鹼-骨材反應(AAR)是一種對混凝土建築物有害的化學反應，反應會形成膠體並膨脹造成混凝土的開裂，因此在量測膨脹量是一個重要的工作。傳統的AAR之膨脹量量測只能得知兩測點間的平均應變，但混凝土建築內含有骨材、水泥砂漿以及孔隙，並非均質材料且由AAR引致之應變亦非均勻，而應變與裂縫發生之機制也有待建立。
本文透過製作兩組試體將其放入ASTM C1293養護環境以及室溫中，利用掃描器掃描影像改善了以傳統相機拍攝造成的問題，並透過數位影像法量測分析試體內部全域應變場及應變不變量，以探討AAR引致應變與裂縫發生之機制。探討的主題包括：(1)AAR引致裂縫之發展過程、(2)裂縫延伸方向與主應變方向之關係、(3)以最大概似法建立羅吉斯迴歸曲線探討裂縫發生之機率與應變大小之關係以及(4)利用ROC曲線找出最佳分界點探討由早期應變預測裂縫發生之可行性。結果顯示由早期的應變建立之裂縫預測圖與晚期實際裂縫圖做比對兩者吻合度頗高。

AAR is one of dangerous chemical reaction in concrete building. The reaction will form the basoid which will expand and make the crack. So it is important to measure the expansion. In the past, the tranditional method of measurement in AAR only can get the average expansion between two observed points. But the concrete building contain aggregate, mortar and pore. It is nonhomogeneous material and the strain which is induced by AAR is not uniform. The mechanism is not established between strain and crack.
In this study, we made two concrete sample and placed into curing of environment with specification and room temperature. The images was scanned by scanner to improve the image which is taken by trainditional camera. We can get the strain filed by digital image correlation and take this to explore the mechanism which is induced by AAR between strain and crack. The discussion of topic include: (1) the crack development process which is induced by AAR; (2) the relationship between the direction of ceack and the direction of maximum major strain; (3) establish the logistic regression cureve with maximum likelihood estimation method to explore the relationship between the probability of crack occurrence and the magnitude of strain; (4) getting the best breakpoint by ROC curve to explore the feasibility of predicting the occurrence of crack from early strain. The result show that the comparison between the crack prediction established by the early strain and the actual crack in the late stage is quite high.

1.田永銘、朱正安、鐘翊展、盧育辰、林奕佐、高嘉偉，「台灣東部公路橋梁混凝土鹼-骨材反應傷害之研究」，交通部公路總局材料試驗所研究報告 (2016)。
2.李惟誠，「以數位影像法量測鹼-骨材反應引致之應變場與裂縫發展之關係」，碩士論文，國立中央大學土木工程學系，中壢 (2018)。
3.林奕佐，「台灣反應性粒料之反應行為及抑制研究」，碩士論文，國立中央大學土木工程學系，中壢 (2017)。
4.陳彥安，「數位影像相關係數法應用於衛星影像崩塌地判釋之研究」，碩士論文，國立高雄大學土木與環境工程學系，高雄 (2011)。
5.郭蔚楓，「邏輯斯迴歸與ROC曲線整合應用於探討道路鋪面破壞影響因素—以台26線為例」，碩士論文，國立屏東科技大學土木工程學系，屏東 (2018)。
6.張弼超，「運用羅吉斯迴歸法進行山崩潛感分析—以台灣中部國姓地區為例」，碩士論文，國立中央大學應用地質研究所，中壢 (2015)。
7.曾舜隆，「數位影像相關性於大變形量測之應用」，國立中興大學，碩士論文，台中 (2014)。
8.葉柏村、張光宗，「山崩淺感模型的逐步建構與決策分析—以陳有蘭集水區為例」，中華水土保持學報，第50卷，第2期，第49-62頁 (2019)。
9.趙崇敏，「應用數位影像相關技術於平面位移量測」，碩士論文，國立中興大學機械工程學系，台中 (2006)。
10.趙逸幃、張光宗，「應用羅吉斯迴歸建立台 21 線公路崩塌潛感預測模式」，水土保持學報，第46卷，第4期，第1223-1234頁 (2014)。
11.謝孟勳，「以網格集群最大概似法建立建築物易損性曲線及其應用於震災緊急運送路徑之選擇」，博士論文，逢甲大學土木及水利工程學系，台中 (2013)。
12.ASTM, “Standard Test Method for Determination of Length Change of Concrete Due to Alkali-Silica Reaction,” ASTM, C1293 − 08b, United States (2008).
13.ASTM, “Standard Test Method for Potential Alkali Reactivity of Aggregates (Mortar-Bar Method),” ASTM, C1260 – 14, United States (2014).
14.Bing, P., Huimin, X., Zhaoyong, W., Kemao Q., and Zhiyong, W., “Study on subset size selection in digital image correlation for speckle patterns,” Optics Express, Vol. 16, No. 10 (2008).
15.Concrete Society, “Non-structural cracks in concrete,” The Concrete Society, Technical Report, No. 22 (1992).
15.Page, C. L., Page, M. M., Durability of concrete and cement composites, CRC Press, pp.81-82 (2007).
16.Walpole, R. E., Myers R. H., Myers, S. L., Ye K., Probability & Statistics for Engineers & Scientists, Prentice Hall, U.S.A., pp. 276-283 (2002).
17.Zhao, J., Zhao, D., “Investigation of Strain Measurements using Digital Image Correlation with a 18.Finite Element Method,” Journal of the Optical Society of Korea, Vol. 17, No. 5, pp. 399-404 (2013).
19.Zhang, K., “Digital Image Correlation Analysis of Alkali Silica reaction in concrete with recycled glass aggregate,” Master Thesis, Illinois, United States (2014).