Acoustic Emission - Standards and Technology Update

S T P 1353

Acoustic Emission: Standards and Technology Update

Sotirios J. Vahaviolos, editor

ASTM Stock #: STP1353

ASTM 100 Barr Harbor Drive West Conshohocken, PA 19428-2959 Printed in the U.S.A.

Library of Congress Cataloging-in-Publication Data Acoustic emission : standards and technology update / Sotirios J. Vahaviolos, editor. p. c m . - (STP : 1353) Includes bibliographical references. "ASTM Stock #: STP1353." ISBN 0-8031-2498-8 1. Acoustic emission testing. I. Vahaviolos, Sotirios J. II. Series: ASTM special technical publication : 1353. TA418.84 .A263 1999 620.1 '27--dc21

99-38512 CIP

Copyright 9 1999 AMERICAN SOCIETY FOR TESTING AND MATERIALS, West Conshohocken, PA. All rights reserved. This material may not be reproduced or copied, in whole or in part, in any printed, mechanical, electronic, film, or other distribution and storage media, without the written consent of the publisher.

Photocopy Rights Authorization to photocopy Items for internal, personal, or educational classroom use, or the Internal, personal, or educational classroom use of specific clients, Is granted by the American Society for Testing and Materials (ASTM) provided that the appropriate fee is paid to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923; Tel: 508-750-8400; online: http..//www.copyrig ht.com/.

Peer Review Policy Each paper published in this volume was evaluated by two peer reviewers and at least one editor. The authors addressed all of the reviewers' comments to the satisfaction of both the technical editor(s) and the ASTM Committee on Publications. To make technical information available as quickly as possible, the peer-reviewed papers in this publication were prepared "camera-ready" as submitted by the authors. The quality of the papers in this publication reflects not only the obvious efforts of the authors and the technical editor(s), but also the work of the peer reviewers. In keeping with long standing publication practices, ASTM maintains the anonymity of the peer reviewers. The ASTM Committee on Publications acknowledges with appreciation their dedication and contribution of time and effort on behalf of ASTM

Printed in Philadelphia,PA October 1999

Foreword This publication, Acoustic Emission: Standards and Technology Update, contains papers presented at the symposium of the same name held in Plantation, Florida, on 22-23 January 1998. The symposium was sponsored by ASTM Committee E7 on Nondestructive Testing. The symposium chairman was Sotirios J. Vahaviolos, Physical Acoustics Corporation.

Contents Overview

vii

AE SOURCES: CHARACTERIZATION

Use of Acoustic Emission to Characterize Focal and Diffuse Microdamage in BonemR. M. RAJACHAR, D. L. CHOW, C. E. CURTIS, N. A. WEISSMAN, AND D. H. KOHN

CONCRETE APPLICATIONS

A Proposed Standard for Evaluating Structural Integrity of Reinforced Concrete Beams by Acoustic E m i s s l o n - - s , YUYAMA,T. OKAMOTO,M. SHIGEISH],M. OHTSU, AND T. KISI-H

25

On the Necessity of a New Standard for the Acoustic Emission Characterization of Concrete and Reinforced Concrete Structures--E. o. NESWJSKI

41

AE Evaluation of Fatigue Damage in Traffic Signal Poles--H. R. HAMILTON,HI, T. J. FOWLER, AND J. A. PUCKETI"

INTmGRrrY AND LEAK

50

DETECTION/LOCATION METHODS

The Development of Acoustic Emission for Leak Detection and Location in LiqnidFilled, Buried Pipelines--R. ~ MILLER, A. A. POLLOCK, P. FINKEL, D. J. WATI'S, J. M. CARLYLE, A. N. TAFURI, AND J. J. y1~7.71 JR.

67

Acoustic Emission and Ultrasonic Testing for Mechanical Integrity--s. J. TERNOWCHEK, T. J. GANDY, M. V. CALVA, AND T. S. PATIT~SON

79

AE SENSORS, STANDAm)S, AND QUANTITATIVEAE

Calibration of Acoustic Emission Transducers by a Reciprocity Method--H. HATANO

93

DIVERSE INDUS1RIAL APPLICATIONS

Acoustic Emission Applied to Detect Workpiece Burn During G r i n d i n g - 107

P. R. DE AGUIAR, P. WILLETI', AND J. WEBSTER

Analysis of Fracture Scale and Material Quality Monitoring with the Help of Acoustic Emission Measurementsms. A. NIKULIN,M. A. SHTREMEL,V. G. KnANZH~, 125

E. Y. KURIANOVA, AND A. P. MARKELOV

Characterization of Micro and Macro Cracks in Rocks by Acoustic Emission-G. M. NAGARAJA RAO, C. R. L. MURTHY, AND N. M. RAJU

Prediction of Slope Failure Based on AE Activity--T. SHIOTANIANDM. OHTSU

141 156

A E SOURCES: RESEARCH T o P I c S

Identification of AE Sources by Using SiGMA-2D Moment Tensor Analysism 175

M. SHIGEISHI AND M. OHTSU

TRANSPORTATION APPLICATIONS, STANDARDS, AND METHODOLOGy Practical AE

Methodology

for Use on AircraftmJ.

M. CARLYLE, H. L. BODINE, S. S. HFa'qLEY,

R. L. DAWES, R. DEMESKI, AND E. v. K. HILL

191

COMPRESSED GAS APPLICATIONS AND STANDARDS

Periodic AE Re-Tests of Seamless Steel Gas Cylindersmp. R. BLACKBURN

209

Field Data on Testing of Natural Gas Vehicle (NGV) Containers Using Proposed ASTM Standard Test Method for Examination of Gas-Filled Filament-Wound Pressure Vessels Using Acoustic Emission (ASTM E070403-95/1)-R. D. FULTINEER, JR. AND J. R. MITCHELL

224

Acoustic Emission Testing of Steel-Lined FRP Hoop-Wrapped NGV Cylinders-A. AKHTAR AND D. KUNG

236

Author Index

257

Subject Index

259

Overview

Acoustic Emission (AE) has been commercially available for more than thirty (30) years. Has any progress been made? The purpose of the Symposium held in January 1998 in Plantation, Florida was to discuss the evolution of the technology of AE over the years in instrumentation, applications, standards and codes and its overall worldwide acceptance. Authors have made comparisons between AE and other Nondestructive Testing (NDT) technologies as to their suitability in solving practical industrial problems worldwide. As the newcomer in the Nondestructive Evaluation (NDE) industry, AE was first tried on applications where other NDT technologies had previously failed or was used where wild financial cost savings were promised. The issue of suitability of AE for an application was never considered until the very late 70's and early 80's, when a new breed of industrial and university researchers entered the field in USA, Europe and Japan. AE "noise counting" was replaced with basic work on source characterization, wave propagation, mode conversion, the study of the inverse problem using a number of Green's functions, pattern recognition and, most importantly, they considered AE as a science, using all available tools at their disposal. While the university academics worked hard to identify certain AE waveform features with source and failure mechanisms, a number of industrial researchers explored a myriad of "Pseudo-sources" of AE and their statistical nature. Instead of absolute one-on-one correlations and exact location of defects, practitioners developed zonal location and data bases based on case studies that enabled them to relate AE to fracture mechanics, corrosion phenomena, and overall part integrity assessment, especially in composite structures first and then in pressurized systems and individual components. The introduction of artificial intelligence, coupled with existent data bases, led to the development of ready-to-use knowledge-based systems based on very complex structures that are found in power utilities, refineries, chemical plants, complex pipelines, wind tunnels, aircraft structures, etc. The hard work of the late 70's and early 80's by CARP (Committee on AE for Reinforced Plastics) and the wide application of AE in testing of Fiberglass (FRP/GRP) vessels and pipes rejuvenated the technology! Eventually they became ASTM Standards now widely in use. The well-publicized early failures of AE in several metal vessels tests, especially in Europe by INEXPERIENCED personnel, were now reconsidered. Unknown to most AE Researchers/ Practitioners a behind the scenes branch of CARP known as CAM (Committee for Acoustic Emission for Metal) start looking carefully utilizing vast experience in Fracture Mechanics, Civil Engineering, NDT and, most importantly, vessel construction maintenance and use, realized early on that the same inexperience that prevented the use of AE in FRP in the early 70's has prevented users to do Metal Vessel Testing by AE. With the help of t h e ' 'core members" of CARP, metal vessel testing was reconsidered, especially after the successes of MONPAC ~ (a commercially available knowledge-based expert system that formed the basis of acceptance of AE by American Society of Mechanical Engineers (ASME) and Department of Transportation (DOT) and, thus, gave credence to the newcomer NDE technology). In addition, the more than ten AE ASTM Standards and AE's acceptance by American Society for Nondestructive Testing (ASNT) as another major NDT technique and the establishment of Level III in AE were major steps forward for the technology worldwide.

vii

viii

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

In this Symposium basic important work is being presented that constitutes the basis for Natural Gas Vehicle (NGV) Cylinder Testing with AE, no matter how controversially some people might view their work. When properly applied, AE can save NGV assets for customers as the ASTM FRP vessel has done for the past 10-plus years. It is interesting to note that infrastructure and slope stability applications worldwide and especially in Japan are now to the point of standardization of existing working procedures. We were very much encouraged by the continuing success of the Reciprocity Method for Calibrating AE Sensors and hope that it eventually will become another ASTM Standard. As for the other applications, I can only comment on their existing uniqueness from micro damage in bones to burning of grinding tools in high speed manufacturing. We hope this publication will prove interesting to a wide spectrum of readers, especailly those who look for new AE Standards and are interested to explore the future directions for the application of the Acoustic Emission Technology.

Sotirios J. Vahavidos, Ph.D.

Physical Acoustics Corporation Princeton Junction,NJ 08550 SymposiumChairmanand Editor

AE Sources: Characterization

Rupak M. Rajachar, 1 Dann L. Chow, l Christopher E. Curtis, 2 Neil A. Weissman, 2 and David H. Kohn 3

USE O F A C O U S T I C E M I S S I O N T O C H A R A C T E R I Z E F O C A L AND D I F F U S E M I C R O D A M A G E IN BONE

REFERENCE: Rajachar, R. M., Chow, D. L., Curtis, C. E., Weissman, N. A., and Kolm, D. H., "Use of Acoustic Emission to Characterize Focal and Diffuse Microdamage in Bone," Acoustic Emission: Standards and Technology Update, ASTM STP 1353, S. J. Vahaviolos, Ed., American Society for Testing and Materials, West Conshohocken, PA, 1999. A B S T R A C T : Fatigue of cortical bone results in the initiation, accumulation, and propagation of microdamage. AE techniques were adopted to monitor damage generated during ex-vivo tension-tension fatigue testing of cortical bone. The primary objectives were to determine the sensitivity of AE in detecting microdamage in cortical bone and to elucidate mechanisms guiding the onset of microdamage. Fatigue cycle data and histological data show that AE techniques are more sensitive than modulus reduction techniques in detecting incipient damage in cortical bone. Confocal microscopy revealed the ability of AE to detect crack lengths and damage zone dimensions as small as 25 gm. Furthermore, measured signal parameters such as AE events, event amplitude, duration, and energy suggest that AE techniques can detect and distinguish microdamage mechanisms spatially and temporally in bone. As fatigue processes continue, AE increases in terms of number of events, event intensities and spatial distribution. Diffuse damage appears to be a precursor to the development of linear microcracks. The spatial and temporal sequence of AE events enables differentiation between linear microcracks and more diffuse damage.

K E Y W O R D S : acoustic emission, bone, microdamage, confocal microscopy

l Graduate Student, Department of Biomedical Engineering, College of Engineering, University of Michigan, Ann Arbor, MI 48109-2125. 2 Undergraduate Student, Department of Biologic and Materials Sciences, School of Dentistry, University of Michigan, Ann Arbor, MI 48109-1078. 3 Associate Professor, Departments of Biologic and Materials Sciences, School of Dentistry and Biomedical Engineering, College of Engineering, University of Michigan, Ann Arbor, MI 48109-1078.

Copyright9

by ASTM International

3 www.astm.org

4

ACOUSTICEMISSION: STANDARDS AND TECHNOLOGY

Age-related skeletal fragility is a significant medical and economic problem. In the United States, an estimated 250,000 hip fractures occur each year among those age 50 and over [t]. The cost involved in the treatment of these fractures is believed to be over 7 billion dollars [2,3]. In younger people, who have a greater physiologic capacity to withstand the functional demands placed on the musculosketetal system, stress fractures and impact/trauma-related fractures are also of significance. However, with age and pathologic conditions, the compensatory mechanisms needed to maintain the mechanical stability of skeletal tissue become significantly more impaired [3]. As a result, diffuse distributions of sub-threshold microcracks, which have been observed even in normal bone, increase exponentially in number with age. These cracks range in length up to 300 ~tm and may ultimately contribute to age-related property degradation and fatigue fracture in skeletally mature cortical bone (e.g. the dense bone of the mid-femur, or outer shell of the proximal femur and spine) and trabecular bone (e.g. the porous, spongy bone inside the cortical shell) [4]. Consequently, the processes guiding the location, orientation, size, and accumulation of microdamage are important in assessing the competence of skeletal tissues and in developing a basic understanding of structurefunction relationships. Cortical bone can be modeled as a fiber-reinforced composite [5]. Osteons or Haversian canals are cylindrical layers of bone around blood vessels. The inter-osteonal bone, which is arranged in lamellae, is called interstitial bone. Thus, osteons in bone, which are primarily oriented in a longitudinal direction, are analogous to unidirectional fibers in a composite. As in many engineering composites, a comparatively high strength/low toughness interface exists between each osteon or "fiber" and its surrounding interstitial bone or "matrix". This region is referred to as the cement line and is thought to be a site of relatively easy crack nucleation [6]. Also of importance to the overall structure of bone are embedded cellular compartments and their connecting network, lacunae and canaliculi, respectively [6]. These regions represent heterogeneous sites of porosity, akin to processing defects in man-made structural materials. Accordingly, these pores may serve as additional sites of crack nucleation and accumulation [7,8]. Bone is therefore a hierarchical composite structure subject to internal microdamage. Many of the difficulties which exist in characterizing the mechanics of structural composites also exist with bone. The long-term objectives of this research program are to model bone as a composite material and: 1) determine the effects of mechanical history on microdamage initiation and growth; 2) provide insight into mechanisms of damage initiation and accumulation; and 3) determine how damage phenomena are modulated by changes in tissue hierarchy. Previous investigations have attempted to use AE to monitor integrity of bone invitro [9-12], and as a non-invasive diagnostic in-vivo [13-15]. These studies made inferences that the AE detected was due to specific failure mechanism(s), but no direct correlations between AE and damage were established. Using more rigorous analyses, we have shown that AE techniques can be effectively used to distinguish mechanisms of crack nucleation, slow crack propagation, and rapid crack propagation, spatially and temporally, in biomaterials, microstructured materials and inhomogeneous materials [16,17]. The critical resolution of AE for detecting crack nucleation was shown to be on the order of 10 ~tm [16]. Because of the complex inhomogeneous nature of cortical bone,

RAJACHAR ET AL, ON MtCRODAMAGE IN BONE

application of these more detailed AE techniques may provide similar insight into the mechanical processes involved in the nucleation and growth of damage in bone. Moreover, the sensitive and non-destructive nature of AE testing may allow multiple crack sites to be distinguished spatially and temporally. In support of our long-term objectives, the specific aims of this project were to: 1) detect and characterize incipient microdamage in cortical bone via AE, 2) verify and quantify the microdamage histologically, 3) compare the sensitivity of AE and modulus reduction (AE) techniques, 4) compare microdmnage in bone of different initial stress intensities, and 5) associate AE signals with microstructural failure mechanisms. Materials and Methods

Cortical Bone Specimens

The flow chart in Figure 1 provides an overview of the experimental design used in preparing and testing cortical bone ex-vivo. Cortical bone specimens were prepared from mature bovine femoral and tibial central diaphyseal sections. Each diaphysis was sectioned on a band saw into paratlelepipeds, such that the longitudinal axis of each parallelepiped was aligned with the long axis of the bone. These rough-cuts were then machined into smooth parallelepiped blanks (L = 120 mm, W = 12 mm, T = 4.5 mm) and gage sections were machined using a precision milling machine. Buffered saline irrigation was used during all machining steps to avoid heating the bone and to maintain tissue saturation. Two gage section geometries were created: V-notched specimens (p = 200 ktm, K t = 2.5), which provided a localized region of strain, and C-notched specimens (K t = 1), which provided a distributed strain region. Specimens not tested immediately after machining were wrapped in moist towels and stored at -65~ Fatigue Loading

Fatigue loading regimes were imposed using a servohydraulic mechanical testing machine and FLAPS mechanical testing computer interface program (Instron Corp., Canton, MA). During loading, specimens were kept moist using an ambient temperature buffered saline gravity drip, since drying results in an increase in the elastic modulus of bone. Initial elastic moduli (E0) were determined by subjecting the specimens to successive uniaxial tensile ramp loading and unloading cycles in the elastic range until a steady state value was reached. Peak stresses were approximately one-quarter of the yield stress for bovine cortical bone (~y - 140 MPa), and a steady state modulus value was typically reached in 5 cycles. The initial elastic modulus was then used to define a fatigue loading regime [18]. Fatigue loading was performed in uniaxial tension under load control. Specimens were loaded parallel to the longitudinal axis of the bone, following a sinusoidal waveform, over an effective strain range (AEeff = A(Y]E0) of 0 to 3000 ge, at a frequency of 1 Hz. The maximum strain was chosen such that it was in the upper range of physiologic strain experienced by cortical bone during normal loading [19].

5

6

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

[

I

Mature Bovine Bone(femur)

I

I

DiaphysealLongitudinalCut

1

I

J

End Milled Paralleleplped I

J

C-Notch , diffusedamagezone I I

v-Notch I localizeddamagezone I I

I Fatigueto

Fatigueto

AEo.set I

I 1%~.~

I

I tf

I

I

I

I

I

I

FatigueLoadingConditions 3000petrain,1 Hzsinewave

I

AE ParameteisMeasured events,Ioc.,ampL,(mergy,etc.

I

10 mmHistology notchedsection bulkstainingandembedding 150 ~ sedalsections

I

Confocal Microscopy grideachsectionandscan crackIoc.,density,INze,etc. 3-D reconstruction

FIGURE

1 -

Unloaded I control

I

RELATEAE TO FAILUREMECHANISMS

Flow chart of experimental design.

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

Experimental Design For each gage section geometry, a 30-specimen experimental matrix was used to study fatigue-related microdamage: 10 specimens were loaded in fatigue until the onset of AE, 10 were loaded in fatigue until there was a measured 1% modulus reduction (AE), and 10 served as unloaded histological controls. Each specimen was randomly assigned to one of these three groups. Modulus measures were made based upon real-time mean strain values, measured with an extensometer.

Acoustic Emission Analysis AE was recorded using a planar array of four Physical Acoustics Corp. (PAC, Princeton, N J) nano-30 piezoelectric transducers having a broad-band frequency range of 125-750 kHz. Two sensors were placed on the shoulders of the 120 mm x 12 mm face, for longitudinal location, and the other two sensors were on opposite sides of the width of the specimen, above and below the notched region, for transverse location. The sensors were coupled to the specimens with an acoustic couplant and fixed in place using waterresistant surgical grade adhesive tape. AE data were collected, stored and analyzed with PAC LOCAN-320 data acquisition and analysis software. The pertinent operating parameters were: variable gain/total gain = 42 dB/80 dB; peak definition time = 500 gtsec; hit definition time = 2 msec; dead time = 1 msec; sample time = 100 msec; threshold = 1 V. A threshold value of 42 dB was used to eliminate background noise produced by specimen irrigation. For the tests stopped at the onset of AE, the first AE signals above 42 dB simultaneously detected at all 4 sensors were taken to signify the onset of microdamage. The following AE parameters were recorded and analyzed: AE source location, number of AE events, and intensities of AE events. Event intensities are a collective term for event amplitude, counts per event, event duration, event energy counts and event rise time. Subsets of events were also created, based on event location, fatigue cycle number, and stress range at which events were generated. Subsets of event intensities generated within different ranges of location, fatigue cycle number and stress level were also analyzed. Subsets of events were then analyzed by evaluating location distribution histograms (LDH) and intensity distribution histograms (IDH) of events. LDH and IDH are general terms for the distribution of events and event intensities as functions of location, fatigue cycle number or stress. Initial analysis of the spectral components of waves was also carried out. Digital transient capture of waveforms was performed using an F4000 Fracture Wave Detector (Digital Wave Corp., Englewood, CO). A maximum digitization sampling rate of 12.5 MHz was used and 1024 points of digitized data were collected from each waveform. Characteristic extensional and flexural waves were generated by breaking a lead pencil at multiple sites on selected bone specimens. Actual AE waves generated during testing of bone were recorded and digitized, and mode shapes and dominant frequency contents were determined and compared to the waveforms generated by the pencil breaks [20,21].

7

8

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

Histological Analysis Following fatigue loading, 10 mm blocks from the region near the dominant AE sources were cut using a diamond wafering blade, and stored at -65~ in buffered saline solution until histological processing. Specimens were stained en-bloc using a graded series of 1% basic fuchsin solutions in ethanol [22]. Basic fuchsin is a fluorescent stain that preferentially marks exposed external and internal defects in biological structures. Bulk staining with basic fuchsin prior to histological sectioning enables visualization of microdamage and differentiation between cracking due to mechanical factors and artifactual cracks induced during histological preparation. Following staining each specimen was embedded in poly-(methyl methacrylate) and serially sectioned (-150 p.m thick) parallel to the 4.5 mm thickness. Each serial section was analyzed using laser scanning confocal microscopy (LSCM) to assess damage at the notch tip and throughout the bulk of the specimen. Three-dimensional data on crack morphology were obtained by using a z-axis reconstruction [23]. A l-ram square grid system was imposed on each histological section. Each grid space was analyzed for number of cracks, crack length, crack density, and crack angle. Results

In support of Specific Aims 1 and 2, we were able to detect and characterize AE generated during fatigue of cortical bone and validate, via histological comparison to unloaded controls, that the sources of this AE were fatigue-induced microdamage. An LDH of AE events generated during fatigue of a V-notched specimen is shown in Figure 2. AE events are localized at the notch-tip (location 0). Representative LSCM photos of the notch-tips of fatigue and control specimens are shown in Figure 3. The highly luminous regions at the notch-tip of the fatigue specimen represent damage nucleation. More well defined linear (Mode I) microcracks are observed ahead of the diffusely damaged region. Quantitative histology revealed crack lengths as small as 25 p.m.

,~ 2O

C

<

10

E z

0

..... . . . -40

-30

.

.

-20

.

m,. -10

.

~1, . . .nn,. nn, . , , 0

10

20

t

tm I

30

40

Location

FIGURE 2 - Location distribution histogram (LDH) of AE events generated in V-

notched cortical bone specimen fatigued until the onset of AE. Events are localized at the notch-tip (location 0), indicating damage nucleation in this region.

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

FIGURE 3 - Confocal micrographs of V-notched specimens: a) control (unloaded) specimen, and b) specimenfatigued until the onset of AE. In the fatigue specimen, damage accumulation is seen at the notch-tip. Diffuse damage is observed at the notchelp, with well-defined linear microcracks observed ahead of the diffuse damage zone. (Both I OOX)

9

10

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

The average number of fatigue cycles to the onset of damage as determined by the two techniques, for V- and C-notched specimens, is presented in Table 1 (Specific Aim 3). For both specimen groups, incipient damage was detected at a significantly lower number of fatigue cycles with AE than AE (p < 0.01, p < 0.02, respectively, via Student's t-tests). TABLE 1 - Number of cycles to the onset of fatigue as determined by acoustic emission (AE) and modulus reduction (AE) techniques. Technique AE AE AE AE

Specimen Type V-Notched V-Notched C-Notched C-Notched

Averase NumberofCycles 8894 • 5317 32548 • 25697 8094 • 4241 40544 • 22828

Figure 4 shows LDHs of AE events generated in V-notched specimens fatigued until the onset of AE and until a 1% AE. Although both tests reveal AE accumulation at the notch-tips, the greater number of events generated in the AE specimen is commensurate with the longer fatigue regime of this specimen. Evaluating IDHs (Fig. 5), shows that peak and maximum values of signal intensities generated in bone that was fatigued until a 1% AE are greater than those generated in bone fatigued until just the onset of AE. Average intensities were also greater in the AE specimens. Comparison of confocal micrographs from AE and AE specimens (Fig. 6) suggests the presence of diffuse microdamage followed by development of linear microcracks as fatigue processes continue. Diffuse damage is characterized by relatively larger zones of staining opacity and minimally resolvable crack dimensions. While many histological sections exhibited Mode I microcracks, several sections revealed the presence of linear microcracks, adjacent to regions of diffuse damage, which were oriented parallel to the direction of loading (Fig. 6c). In support of Specific Aims 4 and 5, we were able to compare microdamage in bone of different initial stress intensities and associate AE signals with microstructural failure mechanisms. The mechanisms, observed histologically, were characterized based upon differences in specific AE signal parameters, as shown in the LDH and IDH in Figures 7 and 8. Specimens with a lower initial stress concentration generated a greater number of events, a greater spatial distribution of events and more high intensity events.

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

4 - Location distribution histograms (LDH) of AE events generated in Vnotched specimens fatigued until the onset of AE and until a 1% modulus loss. The greater number of events generated in the AlE specimen is commensurate with the longer fatigue process in this specimen. FIGURE

11

12

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

FIGURE 5 - hztensity distribution histogl"ams (IDI1) o f A E events generated in Vnotched specimens fatigued until tile onset o f A E and m~til a 1% modulus loss. A E signal hm, nsities in the A E specimens are lower than those o f the AE group, since the tests were stopped at an earlier stage o f fittigue.

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

6 - Confocal micrographs of a) diffuse microdamage (I OOX), b) Mode I linear microcracks (I OOX), and c) combination of diffuse damage and Mode H linear microcracks (60X).

FIGURE

13

14

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

7 - Location distribution histograms (LDH) of AE events generated in Vnotched and C-notched specimens fatigued until the onset of AE. The C-notched specimens exhibited a greater spatial distribution of AE events. FIGURE

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

15

FIGURE 8 - Intensity distribution histograms (IDH) of AE events generated in V-notched and C-notched specimens fatigued until the onset of AE. AE signal intensities are different in the two specimens, indicating different damage mechanisms.

16

ACOUSTICEMISSION: STANDARDS AND TECHNOLOGY

Discussion The objectives of this research were to determine the sensitivity of AE in detecting microdamage in cortical bone and to use the early warning capabilities of AE to help elucidate the underlying mechanisms guiding the nucleation and growth of microdamage in bone. LDHs (Fig. 2) and histology (Fig. 3) provided evidence that damage in cortical bone can be detected and characterized using AE. This is apparent by the prominent AE that was detected and microcracking that was observed at the notch-tips. Quantitative histology revealed that the AE was first generated at crack lengths and damage zone dimensions as small as 25 ~m. The fatigue data shows that AE techniques are more sensitive than modulus reduction techniques for detecting damage in cortical bone for each specimen geometry (Table 1). A greater number of AE events was also detected in specimens which were allowed to fatigue until a 1% AE was achieved (Fig. 3). The IDHs (Fig. 4) reflect the greater number of events, greater peak amplitude and greater number of high intensity events for the AE specimens. The damage inferred by the AE and AE techniques was subsequently verified histologically. All of these data support that conclusion that AE is more sensitive than modulus reduction techniques in discriminating incipient damage in bone. Currently, modulus reduction techniques represent the accepted real-time means of detecting damage in bone [24]. Since the sensitivity of AE is greater than that of AE, it is expected that there should also be a difference in the type and extent of microstmctural damage observed in the two groups. It is therefore hypothesized that different mechanisms of damage initiation and propagation in bone will express distinguishable AE signal profiles. The IDHs suggest these differences. Based on the relative amounts of diffuse and linear damage observed in control, AE and AE specimens, there, qualitatively, appears to be a progression of diffuse damage accumulation which ultimately coalesces into more distinct linear microcracks. There appears to be a greater damage density (i.e. # of cracks or # of diffusely stained regions/unit area) in AE samples compared to AE samples, and a threshold in damage density above which AE is generated. This threshold corresponds to an accumulation of plastic strain energy sufficient to generate detectable AE. Coupled with histological data, AE analysis suggests an initial diffuse matrix damage mechanism in the fatigue failure of cortical bone, analogous to microvoiding and craze formation in fiber reinforced composites. Further fatigue results in the formation of linear microcracks. LDH data allows the differentiation of diffuse matrix damage away from the notch-tip and damage at the primary fracture process zone. Linear microcracks were most often associated with cement lines, an observation made by others as well [7,25,26]. The specific nature of the interactions between microcracks and cement lines is unclear. Microcracks were observed to propagate both along and across cement lines. Combined with the fact that linear microcracks were observed to be oriented in both Mode I and Mode II directions, the overlapping of AE signal parameters in the different test groups is likely due to the combination of ultrastructural failure mechanisms observed.

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

The observation of longitudinal and shear oriented cracks motivated an analysis of AE waveforms. Based on the characteristic extensional and flexural waves generated during pencil breaks (Fig. 9), analysis of spectral components of AE signals generated during fatigue of bone revealed the following. At the notch-tip, longitudinal waves, with a dominant frequency range of 250-400 kHz, were primarily detected, whereas in the bone matrix, flexural waves, with a dominant frequency range of 100-180 kHz, were primarily detected. These characteristics are consistent with those of fiber reinforced plastics [27]. While it is acknowledged that modal analyses are limited to plate specimens and may not be applicable to in-vivo diagnosis of a whole femur, the modes of microfailure observed in this study are consistent with those observed in-vivo. Idealized testing on plate-specimens, which are amenable to modal analysis, may yield significant insight into damage mechanisms. Whether diffuse damage and modes I and II linear microcracking processes can be linked to specific AE signals is a continuing focus of this research. It may ultimately be possible to link damage location with particular initiation and propagation mechanisms, which may provide necessary information to the detection and prevention of catastrophic damage associated with many degenerative orthopaedic conditions. If a continuum can be established between diffuse damage and linear microcrack initiation and propagation mechanisms, it may also be possible to relate mechanical and biological components affecting the overall structure of cortical bone under different loading conditions. Thus, future work will not only attempt to quantify the relationship of the strain history of cortical bone to damage initiation and propagation mechanisms, but also to define local biologic conditions favoring damage mechanisms.

17

18

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

s) extensional w a v e

4.0 .~

3.0t 20

1.0 E

0.0-

>0

-1.0 -

0 n,

-2.0 -3.0 -4.0

1.6E-04

O.OE+O0 Time (seconds)

b) flexural w a v e

6.0 4.0 2.0 a,.

E .< 0 > 0

n,

0.0

I! -2.0-

I,

T

II

II

I I,

j

4.0~.0 O.OE+O0

1.6E-04 Time (seconds)

FIGURE 9 - AE waveforms from lead pencil breaks." a) extensional wave, dominant frequencies- 300-500kHz, b) flexural wave, dominant frequencies ~ 100-200 kHz.

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

Summary and Conclusions AE techniques were employed in association with histological analysis to determine the effects of strain history on the initiation and accumulation of fatigue microdamage in cortical bone. Microstructural damage mechanisms were identified and related to AE parameters spatially and temporally. Specifically, the following conclusions are drawn: (1) (2)

(3)

(4) (5)

AE techniques are effective in detecting incipient microdamage generated during fatigue of cortical bone. Histological comparison between loaded and unloaded specimens validated the AE technique and revealed that AE can detect microcracks in bone as small as 25 ~tm. Multiple microcracks and diffuse regions of damage may also be detected. Incipient fatigue damage is detected at a significantly lower of fatigue cycles with AE techniques, as compared to modulus loss techniques and, as a result, less AE is generated. As fatigue processes continue, AE increases in terms of number of events, event intensities and spatial distribution. Diffuse damage appears to be a precursor to the development of linear microcracks. The spatial and temporal sequence of AE events enables differentiation between linear microcracks and more diffuse damage.

Acknowledgements Supported by NSF BES-9410303, the Whitaker Foundation and the Natural Sciences and Engineering Research Council of Canada. We gratefully acknowledge Mitch Schaffler, Ph.D. of Henry Ford Hospital for his help in the histological processing and analysis, and Gordon Schneider of Digital Wave Corp. for his help performing the modal analyses.

References

[I]

Melton, L.J., Eddy, D.M., and Johnston, C.C. Jr., 1990, "Screening for Osteoporosis," Ann. Int. Med. Vol. 112, pp. 516-528.

[21

Holbrook, T.L., Grazier, K., Kelsey, J.L., and Stauffer, R.N., 1984, American Academy of Orthopedic Surgery, Rosemont, IL.

[31

Birdwood, G., 1996, Understanding Osteoporosis and its Treatment. A Guide for Physicians and their Patients, Pearl River, New York.

[4]

Schaffler, M.B., Choi, K., and Milgrom, C., 1995, "Aging and Matrix Microdamage Accumulation in Human Compact Bone," Bone, Vol. 17, pp. 521525.

19

20

ACOUSTICEMISSION: STANDARDS AND TECHNOLOGY

[51

Katz, J.L., 1981, "Composite Material Models for Cortical Bone," Mechanical Properties of Bone, S.C. Cowin, Ed., American Society of Mechanical Engineers Vol. 45, pp. 171-184.

[61

Martin, R.B., Burr, D.B., 1989, Structure, Function, and Adaptation of Compact Bone, Raven Press, New York.

[7]

Frost, H.M., 1960, "Presence of Microscopic Cracks In Vivo in Bone," Henry Ford Hospital Bulletin, Vol. 8, pp. 25-35.

[8]

Martin, R., 1982, "A Hypothetical Mechanism for the Stimulation of Osteonal Remodelling by Fatigue Damage," J. Biomech., Vol. 15, pp. 137-139.

[9]

Knet-s, I.V., Krauya, U.E., and Vilks, Y.K., 1975, "Acoustic Emission in Human Bone Tissue Subjected to Longitudinal Extension," Mekh. Polim., Vol. 4, pp. 685690.

[lO]

Wright, T.M., Vosburgh, F., and Burstein, A.H., 1981, "Permanent Deformation of Compact Bone Monitored by Acoustic Emission," J. Biomech., Vol. 14, pp. 405-409.

[Ill

Fischer, R.A., Arms, S.W., Pope, M.H., and Seligson, D., 1986, "Analysis of the Effect of Using Two Different Strain Rates on the Acoustic Emission in Bone," J. Biomech., Vol. 19, pp. 119-127.

[12]

Zioupos, P., Currey, J.D., and Sedman, A.J., 1994, "An Examination of the Micromechanics of Failure of Bone and Antler by Acoustic Emission Tests and Laser Scanning Confocal Microscopy," Med. Eng. Phys., Vol. 16, pp. 203-212.

[13]

Wright, T.M., Arnoczky, S.P., and Burstein, A.H., 1978, "In-situ monitoring of ligament damage in the canine knee by acoustic emission," Mater. Eval., Vol. 37.

[14]

Wright, T.M., Hood, R.W., and Flynn, W.J., 1981, "Acoustic emission monitoring in the diagnosis of loosening in total knee arthroplasty," 1981 Biomechanics Symposium, Van Buskirk, W.C. and Woo, S.L.-Y., Eds., American Society of Mechanical Engineers, pp. 203-212.

[15]

Poliakoff, S.J., Miller, R.K., Jones, C.B., and Bright, R.W., 1989, "Acoustic emission monitoring of physeal separation: an experimental study," Trans. Orthop. Res. Soc., Vol. 14, p. 483.

[161

Kohn, D.H., Ducheyne, P. and Awerbuch, J., 1992, "Acoustic Emission During Fatigue of Ti-6A1-4V: Incipient Fatigue Crack Detection Limits and Generalized Data Analysis Methodology," J. Mater. Sci., Vol. 27, pp. 3131-3142.

RAJACHAR ET AL. ON MICRODAMAGE IN BONE

[17]

Kohn, D.H., Ducheyne, P., and Awerbuch, J., 1992, "Acoustic Emission During Fatigue of Porous Coated Ti-6AI-4V Implant Alloy," J. Biomed. Mater. Res., Vol. 26, pp. 19-38.

[18]

Caler, W., 1989, "Bone Creep-Fatigue Damage Accumulation," J. Biomech., Vol. 22, pp. 625-635.

[19]

Lanyon, L.E., Goodship, A.E., Pye, C.J., and McFie, J.H., 1982, "Mechanically Adaptive Bone Remodeling," J. Biomech., Vol. 15, pp. 141-154.

[20]

Kannatey-Asibu, E. and Emel, E., 1987, "Linear Discriminant Function Analysis of Acoustic Emission Signals for Tool Condition Monitoring," J. Mech. Sys. Signal Proc., Vol. 1, pp. 333-347.

[21]

Kohn, D.H., 1995, "Acoustic Emission and Non-Destructive Evaluation of Biomaterials and Tissues," Crit. Rev. Biomed. Eng., Vol. 22, pp. 221-306.

[22]

Burr, D.B., 1995, "Alterations to the En Bloc Fuchsin Staining Protocol for the Determination of Microdamage Produced In Vivo," Bone, Vol. 17, pp.431-433.

[23]

Ross, M., 1995, Histology. A Text andAtlas, Williams and Wilkins, New York.

[24]

Schaffier, M.B., Radin, E.L. and Burr, D.B., 1989, "Mechanical and Morphological Effects of Strain Rate on Fatigue of Compact Bone," Bone, Vol. 10, pp. 207-214.

[25]

Carter, D.R., and Hayes, W.C., 1977, "Compact Bone Fatigue Damage - I Residual Strength and Stiffness," J. Biomech., Vol. 10, pp. 325-337.

[26]

Schaffier, M.B., Pitchford, W.C., Choi, K., and Riddle, J.M., 1994, "Examination of Compact Bone Microdamage Using Back-Scattered Electron Microscopy," Bone, Vol. 15, pp. 483-488.

[27]

Prosser, W.H., Jackson, K.E., Kellas, S., Smith, B.T., McKeon, J., and Friedman, A., 1995, "Advanced Waveform-Based Acoustic Emission Detection of Matrix Cracking in Composites," Mater. Eval., Sept., pp. 1052-1058.

21

Concrete Applications

Shigenori Yuyama, ~ Takahisa Okamoto, ~ Mitsuhiro Shigeishi, 3 Masayasu Ohtsu, aand Teruo Kishi 4 A PROPOSED STANDARD FOR EVALUATING STRUCTURAL INTEGRITY OF REINFORCED CONCRETE BEAMS BY ACOUSTIC EMISSION

REFERENCE: Yuyama, S., Okamoto, T., Shigeishi, M., Ohtsu, M., and Kishi, T., " A Proposed Standard for Evaluating Structural Integrity of Reinforced Concrete Beams by Acoustic Emission," Acoustic Emission: Standards and Technology Update, ASTM STP 1353, S. J. Vahaviolos, Ed., American Society for Testing and Materials, West Conshohocken, PA, 1999. ABSTRACT: A series of studies has been performed to evaluate the structural integrity of reinforced concrete (RC) beams by acoustic emission (AE). Cyclic loadings were applied to RC beams with a single reinforcing bar, large repaired beams, beams deteriorated due to corrosion of reinforcement, and two beams with different damage levels in an aging dock. The test results demonstrated that the Kaiser effect starts to break down when shear cracking starts to play a primary role. It has been also shown that high AE activity is observed during unloadings after serious damage (slips between the concrete and the reinforcement or those between the original concrete and the repaired part) has occurred. A standard for evaluating structural integrity of RC beams by AE is proposed, based on these results.

KEYWORDS: acoustic emission, cyclic loading test, evaluation criteria, Kaiser effect, reinforced concrete, structural integrity

President, Nippon Physical Acoustics Ltd., 8F Okamoto LK Bldg., 2-17-10, Higashi, Shibuya-ku, Tokyo 150, Japan. 2 Chief researcher, Central Research Laboratory, Nihon Cement Co., Ltd., 1-2-23, Kiyosumi, Koto-ku, Tokyo 135, Japan. 3 Associate professor and professor, ,'espectively, Depamnent of Civil and Environmental Engineering, Faculty of Engineering, Kumamoto University, 2-39-1, Kurokami, Kumamoto 860, Japan. 4 Professor, Research Center for Advanced Science and Technology, The University of Tokyo, 4-6-1, Komaba, Meguro-ku Tokyo 153, Japan.

Copyright9

by ASTMInternational

25 www.astm.org

26

ACOUSTICEMISSION: STANDARDS AND TECHNOLOGY

Introduction In recent years the deterioration and cracking of concrete structures such as bridges and buildings has been a significant problem. Proper techniques for the inspection of damaged structures are important in making rational decisions regarding rehabilitation, repair or replacement. Thus, the development of techniques to evaluate degradation of concrete structures in long-term service has been one of the most important issues for an effective maintenance program. A series of studies has been performed to evaluate the structural integrity of rcinforced concrete (RC) beams by acoustic emission (AE). Cyclic loadings were applied to RC beams with a single reinforcing bar [/], large repaired beams [2], beams deteriorated due to corrosion of reinforcement [3l, and two beams with different damage levels in an aging dock [4]. The test results demonstrated that the Kaiser effect starts to break down when shear cracking starts to play a primary role. It has been also shown that high AE activity is observed during unloadings after serious damage (slips between the concrete and the reinforcement or those between the original concrete and the repaired part) has occurred. A concrete beam integrity (CBI) ratio, the ratio of the load at onset of AE and the maximum prior load, has been proposed as an effective criterion to measure the severity of the damage induced in the beams. The high AE activities during unloadings have been shown to be an effective index to estimate the level of deterioration. This paper proposes an AE test method for RC structures, demonstrating four case studies conducted for different types of RC beams. Test procedure and evaluation criteria are presented as guidelines for practical AE tests of RC beams.

Case Study 1: RC Beams with a Single Reinforcing Bar Shown in Fig. 1 is a configuration of the specimen used for the cyclic bending test. A single reinforcing bar of 19 mm dia with lateral lugs is encased eccentrically in the rectangular concrete beam. Concrete cover (depth of reinforcing bar) is 30 mm. Compressive and tensile strengths of the concrete were 36.2 and 3.5 MPa, respectively. Six PAC R15 (150 kHz resonant) sensors were attached on the specimen to perform both a moment tensor analysis using the SIGMA code [.5] and parameter analysis. The specimens were subjected to repeated four-point bending loadings by a strain-control type machine. The maximum load of each loading cycle was increased gradually in order to investigate the relationship between the cracking process and AE behavior. Figure 2 presents the relationship between the number of AE hits and the applied load. AE signals are detected at a lower load than the maximum prior load (49kN) during the second loading. Accordingly, the Kaiser effect breaks down during the second loading. It was shown that the Kaiser effect starts to break down when the

YUYAMA ET AL. ON REINFORCED CONCRETE BEAMS

Z 9 '.

:..:. @.::,.'

e

.',

"..o*..,%

: . : , . . ' , . .

,,

.

Q-

..

,"

.,,"

/:~.:i-:"..:-..t . ". ;.L::.~ DEFORM E D:BAR : [ . . . . . . . . ' -...-...-.~~ _-.....;. ~.... ; .'....'.:-. ~':'.-: ...:..-;. :;

"J J

.l_

150

Ji

FIG.l-Configuration o f the specimen with a single reinforcing bar.

2500)--.T: 2 0 0 0 -

,, o

f_..J

1500-

kl_l

1V

rn 1000z

v

500-

-

!

0

i

49

i

9'8 LOAD

188

'

196

245

(kN)

FIG.2-Relationship between number o f AE hits and the applied load. crack width exceeds 0.12 mm. The breakdown of the effect becomes clearer as the cracking progresses in the third, fourth and fifth loadings. High AE activities are observed during the third, fourth and fifth unloadings. The moment tensor analysis revealed that the contribution of shear cracks increases as the breakdown of the Kaiser effect becomes clearer with the progress of the fracture. It was also indicated that high AE activity is observed during the third, fourth and fifth unloadings after the maximum width of the surface cracks has exceeded about 0.25 mm. The moment tensor analysis found that the shear cracks

27

28

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

generated near the reinforcing bar is responsible fl)r this activity. "lhe origin ,,i the~ emissions was attributed to rubbings between the faces of the existing cracks or ttiction between the reinforcement and concrete.

Case Study 2: Repaired RC Beams The configuration of the specimen and the locations of displacement transducers and AE sensors are shown in Fig. 3. Six PAC R15 sensors were placed lineally on the top plane of the specimen. The repaired part is in the tensile side of the specimen. The depth and length of the repaired part are 100 mm and 2200 mm, respectively. In addition to steel bars as reinforcement, stirrups were embedded in the specimen to prevent beams shear failure. The specimens were subjected to repeated four point bending loadings by a strain-control type machine. During each loading, measurements of AE, crack width, slip length between the repaired part and the original part, and strain of concrete and reinforcement were made by using AE sensors and two different types of displacement transducers. These measurements indicated that the initiation of the early tensile microcracks, main tensile cracks, local slips, and large-scale slips are clearly detected by AE hit measurement. It was shown that once large-scale slips have occurred at the interface between the original concrete and the repaired part, AE starts to emanate at

L CH.,6

480

r

-B-

CH.,5

_1 240_1 240 i ~,-

-

WCH.4

7,

i

C

480

~ ,

3

J 9

CH.m2

CH.II1

I

I

(6)

(7)

1

2

.. - - - - c . . . . . ~

9

....

(8) 3 IF.:"

(9) (10) 4

5

" 5.~ '. ~

"

,-,---.,.----'-----'-----

1_2ooJ2oo.I 2001_2oo.l.2oo 12oo_1

.......

~.

A

I C

[-

Two directional displacement transducer for sllp length Displacement transducer for crack width

mm

AE Sensor

FIG. 3-Configuration o f the repaired RC beams and the locations o f displacement lransduce~ and AE sensors.

YUYAMA ET AL. ON REINFORCED CONCRETE BEAMS

100

•'`''i''''l'``'i`'''•`'''I'''`••'''i•''••''''I''•`•''•`n'''•''''I''''•'`''l''''•','`i•''•''''t ....

MAIN

CRACK

INITIATION SLIP

L A R G E - S C A L E

I ,.

,

.

..

9

v

5o 9 ..

.;.

.

..............

. . . . . . . . . . . . . . . .

1

A

MICRO-CRACK

INITIATION

i0

J

2.5~ l

z 6

'"'

'"'

'"'

'"'

""

""

'"'

'"'

""

""

""

'"'

""

'"'

""

""1'"'1'"

I

3 0 0 0

TIME

I

'

0

6000

(SECOND)

FIG. 4-Amplitude and displacement histories. much lower load than the previous maximum load, that is, the Kaiser effect no longer holds for the next loading and high AE activity can be seen even during unloading. Thus, the breakdown of the Kaiser effect and the high AE activity during unloading can be a good indicator for the occurrence of large-scale slips in repaired RC beam. Amplitudes of all hits are plotted versus time together with the displacement in Fig. 4. It is obvious that the initiation of the early tensile microcracks or the local slips and the mechanical rubbings of the interlocked faces due to large-scale slips gave amplitude levels between 40 and 60 dB, while the initiation of the main tensile crack at 38.2 kN produced very high amplitudes that reached nearly 80 dB. Thus, the different AE sources could be clearly distinguished by comparing the amplitude data with the results of the visual observation and the measurement by displacement transducers. A concrete beam integrity (CBI) ratio, given below, was proposed as a criterion to measure the severity of damage induced in repaired concrete beams. CBI ratio = load at onset of AE / maximum prior load In the field of fiber-reinforced plastic (FRP) structures, AE tests have been widely used to evaluate structural integrity and testing has been standardized by ASME Code, Section V, Article 11. In this code, the Felicity ratio obtained from the ratio of the load at onset of AE and the maximum prior load gives the criterion to measure the severity of previously induced damage. It has been shown that the Felicity effect,

29

30

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

which is referred to as the breakdown of the Kaiser effect, is an indication of defects. The felicity ratio has been well accepted to examine structural integrity of chemical plant equipment such as pressure vessels, tanks and piping. However, beams, pillars, columns, and slabs are inspected in concrete structures. In the case of chemical plant equipment, structures are loaded by pressurization. In contrast, concrete structures are subjected to tensile, shear and bending loadings by jacking or running a heavy vehicle. As shown in the test results, failure mechanisms vary with the progress of damage in RC beams. It is obvious that the decrease of the CBI ratio is related to the generation and propagation of shear cracks. Thus, the ratio can be a practical index for evaluating structural integrity of RC beams. Taking these points into consideration, the CBI ratio was introduced [6]. Listed in Table 1 are CBI ratios for each loading cycle, based on AE hit rate activity. The ratios obtained by AE energy rate activity for all channels are also given in the last column. As shown in Table 1, the CBI ratio decreases from the fifth loading after large-scale slips have occurred due to the fourth loading along the interface between the original concrete and the repaired part. It continues to decrease as the damaged areas grow. It is known that the occurrence of large-scale slips is an essential feature for damage that may result in a serious disaster in repaired concrete structures. As shown above, the CBI ratio has very good correlation with the onset of large-scale slips and the growth of damaged areas. Thus, the CBI ratio can be a very useful and effective criterion to measure the severity of damage induced in repaired RC beams. It was also revealed that high AE activity is observed during unloadings once large-scale slips have initiated between the original concrete and the repaired part. The source of these emissions was ascribed to mechanical rubbings between the interlocked faces introduced by the large-scale slips. TABLE 1-Concrete Beam Integrity (CBI) ratios during the repeated loading tests of

repaired RC beams. CH2

CH3

CH4

CH5

CH1 "~ 6

CH1 ~" 6

(hit)

(hit)

(hit)

(hit)

(hit)

(energy)

2nd

1.25

-

1.25

-

1.25

1.25

3rd

1.16

1.16

1.16

-

1.16

1.16

4th

1.20

1.20

1.20

1.20

1.20

1.20

5th

0.69 0.53 0.68

0.69 0.53 0.40

0.69 0.53 0.50

1.00 0.80 0.68

0.69

0.69

0.53

0.53

0.40

0.25

6th 7th

YUYAMA ET AL. ON REINFORCED CONCRETE BEAMS

Case Study 3: RC Beams Deteriorated Due to Corrosion of Reinforcement Shown in Fig. 5 are dimensions (cm) of the specimen and sensor locations. Six PAC R6 (60 kHz resonant) sensors were attached on the specimen to perform the moment tensor analysis as well as AE parameter analysis. The lower quarter part of the specimen was immersed in a 3% sodium chloride solution and an anodic current was galvano-statically charged to the main steel bars until the maximum width of surface cracks due to corrosion of the bars reached 1 mm or 4 mm. Thus, three different types of specimens i.e. specimen with no corrosion damage and those with the surface cracks determined as above were subjected to repeated four-point bending loadings. Indicated in Fig. 6 are relationships between AE hits and the applied load for the specimens with the different deteriorated levels. It is observed that the Kaiser effect starts to break down during the third loading in the case of the specimen with no corrosion damage. However, it tends to break down during the second loading in the case of the deteriorated specimen (crack width 1 mm) and the breakdown is very clear during the second loading in the heavily deteriorated one (crack width 4 mm). ~_60 dB), and for the data obtained during the second AE test cycle. For each of the ten as received cylinders, the AE data were examined after dividing the test cycle into loading, hold and unloading segments. Table 2 shows for each segment the hits and the corresponding counts (shown in parenthesis) during the first 0-27.6 MPa (0-4000 psi) cycle. The hits and the corresponding counts for each of the three segments listed in Table 2 show variability similar in nature to that found with total hits and hits having amplitudes ~60 dB, the latter two listed in Table 1. TABLE l--Acoustic emission from as-received cylinders Cycle 1 Cylinder

All Hits

Cycle 2

Amplitude ~60 dB

All Hits

Cycle 2/Cycle 1

Amplitude ~60 dB

All Hits

Amplitude z60 dB

Undamaged Cylinders L0529-7/89 L0531-7/89 L1026-3/91 L1123-4/91 L1125-4/91 L1126-4/91 Ll132-4/91

2 212 11 311 72 880 25 362 53 854 25 074 65 174

31 505 2 497 780 1 691 1 498 5403

706 5 277 37 356 14 468 28 589 13 687 33 131

9 171 902 314 635 571 1 955

0.32 0.47 0.51 0.57 0.53 0.54 0.51

0.29 0.34 0.36 0.39 0.37 0.38 0.36

0.61 0.62 0.55

0.32 0.39 0.40

FRP Abrasion Damaged Vessels L0749-10/90 L1122-4/91 L 1131-4/91

91 241 24 958 48 803

6 458 1 446 6 070

55 835 15 516 26 791

2 095 554 2 488

The three damaged cylinders were pressure cycled to simulate fueling in the laboratory with interruptions for AE testing. The results are summarized in Table 3. Total emissions decreased as the number of simulated fueling cycles increased in the case of vessels L0749 and L1131. However, cylinder L1122 was different in this regard. The total emissions decreased upon applying 14 230 simulated fueling cycles (3442 hits from 24 958) but increased gradually upon subsequent pressure cycling to 3779 hits after 29 230 simulated fueling cycles.

243

244

ACOUSTICEMISSION: STANDARDSAND TECHNOLOGY

TABLE 2--AE hits (the corresponding counts in paranthesis) for the various segments of

the first 0-2Z 6 MPa (0-4000 psi) cycle Cylinder

Loading

Hold

Unloading

L0529-7/89 L0531-7/89 L1026-3/91 Ll123-4/91 Lll25-4/91 L1126-4/91 Ll132-4/91

1 280 6600 16000 5000 9500 5000 20000

Undamaged Cylinders ( 9 600) 760 ( 6 000) ( 84000) 3400 ( 112000) ( 340000) 56000 (1140000) ( 60000) 20000 ( 260000) ( 160000) 40500 ( 780000) ( 200000) 16500 ( 600000) ( 580000) 40000 (1180000)

172 1 311 880 362 3854 3574 5174

( ( ( ( ( ( (

L0749-10/90 L1122-4/91 L1131-4/91

20000 7200 12000

FRP Abrasion Damaged Vessels (1000000) 68000 (3200000) (150000) 11600 ( 3 2 0 0 0 0 ) (560000) 31000 (1400000)

3241 6158 5803

(47953) (87674) (45658)

3 076) 14753) 8874) 6253) 19587) 55345) 20925)

Results of Post AE Investigation Hydrostatic testing was carried out on the three FRP damaged cylinders following the last set of pressure cycling and AE tests. The permanent hydrostatic expansion was measured upon pressurization to and depressurizing the vessels from 34.5 MPa (5000 psi). Permanent expansion figures expressed as a percentage of total expansion (elastic and plastic) were as shown in Table 4. These figures suggest that cylinder L1122 underwent a larger permanent expansion than did the other two damaged vessels. Each of the three vessels was subsequently pressurized to successively higher pressures until expansion continued to occur without an increase in pressure. Tests were interrupted at about 32% total expansion. The highest pressure reached ( estimated burst pressure) is shown in Table 4. Table 5 shows the burst pressures of as-received cylinders and those containing through thickness FRP flaws of various lengths.Vessels without apparent fiber breakage, gave burst pressures within a narrow range of 61.9 -+2.6 MPa (8975 -+375 psi). These values are well in excess of the design burst minimum (ANSI/AGA-NGV2) figure of 51.7 MPa (7500 psi). The burst pressure decreased rapidly with about a 52 mm (2 inch) long FRP cut, but a gradual decrease occurred thereafter with the length of cut. It is noteworthy that the three cylinders that sustained in service abrasion damage (Fig. 4) had estimated burst pressures following the simulated fueling in the laboratory (Table 4) that would be equivalent to approximately a 102 mm (4 inch) axial-radial through FRP cut.

AKHTAR AND KUNG ON NGV CYLINDERS

245

TABLE 3 o A E data at intervals offueling simulation

Cylinder

Cycles

L0749-10/90

0* 10 749 15000 0* 14 230 19 230 29230 0* 14 230 19 230 29230

Ll122-4/91

L1131-4/91

*

Test Cycle 1 All 260 Hits dB 91241 6 458 8 404 245 4431 87 24 958 1 446 3 442 97 3 652 98 3 779 86 48 803 6070 5 870 166 3 897 104 3 125 96

Test Cycle 2 All 260 dB Hits 55 835 2095 7 768 161 3056 65 15 516 554 2 320 50 2640 49 2 618 40 26 791 2 488 4 523 116 3 170 122 2 630 84

Cycle 2/Cycle 1 All 260 dB Him 0.61 0.32 0.92 0.68 0.69 0.75 0.62 0.39 0.67 0.51 0.71 0.52 0.69 0.47 0.55 0.40 0.77 0.70 0.81 1.17 0.84 0.87

As-received

TABLE 4--Cylinder characteristics after simulated fueling Simulated Fueling Cycles

Permanent Expansion %

Estimated Burst Pressure MPa (psi)

L0749-10/90

15 000

3.5

45.5 (6 600)

L1122-4/91

29 230

5.4

44.9 (6 500)

L1131-4/91

29 230

3.6

Cylinder

45.5 (6 600) ..........

Liner Flaw Dimensions length x depth mm (inch) 6.4 x 1.2 (0.25 x 0.047) 20 X 0.3 (9.79 X 0.012) 12.4 X 4.2 (0.49 X 0.165)

Cylinders that had in-service abrasion damage were sectioned transversely following the simulated fueling (Table 2) and the hydrostatic test noted earlier. Magnetic particle inspection of the inner surface of the liner revealed bands of superficial flaws that are commonly encountered following NGV service of steel vessels [I]. One significant crack, however, was found on each of the three cylinders examined. The location of that significant liner flaw is shown in Fig. 6 in relation to the FRP-abrasion damage for all three vessels. The dimensions of those major liner flaws are shown in Table 3. The internal diameter of the liner was slightly larger at the locations containing the liner flaw which suggested that these were the locations of the bulge, although the bulge was not visually apparent.

246

ACOUSTICEMISSION: STANDARDSAND TECHNOLOGY

TABLE 5--Burst pressures of as-received cylinders and those with artificially induced FRP damage Cylinder L0529-7/89 L053t-7/89 L1126-4/91 L1132-4/91 L1026-3/91 L1125-4/91 L1123-4/91

Axial Through Thickness FRP Cut mm (inch) ... ... ... ... 51 (2) 102 (4) 152 46)

Burst Pressure MPa (psi) 63.8 (9 250) 64.5 (9 350) 59.3 (8 600) 60.7 (8 800) 49.0 (7 100) 44.7 (6 480) 41.7 46 050)

Discussion

Visual inspection has been proposed recently for the periodic inspection of CNG cylinders on natural gas vehicles (CGA pamphlet C6.4-1996). The evidence gathered in the present work suggests that visual examination has significant shortcomings for the periodic inspection of metal lined FRP hoop-wrapped cylinders on two accounts. Figure 7a shows the liner flaw in cylinder L1131 following 29 230 simulated fueling cycles, 8 AE test cycles, and after hydraulic pressurization to 45.5 MPa (6600 psi). The liner flaw had grown through roughly 67% of the liner thickness. Figure 7b is a schematic drawing of the transverse section of the cylinder through that liner flaw. Along the radius of the cylinder containing the liner flaw, a crack had initiated in the FRP (Fig. 7b). The crack in the FRP, however, was not on the outer surface of the cylinder. Hence, it could not be detected visually. The FRP crack had initiated at the interface between the metallic liner and the FRP (Fig. 7c) and had propagated about 20% (1.3 mm or 0.05 in.) through the thickness of the FRP (Fig. 7d). As shown in Fig. 7, cracks well in excess of the tolerance limit of 0.254 mm (O.010 in.) depth, the latter considered acceptable for cuts, scratches and gouges mentioned in the CGA-C6.4 pamphlet, may remain in the FRP of metal lined hoop-wrapped cylinders, without being detected through visual inspection. Secondly, the inspection criteria (CGA-C6.4) are such that vessels that are fit for usage must be removed from NGV service. As seen from Table 4, cylinders L1122 and L1131, in spite of FRP damage in NGV service, had a remaining life each in excess of 29 230 cycles and had burst pressures equal to or in excess of 44.9 MPa (6500 psi). Such vessels are suitable for NGV service from a fitness for purpose standpoint and yet they must he removed according to CGA-C6.4. Therefore, there exists a need to use alternative methods for the periodic inspection of CNG cylinders in NGV service.

AKHTAR AND KUNG ON NGV CYLINDERS

.................................

1 ~ ...................................

i 9 L079-10/90

II j

L1131-4/91

i I I 1

!

i I

.............................

i-"Damage

.....

'--~5~

1 FIG. 6---Orientation of liner flaws in relation to FRP abrasion damage. This illustration is a two dimensional representation of the curved cylinder surface obtained through making an imaginary axial cut and flattening. Acoustic Emission Testing Acoustic emission remains an attractive option because of its potential for in-situ testing, i.e. without removing the cylinder from the vehicle. However, that potential may be realized only through demonstrated correlation between acoustic emission and structural integrity of the vessel. The total number of hits measured in the present work during the first test cycle (Table 1) are not helpful in that regard. The as-received cylinders subjected to burst tests (Table 3) gave values in the range of 61.9 • 2.6 MPa (8975 • 375 psi). This scatter in the measured values of burst pressure is small being • yet the associated acoustic emission hits ranged from 2212 (L0529) to 65 174 (L1132). More importantly, this wide range of hits did not show a systematic variation with burst pressure within the narrow span of burst pressures measured. This lack of correlation between burst pressure and total acoustic emission noted above, is also observed if one uses loading, hold and unloading segments for the analysis of hits or counts as seen from Table 2. An analysis of "knee pressure" has been proposed by Mitchell and Newhouse [2] based on their work with one type of all composite NGV cylinders. These last mentioned authors have claimed that a lower "knee pressure" implies a lower burst pressure. Some of the data obtained in the present work are shown in Fig. 8 as a plot of cumulative acoustic hits vs pressure during the first test cycle. A

247

248

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

cursory examination of Fig. 8 shows that for a given criterion to define "knee pressure" (such as 1000 hits) one obtains a wide range of knee pressures even though the cylinders have a narrow range of scatter in their burst pressure. One arrives at similar conclusions as regards felicity ratio and emissions during pressure hold at 27.6 MPa (4000 psi).

The Sources of Acoustic Emission Reasons for emissions at low pressures (Fig. 8) and the variability in the quantity of emission from vessels having similar structural integrity may be considered. It is noteworthy that the steel liner and the FRP both produce AE in hoop-wrapped cylinders. Moreover, emissions may occur at their interface due to the cracking of the coating applied sometimes on the liner for corrosion protection prior to the application of the FRP. Remnants of such a coating are seen in Fig. 7c. The AE sources in the FRP are: matrix cracking which may occur in varying degrees in NGV service due to ultraviolet exposure, etc., delaminations which occur in the transverse plane resulting from the axial expansion of the vessel (Fig. 4) and fiber breakage. Of these, the first two have no direct bearing on structural integrity. Although fiber breakage occurs when the structural integrity of the vessel has been compromised (Fig. 7), the occurrence of fiber breakage is not necessarily an indication of structural integrity loss. In an idealized hoop-wrapped cylinder, the fiber axis will lie along the circumference of a circle. However, in practice the roving of fibers used for the fabrication of cylinders contains many misaligned and twisted fibers. Such fibers will fracture at relatively low pressures or even in the absence of cylinder pressurization with time, since the FRP remains under tension due to the autofrettage treatment. An attempt has been made with some success by Walker et al. [3] through neural networks and amplitude distribution of acoustic events to predict the burst pressure of all composite cylinders in their as fabricated state. Matrix cracking and fiber breakage were considered relevant sources by those authors, while debonding was considered not relevant. However, the situation appears to be different with hoop-wrapped cylinders removed from NGV service. As shown in Fig. 5, the amplitude distribution remains similar for vessels with and without gross fiber damage. The explanation for that similarity possibly lies in the varibility of transverse delamination (Fig. 4) observed with NGV cylinders removed from service. Amplitude distribution is therefore unlikely to be of value when applied to cylinders removed from NGV service. Another source of emission is the oxide scale left on the interior surface of the steel due to fabrication heat treatment. In their search for a correlation between the quantity of emission and structural integrity of all-steel NGV cylinders, Akhtar et al. [4] eliminated the oxide scale associated emissions and established a quantitative relationship between crack depth and AE. In hoop-wrapped vessels, the autofrettage treatment forces the steel liner into compression while the vessel remains at low pressures. The cracked and debonded oxide scale interfaces will produce AE as a result even at low pressures.

AKHTAR AND KUNG ON NGV CYLINDERS

FIG. 7--Cylinder L1131-4/91 after 29 230 simulated fuelling cycles, AE tests, and pressurization to 45.5 MPa (6600 psi). (a) Fracture face of the liner flaw. (b) Schematic transverse section of the cylinder through the liner flaw. (c) Underside of FRP showing the crack and two delaminations (vertical). (d) Transverse section shows FRP crack originated from the FRP inner surface.

249

250

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

Those sources in the liner, the FRP and their interface may explain emissions at low pressures (Fig. 8) and the variability in the quantity of emission observed in the present work with vessels having similar burst pressures (Tables 1, 2 and 5). An approach to overcome those difficulties in the AE testing of hoop-wrapped cylinders may be to examine relative emissions during two consecutive test cycles (Tables 1 and 3). The ratio of high amplitude hits (>_60 dB) shows a trend towards higher values when the vessel undergoes degradation through simulated fueling (Table 3). A similar approach is recommended in ASTM E 1888-97, "Standard Test Method for Acoustic Emission Testing of Pressurized Containers Made of Fiberglass Reinforced Plastic with Balsa Wood Cores".

Pressure (psi) 0 12 000

1000 1

....

20O0

3000

1

.............. L. . . . . . . . . . . . . . . .

4000 t-i---

10000

8000

. D

8000J

"r"

Cylinder Burst MPa (psi) # L0531

64.5 (9 350)

; L0529

63.8 (9 250)

9 Ll132 60.7 (8 800) 9 L1126 59.3(8600) . Ll122 FiberDarnaged

0oo

j

I

10

20

30

Pressure ( M P a ) FIG. 8--Cumulative hits vs. pressure for as-received cylinders during the first cycle.

AKHTAR AND KUNG ON NGV CYLINDERS

Conclusion 1.

Visual inspection has shortcomings; it is not capable of detecting certain types of flaws in the FRP and it rejects cylinders which are fit for service with a wide margin of safety as observed in the present study. Hence, there exists a need for other nondestructive test methods for the periodic inspection of NGV cylinders in service.

2.

Acoustic emission data obtained through incremental pressurization to 27.6 MPa (4000 psi) holding pressure for 5 minutes and unloading, do not provide useful information for structural integrity assessment.

3.

Relative emissions during two consecutive test cycles hold promise for the periodic inspection of hoop-wrapped NGV cylinders.

Acknowledgment A portion of the work reported here was carried out under Gas Research Institute (GRI-Chicago) sponsorship. The authors are grateful to Steve Takagishi and Marco Liem, formerly of GRI, for their patience and support. The opinions expressed in this paper are those of the authors and not of Powertech Labs. References

[1]

Akhtar, A. and Kung, D., "An Assessment of All-Steel Cylinders Currently in NGV Service for the Storage of Compressed Natural Gas Fuels on Vehicles," Report NGV200-3.19, Gas Technology Canada, Toronto, Ontario, 1996.

[21

Mitchell, J.R. and Newhouse, N., "Techniques for Using Acoustic Emission to Produce Smart Tanks for Natural Gas Vehicles." Paper presented at the Fifth International Symposium on Acoustic Emission from Composite Materials (AECM 5), Sundvall, Sweden, 1995.

[3]

Walker, J.L., Russell, S.S., Workman, G.L., and Hill, E.V.K., "Neural Network/ Acoustic Emission Burst Pressure Prediction for Impact Damaged Composite Pressure Vessels," Materials Evaluation, Vol. 55, No. 8, August 1997, pp. 903907.

[41

Akhtar, A., Wong, J.Y., Bhuyan, G.S., Webster, C.T., Kung, D., Gambone, L., Neufeld, N., and Brezden, W.J., "Acoustic Emission Testing of Steel Cylinders for the Storage of Natural Gas on Vehicles," Nondestructive Testing and Evaluation International, Vol. 25, No. 3, March 1992, pp. 115-125.

251

252

ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

DISCUSSION

L.R. Gambonel--Comment 1--The conclusion that "Visual inspection has some shortcomings; it is not capable of detecting certain types of flaws in the FRP...." is misleading and without basis given the data provided in the paper. The authors have failed to include pertinent design information regarding the intended service life of natural gas vehicle (NGV) cylinders. Specifically, the ANSI/NGV2-1992 standard for NGV cylinders requires the designs to provide 13 000 pressure cycles to service pressure and 5 000 pressure cycles to 1.25 times service pressure. The July 1997 draft revision to the ANSI/NGV2 standard requires that cylinders intended for a 15 year design life provide only 11 250 pressure cycles to 1.25 times service pressure. These performance test requirements are based on actual pressure cycling service conditions experienced by cylinders in NGV service. The cracks in the composite wrap of the three hoop-wrapped cylinders were artificially created in the laboratory by applying test conditions that imposed an excessive number of pressure cycles (fatigue), and overpressurization cycles, several of which far exceeded the allowable design stress for the glass fibers. For example, the cylinders L1122 and L1131 were subjected to the following: (i) First used in NGV service for 4 years (possibly some 1 000 pressure cycles incurred). (ii) Pressure cycled in the laboratory up to 29 230 cycles to 3 500 psi (equivalent to an additional 39 years of NGV service). (iii) Overpressurized in the laboratory 8 cycles to 4 000 psi for AE measurements (not done in NGV service). (iv) Overpressurized in the laboratory to 5 000 psi for a hydrostatic test (not done in NGV service - there is no requirement for periodic retesting). (v) Overpressurized in the laboratory to some 6 500 psi in an attempt to predict the eventual burst pressure (certainly not done in NGV service). As a result, the damage observed in the FRP wrap would have been caused by stress rupture. Neither the fatigue conditions (excessive pressure cycles), nor the stress rupture damage could occur to the FRP wrap in service. Tens of thousands of hoop-wrapped cylinders have been used in NGV service since 1982. Since that time there has never been an incident or failure associated with "hidden" damage. A review by Powertech Labs in 1997 of the condition of hoopwrapped cylinders removed from NGV service ["Condition Assessment of Glass Fiber Hoop-Wrapped Cylinders Used in NGV Service" - Gas Research Institute Report i Engineer, Materials TechnologiesUnit, PowertechLabs Inc. 12388- 88th Avenue, Surrey, B.C. Canada. V3W 7R7.

AKHTAR AND KUNG ON NGV CYLINDERS

GRI-97/0052] did not provide any evidence of "hidden" damage in the FRP wrap. There has never been any reported instance of any metal or metal-lined composite NGV cylinder falling by fatigue associated with pressure cycling in service. Of the only Type 2 (hoop-wrapped) cylinders that have failed in service, both were associated with extensive external damage to the composite wrap that would have been readily detected by visual inspection. ["Cylinder Safety Revisited" by W. Liss - Gas Research Institute Natural Gas Fuels, November 1996]. Comment 2--The authors claim in their discussion that the visually apparent damage on cylinders L1122 and L1131 would be sufficient to warrant their removal from NGV service in accordance with CGA C-6.4; however, both cylinders still had considerable pressure cycle life remaining. As a result, the authors concluded "...there exists a need for other nondestructive test methods for the periodic inspection of NGV cylinders...". The fact that the visual inspection criteria in CGA C-6.4 is conservative for this particular hoop-wrapped cylinder design does not provide evidence that some other inspection method is required. It is a result that must be expected of inspection criteria erring on the side of conservatism. For other cylinder designs the inspection criteria will be less conservative. The purpose of the inspection criteria in the CGA C-6.4 document is as follows: (i) To prevent essentially any damage to the composite wrap (other than scratches from routine handling) from remaining in service. (ii) To have a single set of inspection criteria applicable to all cylinder types, and not confuse inspection staff by trying to establish different criteria for each different cylinder design and cylinder size. Comment 3---The conclusion that AE emissions generated during two consecutive pressure cycles "...holds promise for the periodic inspection of hoop-wrapped NGV cylinders" cannot be justified since the data has been generated from cylinders subjected to excessive pressure cycling and overpressurization conditions that would not occur in service.

A. Akhtar and D. Kung (authors' closure)--The authors appreciate this opportunity to clarify issues surrounding visual inspection. Mr. Gambone has made two assumptions of which one is false and the other speculative at best. His translation of 29 230 laboratory hydraulic pressure cycles into 39 years of NGV service presumably means that a vehicle cylinder may be fueled to a maximum of 750 times a year. While the authors agree that 750 cycles is a reasonable assumption, the implication of the critic that a fueling cycle in NGV service is equivalent to a hydraulic pressure cycle is incorrect. The synergistic action of the dynamic stresses between refuelings and the corrosive contaminants present in natural gas, principally H2S, CO2 and H20, causes accelerated degradation of the cylinder in NGV service which is only beginning to be documented. For example, in a recent study (Ref I of the paper) approximately 350 all-steel cylinders were removed for ultrasonic scanning after they had been in NGV service for up to 13 years. Fifteen among them were pressure cycled in the laboratory for their remaining life assessment. It was concluded that the incubation period (the fatigue regime leading to the nucleation but prior to the growth of cracks) is reduced by a factor of 5 in NGV service when compared

253

254

ACOUSTICEMISSION: STANDARDSAND TECHNOLOGY

with that obtained through hydraulic pressure cycling. Such information is still lacking for the crack growth regime. However, if one applies that equivalence of 5 laboratory hydraulic pressure cycles to one NGV service pressure cycle, the 29 230 laboratory hydraulic pressure cycles reported in the paper (Table 3) would not translate into 39 years as done by the critic but to a further service life of 5.8 years, placing the cylinder well within its intended service life of 15 years. The statements concerning ANSI/NGV2 document (comment 1) would have one believe that 11 250 cycles were applied over a period of 15 years using natural gas for fueling to a peak pressure of 1.25 times the service pressure through each cycle. Neither a reference to this effect is provided in the said ANSI/NGV2 document nor is it likely that such precise information has been or would be generated. What is certain is that the figures used in the ANSI/NGV2 document (and in other standards as well), which are set with the state of the knowledge at the time of formulation of the standard, will undergo revision as new information (such as that contained in Ref I of the paper) becomes available. In his systematic enumeration of the cylinder treatment used to create the FRP flaw, Mr. Gambone has overlooked the important fact that Type 2 cylinders made from fiber rovings (all the cylinders examined in the paper were made in this manner) are subjected to an autofrettage pressure of approximately 40 MPa (5 800 psi) as apart of the fabrication process. However, the evidence suggests that neither the autofrettage treatment nor the laboratory pressure cycling (including the subsequent overpressurization) caused stress rupture damage. Had there been stress rupture damage, it would be widespread. The damage seen to the underside of the FRP (Fig. 7 in the paper) was confined to the region immediately above the flaw in the steel (which was only part way through the metallic liner). The authors believe that this FRP flaw, which cannot be detected through visual inspection, to have been a result of the liner flaw. The final overpressurization cycle to 44.9 MPa (6 500 psi) did not produce a detectable change in the liner flaw depth. There was only a slight increase in the axial length of the flaw as seen in Fig. 7a. It is conceivable that a smaller FRP flaw would have resulted had there been no overpressurization. The salient point conveyed in the paper, however, is that the creation of a liner flaw over the next 5.8 years of NGV service would produce FRP damage that cannot be detected through visual inspection. The second assumption made by Mr. Gambone that there has never been any incident of failure associated with "hidden" damage is speculative at best. With reference to the two Type 2 (hoop-wrapped) cylinder failures in NGV service, he has stated that both were associated with extensive external damage to the composite wrap that would have been readily detected by visual inspection. When a Type 2 cylinder ruptures in a catastrophic manner under gas pressure as opposed to hydraulic pressure (the former being the case with the two incidents under discussion), the composite material is obliterated adjacent to the region of the FRP flaw which might have caused the failure. Thus the relevant material not being available, post failure analysis is carried out on the adjacent regions of the FRP which have not been obliterated. Hence, the existence of FRP surface flaws at these adjacent regions, revealed through post failure analysis, does not preclude the possibility that the failure occurred as a result of FRP flaws that were not on the surface. The evidence from materials examined in the adjacent regions being

AKHTAR AND KUNG ON NGV CYLINDERS

circumstantial, the conclusion regarding the nature of the FRP flaw that caused the failure may at best be considered speculative. A reference has been made by Mr. Gambone to the project carried out at Powertech Labs on behalf of Gas Research Institute (GRI Chicago) for the evaluation of Type 2 cylinders removed after they had been in NGV service. That work was done by the present authors (A. Akhtar and D. Kung). A conclusion of that investigation, transmitted to GRI, was that visual inspection has shortcomings. The GRI did not wish to see such a conclusion in its report. Upon request from GRI, an alternative interpretation was provided by L.R. Gambone, C.T. Webster and J.Y. Wong of Powertech Labs Inc. who concluded that visual inspection is an acceptable periodic inspection method. The latter interpretation was accepted by GRI. The statement made by Mr. Gambone that the review by Powertech Labs of hoop-wrapped cylinders removed from NGV service did not provide any evidence of "hidden" damage in the FRP is correct to the extent that the draft report submitted by the present authors to the GRI and to Mr. Gambone et al for their reinterpretation did not contain the evidence shown in Figure 7 of this paper. That information was gathered later. A single rationally based acceptance criterion when applied to a number of cylinder designs may reject cylinders of one of those designs with a wider margin than that dictated by fitness for purpose for that specific design. However, this is not the case with the visual inspection of NGV cylinders. That so called "margin" is such that 2 out of 3 cylinders shown in Figure 6 of the paper (L079-10/90 and L1133-4/91) would have life limitation occurring at locations far removed from the band which the visual inspection has identified as being FRP flawed. In other words, not only is the acceptance criterion not rationally based, the visual inspection method itself is not rationally based as far as the NGV cylinders are concerned. If one adds to the shortcomings of visual inspection identified in the present work the fact that certain types of significant impact damage on carbon fiber wrapped vessels can not be detected visually2'3, visual inspection becomes unsuitable indeed for the inspection of NGV cylinders.

2 Christoforou, A.P., and Swanson, S.R., "Strength Loss in Composite Cylinders Under Impact", Trans ASME, JEMT, Vol. 110, April 1988, pp 180-184 3 Kaczmarek, H., and Maison, S., "Comparative Ultrasonic Analysis of Damage in CFRP Under Static Indentation and Low Velocity Impact", Composites ,Scienceand Technology, Vol. 51, 1994, pp 11-26

255

STP 1353-EB/Oct. 1999

Auihor Index

A

M

Akhtar, A., 236 Markelov, A. P., 125 Miller, R. K., 67 Mitchell, J. R., 224 Murthy, C. R. L., 141

B

Blackburn, P. R., 209 Bodine, H. L., 191 C

N

Calva, M. V., 79 Carlyle, J. M., 67, 191 Chow, D. L., 3 Curtis, C. E., 3

Nagaraja Rao, G. M., 141 Nesvijski, E. G., 41 Nikulin, S. A., 125

D

O

Dawes, R. L., 191 De Aguiar, P. R., 107 Demeski, R., 191

Ohtsu, M., 25, 156, 175 Okamoto, T., 25

Finkel, P., 67 Fowler, T. J., 50 Fultineer, R. D., Jr., 224

P

Patterson, T. S., 79 Pollock, A. A., 67 Puckett, J. A., 50

G Gandy, T. J., 79 H

R

Hamilton, H. R., lIl, 50 Hatano, H., 93 Henley, S. S., 191 Hill, E. v. K., 191

Rajachar, R. M., 3 Raju, N. M., 141

K

Khanzhin, V. G., 125 Kishi, T., 25 Kohn, D. H., 3 Kung, D., 236 Kurianova, E. Y., 125 Copyright9

by ASTM International

Shigeishi, M., 25, 175 Shiotani, T., 156 Shtremel, M. A., 125 257 www.astm.org

258 ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY T

Weissman, N. A., 3 Willett, P., 107

Tafuri, A. N., 67 Ternowchek, S. J., 79 Y W Watts, D. J., 67 Webster, J., 107

Yezzi, J. J., Jr., 67 Yuyama, S., 25

STP 1353-EB/Oct. 1999

Subject Index A

Defect detection, 191 Deformation, 125

Aging dock, 25 Aircraft, aging, 191 Amplitude, 3

E

Eigenvalue analysis, 175 Embrittlement, 79

B F

Bone, microdamage, 3 Burn, workpiece, 107 Burst testing, 224, 236 B-value, 156 C Calibration procedure standard, 93 Cold proof testing, 191 Composite wrapped pressure vessels, 224, 236 Concrete beams, structural integrity, 25 Concrete, reinforced beams, structural integrity, 25 structures, characterization, 41 Corrosion, 79 reinforced concrete beams, 25 Cortical bone, fatigue, 3 Crack characterization, rocks, 141 Crack determination, 79 Crack lengths, 3 Crack measurement, 125 Crack mechanisms, twodimensional model, 175 Crack resistance, 41 Crack, shear, 25 Crack volume estimation, 175 Curve-fitting techniques, 156 Cyclic loading test, 25

Fatigue, aging aircraft, 191 Fatigue damage, traffic signal poles, 50 Fatigue testing, bone, 3 Flaw depth, 209 Fracture, 125 brittle, traffic signal structures, 50 rock, 141 G Gas containers, natural, 224, 236 Gas cylinders, seamless steel, 209 Graphical analysis, 156 Green's function, simplified, for moment tensor analysis, 175 Grinding, 107

lnconel, workpiece burn detection, 107 Iron alloys, material quality monitoring, 125 K

Kaiser effect, 25

Japanese Societ~r for NonDestructive Inspection calibration procedure standard, 93

D

Damage zone dimensions, 3 Davies-bar technique, 175 259

260 ACOUSTIC EMISSION: STANDARDS AND TECHNOLOGY

Laser opto-interferometer, bar end oscillation measurement, 175 Leak detection, buried pipelines, 67 Linear location mode, 209 Loading unloading, 25 l_x)cation methods, buried pipelines, 67 la)ngitudinal wave, 93 M Material quality monitoring, 125 Mechanical integrity testing, 79 Metals gas cylinders, seamless steel, inspection, 209 material quality monitoring, 125 steel-lined hoop wrapped, 236 workpiece burn detection, 107 Microscopy, confocal, 3 Mineralogy, rock fault formation, 141 Moment tensor analysis, 175 N Natural gas vehicle, 224, 236 Niobium, material quality monitoring, 125 P

Parametric plot, 141 Peak-amplitude distributions, 156 Pipelines, buried, leak detection, 67 Plastic, continuous fiberreinforced, 236 Pneumatic proof pressurization, 191 Pressure processing, 125 Pressure vessels gas distribution, seamless steel, testing, 209 inspection, ultrasonics, 79

natural gas, filament wound, 224 natural gas, hoop wrapped, steel lined, 236 Process safety management, 79 Proof testing, 191 R

Rate process analysis, 156 Rayleigh wave, 93 Reciprocity method, 93 Reference standards, laboratory, pipeline leaks, 67 Rocks, crack characterization, 141 S Screening, in-service, traffic signal poles, 50 Sensor calibration, 175 Shear cracking, 25 Shear wave analysis, 79 SiGMA-two-dimensional procedure, 175 Signal difference location technique, 67 Signal enhancement, 67 Signalprocessing, 41, 107, 191 Slope failure prediction, 156 Standards characterization, concrete structures, 41 laboratory reference, 67 proposed, gas-filled filament-wound pressure vessels, 224 proposed, reinforced concrete structural integrity, 25 transducer calibration, reciprocity method, 93 Steels alloy, pressure vessel testing, 209 bearing, workpiece burn detection, 107

INDEX 261 dual phase, material quality monitoring, 125 gas cylinders, steel-lined, 236 Stiffness, 41 Storage tank, mechanical integrity, 79 Strength, concrete structures, 41 Stress effects, rock, 141 Superconductors, 125

V Volumetric strain, 141 Volumetric testing technique, 224

W

T Tin alloys, material quality monitoring, 125 Titanium alloy superconductors, 125 Traffic signal poles, 50 Transducer calibration, 93

Wave attenuation, 156 Waveforms analysis, 141, 175 b-value, 156 digital, 191 Welds, 50 Wind-induced vibrations, 50 Workpiece burn, 107

U Ultrasonic imaging, 141 Ultrasonic testmg, 79 U.S. Department of Transportation, seamless pressure vessel inspection, 209

Z

Zirconium-tin-niobium-iron alloys, 125

O~ !

rU

!

ILl

0

D !

Z