Study On The Integrative Forming Quality Assessment And Forming Rules Verification Of Low Temperature Co-fired Ceramic

High integration density,excellent radio frequency performance,quick signal response and strong adaptability to heavy current and high temperature environment are the main characteristics of LTCC(Low Temperature Co-Fired Ceramic).Due to these advantages,LTCC is chosen to manufacture the next generation substrate for electronic module.However,the complicated forming technology and lots of process parameters of LTCC,cause fairly high defective rate of substrate.For addressing this issue,this dissertation studies the assessment methods of LTCC forming quality and verification methods of LTCC forming rules in depth.The main topics of this dissertation include:(1)Based on the construction theory of index system,the initial index system for evaluation of LTCC forming quality has been built.Then,on the strength of screening theory and structure optimization theory,the simplification and optimization analysis of initial index system has been done to reduce the redundancy of indicators.Moreover,the analysis of existing approaches for determining the weights of indicators lays the foundation to define indicators' weights.(2)On the basis of rough set theory and TOPSIS,a mathematic model is established to determine the weights of decision makers in multiple attribute group decision-making(MAGDM)problems.Then,the order of alternatives is defined by combining the weights of indicators.For a deep analysis of the uncertain preference information of decision makers,the mathematic model is extended to the interval number environment,in which the attribute data is expressed in interval number.Furthermore,by utilizing angle cosine and projection method simultaneously,a direct method is developed for determining the weights of experts.Through comparisons with existing methods,the direct method shows high resolution to preference information of decision makers and simple calculation steps.(3)After fully analyzing the advantages,disadvantages and application range of existing comprehensive evaluation methods,a specific comprehensive evaluation approach is proposed for integrative forming quality of LTCC.This method combines the index system and decision makers' weights.Then,this approach is applied to calculate the forming quality of RF(Radio Frequency)module to show its rationality.Finally,an example of forming quality evaluation of various LTCC prototypes,is provided to clarify the availability and validity of the proposed comprehensive evaluation system of LTCC forming quality.(4)Considering the population variance,confidence and error range,the small sample sampling method for the samples from the normal and non-normal distribution population has been researched.For the samples from two different distributions,the mathematic models are built to calculation the number in sampling.Through verifications on the mathematic models of four typical defects' forming rules in micro channels,the sampling steps for small sample have been demonstrated in detail.(5)Taking the sampled data from small sample,Euclidean distance measure method and correlation coefficient method into account,a similarity calculation model is developed between theoretical values and experimental values.Then,the similarity is mapped to goodness of fit by the Gaussian distribution curve.A judgment standard for the goodness of fit is given.Through verifications on the forming rules of LTCC prototypes,the verification system for forming rules of LTCC turns out to be rational and valid.Based on the above research,the main innovation points of this dissertation are as follows:(1)In order to deal with the subjective and heuristic decisions of a group of decision makers in a simple and efficient way,an easy operation method in this dissertation is developed for determining weights of decision makers based on rough set theory.This method is also extended to the interval number environment.By contrast,the proposed method can effectively manage the subjectivity of decision makers in decision process and reflect the vagueness of decision makers objectively in a simple way.(2)A direct approach is developed to determine weights of decision makers in the group decision-making problems via angle cosine and projection method.Comparisons with existing methods show that,the proposed approach has advantages in distinguishing the differences among preference information of all decision makers.Therefore,the proposed method has strong ability to reflect differences of decision information,which are caused by decision makers' knowledge level and cognitive capability.Besides that,as it has only one ideal solution,the proposed method is more effective.(3)By the theoretical analysis of existing small sample sampling methods,the methods for calculating the number in sampling models under different confidence levels and different error ranges are proposed.Comparison with Chebyshev method shows advantages of the proposed method.The illustrative examples of verifications on the forming rules of LTCC prototypes are given to demonstrate the generality and validity of mathematic models of small sample sampling.