Microscopic Measurement of Strain Fields with Digital Image Correlation and Comparison to Finite Element Modelling

Abstract:

Digital Image Correlation (DIC) has become more and more important in the field of material characterization and research, especially for strongly anisotropic fiber reinforced materials. Its big advantage over the conventional methods like strain gauges or point based video-extensometers is the full field strain and displacement measurement and the ability to analyze three-dimensional displacements. Although theoretically, the concept of the DIC as a pure image-based method allows it to work on every imaginable scale, its main field of application is in the range, where the region of interest (ROI) has a size between 10 −2 m to 10 −1 m. In this case, imaging is accomplished with the use of high-resolution black and white digital cameras. This work is focused on a smaller scale with ROI sizes between 10 −4 m to 10 −3 m, where a digital microscope is used to create the images. The innovative idea behind this work is using the natural surface structure of a polished carbon fiber reinforced Polyamide-6 sample, produced by automated fiber placement, as a statistical pattern instead of the usual speckle pattern applied to the area to be investigated. This way the stress and strain distributionin different regions of the investigated sample area can be evaluated and displayed, while the sample is exposed to an increasing mechanical load in form of a three-point bending test. The resulting strain and displacement fields are compared to finite element modeling of the ROI. To provide an accurate model, the image of the sample is first segmented into fiber, matrix and voids using “Trainable Weka Segmentation” and the resulting phases mapped with the corresponding material properties. To compute the resulting strains in the sample, the measured displacements from the DIC on the edges of the ROI were used as boundary conditions for the simulation. Simulation and experimental results clearly point out the inhomogeneity of the strain field in these samples. Due to the presence of fiber rovings and the presence of voids, local strain values exceed the global average by up to 4 %.

• Periodical: Key Engineering Materials (Volume 809)
• Pages: 575-580
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.809.575
• Online since: June 2019
• Authors: Marco Korkisch*, Markus G.R. Sause
• Keywords: Digital Image Correlation, Finite Element Method Simulation, Microscopy

References

[1] W. G. Knauss, G. Vendroux. Exp. Mech, Vol 38, p.86, (1998).

[2] S. Kumari, S. K. Ghosh and M. F. Buchroithner. Rmt. Sensing and Spatial Inf. Sciences, Volume XL-8, p.531, (2014).

[3] T.A. Berfield, J.K. Patel, R.G. Shimmin, P.V. Braun, J. Lambros and N.R. Sottos. Exp. Mech., Vol 47, p.51, (2007).

DOI: 10.1007/s11340-006-0531-2

[4] M:A: Sutton, J.J. Orteu, H. Schreier. Image Correlation for Shape, Motion and Deformation Measurements. Springer US, (2009).

DOI: 10.1007/978-0-387-78747-3

[5] D. Lecompte, A. Smits, S. Bossuyt, H. Sol, J. Vantomme, D. Van Hemelrijck, A.M. Habraken. Optics and Lasers in Eng. Vol 44, p.1132, (2006).

DOI: 10.1016/j.optlaseng.2005.10.004

[6] M.G.R. Sause. In Situ Monitoring of Fiber-Reinforced Composites: Theory, Basic Concepts, Methods, and Applications. Springer Series in Mat. Sc. Springer International Publishing, (2016).

[7] B. Shonkwiler, Y. He, A. Makeev. Composites Sc. and Tech., Vol 73, p.64, (2012).

[8] H. Schreier A. Makeev, Y. He. Strain, Vol 49, p.440, (2013).

[9] S. Preibisch, S. Saalfeld, and P. Tomancak. Bioinformatics, Vol 25, p.1463, (2009).

[10] I. Arganda-Carreras, V. Kaynig, C. Rueden, J. Schindelin, A. Cardona and H.S. Seung. Trainable segmentation: Release v3.1.2. 08 (2016).

DOI: 10.1093/bioinformatics/btx180

[11] H.P. Degischer and S. Lüftl. Leichtbau: Prinzipien, Werkstoffauswahl und Fertigungsvarianten. Wiley Online Library. WILEY-VCH, (2009).