·Home ·Table of Contents ·Materials Characterization and testing

Characterization of the damage area in glass fibre reinforced composites using AE planar localisation C. Santulli, W.J. Cantwell University of Liverpool Dept. of Engineering -Materials Science Liverpool L69 3GH Contact

1. ABSTRACT

The aim of this work was to characterise the damage area in impacted glass woven laminates loaded in three-point bending using the acoustic emission technique. The first system was based on a glass fibre fabric/polyester and the second on an E-glass mat/ polyester laminate. The thicker weaving of the fibre bundles in the second laminate was employed in order to improve the impact resistance of the material.

2. INTRODUCTION

Impact in composite materials has been widely studied in recent years (1, 2). The use of non-destructive techniques to evaluate impact damage has also received extensive coverage in literature (3, 4). In particular, acoustic emission has been deemed to be able to provide quantitative indications on the extent of impact-damaged area (5) and on the characterisation of residual post-impact properties of the composite (6).

3. EXPERIMENTAL

The two materials investigated in this study were based on E-glass fibre reinforced polyester laminates. The resin employed throughout was an unsaturated polyester with styrene as coupling agent. All of the composites were manufactured using the RTM process, yielding square plates with dimensions 600x600 mm, from which the specimens with dimensions 100x100x5 mm were removed. The fibre fraction of all of the composites was 63% by volume. The specimens were impacted at energies up to 20 Joules on a Rosand impact system fitted with an anti-rebound device using an anvil diameter of 40 mm. Quasi-static three points bend tests were carried out by applying a 25.4 mm diameter cylindrical aluminium indentor to the centre of the specimen. The distance between the two hemispherical steel supports was 60 mm. The loading programme involved fifty bending cycles at loads up to the 10% of the ultimate strength of the laminate, followed by fifty bending cycles up to the 20% of the ultimate strength of the laminate and so on, until the specimen failed. During the post-impact bend tests, the acoustic emission activity was monitored using a Vallen AMSY4 acoustic emission system. The AE set-up involved the application of a threshold of 38 dB and an overall amplification of 36 dB. To obtain a planar localisation of the AE signals, four resonant sensors were placed on the sample to form a 70 mm square grid.

4. RESULTS

The mechanical response of the laminates under three-point loading is discussed. Following this, the acoustic emission results are presented. An analysis of the acoustic emission data was undertaken in order to measure the variation of the damaged area with applied stress. In addition, an attempt was made to characterise the damage using other acoustic emission parameters, such as the amplitude and the average frequency of the signals.

4.1 Mechanical properties
The three laminates were loaded in three-point bending test up to failure. The flexural ultimate stress of the three laminates is reported in Table 12. In both the E-glass mat and the E-glass fabric reinforced laminates, a decrease in the ultimate stress with increasing impact energy is evident only in the specimen impacted at 20 Joules.

Impact energy	Ultimate flexural stress (MPa)
Joules	E-glass mat	E-glass fabric
5	498.65	1225.91
7.5	452.27	1292.14
10	450.22	1286.04
12.5	430.18	1163.66
15	411.39	1040.77
20	340.09	746.83
Table 1: Values of the ultimate flexural stress on the laminates impacted at different energies

The values reported in Table 1 were used to define a loading programme for fatigue testing. After fifty bending cycles at 10% of the ultimate flexural stress, fifty bending cycles at 20% were performed and so on, until the specimen failed.

As expected, none of these laminates were able to withstand cycling at loads close to their ultimate stress. The stress at which failure occurred, normalised by the ultimate flexural stress, is shown in Table 2. The E-glass mat reinforced laminates do not appear to suffer any significant reduction in their mechanical properties following post-impact cyclic bend tests. In contrast, failure of the E-glass fabric reinforced laminates occurred at a significantly lower stress.

Impact energy (J)	E-glass mat	E-glass fabric
5	90%	70%
7.5	80%	60%
10	80%	60%
12.5	90%	60%
15	80%	60%
20	80%	70%
Table 2: Percent of ultimate stress at which the failure occurred

4.2 Analysis of the acoustic emission data


Fig 1: Localisation of the AE events on the surface of the E-glass mat/polyester (left) and E-glass fabric/polyester laminates (right) impacted at energies from 5 to 20 Joules.The numbers on the axes correspond to distances in cm.

By placing four acoustic emission sensors on the laminate, it was possible to employ a two-dimensional localisation procedure. This procedure was conducted in order to observe the increase in the size of the damaged area with increasing applied stress in laminates impacted at different energies. Figure 72 shows the spatial disposition of the acoustic emission events on the surface of the impacted E-glass mat/polyester and E-glass fabric/polyester laminates. The numbers at the corners of each plot indicate the position of the sensors, whilst the grey zone indicates the regions beyond the sensors which were excluded from the examination. Here, the displacement stress applied to the laminates was the highest that could be achieved with failure occurring (i.e., increasing the stress by a further 10%, resulted in failure). By comparing the plots of the E-glass mat (left) and the E-glass fabric (right), it is clear that the shape of the damaged area is different in the two cases. The coarser weave of the E-glass mat ensured that the damage extended over a larger area of the laminate. However, smaller amounts of damage were noted in the E-glass mat/polyester laminates. This is shown by the fact that the distribution of the acoustic emission events is more highly concentrated in the E-glass fabric reinforced laminates than in the E-glass mat reinforced laminate. As a consequence, the E-glass mat reinforced laminate was able to withstand flexural loading up to a higher percentage of its ultimate tensile stress than the E-glass fabric reinforced laminate. The different shape of the localisation areas reflects differences in the mode of failure between the two laminates (7).

To measure the size of the damage area from the acoustic emission localisation plot, the region of the laminate, in which the 90% of the acoustic emission events were detected, was identified. This criterion ensured the exclusion of acoustic emission events, which may have no connection with damage. This region is referred to as the Acoustic Emission 90% Area (AE90). Figs.2a and 2b show the AE90 areas for the E-glass fabric reinforced laminate and the E-glass mat reinforced laminate respectively.

Fig 2a: Histogram of the AE90 areas during bend tests on the E-glass fabric reinforced laminate impacted with energies between 5 to 20 Joules

Fig 2b: Histogram of the AE90 areas following bend tests on the E-glass mat reinforced laminate impacted at energies between 5 and 20 Joules

Fig 3: Amplitude distribution histograms during bending fatigue tests on an E-glass fabric reinforced laminate impacted with an energy of 5 Joules

The amplitude distribution of the acoustic emission events during the bend test was also studied. From this study, it was possible to observe two typical amplitude distribution patterns. At low stresses, the amplitude of AE events tended to be randomly distributed over a range between 50 and approximately 90 dB. In contrast, at high stresses, the amplitude distribution histogram assumes a bell-shaped appearance, similar to that observed in a normal statistical distribution. A reason for this is that the total number of events is much greater in this case. A possible interpretation of this behaviour relates to the presence of different damage mechanisms (matrix cracking, delamination and fibre failure) (8). The threshold stress at which the amplitude distribution histogram passes from random to bell-shaped was evident in all of the laminates that were tested. Figure 70 shows the amplitude distribution histogram for an E-glass fabric reinforced laminate impacted with an energy of 5 Joules. In this case, the amplitude distribution during the three point bend tests appears to be random for stresses up to 40 % of the ultimate stress and bell-shaped at 50 and 60 % of the ultimate stress. Failure occurred during fatigue loading at 60% of the ultimate stress.

5. CONCLUSIONS

The coarser weave of the E-glass mat resulted in damage extending over a larger area of the laminate. Smaller amounts of damage were noted in the E-glass mat/polyester laminates. This was demonstrated by the fact that the distribution of the AE events was more highly concentrated in the E-glass fabric reinforced laminates than in the E-glass mat reinforced laminate. As a consequence, the E-glass mat reinforced laminate offered a superior load-bearing capacity in flexure. The two-dimensional AE localisation procedure yielded accurate measurements of the growth of the impact damage area with increasing applied stress. In addition, the different shapes of the localisation areas that were obtained from the two laminates accurately reflected differences in the modes of failure.

REFERENCES

1. S. Abrate, Impact on Composite Structures, Cambridge University Press, 1998.
2. W.J. Cantwell, J. Morton, The Impact Resistance of Composite Materials - a review, Composites, 22, 347,-362, 1991.
3. N. Liu, Q.M. Zhu, C.Y. Wei, N.D. Dykes, P.E. Irving, Impact Damage Detection in Carbon Fibre Composites using Neural Networks and Acoustic Emission, Key Engineering Materials 167-168, 43-54, 1999.
4. D. Almond, G. Busse, E. Grinzato, J.C. Krapez, X. Maldague, V. Vavilov, W. Peng, Infrared Thermographic Detection and Characterisation of Impact Damage in Carbon Fiber Plastics, QIRT Conference, Lodz 1998.
5. K.S. Downs, M.A. Hamstad, Acoustic Emission from Depressurization to Detect/Evaluate Significance of Impact Damage to Graphite/Epoxy Pressure Vessels. Journal of Composite Materials. 32, 258-307, 1998.
6. O.Y. Kwon, J.H. Lee, D.J. Yoon, Acoustic Emission during Tensile Loading of Low Velocity Impact-Damaged CFRP Laminates, Journal of Acoustic Emission 13, 83-88
7. C. Santulli, F. Billi, L'Emissione Acustica su Prove di Trazione dopo Impatto di Laminati Compositi Tessuti, 9a Conferenza AIPnD, Padova, 1997.
8. I. Roman, A. Mittelman, H. Harel, G. Marom, Mechanical Behaviour and Acoustic Emission Characterization of Kevalr-Epoxy Composites Loaded in Tension, 2^nd AECM Conference, Montreal, 85-89, 1986.

© AIPnD , created by NDT.net

|Home|    |Top|