| Summary: | 碩士 === 國立屏東科技大學 === 機械工程系所 === 98 === A reliable and accurate analytical model is desired for the printed circuit board (PCB) with IC package to predict the system response due to loadings such as shock and vibration simulation or even with thermal effect. This work addresses the procedure of model verification by the adoption of experimental modal analysis (EMA) to validate the finite element (FE) model constructed by FE commercial software. The PCB with one package adhered with the heating pad to emulate the heat effect is first considered for completely free boundary condition. The refined FE model of the PCB consists of detail components, such as the chip, substrate, compound and solder balls. The thermal effect on the PCB is simulated to conduct the temperature field analysis as well as the thermal stress. The modal analysis on the PCB with the heating in steady state is then performed to obtain the structural modal parameters, i.e. natural frequencies and mode shapes. The EMA is also carried out to determine the system modal properties that are used to update the analytical FE model. Through the comparison of frequency response functions and modal parameters between the analytical FE model and the real PCB structure, the refined FE model can be verified for material properties and thermal boundary conditions. The same procedure for model verification is then conducted via both EMA and FEA on the PCB in the fixed boundary that complies with the test fixture for the random vibration test of JEDEC specification. The verified equivalent FE model of the PCB can then be adopted to perform spectrum response analysis according JEDEC random vibration test specification. The acceleration power spectral density (PSD) spectrum obtained from FEA is compared and shown good agreement with the experimental results. The stress prediction on the components of PCB can then be evaluated for possible failures due to both thermal effect and random excitation. This work lays out the procedure to validate the analytical model, in particular for the PCB with the IC package in details considering the thermal effects. The response prediction of the PCB subjected to random vibration input with thermal effect is performed to evaluate the component failure. The developed methodology will be beneficial to the design analysis of PCB, in particular for coupling loadings of thermal effects and random excitations.
|
|---|
