Current Proceedings on Technology

Yazarlar: Gozde Inan Sezer, Ahmet Burak Goktepe, Alper Sezer, Kambiz Ramyar, Faruk Tuncer Dilek, Murat Tuyan

Konular:-

Anahtar Kelimeler:PC mortar,Sulfate attack,Cement type,Statistical analysis,Data clustering.

Özet: Sulfate ions derived from ground or seawater may cause undesirable concrete structures. Extent of the damage primarily depends on chemical composition of cement, type of mineral admixture as well as its inclusion level, and the concentration of the sulfate. In this investigation, it was aimed to characterize the sulfate expansion level of portland cement (PC) mortar by use of  three types of portland cements (having different C3A contents), two types of mineral additives (limestone and fly ash), and exposed to two types of sulfate solution (Na2SO4 and MgSO4). In order to distinguish the sulfate resistances of late-age PC mortars in terms of cement type, different clustering algorithms, hard k-means (HKM), fuzzy c-means (FCM) and self organizing maps (SOM), are employed to classify the expansion levels of mortars caused by sulfate attack. Moreover, representative sulfate expansion levels of each cement type are obtained by the analyses. Due to the results of the numerical studies, SOM is found to be unsuccessful; nevertheless, promising outcomes are obtained by HKM and FCM algorithms. The results revealed that the normalized expansion level can be considered as a determining factor for distinguishing the cement type. Consequently, the numerical procedure is presented to determine the effect of cement type as well as to perform backcalculation analysis procedure using late-age sulfate expansion experience. 

---