Search

Steel-concrete composite beams are widely used in building structures due to the excellent mechanical performances. In the application of steel-concrete composite beam, the problem exists in the coordination between concrete slab and steel beam. In this work, based on the Eurocode-4, a series of four 1:1 full-size push-out specimens were designed to study the influence of different napping arrangement styles on the shear performance of the napping joints in the cast-in-place specimens and prefabricated specimens. Based on the test results, finite element analysis was finished for parametric analysis. The test results show that the final failure modes of the specimens include concrete failure and bolt shear. Increasing the concrete grade of the precast specimen can improve the shear resistance of the bolt group. The uniformly arranged specimens show higher shear stiffness and ductility compared to the non-uniformly arranged specimens. The results of finite element analysis showed that the shear capacity of the group nails was increased with the increase of the compressive strength grade of the post-pouring material and the longitudinal spacing of the bolts, and the increase of the spacing of the bolts was more effective in improving the shear capacity of the group nails.

This dissertation addresses rigid body attitude control with piecewise sinusoidal signals. We consider rigid-body attitude kinematics on SO(3) with a class of sinusoidal inputs. We present a new closed-form solution of the rotation matrix kinematics. The solution is analyzed and used to prove controllability. We then present kinematic-level orientation-feedback controllers for setpoint tracking and command following. Next, we extend the sinusoidal kinematic-level control to the dynamic level. As a representative dynamic system, we consider a CubeSat with vibrating momentum actuators that are driven by small $\epsilon$-amplitude piecewise sinusoidal internal torques. The CubeSat kinetics are derived using Newton-Euler's equations of motion. We assume there is no external forcing and the system conserves zero angular momentum. A second-order approximation of the CubeSat rotational motion on SO(3) is derived and used to derive a setpoint tracking controller that yields order O(ε2) closed-loop error. Numerical simulations are presented to demonstrate the performance of the controls. We also examine the effect of the external damping on the CubeSat kinetics. In addition, we investigate the feasibility of the piecewise sinusoidal control techniques using an experimental CubeSat system. We present the design of the CubeSat mechanical system, the control system hardware, and the attitude control software. Then, we present and discuss the experiment results of yaw motion control. Furthermore, we experimentally validate the analysis of the external damping effect on the CubeSat kinetics. Copyright: © Shaoqian Wang 2018 Funding information: This work was supported in part by the National Science Foundation (CMMI-1538782) and the National Aeronautics and Space Administration (80NSSC17M0040).

Wearable cameras are increasingly used in many different applications such as entertainment, security, law enforcement and healthcare. In this thesis, we focus on the application of the police worn body camera and behavioral recording using a wearable camera for one-on-one therapy with a child in a classroom or clinic. To protect the privacy of other individuals in the same environment, we introduce a new visual privacy protection technique called visual bubble. Visual bubble is a virtual zone centered around the camera for observation whereas the rest of the environment and people are obfuscated. In contrast to most existing visual privacy protection systems that rely on visual classifiers, visual bubble is based on depth estimation to determine the extent of privacy protection. To demonstrate this concept, we construct a wearable stereo camera for depth estimation on the Raspberry Pi platform. We also propose a novel framework to quantify the uncertainty in depth measurements so as to minimize a statistical privacy risk in constructing the depth-based privacy bubble. To evaluate our system, we have collected three datasets. The effectiveness of the proposed scheme is demonstrated with experimental results.