Computers and Concrete

  • 제31권2호
  • /
  • Pages.85-95
  • /
  • 2023
  • /
  • 1598-8198(pISSN)
  • /
  • 1598-818X(eISSN)
테크노프레스 (Techno-Press)
권호 표지
DOI QR코드

DOI QR Code

An adaptive neuro-fuzzy inference system (ANFIS) model to predict the pozzolanic activity of natural pozzolans

  • Elif Varol (Department of Geological Engineering, Hacettepe University) ;
  • Didem Benzer (Turkish Cement Manufacturers' Association (TCMB)) ;
  • Nazli Tunar Ozcan (Department of Geological Engineering, Hacettepe University)
  • 투고 : 2021.05.12
  • 심사 : 2022.11.21
  • 발행 : 2023.02.25
⟨ 이전 논문 다음 논문 ⟩

초록

Natural pozzolans are used as additives in cement to develop more durable and high-performance concrete. Pozzolanic activity index (PAI) is important for assessing the performance of a pozzolan as a binding material and has an important effect on the compressive strength, permeability, and chemical durability of concrete mixtures. However, the determining of the 28 days (short term) and 90 days (long term) PAI of concrete mixtures is a time-consuming process. In this study, to reduce extensive experimental work, it is aimed to predict the short term and long term PAIs as a function of the chemical compositions of various natural pozzolans. For this purpose, the chemical compositions of various natural pozzolans from Central Anatolia were determined with X-ray fluorescence spectroscopy. The mortar samples were prepared with the natural pozzolans and then, the short term and the long term PAIs were calculated based on compressive strength method. The effect of the natural pozzolans' chemical compositions on the short term and the long term PAIs were evaluated and the PAIs were predicted by using multiple linear regression (MLR) and adaptive neuro-fuzzy inference system (ANFIS) model. The prediction model results show that both reactive SiO2 and SiO2+Al2O3+Fe2O3 contents are the most effective parameters on PAI. According to the performance of prediction models determined with metrics such as root mean squared error (RMSE) and coefficient of correlation (R2), ANFIS models are more feasible than the multiple regression model in predicting the 28 days and 90 days pozzolanic activity. Estimation of PAIs based on the chemical component of natural pozzolana with high-performance prediction models is going to make an important contribution to material engineering applications in terms of selection of favorable natural pozzolana and saving time from tedious test processes.

키워드

과제정보

This study was financially supported by Hacettepe University (research project number: FUK-2015-6627). We are grateful to Genco Ozcan for his help on the construction of the prediction models, to Sevgi Telsiz for her help in collecting samples, to Selin YONCACI and Serkan TURK for the laboratory test.

참고문헌

  1. Abraham, A. (2001), "Neuro fuzzy systems: State-of-the-art modeling techniques. In Connectionist Models of Neurons, Learning Processes, and Artificial Intelligence", Proceedings of 6th International Work-Conference on Artificial and Natural Neural Networks, Granada, Spain, June.
  2. Ahmad, M., Rashid, K., Tariq, Z. and Ju, M. (2021), "Utilization of a novel artificial intelligence technique (ANFIS) to predict the compressive strength of fly ash-based geopolymer", Constr. Build. Mater., 301, 124251, https://doi.org/10.1016/j.conbuildmat.2021.124251.
  3. Alp, I., Deveci, H., Yilmaz, E., Yilmaz, A.O. and Kesimal, A. (2003), "Investigation of the potential use of the quarry product from tashane, terme as trass raw material in cement industry", Proceedings of the International Symposium on Industrial Minerals and Building Stones, Istanbul, Turkey, September.
  4. Bascetin, A., Eker, H., Adiguzel, D. and Tuylu, S. (2020), "Investigation of pozzolanic properties of some additive materials in cemented paste", Gumushane Univ. J. Sci. Technol., 10(2), 415-424. https://doi.org/10.17714/gumusfenbil.627059.
  5. Bilgehan, M. and Turgut, P. (2010), "The use of neural networks in concrete compressive strength estimation", Comput. Concrete, 7(3), 271-283. https://doi.org/10.12989/cac.2010.7.3.271.
  6. Boukhatem, B., Kenai, S., Hamou, A.T., Ziou, Dj. and Ghrici, M. (2012), "Prediction concrete properties using neural network (NN) with principal component analysis (PCA) technique", Compt. Concrete, 10(6), 557-573. https://doi.org/10.12989/cac.2012.10.6.557.
  7. Cordeiro, G.C., Filho, R.T.D., Tavares, L.M., Fairbairn, E.M.R. and Hempel, S. (2011), "Influence of particle size and specific surface area on the pozzolanic activity of residual rice husk ash", Cement Concrete Compos., 33(5), 529-534. https://doi.org/10.1016/j.cemconcomp.2011.02.005.
  8. Ding, Y., Dong, H., Cao, M., Torgal, F.P. and Azevedo, C. (2020), "Recycled household ceramic waste in eco-efficient cement: A case study", Advances in Construction and Demolition Waste Recycling, Woodhead Publishing.
  9. Ercikdi, B. (2009), "Effect of pozzolanic mineral and chemical admixtures on paste backfill performance", Ph.D Dissertation, Karadeniz Technical University, Institute of Science and Technology, Trabzon.
  10. Erdogan, T.Y. (2003), Concrete, METU Press, Ankara, Turkey.
  11. Gazder, U., Al-Amoudi, O.S.B, Khan, S.M.S. and Maslehuddin, M. (2017), "Predicting compressive strength of bended cement concrete with ANNs", Comput. Concrete, 20(6), 627-634. https://doi.org/10.12989/cac.2017.20.6.627.
  12. Gencoglu, M., Uygunoglu, T., Demir, F. and Guler, K. (2012), "Prediction of elastic modulus of steel-fiber reinforced concrete (SFRC) using fuzzy logic", Comput. Concrete, 9(5), 389-402. https://doi.org/10.12989/cac.2012.9.5.389.
  13. Gokceoglu, C., Yesilnacar, E., Sonmez, H. and Kayabasi, A. (2004), "A neuro-fuzzy model for modulus of deformation of jointedrock masses", Comput. Geotech., 31(5), 375-383. https://doi.org/10.1016/j.compgeo.2004.05.001.
  14. Habert, G. (2014), "Assessing the environmental impact of conventional and 'green' cement production", Eco-Efficient Construction and Building Materials, Woodhead Publishing.
  15. Jalal, M., Grasley, Z., Nassir, N. and Jalal, H. (2020), "Strength and dynamic elasticity modulus of rubberized concrete designed with ANFIS modeling and ultrasonic technique", Constr. Build. Mater., 240, 117920. https://doi.org/10.1016/j.conbuildmat.2019.117920.
  16. Jang, J.S.R. (1993), "ANFIS: Adaptive-network-based fuzzy inference system", IEEE Trans. Syst. Man Cybern., 23(3), 665-685. https://doi.org/10.1109/21.256541.
  17. Khademi F, Jamal S.M., Deshpande, N. and Londhe, S. (2016), "Predicting strength of recycled aggregate concrete using artificial neural network, adaptive neuro-fuzzy inference system and multiple linear regression", Int. J. Sustain. Built. Environ., 5, 355-369. https://doi.org/10.1016/j.ijsbe.2016.09.003.
  18. Khan, K., Amin, M.N., Usman, M., Imran, M., Al-Faiad M.A. and Shalabi, F.I. (2022), "Effect of fineness and heat treatment on the pozzolanic activity of natural volcanic ash for its utilization as supplementary cementitious materials", Cryst., 12(2), 302. https://doi.org/10.3390/cryst12020302.
  19. Massazza, F. (1993), "Pozzolanic cements", Cement Concrete Compos., 15(4), 185-214. https://doi.org/10.1016/0958-9465(93)90023-3.
  20. Massazza, F. (1998), "Pozzolana and Pozzolanic Cements", Lea's Chemistry of Cement and Concrete, 2nd Edition, Elsevier, UK. 
  21. Massazza, F. (2007), "Pozzolana and pozzolanic cements", Lea's Chemistry of Cement and Concrete, 4th Edition, Elsevier, UK.
  22. MATLAB R2020a (2020), The MathWorks, Inc., Natick, Massachusetts, United States.
  23. Mboya, H.A., King'ondu, C.K., Njau, K.N. and Mrema, A.L. (2017), "Measurement of pozzolanic activity index of scoria pumice, and rice husk ash as potential supplementary cementitious materials for portland cement", Adv. Civil Eng., 2017, 1-13. https://doi.org/10.1155/2017/6952645.
  24. Mehta, P.K. (1981), "Studies on blended Portland cements containing Santorin earth", Cement Concrete Res., 11(4), 507-518. https://doi.org/10.1016/0008-8846(81)90080-6.
  25. Mielenz, R.C., White, L.P. and Glantz O.J. (1950), "Effect of calcination on natural pozzolanas", Symposium on the Use of Pozzolanic Materials in Mortars and Concrete, USA.
  26. Nazari, A. and Torgal, F.P. (2013), "Modeling the compressive strength of geopolymeric binders by gene expression programming-GEP", Expert Syst. Appl., 40(14), 5427-5438. https://doi.org/10.1016/j.eswa.2013.04.014.
  27. Neshat, M., Adeli, A., Masoumi, A. and Sargolzae, M. (2011), "A comparative study on ANFIS and fuzzy expert system mod-els for concrete mix design", Int. J. Comput. Sci., 8(3), 196-210.
  28. Saadat, M. and Bayat, M. (2019), "Prediction of the unconfined compressive strength of stabilised soil by adaptive neuro fuzzy inference system (ANFIS) and non-linear regression (NLR)", Geomech. Geoeng., 17(1), 80-91. https://doi.org/10.1080/17486025.2019.1699668.
  29. Salehi, M., Bayat, M., Saadat, M. and Nasri, M. (2022), "Prediction of unconfined compressive strength and California bearing capacity of cement- or lime-pozzolan-stabilised soil admixed with crushed stone waste", Geomech. Geoeng., 2022, 1-12. https://doi.org/10.1080/17486025.2022.2040606.
  30. Sevim, U.K., Bilgic, H.H., Cansiz, O.F., Ozturk, M. and Atis, C.D. (2021), "Compressive strength prediction models for cementitious composites with fly ash using machine learning techniques", Constr. Build. Mater., 271, 121584, https://doi.org/10.1016/j.conbuildmat.2020.121584.
  31. Tahmasebi, P. and Hezarkhani, A. (2012), "A hybrid neural networks - fuzzy logic-genetic algorithm for grade estimation", Comput. Geosci., 42, 18-27, https://doi.org/10.1016/j.cageo.2012.02.004.
  32. TS 25 (2015), Natural Pozzolan (Trass), Recipes Used in Cement and Concrete, Requirements and Compliance Criteria, Turkish Standards Institute, Turkiye.
  33. TS EN 196-1 (2016), Cement Test Methods, Part 1: Strength Determination, Turkish Standards Institute, Turkiye.
  34. TS EN 197-1 (2012), Cement-Part 1: General Cements Composition, Properties and Compatibility Criteria, Turkish Standards Institute, Turkiye.
  35. Umraoa, R.K., Sharmaa, L.K., Singhb, R. and Singha, T.N. (2018), "Determination of strength and modulus of elasticity of heterogenous sedimentary rocks: An ANFIS predictive technique", Measure., 126, 194-201, https://doi.org/10.1016/j.measurement.2018. 05.064.
  36. Uzal, B., Turanli, L., Yucel, H., Goncuoglu, M.C. and Culfaz, A. (2010), "Pozzolanic activity of clinoptilolite: A comparative study with silica fume, fly ash and a non-zeolitic natural pozzolan", Cement Concrete Res., 40(3), 398-404, https://doi.org/10.1016/j.cemconres.2009.10.016.
  37. Vakhshouri, B. and Nejadi, S. (2018), "Prediction of compressive strength of self-compacting concrete by ANFIS models", Neurocomput., 280, 13-22. https://doi.org/10.1016/j.neucom.2017.09.099.
  38. Walker, R. and Pavia, S. (2011), "Physical properties and reactivity of pozzolans, and their influence on the properties of lime-pozzolan pastes", Mater. Struct., 44, 1139-1150. https://doi.org/10.1617/s11527-010-9689-2.
  39. Yesiloglu-Gultekin, N. and Gokceoglu, C.A., (2022), "Comparison among some non-linear prediction tools on indirect determination of uniaxial compressive strength and modulus of elasticity of basalt", J. Nondestruct. Eval., 41(1), 10. https://doi.org/10.1007/s10921-021-00841-2.
  40. Yu, L., Zhou, S. and Li, L. (2017), "Evaluation of pozzolanic activity of volcanic tuffs from Tibet, China", Adv. Cement Res., 29(4), 137-146. http://doi.org/10.1680/jadcr.15.00075.
  41. Yu, L., Zhou, S. and Ou, H. (2015), "Determining the pozzolanic activity component of volcanic rock", Mater. Res. Innov., 19, 881-888. https://doi.org/10.1179/1432891715Z.0000000001830.