during production, not just specified in a procedure. Thermal monitoring gives manufacturers the capability to:
- Verify preheat compliance across the entire joint area before arc strike
- Enforce minimum and maximum interpass temperatures between passes
- Estimate HAZ cooling rates and flag passes outside the WPS-qualified envelope
- Generate automatic, timestamped quality records that satisfy ISO 3834-2, EN 1090, and customer audit requirements
- Integrate HIC-related non-conformances into structured CAPA workflows
For high-strength structural steel, pressure vessel, pipeline, and sour service applications, this is not incremental improvement — it is the difference between a weld you can certify with confidence and one that carries hidden risk.