Connecting Knowledge Compilation Classes Width Parameters

Theory of Computing Systems - The field of knowledge compilation establishes the tractability of many tasks by studying how to compile them to Boolean circuit classes obeying some requirements such...     

Experimental study of the start-up of a fuel cell stack for backup power application

The transient response of proton exchange membrane fuel cells during start-up is an important issue for backup power systems which require a very short start-up time in order to limit the use of batteries during a blackout. The start-up procedure of a ten cells stack was studied: in the first stage the cathode channel initially filled with nitrogen was supplied with oxygen in open circuit then in the second stage it was connected to the load. The influences of the current time-profile (step or ramp), the cell voltage at the connection and the gas flow rates on the voltage variation were investigated. It was found that the voltage value during the filling of the cathode is not sufficient to determine which fraction of the cathode was filled with oxygen. In most cases, high oxygen flow rates allow reducing the start-up time of the stack. Furthermore, for fixed current density and stoichiometric coefficients it was found that a minimum start-up time exits. The analysis of transient response to current steps showed that around 70% of the maximum electrical power was available less than 2 s after the beginning of the start-up procedure.     

Finite element simulation of interfacial segregation in dilute alloys

The finite element software Comsol is used to simulate surface or grain boundary segregation in dilute alloys. The model computes simultaneously the evolution of interfacial concentration and diffusion in the bulk. The solute exchange between bulk and interface is governed by Darken’s equation. The model is able to reproduce thermodynamic and kinetic aspects of the phenomenon, in particular the saturation segregation level and the short-time segregation kinetics expressed by the McLean approximation. It is also able to reproduce experimental trends in the case of surface segregation of sulphur in a Ni superalloy. In the case of the grain boundary segregation of impurities (P or S) in engineering alloys, the present approach provides a practical tool, as it can be coupled to other finite element simulations (heat transfer and/or mechanics). Thus, it becomes possible to predict the risk of synergetic segregation and thermomechanical damage during service or processing (forging, welding,...).     

Behaviour of an immersed boundary method in unsteady flows over sharp-edged bodies

The behaviour of the immersed boundary method proposed by Goldstein et al. [Goldstein D, Handler R, Sirovich L. Modelling a no-slip boundary condition with an external force field. J Comput Phys 1993;105:354-66] as a second-order damped control system is investigated. The natural frequency and the damping coefficient are introduced as driving parameters of the method. The comparison between the velocity response at forced points in the startup flow over a square cylinder with the theoretical response of a second-order damped oscillator is performed. The role of each parameter appears clearly. At the beginning of the startup flow, the response time depends directly on the natural frequency, and this parameter determines the level of residual velocities achieved in an unsteady flow. The damping coefficient drives the oscillation of the velocity response at the beginning of the startup flow, but has negligible influence during the establishment and in the unsteady flow. At forced points facing no unsteady perturbation from the flow, the zero-velocity set point is reached asymptotically, as usual in second-order damped-systems. Through the simulation of the flow over a blunt flat plat at Re=1000, it is observed that the initial thickness of the mixing layer due to the separation at the edge may vary during the simulation because the sharpness of the edge increases as the residual velocities decrease. This insight gained on the behaviour of the response allows a time-step optimisation, which, completed with comparisons to reference literature results, confirms the feedback forcing method a competitive tool for accessing near-wall unsteady flow over sharp-edged bodies.     

FWT2D: A massively parallel program for frequency-domain full-waveform tomography of wide-aperture seismic data—Part 1: Algorithm

This is the first paper in a two-part series that describes a massively parallel code that performs 2D frequency-domain full-waveform inversion of wide-aperture seismic data for imaging complex structures. Full-waveform inversion methods, namely quantitative seismic imaging methods based on the resolution of the full wave equation, are computationally expensive. Therefore, designing efficient algorithms which take advantage of parallel computing facilities is critical for the appraisal of these approaches when applied to representative case studies and for further improvements. Full-waveform modelling requires the resolution of a large sparse system of linear equations which is performed with the massively parallel direct solver MUMPS for efficient multiple-shot simulations. Efficiency of the multiple-shot solution phase (forward/backward substitutions) is improved by using the BLAS3 library. The inverse problem relies on a classic local optimization approach implemented with a gradient method. The direct solver returns the multiple-shot wavefield solutions distributed over the processors according to a domain decomposition driven by the distribution of the LU factors. The domain decomposition of the wavefield solutions is used to compute in parallel the gradient of the objective function and the diagonal Hessian, this latter providing a suitable scaling of the gradient. The algorithm allows one to test different strategies for multiscale frequency inversion ranging from successive mono-frequency inversion to simultaneous multifrequency inversion. These different inversion strategies will be illustrated in the following companion paper. The parallel efficiency and the scalability of the code will also be quantified.     

High-performance finite-element simulations of seismic wave propagation in three-dimensional nonlinear inelastic geological media

We present finite-element numerical simulations of seismic wave propagation in non linear inelastic geological media. We demonstrate the feasibility of large-scale modeling based on an implicit numerical scheme and a nonlinear constitutive model. We illustrate our methodology with an application to regional scale modeling in the French Riviera, which is prone to earthquakes. The PaStiX direct solver is used to handle large matrix numerical factorizations based on hybrid parallelism to reduce memory overhead. A specific methodology is introduced for the parallel assembly in the context of soil nonlinearity. We analyse the scaling of the parallel algorithms on large-scale configurations and we discuss the physical results.     

Tree automata with equality constraints modulo equational theories

This paper presents new classes of tree automata combining automata with equality test and automata modulo equational theories. We believe that these classes have a good potential for application in e.g. software verification. These tree automata are obtained by extending the standard Horn clause representations with equational conditions and rewrite systems. We show in particular that a generalized membership problem (extending the emptiness problem) is decidable by proving that the saturation of tree automata presentations with suitable paramodulation strategies terminates. Alternatively our results can be viewed as new decidable classes of first-order formula.     

Combinatorial mesh optimization

A new mesh optimization framework for 3D triangular surface meshes is presented, which formulates the task as an energy minimization problem in the same spirit as in Hoppe et al. (SIGGRAPH’93: Proceedings of the 20th Annual Conference on Computer Graphics and Interactive Techniques, 1993). The desired mesh properties are controlled through a global energy function including data attached terms measuring the fidelity to the original mesh, shape potentials favoring high quality triangles, and connectivity as well as budget terms controlling the sampling density. The optimization algorithm modifies mesh connectivity as well as the vertex positions. Solutions for the vertex repositioning step are obtained by a discrete graph cut algorithm examining global combinations of local candidates.     

Rigid tree automata and applications

We introduce the class of rigid tree automata (RTA), an extension of standard bottom-up automata on ranked trees with distinguished states called rigid. Rigid states define a restriction on the computation of RTA on trees: RTA can test for equality in subtrees reaching the same rigid state. RTA are able to perform local and global tests of equality between subtrees, non-linear tree pattern matching, and some inequality and disequality tests as well. Properties like determinism, pumping lemma, Boolean closure, and several decision problems are studied in detail. In particular, the emptiness problem is shown decidable in linear time for RTA whereas membership of a given tree to the language of a given RTA is NP-complete. Our main result is the decidability of whether a given tree belongs to the rewrite closure of an RTA language under a restricted family of term rewriting systems, whereas this closure is not an RTA language. This result, one of the first on rewrite closure of languages of tree automata with constraints, is enabling the extension of model checking procedures based on finite tree automata techniques, in particular for the verification of communicating processes with several local non-rewritable memories, like security protocols. Finally, a comparison of RTA with several classes of tree automata with local and global equality tests, with dag automata and Horn clause formalisms is also provided.