Sur une maquette à l’échelle 2/5ième équipée d’un diffuseur et de pneus Michelin. La géométrie du véhicule, basée sur le modèle ASMO, a été modifiée précédemment à ce travail afin d’obtenir un angle d’attaque de l’écoulement sur les roues avant et un équilibre du sillage réaliste en présence de quatre roues tournantes. Cette configuration a servi de référence dans le cadre de cette étude.Il a été mis en évidence que la configuration de base avec un sillage équilibré peut facilement être modifiée d’un point de vue aérodynamique en changeant l’état des roues (en rotation ou pas) et le type de pneumatique, en particulier sur l’essieu arrière. Cela provient d’un effet global et d’une sensibilité importante de l’équilibre du sillage aux changements de débit au soubassement. A contrario, lorsque le sillage du véhicule se trouve déséquilibré, il devient plus robuste par rapport à des perturbations de soubassement comme un changement d’état des roues ou une modification des pneumatiques. Si l’on supprime les quatre roues ou uniquement les deux roues avant, le débit de quantité de mouvement au soubassement est grandement augmenté. Par contre, si l’on supprime le diffuseur (changement important de la géométrie du véhicule), celui-ci s’en trouve nettement réduit. Dans ces deux configurations, le sillage est très fortement déséquilibré vers le sol et devient indépendant aux modifications apportées sur les roues.Il a également été mis en évidence un effet plus local du sillage des roues sur la portance et la traînée du véhicule.En effet, la zone de dépression dans le sillage des roues avant a un effet sur la portance alors que le sillage des roues arrière pilote en partie la pression au culot et donc la traînée. Il a ainsi été observé une augmentation importante de la traînée du véhicule lorsque le sillage des roues arrière, non fermé, venait en interaction directe avec le sillage du véhicule. / The thesis aims to provide a better understanding of the wheel-vehicle interaction, via experimental investigations on a 2/5-th scale vehicle with an underbody diffuser and 2/5-th scale wheels equipped with Michelin tires. The vehicle geometry, based on ASMO model, was modified prior to the PhD work, in order to achieve a reasonable front wheel yaw angle, and a realistic wake balance with four rotating wheels. It is the baseline configuration in the scope of this work.The findings demonstrate that the well-balanced wake of the baseline configuration can be easily modified by different wheel states or tire modifications, especially at the rear axle. This results from a global effect of the underbody momentum modifications, i.e. a high wake sensitivity to the underbody flow. On the contrary, when the vehicle mean wake develops into a non-balanced topology, it is more robust towards underbody perturbations such as different wheel states or tire modifications. By eliminating four wheels or front wheels, the underbody momentum flux is vastly increased; by eliminating the underbody diffuser, which is a vehicle geometry modification, the underbody momentum flux is significantly reduced. In these two circumstances, one can observea robust downwash from the roof, independent of the wheel states or tire modifications. Besides, there is a more local effect of the wheels near wakes on the aerodynamic lift and drag of the vehicle. Low pressure regions in the underbody downstream the front wheels have an effect on vehicle lift. The rear wheels impose pressure conditions on the vehicle base, influencing the vehicle drag. Particularly, the merging of nonclosed mean wakes of the rear wheels with the vehicle wake can give rise to strong penalty in vehicle drag.

In this experiment, average calcaneal fracture force is measured across male and female groups. The purpose of this experiment is an analysis of alternatives exploring the importance of sex-based criteria in models representing injuries typical in underbody blast environments. Seventeen (17) right legs were harvested at the knee from cadavers representing three anthropometries: 50th percentile male (6), 75th percentile female (6), and 5th percentile female (5). Care was taken to preserve anatomically correct geometry as the legs were cut to equal lengths, the tibia and fibula were potted in Dyna-Cast®, flesh and ligaments were excised from the inferior surface of the calcaneus, and a small Dyna-Cast® pad was poured and sanded flat – interfacing with the exposed calcaneal surface. Each test specimen was mounted in a custom fixture and exposed once to high-rate axial loading characterized by a constant acceleration and 25.4mm intrusion, achieving an average speed of 4.7m/s (σ = 0.3m/s) in 10ms. Input acceleration was measured by an Endevco 7264c accelerometer and a Denton 2513 six-axis load cell measured reaction force proximal to the specimen. A VR Phantom v9.1 camera recorded x-ray imagery at 2k frames per second. Data were collected by a TDAS Pro data acquisition system at 20k samples per second and filtered in accordance with SAE J211. Time of fracture, established through x-ray imagery, was used to determined fracture force from the electronically synchronized load-cell data. 100% injury was recorded. Average calcaneus fracture forces were reported as follows: 5406N (σ = 780N) for 50th percentile males, 4130N (σ = 1061N) for 75th percentile females, and 2873N (σ = 1293N) for 5th percentile females. Statistical significance was established between the reported averages according to three ANOVA tests: One-way (p = 0.0054), Brown-Forsythe (p = 0.0091), and Welch's (p = 0.0156). Unpaired Student's t-test confirmed significant differences between 50th percentile male vs 75th percentile female (p = 0.0469) and 50th percentile male vs 5th percentile female (p = 0.0030); the t-test did not show significance between the two female groups (p = 0.1315). Average impulse-to-fracture was calculated for each group and found to be not statistically significant. / Master of Science / A marked shift can be found in combat wound epidemiology towards a predominance of extremity injuries sustained from explosives. The Warrior Injury Assessment Mannequin (WIAMan) Project sought to develop a baseline dataset of post-mortem human surrogate responses to realistic explosive loading and correlate it to a highly instrumented mannequin for the further development of combat vehicles and personal protective gear. The following experiment exists within the WIAMan paradigm as an analysis of alternatives exploring the adequacy of the above mentioned baseline dataset in directly representing both male and female injuries. More specifically, this experiment interrogates the differences in average fracture forces between male and female calcanei across three anthropometries: 50th percentile male, 75th percentile female, and 5th percentile female. Testing was carried out on 17 right cadaver legs cut to equal lengths, potted proximally in Dyna-Cast®, with the inferior surface of their calcanei exposed; a small Dyna-Cast® pad was poured for each calcaneus and sanded flat. Each test specimen was fixed to a Denton 2513 six-axis load cell proximally and exposed to a high-rate, constant acceleration, 25.4mm displacement aligned with the calcaneus along the long axis of the leg bones. Fracture time, established through x-ray images recorded at 2k frames per second with a VR Phantom V9.1 camera, was used to determine load cell force measurement at fracture. Average calcaneus fracture forces were reported as follows: 5406N (σ = 780N) for 50th percentile males, 4130N (σ = 1061N) for 75th percentile females, and 2873N (σ = 1293N) for 5th percentile females. Statistical significance was established between the reported averages according to three ANOVA tests: One-way (p = 0.0054), Brown-Forsythe (p = 0.0091), and Welch's (p = 0.0156). Unpaired Student's t-test confirmed significant differences between 50th percentile male vs 75th percentile female (p = 0.0469) and 50th percentile male vs 5th percentile female (p = 0.0030); the t-test did not show significance between the two female groups (p = 0.1315). Average impulse-to-fracture was calculated for each group and found to be not statistically significant.

No description available.

Aerodynamics has always been a driving force in motorsport and road vehicle design and development, and continues to play an important role. A significant advancement in race car aerodynamics was the development of the vehicle underbody to produce downforce, in particular the implementation of the diffuser. This thesis concentrates on the performance flow mechanisms found in simple rear diffusers commonly seen in motorsport applications. There is little published work on these mechanisms, especially the influence of the more commonly used multiple-channel diffusers.

Cette thèse CIFRE est le fruit d’une collaboration entre Renault Trucks et le LMFA dans la perspective d’une évaluation de la pertinence du contrôle actif pour la réduction de traînée de véhicule industriel. Les deux principaux objectifs de ces travaux expérimentaux consistent à l’analyse de l’influence de l’écoulement de soubassement sur la dynamique du sillage et l’étude d’une stratégie de contrôle d’écoulement combinant des volets déflecteurs (positionnés sur les bords supérieurs et latéraux du culot) à des actionneurs de type jets pulsés dans l’optique d’une réduction de la traînée aérodynamique d’un corps non profilé à culot droit. Le développement du sillage pour différentes vitesses de soubassement évoluant de 10% à plus de 80% de la vitesse de l’écoulement infini amont est étudié sur une maquette simplifiée de véhicule poids lourd à l’échelle 1/43e. Des mesures de pression au culot permettent d’identifier quatre classes d’écoulement associées à des structures de sillage différentes mises en évidence par des mesures PIV 2D-3C. Le sillage de la première classe d’écoulement obtenu pour les très faibles vitesses de soubassement ressemble au sillage derrière une marche 3D. Pour des vitesses de soubassement plus élevées, l’écoulement de soubassement décolle au niveau du sol et impacte soit le culot ou soit la couche de cisaillement supérieure favorisant le développement des instabilités de type Kelvin-Helmoltz dans ce dernier cas. Enfin, la dernière classe est caractérisée par un sillage comparable à celui d’un corps d’Ahmed. L’ajout de volets déflecteurs à l’arrière du culot engendre une augmentation de la pression au culot pour l’ensemble des classes d’écoulement. Cette augmentation réside principalement dans l’effet de vectorisation de l’écoulement. Un système de contrôle actif est intégré sur une maquette 1/8e géométriquement identique à celle à l’échelle 1/43e et équipée de volets déflecteurs. Deux angles de volet supérieur sont testés afin d’obtenir en moyenne (i) un écoulement naturel attaché à la paroi du volet et (ii) un écoulement naturel détaché du volet. Par rapport au cas sans contrôle actif, des gains sur la traînée sont obtenus pour une certaine gamme de fréquence d’actionnement uniquement dans le cas (ii). Ces gains sont obtenus suite au recollement de l’écoulement sur le volet. Enfin, la robustesse des gains sur la pression au culot est testée avec succès en mettant la maquette en dérapage pour simuler un vent de travers. / This PhD thesis was realized in the scope of a collaboration with Renault Trucks and the LMFA in view of an evaluation of the relevance of active flow control for the drag reduction industrial vehicle. The two main objectives of this experimental work are to analyze the impact of the underbody flow on the wake dynamics and to study a flow control strategy combining inclined flaps (located on the upper and lateral edges of the rear base) with pulsed jet actuators for reducing the aerodynamic drag of a square-back bluff body. The wake development for several underbody velocities ranging from 10% to more than 80% of the free-stream velocity is studied on a simplified truck model at scale 1 :43. Rear base pressure measurements lead to the identification of four flow classes associated with different wake structures highlighted by 2D-3C PIV measurements. The wake of the first flow class obtained for very low underbody velocities looks like that of the wake of a 3D backward facing step. For higher underbody velocities, the underbody flow is separated from the ground impaging either the rear base or the upper shear layer triggering Kelvin-Helmoltz instabilities for this last case. Finally, the fourth class is characterized by a wake comparable to that of the Ahmed body. The implementation of inclined flaps at the rear base increases the base pressure for all classes. This increase is mainly due to the vectoring effect of the flow. An active control system is integrated to a 1 :8 scale model geometrically identical to that of the 1 :43 scale model with flaps. Two upper flap angles are tested to have (i) a natural flow attached to the flap and (ii) a natural flow detached from the flap. In comparison to the case without active flow control, drag reductions are obtained only for a specific range of actuation frequencies only in case (ii). These gains are associated with the reattachment of the flow on the flap. Finally, the robustness of the pressure gains is successfully tested in crosswind conditions.