YAPAY SİNİR AĞI VE ÇOKLU REGRESYON YÖNTEMLERİ İLE NH3 YOĞUŞMA ISI TRANSFERİ KATSAYISI TAHMİNİ

Year 2023, , 434 - 444, 31.12.2023

Hakan Aydoğan , Mehmed Rafet Özdemir

https://doi.org/10.55071/ticaretfbd.1383524

Abstract

Soğutma sektöründe kullanılan halokarbon veya sentetik içerikli bileşiklerin çevreye olumsuz etkilerinden dolayı günümüzde kullanımları uluslararası protokollerle sınırlandırılmıştır. Bu sebeple, NH3 gibi düşük küresel ısınma ve ozon tüketme faktörlerine sahip organik bazlı soğutucu akışkanların kullanımı ön plana çıkmıştır. NH3, yüksek ısıl kapasitesi ve düşük viskozite gibi avantajlı termo-fiziksel özelliklerinden dolayı son yıllarda iklimlendirme endüstrisinde tercih edilmektedir. İklimlendirme sistemleri geliştirilirken, belli bağıntılarla sistemin ısı transfer katsayısı ve basınç düşüşü tahmin edilmelidir. Ancak NH3’ün diğer akışkanlardan farklı termo-fiziksel özelliklere sahip olması sebebiyle, literatürde NH3 için yoğuşma rejiminde ısı transfer katsayısını ve basınç düşüşünü yüksek doğruluk oranında veren bir model bulunmamaktadır. Bu çalışmada, yatay ve düz borularda yoğuşma rejiminde bulunan NH3 akışkanı ile yapılmış literatür çalışmalarından veriler alınarak çoklu regresyon ve yapay sinir ağları metotları ile ısı transferi katsayısı tahmini yapılmıştır. Sonuçlar literatürdeki bilgiler ışığında tartışılmış ve öneriler sunulmuştur.

Keywords

Çoklu regresyon, Yapay sinir ağları, Amonyak, Yoğuşma.

PREDICTION OF NH3 CONDENSATION HEAT TRANSFER COEFFICIENT WITH ARTIFICIAL NEURAL NETWORK AND MULTIPLE REGRESSION METHODS

Abstract

The utilization of halocarbon or synthetic-based compounds in the refrigeration sector is limited by international protocols due to their adverse effects on the environment. For this reason, using organic-based refrigerants with low global warming and ozone depletion factors such as NH3 has come to the fore. NH3 has been preferred in the air conditioning industry in recent years due to its advantageous thermo-physical properties, such as high thermal capacity and low viscosity. While developing air conditioning systems, the system's heat transfer coefficient and pressure drop should be predicted with specific correlations. However, since NH3 has different thermo-physical properties from other fluids, no model in the literature gives the heat transfer coefficient and pressure drop in the condensation regime with high accuracy. In this study, heat transfer coefficient prediction was conducted with multiple regression and artificial neural network methods by taking data from literature studies with NH3 fluid in condensation regimes in horizontal and straight pipes. The results were discussed in light of the information in the literature and presented with suggestions.

Keywords

Multiple regression, Artificial neural network, Ammonia, Condensation.

References

• Cavallini, A., Col, D.D., Doretti, L., Matkovic, M., Rossetto, L., Zilio, C. & Censi, G. (2006). Condensation in horizontal smooth tubes: a new heat transfer model for heat exchanger design. Heat Transfer Engineering, 27(8), 31-38.
• Chen, X., Yang, Q., Chi, W., Zhao, Y., Liu, G. & Li, L. (2022). Energy and exergy analysis of NH3/CO2 cascade refrigeration system with subcooling in the low-temperature cycle based on an auxiliary loop of NH3 refrigerants. Energy Reports, 8, 1757-1767.
• Di Filippo, R., Bursi, O. S. & Di Maggio, R. (2022). Global warming and ozone depletion potentials caused by emissions from HFC and CFC banks due to structural damage. Energy and Buildings, 273, 112385.
• Fronk, B. M. & Garimella, S. (2016). Condensation of ammonia and high-temperature-glide zeotropic ammonia/water mixtures in minichannels–Part II: Heat transfer models. International Journal of Heat and Mass Transfer, 101, 1357-1373.
• Haykin, S. (2005). Neural networks: A comprehensive foundation. 2. Baskı, Prentice Hall PTR.
• İnel, M., Eti, S. & Yıldırım, H. (2016). A comparison of artificial neural network and decision tree for profitability in technology sector. International Journal of Development Research, 6(7), 8417-8421.
• James, G., Witten, D., Hastie, T. & Tibshirani, R. (2013). An introduction to statistical learning, Springer, 112, New York.
• Jankovich, D. & Osman, K. (2015). A feasibility analysis of replacing the standard ammonia refrigeration device with the cascade NH3/CO2 refrigeration device in the food industry. Thermal Science, 19(5), 1821-1833.
• Kocak, E., Aylı, E. & Turkoglu, H. (2022). A comparative study of multiple regression and machine learning techniques for prediction of nanofluid heat transfer. Journal of Thermal Science and Engineering Applications, 14(6), 061002.
• Komandiwirya, H. B., Hrnjak, P. S. & Newell, T. A. (2005). An experimental investigation of pressure drop and heat transfer in an in-tube condensation system of ammonia with and without miscible oil in smooth and enhanced tubes. Air Conditioning and Refrigeration Center. College of Engineering. University of Illinois at Urbana-Champaign.
• Li, W., Zheng, B., Lv, T. & Ayub, Z. (2020). A modified correlation for flow boiling heat transfer in plate heat exchangers. Journal of Thermal Science and Engineering Applications, 12(6), 6-14.
• Maggiora, G. M., David, W. E. & Robert, G. T. (1992). Computational neural networks as model-free mapping devices. Journal of chemical information and computer sciences, 32(6), 732-741.
• Nie, F., Wang, H., Zhao, Y., Song, Q., Yan, S. & Gong, M. (2023). A universal correlation for flow condensation heat transfer in horizontal tubes based on machine learning. International Journal of Thermal Sciences, 184, 107994.
• Özdemir, M.R. (2016). Single-phase flow and flow boiling of water in rectangular metallic microchannels [Doktora tezi]. Brunel University London, Londra.
• Park, C. Y. & Hrnjak, P. (2008). NH3 in-tube condensation heat transfer and pressure drop in a smooth tube. International Journal of Refrigeration, 31(4), 643-651.
• Pearson, A. (2008). Refrigeration with ammonia. International Journal of Refrigeration, 31(4), 545-551.
• Qiu, Y., Garg, D., Zhou, L., Kharangate, C. R., Kim, S. M. & Mudawar, I. (2020). An artificial neural network model to predict mini/micro-channels saturated flow boiling heat transfer coefficient based on universal consolidated data. International Journal of Heat and Mass Transfer, 149, 119211.
• Svozil, D., Kvasnicka, V. & Pospichal, J. (1997). Introduction to multi-layer feed-forward neural networks. Chemometrics and intelligent laboratory systems, 39(1), 43-62.
• Tao, X. & Ferreira, C. A. I. (2020). NH3 condensation in a plate heat exchanger: Flow pattern based models of heat transfer and frictional pressure drop. International Journal of Heat and Mass Transfer, 154, 119774.
• Vollrath, J. E., Hrnjak, P. S. & Newell, T. A. (2003). An experimental investigation of pressure drop and heat transfer in an in-tube condensation system of pure ammonia. Air Conditioning and Refrigeration Center. College of Engineering. University of Illinois at Urbana-Champaign.
• Zhang, R., Liu, J. & Zhang, L. (2021). Boiling heat transfer and visualization for R717 in a horizontal smooth mini-tube. International Journal of Refrigeration, 131, 275-285.
• Zhang, J., Elmegaard, B. & Haglind, F. (2021). Condensation heat transfer and pressure drop correlations in plate heat exchangers for heat pump and organic Rankine cycle systems. Applied Thermal Engineering, 183, 116231.