Integration schemes for dissipative particle dynamics simulations: From softly interacting systems towards hybrid models

We examine the performance of various commonly used integration schemes in dissipative particle dynamics simulations. We consider this issue using three different model systems, which characterize a variety of different conditions often studied in simulations. Specifically, we clarify the performance of integration schemes in hybrid models, which combine microscopic and mesoscale descriptions of different particles using both soft and hard interactions. We find that in all three model systems many commonly used integrators may give rise to surprisingly pronounced artifacts in physical observables such as the radial distribution function, the compressibility, and the tracer diffusion coefficient. The artifacts are found to be strongest in systems, where interparticle interactions are soft and predominated by random and dissipative forces, while in systems governed by conservative interactions the artifacts are weaker. Our results suggest that the quality of any integration scheme employed is crucial in all cases where the role of random and dissipative forces is important, including hybrid models where the solvent is described in terms of soft potentials. Regarding the integration schemes, the best overall performance is found for integrators in which the velocity dependence of dissipative forces is taken into account, and particularly good performance is found for an approach in which velocities and dissipative forces are determined self-consistently. Remaining temperature deviations from the desired limit can be corrected by carrying out the self-consistent integration in conjunction with an auxiliary thermostat, in a manner that is similar in spirit to the well-known Nose-Hoover thermostat. Further, we show that conservative interactions can play a significant role in describing the transport properties of simple fluids, in contrast to approximations often made in deriving analytical theories. In general, our results illustrate the main problems associated with simulation methods in which dissipative forces are velocity dependent, and point to the need to develop new techniques to resolve these issues. (C) 2002 American Institute of Physics.PT: J; CR: AHLRICHS P, 1999, J CHEM PHYS, V111, P8225 ALLEN MP, 1993, COMPUTER SIMULATION BESOLD G, 2000, PHYS REV E A, V62, R7611 BOON JP, 1980, MOL HYDRODYNAMICS DENOTTER WK, 2000, INT J MOD PHYS C, V11, P1179 DENOTTER WK, 2001, EUROPHYS LETT, V53, P426 DUNWEG B, 1991, INT J MOD PHYS C, V2, P817 DZWINEL W, 2000, INT J MOD PHYS C, V11, P1 DZWINEL W, 2000, J COLLOID INTERF SCI, V225, P179 ESPANOL P, 1995, EUROPHYS LETT, V30, P191 ESPANOL P, 1999, PHYS REV E, V59, P6340 ESPANOL P, 1999, PHYS REV LETT, V83, P4542 FLEKKOY EG, 1999, PHYS REV LETT, V83, P1775 FLEKKOY EG, 2000, PHYS REV E A, V62, P2140 FORREST BM, 1995, J CHEM PHYS, V102, P7256 FRENKEL D, 1996, UNDERSTANDING MOL SI GARDINER CW, 1983, HDB STOCHASTIC METHO GIBSON JB, 1999, INT J MOD PHYS C, V10, P241 GROOT RD, 1997, J CHEM PHYS, V107, P4423 GROOT RD, 1999, J CHEM PHYS, V110, P9739 HANSEN JP, 1986, THEORY SIMPLE LIQUID HOOGERBRUGGE PJ, 1992, EUROPHYS LETT, V19, P155 LADD AJC, 1993, PHYS REV LETT, V70, P1339 LADD AJC, 1994, J FLUID MECH, V271, P285 LADD AJC, 1994, J FLUID MECH, V271, P311 LECUYER P, 1988, COMMUN ACM, V31, P742 MALEVANETS A, 1999, J CHEM PHYS, V110, P8605 MALEVANETS A, 2000, EUROPHYS LETT, V52, P231 MALEVANETS A, 2000, J CHEM PHYS, V112, P7260 MARSH CA, 1997, EUROPHYS LETT, V38, P411 MARSH CA, 1997, PHYS REV E, V56, P1676 MARTYNA GJ, 1995, J CHEM PHYS, V102, P8071 MASTERS AJ, 1999, EUROPHYS LETT, V48, P1 MURAT M, 1998, J CHEM PHYS, V108, P4340 NIKUNEN P, UNPUB NOVIK KE, 1998, J CHEM PHYS, V109, P7667 PAGONABARRAGA I, 1998, EUROPHYS LETT, V42, P377 PRESS WH, 1972, NUMERICAL RECIPES FO, P271 SERRANO M, 2001, PHYS REV E 2, V64 SHELLEY JC, 2000, CURR OPIN COLLOID IN, V5, P101 SODDEMANN T, 2001, EUR PHYS J E, V6, P409 SPENLEY NA, 2000, EUROPHYS LETT, V49, P534 THIJSSEN JM, 1999, COMPUTATIONAL PHYSIC TUCKERMAN ME, 2000, J PHYS CHEM B, V104, P159 VATTULAINEN I, UNPUB VENTUROLI M, 1999, PHYS CHEM COMM, V10 VERLET L, 1967, PHYS REV, V159, P98 WARREN PB, 1998, CURR OPIN COLLOID IN, V3, P620; NR: 48; TC: 37; J9: J CHEM PHYS; PG: 13; GA: 525TLSource type: Electronic(1