Dye and Pry Analysis for Solder Joint Crack Detection

Dye and pry is the test that answers a question X-ray cannot: where exactly did a solder joint crack, and how far has the crack travelled. It is destructive, it takes a day, and it produces a fracture surface that either shows a clean joint or a mapped history of crack growth. This article covers the procedure and how to read the result.

What Dye and Pry Shows That Other Tests Cannot

X-ray shows voids and shape but struggles with a hairline crack, especially at the interface where the crack is thin and oriented along the beam. Dye and pry separates the joint and stains the surfaces the dye could reach, so the cracked area becomes visible as a coloured region on the fracture face.

The result is a map rather than a pass or fail. A joint with a crack that started at the package corner and progressed partway across is a different problem from one that failed across the whole interface at once, and the two have different causes.

<img src="https://www.gopcba.com/wp-content/uploads/2026/09/224-1.jpg" alt="BGA package removed by prying after dye penetration testing” />

The method is qualitative and comparative, which is exactly what makes it useful during a failure investigation where the question is what changed rather than what the absolute quality is.

Procedure: Dye Penetration and Cure

The assembly is cleaned and dried first, because flux residue can block the crack and prevent the dye from entering. The dye is applied around the package and drawn in by capillary action, often helped by a vacuum step that removes air from the crack before the dye is introduced.

The dye must then be cured before the package is removed. Curing sets the penetrant in place so that the pattern survives the mechanical separation, and a partially cured dye smears during the pry and destroys the evidence.

Penetration time depends on crack width. A wide crack fills quickly, while a hairline crack may need several minutes under vacuum, and a short penetration time is the most common reason for a misleadingly clean result.

Pry Method and Force Control

The package is removed by applying force at a corner, usually with a blade or a wedge, until the joints separate. The force has to be enough to break the remaining joints without tearing the board, and the direction of the pry determines which joints open first.

Consistency matters more than the absolute force. Two operators prying the same board in different directions will leave different fracture patterns, so the procedure should define the tool, the position and the direction for a given package.

Interpreting the Fracture Surface

The fracture face is examined under magnification or with a microscope, and the stained area is compared with the total joint area. A joint that is largely unstained failed cleanly at the pry, while a joint with a stained region failed during the test that preceded the pry.

The location of the stain is the most informative part. Staining at the package interface suggests a thermal expansion problem, staining at the board interface suggests a pad or surface issue, and a stain that runs through the bulk of the joint suggests fatigue in the alloy itself.

Crack Modes in BGA and CSP Joints

The common modes are a crack running around the solder ball at the package side, one at the board side near the pad, and a bulk crack through the ball. Each has a characteristic appearance, and each points to a different root cause in design, materials or process.

Brittle interfacial failure, which defect tracking treats as a priority, is the mode that worries reliability engineers most, because it produces a joint that appears intact and separates under very little load. It is usually associated with excessive intermetallic growth or with contamination at the interface.

Applying Dye and Pry After Thermal Cycling

The test is most valuable when it follows a defined preconditioning sequence, usually thermal cycling or shock with a stated number of cycles. The cycle count becomes part of the result, because a crack pattern after five hundred cycles means something different from the same pattern after two thousand.

Comparing a cycled sample with an uncycled control from the same lot separates the damage caused by cycling from defects that were present after assembly. Without the control, every crack is attributed to the test.

Sample Selection and Statistics

Sample selection decides how much the result means. The joints that fail first are usually at the corners of the package and at the outer rows of the array, so a sample that includes only the centre describes the strongest joints rather than the weakest.

Where the assembly is underfilled, the dye has to travel beneath the underfill before it can reach the joint. That may need a longer vacuum cycle and a warmer penetrant, and it is one of the cases where a clean result should be treated with suspicion rather than accepted at face value.

Several packages from several positions on the panel give a better picture than one package examined in detail. Where the result will be used to accept or reject a process, the number of joints assessed should be stated, because a conclusion drawn from four joints is weak.

Limitations and False Readings

The method cannot detect a crack that is fully closed, and it cannot distinguish a crack that formed during cycling from one that formed during the pry if the dye did not penetrate. Contamination, incomplete cleaning and insufficient cure all produce false clean results.

Preparation should also include a photograph of the board before the pry, because the package position and joint orientation are needed to interpret the fracture faces later. A scale marker in the frame keeps images comparable between samples and between analysts, and a note of the assembly date, the preconditioning and the pry date establishes how much of the crack existed before the test began.

It is also destructive and slow, which makes it unsuitable as a production control. Its role is failure analysis and process qualification, where the information it produces justifies the sample it consumes.

Documentation and Reporting

A useful report includes the sample identity, the preconditioning applied, the dye and cure times, the pry method and photographs of each fracture face with a scale. The percentage of stained area should be recorded per joint rather than as an average, because averaging hides the joint that matters.

gopcb uses dye and pry as one of the tools in a reliability investigation, alongside X-ray inspection and sectioning. Each method shows something the others cannot, and the quality decision is made from the combination rather than from a single image.

FAQ

Can dye and pry be used in production? No. It is destructive and slow, so it belongs in qualification and failure analysis rather than in routine inspection.

Why does a joint sometimes look completely clean? Either it did not crack, or the dye never entered the crack because of residue, insufficient vacuum or too short a penetration time. Check the procedure before concluding the joint is sound.

How should the cracked area be reported? As a percentage of the joint area per joint, with a photograph. Averages across a package hide the corner joints where failure usually starts.