Laser weld cracking is not one defect with one remedy. A crack can form while the weld pool solidifies, in a heat-affected region, at the end of a weld, or after cooling. Effective correction starts by locating and classifying the crack, then separating material, joint and process causes. Simply lowering laser power can hide the symptom while creating incomplete fusion or insufficient penetration.
First determine which crack you have
Cracks should be described by their position, direction, time of appearance and relationship to the fusion boundary. This evidence narrows the mechanism before parameters are changed.
| Crack pattern | Typical location or timing | Questions to investigate |
|---|---|---|
| Solidification or hot crack | Weld centerline or interdendritic path during solidification | Is the alloy or dissimilar combination susceptible? Is the bead shape deep and narrow? Is restraint high? |
| Liquation crack | Partially melted region near the fusion boundary | Are low-melting constituents or segregated phases present? Was peak temperature excessive? |
| Cold or delayed crack | Weld metal or heat-affected zone after cooling | Is the material hardenable? Are hydrogen, high residual stress or an unsuitable thermal cycle involved? |
| Crater or termination crack | Weld stop, overlap or pulse termination | Does energy end too abruptly? Is the termination region underfilled? |
| Interface or intermetallic crack | Dissimilar-metal boundary | Did a brittle reaction layer form? Is dilution or material pairing the real limitation? |
A surface crack may be visible under suitable lighting or liquid penetrant inspection, but absence of a surface indication does not prove the section is crack-free. Cross-sectioning, microscopy or another qualified nondestructive method may be necessary when the risk and acceptance criteria require it.
Why laser welds crack
Material and filler compatibility
Alloy composition controls solidification range, segregation, phase transformation and the ability to accommodate shrinkage. Some combinations are intrinsically more sensitive than others. With dissimilar metals, mixing can produce brittle intermetallic phases even when the bead looks smooth. The material certificates, coating or plating, filler selection and actual joint composition therefore belong in the investigation.
For Kovar laser welding, also consider the mating material and the function of the seal. A correction that produces an attractive bead but increases stress in a glass-to-metal or ceramic-to-metal assembly may not solve the system-level problem.
Joint restraint and fit-up
The molten pool contracts as it cools. A rigid fixture, inconsistent gap, poor edge preparation or a joint geometry that concentrates strain can exceed the ductility available during solidification. Inspect dimensional records and tack or clamping strategy instead of treating restraint as an invisible constant.
Energy distribution and travel conditions
Laser power, travel speed, focus position, beam size, incidence angle and waveform determine the weld pool geometry and thermal cycle. A very deep, narrow pool can make feeding during solidification more difficult. Excessive energy may enlarge the heat-affected zone, while insufficient energy can create lack of fusion that is sometimes mistaken for a crack in a poor image.
Start and stop behavior also matters. Abruptly ending a continuous weld or pulse train can leave a small crater that contracts without enough liquid metal to fill it. Controlled power ramp-down, an overlap strategy or a qualified termination tab may help, but each change must be verified on the actual joint.
Surface condition and hydrogen sources
Oil, moisture, oxide, cleaning residue, marking compounds and contaminated shielding hardware can change pool behavior or introduce gas. Cleaning needs a defined method, a maximum time between cleaning and welding, and controls that prevent recontamination. Reactive materials may require a controlled atmosphere rather than local shielding alone; see the glove-box laser welding machine and vacuum laser welding machine paths for environment options.
A practical crack-diagnosis workflow
- Contain the issue. Identify the affected lot, machine, program, material heat and fixture. Do not mix suspect and accepted parts.
- Record the crack. Capture magnified images and note its exact position, direction, length and relationship to starts, stops, edges and fusion boundaries.
- Confirm the indication. Use the inspection method specified by the drawing or procedure. When destructive analysis is allowed, section through representative locations rather than only the most obvious surface mark.
- Check material evidence. Review alloy certification, temper or heat treatment, coating, storage and dissimilar-material compatibility.
- Check the joint. Measure gap, mismatch, wall thickness, edge condition and fixture restraint. Compare failed and known-good parts.
- Review process history. Verify delivered power, speed, focus, shielding or chamber condition, pulse shape and termination sequence. A programmed setpoint is not the same as a measured process state.
- Change one cause family at a time. Use a controlled trial or design of experiments. Do not simultaneously change power, speed, gas flow and fixture pressure unless the test is designed to separate their effects.
- Requalify the result. Repeat the required surface, section, mechanical and leak tests. Confirm that preventing cracks did not introduce porosity, undercut or inadequate penetration.
Prevention principles
- Select material combinations and filler, when used, on the basis of metallurgical compatibility and the applicable procedure.
- Stabilize joint gap, alignment and clamping before fine-tuning laser settings.
- Use a weld pool shape and thermal cycle supported by cross-section evidence, not appearance alone.
- Define cleaning, handling and atmosphere controls as process variables.
- Program starts, overlaps and stops deliberately; inspect these locations during qualification.
- For high-consequence seals, connect crack inspection to a broader laser weld quality inspection plan.
There is no universal crack-free parameter set. Thickness, alloy heat, surface treatment, joint geometry, beam delivery and acceptance criteria all change the operating window. Published research can explain mechanisms, but production settings should be developed and qualified on representative parts.
