Abstarct: Magneto rheological damper, as a widely used equipment in various industries for the first time has investigated and optimized by using dissipative particle dynamics method as a molecular modeling approach in aspects of molecular properties of magnetic fluid in it and its physical structures. During the studies, optimization process has performed with two approach; the first view point is investigation of a magneto rheological damper by focus on MR fluid and physical structure of damper. In order to selecting a suitable the effect of diameter and weight of magnetic particles on considered damping force of 10 N as favorite damping force is studied. Results show that by increasing in diameter of magnetic particles, damping force firstly increases then trends to a constant value; while, by enhancement in weigh of them damping force firstly increases and then decreases. Results of physical structures of damper modeling show that by increasing in gap size of fluid flow, damping force increases; while, while by enhancement of inner diameter of cylinder damping force decreases. Optimum design of MR damper has probed by investigation on arrangement of electrical coils and step wise magnetic field strength distribution as an innovation. The best operating conditions of MR damper occurs when the electrical energy consumption and hysteresis level are minimum and three electrical coils with length of 3, 5 and 7 mm are employed which the baxse magnetic strength fixed in 40% and height of steps is in 15% and 140-CG MR fluid is utilized. In the second view point of optimization of operating, in addition of study of the characteristics of MR fluid, different four scenarios of arrangements of rough sections on piston surface are investigated. Results of modeling have compared with available experimental data which show good conformity. Weierstrass-Mandelbrot function is used to applying rough surface profile on the piston's body and bounce back boundary condition is utilized as no-slip boundary condition. Results of dissipative particle dynamics modeling show that by increasing in number density of these particles and increasing mass of carrier fluid particles, damping force firstly increases and then decreases; therefore, It is necessary to set optimal values. It is also observed that by decreasing thickness in surfactant laxyer at the surface of the magnetic particles, damping force increases. According to the results of investigation on MR fluid's structure, 140-CG MR fluid is selected as the suitable agent fluid to DPD modeling. Also, it specifies that by increasing in fractal dimension of roughness profile, damping force has initial enhancement and then trends to a constant value. By utilizing Bouc-Wen model and genetic algorithm method and by employing fractal dimension of roughness profile of 1.5 using at the beginning and the end of the piston, by applying 20% of basic magnetic field strength, considered damping force of 10 N achieved within 2 seconds; while, less damping time and electrical energy is required.