We investigated three different automated methods for identification and quantification of meso-scale damage on single-layer composite laminates subjected to ballistic impact by a 5.5 mm (0.22 caliber) right circular cylindrical steel projectile. These methods were based on the combined use of high-resolution optical imaging, fluorescence microscopy, and image detection software. High-resolution images of impacted composite samples were processed using either ImageJ alone (method 1, for sample size of 20 × 20 cm2) or ImageJ plus MATLAB (method 2, for sample size of 20 × 20 cm2), or fluorescence images were processed using ImageJ plus MATLAB (method 3, for sample size of 6 × 5 cm2). These methods were used for identification and grouping of horizontal and vertical transverse tow cracks, and 45° matrix cracks. The identified and grouped damage modes were quantified based on spatial pixel count related to damage modes, and quantified damage modes were used to generate a digital damage map using a separate MATLAB code for all three methods. Finally, the advantages and shortcomings of each of these three automated methods for identification and quantification of meso-scale damage modes were evaluated via comparison to a baseline manual method for identifying and quantifying the damage modes on the same samples.

中文翻译：

---

实现平纹玻璃/环氧树脂复合层压板中尺度损伤模式的自动识别和量化

我们研究了三种不同的自动化方法，用于识别和量化受到 5.5 毫米（0.22 口径）圆柱钢弹丸弹道冲击的单层复合层压板的中尺度损伤。这些方法基于高分辨率光学成像、荧光显微镜和图像检测软件的结合使用。单独使用 ImageJ（方法 1，样本大小为 20 × 20 cm2）或 ImageJ 加 MATLAB（方法 2，样本大小为 20 × 20 cm2）处理受影响复合样本的高分辨率图像，或处理荧光图像使用 ImageJ 加 MATLAB（方法 3，样本大小为 6 × 5 cm2）。这些方法被用于水平和垂直横向拖曳裂纹和45°基体裂纹的识别和分组。基于与损坏模式相关的空间像素计数对识别和分组的损坏模式进行量化，量化的损坏模式用于对所有三种方法使用单独的 MATLAB 代码生成数字损坏图。最后，通过与用于识别和量化相同样品的损伤模式的基线手动方法进行比较，评估了这三种用于识别和量化中尺度损伤模式的自动化方法的优缺点。