Pad Cratering Inspection: Dye and Pry vs Cross Section

Pad cratering is a fracture that begins in the resin-rich region of the laminate directly beneath a surface mount pad, and it grows outward under thermal or mechanical load. The solder joint usually survives, which is what makes the defect so expensive: the assembly passes functional test at build and fails much later in service.

Because the crack sits in the board rather than in the joint, it is invisible to visual inspection and to most optical systems, and X-ray does not see it either. Detecting pad cratering means opening the interface or imaging through the pad, so the inspection method decides what can be found and what is never looked for.

What Pad Cratering Looks Like in the Laminate

The damage starts as a separation between the copper pad and the resin beneath it, running down and outward through the weave. In the early stage it is a hairline a few tens of micrometres long that follows the glass bundles rather than travelling straight down, because the resin-rich area between bundles is the weakest path available. The route it takes is the same one described in our notes on pad cratering and laminate damage.

Pad adhesion is a property of that resin-rich region, and it is the bond that fails first, while the copper and the joint above it often remain intact. A late-stage crater is easy to see, with the pad sitting on a cone-shaped void and lifting under a probe, but early damage leaves no external sign at all.

Dye and Pry: Opening the Interface

Dye and pry draws a penetrating dye into any open path at the pad interface and then mechanically removes the component so the underside of the pad can be examined. The board is usually cycled through vacuum and pressure, or soaked, so the dye reaches the full depth of a crack that may be no wider than a micrometre.

The component is then pried off and both surfaces are inspected at 30x to 50x. Dye on the pad is evidence of a laminate crack, while dye on the component side with a clean pad points to something else, usually a joint defect or flux trapped at the interface during assembly.

Reading the Dye Stain Pattern

The stained area is normally reported as a fraction of the pad area, and the orientation of the stain shows which way the crack ran. A stain confined to one edge suggests bending or connector insertion force, while a stain covering most of the pad suggests a reflow-driven mechanism, because separating a whole pad needs the stored strain energy of a large joint.

Dye and pry is destructive and slow, so it belongs to qualification, supplier comparison and root cause work rather than to lot acceptance. Its value is that it shows the interface directly, without inference from an indirect signal, and it can be aimed at one suspect location on a production board.

Microsection: Depth and Angle of the Crack

A microsection through the pad shows the crack in cross section and allows the depth to be measured against the pad diameter. A crack that has travelled more than about 25 percent of the way across the pad is generally treated as a reject, while a short crack at the pad edge is a warning that should be recorded rather than a failure on its own.

The angle of the crack is diagnostic. Cracks running down and outward at roughly 45 degrees through the resin-rich zone point to reflow thermal stress, while cracks running parallel to the pad surface follow bending or a depaneling load. Recording the angle together with the depth makes the report far more useful than a length on its own.

Scanning Acoustic Microscopy for Screening

Scanning acoustic microscopy images the interface by detecting where ultrasound reflects from a separated layer. It is non-destructive and can be run on production panels, but resolution falls as the transducer frequency drops, and both the pad and the solder above it attenuate the signal that has to travel through them.

The method works best as a screening and comparison tool, because the same measurement conditions have to be repeated for the numbers to mean anything, in the same way that X-ray void measurement only compares like with like. A shift in the reflection signal across a batch is a good reason to cut sections, but an acoustic image alone is rarely accepted as final evidence on a customer report.

Reflow and Thermal History Effects

Cratering is driven by the mismatch between the expansion of the laminate and that of the copper and solder, and most of the damage happens as the assembly cools through the rigid region below roughly 150 °C, when the joint has set but the board is still contracting. A profile with a peak of 245 °C and 45 s above 217 °C is ordinary for lead-free work, and raising the peak to fix something else can push a marginal laminate past its limit.

Moisture makes the mechanism worse, because absorbed water expands sharply at soldering temperature and weakens the resin around the interface. Parts held at moisture sensitivity level 3 or above should follow their floor-life and bake windows, and the thermocouple placement used to confirm the profile deserves a check so that the pad temperature is known rather than assumed.

Mechanical Loads After Assembly

Reflow is not the only load on the pad. Router and punch depaneling, screw fixing, connector mating and board flexure all put strain into the interface, and a pad that survived the oven can crater during a screw-down operation that deflects the board by a few hundred micrometres, which is the same mechanism discussed in our notes on warpage and reflow.

An inspection plan should therefore follow the board through assembly rather than stop at the oven exit. A dye and pry check on parts that have been through a known mechanical operation is often what identifies the real cause, particularly when the reflow profile has already been verified as being inside its window.

Limits, Sampling and Acceptance

There is no universal pass or fail number for a laminate crack, because the limit depends on the pad size, the number of thermal cycles expected in service and the consequence of an open circuit. What matters is that the criterion is written down before the samples are cut, and that the same method is used for the qualification lot and for every later comparison.

Sample size follows from the failure rate that has to be detected, and a check of three parts proves very little. Where the risk comes from a change of laminate supplier or a change in the reflow profile, a larger set taken across the panel, including the corners where the thermal mass differs, gives a result that can be defended.

Correlating Inspection With Field Returns

The final check on any inspection method is whether the defects it finds correspond to the failures that come back from the field. When a returned assembly shows an open pad with a cratered laminate underneath, the method used at build time was either too coarse or applied to too few parts.

Keeping dye images, sections and profile records together in one file makes that comparison possible and gives the next build a baseline. A laminate that craters at a consistent point in the profile is a material or design problem rather than an operator problem, and the evidence for that conclusion comes from the record.

FAQ

Is pad cratering visible on an assembled board under a microscope? No. The crack is in the laminate beneath the pad, so the top surface looks normal. Detection requires dye and pry, a microsection, or an acoustic scan through the pad.

How is dye and pry carried out? A penetrating dye is drawn into the interface under a vacuum and pressure cycle, the component is pried off, and the pad underside is examined at 30x to 50x. The stained area is reported as a fraction of the pad area.

Which method belongs in a qualification record? Use dye and pry or a microsection, since both show the interface directly. Acoustic microscopy is best kept as a screening and comparison tool, because its resolution depends on the stack and on the measurement conditions.