A Multiscale Vision-Illustrative Applications From Biology to Engineering

Abstract

Modeling and simulation have quickly become equivalent pillars of research along with traditional theory and experimentation. The growing realization that most complex phenomena of interest span many orders of spatial and temporal scales has led to an exponential rise in the development and application of multiscale modeling and simulation over the past two decades. In this perspective, the associate editors of the International Journal for Multiscale Computational Engineering and their co-workers illustrate current applications in their respective fields spanning biomolecular structure and dynamics, civil engineering and materials science, computational mechanics, aerospace and mechanical engineering, and more. Such applications are highly tailored, exploit the latest and ever-evolving advances in both computer hardware and software, and contribute significantly to science, technology, and medical challenges in the 21st century.

Multiscale cosimulation design template for neuroscience applications

“…Figure 2c displays the polytrodes with the 32 recording sites. The simulated signal from the ECOG sensor is computing using the model of a point dipole in a homogeneous space as described by Sanz-Leon et al 2015[8] (see Online Methods) and the hybridLFPy[4] software is used for computing the signal from the recording site of the implanted probes (see Online Methods). The latter software uses morphology and spatial position of neurons to generate the underlying local field potential (LFP) for given spike trains of point neurons.…”

“…It is therefore important to bridge different scales. In material science, the study of composite materials requires the description of molecular interactions of individual composites, and a global description for the analysis of the subsequent deformation of the composite plate [4]. In biology, to understand the effect of drugs on tumour growth, it is necessary to model the tissue of cells around the tumour, the tumour cells, and the subcellular transduction signalling pathways [5, 6].…”

Data‐physics driven reduced order homogenization

“…Among popular physics‐based (or model‐based) reduced order multiscale methods are the Voronoi cell method, 14,15 the mesh‐free reproducing kernel particle method, 16 methods of cells, 17,18 the wavelet based reduced order method, 19,20 the reduced order homogenization methods, 21–25 and the nonuniform transformation field methods 26,27 . We refer to References 28–30 for recent review articles on multiscale methods with and without model reduction and their practical applications. For integrated reduced order multiscale methods applied to coupled process‐product design cycle we refer to References 31–35.…”

Inverse modeling of cancellous bone using artificial neural networks

“…In this work, we focus on the evaluation of simulation results and create neural networks to solve the inverse task of detecting the bone structure, which has been used to perform the simulation. We use our model of a two-phase bone material to simulate the application of ultrasound to a bone [19][20][21][22]. Oftentimes, in modeling only mechanical effects were considered.…”

“…In a healthy bone, its fraction is approximately 30%, but it may be reduced up to 5% in a degenerated bone [25, 26]. We showed in the numerical simulations, that the magnetic field strength is reduced for RVEs with lower volume fraction of cortical bone [19, 20, 22]. In this contribution, we randomize the used RVEs along the length of the bone model and calculate the magnetic field strength for different timesteps, which is the input of the neural network model.…”