Browsing by Author "Modarres, Mohammad"

Now showing 1 - 7 of 7
  • Thumbnail Image
    Item
    A Deep Adversarial Approach Based on Multi-Sensor Fusion for Semi-Supervised Remaining Useful Life Prognostics
    (MDPI, 2019-12-27) Verstraete, David; Droguett, Enrique; Modarres, Mohammad
    Multi-sensor systems are proliferating in the asset management industry. Industry 4.0, combined with the Internet of Things (IoT), has ushered in the requirements of prognostics and health management systems to predict the system’s reliability and assess maintenance decisions. State of the art systems now generate big machinery data and require multi-sensor fusion for integrated remaining useful life prognostic capabilities. When dealing with these data sets, traditional prediction methods are not equipped to handle the multiple sensor signals in unison. To address this challenge, this paper proposes a new, deep, adversarial approach to any remaining useful life prediction in which a novel, non-Markovian, variational, inference-based model, incorporating an adversarial methodology, is derived. To evaluate the proposed approach, two public multi-sensor data sets are used for the remaining useful life prediction applications: (1) CMAPSS turbofan engine dataset, and (2) FEMTO Pronostia rolling element bearing data set. The proposed approach obtains favorable results when against similar deep learning models.
  • Thumbnail Image
    Item
    A Thermodynamic Entropy Approach to Reliability Assessment with Applications to Corrosion Fatigue
    (MDPI, 2015-10-16) Imanian, Anahita; Modarres, Mohammad
    This paper outlines a science-based explanation of damage and reliability of critical components and structures within the second law of thermodynamics. The approach relies on the fundamentals of irreversible thermodynamics, specifically the concept of entropy generation as an index of degradation and damage in materials. All damage mechanisms share a common feature, namely energy dissipation. Dissipation, a fundamental measure for irreversibility in a thermodynamic treatment of non-equilibrium processes, is quantified by entropy generation. An entropic-based damage approach to reliability and integrity characterization is presented and supported by experimental validation. Using this theorem, which relates entropy generation to dissipative phenomena, the corrosion fatigue entropy generation function is derived, evaluated, and employed for structural integrity and reliability assessment of aluminum 7075-T651 specimens.
  • Thumbnail Image
    Item
    An Entropy-Based Damage Characterization
    (MDPI, 2014-12-05) Amiri, Mehdi; Modarres, Mohammad
    This paper presents a scientific basis for the description of the causes of damage within an irreversible thermodynamic framework and the effects of damage as observable variables that signify degradation of structural integrity. The approach relies on the fundamentals of irreversible thermodynamics and specifically the notion of entropy generation as a measure of degradation and damage. We first review the state-of-the-art advances in entropic treatment of damage followed by a discussion on generalization of the entropic concept to damage characterization that may offers a better definition of damage metric commonly used for structural integrity assessment. In general, this approach provides the opportunity to described reliability and risk of structures in terms of fundamental science concepts. Over the years, many studies have focused on materials damage assessment by determining physics-based cause and affect relationships, the goal of this paper is to put this work in perspective and encourage future work of materials damage based on the entropy concept.
  • Thumbnail Image
    Item
    Big Machinery Data Preprocessing Methodology for Data-Driven Models in Prognostics and Health Management
    (MDPI, 2021-10-14) Cofre-Martel, Sergio; Lopez Droguett, Enrique; Modarres, Mohammad
    Sensor monitoring networks and advances in big data analytics have guided the reliability engineering landscape to a new era of big machinery data. Low-cost sensors, along with the evolution of the internet of things and industry 4.0, have resulted in rich databases that can be analyzed through prognostics and health management (PHM) frameworks. Several data-driven models (DDMs) have been proposed and applied for diagnostics and prognostics purposes in complex systems. However, many of these models are developed using simulated or experimental data sets, and there is still a knowledge gap for applications in real operating systems. Furthermore, little attention has been given to the required data preprocessing steps compared to the training processes of these DDMs. Up to date, research works do not follow a formal and consistent data preprocessing guideline for PHM applications. This paper presents a comprehensive step-by-step pipeline for the preprocessing of monitoring data from complex systems aimed for DDMs. The importance of expert knowledge is discussed in the context of data selection and label generation. Two case studies are presented for validation, with the end goal of creating clean data sets with healthy and unhealthy labels that are then used to train machinery health state classifiers.
  • Thumbnail Image
    Item
    Conference Proceedings Report: ASME-SERAD and UMD-CRR Interactive seminar & pre-workshop on the intersection of PRA and PHM
    (2020-10) Groth, Katrina; Pourgol-Mohammad, Mohammad; Modarres, Mohammad
    This event is the first in a two-part series exploring the intersection of PRA and PHM in the context of complex engineering systems. Initially the workshop was planned as a fully in-person workshop to be held in April, 2020, but as with many events in 2020, it was postponed due to the travel restrictions resulting from COVID-19 pandemic. The organizers recognized that the online format isn't amenable to the deep discussions which were intended to be at the heart of the in person workshop, but we decided to try an experiment: to see if we could make a “pre-workshop” as interactive possible in an era of webinar fatigue. Thus the workshop was reimagined as an online, interactive pre-workshop in 2020, to be followed with the in person, discussion-heavy workshop to be held when we are able to travel again in 2021.
  • Thumbnail Image
    Item
    Damage Assessment Using Information Entropy of Individual Acoustic Emission Waveforms during Cyclic Fatigue Loading
    (MDPI, 2017-05-30) Sauerbrunn, Christine M.; Kahirdeh, Ali; Yun, Huisung; Modarres, Mohammad
    Information entropy measured from acoustic emission (AE) waveforms is shown to be an indicator of fatigue damage in a high-strength aluminum alloy. Three methods of measuring the AE information entropy, regarded as a direct measure of microstructural disorder, are proposed and compared with traditional damage-related AE features. Several tension–tension fatigue experiments were performed with dogbone samples of aluminum 7075-T6, a commonly used material in aerospace structures. Unlike previous studies in which fatigue damage is measured based on visible crack growth, this work investigated fatigue damage both prior to and after crack initiation through the use of instantaneous elastic modulus degradation. Results show that one of the three entropy measurement methods appears to better assess the damage than the traditional AE features, whereas the other two entropies have unique trends that can differentiate between small and large cracks.
  • Thumbnail Image
    Item
    Measures of Entropy to Characterize Fatigue Damage in Metallic Materials
    (MDPI, 2019-08-17) Yu, Huisung; Modarres, Mohammad
    This paper presents the entropic damage indicators for metallic material fatigue processes obtained from three associated energy dissipation sources. Since its inception, reliability engineering has employed statistical and probabilistic models to assess the reliability and integrity of components and systems. To supplement the traditional techniques, an empirically-based approach, called physics of failure (PoF), has recently become popular. The prerequisite for a PoF analysis is an understanding of the mechanics of the failure process. Entropy, the measure of disorder and uncertainty, introduced from the second law of thermodynamics, has emerged as a fundamental and promising metric to characterize all mechanistic degradation phenomena and their interactions. Entropy has already been used as a fundamental and scale-independent metric to predict damage and failure. In this paper, three entropic-based metrics are examined and demonstrated for application to fatigue damage. We collected experimental data on energy dissipations associated with fatigue damage, in the forms of mechanical, thermal, and acoustic emission (AE) energies, and estimated and correlated the corresponding entropy generations with the observed fatigue damages in metallic materials. Three entropic theorems—thermodynamics, information, and statistical mechanics—support approaches used to estimate the entropic-based fatigue damage. Classical thermodynamic entropy provided a reasonably constant level of entropic endurance to fatigue failure. Jeffreys divergence in statistical mechanics and AE information entropy also correlated well with fatigue damage. Finally, an extension of the relationship between thermodynamic entropy and Jeffreys divergence from molecular-scale to macro-scale applications in fatigue failure resulted in an empirically-based pseudo-Boltzmann constant equivalent to the Boltzmann constant.