Fiber and Resin Calculation Using Micrographic Techniques.

Gustavo Francisco Eichhorn

Master degree in Mechanical Engineer and Materials Science

发布日期: 2024年9月14日

The following explains the procedure followed in my thesis, “Influence of Manufacturing Parameters on Residual Stresses in Carbon Fiber-Epoxy Tubes,” for obtaining the fiber and resin percentage using micrographs. These micrographs were obtained from 8 tubes manufactured using Filament Winding. Each tube corresponds to a different run in which various process parameters were modified according to an Experimental Design: Fiber tension, Angle gradient between successive layers, Winding time, Layer thickness, and Bath temperature.

Samples for Micrographic Analysis.

The procedure for obtaining the micrographs is explained below:?

????1- First, samples are extracted from the tubes using a diamond disc or a Dremel tool. The author of this work recommends the Dremel tool, as the cutting disc is thinner and tends to tear the fiber less. The obtained sample requires sanding prior to the inclusion process. The sample size is approximately 15x15 mm. See Figure 1.

2- Next is the embedding process itself. See Figure 2.

3- After the required time has passed, the grinding process can begin. See Figure 3.

4- Finally, the sample is polished, ensuring no sanding scratches are visible under the microscope. See Figure 4.

Fiber, Resin, and Pore Volume Fraction.?

Below is the methodology for obtaining the volume fraction of fibers, resin, and pores. Micrograph techniques were used to determine the volume fractions of fibers, resin, and pores (This technique has evolved since its early days [M. C. Waterbury and L. T. Drzal, 1989] to today, with modern software allowing the analysis of a large number of images [Brian Hayes & Luther Gammon, 2010]). This technique was employed since previous experiences with acid digestion were not satisfactory.?

Once the samples were prepared according to the procedure described above, micrographs were taken using a camera mounted on the OLYMPUS BX60M microscope at the Materials Management Department of the Constituyentes Atomic Center (CAC). The following micrographs were obtained from the samples: 5 micrographs at 5X magnification, 5 micrographs at 10X magnification, 5 micrographs at 20X magnification, and 5 micrographs at 50X magnification. For 10-layer samples, 25 micrographs at 10X magnification were taken to better cover the entire sample thickness.

The 20X magnification micrographs were used only to cross-check the fiber volume results being obtained. This is because a large number of samples (i.e., essentially taking many more pictures at 20X magnification) would be required to cover all the observable defects. Since in all cases there was good agreement between the fiber percentages obtained from the 5X and 10X micrographs and those from the 20X ones, the results from the latter are not included in this work.

The 50X magnification images are more suitable for checking the orientation of the fibers in each layer. However, given that the fiber orientations were clearly evident, they were also not included in this analysis.

Once the micrographs were obtained, the images were post-processed to extract the needed information, such as the percentages of fiber, resin, pores, as well as their morphology, distribution, and fiber angles. This was done using ImageJ? software (For more information on this software, visit https://imagej.nih.gov/ij/download.html.). Figure 5 shows an example of the post-processing carried out using macros in ImageJ to color the fibers, resin, and pores differently from the original micrograph. This procedure (adapted from a NASA tutorial on the subject: https://ntrs.nasa.gov/api/citations/20170001570/downloads/20170001570.pdf.) was carried out for all images of all samples, which took a total time of approximately one month.

As an example, the micrographic analysis of Run 1 is shown in Table 1. As can be seen, a more detailed analysis of each micrograph is extremely extensive. For document readability purposes (45 pages of micrographs), the micrographic analysis of each run is included in Appendix 1.

The summary of the weight percentage of fiber, resin, and pores for each sample is shown in Table 2.?

Considering all tubes from Run 7, the average percentages of pores, fiber, and resin are 4.54%, 59.28%, and 36.19%, respectively, with an expected standard deviation of 1.58%, 8.80%, and 7.24%.?

Qualitative Results from Micrographs.

The following are personal observations of the author based on the analysis of several micrographs.

????? Cracking in Low-Angle Layers: Generally, cracks were observed in patterns with layers at angles of ±5°. Figure 6 summarizes some of these micrographs. While a few micrographs showed cracks in other areas, these are believed to be caused by sample processing before micrographic analysis. Conversely, in samples with layers at ±5°, these cracks were present in all cases. Refer to Appendix 1 for a more in-depth look at this. The author believes that due to the difficulty in obtaining neat patterns at ±5°, future work should focus on more repeatable low-angle patterns such as ±10° and ±15°.

????? Defect Morphology in Different Layers: Defects (pores) in layers with angles of ±89° were generally more elongated compared to those in layers with angles of ±45° and ±5°. Figure 7 shows that defects in low-angle layers are significantly elongated (in runs 1, 13, and 14). In Run 7A, which has all layers at ±45°, the pore morphology is rounded, allowing for easy distinction of defects from ±89° layers in other micrographs due to their more oblong shape.

????? Exceptional Defect Density: Two runs exhibited an exceptional number of defects: Run 2 (9%) and Run 11 (12.777%). The high defect density in Run 2 is attributed to the aforementioned lack of cohesion in low-angle layers. The reason for the high defect density and size in Run 11 is unknown. See Figure 8.

Figure 6: Cracks in regions with ±5° layers.?

Figure 7: Difference in pore morphology in different layers.

Appendix 1: Summary of Micrographic Analysis.

Below, each 5X and 10X micrograph used to calculate the percentage of fiber, resin, and pores is presented. Pixels corresponding to fibers, pores, and resins are colored red, blue, and green, respectively.