Assessment of Adhesive Bond Strength for Long-Life Performance

One of the great challenges of adhesively bonded systems is to predict their long-term performance when exposed to environmental stressors. Ballistic glass delamination provides a relevant case study.  Ballistic glass is typically a multi-layered composite with several layers of a hard material (glass) and an inner polymeric layer (polycarbonate).  The adhesive between the glass and polycarbonate layers is often a thermoplastic polyurethane (TPU), which has high clarity. The bond between the TPU and polycarbonate is known to fail in service by interlayer delamination.  The delamination allows moisture and contaminants to infiltrate and reduce visibility. As is similar for other systems, the initial bond strength is sufficient to resist immediate failure.  Delamination only occurs after time (a few years). Because delamination occurs so slowly, it is very difficult to correlate specific material and manufacturing processes to improved resistance to delamination.  While various studies have shown that delamination can be accelerated with heat and humidity, they do not quantify the bond strength or predict the rate of failure in the field. Test methods are needed that can accurately quantify bond long-life strength and accelerate failure.

We report successful utilization of a simple wedge test to quantify delamination by measuring the mode I energy release rate in polycarbonate-TPU-polycarbonate specimens exposed to a range of temperatures and humidity. The nature of the wedge test is that the driving force on the sample decreases as the delamination extends.  Over a period of 100-150 days, delamination gradually slowed until arrest.  From these tests, we were able to measure crack growth rate and a delamination threshold stress intensity factor.  The threshold stress intensity factor was measured at 0.12 MPA m^0.5 for samples held at 65 C.  For samples held at 50 C, the threshold stress intensity factor was about three times higher.  The 65 C tests required 100 days, while the 50 C tests required more than 150 days to arrest. Samples held at room temperature did not show any delamination growth, even after several months. Humidity appeared to accelerate the initial delamination rate, but not to significantly affect the final threshold.

By combining environmental exposure and time in the wedge test, we successfully quantified a bond strength that provides a reasonable explanation for widespread delamination in fielded parts.  We propose that this technique might provide a way for more rapid screening of new materials and manufacturing processes.  We also discuss how the test might provide some insight into the nature of the bond failure that leads to delamination and how this could be combined with modeling to begin to predict bond longevity under a range of environmental conditions.