結合立體數位影像相關法與有限元模型更新方法於結構隱藏損傷評估

Abstract

結構的維護與損傷檢測對於保障結構的使用安全與壽命非常重要，然而在檢測與評估過程中隱藏於表面之下的損傷卻容易被忽視，在現有的檢測方法中，有能力對於隱藏於內部進行檢測的檢測方法所需的成本都非常高昂。 在本研究中將打破過往影像方法僅能對結構表面進行檢測與分析的限制，提出將影像方法中的立體數位影像相關法結合有限元模型更新方法對於結構模型進行重建，藉此達到對於結構內部損傷進行檢測與評估的目的，為影像方法與內部損傷檢測方法提出一個全新的發展方向。 本文提出的方法共分為三個步驟，首先使用立體數位影像相關法獲取結構表面的真實位移與應變，此時所獲取的資料僅包含表面資料，因此需藉由下一階段的模型更新進行結構重建，此階段分為兩個步驟，首先透過拓樸優化對於結構的參數分布進行初步的估計，藉此劃分健康區域與損傷區域，接著透過參數掃描對於健康區域與損傷區域進行參數數值的更新，經由此流程便能對於結構隱藏於表面之下的區域進行損傷檢測與評估。 為了驗證本研究所提出方法對於損傷檢測與評估的準確性，本研究透過數值實驗的方式產生已知的損傷區域與參數分布，透過將數值實驗的表面資料進行模型更新流程，對於模型結構進行重建，透過重建結果進行損傷檢測與評估，並與數值實驗已知的正確答案進行比較，驗證所提出方法的準確性。最終為了證實本研究於實際應用上的可行性與適用性，以數值實驗討論了較為複雜的損傷情況，並實際的透過立體數位影像相關法進行表面資料的獲取並應用於所提出之方法，進行結構重建以對於隱藏於表面之下的損傷進行評估。

Maintenance and damage detection of structures are crucial for ensuring their safe usage and longevity. However, damage hidden beneath the surface during inspection and assessment processes can easily be overlooked. Existing detection methods capable of inspecting internal hidden damages incur significantly high costs. In this study aim to break the limitations of conventional imaging methods, which are only capable of detecting and analyzing structures' surface, by proposing the integration of Stereoscopic digital image correlation techniques with finite element model updating methods. This approach aims to reconstruct structural models, thereby enabling detection and assessment of internal damages. This novel approach presents a new direction for both imaging methods and internal damage detection methods. The method proposed in this paper consists of three main steps. Firstly, Stereoscopic digital image correlation is used to obtain real displacements and strains on the surface of the structure. At this stage, the acquired data only includes surface information. Therefore, the next stage involves structural reconstruction through model updating, which consists of two steps. Initially, topological optimization is employed to estimate the distribution of parameters in the structure, thereby delineating healthy and damaged areas. Subsequently, parameter scanning is utilized to update the parameter values in both healthy and damaged areas. Through this process, detection and assessment of damage in regions hidden beneath the surface of the structure can be achieved. To validate the accuracy of the damage detection and assessment method proposed in this study, numerical experiments were conducted. Surface data obtained from numerical experiments underwent a model updating process to reconstruct the structural model. Subsequently, damage detection and assessment were performed based on the reconstructed results, and comparisons were made with the settings of numerical experiments to verify the accuracy of the proposed method. Finally, to demonstrate the feasibility and applicability of the proposed method in real-world applications, Stereoscopic digital image correlation was utilized to obtain surface data, which was then applied to the proposed method for structural reconstruction, allowing for the evaluation of damage hidden beneath the surface.