Modelling, Monitoring and Diagnostic Techniques for Fluid Power Systems

Transcription

1 Modelling, Monitoring and Diagnostic Techniques for Fluid Power Systems

2 John Watton Modelling, Monitoring and Diagnostic Techniques for Fluid Power Systems 123

3 John Watton, BSc, PhD, DSc, CEng, FIMechE Institute of Machines and Structures Cardiff School of Engineering Cardiff University Cardiff CF24 3AA Wales UK British Library Cataloguing in Publication Data Watton, John Modelling, monitoring and diagnostic techniques for fluid power systems 1.Fluid power technology 2.Machinery - Monitoring 3.Fluid power technology - Mathematical models I.Title 620.1'06 ISBN-13: ISBN-10: Library of Congress Control Number: ISBN e-isbn Printed on acid-free paper Springer-Verlag London Limited 2007 MATLAB and Simulink are the registered trademarks of The MathWorks, Inc., 3 Apple Hill Drive, Natick, MA , USA. Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore free for general use. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made Springer Science+Business Media springer.com

4 Preface This book is aimed towards the condition monitoring and diagnostics aspects of fluid power control and is intended to replace and significantly extend my earlier book published in 1992, the majority of copies being destroyed in a warehouse flood shortly after its publication. This salutary lesson in condition monitoring did not persuade the publisher to re-print the book and I suspect that a connection between the topic and the event was never really made. The net result of this has been that a quite different book has emerged. This is perhaps not surprising given the large amount of work on the subject that I and others have done over the intervening years since my first book on the subject was conceived. In particular the application of condition monitoring and fault diagnosis to hydraulic systems has undergone a great deal of experimental validation in research laboratories and on real plant and I have taken much delight from my industrial collaborations. However, research and development on condition monitoring and fault diagnosis of fluid power systems is still not significantly popular in universities. This is not helped by the relative absence of fluid power as a mainstream subject at undergraduate level. This is quite remarkable when one thinks of the diversity of applications from primary materials processing and manufacturing to mobile machines, modern automotive and aerospace engineering. Quite simply there are a vast number of areas where fluid power is the only viable solution. I propose therefore that many undergraduate and graduate engineers are missing a most exciting subject area that involves: new materials technologies solid mechanics, fluid mechanics and thermodynamics component, systems and machine static and dynamic design modern control theory and computer control techniques artificial intelligence techniques new sensor technologies signal processing and algorithms component and systems modelling and simulation condition monitoring and fault diagnosis

5 vi Preface It is rare that a project does not involve several of these aspects and a fluid power engineer really now needs a systems, perhaps mechatronics, approach to new challenges. It is probably fortuitous that since my interest in condition monitoring began in the late 1980s, there has been a fruition of many concepts being applied to real systems from my point of view, and this book is intended to bring these together. Each chapter could be a book in its own right and this has presented a problem in trying to convey essential issues without becoming a research publication with all historical references. It is important to quote appropriate additional reading material but a line has had to be drawn regarding how much should be included; for example it is possible to include over 1000 references on transmission line theory alone. I have attempted to include important references with the view that other works tend, but not always, to be covered in the publications quoted. The reader will note some early references since formative work now tends to be ignored, or certainly not acknowledged, in modern publications. Chapter 1 deals with the general background of the subject matter, illustrating the reason for condition monitoring and some general principles that apply or can be applied. Hopefully this sets out the need for the ensuing chapters which then discuss some of the important details, in my opinion. Chapter 2 considers modelling and computer simulation with appropriate basic theory and its practical application. This chapter was particularly difficult to minimise to a sensible length but most of what is included is supported by practical results. It therefore forms a useful introduction to fluid power with some material taken from my first book on Fluid Power Systems published in 1989 but now out of print. The reader will note the absence of any detailed control theory, beyond some fundamental ideas, since this is not really necessary within the context of this book; also background theory in general is not exhaustive in this area for the same reason. More importantly, what has been included is material that I have found to be useful for real applications particularly from my work with the fluid power and manufacturing industries. The reader will note several applications of artificial neural networks and related ideas such as data-based modelling just hinted at in my previous book on condition monitoring. Chapter 3 considers condition monitoring methods where the theory of Chapter 2 is put into practice. In addition, pragmatic concepts of signal monitoring and processing are included to illustrate the practical reality of combining a sensible amount of both theory and intuition, perhaps experience. Some new algorithms developed at Cardiff are introduced here but an important message is again the practical limitations of each method and the fact that in reality several approaches should be tried to give confidence in the emerging diagnostic. Some sensor information is covered although this is not an attempt to overview the general field of condition monitoring, but only what I have found useful for my contribution to the subject. Perhaps the main themes of this chapter are pressure and flow monitoring, dynamic data analysis including vibration, and oil/wear debris analysis. This chapter also considers expert systems and knowledge-based reasoning. This is a fascinating area, quite complex and useful for situations such as multiple fault conditions, but is still enjoying only a modest

6 Preface vii evolution. It is intended to show how rules may be developed from some rather basic theoretical concepts, which then actually give a great deal of information on the probable fault state of the hydraulic circuit. Again many practical examples are used to illustrate the concepts, from simple drives and lifting systems to a seven-stand steel strip finishing mill. Chapter 4 gives many examples of component faults in pumps and motors taken from industrial sources on the cause and solutions for breakdowns that may occur in a hydraulic circuit. It considers many components and typical failures that have been deduced over many year of experience. A comprehensive list of books, papers and further reading is included together with a detailed Index. Finally I must thank my many industrial contacts and friends who have freely given information and provided funding and equipment that has helped me progress this subject matter. Many of my excellent PhD and EngD research students have contributed to the contents of this book and I congratulate them not only on their work but also on their common sense in selecting such an important research topic. Cardiff 2005 J. Watton

7 Contents 1 Introduction Why Implement Condition Monitoring? Three Maintenance Strategies Some Preliminary Conclusions Potential Benefits of CBM Benefits Applied to Plant Economics Types of Condition Monitoring Systems Methods of Condition Monitoring Failure Modes and Effects Analysis (FMEA) Correcting the Fault Fault Tree Analysis Computer Simulation as a Fault Synthesis/Detecting Tool Modelling and Computer Simulation as an Aid to Understanding Circuit Behaviour Introduction Steady-state Analysis of Components and Circuits Pumps and Motors Cylinders Leakage Flow and Lift Characteristics of Slippers Pressure Ripple in Positive Displacement Pumps and Motors Flow Restrictors Flow Rate and Flow Reaction Force Pressure Drop through Pipes Pressure/Flow Characteristics of Directional Valves, Check Valves, Flow Control Valves, Pressure Relief Valves A Circuit Calculation Example Electrohydraulic Servovalves Principles of Operation Steady-state Performance Characteristics Spool Underlap... 42

8 x Contents Spool Valve Linearised Coefficients Servovalve Dynamic Response Steady-state Control of a Servovalve/Motor Drive The Basic Circuit Open-loop Behaviour with Losses Closed-loop Behaviour with Losses Steady-state Motion of a Servovalve/Linear Actuator Drive The Basic Circuit The Extending Case The Retracting Case Undamped Natural Frequency of Actuators, the Effect of Fluid Compressibility and Load Mass Some General Observations on Line Pressures in Servovalve/ Actuator Control Systems in the Presence of Load Mass and Inertia Linearisation Technique to Estimate the Dynamic Behaviour of Nonlinear Systems Underlapped Servovalve Spools the Effect on Steady-state Behaviour for Closed-loop Control Systems Underlapped Servovalve Spools the Effect on Dynamic Behaviour for Closed-loop Control Systems Proportional Pressure Relief Valve Modelling Concepts Long Lines The Basic Equations Application to a Servovalve/Single-line/Cylinder Modelling Actuator Volume Effects Frequency Response Frequency Response with Actuator Included Response and Stability of a Servovalve/Motor Speed Control System with Long Lines Effective Bulk Modulus Fluid Viscosity and Density Solving the System Equations Simulation Software Transient Response and Stability of a Pressure Rate-controlled Two-stage Pressure Relief Valve Data-based Dynamic Modelling of Components Accumulator Charging Behaviour Data-based Dynamic Modelling of Components Time Series Analysis Data-based Dynamic Modelling of Components Long Lines Data-based Dynamic Modelling of Components Artificial Neural Networks (ANNs) Introduction The Artificial Neuron Forming the ANN

9 Contents xi Training ANNs The Basic Weight Change Equation for a Single Neuron Some Practical Issues Modelling a Long Line in a Pressure Control System Data-based Dynamic Modelling of Components the Group Method of Data Handling (GMDH) Introduction An Introductory Example a Servovalve/Motor Drive Application to a Practical Circuit a Multivariable Problem Circuit Simulation Using Interconnected Artificial Networks Intelligent Control, Improving Closed-loop Performance Condition Monitoring Methods Methods Based Around Steady-state Flow Loss Metering Visual Leak Detection, the Simplest Approach Flow Meter Types Vane Pump Wear Monitoring Assessing the External Leakage from a Pressurecompensated Axial Piston Pump Detection of Cylinder Piston Seal Wear for a Servovalve/ Cylinder Drive Assessing Motor Leakage Characteristics by Coupling to a Servovalve Cylinder Seal Leakage Identification Within a Vehicle Active Suspension Actuator Using Dynamic Data Fluid Borne (FBN), Structural Borne (SBN) and Air Borne (ABN) noise Introduction Application to Pumps, FBN and SBN Spectrum Analysis Fault Data Trending Acoustic Emission Sensor Application and Comparison for Pump Wear and Cavitation Detection, SBN and FBN Spectrum Analysis to Detect Piston Seal Wear within a Cylinder, SBN Frequency Spectrum Archetypes and Fault Correlation FBN Analysis of a Cylinder Pressure Control System with Leakage Component Impedance, FBN Analysis Use of Impedance for Fault Diagnosis of a Pressure Relief Valve Using Frequency Response

10 xii Contents SBN due to Repetitive Fault Phenomena ABN Concepts Time Encoded Signal Processing, TESP Analysis Introduction Example. Pump Torque Data Analysis Example. Application to 28 Servovalves on the Work Roll Bending Control System of a Seven-stand Hot Steel Strip Mill Combined TESP/ANN Approach to Leakage and Servovalve Fault Detection in a Pressure Control System Further Applications of Artificial Neural Networks Leakages Within a Position Control System Weight Changes and the Use of Transmission Line Dynamics to Aid Fault Detection Linear Prediction Coding and Cepstrum (LPC) Features Extraction and ANN Classification Condition Monitoring of a Bent Axis Pump Fluid and Wear Debris Analysis Consideration of the Fluid Wear and Particle Contamination Particle Size Classification, ISO4406, Automatic Particle Counters Wear Debris Analysis Temperature Sensing Data Acquisition Hand-held Equipment Distributed Sensing, Wireless Communication Data Acquisition Cards for System Development Application to a Hot Steel Strip Finishing Mill Expert Systems and Knowledge-based Reasoning What is an Expert System? Generating a Knowledge Base The Inferencing Process Supporting Software Knowledge Elicitation Working Examples Object-oriented Expert Systems Approach for Multiple Subsystems, and its Application to a Hot Steel Strip Rolling Mill

11 Contents xiii 4 Common Faults and Breakdowns that can occur in a Hydraulic Circuit Pumps and Motors Directional Valves Servovalves Check Valves Pressure Limiters Flow Regulators Anti-shock Valves Actuating Cylinders Filters Tanks Couplings Piping Accumulators Oil Cooler Miscellaneous Some Typical Component Failures, Pumps and Motors Further Reading Author Biography Index