The aim of suspension systems in vehicles is to provide the best balance between ride and handling depending on the operating conditions of a vehicle. Active suspensions are far more effective over a variety of different road conditions compared to passive suspension systems. This is because of their ability to store and dissipate energy at different rates. Additionally, they can even provide energy of their own into the rest of the system. This makes active suspension systems an important topic of research in suspension systems. The biggest benefit of having an active suspension system is to be able to provide energy into the system that can minimize the response of the sprung mass. This is done using actuators. Actuator design in vehicle suspension system is an important research topic and a lot of work has been done in the field but little work has been done to estimate the peak control force and bandwidth required to minimize the response of the sprung mass. These two are very important requirements for actuator design in active suspensions. The aim of this study is estimate the peak control force and bandwidth to minimize the acceleration of the sprung mass of a vehicle while it is moving on a compliant surface. This makes the road surface a bi-lateral boundary and hence, the total system is a combination of the vehicle and the compliant road. Generalized vehicle and compliant road models are created so that parameters can be easily changed for different types of vehicles and different road conditions. The peak control force is estimated using adaptive filtering. A least mean squares (LMS) algorithm is used in the process. A case study with fixed parameters is used to show the results of the estimation process. The results show the effectiveness of an adaptive LMS algorithm for such an application. The peak control force and the bandwidth that are obtained from this process can then be used in actuator design. / Master of Science / Active suspension systems have been proven to be a better option compared to passive suspension systems for a wide variety of operating conditions. Active suspensions typically have an actuator system that produces a force which can reduce the disturbance caused by road inputs in the suspension. The sprung mass of a vehicle is the mass of the body and other components supported by the suspension system and the un-sprung mass is the total mass of the components which are not supported by the suspension or are part of the suspension system. The actuator is typically between the sprung mass and the un-sprung mass. When there is a single event disturbance from the road, the energy is transferred to the sprung mass, which contains the occupants, through the un-sprung mass. The actuator produces a force that reduces this acceleration in the sprung mass and hence improves ride comfort for the occupants of the vehicle. In this thesis, the single event disturbance that has been considered is a compliant road surface. This is a bi-lateral boundary since the vehicle interacts with the compliant elements under the surface of the ground. The aim of this thesis is to develop and implement a method to estimate the peak control force and bandwidth that the actuator needs to produce to eliminate or reduce the acceleration of the sprung mass which is caused by the compliant surface single event disturbance.

Most commercial vehicles currently on the market are still equipped with a passive suspension system, while some luxury brands may already use an adaptive suspension. Active suspension systems on the other hand are rarely found, however, they offer great opportunities to close the gap of the well-known trade-off between ride comfort and handling. Besides that, they can also be used to mitigate single event disturbances, an objective of the USA army as announced in a solicitation which initiated and motivated this research. In addition to that, several studies were found stating the impact and danger of potholes and their impact on the vehicle and passenger. Reviewing the literature, several control strategies for controlling active suspension systems were found. However, most of these approaches used feedback control and did not try to mitigate single event disturbances. Since literature also suggested making use of look ahead preview, research at the Performance Engineering Research Lab at Virginia Tech was started in 2015 combining look ahead preview and an adaptive system to generate optimal force profiles. This introductory research succeeded and proved the used approach to be very promising. However, the used adaptive system was not designed to operate in real-time and did not show any correlation between different road profiles. Therefore, the main objective of this research project is to evaluate and analyze each of the adaptive systems by searching for correlations in their solutions. The results then should be used in order to design a control law which emulates the adaptive system and can be used in a real-time environment. First, an overall research methodology was derived. According to this a software application was developed which extracts ideal force profiles from single event disturbance signals in order to mitigate their impact to the vehicle. The application uses a quarter car model with a partially loaded active suspension system, a set of predefined road profiles, a road profile preprocessor, and an adaptive algorithm. The preprocessing includes geometric filtering using a Tandem-Cam Model and the adaptive processor used an iterative version of the Filtered-X Last-Mean-Square algorithm. During evaluation and analysis of several generated data sets, high correlations in the generated and adjusted adaptive systems were discovered. From these an empirical and theoretical universal filter model was derived, which was then used to design an open-loop control law named Optimal Force Control. The original control law and an adjusted version designed for a real-time environment were tested for all predefined road profiles over all considered vehicle velocities and prove to perform much better than the offline solution using the adaptive system. In summary, a control law named Optimal Force Control was designed which can be used and implemented in a vehicle to extract an analytical and ideal force profile given a road profile input. Implementing an active suspension system with tracking controller, this approach can be used in order to mitigate single event disturbance signals by reducing the vertical vehicle acceleration. / Master of Science / Most commercial vehicles currently on the market are still equipped with a suspension system consisting of springs and shock absorbers (passive suspensions), while some luxury brands already use suspension systems including parts which can change their behavior based on the driving situation (active suspensions). While these active suspension systems are still rarely found, they offer great opportunities to make the vehicle stable and at the same time easy to handle. Also, they have the potential to reduce the risk of an accident while driving over a pothole or disturbance in the road, an objective of the USA Army as announced in a solicitation which initiated and motivated this research. Reviewing the literature, several control strategies for controlling active suspension systems were found. However, most of these approaches required measuring the current state of the suspension system. Research at the Performance Engineering Research Lab at Virginia Tech was started in 2015 in order to control active suspension systems by using data of the road profile ahead of the vehicle. This introductory research succeeded and proved the approach used to be very promising. However, the used system was designed to work in a laboratory environment only. Therefore, the main objective of this research project was to evaluate and analyze the used control strategy by searching for intersections and similarities in the different solutions. The results were then used to design a control strategy which can be applied in a real-world vehicle environment. First, an overall research methodology was derived. According to this methodology a software application was developed that generates the ideal control signal for the active suspension system in order to reduce the impact of a disturbance in the road profile. To that end a set of predefined road profiles were used, and a computer algorithm called Filtered-X Last-Mean-Square algorithm calculated the ideal control signal for the active suspension system. During the evaluation and analysis of several generated data sets, a lot of intersections and similarities were discovered. Based on these findings a new control strategy was designed in order to be implemented into a real-world vehicle environment. The new control strategy for the real-world vehicle environment was tested for all predefined road profiles over all considered vehicle velocities and proved to outperform the control strategy for the laboratory environment. In summary, a new control strategy named Optimal Force Control was designed, which can be used and implemented in a vehicle. The implementation of an active suspension system can be used to mitigate disturbances in the road by reducing the vertical vehicle acceleration.

This research will compare the transient response characteristics of a four-degree-of-freedom, roll-plane model, representing a class 8 truck, using passive and semiactive dampers. The semiactive damper control policies that are examined include the previously developed policies of on-off skyhook, continuous skyhook, and on-off groundhook control, along with a newly developed method of fuzzy logic semiactive control. The model input will include body forces and torques, as well as transient displacements at the tires. The model outputs include the vehicle body heave and roll displacements, the vertical displacement of the tire (wheel hop) and the vertical acceleration of the vehicle body. For each output, the maximum peak-to-peak and RMS values of the response are examined. The results of the study show that semiactive dampers have minimal effect on improving the vehicle body and tire transients due to forces or torques applied to the body, as compared to passive dampers. For road inputs, however, semiactive dampers are able to provide a more favorable compromise between the body and axle transient dynamics, when compared to passive dampers. The fuzzy logic semiactive control policy that is proposed in this research is better able to balance the body and axle dynamics than the conventional semiactive damping control policies that are investigated. Further research on the application of fuzzy logic semiactive control concepts is suggested, in order to fully investigate the potential of such control schemes for vehicle suspensions. / Master of Science

A digital controller was designed for a semiactive primary suspension for a class 8 highway truck. The controller used a skyhook policy (where the semiactive damper simulates a damper between the sprung mass and an inertial reference) to control magneto-rheological dampers placed on the truck 's primary suspension in response to measurements made by accelerometers placed on the axle and the truck frame. The completed system was then tested for both random noise (on highway driving) and impulse (speed bump) response. The test results showed that for the damping tuning and controller arrangements used in this study, semiactive dampers do not offer any significant benefits in reducing overall vibration levels at the truck frame or axles. The semiactive dampers, however, provided better control of the dynamic transients, such as roll and pitch induced by hitting speed bumps, as compared to passive dampers. Further assessment of the magneto-rheological damper's tuning and the skyhook control policy is needed to establish any definitive conclusions on the potential benefits of semiactive magneto-rheological suspensions for heavy trucks. / Master of Science

This research examines different techniques and proposes a Computational Fluid Dynamics (CFD) model as a robust tool for flow characterization of load leveling valves. The load leveling valve is a critical component of an air suspension system since it manages air spring pressure, a key function that directly impacts vehicle dynamic performance in addition to maintaining a static ride height. Efficiency of operation of a load leveling valve is established by its flow characteristics, a metric useful in determining suitability of the valve for application in a truck-suspension configuration and for comparison among similar products. The disk-slot type load leveling valve was chosen as the subject of this study due to its popularity in the heavy truck industry. Three distinct methods are presented to model and evaluate flow characteristics of a disk-slot valve. First is a theoretical formulation based on gas dynamic behavior through an orifice; second is an experimental technique in which a full pneumatic apparatus is used to collect instantaneous pressure data to estimate air discharge; and third is a CFD approach. Significant discrepancies observed between theoretically estimated results and experimental data suggest that the theoretical model is incapable of accurately capturing losses that occur during air flow. These variations diminish as the magnitude of discharge coefficient is altered. A detailed CFD model is submitted as an effective tool for load leveling valve flow characterization/analysis. This model overcomes the deficiencies of the theoretical model and improves the accuracy of simulations. A 2-D axisymmetric approximation of the real fluid domain is analyzed for flow characteristics using a Realizable k-ϵ turbulence model, scalable wall functions, and a pressure-based coupled algorithm with a second order discretization function. The CFD-generated results were observed to be in agreement with the experimental findings. CFD is found to be advantageous in the evaluation of flow characteristics as it furnishes precise data without the need to experimentally evaluate a physical model/prototype of the valve, thereby benefitting suspension engineers involved in the development and testing of load leveling valve designs. This document concludes with a sample case study which uses CFD to characterize flow in a modified disk-slot load leveling valve, and discusses the results in light of application on a heavy truck. / MS / A majority of heavy trucks in North America equipped with air suspensions use a device known as a load leveling valve. This is a mechanical control system which manages pressure in air springs to maintain a preset/constant static ride height irrespective of the payload, doing so by sensing the distance between the truck frame and the axle. The rate of airflow to/from air springs in response to a road disturbance or load shift is critical to the stability of the truck when on the road. This rate of airflow for a given set of conditions constitutes flow characteristics of a load leveling valve. Accurate measurement of flow characteristics is necessary to understand the actual effect of the use of a particular valve on a truck-suspension configuration. This research addresses that requirement by presenting three distinct methods to model and evaluate flow characteristics of a load leveling valve, conducted on the disk-slot valve for its popularity in the heavy truck industry. First is a theoretical formulation based on flow of gas through an orifice; second is an experimental technique in which a full pneumatic apparatus is used to collect instantaneous pressure data to estimate air discharge; and third is a Computational Fluid Dynamics (CFD) approach. Significant discrepancies observed between theoretically estimated results and experimental data suggest that the theoretical model is incapable of accurately capturing losses that occur during air flow. The disparities also justify the adoption of CFD as an alternate method. A comprehensive CFD model is proposed as a capable tool for load leveling valve flow analysis/characterization. This model overcomes the deficiencies of the theoretical model and improves the accuracy of simulations. CFD-generated results are found to be in agreement with the experimental findings, highlighting its effectiveness at flow characterization. The ability of a CFD model to furnish precise data without the need to experimentally evaluate a physical model/prototype of the valve promises to benefit suspension engineers involved in the development and testing of load leveling valve designs. This document concludes with a sample case study which uses CFD to characterize flow in a modified disk-slot valve, and discusses the results in light of application on a heavy truck.

Les avantages de la gestion intégrée de l'eau sont connus depuis des années, mais ces approches sont devenues plus importantes que jamais. Cela est reconnu dans l'objectif 6 des 17 objectifs de développement durable des Nations Unies. En ciblant à la fois l'eau potable et l'assainissement, cet objectif demande une approche intégrée car il reconnaît leur interdépendance. Cette thèse vise à faire progresser le domaine de la modélisation intégrée de l'eau, et en particulier en ce qui concerne les matières en suspension. Les émissions globales de l'assainissement ont gagné en intérêt puisque les normes de qualité de l'eau sont étendues de la station de récupération des ressources de l'eau au réseau d'égout. La modélisation intégrée permet d'évaluer les interactions et d'estimer les émissions en plus des mesures de la qualité de l'eau (encore) rares. Les particules peuvent être considérées comme un indicateur de la qualité de l'eau couvrant la pollution particulaire, mais aussi les matières organiques, les nutriments et les substances telles que les micro-polluants hydrophobes. L'approche de modélisation choisie est conceptuelle, pour ses calculs rapides, et basée sur la distribution de la vitesse de décantation des particules, partout où la décantation et la remise en suspension sont les processus caractéristiques. L'approche est complétée par d'autres modèles pour couvrir le système du bassin versant jusqu'à la station de récupération des ressources de l'eau. Un modèle intégré nécessite de nombreuses données. Pour une collecte de données efficace, premièrement, une procédure est établie pour construire un modèle conceptuel d'égout à partir d'un modèle hydraulique détaillé. Deuxièmement, une méthodologie de conception expérimentale optimale est adaptée à l'environnement complexe des égouts pour une campagne de mesure de la qualité de l'eau. L'utilité de l'approche de la distribution de la vitesse de décantation des particules est ensuite démontrée en calibrant et en validant le modèle pour une site pilote. Une procédure est élaborée pour tenir compte de l'incertitude des paramètres et de la variabilité des données d'entrée an d'identifier des points de contrôle ables. La procédure est utilisée pour la réduction d'émission des particules, facilitée par le calcul rapide du modèle car plusieurs analyses de sensibilité sont demandées. Le dernier chapitre termine la thèse par l'évaluation pratique des stratégies visant à réduire les émissions globales. La thèse fait ainsi progresser le domaine de la modélisation intégrée des particules et fournit en même temps des procédures qui permettent de surmonter les obstacles généraux à la modélisation en mettant l'accent sur la collecte de données fiables et efficaces, la modélisation ainsi que l'optimisation. / The advantages of integrated water management have been known for decades, but are more than ever important. This is acknowledged in goal six of the 17 sustainable development goals of the United Nations. By targeting both clean water as well as sanitation, this goal is inherently asking for an integrated approach since it recognizes their interdependence. This dissertation aims at advancing the eld of integrated water systems modelling in general, and in particular with respect to suspended solids. Overall emissions from the integratedurban wastewater system have gained interest since water quality standards are increasing lyextended from the water resource recovery facility to the sewer system. Integrated modelling allows evaluating interactions and estimating overall emissions complementary to the not (yet) abundant water quality measurements. For this evaluation suspended solids can be seen asan indicator for the receiving water quality covering particulate pollution as such, but also undesired organic matter, nutrients and substances such as hydrophobic micropollutants. The modelling approach chosen is conceptual, due to its rapid calculations, and based on the particle settling velocity distribution wherever settling and resuspension are the characteristic processes of suspended solids. The approach is extended with complementary models to coverthe integrated system from the catchment down to the water resource recovery facility. The development of an integrated model however requires vast data sets. First, for efficient data collection a procedure is established to build a fast conceptual sewer model from its detailed hydraulic counter part. Second, an optimal experimental design methodology is adapted to the challenging sewer environment for the efficient planning of a water quality measurement campaign. The usability of the particle settling velocity approach is then shown by calibrating and validating the model for a case study. A procedure is developed to consider parameter uncertainty and input variability to identify reliable control handles. The procedure is appliedfor the abatement of total suspended solid, facilitated by the comparably low computational demand of the model, as the procedure asks for multiple global sensitivity analyses. The last chapter closes the dissertation with the practical application of evaluating different strategies to reduce the total suspended emissions to the receiving water. The dissertation thus advances the eld of integrated modelling for particulates and at the same time provides procedures which overcome barriers general to modelling focusing on reliable and efficient data collection,as well as optimization.

Thesis (MScEng (Mechanical and Mechatronic Engineering))--University of Stellenbosch, 2008. / Suspension seats are commonly used for earth moving machinery to isolate vehicle operators from vibrations transmitted to the vehicle body. To provide the required stiffness and damping for these seats, air-springs are typically used in conjunction with dampers. However, to eliminate the need for additional dampers, air-springs can be used in conjunction with auxiliary air volumes to provide both spring stiffness and damping. The damping is introduced through the flow restriction connecting the two air volumes. In this study, simplified models of an air-spring were derived followed by a model including the addition of an auxiliary volume. Subsequent to simulations, tests were performed on an experimental apparatus to validate the models. The air-spring models were shown to predict the behaviour of the experimental apparatus. The air-spring and auxiliary volume model followed the trend predicted by literature but showed approximately 27 % lower transmissibility amplitude and 21 % lower system natural frequency than obtained by tests when using large flow restriction diameters. This inaccuracy was assumed to be introduced by the simplified mass transfer equations defining the flow restriction between air-spring and auxiliary volume. The models however showed correlation when the auxiliary volume size was decreased by two thirds of the volume actually used for the experiment. This design of a prototype air-spring and auxiliary volume is presented for a suspension seat used in articulated or rigid frame dump trucks. The goal of this study was to design a suspension seat for this application and to obtain a SEAT value below 1,1. The design was optimised by varying auxiliary volume size, flow diameter and load. A SEAT value of less than 0,9 was achieved.

Au cours de ce travail, nous nous sommes attachés à prédire les caractéristiques rhéologiques de suspensions de particules dans des situations non newtoniennes. Nous avons tout d'abord étudié expérimentalement le comportement rhéologique de suspensions bidisperses de particules non colloïdales et non browniennes immergées dans un fluide à seuil dans les situations où les interactions entre particules sont hydrodynamiques (mécaniques). Nous avons montré que l'influence de la polydispersité des particules sur le module élastique et le seuil de contrainte de la suspension pouvait être décrit en utilisant un modèle d'empilement développé pour les matériaux granulaires secs. Deux relations analytiques simples permettant d'estimer le module élastique et la contrainte seuil d'une suspension non newtonienne polydisperse en fonction des caractéristiques des constituants ont alors été proposées et validées. Nous nous sommes ensuite intéressés à la prédiction des caractéristiques globales de suspensions contenant une forte concentration de particules de taille différente dans le cadre d'une approche par changement d'échelle. Pour cela, nous avons mis en uvre une approche par motifs morphologiques permettant la prise en compte de paramètres géométriques tels que la distribution de tailles des inclusions ou les distances entre particules voisines. Les résultats de ce modèle ont été validés par comparaison avec des données expérimentales de la littérature. Nous avons ensuite montré que les modèles morphologiques permettaient également de décrire l'évolution des modules complexes d'un matériau thixotrope subissant une transition de phase liquide-solide due à la croissance et la percolation d'un réseau de particules attractives. Dans la dernière partie de cette étude, nous avons étudié expérimentalement l'influence de l'histoire de cisaillement sur la structuration et la déstructuration d'une suspension colloïdale thixotrope / In this work, we aimed to predict the rheological properties of suspensions of particles in non-Newtonian fluids. We first studied experimentally the rheological behaviour of bidisperse suspensions of non colloidal particles and non-Brownian immersed in a yield stress fluid in situations where interactions between particles are hydrodynamic (mechanical). We showed that the influence of the polydispersity on the elastic modulus and yield stress of the suspension could be described using a packing model developed for dry granular materials.Two simple closed form estimates for both the elastic modulus and the yield stress of non-Newtonian polydisperse suspension as a function of the characteristics of the constituents were proposed and validated. This study was then extended in a micromechanical analytical framework to predict the macroscopic characteristics of suspensions containing highly concentration of particles with different size. Our ne w model, called a "Morphologically Representative Pattern - Based Approach" (MRP-based approach), successfully predicts the behaviour of such materials by taking into account some geometrical parameters such as the particle size distribution or the distance between the nearest-neighbour particles. This model has been validated by a comparison with the experimental data in the literature. We then showed that the MRP-based approach also allowed describing the evolution of complex modules of a thixotropic material undergoing a liquid-solid transition due to the growth and the percolation of an attractive particle network. In the last part of this work, we studied experimentally the influence of the flow history on the structure of a thixotropic colloidal suspension

The research work performed in this thesis has been carried out at the Production Tech-nology Centre where the Thermal Spray research group of University West has its work-shop and labs. This research work has been performed in collaboration with the International Advanced Research Centre for Powder Metallurgy and New Materials (ARCI), Hyderabad, India. First of all, I would like to express my sincere thanks and gratitude to my supervisors Dr. Nicolaie Markocsan and Dr. Nicholas Curry for their guidance, great support and valuable suggestions without which this work could not have been possible. I would also like to thanks Prof. Per Nylén for keeping faith in me and providing me an opportunity to work at PTC, which is a great place to perform research. It is my pleasure being their student and I wish I would keep learning from all of them, both on academic and personal grounds. I would also like to thank my colleagues at PTC Mr. Mohit Gupta and Mr. Stefan Björklund, for their help and support during this work. I would like to acknowledge the H.C. Starck Company for its financial support for the pro-ject; Dr. Filofteia-Laura TOMA at Fraunhofer IWS, Dresden to help us in spraying suspen-sion sprayed YSZ top coats, G Shivkumar from ARCI to help us in spraying solution pre-cursor sprayed top coats and Toni Bogdanoff, Jönköping University to help us in conduct-ing the LFA experiment

Les granulats naturels alluvionnaires sont utilisés actuellement à hauteur de 70 à 80% en volume dans la composition d'un béton et représentent désormais une ressource non renouvelable dont l'accessibilité diminue. Les formulations actuelles de béton visent à augmenter la part de granulats de roche massive concassés. Ces matériaux, de par leur histoire de fabrication, n'ont pas du tout la même morphologie. La substitution des granulats roulés par des granulats concassés affecte considérablement la mise en œuvre du matériau. Dans un premier temps, nous étudions spécifiquement l'influence de la morphologie des granulats sur les propriétés d'empilement des grains. Nous montrons que, le facteur d'aspect des grains affecte fortement la compacité dense et lâche aléatoire. À l'inverse, pour des grains de facteur d'aspect identique, l'impact de la convexité sur la mesure de la compacité dense est faible. Nous étendons, dans un deuxième temps, notre étude au comportement rhéologique de nos systèmes composés d'émulsion inverse et de sables monodisperses. Pour cela, nous construisons des courbes d'écoulement à l'échelle locale grâce à un imageur à résonance magnétique. Nous montrons que cette méthode est, de nos jours, la plus précise pour des mesures de viscosités, car elle prend en compte la migration et la localisation du matériau durant l'écoulement. Nous montrons également que le facteur d'aspect contrôle tout d'un point de vue rhéologique dans le régime où il n'y a pas de réseau percolé de contacts entre grains / Natural alluvial aggregates compose up to 70 at 80% of the volume of today's concrete and now represent for a non-renewable resource whose accessibility is reduced. Actually, concrete mix designs tend to increase the share of aggregate crushed rock mass. These materials have a very different morphology because of their manufacturing process. The substitution of rolled aggregates by crushed aggregates affects the concrete workability. At first, we study specifically the influence of the morphology of the aggregates on the packing properties. We show that the aspect ratio strongly affects the dense and random loose packing. However, for particles of the same aspect ratio, the impact of the convexity of the measure of dense packing is low. In a second step, we go further in our study by measuring the rheological behavior of our systems composed of inverse emulsion and monodisperse sands. To that purpose, we construct flow curves at the local scale thanks to a magnetic resonance imager. We show that this method is, nowadays, the most accurate for viscosity measurements because it takes into account the migration and location of the material during flow. We also show that the aspect ratio control everything of a rheological point of view in the regime where there is no percolated network of contacts between particles