May 3, 2024
Mohammad Reza Mahini

Mohammad Reza Mahini

Academic Rank: Assistant professor
Address: -
Degree: Ph.D in -
Phone: -
Faculty: Faculty of Engineering

Research

Title Procedures for characterization of compressive material parameters for GRE pipes, using flexural modulus test results and inverse analysis
Type Article
Keywords
Inverse analysis, GRE pipe, Material characterization, POD + RBF, Flexural modulus test
Journal MECCANICA
DOI https://doi.org/10.1007/s11012-023-01659-0
Researchers Saeid Hosnifam (First researcher) , Mohammad Reza Mahini (Second researcher) ,

Abstract

Nowadays, the use of pipes manufactured with composite materials is preferred for transferring fluid by the oil and gas industries. For the correct design of such infrastructures, determining their material parameters makes recourse to specifically regulated experimental tests. The tensile properties of glass resin epoxy (GRE) pipes can be determined using conventional direct tensile and internal hydrostatic pressure tests. However, the conventional shear and compressive test methods for determining the compressive and shearing properties of a single GRE layer cannot be followed reliably due to uncertainties relevant to buckling, stress concentrations and material locality. In addition, the filament winding’s direction, characterizing the orthotropic behaviour of such (GRE) composite material, can be chosen in order to optimize the mechanical behaviour of the pipe. Herein, an inverse analysis approach, previously applied successfully by authors to different mechanical contexts, is proposed to overcome the possible above-mentioned uncertainties relevant to the compressive and shearing properties of GRE material and to optimize the filament winding’s direction. The load–displacement response of the flexural modulus test ASTM D 2412 is considered as a measurable quantity to be used in the inverse procedure. The efficiency of the proposed method, in the sense of accuracy and stability, is verified after creating some pseudo-experimental tests, whose numerical results have been statistically perturbed in order to simulate experimental data’s noise. The proposed identification procedure would require a relatively high computing time, an uncommon knowledge in computational mechanics, and a sophisticated computing system. Therefore, in order to speed up the characterization procedure and remove any need of deep computational mechanics knowledge and advanced computing devices, an “Offline” analysis tool, based on proper orthogonal decomposition and radial basis functions whi