Liquid Penetrant Testing for Welding: A Complete Guide

Two inspectors can look at the same fillet weld on a stainless pressure vessel and see the same thing: clean metal, no undercut, no visible crack. Then one of them sprays on a fluorescent penetrant, waits, wipes it back, and lays on a developer — and a hairline crack running the full length of the heat-affected zone lights up green in a darkened room. That moment is the whole reason liquid penetrant testing has survived since the railroads first dipped cast wheels in used oil and chalked them: it turns an invisible flaw into something you can see, measure, photograph, and argue about on a report.

What Liquid Penetrant Testing Is (and What It Is Not)

Liquid penetrant testing, usually shortened to PT, is a nondestructive examination method that uses a low-viscosity liquid to reveal cracks and other flaws that break through the surface of a weld or weldment. Penetrant is applied to a prepared surface, soaks into any open discontinuity, then gets drawn back out with a developer. Where liquid traveled into a flaw and came back out, you see a visible mark proving a surface-breaking defect exists. Nothing more, nothing less.

PT shows up under many names for the same process. The National Board notes it is called penetrant testing (PT), liquid penetrant testing (LP) or dye penetrant testing (DP), while ASME and the National Board prefer the formal phrase liquid penetrant examination. Dye penetrant often refers to the visible-red variety, and older shops may still call the whole thing the “oil and whiting” method.

The first documented use of penetrant testing was in the railroad industry, where cast steel wheels were dipped in used oil, dried, then chalked. Oil trapped in a surface flaw bled back out through the chalk and left a dark mark, revealing a crack nobody could see by eye. Crude, but the physics were right, and ASNT traces the method to that oil-and-chalk lineage. Modern penetrant testing is a controlled, repeatable version of it.

Where PT Sits Among NDT Methods

OSHA’s Technical Manual groups examination methods into two families. Visual testing, liquid penetrant testing and magnetic particle testing are surface methods, meaning they find only conditions open to or very near the surface. Radiography and ultrasonic testing are volumetric methods, detecting conditions buried inside the part. PT belongs on the bench with a magnifier and a UV-A lamp, not in the radiographic bay.

What PT Is Actually For

According to ASNT, the core objectives of liquid penetrant testing fall into three buckets:

• Detecting surface defects such as cracks, seams and porosity open to the surface
• Quality control of manufactured parts and completed welds during fabrication
• Maintenance and in-service inspection of components in service, including welds that are hard to reach

That third category is where PT earns its keep. An aerospace weld, a pressure vessel nozzle, a pipeline girth weld or a bridge girder connection all raise the same problem: something may be cracked, but you cannot see it and cannot take it apart.

What PT Cannot Do

The limits matter, because the most common field mistake with penetrant testing is expecting it to find something it structurally cannot find.

• No subsurface detection. A crack buried inside the weld metal, internal porosity, lack of fusion under the root bead or trapped slag will never show up. There is no path in and no path out.
• No inspection of porous surfaces. Concrete, wood, paper, cloth, rough castings with connected porosity and some fiberglass soak penetrant into the surface itself, so the background turns to a mess and contrast disappears. PT needs a nonporous surface.
• Nothing on dirty, coated or smeared metal. Paint, rust, scale, grease, and metal smeared over a crack by power wire brushing or grit blasting all block or mask the flaw. Surface condition is the detection mechanism.

Who Uses It and Why They Like It

PT turns up across the trade: informal dye checks by welders repairing porosity in a stainless spool, acceptance testing by certified inspectors on code jobs, boiler shop tube sheets and nozzle welds, fire damage assessments, pipeline girth welds, and automotive or aerospace work in fluorescent mode under controlled lighting.

Its appeal is simple and hard to argue with:

• Cheap. A can of penetrant and developer costs far less than radiography or a UT machine.
• Portable. A kit fits in a toolbox, so you can run PT on a derrick, in a tank or on a bridge at night.
• Fast. With short dwell times on thin parts, a complete test fits inside a lunch break.
• Sensitive to fine surface breaks. It finds hairline cracks that visual inspection and most other methods miss.
• Works on every alloy. PT applies to essentially any nonporous material: steel, stainless steel, cast iron, aluminum, brass, bronze, titanium, rubber, plastics and glass. Its biggest edge over magnetic particle testing, which needs a ferromagnetic base metal.

Aluminum MIG welding covers surface preparation habits that pay off later, and prepping aluminum correctly before welding matters because aluminum demands a clean, oxide-free surface.

Six Recognized Techniques

ASME’s Boiler and Pressure Vessel Code recognizes six liquid penetrant examination techniques, and they are not arbitrary. The Code builds them from two penetrant types (fluorescent, or color contrast/dye) and three cleaning and removal methods (water washable, post-emulsifying, solvent removable). Two times three equals six, and every combination is code-recognized.

Those six techniques structure the rest of this guide.

How the Dye Works: Capillary Action, Surface Tension and Wetting

Liquid penetrant testing looks like the simplest test in the NDT shop: spray, wait, wipe, spray, look. But when a hairline crack lights up on a dark bead, you are watching a physics demonstration. Every rule in the procedure comes straight out of how liquids behave on metal.

Capillary action: the liquid climbs

Capillary action occurs when the adhesive forces between a liquid and a solid surface are greater than the cohesive forces holding the liquid molecules together. The liquid is pulled into a narrow gap and climbs it, even upward against gravity.

Think of dye worked into the fibers of a rope. Squeeze it and color emerges at the far end, because the fiber bundle acts as a wick. A crack is a wick too, only smaller and less forgiving.

The Federal Highway Administration puts it plainly: dye that has penetrated into a crack is not removed by the cleaning step and emerges due to capillary action, producing a visible indication. Dye that goes in during the penetrant dwell comes back out during the developer dwell, where a technician can see it.

Surface tension and wetting: why droplets bead and grease fights the dye

Surface tension is the way a liquid’s surface pulls on itself, and it is why a drop of water on a tabletop pulls itself into a tidy round bead instead of spreading into a thin puddle. Wetting ability is the opposite: how readily a liquid spreads across and clings to a solid. Water films out on clean glass and beads up on a waxed car.

Weld surfaces almost never look like clean glass. They carry mill scale, heat tint, spatter, rust, grease, and grinding residue. Oil or heavy oxide behaves like that wax: the dye beads up, sits on the surface, and never enters the flaw. That is why preparing aluminum and other metals properly before welding or inspection is not a courtesy. ASNT names surface tension, wetting ability, and capillary action as the three liquid behaviors that decide whether penetrant covers the surface and enters an opening, even when the part is inverted.

Liquid behavior	What it does	Why it matters in penetrant testing
Surface tension	The surface of the liquid pulls on itself, forming droplets	Penetrant that does not wet the metal beads up and is wiped off instead of entering a flaw
Wetting ability	How readily a liquid spreads and clings to a solid	Oil, grease, and oxide scale repel dye; clean metal accepts it
Capillary action	Liquid climbs a narrow gap when adhesion beats cohesion	Draws penetrant into a crack, then pulls it back out onto the developer film

The two-part timing: going in, then coming out

Penetrant testing is a two-act story, and both acts depend on capillary action.

Act one: the penetrant goes in during the penetrant dwell. It is a low-viscosity liquid formulated to wet metal and enter fine openings, so it wets the crack walls and is drawn inward. Act two: the developer pulls it out during the developer dwell, drawing the retained dye back through the same narrow path and spreading it across the surface as a visible indication against a white background.

The crack is a two-way valve and the developer is the pump that empties it. That also explains the limited inspection window after developer application: dye is still moving out of the flaw after the dwell ends, so an acceptable indication can grow into a rejectable one minutes later.

What the method is built to reveal

ASTM E1417/E1417M-21 defines the purpose of liquid penetrant testing as detecting discontinuities that are open or connected to the surface, and names the flaws it targets: lack of fusion, corrosion, cracks, laps, cold shuts, and porosity open to the surface.

Notice the common thread. Each of those is a channel with a door at the outside. The liquid needs a path in and a path out, and where the door exists, the method works. Turn that logic around and you get the limits: a subsurface crack, an internal slag, or a root defect buried under intact metal has no path at all.

Crack width, orientation, and the smear trap

Penetrant needs an opening, and the more directly connected the flaw is to the surface, the better the detection chance. A crack running out to the surface gives the dye a clean highway. One that is tight, sealed by corrosion products, or smeared shut by power wire brushing, grinding, or grit blasting may show nothing at all. A missed crack is far more expensive than a false call.

Orientation matters for the same reason. A crack running parallel to the surface may barely open at the surface, while one crossing it nearly at a right angle offers the best path. The surface condition of the joint is not housekeeping. It is the detection mechanism.

The Six-Step Procedure for Testing a Weld

Penetrant testing looks deceptively simple. You clean the metal, spray on some dye, wait, wipe it off, spray on the developer, and look for marks. No radiation, no magnet, no expensive electronics. Even so, the American Society of Nondestructive Testing still lays the method out as a fixed six-step sequence, because the technique only works when every step is respected and timed correctly:

1. Surface preparation
2. Penetrant application, followed by dwell
3. Excess penetrant removal
4. Developer application, followed by dwell
5. Inspection under appropriate lighting
6. Post-cleaning

Each step depends on the one before it. Skip preparation and the dye never reaches the crack. Let the penetrant dry and it seals itself out of the flaw. Rush the developer and the indication never bleeds back to the surface.

Step 1: Surface Preparation

Everything that can hide, mask, or trap a flaw comes off first: paint, primer, rust, mill scale, slag, spatter, oil, grease, and any coating of any kind. That includes metal smeared across the crack by power wire brushing, grinding, or grit blasting. Smearing is metallic and continuous, and no amount of dye will see through it.

Match the cleaning method to the base material and the coating. What works on steel can behave differently on aluminum, which is why prep deserves its own guide, such as how to prep aluminum for welding. Abrasive blasting, grinding, or wire brushing followed by a solvent wipe is typical. The surface has to be free of debris before the penetrant goes on.

Step 2: Penetrant Application

Apply a thin, even, continuous film. Spray, brush, or dip all work; what matters is coverage without thick buildup. Hold the can close enough to wet the surface but not so close that you lay down a heavy wet mat. Over-application is a real problem, because excess penetrant has to be removed later, and removing it too aggressively can pull good penetrant right back out of the flaw.

Keep the film wet for the entire dwell. Penetrant must never be allowed to dry on the part, and the National Board procedure explicitly warns against a dwell long enough for the penetrant to dry.

Step 3: Dwell (Penetration Time)

Dwell is simply how long the penetrant soaks into the flaw while it is still wet. Published ranges differ by source, process, and material, which is exactly why dwell belongs in the written procedure and the penetrant manufacturer’s data rather than in the technician’s memory.

Source	Penetrant dwell	Developer dwell
FHWA bridge practice	10 to 30 minutes	Additional 5 to 15 minutes after developer application, with roughly 30 to 60 minutes total application time
National Board	5 to 30 minutes	10 to 60 minutes

Step 4: Removing Excess Penetrant

Remove the surface film without disturbing what is inside the flaw. The usual sequence is a solvent remover applied with clean, lint-free rags, taking off the bulk film, followed by the emulsifier and water rinse for methods that use one. Method A, water washable, rinses clean with water. Method B, post-emulsifiable lipophilic, uses an emulsifier and water rinse. Method C, solvent removable, the most common field choice, uses a solvent remover to lift the excess. Method D, post-emulsifiable hydrophilic, uses an emulsifier and rinse as well.

Use clean rags and change them often. A dirty rag smears the penetrant back across the surface and can also drive it out of open flaws.

Step 5: Developer Application and Dwell

The developer does two jobs. It pulls retained penetrant back out of the flaw, and it provides the contrasting background that makes the indication visible. Spray a thin, uniform coat over the whole area and let it stand for the specified developer dwell. Too thick and you are looking at a heavy white film with poor contrast. Too thin and the bleed-out is weak and hard to read.

Step 6: Inspection, Then Post-Clean

Examine the part under the lighting the standard requires. ASME Section V requires dye penetrant solvent-removable evaluation with a minimum light intensity of 100 foot-candles on the part surface, verified with a light meter. For fluorescent penetrant, the room is darkened, and the FHWA cites ambient light in the darkened area kept at roughly 2 foot-candles or less so the fluorescence reads cleanly.

Then post-clean. Trapped penetrant has to be removed before any further welding, painting, or coating. NASA’s penetrant testing practice covers this in detail, along with the cleaning steps that precede the test.

Finally, record the result on the report your code requires. The National Board NB-7 inspection report form is one example, and ASTM E1417/E1417M-21 sets the standard test method requirements behind a properly executed procedure.

Choosing a Method: Visible vs Fluorescent, Type 1/2/3 Penetrants and Developer Forms

Walk into any supply counter and the shelves look the same: penetrant, developer, cleaner, wildly different prices. Small labels, dense jargon, and picking wrong can cost you an indication, a shift, or a rejected weld. You do not need a chemistry degree. Answer three questions: does the job demand maximum sensitivity, what are the lighting and working conditions like, and how much surface area is involved?

The ASNT liquid penetrant testing overview lays out the classification that most codes and procedures borrow from. Everything else here hangs off it.

Penetrant Types: 1, 2 and 3

Type	How You Read It	Sensitivity	Where It Fits
1, Fluorescent	Dye fluoresces yellowish-green under UV-A light	Most sensitive family, sold in graded levels, roughly Level 1/2 at the bottom through Level 4 at the top	Aerospace, nuclear and other critical work, where it is the default
2, Visible Dye	Red dye bleeds onto a white developer, giving a clear red indication under ordinary white light	Lower, Level 1 only	General fabrication, pressure vessels, structural steel; no darkened room, no UV-A lamp, no UV eye protection
3, Dual-Use	Reads as a visible dye under white light and fluoresces under UV-A	Reduced in both viewing modes	A compromise that costs sensitivity either way

One terminology note. ASTM E1417/E1417M-21 changed the term “blacklight” to UV-A, matching ASTM E1316, because fluorescent penetrant examination uses only the UV-A range. If a procedure or purchase order still says “blacklight,” it means the same thing, older phrasing.

How Sensitive Do You Actually Need to Be?

Sensitivity is graded against a reference block with known artificial flaws. For general fabrication, pressure vessel work and structural steel, Type 2 visible dye at Level 1 is usually adequate, and it is what you see on most jobsites. Aerospace turbine blades, critical castings and high-cycle fatigue components specify the highest fluorescent sensitivity level, since catching fine cracks early is the point. On aluminum, prep rules matter more than people expect, so read the Aluminum Cleaner for Welding: How to Prep Aluminum Right guide before buying penetrant.

Processing Methods: A, B, C and D

Processing method describes how excess penetrant is removed before developer goes on.

1. Method A, Water Washable: the penetrant already contains an emulsifier, so it rinses off with water. Fast and clean, but over-washing can pull retained penetrant out of a flaw and erase a real indication.
2. Method B, Post-Emulsifiable Lipophilic: an oil-based emulsifier is applied after the dwell, then water breaks it up. Strong on heavily contaminated work, but it adds a step and a bath.
3. Method C, Solvent Removable: excess penetrant is removed with a solvent remover and lint-free rags. The common field choice, because it needs no water source and no rinse tank.
4. Method D, Post-Emulsifiable Hydrophilic: a water-based emulsifier applied after dwell, with a pre-rinse before emulsifying. Effective and gentle, but the pre-rinse step is easy to skip.

Developer Forms A Through E

• Form A, Dry Powder: shaken over the part to form a tight white background. Ideal in dry or dusty environments, though wind can carry it off and the dust hazard is real.
• Form B, Water Soluble: a liquid developer sprayed, brushed or dipped. Good for large areas and production work.
• Form C, Water Suspendable: applies a uniform, high-contrast coating that stays on the surface long enough for easy inspection.
• Forms D and E, Nonaqueous: solvent-based spray developers for localized or field work, where you want a tight film on one specific weld without wetting down the whole assembly.

Cost and Practical Reality

Visible dye is cheap by design. The National Board notes that solvent-removable dye penetrant kits are typically three aerosol cans (cleaner, penetrant and developer) at roughly $5 to $15 per can, so a complete setup lands under $50. Fluorescent work adds the cost of a UV-A lamp and, more importantly, a darkened space. The FHWA notes that fluorescent DPT generally needs ambient light around 2 foot-candles or less to read the contrast, which is exactly why it is seldom used for routine bridge field work.

Temperature matters too. EN ISO 3452-1 practice commonly keeps the penetrant and part surface within roughly 10 to 50 °C. Cold steel slows penetration, and warm steel can shorten the working window on the aerosol.

Shops working aluminum regularly should also see the Aluminum MIG Welding: The Complete Practical Guide, which covers how surface condition changes everything downstream.

What PT Can and Cannot Find: Limitations, Blind Spots and False Calls

One limitation defines everything else about liquid penetrant testing: PT finds only discontinuities that are open to the surface. ASNT lists surface-only detection among the method’s limitations, noting that it is not effective for subsurface defects. If the flaw does not break through the metal, the penetrant has no path in and no path out.

That rules out a long list of things people hope PT will find:

• Subsurface cracks, including fatigue cracks growing down into the base metal
• Internal or buried porosity
• Lack of fusion sealed beneath the root or between layers
• Slag, tungsten inclusions and other volumetric defects
• Laminations and delaminations trapped under the surface

The Federal Highway Administration puts it plainly: dye penetrant testing cannot detect subsurface cracks or volumetric weld defects such as porosity, lack of fusion, or slag inclusions.

Surfaces That Cannot Be Inspected at All

PT does not work on porous surfaces. Concrete, wood, paper, cloth and some fiberglass with exposed fibers drink the penetrant in, where it can never be washed out, so nothing is left to read against the white developer. Rough, corroded or scaled surfaces do the same in milder form: too much penetrant sits in the texture instead of the flaw, so contrast collapses.

Smearing, Coatings and the Price of a Clean Surface

One of the most common ways a real crack gets missed is a dirty procedure, not a faulty one. The National Board is explicit: metal smearing from power wire brushing, shot blasting or grit blasting must be removed before liquid penetrant examination. A power wire brush can smear a film of soft, ductile metal across a hairline crack and seal it shut. The same board notes that surface finish and roughness affect sensitivity enough that grinding may be necessary first.

On aluminum the problem starts earlier, with oxide, grease and residue from forming and handling. The prep discipline in our guide to how to prep aluminum for welding is the same discipline PT depends on. Coatings cannot survive the test either: each one must come off before penetrant reaches the metal, and removing it removes the corrosion protection it provided. FHWA lists that tradeoff among the method’s limitations, along with the point that extensive surface preparation can produce hazardous material such as lead paint.

The Access Problem in Real Assemblies

Then there is geometry. On a real spool, structural assembly or pipe rack, the weld of interest sits next to a stiffener, a nozzle, a support ring or an adjacent run of pipe. You cannot clean what you cannot reach, reach what you cannot see, or evaluate an indication you cannot photograph. FHWA notes that areas with interfering members are difficult to assess, so a good result on the accessible 80 percent of a joint says little about the buried 20 percent.

False Calls: Indications That Are Not Cracks

False positives are the other side of the coin. Weld surface relief caused by undercut, surface porosity or a rough bead can produce non-relevant indications that must be separated from genuine crack indications during evaluation. The bead of a poorly shaped fillet weld has natural pockets of undercut at the toes and lumps of reinforcement at the crown, and both can hold a neat line of penetrant that reads convincingly like a crack under UV-A. Grinding marks, chatter from a wandering wheel and handmade bead roughness do the same.

In aerospace work, where aerospace TIG welding demands tight profile control, telling a real crack from surface relief is a trained eye working to a written procedure, not luck.

Trapped Penetrant: The Problem That Comes Back Later

The National Board raises a failure mode that catches people by surprise. If an assembly has stitch welds, or material not sealed by a weld, penetrant travels behind the welds and between layers of unfused material, where it is nearly impossible to remove. The Board warns that trapped penetrant will cause defects in subsequent welds and can bleed out over time, contaminating paint and process fluids. That is a defect in tomorrow’s joint, created by today’s test.

The Depth MT Reaches and PT Never Will

Compare with magnetic particle testing. OSHA’s Technical Manual notes that MT has a crude approximate sensitivity depth of no more than 1.5 to 3 mm (1/16 to 1/8 in), a capability PT cannot offer at all. On nonferrous alloys such as aluminum or stainless, MT is not an option, so the entire surface inspection burden falls on PT. That is why prep discipline is non-negotiable on stainless and aluminum work, and why our aluminum MIG welding guide stresses the same cleanliness habits for the same reason.

Weld Defects PT Is Designed to Catch — and the Ones It Misses

Penetrant testing is not a general-purpose flaw detector. It is a targeted surface-break method, and ASTM E1417/E1417M-21 keeps the list short: lack of fusion, corrosion, cracks, laps, cold shuts, and porosity that is open or connected to the surface. Every item shares one trait. There must be a path from the flaw to the outside of the metal for a low-viscosity liquid to travel. That single requirement explains both the strength and the limit of the method.

The Defect List, in Welding Terms

Here is what an inspector is actually hunting for:

• Surface cracks in weld metal and HAZ. One category, several mechanisms: fatigue cracks from cyclic loading, distortion cracks from restrained shrinkage, hydrogen-type cracking in restrained high-hardness joints within days, quench cracks in hardenable steels quenched but not tempered. Aerospace TIG welding of nickel and titanium alloys is a familiar home for several at once.
• Root, toe and end cracks. A root crack runs along the fusion line at the weld root, a toe crack starts where the bead meets the parent metal, an end crack stops at a weld termination or edge. All are open to the surface, so all are fair game for dye.
• Surface-breaking lack of fusion. When a bead fails to fuse and leaves a crevice open to the outside, penetrant walks in and the developer pulls it back out as a clear linear indication.
• Cold shuts, laps and surface laminations. Unfused laps left by a welding discontinuity, plus rolled-in scale pressed into the surface. The National Board requires PT for this kind of work: castings for surface defects, plates for laminations in corner joints, head spin hole plug welds, weld metal build-up, and defect removal before welding repair.
• Surface porosity. Open gas porosity and wormhole porosity are detectable. Porosity sealed beneath the surface is not, and neither is internal porosity. That is a line in the sand, not a limit of the method.
• Corrosion and corrosion fatigue cracks. Dimensional loss and cracking from wet H2S, sour service, or general atmospheric attack that has eaten into a weld.
• Arc strikes, burns and post-weld ground cracks. A stray arc strike on base metal is a quench-crack-like defect PT finds easily, and one of the most common findings on repair work. Grinding out a crack can leave a fresh one at the gouge edge, so ground surfaces get inspected too.

Location Is the Risk: Where Cracks Actually Turn Up

Knowing the defect list is half the job. The rest is knowing where cracking starts, because that is where you point the penetrant. Fatigue cracks rarely begin in sound plate. They start at geometric stress risers: weld toes, attachment welds, gusset plates, nozzle openings and interrupted welds where the profile changes abruptly. That is why toe-crack PT on structural connections is routine.

For corrosion-driven damage, the survey data in the OSHA Technical Manual is sobering. In amine service, all observed cracks were in or near welds, mostly stressed welds without post-weld heat treatment, and across the survey cracking incidence was about 40% over 294 plants. In pulp digester service roughly 65% of properly inspected ASME Code vessels showed cracking, all weld-associated, and wet fluorescent magnetic particle testing with proper surface preparation proved most effective. OSHA reports LPG vessel cracking predominantly in the weld HAZ, cracks in or near welds in as-welded ammonia vessels, and corrosion fatigue as the predominant mechanism in deaerator cracking.

Those findings shape inspection strategy directly. ASNT frames the division of labor: PT covers accessible surface-breaking flaws on any alloy, magnetic particle testing is preferred for near-surface cracking in ferromagnetic steel, and RT or UT is required for internal condition.

Nonferrous equipment has its own signature. Caustic service and high-humidity ammonia environments produce stress corrosion cracking that frequently initiates at or adjacent to welds in aluminum, stainless and high-strength steels alike. On aluminum, surface condition is half the battle: smearing a crack shut during cleaning negates the test, which is why prep technique matters (see our guide to cleaning aluminum for welding).

Defect to Method: A Working Reference

Defect or condition	What PT will do	What to add or use instead
Surface crack in weld metal or HAZ	Reliable if open to the surface	MT for near-surface cracks in steel; RT or UT inside
Surface-breaking lack of fusion, cold shut, lap	Detectable	UT or RT for depth and internal extent
Open surface porosity	Detectable	RT where subsurface porosity matters
Subsurface porosity, buried lack of fusion, slag	Not detectable	RT or UT
Tight or smearing-over crack	May not show; no liquid path in	Remove smearing, re-test; MT for steel
Coated, dirty or rough surface	Masked, not testable	Strip and clean first; MT needs clean metal

Liquid Penetrant vs Magnetic Particle vs Ultrasonic vs Radiographic

Most field errors in nondestructive testing are not caused by a bad technician. They are caused by a good technician running the wrong method. A dye penetrant test on a carbon steel weld that has a buried lack of fusion will come back clean, and it will come back clean with real confidence, because the method did exactly what it was built to do. The problem is what was asked of it.

OSHA organizes inspection methods into two families, and that vocabulary is the foundation for every selection decision: surface examination methods (visual testing, liquid penetrant testing and magnetic particle testing) and volumetric methods (radiographic testing and ultrasonic testing), per the OSHA Technical Manual on NDT methods. A surface method sees what is broken open to the air. A volumetric method reaches into the metal.

Detection depth and what each method can actually see

Liquid penetrant testing is strictly a surface method. It finds discontinuities that are open to the surface, full stop. It cannot see a subsurface crack, internal porosity, slag, or a lack of fusion buried under the root pass. Its one advantage over the other surface methods is that it isindependent of the type and composition of the metal alloy, so it works on austenitic stainless steels and nonferrous alloys where magnetic particle testing is not applicable, according to OSHA.

Magnetic particle testing shares the surface category but reaches further down. It is essentially limited to carbon and low-alloy steels by the ferromagnetic material requirement, and it is most sensitive to discontinuities oriented transverse to the magnetic flux lines, OSHA notes. Practical guidance puts its useful sensitivity at roughly 1.5 to 3 mm below the surface, so MT catches near-surface defects that PT would miss entirely. It also needs no dwell time, no cleaner, no developer, and no post-clean, which makes it faster on a continuous production line.

Ultrasonic testing and radiographic testing work on a different premise entirely. RT passes a radiation beam through the part and captures a permanent film. UT sends sound through the metal and reads echoes from interfaces. Both can locate internal flaws, and UT can detect and measure general material loss such as corrosion and erosion, OSHA points out. OSHA also notes RT is effective at showing defect dimensions on a plane normal to the beam, which is why film radiography remains the reference standard for volumetric sizing in many codes.

PT and MT are partners, not competitors

The most useful thing to understand about the two surface methods is that they are usually deployed together on the same assembly. ASNT describes the pattern with an aerospace manufacturer that inspects landing gear: magnetic particle testing on the ferromagnetic steel parts, and liquid penetrant testing on the nonferromagnetic aluminum alloy parts, for surface defects such as cracks and corrosion pits.

That pairing solves a real problem. A landing gear strut is steel. A fitting bolted to it might be aluminum or a stainless fastener. You cannot run MT over the aluminum. You cannot run PT on the steel and expect to catch a subsurface initiation. Run both and you cover both materials and both depths.

Diverting attention to the alloy side of that equation, prep work matters more than most people expect. how to prep aluminum correctly before and after inspection is worth getting right, since oxidation and grease can mask indications on exactly the nonferrous surfaces where PT is the only option.

Comparison table

Factor	PT	MT	UT	RT
Detection category	Surface only	Surface and near-surface	Volumetric	Volumetric
Typical detection depth	Open to surface	~1.5 to 3 mm	Throughout the wall	Throughout the wall
Material applicability	Any alloy	Ferromagnetic only	Most metals	Most metals
Sensitivity	High at the surface	High at and near surface	High, depth dependent	Moderate to high
Relative cost	Low	Low to moderate	Moderate to high	High, plus isolation
Speed in the field	Slower, dwell time	Fast, no dwell	Moderate	Slow
Portability	Excellent, aerosol kits	Good, portable yoke	Limited	Poor, fixed exposure
Access required	Direct to surface, all sides	Direct contact at least one position	Scan surface access	Both sides of the part
Documentation	Written report, photos	Written report, photos	Indications logged, no permanent record in basic form	Permanent radiographic film
Typical use	Weld toe and root surface cracks, nonferrous	Steel welds, near-surface cracking	Internal flaws, corrosion/erosion wall thickness	Internal flaws, weld volumetric sizing

Cost, access and safety

PT wins on logistics. A fluorescent setup needs a darkened room and UV-A lamps, and a visible dye kit needs cleaner, penetrant and developer. Radiography needs an isolation area, controlled access and a qualified radiographer, which makes it the most expensive and least portable option by a wide margin. Ultrasonic testing requires trained operators and a good couplant surface, and it delivers the best flaw location and depth data of any of the four, plus wall-thickness measurement for corrosion and erosion, though OSHA notes UT in its basic form does not produce a permanent record of the examination while RT produces a permanent radiographic record. On a pipeline or pressure vessel that must be re-inspected years later, that film has real value.

Access drives many field decisions. PT needs every suspect surface directly reachable and clean, MT needs contact on at least one position and sometimes two to bias the field, UT needs a scanning surface plus a reliable path into the material, and RT needs radiation to pass through the part with minimal interference from nearby members.

Safety profiles differ too. PT and MT put solvents, penetrant, developer and aerosols in the technician’s hands, soOSHA Hazard Communication requirements for labels, SDS and training apply. MT adds high current and a strong magnetic field. UT is quiet but adds confined-space and ladder work. RT brings the strictest controls of all: radiation safety programs, dosimetry, exclusion zones and a licensed operator.

The honest answer on reliability

Here is the point that should temper every “clean” result, including a clean PT result. OSHA frames safety assessment around the largest flaw that may be missed, not the smallest one that can be found. A method’s headline sensitivity is its best case, not its guarantee.

OSHA attributes high NDE reliability to validated procedures, equipment and test personnel, the use of diverse methods, and redundancy through repetitive and independent tests. In plain terms, no single technique is a guarantee. Reliable inspection is a system, not a product.

The practical takeaway is not that one method is best. It is that on an assembly containing steel and aluminum or stainless, PT and MT should be planned as a pair, that UT and RT belong in the plan whenever the failure mode being hunted for is internal, and that the written procedure for the specific job decides which combination is required. If you are working through the alloy-specific prep and cleaning steps first, our guides on aluminum MIG welding and aluminum flux core welding cover the surface condition questions that determine whether an indication can even be seen.

Codes, Standards, Acceptance Criteria and Who May Perform the Test

With liquid penetrant testing being used for everything from structural fabrication to boiler repairs, compliance isn’t optional. For code governed work, the method must follow approved standards, a written procedure, and qualified personnel. Without all three, a passing indication report holds no weight under most construction or repair codes.

The ASTM Backbone

The majority of modern penetrant requirements stem from ASTM E1417/E1417M-21. This 12 page practice, developed by ASTM Committee E07 and its subcommittee E07.03, was last updated on November 25, 2021, replacing MIL-STD-6866 at a price of $96 for a single-user PDF copy. Most importantly, ASTM is very blunt about its intent: E1417/E1417M-21 is not a detailed how-to procedure to be used by the inspector and must be supplemented by a detailed procedure conforming to the practice. Its companion, ASTM E165/E165M, is used to supply information for developing those detailed requirements.

Within any approved procedure, variables are split into two categories. Essential variables are changes that would affect the sensitivity or reliability of the examination (such as changing penetrant type or method), and nonessential variables are changes that may not affect results, but still must be controlled or noted per the procedure. Also worth noting under Scope 1.3 of E1417, some cleaners, penetrants, and developers can harm nonmetallic materials such as plastics, so compatibility tests should be conducted prior to examination. These precautions are especially relevant for repairs seen within more specialized fabrications, such as alloy wheel welding, where plastics and delicate coatings may be present near the repair area.

Pressure, Power and Structural Codes

For boilers and pressure vessels, ASME BPVC Section V Article 6 is the examination method of record, with construction requirements often falling under ASME Section VIII. Section V recognizes the six processing techniques for penetrant testing and, for dye penetrant solvent removable evaluations, requires a minimum of 100 foot-candles of white light, verified with a light meter.

The National Board Inspection Code (NBIC) takes this a step further. For ASME Code construction, or for NBIC repairs and alterations, the National Board requires a written procedure complying with ASME BPVC Section V, Article 6 that addresses all essential and nonessential variables. Examinations are recorded on the NB-7 Inspection Report Form. The National Board specifically recommends PT for firetube boiler tube sheets, external weld joint inspection, fire-damage evaluation, and LPG service. One of their well known examples is a head spin hole plug weld that looked acceptable to the naked eye, but showed many defects once liquid penetrant examined.

For structural work, AWS D1.1/D1.1M (Structural Welding Code – Steel, approved by ANSI) and AWS D1.5 (Bridge Welding Code) are used most frequently. AWS has authored over 350 welding standards in total. A key difference here is acceptance: NDE acceptance criteria for these codes generally follow the visual inspection criteria, and AWS D1.1’s approach is to size the actual discontinuity for acceptance, not the larger, often bloated PT indication. For international work, ISO 3452-1 serves as the predominant companion standard.

Who May Perform the Test

For code work, “qualified personnel” is not a matter of experience alone, it’s a matter of certification to a recognized program. The most common is ASNT SNT-TC-1A, with the industry note that ASNT is currently modernizing it. For in-service crack surveys, OSHA’s Technical Manual points to ASNT SNT-TC-1A minimum Level I for inspection, with interpretation by a minimum Level II as a common industry requirement. Other programs utilized are NAS-410 (prevalent within aerospace), and for those already holding a CWI or SCWI credential, the AWS Penetrant Testing (PT) endorsement is available, alongside its Magnetic Particle and ISO Standard endorsements.

There is one critical distinction made by the National Board: a welder grinding out a crack for repair may use PT informally to verify removal, however, Code work requires a written procedure followed by qualified personnel. Just the same, for more advanced, regulated work such as aerospace welding, the expectations for documentation and qualification are held to their strictest degree.

At the end of the day, the governing factor will always be the contract. Between ASTM, ASME, NBIC, AWS, ISO, and your client’s specification, the most restrictive requirement wins. If there is any conflict between a manufacturer’s sales sheet and your approved written procedure, the approved written procedure is the only one that matters on code work.

Safety, Chemicals and Real-World Working Conditions

Dye penetrant testing may look simple, but the chemicals demand respect. On any job site, you’re working with a cleaner, penetrant, developer, and often an aerosol propellant. Get the handling wrong and you risk flammable vapours, toxic exposure, or hazardous debris. All of it needs to be accounted for before a single can is opened.

The Chemical Inventory

Each PT consumable carries its own hazard:

• Cleaner and solvent remover: Fast-evaporating solvents, often petroleum distillate or chlorinated based. Many flash below or near room temperature.
• Penetrant: A low-viscosity carrier containing either a visible red dye or a fluorescent dye designed to seep into fine openings.
• Developer: Available as a fine powder, a water-soluble liquid, or a nonaqueous solvent-based spray. Its job is to pull retained penetrant back to the surface for viewing.
• Emulsifier and propellant: Used with post-emulsifiable methods or in aerosol kits. These can be flammable and will add vapour load to the work area.

The National Board warns that without proper ventilation, penetrant testing fumes can be hazardous and flammable. These vapours are heavier than air, meaning they settle in pits, flanges, and tank bottoms where they’re easy to miss.

Labels, SDS and Responsibility

Before use, pull the safety data sheet for every product on the truck. Under OSHA’s Hazard Communication Standard, aligned with the Globally Harmonized System, employers must provide labels, safety data sheets, and training. A compliant 16-section SDS will cover composition, hazards, handling, storage, first aid, fire response, spills, and disposal.

Responsibility is also spelled out by the standard itself. ASTM E1417/E1417M-21 states that it does not purport to address all safety concerns, and that the user is responsible for establishing appropriate safety, health, environmental practices and regulatory compliance. If a container has lost its label, treat its contents as unknown and do not use it until identified or properly disposed of.

PPE, Respiratory Protection and UV-A

At a minimum, use chemical-resistant gloves, eye protection, and appropriate clothing. Nitrile or butyl gloves may work for many products, but always check the glove compatibility chart in the SDS, as some chlorinated solvents require more resistant materials.

When ventilation is inadequate, respiratory protection falls under 29 CFR 1910.134. This requires a written respiratory protection program, respirator selection with NIOSH certification, medical evaluations, fit testing, training, and program evaluation, all provided at no cost to the employee. A simple dust mask is not a substitute for this program.

For fluorescent penetrant inspections, UV-A radiation at 365 nm poses its own hazard. The light is invisible, so there is no warning arc. Wear the rated UV-A eye protection specified by the lamp manufacturer, follow your darkroom procedure, and consider skin exposure for long inspections or if taking photosensitising medications.

Lead Paint, Conditions and Containment

On older bridges, tanks, ship hulls, and water towers, cleaning to bare metal can mean grinding through lead-containing paint. The FHWA warns that extensive surface preparation can produce hazardous materials in these cases. If the coating history is unknown, treat it as lead until tested. This will often require containment, specific respiratory protection, and a regulated disposal plan.

Environmental factors can also invalidate an inspection. Per ASNT, high temperatures or wind can adversely affect penetrant application and effectiveness. Keep parts and products within roughly 10 to 50 °C as referenced by ISO 3452-1, shield from wind and dust, prevent condensation, and never open solvent-based materials inside a permit-required confined space without full ignition control and ventilation.

Disposal and Post-Clean

Used rags, contaminated absorbents, emulsified rinses, and aerosol cans are often hazardous waste. Collect them in closed, labelled containers and dispose of them through an approved waste stream. Never pour them down a floor drain or burn solvent residue.

Equally important is post-cleaning. Trapped penetrant can migrate, contaminate paint, or introduce porosity when welding over an uncleaned surface. Cleaning after inspection is required by procedure, not optional. For a solid foundation, review welding supply and safety basics and consider surface sensitivity for reactive metals in how to prep aluminum for welding. The safest inspection is always a planned one, with time built in for every one of these precautions.

Field Tips, Common Mistakes and a Working Checklist

Most liquid penetrant testing does not fail because the chemistry is wrong. It fails in the ten minutes before the can comes out and the ten minutes after the developer dries. That is where the money goes: a rejected inspection report, a spool that has to be re-welded, or a crack that was never really there in the first place. The Federal Highway Administration notes on dye penetrant testing make the point that inadequate surface cleaning produces non-relevant indications, because dye emerges from debris or from under coatings that were never fully removed.

The mistakes that cost real money, and how to fix them

• Rushing or skipping surface preparation. The fix is to budget time for cleaning before you budget time for anything else. Paint, rust, scale, grease and shop dust all mask flaws and all produce misleading indications.
• Leaving grinding or wire-brush smearing over the crack. Power wire brushing and grit blasting smear soft metal across the crack face and seal it. Fix: mechanically abrade or clean the area, then re-clean with solvent and a lint-free rag before applying penetrant.
• Letting the penetrant dry. Once the film turns chalky, it will not flow into a tight flaw. Fix: keep the surface visibly wet, work in small sections, and re-wet before the dwell ends.
• Over-washing Method A penetrant. A heavy rinse can pull retained penetrant back out of the flaw before the developer ever reaches it. Fix: use the specified rinse time and pressure, and prefer Method C solvent-removable consumables when you are new to the method.
• Cleaning excess penetrant poorly. The National Board describes the solvent-removable step as wiping with clean, dry, lint-free rags until no penetrant is visible, then applying cleaner/remover on a second clean rag and rubbing vigorously until the rag shows no penetrant. A halo of residual penetrant around the flaw mimics a real indication.
• Applying developer too thick or too thin. Too little gives no contrast, too much buries the indication under a heavy white film. The National Board specifies a thin, light coating of developer followed by a developer dwell typically of 10 to 60 minutes. Fix: apply in two light passes rather than one wet coat.
• Examining too early or too late. Too early, the developer has not pulled the penetrant out. Too late, the indication has bled out and shrunk. Fix: work the written procedure’s inspection window and record the time.
• Checking under inadequate light. The ASME Section V minimum of 100 foot-candles for dye penetrant solvent-removable evaluation means a light meter check belongs in a defensible procedure, not in the “if you remember” column.
• Failing to post-clean before welding or painting. The National Board cautions that post-cleaning is required after evaluation, especially if welding is to be performed afterward. Trapped penetrant left on the joint contaminates the weld and the paint.
• Using one can until it dies. Fix: check shelf life and storage temperature, keep containers tightly closed, and buy a fresh kit rather than a new inspection.
• Ignoring temperature limits. Cold steel and warm penetrant give poor and uneven results. Fix: keep the part and the consumables inside the range your procedure specifies.
• Testing in rain, wind or blowing dust. Wind scatters developer aerosol, dust traps penetrant, and moisture ruins the film. Fix: wait, shelter the area, or come back.
• Calling every blob a crack. Fix: clean the area, re-apply, and compare the indication against the acceptance criteria before you condemn the part.

Troubleshooting: what you are actually looking at

Symptom	Likely cause	Corrective action
No indications at all	Penetrant dried before it could enter the flaw, dwell too short, or the flaw is not open to the surface	Restore the published dwell, keep the film wet, and if the crack may be subsurface, consider MT or UT instead
Broad, diffuse, fuzzy indications	Over-applied developer, porous or very rough surface, dirty substrate	Apply a thinner developer film, clean and smooth the surface, and remember that porous materials such as concrete, wood, paper and cloth should not be examined by PT
Indications that fade or bleed out quickly	Poor cleaning, over-washing during removal, or excessive dwell	Clean thoroughly, control the rinse, and evaluate within the specified window
Linear indication running along the bead	Undercut or rough bead profile, though linear indications are generally read as cracks	Grind the profile smooth and re-test; then confirm or clear by a second, careful inspection
Indication under a stitch weld or between layers	Penetrant trapped and migrating	Grind out the stitch, clean, and re-test before any weld is made
Red dye visible but little developer contrast	Developer contaminated or applied too heavily	Use fresh developer, new applicator, and a thin coat

Pre-test and post-test checklist

Before the test

• Written procedure in hand, with the correct consumables, dwell times and acceptance criteria.
• Technician qualified for the method and the applicable code.
• Surface free of paint, rust, scale, grease, smearing and embedded debris.
• Surface nonporous and temperature within the specified range.
• Cleaning materials and lint-free rags clean and dedicated to this job.
• Lighting verified at 100 foot-candles, or darkened area set up and measured for fluorescent work.

After the test

• Post-cleaned so no penetrant remains before welding, painting or coating.
• Indications photographed, measured, compared against acceptance criteria and dispositioned.
• Report and record completed in the form the job requires, such as the National Board inspection report.
• Coating or corrosion protection restored where required.

A field repair, start to finish

A technician is called to a stainless steel pipe spool in a food plant after a hairline crack appears beside a weld. Stainless is nonferromagnetic, so magnetic particle testing is off the table, and the repair decision rests on a visible dye penetrant. The technician grinds the crack out, cleans the area, sprays on the dye, waits, wipes back, applies developer, and finds a clear linear indication still sitting in the toe of the weld. The question is no longer “is there a crack” but “how deep does it go.”

Here is the trap. Penetrant only proves what is open to the surface. Even a bright, unmistakable indication says nothing about whether the crack has run down into the root or stopped in the fillet. Grinding deeper until the indication vanishes is not a technical decision, it is a guess, and a guess that gets welded over. The defensible path is to grind progressively, re-test after each pass, and then bring in a method that can look at the root. That is where ultrasonic testing earns its place, and where an SMAW root pass or a filler metal selected for the joint position comes into play, covered in the all-position welding rod guide.

For work on aluminium spools where cleanliness before welding is the recurring theme, the guide to preparing aluminium for welding covers the same logic from the welding side: what you leave on the surface decides what you get in the joint.

Frequently Asked Questions

What is liquid penetrant testing for welding, in simple terms?

It is a nondestructive inspection method in which a low-viscosity liquid is applied to a cleaned weld surface, allowed to soak into any crack or open-to-surface flaw, then drawn back out with a developer so the flaw becomes visible. Because it relies on capillary action, it works on any nonporous material — steel, stainless, aluminum, titanium — and is cheap, portable, and fast, which is why it remains a workhorse method for shop and field weld inspection.

Can liquid penetrant testing find defects inside a weld?

No. PT only detects discontinuities that are open or connected to the surface. It will not find subsurface cracks, buried lack of fusion, internal porosity, or slag inclusions — for those you need ultrasonic or radiographic testing. Even the National Board and OSHA’s Technical Manual classify PT as a surface method alongside visual and magnetic particle, with RT and UT as the volumetric methods.

Is dye penetrant testing better than magnetic particle testing?

Neither is better in general — they are complementary. Magnetic particle testing can detect near-surface flaws (roughly 1.5 to 3 mm deep) but only in ferromagnetic materials, while PT reaches the surface of any material but cannot look beneath it. On a mixed assembly such as aircraft landing gear, MT is used on the steel and PT on the aluminum, exactly as ASNT describes.

How long should penetrant dwell and developer dwell times be?

Typical published ranges are roughly 5 to 30 minutes for penetrant dwell and 10 to 60 minutes for developer dwell, with the Federal Highway Administration citing 10 to 30 minutes and 5 to 15 minutes for field bridge work. Always follow your written procedure and the penetrant manufacturer’s data sheet, because the required dwell depends on the penetrant class, method and sensitivity level.

Why do I get false indications on my welds?

Most false calls come from surface condition, not from the penetrant: undercut, surface porosity, rough bead profile, grinding marks, wire-brush smearing, or debris and old coating left on the surface that traps dye and bleeds it out through the developer. Fix the surface preparation, keep developer coats thin and even, and evaluate indications in the same lighting and within the time window the procedure allows.

Can a welder do their own penetrant testing?

For informal checks — for instance, confirming that a crack has been fully ground out before a repair weld — yes. But for ASME Code construction, National Board work, or contract-governed inspection, a written procedure complying with ASME Section V, Article 6 must be followed by qualified NDE personnel, and the results recorded on the required report form.

Do I need to remove paint and coating before liquid penetrant testing?

Yes. The surface must be clean, smooth, and free of paint, rust, scale, oil, grease, and smearing, because coatings both hide the surface and can trap dye that later bleeds out as a false indication. The Federal Highway Administration notes this compromises corrosion protection and can generate hazardous material such as lead paint dust when old coatings are disturbed.

What lighting is required to evaluate penetrant indications?

For dye penetrant evaluated by the solvent-removable method, ASME Section V requires a minimum of 100 foot candles on the part surface, verified with a light meter. Fluorescent penetrant is viewed under UV-A in a darkened area — the FHWA cites ambient conditions of about 2 foot-candles or less — and ASTM has standardized the term ‘UV-A’ in place of the older ‘blacklight’ wording.

What standards govern liquid penetrant testing?

In the United States, ASTM E1417/E1417M-21 sets the minimum requirements for liquid penetrant examination of nonporous components (and replaced MIL-STD-6866), while ASTM E165/E165M supports writing detailed procedures. ASME BPVC Section V, Article 6 governs the examination method for Code work, the National Board Inspection Code applies to boiler and pressure vessel work, and AWS codes such as D1.1 and D1.5 govern structural and bridge welding, with ISO 3452-1 as the international counterpart.

Is liquid penetrant testing safe to do on a job site?

It is safe when the basics are respected: read the SDS, work with ventilation, wear chemical-resistant gloves and eye protection, and follow OSHA’s Hazard Communication and respiratory protection requirements where vapors are a concern. The National Board warns specifically that fumes can be hazardous and flammable without proper ventilation, and that chemical handling and disposal must be planned for.