Rupert W. Nash

Run-and-Tumble Particles with Hydrodynamics: Sedimentation, Trapping, and Upstream Swimming

We simulate by lattice Boltzmann the nonequilibrium steady states of run-and-tumble particles (inspired by a minimal model of bacteria), interacting by far-field hydrodynamics, subject to confinement. Under gravity, hydrodynamic interactions barely perturb the steady state found without them, but for particles in a harmonic trap such a state is quite changed if the run length is larger than the confinement length: a self-assembled pump is formed. Particles likewise confined in a narrow channel show a generic upstream flux in Poiseuille flow: chiral swimming is not required.

Choice of boundary condition for lattice-Boltzmann simulation of moderate-Reynolds-number flow in complex domains

Modeling blood flow in larger vessels using lattice-Boltzmann methods comes with a challenging set of constraints: a complex geometry with walls and inlets and outlets at arbitrary orientations with respect to the lattice, intermediate Reynolds (Re) number, and unsteady flow. Simple bounce-back is one of the most commonly used, simplest, and most computationally efficient boundary conditions, but many others have been proposed. We implement three other methods applicable to complex geometries [Guo, Zheng, and Shi, Phys. Fluids 14, 2007 (2002); Bouzidi, Firdaouss, and Lallemand, Phys. Fluids 13, 3452 (2001); Junk and Yang, Phys. Rev. E 72, 066701 (2005)] in our open-source application hemelb. We use these to simulate Poiseuille and Womersley flows in a cylindrical pipe with an arbitrary orientation at physiologically relevant Re number (1-300) and Womersley (4-12) numbers and steady flow in a curved pipe at relevant Dean number (100-200) and compare the accuracy to analytical solutions. We find that both the Bouzidi-Firdaouss-Lallemand (BFL) and Guo-Zheng-Shi (GZS) methods give second-order convergence in space while simple bounce-back degrades to first order. The BFL method appears to perform better than GZS in unsteady flows and is significantly less computationally expensive. The Junk-Yang method shows poor stability at larger Re number and so cannot be recommended here. The choice of collision operator (lattice Bhatnagar-Gross-Krook vs multiple relaxation time) and velocity set (D3Q15 vs D3Q19 vs D3Q27) does not significantly affect the accuracy in the problems studied.

Flexible composition and execution of high performance, high fidelity multiscale biomedical simulations.

Multiscale simulations are essential in the biomedical domain to accurately model human physiology. We present a modular approach for designing, constructing and executing multiscale simulations on a wide range of resources, from laptops to petascale supercomputers, including combinations of these. Our work features two multiscale applications, in-stent restenosis and cerebrovascular bloodflow, which combine multiple existing single-scale applications to create a multiscale simulation. These applications can be efficiently coupled, deployed and executed on computers up to the largest (peta) scale, incurring a coupling overhead of 1–10% of the total execution time.

Analysing and modelling the performance of the HemeLB lattice-Boltzmann simulation environment

We investigate the performance of the HemeLB lattice-Boltzmann simulator for cerebrovascular blood flow, aimed at providing timely and clinically relevant assistance to neurosurgeons. HemeLB is optimised for sparse geometries, supports interactive use, and scales well to 32,768 cores for problems with ∼81 million lattice sites. We obtain a maximum performance of 29.5 billion site updates per second, with only an 11% slowdown for highly sparse problems (5% fluid fraction). We present steering and visualisation performance measurements and provide a model which allows users to predict the performance, thereby determining how to run simulations with maximum accuracy within time constraints.

Computer simulations reveal complex distribution of haemodynamic forces in a mouse retina model of angiogenesis

There is currently limited understanding of the role played by haemodynamic forces on the processes governing vascular development. One of many obstacles to be overcome is being able to measure those forces, at the required resolution level, on vessels only a few micrometres thick. In this paper, we present an in silico method for the computation of the haemodynamic forces experienced by murine retinal vasculature (a widely used vascular development animal model) beyond what is measurable experimentally. Our results show that it is possible to reconstruct high-resolution three-dimensional geometrical models directly from samples of retinal vasculature and that the lattice-Boltzmann algorithm can be used to obtain accurate estimates of the haemodynamics in these domains. We generate flow models from samples obtained at postnatal days (P) 5 and 6. Our simulations show important differences between the flow patterns recovered in both cases, including observations of regression occurring in areas where wall shear stress (WSS) gradients exist. We propose two possible mechanisms to account for the observed increase in velocity and WSS between P5 and P6: (i) the measured reduction in typical vessel diameter between both time points and (ii) the reduction in network density triggered by the pruning process. The methodology developed herein is applicable to other biomedical domains where microvasculature can be imaged but experimental flow measurements are unavailable or difficult to obtain.

Singular forces and pointlike colloids in lattice Boltzmann hydrodynamics

We present an accurate method to include arbitrary singular distributions of forces in the lattice Boltzmann formulation of hydrodynamics. We validate our method with several examples involving Stokeslet, stresslet, and rotlet singularities, finding excellent agreement with analytical results. A minimal model for sedimenting particles is presented using the method. In the dilute limit, this model has accuracy comparable to, but computational efficiency much greater than, algorithms that explicitly resolve the size of the particles.

Impact of blood rheology on wall shear stress in a model of the middle cerebral artery

Perturbations to the homeostatic distribution of mechanical forces exerted by blood on the endothelial layer have been correlated with vascular pathologies, including intracranial aneurysms and atherosclerosis. Recent computational work suggests that, in order to correctly characterize such forces, the shear-thinning properties of blood must be taken into account. To the best of our knowledge, these findings have never been compared against experimentally observed pathological thresholds. In this work, we apply the three-band diagram (TBD) analysis due to Gizzi et al. (Gizzi et al. 2011 Three-band decomposition analysis of wall shear stress in pulsatile flows. Phys. Rev. E 83, 031902. (doi:10.1103/PhysRevE.83.031902)) to assess the impact of the choice of blood rheology model on a computational model of the right middle cerebral artery. Our results show that, in the model under study, the differences between the wall shear stress predicted by a Newtonian model and the well-known Carreau–Yasuda generalized Newtonian model are only significant if the vascular pathology under study is associated with a pathological threshold in the range 0.94–1.56 Pa, where the results of the TBD analysis of the rheology models considered differs. Otherwise, we observe no significant differences.

Role of Correlations in the Collective Behavior of Microswimmer Suspensions

In this Letter, we study the collective behavior of a large number of self-propelled microswimmers immersed in a fluid. Using unprecedentedly large-scale lattice Boltzmann simulations, we reproduce the transition to bacterial turbulence. We show that, even well below the transition, swimmers move in a correlated fashion that cannot be described by a mean-field approach. We develop a novel kinetic theory that captures these correlations and is nonperturbative in the swimmer density. To provide an experimentally accessible measure of correlations, we calculate the diffusivity of passive tracers and reveal its nontrivial density dependence. The theory is in quantitative agreement with the lattice Boltzmann simulations and captures the asymmetry between pusher and puller swimmers below the transition to turbulence.

Preparing scientific application software for exascale computing

Many of the most widely used scientifc application software of today were developed largely during a time when the typical amount of compute cores was calculated in tens or hundreds. Within a not too distant future the number of cores will be calculated in at least hundreds of thousands or even millions. A European collaboration group CRESTA has recently been working on a set of renowned scientific software to investigate and develop these codes towards the realm of exascale computing. The codes are ELMFIRE, GROMACS, IFS, HemeLB, NEK5000, and OpenFOAM. This paper contains a summary of the strategies for their development towards exascale and results achieved during the first year of the collaboration project.

Enabling In Situ Pre- and Post-processing for Exascale Hemodynamic Simulations - A Co-design Study with the Sparse Geometry Lattice-Boltzmann Code HemeLB

Todays fluid simulations deal with complex geometries and numerical data on an extreme scale. As computation approaches the exascale, it will no longer be possible to write and store the full-sized data set. In situ data analysis and scientific visualisation provide feasible solutions to the analysis of complex large scaled CFD simulations. To bring pre- and postprocessing to the exascale we must consider modifications to data structure and memory layout, and address latency and error resiliency. In this respect, a particular challenge is the exascale data processing for the sparse geometry lattice-Boltzmann code HemeLB, intended for hemodynamic simulations. In this paper, we assess the needs and challenges of HemeLB users and sketch a co-design infrastructure and system architecture for pre- and post-processing the simulation data. To enable in situ data visualisation and analysis during a running simulation, post-processing needs to work on a reduced subset of the original data. Particular choices of data structure and visualisation techniques need to be co-designed with the application scientists in order to achieve efficient and interactive data processing and analysis. In this work, we focus on the hierarchical data structure and suitable visualisation techniques which provide possible solutions to interactive in situ data processing at exascale. Architectural challenges and road-maps will be presented as the major focus of this paper. We sketch a software architecture which integrates pre- and post-processing techniques that can provide in situ analysis and ultimately computational steering to HemeLB.