M.S. Lebold

Hybrid reasoning for prognostic learning in CBM systems

Reasoning systems that integrate explicit knowledge with implicit information are essential for high performance decision support in condition-based maintenance and prognostic health management applications. Such reasoning systems must be capable of learning the specific features of each machine during its life cycle. In this paper, a hybrid reasoning approach that is capable of integrating domain knowledge and test and operational data from the machine is described. This approach is illustrated with an industrial gearbox example. In this approach explicit domain knowledge is expressed as a rule-base and used to train a feedforward neural network. The training process results in a parsimonious representation of the explicit knowledge by combining redundant rules. A significant added practical benefit of this process is that it also is able to identify logical inconsistencies in the rule-base. Such inconsistencies are notorious in causing deadlock in large-scale expert systems. The neural network can be periodically updated with test and operational data to adapt the network to each specific machine. The flexibility and efficiency of this hybrid approach make it very suitable for practical health management systems designed to operate in a distributed environment.

Using torsional vibration analysis as a synergistic method for crack detection in rotating equipment

A non-intrusive torsional vibration method for monitoring and tracking small changes in crack growth of reactor coolant pump shafts is presented in this paper. This method resolves and tracks characteristic changes in the natural torsional vibration frequencies that are associated with shaft crack propagation. The focus of this effort is to develop and apply the torsional vibration shaft cracking monitoring technique on a Westinghouse 93A reactor coolant pump. While this technique is being applied to reactor coolant pumps, it is generally applicable to many types of rotating equipment, including centrifugal charging pumps, condensate and feed water pumps, and may be used to detect and track changes in blade natural frequencies in gas or steam turbines. A laboratory scale rotor test bed was developed to investigate shaft cracking detection techniques under controlled conditions. The test bed provides a mechanism to evaluate sensing technologies and algorithm development. For accurate knowledge of the crack characteristics (crack depth and front), a sample shaft was seeded with a crack that was propagated using a three-point bending process. Following each crack growth step, the specimen was evaluated using ultrasonic inspection techniques for crack characterization. After inspection, the shaft was inserted in the rotor test bed for analysis and to track changes in shaft torsional vibration features. The torsional vibration measurement method has demonstrated the ability to reliably detect changes in the first natural shaft frequency in the range of 0.1 to 0.2 Hz. This technique shows the potential to enable online structural health diagnostics and ultimately the prevention of shaft or even possibly blade failure due to crack growth.

Utilizing dcom in an open system architecture framework for machinery monitoring and diagnostics

A communications framework for nextgeneration machinery monitoring and diagnostic systems has been addressed through the Open System Architecture for Condition Based Maintenance (OSA-CBM) development team. This framework employs the use of an open system architecture standard utilizing a distributed software model approach. The OSA-CBM team investigated different middleware technologies and defined data models in C O D A and COMIDCOM as well as XML. Even though IDL files are provided, the protocol logic or skeleton code isn’t defined; only conceptual descriptions are provided. The Applied Research Laboratory (a) has taken the COIWDCOM IDL files along with supporting OSA-CBM documentation and has developed skeleton code to implement a functional CBM system. This work focused on writing the necessary handshaking and data management code needed for communications between OSA-CBM modules. From this work, a user can develop their own working version and expand the theory to all other OSA-CBM layers. This paper provides an overview of the OSA-CBM development models and insight into a DCOMbased OSA-CBM implementation. TABLE OF CONTENTS

Wireless Technology Study and the Use of Smart Sensors for Intelligent Control and Automation

With advances in processing power of embedded smart sensors, it is possible to incorporate enough intelligence into a small package to enhance an existing platform while reducing data flow requirements on legacy systems. The addition of a wireless interface to these sensors is possible as well. This paper covers information on current wireless technologies and examples on how smart wireless sensors can be used to develop diagnostic and prognostic wireless health monitoring nodes. This study provides the knowledge needed to design the wireless infrastructure for a wireless health monitoring system. This paper presents the current wireless technologies which may be used for wireless sensors and the advantages and disadvantages of each standard. The technology presented within this paper was successfully demonstrated on various health monitoring applications. The last section of this paper shows a few wireless health monitoring applications developed by the applied research laboratory. The first application demonstrates a Windows based application developed on a PC-104 platform for monitoring fuel system health and reconfigures system for optimal usage of pump and filters. The other applications were developed using a SHARC-based system for providing diagnostic and prognostic assessments of gearbox systems

Diagnostic fault detection for internal combustion engines via pressure curve reconstruction

One proven technique for monitoring the health of a sealed internal combustion engine is to analyze the combustion pressure cycle curves of the individual cylinders. Current techniques for doing this require a pressure sensor mounted directly in the combustion chamber. This necessitates maintenance and design considerations that may be unacceptable especially on legacy systems. This paper describes a non-invasive technique developed for monitoring combustion pressure cycle related faults. This method has been developed and tested on a diesel engine test bed at Penn State Universitys Applied Research Laboratoq (ARL), Condition Based Maintenance Department. The diesel engine test bed was used to gather all forms of data under different engine operating conditions. Using crdshaft angular velocity data fiom a high-resolution encoder, a trained neural network is used to reconstruct the combustion pressure cycle curves. These reconstructed combustion pressure curves are then passed into another trained neural network for fault detection analysis.

Diagnostic End to End Monitoring a Fault Detection for Braking Systems

The reliable and effective performance of a braking system is fundamental to the operation of most vehicles. Any failure in the braking system that impacts the ability to retard a vehicles motion have an immediate and frequently catastrophic effect on a vehicles safety. There are very few diagnostic systems that monitor the health of multiple individual components in a braking system and even fewer that can automatically detect early stage failures and component wear with the reliability required for such an important vehicle subsystem. The systems that can accomplish these tasks are restricted in application due to size, weight and there own unique maintenance loads. Current diagnostic techniques call for careful maintenance of braking system components. These techniques tend to be backward thinking in that they are based on previous experience which is not always a good indicator for future systems. In addition the need to perform constant maintenance on the braking system puts restrictions on its design and heavy loads on maintenance personnel. This paper describes a technique developed at the Penn State Applied Research Laboratories Systems Operations and Automation Department for low physical impact fault detection and monitoring of hydraulic vehicular braking systems without compromising reliability. This technique is being applied to severe duty tactical vehicles with brake systems that have maintenance accessibility issues. The brake monitoring system (BMS) directly monitors the linings of the friction surfaces for wear using environment tolerant sensors. Additionally using pressure sensors, level sensors and high band width data gathering the hydraulic system is monitored for transients that are consistent with early stage failures

Diagnostic fault detection & intelligent reconfiguration of fuel delivery systems

The reliable operation of an engines fuel delivery system is fundamental. A failure in the fuel system that impacts the ability to deliver fuel to the engine will have an immediate effect on system performance and safety. There are very few diagnostic systems that monitor the health of the fuel system and even fewer that can accommodate for detected faults. Current diagnostic techniques call for careful maintenance of fuel system components. These techniques tend to be backward thinking in that they are based on previous experience which is not always a good indicator for future systems. This paper describes a technique developed at the Penn State Applied Research Laboratories Condition Based Maintenance Department for fault detection and reconfiguration for fuel delivery system components. This technique has been applied to a diesel engine test rig. The test rig is fully instrumented with sensors including those for fuel pressure. Even though this technique is being applied on a diesel engine, the approach is fully compatible to any fuel delivery system

Platform degrader analysis for the complex systems for the design and application of condition based maintenance

The US Armys Heavy Brigade Combat Team (HBCT) is investigating the implementation of Vehicle Health Management Systems across its fleet of platforms. A vehicle degrader analysis was conducted to determine where the application of vehicle health management systems (VHMS) could potentially have the greatest impact on increasing maintainer effectiveness and operational availability. The analysis was conducted for M1 Abrams tanks and M2/M3 Bradley Fighting Vehicles using data from three sources: 1) multi-year statistical data on repairs, operating and support costs, and logistics data, 2) interviews with maintainers, service representatives and operators, and 3) questionnaires completed by the platform OEMs. The paper will describe how these alternative data and information sources were used when ideal data and information sources were not available. The statistical data provided an indication of vehicle consumable and repairable cost drivers, part replacement quantity and some indication of component failure isolation for vehicle systems. The interviews with vehicle maintainers, field service representatives and operators provided insight into platform reliability, maintainability and human factors issues. The maintainability, reliability and diagnostic coverage information from the OEMs, provided information about mission criticality and general diagnostic capability for specific vehicle sub-systems and components. The results of the degrader analysis include a list of components and sub-systems that contribute most to maintainability, reliability and vehicle operational availability issues. The degrader results, were used to develop the conceptual design of the optimum application of embedded vehicle diagnostics and prognostics for more effective and efficient HCBT platform maintenance, logistic support and platform life cycle management.

Minimally Intrusive Torsional Vibration Sensing on Rotating Shafts

Time Interval Measurement System (TIMS) is a common torsional vibration sensing method used for rotating equipment. The technique uses high-speed counters to detect “zero crossings” in a carrier signal that is generated by a multiple pulse per revolution encoder on the shaft. The “zero crossings” are based on the passage timing of discrete intervals from an incremental geometric encoder (i.e., gear, optical encoder) on a rotating shaft. A variety of transducers have been used to sense the encoder interval passages, including Hall effect and reflective light intensity transducers. The encoder and sensor require physical attachment to the rotating shaft and surrounding hardware. Furthermore, the combinations must work in concert with each other. Usually the sensing transducer requires precise positioning with respect to the shaft encoder. The physical attachment and installation of the encoding and sensing devices can range from problematic to extremely difficult depending on the application. To make sensing of torsional vibration on a rotating shaft easier and minimally intrusive a combination of an adhesive backed “zebra” tape and a specially modified laser tachometer is used. The laser tachometer is considerably easier as permits a greater range of standoff, targeting and mounting options. Tests are performed a on mechanical diagnostics test bed with a 30 hp electrical drive connected to a 70 hp load motor by a shaft and couplers to demonstrate the issues. Torsional vibration measurements are acquired with the laser tachometer-“zebra” tape, Hall effect-gear and a precision optical encoder. Results are presented and discussed from the various torsional vibration sensing systems. The application illustrates the ease of set up for the laser tachometer-“zebra tape” combination and the high data quality that can be obtained.

Using wireless sensor technologies for sense and respond logistic applications

Our team developed and demonstrated a new system design approach that can diagnosis and track overall machinery health, inform operators shortly before maintenance action is needed, place orders for parts needed to perform the specific repair or maintenance action and schedule needed manpower for repair action. This embedded wireless system was integrated into legacy Navy supply and maintenance systems. This paper discusses the wireless device, architecture, and processing used to meet the new objective of a sense and respond applications