Liquid penetrant inspection finds defects that break the surface. A coloured or fluorescent liquid is applied to a cleaned part and left to dwell, so it can work its way into any opening narrow enough to draw it in. After the surface is wiped clean, a developer draws the trapped liquid back out, and the defect shows as a visible mark.
The method works by capillary action rather than by anything electrical or ultrasonic, which is why it needs no power supply and no special geometry. It is a purely surface technique, and that limitation defines both where it is useful and where it must be replaced by something else.
It finds cracks that reach the surface, incomplete fusion lines open to the surface, porosity that has broken through, laps, seams and grinding cracks. It is very good at picking up the fine, tight cracks that a visual inspection under workshop light will miss entirely.
It cannot find anything below the surface. A void inside a weld, a lack of penetration at the root, or a crack that has not yet reached the outside stays invisible to it. It also cannot report depth or size in any useful sense; a strong indication and a weak one look similar, and neither tells you how far the defect runs into the metal.
The surface has to be clean and dry before the penetrant is applied. Oil, paint, scale, machining residue and moisture all block the opening the liquid is meant to enter, and a blocked defect produces no indication at all. That is the most common way this test misleads someone: a false pass caused by poor preparation rather than by a sound part.
Removal matters just as much. If the excess is not cleaned properly, the developer lifts residual liquid out of surface texture and shows marks that mean nothing. Rough castings, welds with heavy ripple and shot-blasted surfaces are the hardest to read for exactly this reason.
An indication has to be interpreted. A mark in a line along a weld toe is likely a crack. A cluster of small dots usually points to porosity that has broken through. A broad, faint stain may be nothing more than a surface irregularity holding extra liquid. The reading is a judgement, and it improves with experience on the same type of part.
Where the answer matters, the finding should be recorded with its location and then checked by another method. Penetrant is at its best as a screen that sorts parts into three groups: clearly sound, clearly suspect, and needing a closer look with something that can see below the surface.
Hand-welded tubes, headers, tanks and brackets are natural candidates, because their critical defects are surface-breaking cracks at weld toes and heat-affected zones. A part that has been dressed back with a grinder is worth checking afterwards, since grinding can smear metal over a crack and hide it until the part is loaded.
It also suits parts that have seen service. A component returned for inspection, or a welded item that has been heat cycled, can be checked for cracks propagating from a stress point without cutting it up. That makes it a practical field technique rather than only a workshop one.
No single method covers everything, which is why inspection is normally layered. Visual and penetrant work on the surface, pressure testing proves containment, and dimensional work proves the geometry. Each answers a question the others cannot, and a result from one should never be presented as though it covered the rest.
On welded assemblies, the method is chosen to match the joint and the failure it could produce. For the titanium welding parts side of the work, where a defect at the weld toe is the realistic risk, a surface technique covers most of the concern. On heavier assemblies, the titanium fab components range is judged on a combination of checks rather than on any one of them.
