Weld cracks are not all the same problem. A hot crack forming at the centerline of a weld bead as it solidifies has a completely different mechanism — and a different WPS fix — than a cold crack found in the HAZ three days after the weld passed visual inspection. Treating them as interchangeable leads to wrong countermeasures: increasing preheat addresses cold cracking but does little for solidification cracking driven by contaminated filler.

For a CWI or QC manager, understanding the classification and mechanism of each crack type is essential for writing a WPS that prevents the right failure mode on a given material and joint configuration.

Hot cracking and solidification cracking

Hot cracking occurs during and immediately after solidification. As the weld pool cools, the last liquid to solidify is concentrated at the center of the bead and at grain boundaries. If that residual liquid contains elevated levels of sulfur, phosphorus, or other elements that form low-melting-point films, the solidifying grains cannot knit together under the shrinkage stresses of cooling. The result is a crack that follows the weld centerline longitudinally, or appears as a crater crack if the arc is broken abruptly.

Key WPS-level controls:

  • Filler metal selection. Use filler metals with low sulfur and phosphorus in the classification limits — AWS A5 filler metal standards specify maximum levels. For SMAW, the lime-based E7018 flux system is less susceptible to hot cracking than acid-flux types.
  • Heat input. High heat input produces a wider, flatter bead profile with more centerline segregation. Controlling travel speed and amperage to produce a convex-to-flat bead cross-section reduces susceptibility. This is part of why the WPS specifies current and travel speed ranges, not just electrode type.
  • Base metal chemistry. High-sulfur structural steels (older materials, free-machining grades) are susceptible. For base metals with elevated sulfur, the WPS should specify a filler metal with a high Mn:S ratio that neutralizes the sulfur's grain-boundary effect.

Hot cracks found during production are highly visible in the weld bead surface — they open during cooling and are typically wide enough for unaided visual inspection. A CWI who sees longitudinal centerline cracking in weld beads should investigate base metal chemistry and heat input before accepting the weld or issuing a repair.

Hydrogen-assisted cold cracking (HACC)

HACC is the most prevalent crack type in structural carbon and low-alloy steel fabrication. Three conditions must coexist for it to occur:

  1. Diffusible hydrogen in the weld metal (from moisture in electrode coatings, shielding gas contamination, damp base metal surfaces, or flux)
  2. Tensile residual stress from welding shrinkage and restraint
  3. A susceptible microstructure — typically martensite or bainite in the HAZ, more common in higher-carbon-equivalent steels

The crack commonly forms in the HAZ rather than the weld metal, because the HAZ in hardenable steels produces exactly the brittle microstructure hydrogen needs to cause fracture under residual stress. HACC can be delayed — it may not appear for hours or days after welding, which means a visual acceptance done the same day as welding can miss it.

WPS-level controls from AWS D1.1:

  • Preheat and minimum interpass temperature. Heat keeps the HAZ above the martensite-start temperature longer, allowing hydrogen time to diffuse out before transformation. AWS D1.1 prequalified preheat requirements (based on base metal group and plate thickness) are the code's HACC prevention threshold. Do not arbitrarily lower preheat to speed production.
  • Low-hydrogen filler metals. The H-designator on the electrode (H4, H8, H16) is the maximum allowable diffusible hydrogen in mL per 100 g of deposited weld metal. For crack-sensitive joints, the WPS should specify H4 or H8. The WPS must list the H-designator — a CWI should verify the electrode packaging matches.
  • Moisture control. Base metal must be dry at the weld joint. Rain, dew, and surface condensation introduce hydrogen. AWS D1.1 Clause 5 prohibits welding on wet surfaces; preheat to a minimum temperature (typically 50°F above ambient for moisture) before striking an arc.
  • Post-weld hydrogen release. For restrained high-strength steel joints, PWHT or a post-heat soak (below PWHT temperatures) immediately after welding drives out diffusible hydrogen before cracking initiates.

Rule library based on AWS D1.1:2025; verify against your governing edition. The AHJ or project contract may specify AWS D1.1:2020 or an earlier edition — confirm before issuing procedures.

For a complete treatment of H-designators and electrode storage, see Hydrogen Cracking Prevention and WPS Documentation and Low-Hydrogen Electrode Conditioning: H4, H8, H16 on a WPS.

Lamellar tearing

Lamellar tearing is a base metal failure, not a weld metal failure. It occurs through the thickness of a plate in the short-transverse (Z) direction, driven by the through-thickness tensile stress that welding places on a restrained plate. The cracks propagate along flat manganese sulfide (MnS) inclusions that are elongated in the rolling direction of structural plate.

Structurally, it appears most often in:

  • Heavy T-joint connections where a connecting member bears down through the thickness of a base plate (column base plates, continuity plate-to-column-web connections)
  • Repair welds in thick plate
  • Corner joints in box columns where CJP groove welds pull transversely on the adjacent plate

The WPS alone cannot always prevent lamellar tearing — the metallurgy is in the base metal, not the weld. But WPS-level decisions reduce risk:

  • Buttering the through-thickness surface. Depositing a low-strength, ductile buffer layer on the through-thickness face before making the joint weld distributes the strain over more material.
  • Preheat and heat input. Adequate preheat reduces shrinkage stresses during cooling, which reduces the peak through-thickness strain.
  • Sequencing weld passes. Completing full-penetration weld segments in a balanced sequence reduces the net shrinkage force on the susceptible plate.

When lamellar tearing risk is high (thick plates, restrained T-joints, elevated sulfur base metal), the project specification may call for through-thickness tension test coupons (ASTM A770) or UT inspection of the base metal prior to welding. See Lamellar Tearing Risk in Heavy Plate WPS Design for a detailed treatment.

Reheat cracking and stress-relief cracking

Reheat cracking occurs during PWHT or multi-pass welding when the weld metal or HAZ loses ductility due to carbide precipitation in the grain boundaries. It is primarily a concern in alloy steels containing chromium, molybdenum, vanadium, or niobium (Cr-Mo and CrMoV steels), where these elements stabilize carbides at PWHT temperatures.

For structural carbon steel (A36, A572), reheat cracking is not a significant concern. For low-alloy steels used in pressure vessels, high-temperature service equipment, or bridge structures, the WPS must address:

  • Controlled PWHT heat-up and cool-down rates
  • Holding temperature ranges that avoid the cracking-susceptible temperature zone
  • Post-PWHT NDE (typically PWHT is followed by VT and often UT or MT)

Quick decision guide for WPS writers

Crack type Primary WPS control Timing
Hot / solidification Filler metal sulfur content, bead shape, heat input During welding
HACC / cold Preheat, H-designator, moisture control Hours–days post-weld
Lamellar tearing Buttering, preheat, pass sequence During/after welding
Reheat PWHT rate/temp control, alloy awareness During PWHT

Writing a WPS that addresses the right crack type for the material and joint configuration is foundational quality control. For WPS and PQR document generation with built-in AWS D1.1:2025 compliance checks, see pricing.