Using a hybrid simulation approach that combines a finite difference method with a Brownian dynamics,we investigated the motion of charged polymers.Owing to the fact that polymer-solution systems often contain a large...Using a hybrid simulation approach that combines a finite difference method with a Brownian dynamics,we investigated the motion of charged polymers.Owing to the fact that polymer-solution systems often contain a large number of particles and the charged palymer chains are in a state of random motion,it is a time consuming task to calculate the electrostatic interaction of the system.Accordingly,we propose a new strategy to shorten the CPU time by reducing the iteration area.Our simulation results ilustrate the effect of preset parameters on CPU time and accuracy,and demonstrate the feasibility of the"local iteration"method.Importantly,we find that the increase in the number of charged beads has no signifiant infuence on the time of global iterations and local iterations.For a number of 80×80×80 grids,when the relative error is controlled below 1.5%,the computational efficiency is increased by 8.7 times in the case that contains 500 charged beads.In adition,for a number of 100×100×100 grids with 100 charged beads,the computational efciency can be increased up to 12 times.Our work provides new insights for the optimization of iterative algorithms in special problems.展开更多
基金the National Natural!Science Foundation of China(No.21604086)Jilin Provincial science and technology development program(No.20190103124JH)Key Research Program of Frontier Sciences,CAS(No.QYZDY-SSW-SLH027).
文摘Using a hybrid simulation approach that combines a finite difference method with a Brownian dynamics,we investigated the motion of charged polymers.Owing to the fact that polymer-solution systems often contain a large number of particles and the charged palymer chains are in a state of random motion,it is a time consuming task to calculate the electrostatic interaction of the system.Accordingly,we propose a new strategy to shorten the CPU time by reducing the iteration area.Our simulation results ilustrate the effect of preset parameters on CPU time and accuracy,and demonstrate the feasibility of the"local iteration"method.Importantly,we find that the increase in the number of charged beads has no signifiant infuence on the time of global iterations and local iterations.For a number of 80×80×80 grids,when the relative error is controlled below 1.5%,the computational efficiency is increased by 8.7 times in the case that contains 500 charged beads.In adition,for a number of 100×100×100 grids with 100 charged beads,the computational efciency can be increased up to 12 times.Our work provides new insights for the optimization of iterative algorithms in special problems.
