Proposal of Novel Icing Simulation Using a Hybrid Grid- and Particle-Based Method

ABSTRACTIcing on aircraft can drastically reduce aerodynamic performance and lead to serious accidents. Therefore, prediction of the accreted ice shape and area and its effects on aerodynamic performance is crucial during the design phase of an aircraft. However, numerical simulations based on conventional grid-based methods such as the finite volume method cannot accurately reproduce the complex ice shapes, which involve horn growth, feather growth, air voids, and severe surface roughness. In the present study, instead of the grid-based method, a hybrid grid- and particle-based method was newly proposed and applied to the icing problem on a NACA0012 airfoil. The explicit moving particle semi-implicit method was employed as the particle-based method due to its short computing time. The numerical simulations effectively reproduced feather-shaped ice, air voids, and surface roughness. Finally, by computing the flow around the iced airfoil, it was confirmed that flow separation around the leading edge occurred due to the ice layer, which resulted in a thicker boundary layer and wake and an increase in the drag coefficient of approximately 70% after a residence time of only 60 seconds.

Brief communication: Vehicle routing problem and UAV application in the post-earthquake scenario

In this paper we simulate unmanned aerial vehicle (UAV) recognition after a possible case of diffuse damage after a seismic event in the town of Acireale (Sicily, Italy). Given a set of sites (84 relevant buildings) and the range of the UAV, we are able to find a number of vehicles to employ and the shortest survey path. The problem of finding the shortest survey path is an operational research problem called the vehicle routing problem (VRP). which has a solution that is known to be computationally time-consuming. We used the simulated annealing (SA) heuristic, which is able to provide stable solutions in a relatively short computing time. We also examined the distribution of the cost of the solutions varying the depot on a regular grid in order to find the best area for executing the survey.

Brief communication: Vehicle routing problem and UAV application in the post-earthquake scenario

In this paper we simulate a Unmanned Aerial Vehicle's (UAV) recognition after a possible case of diffuse damage after a seismic event in the town of Acireale (Sicily, Italy). Given a set of sites (84 relevant buildings) and the range of the UAV, we are able to find the number of vehicles to employ and the shortest survey path. The problem of finding the shortest survey path is an operational research problem called Vehicle Routing Problem (VRP) whose solution is known to be computationally time-consuming. We used the Simulated Annealing (SA) heuristic that is able to provide stable solutions in relatively short computing time. We also examined the distribution of the cost of the solutions varying the depot on a regular grid in order to assess the best area where to execute the survey.

Modification and Numerical Method for the Jiles-Atherton Hysteresis Model

The Jiles-Atherton (J-A) model is a commonly used physics-based model in describing the hysteresis characteristics of ferromagnetic materials. However, citations and interpretation of this model in literature have been non-uniform. Solution methods for solving numerically this model has not been studied adequately. In this paper, through analyzing the mathematical properties of equations and the physical mechanism of energy conservation, we point out some unreasonable descriptions of this model and develop a relatively more accurate, modified J-A model together with its numerical solution method. Our method employs a fixed point method to compute anhysteretic magnetization. We obtain the susceptibility value of the anhysteretic magnetization analytically and apply the 4th order Runge-Kutta method to the solution of total magnetization. Computational errors are estimated and then precisions of the solving method in describing various materials are verified. At last, through analyzing the effects of the accelerating method, iterative error and step size on the computational errors, we optimize the numerical method to achieve the effects of high precision and short computing time. From analysis, we determine the range of best values of some key parameters for fast and accurate computation.

Rapid Predicting the Impact Behaviors of Marine Composite Laminates

There are challenges of using composite laminates in the marine engineering, i.e., composites are frequently suffering from the effects of impaction including wave impaction, ship or other objects hitting, missiles or bullets hitting and other especially conditions. It is significant to understand the impact behaviors of laminates, in this research, the impact responses of typical laminates are investigated numerically. The delamination responses among the plies and fibre and/or matrix damage responses within the plies are simulated to understand the impaction behaviours of laminates under impaction conditions. The impact damage of composite laminates in the form of intra-and/or inter-laminar cracking is modelled by using stress-based criteria for damage initiation, and fracture mechanics technique is used to capture its evolution. Interface cohesive elements are inserted between plies with appropriate mixed-mode damage laws to predict the delamination. A group of graphite fibre/epoxy laminates with impact energies of 5, 10, 15 and 20 J, respectively, are simulated with a full scale FE model and a simplified FE model respectively. Through comparing the simulation results with each other, we find out that the impact behaviors obtained in the simplified FE model is comparable to experiments with a short computing time, but the simplified model cannot represent the properties of laminate after impact.

FRACS: modelling of the dust disc of the B[e] CPD-57° 2874 from VLTI/MIDI data

The physical interpretation of spectro-interferometric data is strongly model dependent. On one hand, models involving elaborate radiative transfer solvers are in general too time consuming to perform an automatic fitting procedure and derive astrophysical quantities and their related errors. On the other hand, using simple geometrical models does not give sufficient insights into the physics of the object. We developed a numerical tool optimised for mid-infrared (mid-IR) interferometry, the Fast Ray-tracing Algorithm for Circumstellar Structures (FRACS). Thanks to the short computing time required by FRACS, best-fit parameters and uncertainties for several physical quantities were obtained, such as inner dust radius, relative flux contribution of the central source and of the dusty CSE, dust temperature profile, disc inclination.

Parameter Based Combustion Model for Large Prechamber Gas Engines

Challenging requirements for modern large engines regarding power output, fuel consumption, and emissions can only be achieved with carefully adapted combustion systems. With the improvement of simulation methods simulation work is playing a more and more important role for the engine development. Due to their simplicity and short computing time, one-dimensional and zero-dimensional calculation methods are widely applied for the engine cycle simulation and optimization. While the gas dynamic processes in the intake and exhaust systems can already be simulated with sufficient precision, it still represents a considerable difficulty to predict the combustion process exactly. In this contribution, an empirical combustion model for large prechamber gas engines is presented, which was evolved based on measurements on a single cylinder research engine using the design of experiment method. The combustion process in prechamber gas engines is investigated and reproduced successfully by means of a double-vibe function. The mathematical relationship between the engine operating parameters and the parameters of the double-vibe function was determined as a transfer model on the base of comprehensive measurements. The effects of engine operating parameters, e.g., boost pressure, charge temperature, ignition timing, and air/fuel ratio on the combustion process are taken into account in the transfer model. After adding modification functions, the model can be applied to gas engines operated with various gas fuels taking into account the actual air humidity. Comprehensive verifications were conducted on a single-cylinder engine as well as on full-scale engines. With the combination of the combustion model and a gas exchange simulation model the engine performance has been predicted satisfactorily. Due to the simple phenomenological structure of the model, a user-friendly model application and a short computing time is achieved.

Parameter Based Combustion Model for Large Pre-Chamber Gas Engines

Challenging requirements for modern large engines regarding power output, fuel consumption and emissions can only be achieved with carefully adapted combustion systems. With the improvement of simulation methods simulation work is playing a more and more important role for the engine development. Due to their simplicity and short computing time, one-dimensional and zero-dimensional calculation methods are widely applied for the engine cycle simulation and optimization. While the gas dynamic processes in the intake and exhaust system can already be simulated with sufficient precision, it still represents a considerable difficulty to predict the combustion process exactly. In this contribution, an empirical combustion model for large pre-chamber gas engines is presented, which was evolved based on measurements on a single cylinder research engine using the DOE (Design of Experiments) method. The combustion process in pre-chamber gas engines is investigated and reproduced successfully by means of a Double-Vibe function. The mathematical relationship between the engine operating parameters and the parameters of the Double-Vibe function was determined as a transfer model on the base of comprehensive measurements. The effects of engine operating parameters e.g. boost pressure, charge temperature, ignition timing, air/fuel ratio on the combustion process are taken into account in the transfer model. After adding modification functions, the model can be applied to gas engines operated with various gas fuels taking into account the actual air humidity. Comprehensive verifications were conducted on a single cylinder engine as well as on full scale engines. With the combination of the combustion model and a gas exchange simulation model the engine performance has been predicted satisfactorily. Due to the simple phenomenological structure of the model, a user-friendly model application and a short computing time is achieved.

Ab initioprotein phasing: the Patterson deconvolution method inSIR2002

An automated Patterson deconvolution technique, based on the minimum superposition function, has been implemented into theSIR2002package, a program for theab initiodetermination of small and large molecules. The procedure, combined with the direct-space phase refinement routines present in theSIR2002program, has been designed to solve proteins diffracting at atomic resolution. The procedure has been validated using a large set of test structures, containing either heavy or intermediate atoms (e.g.P, S and Cl) and with structural complexity up to 1500 non-hydrogen atoms in the asymmetric unit. The success, obtained in a completely automatic way, requires a very short computing time and indicates that Patterson deconvolution techniques are a valid alternative to current direct-methods approaches.

Enrichment of Extremely Noisy High-Throughput Screening Data Using a Naïve Bayes Classifier

The noise level of a high-throughput screening (HTS) experiment depends on various factors such as the quality and robustness of the assay itself and the quality of the robotic platform. Screening of compound mixtures is noisier than screening single compounds per well. A classification model based on naïve Bayes (NB) may be used to enrich such data. The authors studied the ability of the NB classifier to prioritize noisy primary HTS data of compound mixtures (5 compounds/well) in 4 campaigns in which the percentage of noise presumed to be inactive compounds ranged between 81% and 91%. The top 10% of the compounds suggested by the classifier captured between 26% and 45% of the active compounds. These results are reasonable and useful, considering the poor quality of the training set and the short computing time that is needed to build and deploy the classifier. ( Journal of Biomolecular Screening 2004:32-36)