A Coupled Approach to Developing Damage Prognosis Solutions

Sohn, Hoon; Farrar, Charles R.; Hemez, François M.; Park, Gyuhae; Robertson, Amy; Williams, Todd O.

doi:10.4028/www.scientific.net/kem.245-246.289

Cited by 10 publications

(5 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If no such model exists the damage metrics have to be determined by calibration experiments. It should be mentioned that sometimes the determination of the type of damage is included as an extra step between Level II and III [4,5,6]. The highest, and most sophisticated level, is the prognosis of the remaining lifetime.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Vibration-Based Structural Health Monitoring – Concepts and Applications

Fritzen

2005

KEM

View full text Add to dashboard Cite

This paper gives an overview on the current status of vibration-based methods for Structural Health Monitoring. All these methods have in common that a structural change due to a damage results in a more or less pronounced change of the dynamic behavior. The use of modal information is discussed, as well as the direct use of forced and ambient vibrations. From this information, different strategies can be deduced which depend on the type of measurement data (time/frequency domain) but also on the frequency spectrum. The incorporation of actuation and sensing devices into the structure leads to modern concepts of Smart Structural Health Monitoring. Examples from civil and aerospace engineering show the applicability of these methods.

show abstract

Section: Introductionmentioning

confidence: 99%

“…Due to lack of space this article does not claim to give a complete literature compilation. An overview can be obtained by [4,6,8,9,10]. However, the paper intends to shortly discuss the basic physical ideas of different vibration-based methods.…”

Section: Introductionmentioning

confidence: 99%

Vibration-Based Structural Health Monitoring – Concepts and Applications

Fritzen

2005

KEM

View full text Add to dashboard Cite

show abstract

“…In some situations, especially when a failure is potentially catastrophic, it is more than desirable to predict the probability that an equipment operates without fault or failure up to some future point in time (the next inspection interval), given the current equipment condition and its past operational profile. In any situation, the probability that equipment operates without a fault until the next inspection interval is a good reference point for a maintenance engineer to determine the appropriateness of the inspection interval [102,103]. However, this is not a popular method due to the limitation in gathering the required data for prediction.…”

Section: Fault Prognosismentioning

confidence: 99%

Predictive Maintenance of an External Gear Pump using Machine Learning Algorithms

Lakshmanan¹

View full text Add to dashboard Cite

The importance of Predictive Maintenance is critical for engineering industries, such as manufacturing, aerospace and energy. Unexpected failures cause unpredictable downtime, which can be disruptive and high costs due to reduced productivity. This forces industries to ensure the reliability of their equip-ment. In order to increase the reliability of equipment, maintenance actions, such as repairs, replacements, equipment updates, and corrective actions are employed. These actions aﬀect the ﬂexibility, quality of operation and manu-facturing time. It is therefore essential to plan maintenance before failure occurs.Traditional maintenance techniques rely on checks conducted routinely based on running hours of the machine. The drawback of this approach is that maintenance is sometimes performed before it is required. Therefore, conducting maintenance based on the actual condition of the equipment is the optimal solu-tion. This requires collecting real-time data on the condition of the equipment, using sensors (to detect events and send information to computer processor).Predictive Maintenance uses these types of techniques or analytics to inform about the current, and future state of the equipment. In the last decade, with the introduction of the Internet of Things (IoT), Machine Learning (ML), cloud computing and Big Data Analytics, manufacturing industry has moved forward towards implementing Predictive Maintenance, resulting in increased uptime and quality control, optimisation of maintenance routes, improved worker safety and greater productivity.The present thesis describes a novel computational strategy of Predictive Maintenance (fault diagnosis and fault prognosis) with ML and Deep Learning applications for an FG304 series external gear pump, also known as a domino pump. In the absence of a comprehensive set of experimental data, synthetic data generation techniques are implemented for Predictive Maintenance by perturbing the frequency content of time series generated using High-Fidelity computational techniques. In addition, various types of feature extraction methods considered to extract most discriminatory informations from the data. For fault diagnosis, three types of ML classiﬁcation algorithms are employed, namely Multilayer Perceptron (MLP), Support Vector Machine (SVM) and Naive Bayes (NB) algorithms. For prognosis, ML regression algorithms, such as MLP and SVM, are utilised. Although signiﬁcant work has been reported by previous authors, it remains diﬃcult to optimise the choice of hyper-parameters (important parameters whose value is used to control the learning process) for each speciﬁc ML algorithm. For instance, the type of SVM kernel function or the selection of the MLP activation function and the optimum number of hidden layers (and neurons).It is widely understood that the reliability of ML algorithms is strongly depen-dent upon the existence of a suﬃciently large quantity of high-quality training data. In the present thesis, due to the unavailability of experimental data, a novel high-ﬁdelity in-silico dataset is generated via a Computational Fluid Dynamic (CFD) model, which has been used for the training of the underlying ML metamodel. In addition, a large number of scenarios are recreated, ranging from healthy to faulty ones (e.g. clogging, radial gap variations, axial gap variations, viscosity variations, speed variations). Furthermore, the high-ﬁdelity dataset is re-enacted by using degradation functions to predict the remaining useful life (fault prognosis) of an external gear pump.The thesis explores and compares the performance of MLP, SVM and NB algo-rithms for fault diagnosis and MLP and SVM for fault prognosis. In order to enable fast training and reliable testing of the MLP algorithm, some predeﬁned network architectures, like 2n neurons per hidden layer, are used to speed up the identiﬁcation of the precise number of neurons (shown to be useful when the sample data set is suﬃciently large). Finally, a series of benchmark tests are presented, enabling to conclude that for fault diagnosis, the use of wavelet features and a MLP algorithm can provide the best accuracy, and the MLP al-gorithm provides the best prediction results for fault prognosis. In addition, benchmark examples are simulated to demonstrate the mesh convergence for the CFD model whereas, quantiﬁcation analysis and noise inﬂuence on training data are performed for ML algorithms.

show abstract

“…In order to accurately assess safety conditions of existing infrastructure systems, and to identify structural vulnerabilities to extreme events, strong interests in various structural health monitoring technologies have been growing rapidly in recent years (Sohn, et al 2003). Structural health monitor can provide insights into the real performance of a structure, and offer empirical data for refining structural models and existing building codes.…”

Section: Figure 11 Scene Of the Collapse Of The Minneapolis's I-35w mentioning

confidence: 99%

Wireless Sensing System for Load Testing and Rating of Highway Bridges

Luo¹

View full text Add to dashboard Cite

Structural capacity evaluation of bridges is an increasingly important topic in the effort to deal with the deteriorating infrastructure. Most bridges are evaluated through subjective visual inspection and conservative theoretical rating. Diagnostic load test has been recognized as an effective method to accurately assess the carrying capacity of bridges. Traditional wired sensors and data acquisition (DAQ) systems suffer drawbacks of being labor intensive, high cost, and time consumption in installation and maintenance. For those reasons, very few load tests have been conducted on bridges. This study aims at developing a low-cost wireless bridge load testing & rating system that can be rapidly deployed on bridges for structural evaluation and load rating. Commercially available wireless hardware is integrated with traditional analogue sensors and the appropriate rating software is developed. The wireless DAQ system can work with traditional strain gages, accelerometers as well as other voltage producing sensors. A wireless truck position indicator (WVPI) is developed and used for measuring the truck position during load testing. The software is capable of calculating the theoretical rating factors based on AASHTO Load Resistance Factor Rating (LRFR) codes, and automatically produces the adjustment factor through load testing data. A simplified finite element model was used to calculate deflection & moment distribution factors in order to reduce the amount of instrumentation used in field tests. The system was used to evaluate the structural capacity of Evansville Bridge in Preston County, WV. The results show that the wireless bridge load testing & rating system can effectively be implemented to evaluate the real capacity of bridges with remarkable advantages: low-cost, fast deployment and smaller crew. I would like to express my deepest gratitude and respect for my advisor, Professor Samir Shoukry, for supporting my research and helping me find an area of intellectual pursuit that fit my abilities and interests well. His endless contribution of ideas and insights resulted in bettering my efforts to design the wireless bridge load testing & rating system. Having learned under his tutelage helped me to grow both intellectually and as a person. Many thanks are owed to Dr. Mourad Riad and Dr. Gergis William, who provided me with lots of selfless help and precious experiences in civil engineering and bridges. I would like to thank Professor Michael Palmer for unselfishly sharing with me his encyclopedic knowledge of printed circuit board. My thanks also go to Mr. Hermman Alcazar and Mr. Joseph Greenfeder for helping me with the field tests. I recall the great fun and experience that I gained when working with Mr. Musat Crihalmeanu. Especial thanks to my friend Joe Watring for some of the best experiences of daily life in Morgantown, WV. Lastly, I would like to thank the committee members for this research. Each of you has provided me with your valuable time and precious advices and for that I am indebted to you. iii

show abstract

A Coupled Approach to Developing Damage Prognosis Solutions

Cited by 10 publications

References 35 publications

Vibration-Based Structural Health Monitoring – Concepts and Applications

Vibration-Based Structural Health Monitoring – Concepts and Applications

Predictive Maintenance of an External Gear Pump using Machine Learning Algorithms

Wireless Sensing System for Load Testing and Rating of Highway Bridges

Contact Info

Product

Resources

About