Comparison of solder joint fracture behavior in Arcan and DCB specimens

Predicting fracture of solder joints with different constraint factors

“…DCB specimens of 2.5‐mm‐long solder joints were prepared as described in Nourani et al, Nourani and Spelt, Akbari et al, Nourani and Spelt, and Nadimpalli and Spelt Copper bars of C110 alloy were cut to the dimensions indicated in Figure . Milling process with a depth of 0.2 mm was performed on the surfaces to be soldered in order to eliminate the roundness of the original copper bars.…”

“…A piece of bleached cheese cloth was used to wipe out any dirt remained on the surfaces. This procedure produced an organic solderablility preservative (OSP) surface finish . The soldering process was done on a hot plate with a constant temperature of 220°C, almost equal to the melting point of the flux‐cored Sn3.0Ag0.5Cu (SAC305) solder wire.…”

“…The copper bars were clamped together against the steel wires with a thickness equal to the desired thickness of the solder joints (ie, 250 μm). Time above liquidus (TAL), measured right after the soldering process started, was adjusted to be 120 seconds as done in Nourani et al, Nourani and Spelt, Akbari et al, Nourani and Spelt, and Nadimpalli and Spelt. Consistent with Nourani et al, Nourani and Spelt, Akbari et al, and Nourani and Spelt, the specimen was cooled in a small wind tunnel with a cooling rate of 1.4 to 1.6°C/second, which is typical of microelectronic manufacturing .…”

“…The structure of adherends has been shown to influence the fracture behaviour of these joints. For example, the critical strain energy release rate for crack initiation, J ci , was obtained as a significant function of adherend structure and type of loading in Nourani et al where the model 2‐mm‐long SAC305 solder joints between 2 copper substrates were examined using double cantilever beam (DCB) and Arcan specimens.…”

Bending strength of adhesive joints in microelectronic components: Comparison of edge-bonding and underfilling

“…As explained in Section 4.2.2, the load-displacement curve was linear to failure in all DCB specimens, and no crack extension was observed before the attainment of the maximum load, which was followed by rapid fracture. Therefore, the crack initiation strain energy release rate (G ci ) of the underfills and the PCB were determined using these maximum loads as inputs to a finite element model following [26,27]. A 250-lm-long crack was modeled using singular elements in the mid-plane of the underfill or within the PCB, and G ci was calculated using the J-integral method.…”