8th Conference on Industrial Computed Tomography, Wels, Austria (iCT 2018) www.3dct.at 1 IN-SITU DAMAGE COMPARISON BETWEEN FABRIC-FABRIC AND FABRIC-UD BONDED CFRP Christian Hannesschläger 1 , Stefanie Rauchenzauner 2 , Christian Gusenbauer 1 , Dietmar Salaberger 1 and Johann Kastner 1 1 University of Applied Sciences Upper Austria, Stelzhamerstrasse 23, 4600 Wels, Austria, e-mail: [email protected], [email protected], [email protected], johann.kastner@fh- wels.at 2 FACC, Fischerstraße 9, 4910 Ried im Innkreis, Austria, e-mail: [email protected]Abstract An interrupted in-situ X-Ray computed tomography (XCT) was used to investigate the influence of different Carbon fibre reinforced polymers (CFRP) material systems on the crack progression and propagation of adhesive jointed CFRP plates. The damage behaviour of adhesive jointed CFRP is decisive for the mechanical strength. For this paper the material systems (I) CFRP fabric plate with another CFRP fabric plate and (II) a CFRP fabric plate with a unidirectional (UD) CFRP plate were in- situ investigated with a high resolution X-Ray computed tomography (XCT) with a voxel size of (6 μm)³ at different load steps. The damage volumes of both material systems were segmented and compared with each other. The results confirmed the stress concentration at the end of the specimens [14]. In general the CFRP-cracks in the plates can be identified as major damage mechanism. The different strength of the two material systems is caused by the missing damage in the UD-plate. This lack of damage is caused by the high crack resistance of the UD-plies in respect to the force direction and results in a higher strength of the system fabric-UD. Keywords: x-ray tomography, in-situ, lap shear testing, adhesive bonded CFRP 1 Introduction Carbon fibre reinforced polymers (CFRP) are widely used in aeronautic industries for their special properties such as light weight, high specific stiffness, high specific strength and the high corrosion resistance [1] . An important manufacturing issue is the bonding of CFRP plates. Therefore adhesives are widely used. A main advantage of the adhesives is the homogenous stress distribution in comparison to other joining techniques [2]. The microstructure of the jointed area is decisive for the mechanical strength of adhesive jointed CFRP plates. In the past several studies investigated the failure behaviour of bonded composites joints [3-6]. Therefore, light optical microscopy or a visual inspection [7] was used to characterize the failures. With these methods a non-destructive 3D tracking of the crack propagation is not possible. High resolution X-Ray computed tomography (XCT) as common non-destructive technique (NDT) is used to characterise defects in CFRP specimens [8-9]. In the past in-situ XCT has already been used to investigate damage mechanism in CFRP [10-11]. H. Kunz et al. showed a technique for particle tracking in polyurethane adhesive bonded CFRP and steel with in-situ XCT while performing a lap shear test [12]. In this paper, the method in-situ XCT is used to determine the damage progression in adhesive during an adapted slotted single lap shear test [13]. With interrupted in-situ XCT measurements under different load conditions it is possible to mechanically test the adhesive jointed CFRP plates and to visualise and study the influence of different CFRP material combinations on the crack progression and propagation. 2 Experimental Method 2.1 Material and Specimen A supported epoxy adhesive was used to bond two different material systems: (I) CFRP fabric plate with another CFRP fabric plate and (II) a CFRP fabric plate with a unidirectional (UD) CFRP plate. In both cases the adhesive layer is 0.2 mm thick. An
10
Embed
IN-SITU DAMAGE COMPARISON BETWEEN FABRIC-FABRIC AND FABRIC … · different load steps with the unloaded specimen allows a determination of the influence of different damage types
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
8th Conference on Industrial Computed Tomography, Wels, Austria (iCT 2018)
www.3dct.at 1
IN-SITU DAMAGE COMPARISON BETWEEN FABRIC-FABRIC AND
FABRIC-UD BONDED CFRP
Christian Hannesschläger1, Stefanie Rauchenzauner2, Christian Gusenbauer1, Dietmar Salaberger1 and Johann Kastner1
Figure 9: Comparison of different load steps of top and bottom regions: damage propagation is visualised minimum intensity projections (thickness: 300 µm) of (a) fabric-fabric and (b) fabric-UD; failure mode: debonding of polyester fabric (green), CFRP-cracks (red), initial
cracks (blue) and pores (violet).
In particular four different defect types can be classified that are colour coded. To understand the different breakage behaviour
of the two different material systems, a segmentation and quantification of each load step is essential.
First visual cracks can be seen at a load of ~2/3 Fmax. These first CFRP-cracks and debonding failures are too small for
segmentation. Nevertheless a quantitative comparison of the load steps 0, 90 and 95% of Fmax,reference was performed. The
volume of each defect type at every position represents the damage evolution. Figure 10 shows colour coded defect volume
progression of the system fabric-fabric (pores = violet, debonding of polyester fabric = green, CFRP-cracks = red). It is
obvious that the initial pore volume is nearly constant over all load steps. The shape of the debonding is defined by the
polyester fabric (see Figure 11: row (a)). CFRP-cracks can be identified as the main defect type. It can be observed that they
are growing from top and bottom, in an angle of 45° to the y-axis, to the specimen centre. This behaviour is expected, since the
orientation correlates to the ply orientation.
0 N 1400 N 1500 N
Figure 10: Fabric-Fabric system colour coded defect type 3D comparison at different load steps (pores = violet, debonding of polyester fabric
= green, CFRP-cracks = red). The regions (a) and (b) are visualized in detail in Figure 11.
bo
tto
m
to
p
Z
Y
(b
) fa
bri
c-U
D
(a)
fab
ric-
fabri
c
Y
X 1500 µm
(b)
(a)
CF
RP
-cra
cks
deb
on
din
g o
f p
oly
este
r fa
bri
c
init
ial
crac
ks
po
res
8th Conference on Industrial Computed Tomography, Wels, Austria (iCT 2018)
www.3dct.at 8
Figure 11 shows the visualized failure types, debonding and CFRP cracks in detail.
unsegmented slice segmented slice segmented: 3D
Figure 12 shows the defect volume progression of the fabric-UD. In the unloaded stage, larger pores and smaller cracks in the
fabric-plies are visible. A propagation of these initial cracks cannot be detected. It also can be observed that these initial
defects are not connected with the CFRP-cracks and debonding volume. Due to this fact, an influence of these defects as a
crack origin can be neglected. Furthermore, a growth of debonding volume from load step 2100 N to 2200 N can be observed.
It is noticeable that CFRP-cracks can only be detected in the fabric in the bottom region. The orientation of these cracks
correlates with the orientation of the fabric plies.
0 N 2100 N 2200 N
Figure 13 shows a defect volume comparison between the investigated material systems over the segmented load steps. It can
be clearly seen that there is no noticeable increase of initial cracks and pores at the final breakage. In general the CFRP-crack
volume is always larger than the debonding volume. At the fabric-UD system a nearly parallel progression of debonding and
CFRP-cracks can be observed. In opposite to this, a much higher increase of CFRP-crack volume than debonding volume can
be observed at the fabric-fabric system, from load step 1400 to1500 N. Also the volume proportion of debonding to CFRP
cracks is at the system fabric-fabric higher than at the system fabric-UD. The difference can be explained by the missing cracks
in the UD-plate.
Z
Y
Z
Y
Figure 11: Visualization (slice- and 3D-images) of the segmented defects: (a) debonding and (b) CFRP-crack
Figure 12: Fabric-UD system colour coded defect type 3D comparison at different load steps (initial cracks = blue, pores =
violet, debonding of polyester fabric = green, CFRP-cracks = red)
Z
X
UD Fabric
300 µm
300 µm
(a)
(b)
8th Conference on Industrial Computed Tomography, Wels, Austria (iCT 2018)
www.3dct.at 9
Figure 13: Defectvolume comparison of the systems fabric-fabric and fabric-UD over the segmented load steps
In both material systems the crack initiation occurs mainly in the adhesive layer (see Figure 9). But only in the fabric-fabric
system CFRP cracks are in both plates visible. This lack of CFRP-cracks in the UD-plate explains the lower CFRP-crack
gradient (step from 1400 to1500 N) at the fabric-UD system in comparison to fabric-fabric system. It can be concluded that the
stress is not high enough for a crack initiation in the UD-plies. The main advantage of UD in this case is that the force and the
plies are orientated in the same direction.
5 Conclusion and Outlook
In general the interrupted in-situ XCT measurements allowed a detailed look on the defect propagation of adhesively bonded
CFRP during an adapted slotted single lap shear test. It has been seen shown that the influence of the substrate on the damage
mechanism can be determined in a qualitative and quantitative manner. The results of the interrupted in-situ scanned fabric-
fabric specimen were compared with the fabric-UD specimen. The stress concentration at the end of the specimen [14], in the
overlapping region, can be seen in both materials. A voxel size of (6 µm)3 is sufficient to detect visually the first defects at ~2/3
of Fmax,reference. At a load step of 90% of Fmax,reference the defects could be segmented and quantified. Four different failure types
could be observed. An influence of initial defects (pores, cracks) from the unloaded step, on the final breakage cannot be seen.
The crack initiation event can be determined as the debonding of the polyester fabric. After this first damage step, the cracks
are growing mainly through the adhesive nearest fabric ply until the final fracture occurs. This major influence of the CFRP
cracks is determined by the high defect volume ratio of CFRP cracks to debonding of polyester. It can be concluded that the
higher strength of the fabric-UD system is caused by the higher crack resistance of the UD in comparison to the 45° plies of the
fabric. All plates show a potential crack initiation behavior by debonding of the polyester fabric, but only the strength of the
UD-plate is high enough to prevent a growth of those early cracks. The lack of damage in the UD plates leads to a significant
higher strength of the system fabric-UD. In a further step, an improved system of fabric-fabric with one UD-layer next to the
adhesive and the system UD-UD should be investigated. This single UD-layer may increase the crack resistance of the plate.
To sum up in case of single lap shear strain the system fabric-UD shows a better resistance against crack growth. With the
usage of fabric-UD instead of fabric-fabric the maximum force can be increased by ~37 percentages.
Acknowledgements
The authors like to acknowledge FACC Operations GmbH and CoLT Prüf und Test GmbH for the manufacturing of the
specimen. This work was supported by the K-Project for “non-destructive testing and tomography plus” (ZPT+) and by the
COMET program of FFG and the federal government of Upper Austria and Styria and supported by the project “multimodal
and in-situ characterization of inhomogeneous materials” (MiCi) and by the European Regional Development Fund (EFRE) in
the framework of the EU-program IWB2020.
References
[1] B. Jahn, D. Karl. Composites-Marktbericht, 2012
[2] G. Habenicht. Kleben: Grundlagen, Technologien, Anwendungen. Springer Berlin Heidelberg New York, 2006
[3] K. S. Kim, J. S. Yoo, Y. M. Yi and C. G. Kim, Failuremode and strength of uni-directional composite single
lap bonded joints with different bonding methods, Compos. Struct., vol. 72, no. 4, pp. 477–485, 2006.
[4] J. K. Kim, H. S. Kim and D. G. Lee, Investigation ofoptimal surface treatments for carbon/epoxy composite
adhesive joints., Journal of adhesion science and Technology, vol. 17, no. 3, pp. 329–352, 2003.
8th Conference on Industrial Computed Tomography, Wels, Austria (iCT 2018)
www.3dct.at 10
[5] T. Cheuk and L. Tong, Failure of adhesive bonded composite lap shear joints with embedded precrack, Compos. Sci.
Technol.,vol. 62, no. 7-8, pp. 1079–1095, 2002.
[6] K. D. Potter, F. J. Guild, H. J. Harvey, M. R. Wisonm and R. D. Adams, Understanding and control ofadhesive crack
propagation in bonded joints between carbon fibre composite adherends I. Experimental, International Journal of