پيش بيني سفتي كامپوزيت هاي الياف كوتاه با استفاده از روش هاي هوش مصنوعي

چكيده انگليسي :

Prediction of Stiffness of Short Fiber Composites Using Artificial Intelligence Method Fatemeh Kowsar f kosar@tx iut ac ir Department of Textile Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language FarsiSupervisor Dr Ali Zadhoush azadhoush@cc iut ac irAbstractShort Fiber Composites have widespread use in different areas because of their easy and low costproduction On the other hand the conditions of the processes such as injection molding throughwhich they are produced lead to fracture of fibers and cause to randomly distribute them in thematrix Consequently there are fiber length and fiber orientation distribution in short fibercomposites Mechanical properties of these materials are markedly affected by these two factors So they must be taken into account in calculation of mechanical properties of short fibercomposites Besides these two factors since short fibers carry much less stress than would longfibers in the same system they are not providing very sufficient reinforcement so they have lowermechanical properties than long fiber composites Given all above reasons it is necessary to find anew theory for prediction of mechanical properties of short fiber composites To serve this purpose a lot of researches have been done so far Among them it can be mentionedto using finite element method correction of Mixture rule or micromechanical theories forunidirectional short fiber composites Bear in mind that some of these theories have limitingpreconditions and in some others the effects of fiber length and fiber orientation distribution are notconsidered in order to develop an efficient model which is able to predict stiffness of short fibercomposites ANN approach is applied in current study The main advantage of this method overprevious methods is that there aren t any limiting preconditions in it In order to train ANN thesesteps have been followed production of polypropylene reinforced with short glass fibers measurement of fiber length and fiber orientation distribution through image analysis measurementof longitudinal elastic modulus of the produced composites training a variety of ANN withdifferent structures to find the best network which can do the prediction most efficiently applyingfour different models into obtained data and evaluation the accuracy of these models and the modelcreated using ANN Final step has been done through comparison the results of calculated modulusby the models with experimental results to examine the agreement between them Based on theresults obtained in this study paper physics approach PPA can predict longitudinal elasticmodulus more efficiently than others Laminate analogy approach LAA and ANN get the secondrank rule of thumb and Cox Krenchel methods get the third and fourth rank respectively Thesefacts are revealed through root mean square error RMSE which is calculated for each model Thevalues of RMSE are 0 7206 1 0673 1 1331 1 2643 and 1 4230 for PPA LAA ANN rule ofthumb and Cox Krenchel respectively So the best model for prediction of longitudinal modulusof short fiber composites is PPA It is also found that rule of thumb is a suitable and simple methodfor estimation of longitudinal modulus of short fiber composites with low amount of fiber forexample 10 or 20 percent of weight fraction The main result of this research is that ANN is an appropriate approach to predict stiffness of shortfiber composites This is proved by the model created using ANN Keywords Short fiber composites ANN fiber orientation distribution fiber length distribution

كوثر، فاطمه

پيش بيني سفتي كامپوزيت هاي الياف كوتاه با استفاده از روش هاي هوش مصنوعي

توزيع طولي الياف , توزيع آرايش يافتگي الياف , شبكه عصبي