High temperature loads were applied under isothermal conditions to a film type epoxy resin, which is utilized in ultra-thin semiconductor packages. Dynamic viscoelastic properties, such as storage modulus, loss modulus and glass transition temperature, were measured after these accelerated thermal aging tests. An increase in the glass transition temperature upon thermal aging could be described by an Arrhenius process relating the aging temperature and a thermal aging rate coefficient.

View full abstract

Heat transfer characteristics of flowing polymer melts are examined by insertion of a probe into the melt under laminar flow at 160-280°C. It is found that the heat transfer coefficients of high-density polyethylene (HDPE) and polycarbonate (PC) melts are only weakly dependent on temperature, but very strongly on flow velocity. The thickness of the equivalent conduction layer, which forms in the vicinity of the probe surface where the polymer melt is virtually in a nonflow state, decreases with flow velocity as a power function. The boundary flow velocity, which dominates the heat transfer behavior of flowing polymer melts, is 0.08-0.12mm/s and 0.07-0.12mm/s in the case of HDPE and PC, respectively. Above a flow velocity of 0.07-0.12mm/s, the dominant form of heat transfer in HDPE and PC melts is convection, whereas under lower flow velocities, heat transfer by conduction is found to dominate. The equivalent conduction layer thickness at any given temperature decreases with flow velocity and Reynolds number (Re). The Prandtl number (Pr) of HDPE and PC melts increases with melting viscosity (η′). In addition, the Pr of a polymer melt depends on temperature and flow velocity. The relationship between the Pr and η′ of a polymer melt is clearly linear, the slope β (1/Pa·s) of the straight line 3.2×10³ for HDPE, 10.7×10³ for PC being equal to the ratio of specific heat and thermal conductivity. The heat transfer coefficient measured by this probe method is useful as input resin data in flow and thermal analysis CAE programs.

View full abstract