Search Results

Near-Field Acoustical Holography (NAH) is an inverse process in which sound pressure measurements made in the near-field of an unknown sound source are used to reconstruct the sound field so that source distributions can be clearly identified. NAH was originally based on performing spatial transforms of arrays of measured pressures and then processing the data in the wavenumber domain, a procedure that entailed the use of very large microphone arrays to avoid spatial truncation effects. Over the last twenty years, a number of different NAH methods have been proposed that can reduce or avoid spatial truncation issues: for example, Statistically Optimized Near-Field Acoustical Holography (SONAH), various Equivalent Source Methods (ESM), etc.

This paper explains the basic principles behind two platform technologies that my research team has developed in the field of optical metrology and optical information processing: 1) high-speed 3D optical sensing; and 2) real-time 3D video compression and streaming. This paper will discuss how such platform technologies could benefit the automotive industry including in-situ quality control for additive manufacturing and autonomous vehicle systems. We will also discuss some of other applications that we have been working on such as crime scene capture in forensics.

Nested planetary gear trains, which consist of two integrated co-axial single-stage planetary gearsets, have recently been widely implemented in automobile transmissions and various other applications. In the current study, a non-destructive vibrational and acoustical monitoring technique is developed to detect a common type of gear grinding defect for a complex nested gear train structure. A nested gear train which has an unground pinion with unpolished teeth profile is used to exemplify the developed methodology. An experimental test stand with an open and vertical mounting configuration has been designed to acquire both vibrational and acoustical data. The measured data are investigated using several signal processing techniques to identify unground pinions in the gear system. A general frequency spectrum analysis is performed initially, which is then followed by a peak finding algorithm to identify the peaks in the spectrum.

The basic idea behind large-eddy simulation (LES) is to accurately resolve the large energy-containing scales and to use subgrid-scale (SGS) models for the smaller scales. The accuracy of LES can be significantly impacted by the numerical discretization schemes and the choice of the SGS model. This work investigates the accuracy of low-order LES codes in the simulation of a turbulent round jet which is representative of fuel jets in engines. The turbulent jet studied is isothermal with a Reynolds number of 6800. It is simulated using Converge, which is second-order accurate in space and first-order in time, and FLEDS, developed at Purdue University, which is sixth-order accurate in space and fourth-order in time. The high-order code requires the resolution of acoustic time-scales and hence is approximately 10 times more expensive than the low-order code.

The vertical force generated from terrain-tire interaction has long been of interest for vehicle dynamic simulations and chassis development. To improve simulation efficiency while still providing reliable load prediction, a terrain pre-filtering technique using a constraint mode tire model is developed. The wheel is assumed to convey one quarter of the vehicle load constantly. At each location along the tire's path, the wheel center height is adjusted until the spindle load reaches the pre-designated load. The resultant vertical trajectory of the wheel center can be used as an equivalent terrain profile input to a simplified tire model. During iterative simulations, the filtered terrain profile, coupled with a simple point follower tire model is used to predict the spindle force. The same vehicle dynamic simulation system coupled with constraint mode tire model is built to generate reference forces.

A common scenario in engineering design is the evaluation of expensive black-box functions: simulation codes or physical experiments that require long evaluation times and/or significant resources, which results in lengthy and costly design cycles. In the last years, Bayesian optimization has emerged as an efficient alternative to solve expensive black-box function design problems. Bayesian optimization has two main components: a probabilistic surrogate model of the black-box function and an acquisition functions that drives the design process. Successful Bayesian optimization strategies are characterized by accurate surrogate models and well-balanced acquisition functions. The Gaussian process (GP) regression model is arguably the most popular surrogate model in Bayesian optimization due to its flexibility and mathematical tractability. GP regression models are defined by two elements: the mean and covariance functions.

In the dense fabric of urban areas, electric scooters have rapidly become a preferred mode of transportation. As they cater to modern mobility demands, they present significant safety challenges, especially when interacting with pedestrians. In general, e-scooters are suggested to be ridden in bike lanes/sidewalks or share the road with cars at the maximum speed of about 15-20 mph, which is more flexible and much faster than pedestrians and bicyclists. Accurate prediction of pedestrian movement, coupled with assistant motion control of scooters, is essential in minimizing collision risks and seamlessly integrating scooters in areas dense with pedestrians. Addressing these safety concerns, our research introduces a novel e-Scooter collision avoidance system (eCAS) with a method for predicting pedestrian trajectories, employing an advanced Long short-term memory (LSTM) network integrated with a state refinement module.