Four Point Bend Testing for Solder Joint Reliability

Four point bend testing loads a populated board so that it curves between two inner supports while being pushed by two outer rollers. The arrangement produces a region of uniform bending moment between the inner supports, which means the joints inside that region all see a similar load.

The test answers a practical question: how much board flexure can an assembly survive before a solder joint cracks? It is used to compare designs, to qualify a component package and to set handling limits for the production line.

What the Test Measures

The output is a load and deflection curve recorded until a defined failure occurs, or a strain measurement taken at a chosen location while the board is loaded to a stated deflection. The first approach describes the assembly, the second describes the local strain that a joint experiences.

Both are useful, and they answer different questions. A load curve is suited to comparing two designs or two soldering processes, while a strain limit is what a production engineer needs when deciding how much a board may be allowed to bow during depaneling or connector insertion.

Fixture Geometry and Setup

The span between the supports sets the moment arm and therefore the strain for a given deflection. A span of about 100 mm is common for small assemblies, with the inner supports separated by roughly half the outer span so that the uniform moment region covers the components of interest.

Support rollers should allow the board to slide slightly as it bends, because a board that is constrained horizontally develops membrane tension as well as bending, and that additional load is not part of the intended measurement. The rollers should also be free of nicks, since a point contact can start a crack of its own.

Strain and Deflection Measurement

Deflection is read from the crosshead displacement, which includes the compliance of the fixture and the supports. Strain is read from gauges bonded to the board surface close to the joints of interest, and it is the more transferable number because it does not depend on the stiffness of the test machine.

Gauges should be placed on the component side, aligned with the direction of bending, and as close to the package corner as the layout allows. Metal foil gauges give the cleanest signal, while the small gauges used for this work require careful bonding and a stable bridge supply.

Loading Rate and Time to Failure

The crosshead speed sets the strain rate, and solder behaves differently at different rates. A slow test, in the range of a few millimetres per minute, allows the alloy to creep and produces a lower failure load than a fast test on the same assembly.

The rate should therefore be stated with the result and held constant between tests that are to be compared. A common starting point is 1 to 5 mm per minute for monotonic testing, with the choice recorded together with the span and the support geometry so that another site can reproduce the setup.

Where the Board Flexes and Why Joints Fail

Bending puts the outer fibres of the board into tension on one side and compression on the other, and the joints on the tension side are the ones at risk. The load path runs from the board through the solder into the component, and the failure usually occurs where that path is narrowest.

For a ball grid array, the critical joints are the corner balls, which carry the highest strain. For a leaded package, the failure normally appears in the solder fillet or at the interface between the fillet and the pad. Package stiffness changes the result as well, because a rigid body concentrates strain in the joints nearest its corners. Damage that starts in the laminate under a pad is the mechanism described under pad cratering.

Interpreting Failure Locations

After the test, the assembly is examined to find which joints failed and how. Optical inspection under magnification identifies cracked fillets, and a cross section of a failed joint shows whether the crack ran through the bulk alloy, along the intermetallic layer or into the laminate.

The location points to the cause. A crack in the bulk alloy with a ductile appearance suggests that the joint was overloaded, a crack at the intermetallic interface suggests a soldering or ageing issue, and a crack that has propagated into the resin under the pad indicates that the board, not the joint, was the weakest link.

Board Design Factors That Change the Result

Board thickness, the position of the neutral axis and the amount of copper under the components all change the strain for a given deflection. A thick board with heavy planes flexes less, and a thin board with a large cut out flexes more, so results from two different designs are not comparable unless the geometry is matched.

Component placement has an equally strong effect. A heavy component in the middle of the bend region applies a point load that localises strain, and two large components placed close together stiffen the assembly locally and move the failure elsewhere. Test coupons should mirror the production layout, not a simplified version of it.

Comparing Materials and Layouts

The test is most valuable as a comparison run under identical conditions. Two laminate types, two surface finishes, two reflow profiles or two pad designs can be ranked by the deflection at which the first joint fails, provided the coupon geometry, the loading rate and the failure criterion are the same.

The failure criterion has to be defined before the test begins. A drop in load of a stated percentage, a measured strain at a chosen location, or the first detectable fracture found by inspection are all valid, but mixing them across a data set makes the comparison meaningless. The fracture criterion should be written down before the first specimen is loaded, because the decision becomes much harder once a curve is in front of you.

Records and Specification Limits

A bend test record should contain the assembly identification, the span, the crosshead rate, the deflection and strain at failure, the failure location and the inspection method used to find it. Photographs of the failure add the context that numbers alone cannot carry. Where a joint is sectioned for confirmation, the preparation follows the same rules as any other microsection.

The result is then translated into a handling limit. If the first failure occurs at 2000 microstrain and the production line can be shown to stay below 800 microstrain, there is a defensible margin. That translation is the reason for running the test at all, and it is worth more than a single pass or fail number.

FAQ

What does a four point bend test measure? It applies a uniform bending moment between two inner supports and records how much deflection or strain the assembly withstands before a solder joint fails. It is a comparative test of assembly robustness.

Why use four points instead of three? Four point loading keeps the bending moment constant between the inner supports, so every joint in that region sees the same load. A three point setup concentrates the moment under the single load point.

Can the result predict field reliability? It ranks designs and sets handling limits. Translating a monotonic bend result into a field life needs a validated model and additional testing such as thermal cycling or vibration.

}