Printer Friendly

Acoustic emission rate behavior of laminated wood specimens under tensile loading.

Abstract

Quasi-static tensile tests on two different types of laminated wood specimens are monitored with acoustic emission (AE). AE signal parameter data are analyzed with respect to AE rate behavior under quasi-static tensile loading to failure. For various AE signal parameters an exponential increase of the cumulative AE and of the AE rate curves is observed with time or load. The time-constant of this exponential increase is material-dependent. This is analogous to the AE rate behavior observed for various glass-fiber reinforced polymer-matrix composite specimens under similar tensile load conditions [1]. Implications of these results for damage accumulation in composites under tensile loading are discussed.

Keywords: Laminated wood, tensile test, damage accumulation, AE rate analysis

Introduction

Acoustic emission (AE) monitoring is an established technique for assessing the integrity of structures made from various materials (see, e.g., [2]). There are several standardized applications developed specifically for structures made from fiber-reinforced polymer-matrix composites (see, e.g., [3-5]). Typically, such structures are loaded by appropriate means and the resulting AE is analyzed with respect to AE activity and AE intensity, yielding an indication of the level of structural integrity. It is less common to apply AE monitoring to (standardized) materials tests in order to monitor damage accumulation. A series of such tests with different loading configurations, material and specimen types have been reported by Brunner et al. [1]. The main conclusion was that for quasi-static tests, a specific, exponential increase of the AE activity was observed with increasing load. The exponential increase (proportional to exp([alpha]t)) could even be quantified by evaluating the exponential factor [alpha]. This factor was empirically shown to be independent of the AE signal parameter (AE hit, counts, energy) used to analyze the AE activity pattern, and also to be independent of the test temperature (within a specified range), but dependent on the strain rate.

Since wood can be interpreted as being a composite material also, it is of interest to see whether an analogous behavior is observed for the AE activity during quasi-static tests on wood specimens. The present paper reports on a first series of tensile tests with two different types of wood specimens.

Experimental Procedures

The specimens were of dog-bone shape, about 400 mm long, 20 mm wide gauge, 18 mm thick, and made of two different types of laminated wood (Fig. 1). The first type was medium density fiber-board (labeled Medium Density Fiberboard = MDF) with a density of 720 kg/[m.sup.3], the second particle board (labeled SP) with a density of about 650 kg/[m.sup.3]. Three specimens per type were tested with quasi-static tensile crosshead displacements of either 1.7 mm/min or 2.5 mm/min.

AE was monitored with AE equipment (AMS-3, Vallen Systeme GmbH) with four resonant AE sensors (SE-150M, Dunegan Engineering Corp. Inc.). Two sensors each were mounted on either side of the specimens, at the top and bottom near the mechanical clamps, respectively (Fig. 1). The AE signal parameter set was recoded with a threshold of 31.2 dBAE, a rearm-time of 1.382 ms, and with 100 kHz to 1,000 kHz band-pass frequency filters. Preamplifier gain was set at 34 dB. A silicone-free vacuum grease was used as coupling agent and the AE sensors were mounted with metal springs. AE source location was performed by linear interpolation between the two sensors (top and bottom) on either side of the specimens. An average signal speed was used, derived from Hsu-Nielsen sources applied on the surface of the specimens prior to the test.

Tensile loads were applied in a screw-driven test machine (Z100, Zwick) with a 50-kN load cell (HBM Z4). Strain was measured with a video extensometer (ME-46, Messphysik, Austria) between two marks (Fig. 1) applied within the gauge length of the specimens. Load and crosshead displacement were also recorded as analog external parameters with the AE equipment.

[FIGURE 1 OMITTED]

Results and Discussion

The tensile properties of the specimens are summarized in Table 1. Because of different strain-rates (crosshead speeds) used in the tests, no averages and standard deviations are quoted.

[FIGURE 2 OMITTED]

The ultimate tensile strength in board plane of the two specimen types differs by about a factor of four, i.e., 18-19 MPa for MDF versus 4-5 MPa for particle board (SP), while the failure strains differ by about a factor of two, i.e. 0.7-1.0 % for MDF versus 0.4-0.5 % for particle board (SP). Typical macroscopic failure patterns are shown in Fig. 1. Strain rates are estimated from measured strains at maximum load and duration of the test (note that the specimens shown in Fig. 1 failed outside the measurement range).

An initial analysis checks whether an exponential behavior of AE rate and cumulative AE with time (and load in quasi-static tests) is observed. Selected examples of different AE signal parameters for one MDF specimen are shown in Fig. 2. If there is an overlap of the AE rate and the cumulative AE curves, this indicates an exponential behavior with time. The data as measured show some cases of overlap and others, where the two curves clearly deviate. If, however, the plot scales are adjusted, it can be seen that for the latter curves, an exponential behavior is observed at least for the initial part of the test, with a deviation of the two curves after a certain time (Fig. 3).

[FIGURE 3 OMITTED]

From the ratios between the maximum scale values of the overlap plot of the cumulative and rate curves, the exponential factor [alpha] can be calculated. It is simply the ratio between the values of the AE rate curve and of the cumulative AE curve. This ratio has been calculated for all AE sensors on each specimen, using the ring-down count ("AE counts") as AE signal parameter. These values are listed in Table 2, together with the measured strain-rate. For selected specimens, the value of [alpha] has been calculated also based on the overlap plot of AE hits with time. The results (examples shown in Fig. 3) essentially yield the same values for the same value of strain rate. This confirms that as was the case for the GFRP specimens, the exponential behavior does not depend on the choice of the AE signal parameter for the analysis.

The values in Table 2 also show that for comparable crosshead speeds or strain-rates the MDF specimen type yields higher values of [alpha] than the SP type. This is similar to the situation observed for the GFRP materials, which yielded distinctly different values of [alpha] for a given strainrate as well. Experiments varying the strain-rate over a fairly large interval indicated a linear increase of [alpha] with increasing strain-rate [1]. With the assumption that this linear behavior should include the origin (as it does seem to be the case for the GFRP), the dependence of [alpha] for the wood specimens can be plotted (Fig. 4). From that graph, a reference value for a fixed strain-rate (or crosshead speed) can be evaluated and compared with corresponding values for GFRP. Taking a crosshead speed of 1 mm/min, for example, yields values of a around 0.25 for MDF and around 0.35 for particle board (SP). This can be compared with values of 0.010, 0.020, and 0.014 [s.sup.-1] for unidirectional glass-polyester, combined glass-mat polyester, and glass-roving epoxy (at 1 mm/min), respectively [1].

Table 2 AE rate behavior analysis: Determination of exponential coefficient[alpha] from hits and counts for the two types of specimens (strain-rate average, where applicable, average [alpha] calculated from all four sensors, stand. dev. = standard deviation)

[FIGURE 4 OMITTED]

The significance of the empirical analysis deriving the value of [alpha] from AE monitoring of simple, quasi-static tensile tests has not been discussed much in the original reference on the GFRP tensile specimens [1]. The exponential increase in AE rate and AE cumulative curves with time or load as shown by the analysis, however, differs qualitatively from previous models for composites, interpreting the steep increase in AE observed during the test as an "AE knee" [6]. Of course, drawing tangents on the initial and final part of the AE cumulative or rate curves will yield an apparent "knee" at the intersection of the two tangents. This intersection point, however, will not separate the curves into a "low" and "high" rate regime with a distinct transition of the activity, since the rate is following an exponential curve.

It is also important to note that not all quasi-static tensile tests on composites will yield the exponential AE rate behavior described above. Using a fiber-reinforced polymer-matrix specimen with a cross-ply lay-up (alternating 0[degrees]/90[degrees] plies, with uniaxial tensile loading along the 0[degrees]- fiber direction) yields an initial exponential increase, followed by a drop in activity and, later, the second increase in the AE rate (the second increase presumably again following an exponential behavior). This type of behavior can be explained in terms of distinctly differing dominant damage mechanisms [7] that occur sequentially. In the case of the cross-ply laminate, the first peak in the AE activity relates to transverse matrix cracking (up to saturation level), later followed by longitudinal matrix splitting upon increasing loading of the fibers along the loading direction. It is hypothesized that the exponential AE rate behavior indicates the action of one single dominant damage mechanism under the specified load type and load-rate. It is interesting that at least for the limited range of materials investigated in the two studies to date an analogous AE rate behavior is observed in spite of the apparently dissimilar nature of the test materials. The common link producing the analogous behavior does seem to be the composite nature of the materials, combining fibers inside a suitable matrix.

Summary

It has been shown that an AE rate behavior analysis applied to tensile tests on different types of glass fiber-reinforced polymer-matrix laminates is also applicable to different types of wood specimens. The AE rate determined from various AE signal parameters (e.g., hits, counts, duration, and energy) essentially indicates an exponential increase with time, or, equivalently with load in the quasi-static tests. The exponential factor [alpha] governing this AE rate increase scales linearly with crosshead speed (or strain-rate). A larger value of [alpha] for a given crosshead speed or strain-rate indicates higher AE rate increase, consistent with larger damage accumulation and lower ultimate tensile strength. Deviations from exponential AE rate behavior in quasi-static tensile tests on these materials are probably related to damage accumulation interfering with AE signal propagation, or a change in dominant damage mechanism.

References

[1] A.J. Brunner, R. Nordstrom, P. Flueler "A Study of Acoustic Emission-Rate Behavior in Glass Fiber-Reinforced Plastics", J. Acoust. Em., 13, (3-4), (1995), 67.

[2] R.K. Miller, E.v.K. Hill, P.O. Moore, Nondestructive Testing Handbook, Vol. 6 Acoustic Emission Testing, 3rd ed., American Society for Nondestructive Testing ASNT, Columbus, USA (2005), Chapters 8 and 11.

[3] Standard Practice for Acoustic Emission Examination of Fiberglass Reinforced Plastic Resin (FRP) Tanks/Vessels, E 1067, Annual Book of Standards, Vol. 03.03, American Society for Testing and Materials, ASTM Intl., West Conshohocken, USA (2001).

[4] Standard Test Method for Examination of Gas-Filled Filament-Wound Composite Pressure Vessels Using Acoustic Emission, E 2191, Annual Book of Standards, Vol. 03.03, American Society for Testing and Materials, ASTM Intl., West Conshohocken, USA (2002).

[5] Standard Test Method for Acoustic Emission Examination of Pressurized Containers Made of Fiberglass Reinforced Plastic with Balsa Wood Cores, E 1888/1888M, Annual Book of Standards, Vol. 03.03, American Society for Testing and Materials, ASTM Intl., West Conshohocken, USA (2002).

[6] J.R. Mitchell, "Standard Test to Quantify the Knee in the AE versus Load Curve as a Material Parameter for Composites", Proceedings 3rd International Symposium on Acoustic Emission from Composites, American Society for Nondestructive Testing ASNT, Columbus, USA (1989), pp. 207-219.

[7] A.J. Brunner, R.A. Nordstrom, P. Flueler, "Fracture Phenomena Characterization in FRP-Composites by Acoustic Emission", Proceedings European Conference on Macromolecular Physics: Surfaces and Interfaces in Polymers and Composites (ed. R. Pick), European Physical Society EPS, 21B (1997), 83-84.

ANDREAS J. BRUNNER, MARTIN T. HOWALD * and PETER NIEMZ *

Empa--Swiss Federal Laboratories for Materials Testing and Research, Laboratory for Mechanical Systems Engineering, Uberlandstrasse 129, CH-8600 Dubendorf, Switzerland.

* ETH Swiss Federal Institute of Technology, Institute for Building Materials, Schafmattstrasse 6, CH-8093 Zurich, Switzerland.
Table 1 Summary of mechanical properties from tensile
tests on laminated wood specimens ([[sigma].sub.max] =
ultimate tensile strength, [epsilon] at [F.sub.max] =
failure strain, [F.sub.max] = maximum load)

Specimen     E-modulus   [[sigma].sub.max]   [epsilon] at
Type/label   (tensile)   [MPa]               [F.sub.max]
             [MPa]

MDF / 1        3,100           18.7              0.87
MDF / 2        2,620           18.9              1.02
MDF / 3        4,240           18.6              0.65
SP / 1         1,700            5.3              0.38
SP / 2         1,350            4.8              0.45
SP / 3         1,170            4.6              0.40

Specimen        [F.sub.max]   cross-     crosshead
Type/label                    section    speed
                [N]           [mm2]      [mm/min]

MDF / 1            7172        380.3        1.7
MDF / 2            7233        381.9        2.5
MDF / 3            6987        375.8        2.5
SP / 1             1951        368.4        2.5
SP / 2             1701        354.7        1.7
SP / 3             1634        357.7        1.7

Table 2 AE rate behavior analysis: Determination of exponential
coefficient a from hits and counts for the two types of specimens
(strain-rate average, where applicable, average [alpha] calculated
from all four sensors, stand. dev. = standard deviation)

Specimen   crosshead   strain-   AE          average   stand.   stand.
Type       [mm/min]    rate      parameter   [alpha]   dev.     dev.
                       [%/min]               [-]       [-]      [%]

MDF        1.7         0.31      hit         0.0432    0.0045   10.4
MDF        1.7         0.31      counts      0.0452    0.0068   15.1
MDF        2.5         0.40      hit         0.0561    0.0046    8.1
MDF        2.5         0.40      counts      0.0627    0.0097   15.5
SP/2       1.7         0.41      hit         0.0614    0.0103   16.8
SP/2       1.7         0.41      counts      0.0592    0.0090   15.2
SP/3       2.5         0.56      hit         0.0796    0.0056    7.0
SP/3       2.5         0.56      counts      0.0898    0.0112   12.5
COPYRIGHT 2006 Acoustic Emission Group
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Brunner, Andreas J.; Howald, Martin T.; Niemz, Peter
Publication:Journal of Acoustic Emission
Date:Jan 1, 2006
Words:2364
Previous Article:Elastic waves from fast heavy-ion irradiation on solids.
Next Article:AE measurements for superconducting devices.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters