ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part B: Mechanical Engineering

Predictive maintenance, as a form of pro-active maintenance, has increasing usage and shows significant superiority over the corrective and preventive maintenance. However, conventional methods of predictive maintenance have noteworthy limitations in maintenance optimization and reliability improvement. In the last two decades, machine learning has flourished and overcome many inherent flaws of conventional maintenance prediction methods. Meanwhile, machine learning displays unprecedented predictive power in maintenance prediction and optimization. This paper compares the features of corrective, preventive, and predictive maintenance, examines the conventional approaches to predictive maintenance, and analyzes their drawbacks. Subsequently, this paper explores the driving forces, and advantages of machine learning over conventional solutions in predictive maintenance. Specifically, this paper reviews popular supervised learning and reinforcement learning algorithms and the associated typical applications in predictive maintenance. Furthermore, this paper summarizes the four critical steps of machine learning applications in maintenance prediction. Finally, the author proposes the future researches concerning how to utilize machine learning to optimize maintenance prediction and planning, improve equipment reliability, and achieve the best possible benefit. This article is available in the ASME Digital Collection at .

Many fault detection, optimization, and control logic methods rely on sensor feedback that assumes the system is operating at steady-state conditions, despite persistent transient disturbances. While filtering and signal processing techniques can eliminate some transient effects, this paper proposes an equilibrium prediction method for first-order dynamic systems using an exponential regression. This method is particularly valuable for many commercial and industrial energy system, whose dynamics are dominated by first-order thermo-fluid effects. To illustrate the basic advantages of the proposed approach, Monte Carlo simulations are used. This is followed by three distinct experimental case studies to demonstrate the practical efficacy of the proposed method. First, the ability to predict the carbon dioxide level in classrooms allows for energy efficient control of the ventilation system and ensures occupant comfort. Second, predicting the optimal time to end the cool-down of an industrial sintering furnace allows for maximum part throughput and worker safety. Finally, fault detection and diagnosis (FDD) methods for air conditioning systems typically use static system models; however, the transient response of many air conditioning signals may be approximated as first-order, and therefore, the prediction model enables the use of static fault detection methods with transient data (a need that has not been addressed in over 20 years of air conditioning FDD research). In this paper, the equilibrium prediction method's performance will be quantified using both Monte Carlo simulations and case studies. This article is available in the ASME Digital Collection at .

The purpose of this work is to develop a computationally efficient model of viral spread that can be utilized to better understand the influences of stochastic factors on a large-scale system – such as the air traffic network. A particle-based model of passengers and seats aboard a single-cabin 737-800 is developed for use as a demonstration of the concept on tracking the propagation of a virus through the aircraft's passenger compartment over multiple flights. The model is sufficiently computationally efficient to be viable for Monte Carlo simulation to capture various stochastic effects, such as number of passengers, number of initially sick passengers, seating locations of passengers, and baseline health of each passenger. The computational tool is then exercised in demonstration for assessing risk mitigation of intervention strategies, such as passenger-driven cleaning of seating environments and elimination of middle seating. This article is available in the ASME Digital Collection at .

Standup high lift trucks have an open operator compartment designed to allow for easy egress from the truck in the event of an off-dock, off-trailer, or tip-over accident. The open compartment design also protects the operator during nontip-over collision accidents, provided the operator maintains him or herself within the compartment before impact. However, it has been suggested that an additional rear operator guard would better protect operators during nontip-over collision accidents. This study aimed to investigate operator injury potential and kinematics during off-dock and off-trailer accidents where the presence of a rear operator guard prevents the operator from exiting the truck and to evaluate the potential for an operator to restrain him or herself within the operator compartment of a standup lift truck during an off-dock or off-trailer accident. An instrumented anthropomorphic test device (ATD) was subjected to a series of off-dock accidents and a single off-trailer accident utilizing three different standup lift trucks. The kinematics and loads experienced by the ATD were quantified and correlated to human injury potential. Additionally, this study examined the potential for a rear operator guard to impede operator egress during an off-dock or off-trailer accident. Testing demonstrated the potential for serious or greater injury to an operator subjected to an off-dock or off-trailer accident. The authors conclude that a rear guard would negatively impact the overall safety of a standup lift truck in most working environments. This article is available in the ASME Digital Collection at .

Dry powder inhalers (DPI), used as a means for pulmonary drug delivery, typically contain a combination of active pharmaceutical ingredients (API) and significantly larger carrier particles. The microsized drug particles—which have a strong propensity to aggregate and poor aerosolization performance—are mixed with significantly large carrier particles that cannot penetrate the mouth-throat region to deagglomerate and entrain the smaller API particles in the inhaled airflow. Therefore, a DPI's performance depends on the carrier-API combination particles' entrainment and the time and thoroughness of the individual API particles' deagglomeration from the carrier particles. Since DPI particle transport is significantly affected by particle-particle interactions, particle sizes and shapes present significant challenges to computational fluid dynamics (CFD) modelers to model regional lung deposition from a DPI. We employed the Particle-In-Cell method for studying the transport/deposition and the agglomeration and deagglomeration for DPI carrier and API particles in the present work. The proposed development will leverage CFD-PIC and sensitivity analysis capabilities from the Department of Energy laboratories: Multiphase Flow Interface Flow Exchange and Dakota UQ software. A data-driven framework is used to obtain the reliable low order statics of the particle's residence time in the inhaler. The framework is further used to study the effect of drug particle density, carrier particle density and size, fluidizing agent density and velocity, and some numerical parameters on the particles' residence time in the inhaler. This article is available in the ASME Digital Collection at .

In this paper, a four-step safety integrity level (SIL) analysis is developed to analyze numerous components of the subsea control system based on the OREDA database. For this purpose, initially, a failure mode classification table is provided, aiming to identify the number of dangerous as well as the prevalent failure modes. Then, several parameters including component inventory, installation inventory, equipment units, failure events, and total time in service (per hour) are calculated. Also, failure rates are evaluated and the software checking is provided as the final step of the proposed framework. The results show that the dangerous failure modes occur in about 80 cases, and the leakage in the closed position is the failure mode occurring more than the other ones. Moreover, calculations indicate that the process isolation valve of the subsea X-mass tree has the largest values in all the referred parameters. Moreover, hydraulic coupling of the choke module and HC leak sensor are the two components with the highest value of failure rates. This article is available in the ASME Digital Collection at .

Based upon a novel control-based dynamic modeling framework, this paper proposes two new indicators, i.e., resilience by mitigation and resilience by recovery, for the resilience analysis of interdependent critical infrastructures (ICIs) under disruptions. The former is built from the protection activities before and during the mitigation phase of a disruptive event, and the latter is the result of the restoration efforts, which take place at the recovery phase. The total resilience of ICIs combines both of these two aspects by taking into account the preferences of the decision makers. We demonstrate the applicability of the proposed modeling framework and metrics in a case study involving ICIs made of a power grid and a gas distribution system. Owing to the new resilience indicators, the priorities of subsystems and links within ICIs at different phases can be ranked; therefore, different resilience strategies at different phases of disruptive events are compared. The results show that proposed metrics can be used by stakeholders of ICIs on improving the effectiveness of system protection measurements. This article is available in the ASME Digital Collection at .

Since the last decade, gearbox systems have been requiring increasing power, and consequently, the complexity of systems has escalated. Inevitably, this complexity has resulted in the need for the troubleshooting of gearbox systems. With a growing trend of health monitoring in rotating machines, diagnostic and prognostic studies have become focused on diagnosing existing and potential failures in gearbox systems. In this context, this study develops the architecture of the cloud-based cyber-physical system (CPS) for condition monitoring of gearbox. Empirically collected vibration signals of gear wear at various time intervals are processed using empirical mode decomposition (EMD) algorithm. A Euclidian-based distance evaluation technique is applied to select the most sensitive features of car gear wear. Artificial neural network (ANN) is trained using extracted features to monitor the gearbox for the future dataset. Comparison of the performance results revealed that the ANN is superior to the other EMD methods. The present methodology was found efficient and reliable for condition monitoring of industrial gearbox. This article is available in the ASME Digital Collection at .

The 6061-T6 10-gauge sheet-type flat fatigue specimen was designed, manufactured, and tested on an Instron 8081 fatigue test machine. The fatigue tests were performed under five different cyclic axial loads with 195 tests. This paper will display how to use these 195 test data under different cyclic axial loads to build a probabilistic fatigue damage model. The model is validated by the results obtained from the traditional P-S-N curve approach. One advantage of this probabilistic fatigue damage model is that it can calculate the reliability of a component under any type of cyclic loads at any cyclic load level. This article is available in the ASME Digital Collection at .