Fillet Lifting in Wave Soldered Joints: Why It Happens

Fillet lifting is the separation of the solder fillet from the wall of a plated through-hole during wave soldering, as the joint solidifies. The joint looks complete from above and it is electrically continuous, which is why it is usually found by a microsection or by a thermal cycle test rather than by an inspector. The lift is a crack between the solder and the barrel that follows the circumference of the hole.

The mechanism is a mismatch of movement during freezing. The solder contracts as it cools, the laminate and the copper barrel contract at a different rate, and the resulting stress is concentrated at the point where the solder meets the barrel wall. Where the alloy solidifies over a wide range and the joint is constrained, the stress exceeds the strength of the bond and the fillet separates.

What Fillet Lifting Looks Like

In section the fillet appears to have pulled away from the barrel wall on one or both sides of the hole, leaving a gap that narrows as it approaches the surface. The solder itself is sound, and the barrel plating is intact, which distinguishes the defect from a cracked barrel or an incompletely filled hole.

On the board the defect is nearly invisible. There may be a slight step at the top of the fillet or a dull line where the lift reaches the surface, and both are easy to miss under normal inspection lighting. That is why a process that produces fillet lifting is usually discovered by a thermal cycle test rather than by visual inspection.

The Mechanism

Three events happen at once as the board leaves the wave. The alloy cools through its solidification range, the board and the barrel cool at their own rates, and the joint is constrained by the laminate that surrounds it. The alloy contracts on cooling, and because the barrel constrains it in one direction and the pad in another, the contraction is converted into a shear stress at the interface.

The interface is the weakest link when the alloy has not fully wet the barrel wall or when an intermetallic layer is thick and brittle. A well wetted joint can accommodate the movement, while one with a thin or contaminated bond cannot. The failure is therefore a combination of the stress and the quality of the interface rather than of the stress alone.

Alloy Composition and Contamination

The alloy sets the solidification range and therefore the amount of movement that happens while the joint is partly liquid. A lead-free alloy such as a tin-silver-copper composition solidifies over a range of a few degrees, while a tin-copper alloy with a higher melting point solidifies differently and a contaminated bath can widen the range considerably.

Copper in the bath is the most common contaminant that matters, because it raises the liquidus and widens the gap between the liquidus and the solidus. A bath that has drifted out of its composition produces joints that freeze over a longer interval, and those joints are the ones that lift. The analysis side of that drift is described in the notes on solder bar purity checks.

Cooling and Solidification

The cooling rate after the wave determines how much time the joint spends in the range where it is partly liquid and mechanically weak. A slow cooling rate from a hot board or from a warm ambient lets the stress develop slowly and gives the alloy more time to accommodate it, while a fast cooling rate freezes the joint in a stressed state.

The pot temperature and the contact time set the starting point of the cooling. A joint that leaves the wave hot has further to cool, and a joint with a large pad has more mass to hold that heat. Both change the timing of the stress relative to the strength of the bond, and both are recorded with the profile rather than adjusted afterwards.

Board and Pad Geometry

Geometry sets the constraint. A pad that is large relative to the hole gives the alloy more area to bond and more constraint at the same time, while a small pad with a heavy barrel concentrates the stress. The thermal relief pattern around a pad changes the heat flow and therefore the cooling, and the notes on barrel fill describe the same geometry from the fill side.

The board thickness and the hole diameter set the aspect ratio, and a deep, narrow hole is more constrained than a shallow one. Where a design has both a high aspect ratio and a large pad, the process window for the fillet is narrow, and the manufacturing engineer has fewer parameters available to open it.

Laminate and Finish Effects

The laminate expands in the direction of its thickness far more than the copper does, and that difference in thermal expansion is what loads the joint. A laminate with a high coefficient of expansion in the through-thickness direction moves further during the cooling and stresses the bond harder, and a thick board amplifies the effect because the movement accumulates over the depth.

The surface finish on the barrel and the pad affects wetting and therefore the quality of the interface. A finish that oxidises or that is too thick interferes with wetting, and the resulting joint has a weaker bond at exactly the location where the stress is highest. The finish side is a design decision, and it is compared in the notes on the surface of the copper before assembly.

Detection and Measurement

Detection is by microsection, with the cut taken through the axis of the hole and the joint examined on both sides. The measurement is the length of the separation and its position relative to the surface: a lift that reaches the board surface is more significant than one that is confined to the lower part of the barrel, because the surface end is where the environment reaches it.

Thermal cycling is the production-level detector. A joint with a lift that is mechanically marginal will open further under a few hundred cycles, and a resistance measurement taken during the test detects the change before a functional failure appears. The test itself is described in the notes on intermetallic growth and joint ageing.

Process Controls

The controls that reduce fillet lifting are the alloy composition, the pot temperature, the contact time, the cooling after the wave and the flux activity. Each of them can be adjusted, and each has a consequence: a higher pot temperature improves wetting and increases the cooling stress, and a longer contact improves fill and heats the board further.

The most effective single control is the alloy, because the solidification range is a property of the material rather than of the machine. A bath that is analysed on schedule and topped up correctly holds that range, and the defects that follow a composition drift disappear when the drift is corrected. The bath practice that supports it is described in the notes on wave solder blowhole control.

Records and Limits

The records are the bath analysis, the pot temperature, the profile with the contact time and the cooling, the flux type and volume, and the laminate lot. A fillet lifting investigation that starts from those fields usually finds a single changed variable within a day, while an investigation that starts from the defect alone has to reproduce it in a trial.

The acceptance limit for a lifted fillet should be tied to a thermal cycle result rather than to a visual criterion. A joint that passes several hundred cycles with a small lift is different from one that opens in fifty, and the difference is not visible on the board. Writing the limit in terms of the test that established it is what makes it defensible.

FAQ

Is fillet lifting the same as a cracked barrel? No. The barrel plating is intact in a lifted fillet and the separation is at the interface between the solder and the barrel wall.

Can it be seen on the assembly line? Rarely. It is found by sectioning or by a thermal cycle test, because the surface of the joint usually looks complete.

Does a lower pot temperature help? It reduces the cooling stress and it also reduces wetting, so the effect on the defect depends on which of the two dominates. The alloy is the more reliable lever.