5 Ways Gases Form Porosity in Welds
Weld porosity forms when gas enters, develops inside or becomes mechanically trapped in molten metal—and the solidification front closes before the bubbles escape. The useful question is not simply “Which gas caused it?” but “Where did the gas come from, and what stopped it leaving?”
Welding process visual
What causes weld porosity?
Weld porosity is a cavity-type discontinuity created when gas remains in weld metal as it solidifies. Gas may be drawn in from the atmosphere, released by moisture or hydrocarbons, evolved from a coating, generated or rejected by weld-pool chemistry, or isolated by an unstable molten pool or laser keyhole.
A visible pinhole, a cluster on radiography and a centerline cavity may all be called porosity, but they do not automatically share one cause. Shape and location are useful clues. They must be checked against the welding process, base and filler materials, coating, shielding delivery, consumable history, joint geometry and process record.
Lack of fusion, slag inclusion, shrinkage cavities and cracks have different formation mechanisms. Confirm the discontinuity type before changing the process.
Keyhole porosity is an entrapment problem—not only a contamination problem
Deep penetration depends on a vapor cavity that continually opens, changes shape and closes behind the beam. If the lower keyhole necks or collapses, a pocket of vapor can detach. The moving solidification front may then capture it as a centerline or root pore.
A clean surface is still necessary, but it does not guarantee a stable keyhole. Power, speed, focus, beam profile or wobble, penetration state, fit-up and shielding must be treated as one process window. A single increase or decrease in power is not a universal cure.
Recoil pressure and liquid flow maintain a deep keyhole.
Material, gap and process changes disturb the balance.
The detached bubble moves through the liquid weld pool.
An internal pore remains at the root, centerline or transition.
Industrial laser welding visual
Use pore shape as a clue—not a verdict
Fine round pores often support a dissolved-gas hypothesis. Elongated or piping pores suggest gas continued to evolve while the pool was freezing. A localized cluster suggests a local condition. None of these patterns uniquely identifies the gas source.
| Mechanism | Where gas may come from | Common clues | Confirm first | First controlled action |
|---|---|---|---|---|
| Shielding failure | Air entering an unstable gas envelope | Scattered or surface-opening pores; issue responds to drafts, torch position or setup changes | Nozzle flow, leaks, diffuser/nozzle condition, torch angle and distance | Restore the qualified gas setup and repeat the same coupon |
| Hydrogen source | Moisture, oil, hydrated oxide, filler, electrode or flux | Often fine internal round pores; aluminum is particularly sensitive | Storage, condensation, cleaning, filler lot and gas-line moisture | Remove one suspected source, preserve all other variables and re-test |
| Residue or coating vapor | Paint, primer, zinc, oil, sealant or contamination trapped at an interface | Blowholes, piping, clusters near coated laps; spatter or expulsion may accompany the defect | Coating identity, interface location, gap and escape path | Compare a prepared or purpose-vented coupon under an approved procedure |
| Metallurgical reaction | Gas rejection during cooling or chemistry involving oxygen, carbon and deoxidizers | Fine distributed pores that may follow a material heat or consumable lot | Base/filler certificates, surface oxide, gas and qualified consumable combination | Escalate for metallurgy and consumable review before broad parameter changes |
| Pool or keyhole entrapment | Bubble isolation caused by unstable transfer, turbulence, crater closure or keyhole collapse | Centerline/root pores, start/end pores or defects linked to process instability | Transfer, wire feed, termination, power-speed-focus relationship and fit-up | Stabilize one qualified process variable and compare macrosection or NDE evidence |
Weld radiograph and cross-section comparison
Internal evidence can change the diagnosis
This research figure compares radiographs and longitudinal sections from a controlled laser-welding study. It demonstrates why a top-bead photograph is not enough: internal pore number, size and location can change even when the surface offers limited information.
The figure is evidence from one experimental design, not a universal porosity chart. Use the examination method and acceptance criteria required for the actual component.
Figure: Pang, Dai, Zhang and Zhang, Photonics 2021, CC BY 4.0.The gas envelope is a flow system
The regulator is only one part of the system. Gas must travel through the line, valve, diffuser and nozzle or cup, then remain coherent around the electrode, arc or laser interaction zone.
If the jet becomes turbulent, it can entrain air. Verify coverage at the torch and on the actual joint instead of compensating for a draft with an arbitrary flow increase.
GMAW shielding diagram
Diagram: Spangineer and Razorbliss, via Wikimedia Commons, CC BY-SA 3.0.
Porosity checks change with the welding process
A gas-shielded arc, a flux-generated shield and a laser keyhole do not fail in the same way. Start with the variables that actually control gas entry and bubble escape for the selected process.
Gas delivery and transfer stability
Check: gas type and delivery, leaks, nozzle and diffuser, drafts, torch angle, electrode extension, base/wire cleanliness, wire-feed consistency and transfer mode. Spatter obstructing the nozzle can disturb coverage.
Cup coverage and a clean gas path
Check: cup or gas lens, torch assembly, leaks, purge, pre-flow/post-flow per procedure, long-line surge, moisture and possible water-cooled torch leakage. Tungsten contact is contamination and arc instability—not a gas source by itself.
Electrode condition and arc length
Check: base-metal cleanliness, excessive arc length and electrode exposure. Storage, holding and redrying depend on the exact classification and manufacturer; do not treat cellulosic electrodes like low-hydrogen electrodes.
Distinguish the shielding route
Check: wire moisture or damage, contamination, angle, stickout and parameters. FCAW-G also depends on external gas delivery; FCAW-S does not. Avoid assuming every flux-cored pore is “CO trapped under slag.”
Keyhole, fit-up and shielding
Check: surface preparation, shield route, power, speed, focus, wobble or beam profile, partial/full penetration behavior, joint gap and seam tracking. Aluminum hydrogen and zinc vapor can coexist with keyhole entrapment.
Give vapor a controlled route
Check: coating identity and thickness, zero-gap interfaces, trapped volumes, vent features, fume control and whether the approved procedure permits coating removal or a designed gap.
Build a porosity root-cause hypothesis
Choose the closest evidence. The result ranks a starting route for investigation; it does not identify the gas with certainty or decide whether the weld meets a code.
Change one field at a time to see which evidence drives the hypothesis. Preserve the original failed part and the process record before running trials.
Collect evidence first
The current inputs do not support one strong mechanism. Start by confirming the process, material, pore location and the most recent production change.
An eight-step porosity investigation
A disciplined sequence prevents several variables from moving at once. It also creates evidence that can be transferred from the shop floor to welding engineering, inspection and equipment support.
Confirm acceptance criteria
Identify the applicable code edition, contract, joint category, load condition, quality level and examination method before calling a pore acceptable or rejectable.
Map the evidence
Record pore size, shape, distribution, depth and relation to the start, stop, root, centerline, interface, pass or one material region.
Separate surface from internal
Visual inspection finds surface-opening evidence. Use the required NDE or sectioning method when internal volumetric porosity matters.
Verify shielding delivery
Trace gas from the source to the torch and actual weld zone. Inspect leaks, restrictions, timing, hardware, position and airflow.
Audit material and consumables
Check base metal, filler, electrodes or flux, storage history, condensation, lot changes, certificates and cleaning records.
Identify coatings and traps
Confirm paint, zinc, primer, oil, sealant, corrosion products, lap interfaces and any volume that can hold gas or vapor.
Change one variable
Choose the highest-value hypothesis. Hold material, joint, inspection and all other qualified variables constant for the comparison.
Validate and document
Compare equivalent coupons with the required visual, RT, CT, UT or section evidence, then update the approved procedure if justified.
Classification is not the same as acceptance
ISO 6520-1 classifies geometric welding imperfections. It does not, by itself, tell a fabricator whether a pore is acceptable. Acceptance depends on the product standard, quality level, process, material, loading, examination method, contract and code edition.
For example, ISO 5817 provides quality levels for certain fusion-welded steel, nickel, titanium and their alloys but excludes beam welding from its scope. Beam-welded joints may be governed by ISO 13919-1 or another applicable product specification. AWS and ASME requirements likewise depend on the governing construction code and application; ASME Section V addresses NDE methods and Section IX addresses qualification, not one generic production-weld pore allowance.
Useful for surface-opening pores, bead profile and contextual clues. A sound-looking face does not rule out internal porosity.
Often effective for volumetric discontinuities. Suitability depends on joint geometry, thickness, material, access and the inspection plan.
Can be part of an approved examination strategy, but detectability and interpretation depend strongly on geometry, technique and calibration.
Useful for development and root-cause work because pore depth, interface relation and penetration behavior become easier to compare.
Removing the gas source can create another hazard
Cleaning, grinding, heating or welding coated metal changes the exposure profile. Identify the actual base metal, coating, residue, filler and flux before selecting controls.
Heat and arc radiation can form highly hazardous decomposition products, including phosgene. Follow the SDS and keep incompatible degreasing operations away from hot work.
Galvanized, painted and unknown surfaces require a hazard assessment and suitable local exhaust close to the plume. Zinc oxide fume can cause metal fume fever.
Argon, helium and other gases can displace oxygen, particularly in enclosed or confined spaces. Ventilation and confined-space controls must match the site.
Flammable solvents require evaporation, ventilation and fire controls. Material compatibility, SDS instructions and the hot-work procedure govern cleaning.
Send the evidence needed to diagnose the pore—not only a bead photo
Oceanplayer can review laser-welding applications and build a sample test around the material, joint and target result. A stronger evidence package makes the first test more useful and reduces random parameter changes.
Include these details
- Welding process and machine or laser-source model
- Base grade, filler, thickness, joint and fit-up
- Coating, cleaning and consumable-storage method
- Shielding gas, nozzle or cup and delivery details
- Current/power, wire feed, speed, focus and CTWD as applicable
- Pore location, when the issue began and recent changes
- RT, CT, macrosection or close-up evidence
- Required code, specification and production target
Related welding guides and tools
Common questions about gas porosity
What is weld porosity?
Weld porosity is a cavity-type discontinuity formed when gas remains inside or opens to the surface of weld metal as it solidifies. It may appear as isolated pores, clusters, piping or other distributions. Confirm that the indication is porosity rather than slag, lack of fusion, shrinkage or cracking.
What gases commonly cause porosity in welds?
Hydrogen, nitrogen, oxygen-bearing atmosphere and reaction products such as carbon monoxide may be involved, depending on the material and process. Coatings and residues can also generate vapor. A gas name alone is not a diagnosis; identify both its source and why it could not escape.
Can too much shielding gas cause porosity?
Yes. Too little flow may leave the pool exposed, while excessive velocity can create turbulence and entrain air. The correct flow depends on the process, gas, cup or nozzle, position, joint, airflow and qualified procedure. Verify delivery at the torch instead of relying only on the regulator.
Why is aluminum welding especially sensitive to hydrogen?
Molten aluminum can hold substantially more hydrogen than solid aluminum. As the pool solidifies, hydrogen may leave solution and form pores. Moisture, hydrated oxide, hydrocarbons, filler handling and the gas system are therefore important investigation areas.
Why can galvanized steel create blowholes or wormhole-like pores?
Zinc can vaporize before the steel melts. In a tight lap joint, pressurized vapor may pass through or destabilize the molten pool, producing porosity, blowholes, spatter or expulsion. Pore shape alone does not prove zinc; confirm coating location and joint geometry.
Can pore shape identify the exact root cause?
No. Shape and distribution help prioritize hypotheses, but similar patterns can come from shielding, contamination, coatings, chemistry and pool instability. Combine morphology with process history, material records, gas checks and controlled trials.
Can I weld over existing porosity?
Do not assume an over-weld is acceptable. The repair method, defect removal, re-examination and procedure qualification depend on the governing code and repair plan. Welding over a gas cavity without removing the cause can trap more gas or introduce lack of fusion.
Does preheating always reduce weld porosity?
No. Preheat may help manage moisture or thermal behavior in some qualified procedures, but the required temperature depends on material, thickness, consumable, hydrogen control and metallurgy. Uncontrolled preheat can damage coatings, alter properties or create other defects.
Which inspection method finds internal porosity?
Radiography is commonly effective for volumetric porosity, while CT and macrosections are powerful development tools. Ultrasonic techniques may also be specified. The correct method depends on geometry, thickness, material, access and the governing inspection plan.
How is laser welding porosity different from MIG or TIG porosity?
All can involve shielding and contamination, but deep-penetration laser welding adds a dynamic keyhole that can collapse and trap vapor. Laser diagnosis therefore includes focus, power-speed balance, penetration mode, beam profile, fit-up and molten-pool dynamics in addition to cleanliness and gas delivery.
Sources used for this guide
- TWI — What is porosity and how can it be prevented?
- TWI — Avoiding porosity when welding aluminium
- AWS Welding Digest — What to know about aluminum GMAW setup
- Miller — Best practices for proper GTAW shielding
- TWI — Typical defects in laser welds
- Materials 2024 — Keyhole stability and porosity in laser welding of aluminum alloy
- Journal of Materials Processing Technology — Bubble formation and porosity in laser welding
- ISO 6520-1 — Classification of geometric imperfections
- ISO 5817:2023 — Quality levels for imperfections in fusion-welded joints
- OSHA — Welding, cutting and brazing chemical hazards
- NIOSH Pocket Guide — Zinc oxide
- Lincoln Electric — Electrode storage and redrying guidance
A gas source alone does not guarantee a pore
The molten pool must first receive or create gas. A bubble then nucleates, grows and moves. Porosity remains only when the solidification front captures that bubble before it reaches a free surface.
This explains why a change in travel speed, penetration mode, joint gap or pool turbulence can alter porosity even when the contamination source has not changed.
Atmosphere, moisture, chemistry, coating or vapor cavity.
Local pressure and solubility conditions permit a gas phase.
Buoyancy and liquid flow compete with the moving solid front.
The remaining cavity becomes a surface or internal pore.
Trace the gas from source to capture
Each route below includes the evidence to collect before changing settings. The first controlled action is deliberately conservative: restore or verify one variable, repeat a coupon and compare the same inspection result.
Atmospheric air enters through an unstable shielding envelope
GMAW, GTAW, gas-shielded FCAW and many laser welding setups rely on a controlled gas envelope. A leaking connection, damaged hose, blocked diffuser, contaminated nozzle, unsuitable cup or nozzle, excessive stand-off, poor torch angle or a local draft can allow air to reach hot metal. Insufficient flow can leave the pool uncovered; excessive flow can become turbulent and draw surrounding air into the stream.
Do not diagnose shielding only from the regulator setting. Verify the gas type, supply, connections and actual delivery at the torch or nozzle. Compare the result with the qualified WPS and equipment guidance; there is no single flow rate that applies across cup sizes, nozzle geometries, processes and positions.
The defect began after a nozzle change, hose repair or relocation; it changes with airflow; surface-opening or scattered pores appear over much of the weld.
Restore the qualified gas path, inspect seals and nozzle condition, measure delivery at the torch and repeat the same coupon without changing heat or travel.
Shielding-gas supply visual
Check the complete gas path
- Cylinder or bulk supply, gas identification and contamination risk
- Regulator, flow control, fittings, O-rings and hose integrity
- Solenoid timing, purge condition and long-line surge
- Diffuser, gas lens, cup or nozzle cleanliness and condition
- Torch angle, distance, travel direction and surrounding airflow
- Backside or trailing shielding when the material and WPS require it
Moisture and hydrocarbons supply hydrogen
Water, condensation, oil, grease, fingerprints, cutting fluid, dirty filler, damp electrodes or flux and moisture in the gas-delivery system can introduce hydrogen-bearing material. Aluminum is especially sensitive because molten aluminum can dissolve far more hydrogen than solid aluminum. As the pool freezes, the excess gas may form fine internal pores instead of remaining in solution.
For steel welds, moisture in low-hydrogen electrodes or fluxes can raise diffusible hydrogen. Porosity may occur, but hydrogen-assisted cracking can be the more serious risk. Storage, holding and redrying are not interchangeable: follow the exact consumable classification, manufacturer instructions and qualified procedure rather than applying a universal oven temperature.
Oxide, wire and condensation
Hydrated oxide, contaminated filler, cold material moved into humid air and wet gas lines can all contribute. Clean and store materials with a documented, material-compatible method.
Porosity is not the only hydrogen risk
Review base-metal condition, consumable exposure and preheat/interpass requirements. Do not treat the absence of visible pores as proof that hydrogen is controlled.
History matters
Record storage time, packaging condition, oven logs, ambient changes, filler lot, cleaning method and whether the issue follows one batch or workstation.
A clean top bead can still hide internal porosity
Surface color and appearance cannot establish whether aluminum contains internal hydrogen pores. Use the examination method required by the product quality plan, then compare equivalent coupons after controlling cleanliness and material handling.
Fine, round internal pores may support a dissolved-gas hypothesis, but they do not prove the source. Shielding leaks, local contamination and process instability can create similar evidence.
Aluminum weld-surface visual
Image: W. S. Yerazunis, via Wikimedia Commons, public domain.
Residues decompose or volatile coatings create high-pressure vapor
Oil, grease, excessive primer, paint, sealant, moisture-bearing corrosion products and contaminated consumables can decompose when heated. Volatile metallic coatings deserve separate attention. Zinc, for example, can vaporize before steel melts. In a tight galvanized lap joint, that vapor may have no easy route except through the molten pool or laser keyhole.
The result may include blowholes, piping or wormhole-like cavities, spatter and metal expulsion. Those shapes are clues—not proof of zinc. A coating-compatible solution may involve controlled removal, a designed escape path, joint-gap control, altered sequencing or a validated process window. The correct action depends on the part, coating thickness, joint and applicable procedure.
Identify the coating and SDS, map where pores intersect the coated interface, inspect trapped volumes, record fit-up and compare coated versus prepared coupons.
A fixed grind width, weave width or percentage change in travel speed can introduce coating damage, fusion problems or excessive heat.
Review the dedicated galvanized steel laser welding guide for coating, joint-gap and process-validation considerations.
Weld-pool chemistry rejects or generates gas during cooling
Liquid and solid metal do not hold gases in the same way. When the solubility limit falls during solidification, dissolved gas can leave solution and nucleate bubbles. Aluminum and hydrogen are the familiar example. In some steel and nickel welding conditions, oxygen, carbon and deoxidizing elements can also influence carbon-monoxide formation and whether a sound weld is produced.
This route cannot be reduced to “dirty steel makes CO.” Base-metal composition, filler chemistry, deoxidizer level, shielding gas, surface oxide and the thermal cycle interact. If porosity follows a material heat, wire lot or consumable change, preserve samples and certificates before experimenting with many machine settings.
Dissolved gas, oxygen-bearing surface or reaction ingredients enter the molten pool.
Temperature, composition and local equilibrium change as the weld cools.
Gas is rejected from solution or generated by reaction within the pool.
The bubble remains when the surrounding weld metal becomes solid.
Pool turbulence, unstable transfer or rapid freezing traps bubbles
A known gas source is only half the mechanism. Erratic droplet transfer, inconsistent wire feed, an unstable arc, unsuitable heat input or travel, turbulent liquid flow, poor fit-up and an abrupt weld termination can keep bubbles inside the pool. Changing heat or speed without understanding the mechanism can move the pore instead of removing it—and may create fusion, undercut or distortion problems.
In deep-penetration laser welding, the vapor cavity is a dynamic keyhole. Oscillation, necking or collapse can isolate a bubble near the root or centerline. Whether that bubble escapes depends on coupled variables such as laser power, travel speed, focus position, beam profile, penetration mode, joint gap, material condition and molten-pool flow.
Record arc sound, transfer mode, wire-feed stability, CTWD or arc length, starts/stops, joint fit and the exact location of the pores.
Record power, speed, wobble, focus, spot size, penetration state, seam tracking, joint gap, shielding and pore depth on a macrosection or CT/RT result.