Smail Khalfallah's research while affiliated with Ecole Militaire Polytechnique and other places

Microgas turbines are still limited in performance due to the difficulty of applying cooling systems to the radial turbines (RT). This paper proposes and investigates a cooling solution, which is a new scheme of circulating cooling air inside the RT disc. This solution can be easily manufactured or applied to existing RTs. Therefore, this solution is tested on an existing RT, starting by analyzing the temperature distribution inside the turbine blades without the cooling solution, which reveals the zones of the highest temperature where the cooling air circulation must be applied. The cooling air is supplied from the compressor through the turbine shaft. This study was carried out using both numerical heat transfer simulation and experimental tools. An experimental microgas turbine test bench using a turbocharger was built to estimate the boundary conditions and validate CFD simulations. The latter, which are based on the conjugate heat transfer (CFD_CHT), were applied to provide the temperature distributions and the flow behavior. The results show that the proposed solution can significantly reduce the disc temperature (about 60 °K) near the inlet, which is detected as the hottest zone, in contrast to the temperature of the blade tip. The analyses of the turbine performance show that the proposed solution reduces slightly the isentropic efficiency (lower than 2%) for high TITs (1100 °K).

Many types of optimization problems involving expensive functions have been handled successfully by meta-model-based optimization (MBO) methods. However, some particular difficult problems, such as multimodal functions, which require an accurate approximation even around local optima, still prove challenging for MBO. Generally, this type of problems is solved by combining or assembling meta-models. This paper proposes a new strategy of combining a global meta-model with many mid-range ones. The latter are constructed on sub-regions where the global meta-model is inaccurate. These mid-range meta-models are constructed in two phases. The first phase aims to explore accurately all the design space; whereas the second phase aims to provide more accurate solutions than those of the global meta-model on regions of interest (exploitation). The set of sub-regions can be updated adaptively until reaching the meta-model target accuracy. This two phases’ adaptive process prevents the over sampling of non-interesting regions, which allows gaining significantly in the cost. The proposed strategy is tested using the following techniques: Radial basis function (RBF) and Non-dominated Sorting Genetic Algorithm (NSGA-II). This strategy is tested both on difficult mathematical benchmarks and on an engineering application giving satisfactory results. The considered engineering application aims to optimize the operation of a vertical axe wind turbine.

Vertical axis wind turbines present many advantages compared with horizontal axis ones despite their low performance. Thus, mechanisms, which aim to improve VAWT performance, are still in continuous development and investigation. The present paper aims to contribute to this improvement by proposing a mechanism for an H-Darrieus wind turbine and aims to gain insights about this improvement by performing numerical and experimental investigations. The proposed solution is a combined scheme of two existing mechanisms, namely a variable blade-pitch mechanism and the omni-directional-guide-vanes (ODGV). Therefore, an experimental test bench with a wind tunnel was built to measure the different combinations performance. Computational fluid dynamics (CFD) simulations were used to provide local properties and flow patterns that give more insight about the flow behaviour. Among the results, we found that the proposed combined-mechanism increases the maximum power at the optimal tip speed ratio compared to the blade-pitch mechanism alone. Furthermore, this combined-mechanism with non-uniform distribution of ODGV provides better performance and increases the maximum extracted power about 35.16% as the experiments show. On the other hand, CFD simulations show that the maximum power coefficient is obtained with two configurations, the non-uniform configuration and the uniform configuration with 4 guide-vanes at 20° of tilt angle. It is found that the power production of an H-Darrieus VAWT with a variable blade pitch can be improved using a non-uniform ODGV, which can improve considerably the power extraction in the downstream turbine half.

Improvised Explosive Devices (IEDs) represent a serious threat to vehicles. These IEDs can partially or fully damage the vehicle and the onboard equipment and, threaten the lives of its crew. To protect against these devices, modifications to specialized vehicles are made in different parts. Among these latter, the present work focuses on the effects of the suspension system on the dynamic response of the vehicle. The behaviour of a McPherson suspension system of a light vehicle is numerically studied in order to optimize the dimensions of selected components. Three models based on different approaches are developed. The first is based on an Arbitrary Lagrangian Eulerian (ALE) formulation with fluid-structure coupling, which gives a more realistic representation of the blast phenomena. However, it needs significant calculation time and computational power. The second is a simplified, purely Lagrangian, model which requires less computational resources but includes several assumptions. The third model is based on a meta-modelling approach. This latter is used to optimize the dimensions of selected sub-systems of the vehicle suspension via a genetic optimization algorithm.

The understanding and accurate prediction of the flow behavior in a Kaplan turbine are important to the design work enhancing the turbine performance including the elongation of the operation life span and the improvement of turbine efficiency. The present work concerns a stationary numerical investigation of an axial hydraulic turbine of the Kaplan type in order to study the influence of the numbers of stator and rotor blades on the flow behavior related to cavitation through the commercial code ANSYS CFX. The operating conditions of this axial turbine, similar to those encountered in a previous experiment, made it possible both to validate the numerical model and to identify the nominal operating regime. In addition, variations in the numbers of stator and rotor blades have shown observations regarding the influence of the numbers of stator and rotor blades on the cavitation phenomenon.

Purpose The purpose of this paper is the development of a new turbocharger compressor is a challenging task particularly when both wider operating range and higher efficiency are required. However, the cumbersome design effort and the inherent calculus burden can be significantly reduced by using appropriate design optimization approaches as an alternative to conventional design techniques. Design/methodology/approach This paper presents an optimization-based preliminary-design (OPD) approach based on a judicious coupling between evolutionary optimization techniques and a modified one-dimensional mean-line model. Two optimization strategies are considered. The first one is mono-objective and is solved using genetic algorithms. The second one is multi-objective and it is handled using the non-dominated sorting genetic algorithm-II. The proposed approach constitutes an automatic search process to select the geometrical parameters of the compressor, ensuring the most common requirements of the preliminary-design phase, with a minimum involvement of the designer. Findings The obtained numerical results demonstrate that the proposed tool can rapidly produce nearly optimal designs as an excellent basis for further refinement in the phase by using more complex analysis methods such as computational fluid dynamics and meta-modeling. Originality/value This paper outlines a new fast OBPD approach for centrifugal compressor turbochargers. The proposal adopts an inverse design method and consists of two main phases: a formulation phase and a solution phase. The complexity of the formulated problem is reduced by using a sensitivity analysis. The solution phase requires to link, in an automatic way, three processes, namely, optimization, design and analysis.

The flow around a circular-cone is 3D, unsteady and turbulent. As reported in many research works on this type of flow, boundary layer separations, vortex structures and the effect of surface roughness are strongly related. The effect of roughness is still not completely understood and proves challenging both for numerical and experimental approaches. The present work aims to gain more insights in understanding and numerically predicting this type of flows. Particularly, it attempts to improve the accuracy of numerical predictions of the effects of roughness on the asymmetric vortex flow around a circular-cone at subsonic flow conditions. This is performed using numerical simulations, which predict the flow field details across different transversal stations along the cone. Therefore, we have analyzed the effect of many turbulence models, notably, the transition SST, using ANSYS Fluent code. Flows characteristics are highlighted and validated against the well-known theory and some experimental measurements. The findings depict that a surface roughness height about 0.32 mm has efficaciously triggered the boundary layer to change from a laminar to a fully turbulent state in both sides of the body, in addition to a significant reduction in the global lateral force. Moreover, it seems that the roughness reduces the asymmetry of the vortex in the leeward side of the cone. Testing many roughness heights has allowed us to determine the way in which it is possible to reduce the side force effectively at high angles of attack without affecting the lift of the cone cylinder.

Preliminary design of centrifugal compressors is an emphasized step, which initiates the design process. Even with the use of quick analysis methods such as one-dimensional models, preliminary design still occupies a substantial part of the total design time. Among the factors that can lengthen this time or even cause the design failure is the inappropriate selection of the design input parameters. The present paper proposes a methodology to generate optimal inputs for the preliminary design, which reduces the design time and optimizes its overall performance. This is achieved, firstly, by performing an aerothermodynamic analysis that defines the appropriate input parameters, which will be used by a preliminary design code. This analysis has allowed identifying three pilot parameters (inlet relative Mach number, work input factor, and slip factor) to guide the generation of adequate input parameters. This input parameter generator, which reduces considerably the failure rate, is then exploited efficiently in an optimization process considering the pilot parameters as decision variables. Therefore, the proposed input parameter generator is coupled with a preliminary design code and an optimization algorithm. The proposed input parameter generator was validated on four existing compressors, showing a gain of more than 90% of the design time. Mainly, the proposed optimization has created a preliminary design trade-off having the target requirements and with optimized off-design performance.