An essential tool in the truck development process is the ability to quantify and grade vehicle dynamic behavior. Today this is performed either through subjective or objective tests. Subjective tests have the disadvantage that numerous factors influence test drivers’ opinions while objective measures have the advantage of repeatability. However, objective methods of today are often only able to provide a rough grading of vehicles. The main objective with this thesis is to develop more sensitive objective methods for ride comfort evaluation. An effective test procedure to measure driver perception sensitivity to small differences in vehicle ride is suggested and utilized. The driver sensitivity is tested on dynamic behavior that is typically graded in vehicle development. Cab motions from a truck are first measured and then recreated in a simulator where a test driver is seated. The perception threshold for small changes in typical vehicle motion is established in this way for each test person. The perception sensitivity tests indicate that humans are quite sensitive to transients in vehicle motion. One problem with many common vehicle ride measures is that the impact of transient behavior is small due to the averaging used to condense the measurement data into scalar measures. A new evaluation method for ride comfort, with influences from the well known handling diagram, is suggested. This method has four main advantages: it is fairly simple to interpret, it shows the absolute vibration level, it considers transient events separately and it shows changes in vehicle character with increasing excitation. Promising results from both measurements and simulations are derived. New technology has made it possible to vary vehicle suspension parameters during vehicle ride. In order to prescribe different damping for different vehicle modes, modal motion estimates are needed. A system identification approach is suggested. It yields improved estimates of vehicle modal motion compared to previous work. / QC 20101221

In motorsport, simulating road vehicles driving at the limit of handling is a valuable tool to study and optimize their overall performance during the design and set-up phases. Along with Quasi-Steady-State optimization, optimal control (OC) is the most utilized technique to simulate the control and states of a vehicle during minimum-time maneuvers and has been used for offline lap-time optimization for more than twenty years now. Since the first applications of optimal control in this field, it has been clear that the solution of the minimum-time optimization does not represent a model of the human driver but instead substitutes him/her. However, the common points or divergences between the minimum-time strategy of human race drivers and the OC one are still unclear. Moreover, it seems that in the literature there is no agreement about what vehicle models must be used, and in general the choice of one model or the other is not clearly justified. Finally, thanks to the rise in popularity of autonomous driving and racing, optimal control has been used as path planner for automated vehicles: %nonetheless, the application of free-trajectory real-time nonlinear optimal control in Model Predictive Control (MPC) schemes, where the optimal controls are directly fed to the vehicle, is still an unexplored topic. nonetheless, the application of free-trajectory real-time nonlinear optimal control in Model Predictive Control (MPC) schemes, where the optimal controls are computed from a single optimization and directly fed to the vehicle, is a topic still open for exploration. The first aim of this thesis is to provide an objective comparison of several vehicle, tire, powertrain and road models to be used in minimum-time OC. In the first part of this work we thus detail several models of the vehicle and its subsystems. We then solve minimum-time OC problems on a series of test tracks adopting most of the model combinations and discuss the differences in the solutions. We then draw conclusions on the best model combinations to obtain realistic and reliable minimum-time maneuvers. The second part of the thesis aims to prove that the solutions of minimum-time OC problems are indeed different from the driving behavior of professional drivers, but they can be employed to coach the human driver and improve his/her racing performance. After modeling a high-performance vehicle manufactured by Ferrari, we again use optimal control to compute minimum-time maneuvers on two different tracks. A professional racer driving is then coached in following the OC strategy on the Ferrari driving simulator, and we objectively prove that the driver can outperform his previous lap times. In the third and last part of the thesis, we aim to prove that free-trajectory real-time optimal control is a valid alternative to hierarchical MPC frameworks based on high-level path planning and low-level path tracing. We first develop a novel kineto-dynamic vehicle model able to satisfy the trade-off between computational lightness and accuracy in representing the vehicle's pure and combined dynamics. Then, by solving a minimum-time OC in real-time, we are able to autonomously drive a real scaled vehicle around a track at the limits of tire adherence.

This thesis aims at estimating vital vehicle states and developing control algorithms for automotive suspensions and vehicle stability. A parametric model of an automotive monotube damper is developed and several control algorithms for semi-active suspensions have been developed. An extensive comparison of different control algorithms has been done. Skyhook, Groundhook, Hybrid, Acceleration-driven, Power-driven, Groundhook-linear, Linear Quadratic Regulator (LQR) optimal, Genetic algorithm optimized Linear Quadratic Regulator optimal, Model-reference adaptive, H∞ robust, µ-synthesis, fuzzy-logic based, and Deep Reinforcement learning based control algorithms have been developed and simulated. A shock dyno is instrumented and skyhook and groundhook control algorithms have been implemented as well. In addition to this, a semi-active suspension switching based control algorithm is developed for reducing the effort of a direct moment yaw rate controller, and improve stability of a vehicle when turning. Accelerometer based intelligent tires have been used to estimate vehicle states like vertical load on tire, velocity of the vehicle, unsprung mass acceleration, and forces on a tire. All these estimations would be helpful in observing various parameters of a vehicle using data from only a tri-axis accelerometer inside the tire. Data was collected in an instrumented Volkswagen Jetta and a Trailer setup as well. The test vehicle was instrumented with a tri-axis accelerometer inside the tire, encoder, Inertial Measurement Unit (IMU), and VBOX Racelogic Global Positioning System (GPS) based velocity measurement unit. For payload estimation, the data collected by the in-tire accelerometer was converted into frequency domain using Welch's method of averaging, followed by feature extraction. The extracted features were fed to a trained bagged trees model. Root mean squared error of 11% was observed on the test dataset. For velocity estimation, the data collected by the accelerometer was fed to a variational mode decomposition process. The extracted mode was converted to time-frequency domain using Hilbert transform and features for machine learning were extracted. A root mean squared error of 1.02kmph was observed on the trained dataset. A Gaussian process model was trained for this application. For unsprung mass acceleration estimation, the test vehicle was instrumented with an accelerometer near the wheel spindle as well. For this estimation problem, Convolutional neural networks (CNN) were used. The time-frequency spectrogram of x, y, and z axis data of the in-tire accelerometer were considered as the three color channels of an image. With this, an image of 224 x 224 x 3 dimensions was generated, which represented the time and frequency variation of data. These images were used for training the CNN and a 96.8% coefficient of correlation was obtained for this regression task. For the last wheel force estimation problem, the concept of training the images generated by overlapping time-frequency matrices was used and an accuracy of 90.1% was achieved. With these estimation of vehicle states, better control algorithms can be developed and deployed for better handling, safety and comfort of vehicles using data from only tri-axis accelerometer in the tire. / Master of Science / The main objective of this thesis is to aid in the development of better control systems for vehicles, using data from accelerometer-based intelligent tire. Payload on the vehicle's tire, vehicle velocity, wheel acceleration, and wheel forces are vital parameters, which if estimated correctly can be instrumental in having better understanding of the vehicle's condition. A tri-axis accelerometer is mounted inside the tire, and is used for estimating these vehicle parameters. Statistical models are developed based on features extracted from the accelerometer data. The main challenge was to use the data collected by only intelligent tire to estimate vehicle states. This makes the developed algorithms independent of other sensors and hence economic. Tires are the only component which serve as a link between the vehicle and road. Hence, these parameter estimations can be accurately observed simultaneously using the in-tire accelerometer data. Testing is done on an instrumented trailer-test setup and a Volkswagen Jetta. The vehicle is instrumented with the intelligent tire, a Global positioning system (GPS) based velocity measuring unit, Inertial measurement unit (IMU), and encoder. Testing is done for different loading conditions, road surfaces, inflation pressures, and vehicle velocities. In this way, it has been attempted to make the developed statistical models robust and expose them to a multitude of test conditions. In addition to this, several suspension semi-active control algorithms have been developed for improving vehicle ride comfort and road holding. A parametric damper model has been developed, and several control algorithms have been simulated. A shock dyno experimental setup has been instrumented and some of the control algorithms have been implemented. With this, several suspension semi-active control algorithms have been developed, and statistical models have been developed for estimation of various vehicle parameters. This research can be helpful for developing accurate control algorithms for active safety systems in a vehicle.

A significant number of studies address various aspects related to tire modeling; most are dedicated to the development of tire models for on-road conditions. Such models cover a wide range of resolutions and approaches, as required for specific applications. At one end of the spectrum are the very simple tire models, such as those employed in real-time vehicle dynamic simulations. At the other end of the spectrum are the very complex finite element models, such as those used in tire design. In between these extremes, various other models have been developed, at different levels of compromise between accuracy and computational efficiency. Existing tire models for off-road applications lag behind the on-road models. The main reason is the complexity added to the modeling due to the interaction with the soft soil. In such situations, one must account for the soil dynamics and its impact on the tire forces, in addition to those aspects considered for an on-road tire. The goal of this project is to develop an accurate and comprehensive, while also efficient, off-road tire model for soft soil applications. The types of applications we target are traction, handling, and vehicle durability, as needed to support current army mobility goals. Thus, the proposed approach is to develop a detailed semi-analytical tire model for soft soil that utilizes the tire construction details and parallels existing commercially available on-road tire models. The novelty of this project relies in developing a three-dimensional three-layer tire model employing discrete lumped masses and in improving the tire-soil interface model. This will be achieved by enhancing the resolution of the tire model at the contact patch and by accounting for effects and phenomena not considered in existing models. / Master of Science

A racing vehicle suspension system is a kinematic linkage that supports the vehicle under complex loading scenarios. The suspension also defines the handling characteristics of the vehicle. Understanding the loads that the suspension carries in a variety of loading scenarios is necessary in order to properly design a safe and effective suspension system. In the past, the Formula SAE team at Virginia Tech has used simplified calculations to determine the loads expected in the suspension members. This approach involves several large assumptions. These assumptions have been used for years and the justification for them has been lost. The goal of this research is to determine the validity of each of the assumptions made in the method used for calculating the vehicle suspension loads by hand. These assumptions include modeling the suspension as pinned-pinned truss members to prevent bending, neglecting any steering angle input to the suspension, and neglecting vertical articulation of the system. This thesis presents an approach to modeling the suspension member loads by creating a finite element (FE) model of the entire suspension system. The first stage of this research covers the validation of the current calculation methods. The FE model will replicate the suspension with all of the current assumptions and the member loads will be compared to the hand calculations. This truss-element-based FE model resulted in member loads identical to the hand calculations. The next stage of the FE model development converts the truss model to beam elements. This step is performed to determine if the assumption that bending loads are insignificant is a valid approach to calculating member loads. In addition to changing the elements used from truss to beam element, the suspension linkage was adapted to more accurately model the methods by which each member is attached to the others. This involves welding the members of each control arm together at the outboard point as well as creating a simplified version of the pull rod mounting bracket on the upper control arm. The pull rod is the member that connects the ride spring, damper, and anti-roll bar to the wheel assembly and had previously been mounted on the upright. This model reveals reduced axial components of load but increases in bending moments sizable enough to reduce the resistance to buckling of any member in compression. The third stage of model development incorporates the steer angle that must be present in loading scenarios that involve some level of cornering. An analysis of the vehicle trajectory that includes the effects of slip angle is presented and used to determine the most likely steer angle the vehicle will experience under cornering. The FE model was adapted to include the movement of the steering linkage caused by driver input. This movement changes the angle of the upright and steering linkage as well as the angle at which wheel loads are applied to the suspension. This model results in a dramatic change in member loads for loading cases that involve a component of steering input. Finally, the FE model was further enhanced to account for vertical movement of the suspension as allowed by the spring and damper assembly. The quasi-static loading scenarios are used to determine any member loading change due to vertical movement. The FE model is also used to predict the amount of vertical movement expected at the wheel center. This data can be used by the suspension designer to determine if changes to the spring rate or anti-roll bar stiffness will result in a more desirable amount of wheel movement for a given loading condition. This model shows that there is no change in the member loads due to the vertical movement of the wheel. This thesis concludes by presenting the most important changes that must occur in member load calculations to determine the proper suspension loading under a variety of loading scenarios. Finally, a discussion of future research is offered including the importance of each area in determining suspension loads and recommendations on how to perform this research. / Master of Science

This experimental study evaluates the dynamic response of five semi-active control policies as tested on a single suspension quarter-car system. Incorporating a magneto-rheological damper, the full-scale 2DOF quarter-car system was used to evaluate skyhook, groundhook, and hybrid control. Two alternative skyhook policies were also considered, namely displacement skyhook and relative displacement skyhook. As well as exploring the relative benefits of each of these controllers, the performance of each semi-active controller was compared to the performance of conventional passive damping. Each control policy is evaluated for its control performance under three different base excitations: chirp, step, and pure tone. Corresponding to the chirp input, transmissibilities and auto spectrums are considered for each control policy. Specifically, transmissibilities between the sprung mass displacement and the unsprung mass displacement are generated relative to the input displacement. Further, the ratio between the relative displacement across the damper and the input displacement is evaluated for each control technique. The chirp input also reveals the results of the auto spectrums of the sprung and unsprung mass accelerations. Both the step input and the pure tone input were used to generate time domain values of RMS and peak-to-peak displacements and accelerations. This study shows that semi-active control offers benefits beyond those of conventional passive damping. Further, traditional skyhook control is shown to outperform the less conventional alternative skyhook policies. / Master of Science

This analytical study evaluates the response characteristics of a two-degree-of freedom quarter-car model, using passive and semi-active dampers, along with a seven-degree-of-freedom full vehicle model. The behaviors of the semi-actively suspended vehicles have been evaluated using skyhook, groundhook, and hybrid control policies, and compared to the behaviors of the passively-suspended vehicles. The relationship between vibration isolation, suspension deflection, and road-holding is studied for the quarter-car model. Three main performance indices are used as a measure of vibration isolation (which can be seen as a comfort index), suspension travel requirements, and road-holding quality. After performing numerical simulations on a seven-degree-of-freedom full vehicle model in order to confirm the general trends found for the quarter-car model, these three indices are minimized using optimization techniques. The results of this study indicate that the hybrid control policy yields better comfort than a passive suspension, without reducing the road-holding quality or increasing the suspension displacement for typical passenger cars. The results also indicate that for typical passenger cars, the hybrid control policy results in a better compromise between comfort, road-holding and suspension travel requirements than the skyhook and groundhook control policies. Finally, the numerical simulations performed on a seven-degree-of-freedom full vehicle model indicate that the motion of the quarter-car model is not only a good approximation of the heave motion of a full-vehicle model, but also of the pitch and roll motions since both are very similar to the heave motion. / Master of Science

This study experimentally evaluates the dynamic response of a single vehicle suspension incorporating a magneto-rheological (MR) damper. A full-scale two-degree-of-freedom (2DOF) quarter-car test apparatus has been constructed at the Advanced Vehicle Dynamics Lab at Virginia Tech to evaluate the response of a vehicle suspension under the different control schemes of skyhook, groundhook, and hybrid semiactive control. The quarter-car apparatus was constructed using materials from 80/20 Incorporated and a hydraulic actuation system from MTS. A dSPACE AutoBox was used both for controlling the MR dampers and acquiring data. The first task was to understand the baseline dynamic response of the quarter-car system with only a passive damper. Next, the passive damper was replaced with a controllable MR damper. The control schemes of skyhook, groundhook, and hybrid semiactive control were applied to the MR damper. The physical response of the quarter-car with the different control schemes was then compared to the analytical prediction for the response, with favorable results. The response of the quarter-car with the semiactive damper was also compared to the response of the quarter-car with a passive damper, and the resulting limitations of passive damping are discussed. Finally, the practical implications of this study are shown in a discussion of the physical implementation of the MR dampers in the Virginia Tech FutureCar, a full-size Chevrolet Lumina. Although the actual skyhook, groundhook, and hybrid semiactive control schemes were not implemented on the vehicle, the results were promising and generated several recommendations for future research. / Master of Science

One seminal question that faces a vehicle's driver (either human or computer) is predicting the capability of the vehicle as it encounters upcoming terrain. A Location-Aware Adaptive Vehicle Dynamics (LAAVD) System is being developed to assist the driver in maintaining vehicle handling capabilities through various driving maneuvers. In contrast to current active safety systems, this system is predictive, not reactive. The LAAVD System employs a predictor-corrector method in which the driver's input commands (throttle, brake, steering) and upcoming driving environment (terrain, traffic, weather) are predicted. An Intervention Strategy uses a novel measure of handling capability, the Performance Margin (PM), to assess the need to intervene. The driver's throttle and brake control are modulated to affect desired changes to the PM in a manner that is minimally intrusive to the driver's control authority. This system depends heavily on an understanding of the interplay between the vehicle's longitudinal, lateral, and vertical forces, as well as their resulting moments. These vehicle dynamics impact the PM metric and ultimately the point at which the Intervention Strategy will modulate the throttle and brake controls. Real-time implementation requires the development of computationally efficient predictive models of the vehicle dynamics. In this work, a method for predicting future vehicle states, based on current states and upcoming terrain, is developed using perturbation theory. An analytical relationship between the change in the spindle forces and the resulting change in the PM is derived, and the inverse relationship, between change in PM and resulting changes in longitudinal forces, is modeled. This model is implemented in the predictor-corrector algorithm of the Intervention Strategy. Corrections to the predicted states are made at each time step using a detailed, full, non-linear vehicle model. This model is run in real-time and is intended to be replaced with a drive-by-wire vehicle. Finally, the impact of this work on the automotive industry is discussed and recommendations for future work are given. / Master of Science

This study provides a comprehensive performance analysis of Wave-Adaptive Modular Vessels (WAM-V) using simulations and testing data. WAM-Vs are a new class of marine technology that build upon the advantages of lightweight, low-draft, catamaran construction. Independent suspensions above the hulls isolate the passengers and equipment from the harsh sea environment. Enhanced understanding of the relationship between suspension and vehicle performance is critical for future missions of interest to the U.S. Navy. Throughout this study, the dynamic properties of three different WAM-Vs were evaluated. A multi-body dynamics simulation was developed for the 100-ft WAM-V 'Proteus' based on an automotive 4-post shaker rig. The model was used to characterize the sensitivities of different suspension parameters and as a platform for future models. A 12-ft unmanned surface vessel (USV) was instrumented and sea trials were conducted in the San Francisco Bay. A dynamic 4-post simulation was created for the USV using displacement inputs calculated from acceleration data via a custom integration scheme. The data was used to validate the models by comparing the model outputs to sensor data from the USV. A vertical hydrodynamics testing rig was developed to investigate the interaction between the pontoons and the water surface to improve the understanding of how hydrodynamic forces affect suspension performance. A model was created to accurately simulate the hydrodynamic forces that result from vertical pontoon motion. The model was then scaled to fit a 33-ft WAM-V prototype. The 33-ft WAM-V was instrumented and sea trials were conducted in Norfolk, VA. The WAM-V's suspension was upgraded based on the testing results. A 2-post rig was also built for evaluating the 33-ft WAM-V's dynamics. Two dynamic models were made for the 33-ft WAM-V to evaluate different suspension designs. The results from this study have numerous impacts on the naval community and on the development of WAM-Vs. The methodology for testing and evaluation will allow for future WAM-V designs to be compared under controlled circumstances. The performance of WAM-Vs can then be compared against conventional platforms to determine their suitability for future missions. Simulation development will enable future WAM-Vs to be evaluated prior to undergoing sea trials. The hydrodynamic models become a powerful design tool that can be easily scaled and combined with the 4-post models. By providing the simulations and test data to future vessel designers, the designers will be able to intelligently evaluate numerous iterations early in the design phase, improving performance and safety. / Ph. D.