Application of adaptive multilevel substructuring technique to model CMOS micromachined thermistor gas sensor, part (II): effect of manufacturing uncertainties in the reliability of MEMS

For pt.I, see ibid., no.54, p.279-84 (2003). A study has been conducted to investigate the effect of the manufacturing uncertainties on the reliability of MEMS. The parameters investigated, include the uncertainties in the characterization of the Young´s modulus and the values of the residual stresses generated during the deposition process of MEMS thin films. The study was conducted on a CMOS micromachined thermistor gas sensor, recently proposed in the literature. A novel technique called adaptive multilevel substructuring was used to reduce the computational cost of the analysis. The numerical results suggest that, the uncertainty in the characterization of Young´s modulus has a reduced effect on the fatigue life. At the other hand the change in the value of the residual stress has a significant effect in the maximum operational stress level encountered during the operation, equivalent alternating stress value and consequently on the expected operational life of the MEMS component. The maximum expected life was found to occur at residual stresses values ranging from 0 to 400 Mpa. At these residual stresses values, the equivalent alternating stress is found to be lower than the endurance limit of the material. These values of the residual stresses correspond to a deposition temperature of 850°C and a SiH₂ Cl₂/NH₃ ratio ranging from 2 to 4 for the Si₃N ₄ film deposition process. The achieved results emphasize the important role that can be played by the numerical modeling of the end product. Using the numerical modeling, conclusions for the process parameters can be evaluated before proceeding to the actual microfabrication process.