6.2 Porosity, Inclusions, and Lack of Fusion
Key Takeaways
- Gas porosity is typically rounded; shrinkage porosity is angular or dendritic and sits at last-to-freeze thermal centers
- Slag, oxide, and tungsten inclusions differ by process source (SMAW slag, oxidized films, GTAW electrode) and by density/morphology on RT and UT
- Lack of fusion (LOF) is incomplete joining of weld metal to base metal or between beads; lack of penetration (LOP) is incomplete fill of joint root or groove depth
- RT is generally strong for volumetric gas porosity and many inclusions; UT is strong for planar LOF when the beam addresses the unfused interface
- VT finds surface-breaking open defects and weld profile issues but cannot alone prove internal porosity or buried LOF
6.2 Porosity, Inclusions, and Lack of Fusion
Quick Answer: Porosity is a void; inclusions are foreign solid material; lack of fusion is a planar incomplete bond. Gas pores are rounded, shrinkage voids are jagged and location-specific. RT loves volumetric voids and many inclusions; UT loves planar LOF when oriented for reflection; VT only sees what opens to a surface. LOF ≠ LOP—know the geometry difference.
These discontinuity types dominate weld quality discussions on the ASNT NDT Level III Basic exam and appear in casting and brazing contexts as well. Master morphology first; method choice follows.
Porosity: Gas vs Shrinkage
Porosity means cavities left by gas or by solidification shrinkage without adequate liquid feed.
Gas porosity
Dissolved gases (hydrogen is infamous in aluminum and also important in steels and welds), entrained air, or gases from moisture and contamination nucleate bubbles as solubility drops during freezing. Morphology:
- Rounded or slightly elongated spheres/ellipsoids
- Smooth internal surfaces
- Scattered, clustered, or aligned with solidification or weld progression (wormhole / piping porosity when elongated)
- In welds: surface-breaking pores (visible) or subsurface pores
Process causes include wet electrodes or flux, contaminated joint surfaces, inadequate shielding gas, long arc, and base metal with high gas content.
Shrinkage porosity
When liquid cannot feed contraction, voids form in last-to-freeze regions. Morphology:
- Angular, dendritic, sponge-like, or interdendritic networks
- Located at thermal centers: heavy sections, weld crater ends, bosses, junctions
- May connect into larger shrinkage cavities
| Attribute | Gas porosity | Shrinkage porosity |
|---|---|---|
| Shape | Rounded | Angular / dendritic |
| Typical location | Anywhere gas is trapped; often near surfaces or upper regions | Thermal centers, last-to-freeze zones |
| Primary cause | Gas evolution / entrapment | Inadequate feeding of solidification shrinkage |
| RT appearance | Dark rounded spots (film/DR density increases) | Dark irregular / spongy zones |
| UT appearance | Discrete reflectors, often weaker planar character | Scattered/complex; not a single smooth plane |
Exam items love the shape-and-location rule: round + scattered → gas; jagged + hot-spot → shrinkage.
Inclusions: Slag, Oxide, and Tungsten
An inclusion is solid foreign material trapped in the metal. Unlike porosity (empty void), inclusions have material that may be denser or less dense than the matrix—important for RT contrast and UT scattering.
Slag inclusions
Common in SMAW, SAW, and flux-cored processes when slag is not fully removed between passes or is trapped by poor technique. Morphology is often elongated or irregular along the weld axis or between beads. On RT, slag may appear less dense (darker on negative film) than steel if composition is lighter; UT may show elongated reflectors. VT after grinding or on incomplete cleaning can reveal slag pockets at the surface.
Oxide inclusions and films
Oxide films form from inadequate shielding, turbulent welding of aluminum, or reoxidation of melt. They can be film-like and planar—more LOF-like in NDT response than a round slag blob. Aluminum welds are notorious for oxide-related incomplete bonding. In castings, dross and oxide skins act similarly.
Tungsten inclusions
GTAW (TIG) can deposit tungsten particles if the electrode dips into the puddle or is overheated/contaminated. Tungsten is much denser than steel or aluminum, so on RT it appears as bright (white) spots on negative film—opposite of typical gas porosity (dark). That density contrast is a classic exam discriminator.
| Inclusion type | Typical process | RT density cue (concept) | Notes |
|---|---|---|---|
| Slag | SMAW / SAW / FCAW | Often darker than metal if lower density | Elongated between passes |
| Oxide film | GMAW/GTAW Al, turbulent melts | May be subtle; planar | Can behave like LOF for UT |
| Tungsten | GTAW electrode contamination | Brighter (higher density) | Discrete particles |
| Sand / refractory | Casting | Variable density | Near mold surface often |
Lack of Fusion (LOF) vs Lack of Penetration (LOP)
These terms are not interchangeable.
Lack of fusion (LOF)
Incomplete fusion between weld metal and base metal (sidewall or groove face), or between successive weld beads/passes. Causes: low heat input, wrong angle, travel speed too high, arc not directed into the sidewall, oxide films, or cold metal. Geometry is planar along the unfused interface—often parallel to the original joint face or bead boundary.
NDT: LOF is crack-like in ultrasonic behavior. Angle-beam UT is a primary volumetric method for weld sidewall LOF. RT may miss tight LOF with little gap. Surface LOF at toes or faces may show with VT/PT/MT if open.
Lack of penetration (LOP) / incomplete penetration
Incomplete penetration means the weld metal did not reach the required root or full joint thickness—unfilled root gap, incomplete groove fill, or failure to melt through the root face as the design/WPS requires. Geometry is a root or mid-thickness unfilled region, often elongated along the weld length, which may look like a dark straight indication on RT when open.
| Feature | Lack of fusion (LOF) | Lack of penetration (LOP) |
|---|---|---|
| Definition | Incomplete bonding of metal to metal that should have fused | Incomplete fill/penetration of joint depth or root |
| Typical plane | Sidewall, interpass, groove face | Root centerline / design throat path |
| Common cause | Technique, oxide, heat input | Inadequate amperage, root opening, joint access |
| Preferred NDT | UT (planar); surface methods if open | RT often useful if open gap; UT; VT of root if accessible |
Both can reduce cross-section and act as stress raisers. Codes define acceptance differently; Level III applies the specified criteria rather than inventing severity ranks.
Detection Strengths: RT, UT, and VT
Radiographic testing (RT)
Strengths: Volumetric gas porosity, many slag inclusions, tungsten (by density), open root defects and incomplete penetration with gap, some elongated wormhole porosity.
Limitations: Tight LOF and tight cracks with minimal opening and poor alignment; very small porosity below IQI/sensitivity limits; interpretation depends on technique (kV, density, geometry).
Ultrasonic testing (UT)
Strengths: Planar LOF and crack-like interfaces when beam orientation is correct; depth location; often better than RT for tight planar weld flaws.
Limitations: Small scattered porosity may give weak or noisy responses; coarse grain and austenitic welds complicate signals; requires access for probe coupling and skilled interpretation under procedure.
Visual testing (VT)
Strengths: Surface pores, open slag, undercut, profile, incomplete fill at accessible surfaces, misalignment; cheapest first filter.
Limitations: No internal porosity or buried LOF detection; surface condition and lighting control quality; borescopes help only where optical access exists.
Practical pairing: VT + RT for many pressure-vessel weld quality schemes emphasizing volumetric weld metal quality; VT + UT for crack/LOF-sensitive joints; MT/PT added for surface crack detection on appropriate materials. Level III procedures state the combination required by code and risk—not a single universal method.
Process → Discontinuity → Method Map (Weld Focus)
| Observation / process clue | Likely discontinuity | First-line NDT thinking |
|---|---|---|
| Wet electrode, poor gas cover | Gas porosity | RT (volume); VT if surface |
| Slag not cleaned between SMAW passes | Slag inclusion | RT; UT; VT if exposed |
| Tungsten dipped in GTAW puddle | Tungsten inclusion | RT (high-density spots) |
| Cold sidewall, steep groove, fast travel | Sidewall LOF | Angle-beam UT; RT secondary |
| Root opening closed, low amperage | LOP / incomplete penetration | RT root; UT; VT root side |
| Crater left unrefilled | Crater crack / shrinkage | VT/PT/MT; UT as needed |
Level III Evaluation Notes
Acceptance is never "I see porosity, so reject." Charts and code tables specify size, spacing, and accumulated length. Clustered porosity near a surface may be more serious for fatigue than sparse internal pores of the same total area. LOF is often more structurally concerning than rounded porosity of similar length because it is sharp and planar. Document indication type, location, size method, and code reference. When methods disagree (RT clear, UT shows planar reflector), investigate orientation and technique before dismissing either—do not average them into a vague "maybe."
Exam traps: Calling every dark RT spot "LOF"; calling every planar UT signal "porosity"; assuming VT alone qualifies a full-penetration weld; confusing tungsten (bright on film) with gas porosity (dark). Keep morphology and process cause linked to method physics.
On a radiograph of a multipass steel weld, several small, rounded dark indications are scattered through the weld metal. Which discontinuity is most consistent with this description?
Which statement correctly distinguishes lack of fusion from lack of penetration?
Bright (high-density) discrete spots on a radiographic image of a GTAW weld most likely represent:
For a suspected tight sidewall lack of fusion in a thick groove weld, which method is generally more effective than film radiography alone?