Dimensional stability of multilayer circuit boards

The present work investigates the time and temperature dependent response of the woven composite substrate used in multilayer circuit board applications and its influence on the residual stress development during processing and post-processing of the circuit boards. The fabric architecture of one commonly used substrate (7628 fabric style) is characterized using optical microscopy. The creep compliance and stress relaxation of the composite substrate and the FR-4 matrix are determined through accelerated viscoelastic characterization. Both finite element analysis (FEA) and theoretical analysis models are developed for prediction of the substrate response from the measured matrix and fabric properties. Comparison of the micromechanical model predictions with the measured response reveals the influence of fabric architecture and boundary conditions on the composite viscoelastic properties. Moiré interferometry is utilized to investigate deformation of the composite unit cell and verify the physical basis for kinematic assumptions of the micromechanical models.; Numerical and experimental studies are performed to study residual deformation and warpage in a model multilayer circuit board construction of a common composite substrate (7628 fabric style). A numerical procedure based on classical lamination theory with non-isothermal viscoelastic constitutive relations is developed to predict the deformation and residual stress state due to relamination. Experimental values of the substrate stress relaxation modulus and coefficients of thermal expansion (CTE) are used as inputs in the numerical procedure to predict warpage of model circuit boards with a non-symmetric lay-up of 7628 style composite substrate. Boards with the exact same construction as used in the numerical analysis were fabricated according to the prescribed pressing cycle and the time dependent warpage measured using an ultrasonic contour scan technique. Comparison of the experimental warpage data with numerical predictions provides insight into the effects of processing cycle and substrate properties on residual stress development.