6932

6474

كارشناسي ارشد

مكانيك سنگ

اصفهان: دانشگاه صنعتي اصفهان، دانشكده معدن

1390

سيزده،73ص.: مصور،جدول،نمودار

ص.ع.به فارسي و انگليسي

لهراسب فرامرزي، جعفر قيصري

مريم ذكري، عزت اله قطره ساماني

11/5/91

نادر فتحيان پور، محمود هاشمي

مهندسي معدن

ID6474

به فارسي و انگليسي: قابل رويت در نسخه ديجيتالي

Development an Empirical Model for Estimating the Deformation Modulusof Rock Masses Using the Result of Laboratory and in Situ Test with FuzzyInference System Case study Beheshtabad Dam Kaveh Soleymani K Soleymani@mi iut ac ir Date of Submission 2012 02 25 Department of mining Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language Farsi Supervisors Lohrasb Faramarzi lfaramarzi@cc iut ac ir Jafar Gheisari ghaisari@cc iut ac irAbstractDeformation modulus of rock mass which is composed of plastic and elastic deformations are defined as theratio of stress to corresponding deformation during the loading of rock mass This parameter is of fundamentalimportance for a safe and economical design of rock structures i e large underground opening like undergroundpowerhouses Among the rock mass properties the deformation modulus has vital importance for the design ofrock engineering projects because it is the best representative parameter for the pre failure mechanical behaviorof the rock material and of rock mass There are direct methods to determine the rock mass modulus i e in situtests and also empirical relations for indirect estimation of the deformation modulus In large geotechnicalprojects determination of this parameter is carried out by in situ tests which can cover a large volume of rockmasses subjected to load The plate loading test is a mostly familiar in situ test to measure the deformationmodulus of rock mass The determination of this parameter by in situ tests requires considerable costs andinvolves dif3cult operational processes On the other hand there are several emprical relations to estimatedeformation modulus but none of them alone is recommended In this study Neuro Fuzzy modeling isperformed in order to evaluate the deformation modulus of rock mass in Beheshtabad dam site At first theresults of the plate loading tests have been analyzed using different methods and the most appropriate one isapplied in the output of the model Then all the effective parameters in the modulus of the rock mass includingfield and laboratory tests evaluated by linear regression To evaluate the existing empirical equations RMRclassification system was performed at each test location and the error of each equation was calculated In thefinal step Adoptive Neuro Fuzzy Inference System ANFIS was used to predict the deformation modulus indifferent parts of the site For improve the neuro fuzzy model clustering algorithm was used to obtain the initialmodel Selecting the best combination of required input variables including of weathering degree density number of joints per meter Young s modulus strength of intact rock and porosity was performed by a methodbased on the gradual elimination of input data Modeling results show that the proposed method has highefficiency since the coe cient of cross correlation between predicted and measured deformation modulus is0 86 and the error of the model is 12 which illustrate higher accuracy whit respect to other models Thereforethe number of in situ tests can be reduced by means of this method Keywords Deformation modulus Fuzzy inference system Plate loading test ANFIS system Emprical equations

لهراسب فرامرزي، جعفر قيصري

مريم ذكري، عزت اله قطره ساماني

نادر فتحيان پور، محمود هاشمي