Non-Destructive Testing of Industrial Materials Using Inverse Techniques

当感兴趣的数量可以独立测量时，工业质量控制测试就能直接工作。然而，实际测量通常将几个量组合在一起，然后需要分离。这可以通过使用“逆”计算来实现。“向前”计算将是将所有目标量相加，如果它们是已知的，以确定将进行的综合测量。逆计算将各种量分开，难度大得多，在数学上也很敏感。这里讨论的两种反计算应用在物理上不同，但它们在数学上有很大的共性和互补性。锯木厂使用X射线计算机断层扫描技术进行原木分选：为了在国际市场上有效竞争，加拿大锯木厂必须利用可用的原材料生产最高价值的产品。这可以通过检查进入锯木厂的每一根原木，并战略性地对其进行处理，以获得最高价值的产品来实现。这种分类必须做好，否则就会出现大量浪费，要么是优质材料被用于低价值产品，要么是低质量材料无法制造高价值产品。X射线计算机层析成像(CT)为识别原木内部的质量控制特征提供了一种很有前途的方法。CT测量在医学领域得到了很好的应用，但由于其复杂性、成本和机械敏感性，不太适合锯木厂的工业应用。这里一直在努力开发一种简化、经济和坚固耐用的CT扫描仪设计。所采用的方法的关键是观察到测量的原木及其内部特征具有非常特定的几何形状，例如，原木是圆柱形的，心材和边材轴对称排列，节径向排列。因此，可以开发专门的CT反演方案来利用该提前信息，从而降低对所需测量精度和计算量的要求。在实验室测试中取得了非常有希望的结果，目前的建议是针对进一步的功能开发，如螺旋扫描和实际演示。工业部件中的实际残余应力测量：残余应力是指在没有任何外部载荷的情况下存在于材料中的锁定应力。这些应力很重要，因为它们会显著影响工业部件的尺寸稳定性和材料强度。如果在制造过程中没有发现和控制，可能会发生大量的材料损耗和过早的产品故障。残余应力的“隐蔽性”使其难以可靠地测量，需要使用“逆”评估。残余应力的测量通常是通过测量当某些受力材料被移除时发生的变形来完成的，例如，通过钻一个小孔。数字图像相关(DIC)和电子散斑干涉(ESPI)等光学技术因其非接触、快速、经济等优点而备受关注。它们提供了全场光学数据，使复杂的解释大大超出了更典型的最小数据测量方法。传统上，DIC和ESPI使用单色光。这项拟议的研究旨在利用多色光学测量中可用的更大数据量。这允许从名义上的2-D数据进行3-D变形识别，并能够更复杂地评估残余应力，包括一些目前的测量技术所不能获得的平面外应力分量。

Industrial quality control testing works straightforwardly when the quantity of interest can be measured independently. However, practical measurements often combine several quantities that then need to be separated. This can be done by using an “inverse” calculation. The “forward” calculation would be to sum all the target quantities, if they were known, to determine the combined measurement that would be made. The inverse calculation separates the various quantities and is much more difficult and mathematically sensitive. The two applications of inverse calculations explored here differ physically but they share substantial mathematical commonality and complementary features. Log Sorting in Sawmills using X-Ray Computed Tomography: To compete effectively in the international market, Canadian sawmills must produce the highest value products from the available raw material. This can be done by examining each log entering a sawmill, and strategically processing it for the highest value product. This sorting must be done well, else substantial wastage will occur either as high-quality material getting used for low-value products, or low-quality material failing to make high-value products. X-ray Computed Tomography (CT) provides a promising way to identify the quality-controlling interior features of logs. CT measurements are well established in the medical field, but are not well suited for industrial use in sawmills because of their complexity, cost and mechanical sensitivity. Work has been ongoing here to develop a simplified, economical and rugged CT scanner design. The key to the approach taken is the observation that the measured logs and their internal features have very specific geometries, for example, logs are cylindrical, heartwood and sapwood are arranged axi-symmetrically, and knots are arranged radially. Thus, a specialized CT inversion scheme can be developed to take advantage of this advance information and thereby reduce the demand on both required measurement accuracy and computational size. Very promising results have been obtained in laboratory tests, and the present proposal is directed towards further functional developments such as spiral scanning and to practical demonstrations. Practical Residual Stress Measurement in Industrial Components: Residual stresses are locked-in stresses that exist in materials without the presence of any external loads. These stresses are important because they significantly affect the dimensional stability and material strength of industrial components. If not detected and controlled in manufacturing processes, substantial material wastage premature product failures can occur. The “hidden” character of residual stresses makes them difficult to measure reliably and requires the use of “inverse” evaluations. Residual stress measurements are typically done by measuring the deformations that occur when some stressed material is removed, for example, by drilling a small hole. Optical techniques such as Digital Image Correlation (DIC) and Electronic Speckle Pattern Interferometry (ESPI) are attractive measurement techniques because they are non-contact, rapid and economical. They provide full-field optical data that allow sophisticated interpretation significantly beyond the more typical minimal data measurement approaches. Traditionally, DIC and ESPI have used monochromatic light. The proposed research is aimed at exploiting the larger data content available in multi-colour optical measurements. This allows 3-D deformation identifications from nominally 2-D data and enables more sophisticated evaluations of residual stresses, including some out-of-plane stress components that are not available with present measurement techniques.