In-situ Tensile Test in ESEM with AE

Sprachauswahl/Language selection

Dieser Artikel ist auch auf Deutsch verfügbar In-situ-Zugversuch im ESEM mit SEA

A service provided by
Polymer Service GmbH Merseburg
Tel.: +49 3461 30889-50 E-Mail: info@psm-merseburg.de Web: https://www.psm-merseburg.de
Our further education offers: https://www.psm-merseburg.de/weiterbildung
PSM on Wikipedia: https://de.wikipedia.org/wiki/Polymer Service Merseburg

In-situ tensile test in ESEM with AE

In-situ tensile test in the environmental scanning electron microscope (ESEM) with simultaneous recording of damage-sensitive acoustic emissions (SEA)

Introduction

To evaluate the damage kinetics, quasi-static in-situ tensile tests are performed on notched test specimens in an environmental scanning electron microscope (ESEM). Due to its mode of operation, an ESEM (environmental-SEM) is suitable for performing in-situ tests on electrically non-conductive materials [1, 2]. The additional inclusion of acoustic emission analysis should allow a morphological evaluation of the damage mechanisms occurring at the crack tip (see: deformation mechanisms). This combination of three methods of polymer testing and diagnostics represents a novel development, and the application of this method should contribute to explaining the relationship between acoustic events and damage mechanisms in short-glass fibre reinforced polymers (see: hybrid methods, examples).

Experimental

A polypropylene reinforced with 20 per cent by mass of short glass fibres (PP/20) was investigated. Due to the non-polar nature of polypropylene, maleic anhydride (MSA) was used as a coupling agent to optimise the bonding of the fibres (see: fibre–matrix adhesion) to the matrix. Kardelky [3] and Schröder [4] demonstrated that Echtblau, compared to other nucleating agents for PP/GF composites, achieves the best mechanical properties at a content of 0.01 % by mass. For this reason, Echtblau was used as the nucleating agent.

The in-situ tensile test with simultaneous recording of damage-sensitive sound emissions was performed at a test speed of 0.2 mm/min on notched test specimens made of PP/20 with dimensions of 60 x 9.6 x 3.6 mm³ and a notch depth of 2 mm in accordance with ISO 527-1 [5] (see: notch geometry) in a Quanta 600 FEG (FEI – EINDHOVEN, THE NETHERLANDS) ESEM. For the SEM images, the skin typical of injection-moulded test specimens was removed by polishing with 4000-grit silicon carbide sandpaper (grain size 5 µm) by 200 µm using a specially manufactured aluminium block under constant cooling with water to prevent the surface of the test specimens from heating up. The test specimens are clamped in the aluminium block and protrude exactly 200 µm at the edge [6]. This allows direct observation of the interaction between the fibres and the matrix (see also: fibre–matrix adhesion). The acoustic emissions were recorded using the AMSY-4 measuring system and the tensile stress was applied using a Deben MT 5000 tensile testing machine (SUFFOLK, UK) installed in the ESEM sample chamber. The tensile testing machine can be controlled via five coordinates, which allows positioning in the X, Y and Z directions as well as rotation and tilting of the clamped test specimen. Figure 1a shows the clamped test specimen with the applied ultrasonic sensor after the end of the test, and Figure 1b shows the built-in and tilted tensile testing machine in the sample chamber.

The notch radius was 0.1 mm, which corresponds to type C according to ISO 179-1 [7], and the strain rate resulting from the test speed and the clamping length of 42.5 mm is 0.0047 min^-1. This theoretical value for the strain rate can be verified by the increase in the traverse path versus time diagram. It was found that a lower strain rate of 0.0027 min^-1 occurs during the tensile stress of the notched test specimen.

A –	piezoelectric force transducer	D –	specimen
B –	extensometer	E –	acoustic sensor; secured with cable ties
C –	clamping jaws	F –	scattered electron detector

Fig. 1:

Tension platform with clamped test specimen and applied acoustic emission sensor after the end of the test (a) and in the sample chamber in a tilted position (b)

Reproducible clamping of the test specimens was achieved with an applied torque of 1 Nm. The force measurement and the measurement of the elongation were carried out with a 5 kN force transducer and via the traverse path. Suitable application of the AE-204A broadband acoustic emission sensor to the test specimen was ensured by using beeswax as a coupling medium. The sensor was secured in place, i.e. prevented from slipping during loading, using a cable tie (Fig. 1), as it was not possible to attach it using a clamp due to the experimental conditions. The tensile testing platform with the test specimen clamped in place was tilted by 30° in order to observe the damage mechanisms occurring at the notch tip during loading (Fig. 2). The position of the backscattered electron detector (solid state detector – SSD) directly on the underside of the pole shoe (Fig. 1b) required a minimum distance for the acoustic sensor, so that the distance between the sensor and the asymmetrical position of the notch was 30 mm.

Fig. 2:

Schematic presentation of the tensile platform in orthogonal and tilted positions

In a scanning electron microscope (SEM), contrast is created by the interactions between the primary electrons (PE) and the test specimen surface. The interaction mechanisms between the primary electrons and the test specimen shown in Fig. 3a can be divided into an analysis part and an imaging part. With the aid of cathodoluminescence, characteristic X-rays and Auger electrons, information about the material, i.e. about the elements and the structure, can be derived [8]. Important types of imaging result from the secondary electrons (SE) and the backscattered electrons (RE). Due to the different output depending on the angle of inclination, the secondary electrons map the topography of the test specimen (Fig. 3b), while the backscattered electrons can reflect material differences due to their dependence on the atomic number Z. To prevent interactions with the medium, the sample chamber must be evacuated. In the case of non-conductive materials, it is necessary to coat the test specimen with an electrically conductive material, e.g. gold, or to contact it in such a way that electrical charging is avoided [9]. This means that biological preparations, curing processes, swelling, drying or in-situ experiments cannot be carried out in the SEM. These investigations can be carried out in the ESEM, as the operating principle is based on the neutralisation of negative surface charge by positively ionised gas, e.g. water vapour, nitrogen or air at low pressures of 0.1–10 Torr.

Fig. 3:

Interaction of primary electrons (PE) with the test sample (a) and dependence of secondary electron yield on surface inclination for mapping topography [8]

The secondary electrons are detected using the Large Field Detector (LFD) and the backscattered electrons using a semiconductor detector (SSD). The contrast is created because the signal differs from point to point due to the varying number of electrons emitted. The more electrons that can be emitted, the brighter the areas appear in the image. Figure 4 shows examples of the material and topography contrast of a short glass fibre reinforced PP material. The ESEM images show the damage processes occurring in the area of the crack tip under uniaxial tensile stress.

Fig. 4:

Different types of contrast in ESEM

As the backscattering coefficient increases with increasing atomic number, as explained above, the glass fibres in the polymer matrix (left section of Fig. 4) are very clearly visible. With the aid of topography contrast, conclusions can be drawn about the plastic deformations in the matrix and at the crack tip (right-hand section of Figure 4). By using the different types of contrast in parallel, more comprehensive conclusions can be drawn about the notched test specimens examined here.

Example

Damage kinetics, fractography and sound emission analysis of PP/20

The damage mechanisms occurring during the tensile test for PP/20 are documented in Figures 5–7, and the correlations between the acoustic emission measured variables and the force–time diagram are shown in Fig. 8. In order to enable comparison of the micromechanical damage mechanisms shown in Figs. 5–7 at selected points in time with the graphical representation of the acoustic emission measured variables in Fig. 8, the points in time were marked with the letters A–C. Figures 5a and b show that small forces in the area in front of the crack tip lead to elastic deformations around the glass fibres and to elastic widening of the crack flanks. At the transition point from linear-viscoelastic to non-linear viscoelastic behaviour (see: elasticity), the load causes irreversible pitting on the glass fibres (Fig. 5a–i and b–i) and blunting (see: stretch zone, ICIT with AE and crack opening) of the sharp surface crack (Fig. 5a–ii and b–ii), which was caused by notching followed by polishing. The onset of irreversible damage to the material after a period of 81 s correlates with the increased number of hits occurring at this point in time. As the load continues to increase, fibrillation of the matrix occurs (Fig. 5c–iii and d–iii) as a result of further widening of the notch, as well as plastic deformation around the glass fibres, corresponding to the lines of force (Fig. 5b–i).

These significant plastic deformations in the region of the glass fibre ends indicate good adhesion of the glass fibres (see: fibre–matrix adhesion) within the polymer matrix, which enables force transfer between the matrix and the fibres and thus leads to local ductility in the adjacent matrix [10, 11]. At this point, the dissipation zone covers a larger volume and an increased number of hits are detected, which is marked with the letter A in Fig. 8 and correlates with the ESEM image in Fig. 5b. In accordance with the literature [10], the dissipation zone is the region in front of the crack tip where the failure processes are distributed over a more or less extensive area in which the stress field acts due to the external loading.

Crack opening and plastic deformation around the glass fibres in the dissipation zone (see: fracture process zone) increase as the force is applied, and this increase in force causes significant plastic deformation ahead of the crack tip in the form of a dimple (Fig. 5c–iv and d–iv).

Fig. 5:

Comparison of the opening of the notch and the blunting of the sharp surface crack (ii) as well as the plastic deformation corresponding to the lines of force on the glass fibre in the area of the dissipation zone (i) after a test duration of 116 s (b) compared to the unloaded state before the start of the test (a) and fibrillation of the polymer matrix (iii) and plastic deformations in front of the crack tip in the form of trough formation (iv) with fragmentation of a transverse glass fibre (v) after a test duration of 280 s (c) and increase in plastic deformations associated with the exposure of glass fibres (vi) after 342 s; the double arrow represents the lines of force (direction of stress) for PP/20

The glass fibre lying directly across from the crack tip is destroyed by the increase in the size of the dissipation zone or the magnitude of the stress field, a process facilitated by pre-existing damage resulting from polishing (see Figs. 5c–v and d–v). The increase in crack opening, in addition to further fibrillation of the matrix, leads to debonding and subsequently to the pull-out of the glass fibres (see: Fibre-reinforced plastics fracture model), i.e. sliding processes occur along the fibres. This manifests itself in an increase in acoustic emission events (Fig. 8–B).

The damage ahead of the crack tip is characterised by the formation of a hole and the ingress of matrix material into the hole, comparable to typical craze growth mechanisms (Fig. 6a–i) [10]. The pull-out of the fibres from the matrix due to crack opening is coupled with friction processes because of the fibres’ good adhesion to the matrix. After 434 s (Fig. 8–C), the plastic deformation has reached a magnitude at which loose ends of the fibres are visible, i.e. the fibres have completely detached from the matrix on one side (Fig. 5d–vi and Fig. 6b–ii), which manifests itself in a peak in the force.

Fig. 6:

Hole formation and matrix material pulling into the holes (i) combined with the exposure of glass fibres in the dissipation zone (ii) after 396 s (a) and local trough formation in front of the crack tip (iii) and pulling out of fibres (iv – compare with b–iv) and visible loose fibre ends (v) due to severe crack tip opening after 434 s; Tearing of the fibre/matrix interface with subsequent fibrillation of the remaining matrix bridges (vi) and alignment of the glass fibre in the direction of the acting load lines (vii) after a test time of 453 s (a) and 474 s (b); double arrow represents the load lines for PP/20 [6]

The damage mechanisms at play at this stage are the fibrillation of the crack flanks, visible surface damage in the form of dimpling, and the pull-out of fibres accompanied by severe plastic deformation of the adjacent matrix. In this process, tearing of the fibre/matrix interface followed by fibrillation of the remaining matrix bridges can be observed (Figs. 6c–vi) . Due to the lower deformation hindrance compared to the fibres inside the test specimen, the glass fibres near the surface can align themselves in the direction of the lines of force (see: glass fibre orientation), which is clearly visible in Figs. 6c–vii and d–vii. Fracture occurs after the maximum force is reached, following a stable crack propagation phase preceding the unstable crack propagation, accompanied by plastic deformation of the matrix. Microfractographic images at various magnifications taken after the end of the test are shown in Fig. 7a–c.

Fig. 7:

Trough and fringe formation in the dissipation zone along the fracture flanks (a–i) as well as tips formation of the matrix material (b–ii and c–ii) and aligned glass fibres not wetted with matrix material (b–iii and c–iii) with good matrix bonding (c–iv); double arrow represents the lines of force for PP/20 [6]

The fracture flanks are characterised by pronounced dimples and fringe formation in the dissipation zone, as can be seen in Figs. 7a–i. Detailed images of the notch base are shown in Figs. 7b and c. Due to the good fibre–matrix adhesion and the associated force transmission in the region of the fibre/matrix interfaces, significant plastic deformation of the matrix in the form of tips formation is evident (Figs. 7b–ii and c–ii). Furthermore, particularly in the region close to the surface, the fibres align in the direction of the lines of force, and as a result of the pull-out of the glass fibres, no interfacial layer remains, i.e. the fibres are not wetted by the matrix material (Figs. 7b–iii and c–iii). The fibres are very well bonded to the matrix (Fig. 7c–iv), causing the matrix material to stretch at the notch root. The sound emission measured variables used for the evaluation are the number of sound emission events (hits), the energy E_AE, the event duration t_ED, and the peak amplitude values A_p of all recorded hits. The functional relationship between the respective measurement parameter and the force–time curve is shown in Fig. 8a–d. The distributive plot, with a time window (Δ_t – bar chart width) of 5 s, corresponds to a division into 100 sections for a test duration of 500 s. Based on the characteristic curves and in light of the results discussed above, the sound emission characteristics can be divided into three sections. In the linear-viscoelastic range (I), there is low acoustic emission activity, and at a test duration of 81 s, the transition from linear viscoelastic to non-linear viscoelastic behaviour occurs (see: elasticity), which is characterised by a continuous increase in acoustic emission events (II). In the third section, after a duration of 434 s, the sound emission activity increases significantly at the peak force due to the stable crack propagation visible in the ESEM images, with the maximum being reached at the point of unstable crack propagation, which is attributable to the fracture of the test specimen and the associated accumulation of damage (III). The division of the sections is highlighted by vertical lines in Figs. 8a–d.

Fig. 8:

Load–time curve of PP/20 at a test speed of 0.2 mm/min with the frequency distribution of hits (a), the energy EAE (b) and the event duration tED (c) as well as the peak amplitude values Ap of all hits; distributive representation with a time window of 5 s [6]

When examining the functional relationship between the increase in force and the distribution of energy (Fig. 8b), a linear regression function was used to mathematically describe sections II and III. The rise in the curves (m_II and m_III) is a measure of the increase in energy resulting from the heightened acoustic emission activity, with a value two orders of magnitude higher being evident for section III. This is attributable to the stable crack propagation preceding the unstable crack propagation and thus the inclusion of a larger volume. The division into three sections of differing acoustic emission activity can be applied to the presentation of the peak amplitude values in Fig. 8d, where the amplitudes lie in the range of 40–80 dB.

See also

• Fibre-reinforced plastics fracture model
• Micro-damage limit
• Fracture process zone
• Tensile test and sound emission analysis
• Tensile test/Sound emission/Thermography
• Scanning electron microscopy (SEM)

References

[1]	Zankel, A., Pölt, P., Ingolic, E., Gahleitner, M., Grein, C.: Fracture Behaviour of Polymers – in situ Investigations in the ESEM. Imaging & Microscopy 7 (2005) 16–18; Download as pdf (access on march 31, 2026)
[2]	Zankel, A., Pölt, P., Gahleitner, M., Ingolic, E., Grein, C.: Tensile Tests of Polymers at Low Temperatures in the Environmental Scanning Electron Microscope: An Improved Cooling Platform. Scanning 29 (2007) 261–269; https://doi.org/10.1002/sca.20075 (access on march 31, 2026)
[3]	Kardelky, S.: Einfluss der Nukleierungsmittelart auf die Deformations- und Bruchmechanismen von medial beanspruchten PP/GF-Verbunden. Master Thesis, Martin-Luther-Universität Halle-Wittenberg (2002), (see AMK-Library under B 3–101)
[4]	Schröder, D.: Kombinierte Wirkung des Faservolumen- und Nukleierungsmittelgehaltes auf das mechanische Eigenschaftsniveau von PP/GF-Verbunden. Master thesist, Martin-Luther-Universität Halle-Wittenberg (2003), (see AMK-Library under B 3–102)
[5]	ISO 527-1 (2019-07): Plastics – Determination of Tensile Properties – Part 1: General Principles
[6]	Schoßig, M.: Schädigungsmechanismen in faserverstärkten Kunststoffen – Quasistatische und dynamische Untersuchungen. Vieweg+Teubner / GWV Fachverlage GmbH, Wiesbaden (2010), (ISBN 978-3-8348-1483-8; see AMK-Library under B 1–21) Open-Access-Publication
[7]	ISO 179-1 (2026-03): Plastics – Determination of Charpy Impact Properties – Part 1: Non-instrumented Impact Test
[8]	Schmidt, P. F. (Ed.): Praxis der Rasterelektronenmikroskopie und Mikrobereichsanalyse. Band 444, Expert Publishing, Renningen, 2nd fully revised Edition (2018) (ISBN 978-3-8169-1597-3)
[9]	Göcke, R., Präparation – Überblick über Präparationsmethoden. In: Schmidt, P. F. (Ed.): Praxis der Rasterelektronenmikroskopie und Mikrobereichsanalyse. Expert Publishing, Renningen (1994) pp. 672–678 (ISBN 978-3-8169-1038-1)
[10]	Michler, G. H.: Kunststoff-Mikromechanik: Morphologie, Deformations- und Bruchmechanismen. Carl Hanser, Munich Vienna (1992), (ISBN 3-446-17068-5; see AMK-Library under F 4)
[11]	Cantwell, W. J., Roulin-Moloney, A. C., Fractography and Failure Mechanisms of Unfilled and Particulate Filled Epoxy Resins. In: Roulin-Moloney, A. C. (Ed.): Fractography and Failure Mechanisms of Polymers and Composites. Elsevier Science Publishers B. V., Amsterdam (1989) pp. 233–290 (ISBN 978-1-8516-6296-8)