Abstract. Self-generated chaotic current fluctuations in the post-breakdown regime of a n-GaAs layer at 4.2 K have been analyzed in detail. Without an external magnetic field only regular oscillations were observed. Increasing the magnetic field strength up to 100mT generates a sequence of quasiperiodic and frequency-locking current oscillations and finally a Ruelle-Takens-Newhouse scenario with chaos. This may be understood by assuming two coupled oscillatory processes caused by dielectric relaxation and energy relaxation in the distribution of free carriers. 05.40., 72.70., 72.20J High-purity semiconductors at low temperature show highly nonlinear current-voltage characteristics. For small electric fields almost all carriers are bound to shallow impurities yielding a low conductance of the sample. At a critical field of a few volts per cm the impact ionization rate of shallow impurities exceeds the capture rate for low carrier concentration resulting in a rapid increase of the current. The steady-state properties of the transition from the low-conducting state to the high-conducting state have been analyzed in terms of nonequilibrium phase transformations 1-14]. In the course of the transition, spontaneous oscillations and chaotic current fluctuations have been observed in several semiconductor materials 1,5-17]. Different types of, and routes to, chaos were recognized and discussed in terms of nonlinear dynamics [18][19][20][21][22][23][24][25][26][27][28]. Current fluctuations in semiconductors may occur spontaneously I-5-13] or be induced by an external periodic driving force 1-13-17]. PACS:In the present paper we report on a detailed study of self-generated current fluctuations in high purity n-GaAs epitaxial layers which occur within a limited bias voltage interval in the post-breakdown regime of the material. The observed phenomena depend critically on the strength of an external magnetic field. At zero field, B = 0, only regular oscillations were found. Increasing B up to not more than 100mT causes a sequence of quasiperiodic and frequency-locking current phenomena, finally undergoing a Ruelle-TakensNewhouse scenario to chaos. This behavior may be attributed to the coupling of two oscillatory processes, in the present case dielectric relaxation and an oscillation of the nonequilibrium electron distribution. The experimental results are in excellent agreement with the predictions of the circle-map theory. The coupling strength, frequencies and amplitudes of both selfsustained processes depend strongly on the magnetic field strength.

In the computation of turbulent flows via turbulence modeling, the treatment of the convective terms is a key issue. In the present work, we present a numerical technique for simulating two-dimensional incompressible turbulent flows. In particular, the performance of the high Reynoldsκ-ɛmodel and a new high-order upwind scheme (adaptative QUICKEST by Kaibara et al. (2005)) is assessed for 2D confined and free-surface incompressible turbulent flows. The model equations are solved with the fractional-step projection method in primitive variables. Solutions are obtained by using an adaptation of the front tracking GENSMAC (Tomé and McKee (1994)) methodology for calculating fluid flows at high Reynolds numbers. The calculations are performed by using the 2D version of theFreeflowsimulation system (Castello et al. (2000)). A specific way of implementing wall functions is also tested and assessed. The numerical procedure is tested by solving three fluid flow problems, namely, turbulent flow over a backward-facing step, turbulent boundary layer over a flat plate under zero-pressure gradients, and a turbulent free jet impinging onto a flat surface. The numerical method is then applied to solve the flow of a horizontal jet penetrating a quiescent fluid from an entry port beneath the free surface.

No abstract

No abstract

No abstract

No abstract

Advances in multi-cores CPUs and in Graphics Processors Units (GPUs) are attracting a lot of attention of the scientific community due to their parallel processing power in conjunction with their low cost. In recent years the resolution of inverse thermal problems (ITP) is gaining increasing importance and attention in simulation-based applied science and engineering. However, the resolutions of these problems are very sensitive to random errors and the computer cost is high. In an attempt to improve the computational performance to solve an ITP, the computational power of multi-core architectures was used and analysed; mainly those offered by the GPU via Compute Unified Device Architecture (CUDA) and multi-cores CPUs via Pthreads. Also, we developed the implementation of the Preconditioned Conjugate Gradient method as a kernel on GPU to solve several sparse linear systems. Our CUDA and Pthreads-based systems are, respectively, two and four times faster than the serial version, while maintaining comparable convergence behaviour.

No abstract