We demonstrate that Multi-Body Dissipative Particle Dynamics (MDPD) can be used as an efficient computational tool for the investigation of nanoscale capillary impregnation of confined geometries. As an essential prerequisite, a novel model for a solid-liquid interface in the framework of MDPD is introduced, with tunable wetting behaviour and thermal roughening to reduce artificial density-and temperature oscillations. Within this model, the impregnation dynamics of a water-like fluid into a nanoscale slit pore has been studied. Despite the coarse graining implied with the model fluid, a sufficient amount of non-equilibrium averaging can be achieved allowing for the extraction of useful information even from transient simulations, such as the dynamic apparent contact angle. Although it is found to determine the capillary driving completely, it cannot be intepreted as a simple function of the capillary number.Introduction. Over the last decade, continuous, mesoscale particle simulation methods such as Dissipative Particle Dynamics (DPD)[1, 2] and many variants thereof [3,4,5] have received considerable attention. Originally invented to include hydrodynamic effects in meso scale simulations of simple and complex fluids [1,6], it also has successfully been employed for studying polymeric systems or melts [7,8,9] or lipid membranes [10], and for colloidal suspensions [11,12]. As a model for solvents, one of the most important recent developments is the introduction of cohesive properties [13,14,15], extending the simple quadratic dependance on density in the equation of state (EOS) in early formulations [2]. The approach of Warren [15] leads to particularly stable liquidvapor interfaces and could be of great value in studying free-surface fluid dynamics problems where thermal capillary fluctuations are important, such as the intriguing phenomena related to the nanoscale Rayleigh instability (e.g. the break-up of liquid nano jets [16]). In most numerical studies involving fluid particle (FP) [17] methods, the investigations have largely been kept generic. Specific interactions, e.g., between solid-liquid interfaces, have only been accounted for in crude ways, since in the majority of cases, the FP-interactions do not arise from a systematic coarse graining procedure starting at the atomistic scale. We nevertheless suggest that a FP method can still be suitable for studying the aforementioned phenomena, provided that adhesive and interfacial properties are introduced carfully and with respect to the
