This research aims at characterizing and predicting the Young’s Modulus of thin film materials that are utilized in Microelectromechanical systems (MEMS). Recent studies indicate that the mechanical properties such as Young’s Modulus of thin films are significantly different from the bulk values. Due to the lack of proper understanding of the physics in the micro-scale domain the state-of-art estimation techniques are unreliable and often unfit for use for predicating the mechanical behavior of slight modifications of existing designs as well as new designs. This disadvantage limits the MEMS designers to physical prototyping which is cost ineffective and time consuming. As a result there is an immediate need for alternative techniques that can learn the complex relationship between the various parameters and predict the effective Young’s Modulus of the thin films materials. The proposed technique attempts to solve this problem using empirical estimation techniques that utilize soft computing techniques for the estimation as well as the prediction of the effective Young’s Modulus. As a proof of concept, effective Young’s Modulus of Aluminum and TetrathylOrthoSilicate (TEOS) thin films were computed by fabricating and analyzing self-deformed micromachined bilayer cantilevers. In the estimation phase, 2D search and micro Genetic algorithm were studied and in the prediction phase, back propagation based Neural networks and One Dimensional Radial Basis Function Networks (1D-RBFN) were studied. The performance of all combinations of these soft computing techniques is studied. Based on the results, we conclude that performance of the soft computing techniques is superior to the existing methods. In addition, the effective values generated using this methodology is comparable to the values reported in the literature. Given a finite number of data samples, the combination of 1D-RBFN (prediction phase) and GA (estimation phase) presented the best results. Due to these advantages, this methodology is foreseen to be an essential tool for developing accurate models that can estimate the mechanical behavior of thin films.
Department, Program, or Center
Microelectronic Engineering (KGCOE)
Engineering Letters 14N2 (2007) EL_14_2_12
RIT – Main Campus