Search Results

A Machine Learning (ML) approach is presented to correlate in-cylinder images of early flame kernel development within a spark-ignited (SI) gasoline engine to early-, mid-, and late-stage flame propagation. The objective of this study was to train machine learning models to analyze the relevance of flame surface features on subsequent burn rates. Ultimately, an approach of this nature can be generalized to flame images from a variety of sources. The prediction of combustion phasing was formulated as a regression problem to train predictive models to supplement observations of early flame kernel growth. High-speed images were captured from an optically accessible SI engine for 357 cycles under pre-mixed operation. A subset of these images was used to train three models: a linear regression model, a deep Convolutional Neural Network (CNN) based on the InceptionV3 architecture and a CNN built with assisted learning on the VGG19 architecture.

The classical methods of measuring acoustic absorption coefficient using an impedance tube and a reverberation chamber are well established [1, 2]. However, these methods are not suitable for in-situ applications. The two in-situ methods; single channel microphone (P- probe) and dual channel acoustic pressure and particle velocity (Pu-probe) methods based on measurement of impulse response functions of the material surface under test, provide considerable advantage in data acquisition, signal processing, ease and mobility of measurement setup. This paper evaluates the measurement techniques of these two in-situ methods and provides results of acoustic absorption coefficient of a commercial artificial Astroturf, a Dow quash material, and a grass surface.

The Time Variant Discrete Fourier Transform was implemented as a real-time order tracking method using developed software and commercially available hardware. The time variant discrete Fourier transform (TVDFT) with the application of the orthogonality compensation matrix allows multiple tachometers to be tracked with close and/or crossing orders to be separated in real-time. Signal generators were used to create controlled experimental data sets to simulate tachometers and response channels. Computation timing was evaluated for the data collection procedure and each of the data processing steps to determine how each part of the process affects overall performance. Many difficulties are associated with a real-time data collection and analysis tool and it becomes apparent that an understanding of each component in the system is required to determine where time consuming computation is located.

This study investigates a methodology in which the general public's subjective interpretation of vehicle handling and performance can be predicted. Several vehicle handling measurements were acquired, and associated metrics calculated, in a controlled setting. Human evaluators were then asked to drive and evaluate each vehicle in a winter driving school setting. Using the acquired data, multiple linear regression and artificial neural network (ANN) techniques were used to create and refine mathematical models of human subjective responses. It is shown that artificial neural networks, which have been trained with the sets of objective and subjective data, are both more accurate and more robust than multiple linear regression models created from the same data.

The effect of flow direction towards the spark plug electrodes on ignition parameters is analyzed using an innovative spark aerodynamics fixture that enables adjustment of the spark plug gap orientation and plug axis tilt angle with respect to the incoming flow. The ignition was supplied by a long discharge high energy 110 mJ coil. The flow was supplied by compressed air and the spark was discharged into the flow at varying positions relative to the flow. The secondary ignition voltage and current were measured using a high speed (10MHz) data acquisition system, and the ignition-related metrics were calculated accordingly. Six different electrode designs were tested. These designs feature different positions of the electrode gap with respect to the flow and different shapes of the ground electrodes. The resulting ignition metrics were compared with respect to the spark plug ground strap orientation and plug axis tilt angle about the flow direction.

In today’s race for improved fuel economy and lower emissions from gasoline engines, precise metering of delivered fuel is essential. Gasoline Direct Injection fuel systems provide the means for improved combustion efficiency through mixture preparation and better atomization. These improvements can be achieved from both increasing fuel pressure and using multiple injection events, which significantly reduce the required energizing time per injection, and in a number of cases, force the injector to operate at less than full stroke. When the injector operates in this condition, the influence of variation in injector dynamics account for a large percentage of the delivered fuel and require compensation to ensure accurate fuel delivery. Injector dynamics such as opening delay and closing time are influenced by operating conditions such as fuel pressure, energizing time, and temperature.

Stochastic Preignition (SPI) is an abnormal combustion event that occurs in a turbocharged engine and can lead to the loss in fuel economy and engine hardware damage, and in turn result in customer dissatisfaction. It is a significant limiting factor on the use and continued downsizing of turbocharged spark ignited direct injection (SIDI) gasoline engines. Understanding and mitigating all the factors that cause and influence the rate and severity of SPI occurrence are of critical importance to the engine’s continued use and fuel economy improvements for future designs. Previous studies have shown that the heavy molecular weight components of the fuel formulations are one factor that influences the rate of SPI from a turbocharged SIDI gasoline engine. All the previous studies have involved analyzing the fuel’s petroleum hydrocarbon chemistry, but not specifically the additives that are put in the fuel to protect and clean the internal components over the life of the engine.

Premium instrument panels (IPs) contain passenger airbag (PAB) systems that are typically comprised of a stiff plastic substrate and a soft ‘skin’ material which are adhesively bonded. During airbag deployment, the skin tears along the scored edges of the door holding the PAB system, the door opens, and the airbag inflates to protect the occupant. To accurately simulate the PAB deployment dynamics during a crash event all components of the instrument panel and the PAB system, including the skin, must be included in the model. It has been recognized that the material characterization and modeling of the skin tearing behavior are critical for predicting the timing and inflation kinematics of the airbag. Even so, limited data exists in the literature for skin material properties at hot and cold temperatures and at the strain rates created during the airbag deployment.

Predictive Signal Phase and Timing (SPAT) message set is one fundamental building block for vehicle-to-infrastructure (V2I) applications such as Eco-Approach and Departure (EAD) at traffic signal controlled urban intersections. Among the two complementary communication methods namely short-range sidelink (PC5) and long-range cellular radio link (Uu), this paper documents the work with long-range link: the complete data chain includes connecting to the traffic signals via existing backhaul communication network, collecting the raw signal phase state data, predicting the signal state changes and delivering the SPAT data via a geofenced service to requests over HTTP protocols. An Application Programming Interface (API) library is developed to support various cellular data transmission reduction and latency improvement techniques.

Due to growing interest in automated driving, the need for better understanding of human driving behavior in uncertain environment, such as driving behavior at un-signalized crossroad and roundabout, has further increased. Driving behavior at roundabout is greatly influenced by different dynamic factors such as speed, distance and circulating flow of the potentially conflicting vehicles, and drivers should choose whether to leave or wait at the upcoming exit according to these factors. In this paper, the influential dynamic factors and driving behavior characteristics at the roundabout is analyzed in detail, random forest method is then deployed to predict the driving behavior. For training the driving behavior model, four typical roundabout layouts were created under a real-time driving simulator with PanoSim-RT and dSPACE. Traffic participants with different motion style were also set in the simulation platform to mimic real driving conditions.

Large channel count data acquisition systems have seen increasing use in the acquisition and analysis of rotating machinery, these systems have the ability to generate very large amounts of data for analysis. The most common operating measurement made on powertrains or automobiles on the road or on dynamometers has become the order track measurement. Order tracking analysis can generate a very large amount of information that must be analyzed, both due to the number of channels and orders tracked. Analysis methods to efficiently analyze large numbers of Frequency Response Function (FRF) measurements have been developed and used over the last 20 years in many troubleshooting applications. This paper develops applications for several FRF based analysis methods as applied for efficient analysis of large amounts of order track data.

This paper presents the development of a reduced-order powertrain model for energy and SOC estimation of a multi-mode plug-in hybrid electric vehicle using only vehicle speed profile and route elevation as inputs. Such a model is intended to overcome the computational inefficiencies of higher fidelity powertrain and vehicle models in short and long horizon energy optimization efforts such as Coordinated Adaptive Cruise Control (CACC), Eco Approach and Departure (EcoAND), Eco Routing, and PHEV mode blending. The reduced-order powertrain model enables Connected and Automated Vehicles (CAVs) to utilize the onboard sensor and connected data to quickly react and plan their maneuvers to highly dynamic road conditions with minimal computational resources.

As the world searches for ways to reduce humanity’s impact on the environment, the automotive industry looks to extend the viable use of the gasoline engine by improving efficiency. One way to improve engine efficiency is through more effective control. Torque-based control is critical in modern cars and trucks for traction control, stability control, advanced driver assistance systems, and autonomous vehicle systems. Closed loop torque-based engine control systems require feedback signal(s); indicated mean effective pressure (IMEP) is a useful signal but is costly to measure directly with in-cylinder pressure sensors. Previous work has been done in torque and IMEP estimation using crankshaft acceleration and ion sensors, but these systems lack accuracy in some operating ranges and the ability to estimate cycle-cycle variation.

The proliferation of small silicon micro-chips has led to a large assortment of low-cost transducers for data acquisition. Production vehicles on average exploit more than 60 on board sensors, and that number is projected to increase beyond 200 per vehicle by 2020. Such a large increase in sensors is leading the fourth industrial revolution of connectivity and autonomy. One major downfall to installing many sensors is compromises in their accuracy and processing power due to cost limitations for high volume production. The same common errors in data acquisition such as sampling, quantization, and multiplexing on the CAN bus must be accounted for when utilizing an entire array of vehicle sensors. A huge advantage of onboard sensors is the ability to calculate vehicle parameters during a daily drive cycle to update ECU calibration factors in real time. One such parameter is driveline damping, which changes with gear state and drive mode. A damping value is desired for every gear state.

Analysis-driven pre-calibration of a modern automotive engine is extremely valuable in significantly reducing hardware investments and accelerating engine designs compliant with stricter emission regulations. Advanced modelling tools, such as a Virtual Engine Model (VEM) using Computational Fluid Dynamics (CFD), are often used within the framework of a Design of Experiments for Powertrain Engineering (DEPE) with the goal of streamlining significant portions of the calibration process. The success of the methodology largely relies on the accuracy of analytical predictions, especially engine-out emissions. Results show excellent agreements in engine performance parameters (with R2 > 98%) and good agreements in NOx and combustion noise (with R2 > 87%), while the Carbon Monoxide (CO), Unburned Hydrocarbons (HC) and Smoke emissions predictions remain a challenge even with a large n-heptane mechanism consisting of 144 species and 900 reactions and refined mesh resolution.

Third generation advanced high strength steels (AHSS) that rely on the transformation of austenite to martensite have gained growing interest for implementation into vehicle architectures. Previous studies have identified a dependency of the rate of austenite decomposition on the amount of strain and the associated strain path imposed on the sheet. The rate and amount of austenite transformation can impact the work hardening behavior and tensile properties. However, a deeper understanding of the impact on toughness, and thus crash performance, is not fully developed. In this study, the strain path and strain amounts were systematically controlled to understand the associated correlation to impact toughness in the end application condition (strained and baked). Impact toughness was evaluated using an instrumented Charpy machine with a single sheet v-notch sample configuration.