HAD Çözümlerini Kullanarak Endüstriyel Cam Ergitme Fırınlarının İşletme Analizi için Makine Öğrenimi Tabanlı İndirgenmiş Model Geliştirilmesi

Öz

Endüstriyel cam ergitme fırınları için enerji tüketimi, cam kalitesi, sıcaklık dağılımı ve refrakter aşınması hakkında bilgiler sunan hesaplamalı akışkanlar dinamiği (HAD) modelleri, fırınların tasarımında kritik öneme sahiptir. Ancak, cam ergitme prosesinde geniş bir aralığa sahip olan zaman ve uzunluk boyutlarına bağlı olarak ortaya çıkan yüksek hesaplama maliyeti, bu modellerin fırınların günlük operasyonlarında doğrudan kullanılmasını engellemektedir. Bu zorluğu aşabilmek için bu çalışma, bir cam ergitme fırınının cam banyosu HAD modelinden elde edilen parametrik sonuçları kullanarak, makine öğrenimi tabanlı bir indirgenmiş model geliştirilmesi yoluyla yeni bir yaklaşım sunmaktadır. HAD modelinin de sınır koşullarına etki ederek çözüm sonucunu değiştiren çekiş hızı (cam debisi), doğal gaz kaynaklı ısı akısı ve elektrik potansiyel fark, işletme parametreleri olarak seçilerek HAD çözümü bir veri seti oluşturulmuştur. HAD çıktısı sıcaklık ve hız alanı verilerini öğrenmek ve tahmin etmek için konvolüsyonel nöral ağları içeren bir otokodlayıcı yapı oluşturulmuştur. Sonrasında, otokodlayıcının boyutsal yükseltme yapan bölümü, ek bir tam bağlantılı nöral ağ aracılığıyla, işletme parametreleri olan sınır koşullarıyla ilişkilendirilmiştir. İndirgenmiş modelin performansı, ek HAD çözümleri kullanılarak hem interpolasyon hem de ekstrapolasyon için değerlendirilmiştir. İndirgenmiş model tarafından üretilen sonuçlar ile gerçek HAD çözümleri arasındaki karşılaştırma, %1'den az sapma olduğunu göstermekte ve indirgenmiş modelin günlük fırın operasyonları için etkili bir analiz aracı olarak hizmet etme kabiliyetini kanıtlamaktadır. Ayrıca, indirgenmiş model çözüm süresinde de üçüncü mertebeye kadar hızlanma sağlamaktadır. Bu hızlanma fırın işletmesi açıdan geliştirilen yaklaşımın oldukça fayda sağlayacağını göstermektedir.

Anahtar Kelimeler

• Cam Ergitme Fırınları
• Hesaplamalı Akışkanlar Dinamiği
• Makine Öğrenimi
• İndirgenmiş Model
• Otokodlayıcı

Machine-Learning-Based Reduced Order Modeling for Operational Analysis of Industrial Glass Melting Furnaces Using CFD Solutions

Öz

Computational Fluid Dynamics (CFD) models play a vital role in the design of industrial glass melting furnaces, offering insights into energy consumption, glass quality, temperature distribution, and refractory wear. However, the considerable computational expense associated with the large time and length scales involved in the glass melting process prevents practical utilization of those models in daily operation of the furnaces. This study presents a novel approach to address this challenge through the development of a machine-learning-based Reduced-Order Model (ROM) utilizing parametric data obtained from a CFD model of a glass melting tank of a furnace. Key operational parameters, namely pull rate, heat flux from combustion space, and electrical potential difference to supply electrical power, are chosen to create a CFD solution dataset, as they change the boundary conditions of the CFD model and, consequently, the field solution data. An autoencoder structure incorporating convolutional neural networks is established to learn and predict temperature and velocity field data. Then, the decoder section of the autoencoder is connected to the operational parameters through an auxiliary neural network. The performance of the reduced-order model is assessed for both interpolation and extrapolation using additional CFD solutions. Comparison between the field data generated by the ROM and the ground-truth CFD solutions indicates less than 1\% deviation, proving that the ROM’s capability to serve as an effective analysis tool for daily furnace operation. Furthermore, the ROM demonstrates significant advancements in solution time, up to third order, further enhancing its practical utility.

Anahtar Kelimeler

• Glass Melting Furnaces
• Computational Fluid Dynamics
• Machine Learning
• Reduced Order Model
• Autoencoders

Kaynakça

1. Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Corrado, G. S., Davis, A., Dean, J., Devin, M., Ghemawat, S., Goodfellow, I., Harp, A., Irving, G., Isard, M., Jia, Y., Jozefowicz, R., Kaiser, L., Kudlur, M., … Zheng, X. (2015). TensorFlow: Large-Scale Machine Learning on Heterogeneous Systems, Version 1.15. https://www.tensorflow.org/
2. Abbassi, A., & Khoshmanesh, K. (2008). Numerical simulation and experimental analysis of an industrial glass melting furnace. Applied Thermal Engineering, 28(5–6), 450–459. https://doi.org/10.1016/j.applthermaleng.2007.05.011
3. Abooali, D., & Khamehchi, E. (2019). New predictive method for estimation of natural gas hydrate formation temperature using genetic programming. Neural Computing and Applications, 31(7), 2485–2494. https://doi.org/10.1007/s00521-017-3208-0
4. ANSYS Inc. (2022). Ansys Fluent User’s Guide. https://www.ansys.com
5. Atzori, D., Tiozzo, S., Vellini, M., Gambini, M., & Mazzoni, S. (2023). Industrial Technologies for CO2 Reduction Applicable to Glass Furnaces. Thermo, 3(4), 682–710. https://doi.org/10.3390/thermo3040039
6. Bhatnagar, S., Afshar, Y., Pan, S., Duraisamy, K., & Kaushik, S. (2019). Prediction of aerodynamic flow fields using convolutional neural networks. Computational Mechanics, 64(2), 525–545. https://doi.org/10.1007/s00466-019-01740-0
7. Brunton, S. L., Noack, B. R., & Koumoutsakos, P. (2020). Machine learning for fluid mechanics. Annual Review of Fluid Mechanics, 52(1), 477–508. https://doi.org/10.48550/arXiv.1905.11075
8. Cassar, D. R., de Carvalho, A. C. P. L. F., & Zanotto, E. D. (2018). Predicting glass transition temperatures using neural networks. Acta Materialia, 159, 249–256. https://doi.org/10.1016/j.actamat.2018.08.022

9. Castillo, V., & Kornish, B. (2017). Development of Reduced Glass Furnace Model to Optimize Process Operation, Final Report CRADA No. TC02241.
10. Catsoulis, S., Singh, J.-S., Narayanan, C., & Lakehal, D. (2022). Integrating supervised learning and applied computational multi-fluid dynamics. International Journal of Multiphase Flow, 157, 104221. https://doi.org/10.1016/j.ijmultiphaseflow.2022.104221
11. Celik, I. B., Ghia, U., Roache, P. J., Freitas, C. J., Coleman, H., & Raad, P. E. (2008). Procedure for estimation and reporting of uncertainty due to discretization in CFD applications. Journal of Fluids Engineering, Transactions of the ASME, 130(7), 0780011–0780014. https://doi.org/10.1115/1.2960953
12. Cho, G., Wang, M., Kim, Y., Kwon, J., & Su, W. (2022). A physics-informed machine learning approach for estimating lithium-ion battery temperature. IEEE Access, 10, 88117–88126. https://doi.org/10.1109/ACCESS.2022.3199652
13. Chollet, F. (2021). Deep learning with Python. Simon and Schuster.
14. Choudhary, M. (1985). Three-dimensional Mathematical Model for Flow and Heat Transfer in Electric Glass Furnaces. Heat Transfer Engineering, 6(4), 55–65. https://doi.org/10.1080/01457638508939639
15. Choudhary, M. K., & Potter, R. M. (2005). Heat transfer in glass-forming melts. In L. D. Pye, A. Montenero, & I. Joseph (Eds.), Properties of glass-forming melts (pp. 249–294). Taylor & Francis Publications. https://doi.org/10.1201/9781420027310
16. Choudhary, M. K., Purnode, B., Lankhorst, A. M., & Habraken, A. F. J. A. (2018). Radiative heat transfer in processing of glass-forming melts. International Journal of Applied Glass Science, 9(2), 218–234. https://doi.org/10.1111/ijag.12286
17. Choudhary, M. K., Venuturumilli, R., & Hyre, M. R. (2010). Mathematical modeling of flow and heat transfer phenomena in glass melting, delivery, and forming processes. International Journal of Applied Glass Science, 1(2), 188–214. https://doi.org/10.1111/j.2041-1294.2010.00011.x
18. Cravero, C., & Marsano, D. (2023). Numerical Simulation of Melted Glass Flow Structures inside a Glass Furnace with Different Heat Release Profiles from Combustion. Energies, 16(10), 4187. https://doi.org/10.3390/en16104187
19. Daurer, G., Raič, J., Demuth, M., Gaber, C., & Hochenauer, C. (2022). Comprehensive and numerically efficient CFD model for bubbling in an industrial glass tank. Chemical Engineering Research and Design, 186, 82–96. https://doi.org/10.1016/j.cherd.2022.07.044
20. Erichson, N. B., Mathelin, L., Yao, Z., Brunton, S. L., Mahoney, M. W., & Kutz, J. N. (2019). Shallow learning for fluid flow reconstruction with limited sensors and limited data. ArXiv Preprint ArXiv:1902.07358. https://doi.org/10.1098/rspa.2020.0097
21. Faber, A.-J., Rongen, M., Lankhorst, A., & Meneses, D. D. S. (2020). Characterization of high temperature optical spectra of glass melts and modeling of thermal radiation conductivity. International Journal of Applied Glass Science, 11(3), 442–462. https://doi.org/10.1111/ijag.15116
22. Gao, H., Sun, L., & Wang, J.-X. (2021). PhyGeoNet: Physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain. Journal of Computational Physics, 428, 110079. https://doi.org/10.1016/j.jcp.2020.110079
23. Han, J., Li, L., Wang, J., Chen, S., Liu, C., & Li, C. (2022). Simulation and evaluation of float glass furnace with different electrode positions. Journal of the American Ceramic Society, 105(12), 7097–7110. https://doi.org/10.1111/jace.18700
24. He, X.-J., Yu, C.-H., Zhao, Q., Peng, J.-Z., Chen, Z.-H., & Hua, Y. (2022). Reduced order modelling of natural convection of nanofluids in horizontal annular pipes based on deep learning. International Communications in Heat and Mass Transfer, 138, 106361. https://doi.org/10.1016/j.icheatmasstransfer.2022.106361
25. Jebava, M., & Němec, L. (2018). Role of glass melt flow in container furnace examined by mathematical modelling. Ceram-Silik, 62, 86–96. https://doi.org/10.13168/cs.2017.0049
26. Kim, B., Azevedo, V. C., Thuerey, N., Kim, T., Gross, M., & Solenthaler, B. (2019). Deep fluids: A generative network for parameterized fluid simulations. Computer Graphics Forum, 38(2), 59–70. https://doi.org/10.1111/cgf.13619
27. Lankhorst, A. M., Thielen, L., Simons, P. J. P. M., & Habraken, A. F. J. A. (2013). Proper modeling of radiative heat transfers in clear glass melts. 73rd Conference on Glass Problems, 249–258. https://doi.org/10.1002/9781118710838.ch19
28. Li, L., Han, J., Lin, H.-J., Ruan, J., Wang, J., & Zhao, X. (2020). Simulation of glass furnace with increased production by increasing fuel supply and introducing electric boosting. International Journal of Applied Glass Science, 11(1), 170–184. https://doi.org/10.1111/ijag.13907
29. Masoumi-Verki, S., Haghighat, F., & Eicker, U. (2022). A review of advances towards efficient reduced-order models (ROM) for predicting urban airflow and pollutant dispersion. Building and Environment, 216, 108966. https://doi.org/10.1016/j.buildenv.2022.108966
30. Matsuno, S., Iso, Y., Uchida, H., Oono, I., Fukui, T., & Ooba, T. (2008). CFD modeling coupled with electric field analysis for joule-heated glass melters. Journal of Power and Energy Systems, 2(1), 447–455. https://doi.org/10.1299/jpes.2.447
31. Molinaro, R., Singh, J.-S., Catsoulis, S., Narayanan, C., & Lakehal, D. (2021). Embedding data analytics and CFD into the digital twin concept. Computers & Fluids, 214, 104759. https://doi.org/10.1016/j.compfluid.2020.104759
32. Mücke, N. T., Bohté, S. M., & Oosterlee, C. W. (2021). Reduced order modeling for parameterized time-dependent PDEs using spatially and memory aware deep learning. Journal of Computational Science, 53, 101408. https://doi.org/10.1016/j.jocs.2021.101408
33. Paszke, A., Gross, S., Chintala, S., Chanan, G., Yang, E., DeVito, Z., Lin, Z., Desmaison, A., Antiga, L., & Lerer, A. (2017). Automatic differentiation in pytorch.
34. Pazarlioglu, H. K., Tepe, A. Ü., & Arslan, K. (2022). Optimization of Parameters Affecting Anti-Icing Performance on Wing Leading Edge of Aircraft. European Journal of Science and Technology. https://doi.org/10.31590/ejosat.1062495
35. Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., & Duchesnay, É. (2011). Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research, 12(85), 2825–2830. http://jmlr.org/papers/v12/pedregosa11a.html
36. Pfaff, T., Fortunato, M., Sanchez-Gonzalez, A., & Battaglia, P. W. (2020). Learning mesh-based simulation with graph networks. ArXiv Preprint ArXiv:2010.03409.
37. Pigeonneau, F., & Flesselles, J.-M. (2012). Practical laws for natural convection of viscous fluids heated from above in a shallow cavity. International Journal of Heat and Mass Transfer, 55(1–3), 436–442. https://doi.org/10.1016/j.ijheatmasstransfer.2011.09.057
38. Pigeonneau, F., Pereira, L., & Laplace, A. (2023). Dynamics of rising bubble population undergoing mass transfer and coalescence in highly viscous liquid. Chemical Engineering Journal, 455, 140920.
39. Pokorny, R., Hrma, P., Lee, S., Klouzek, J., Choudhary, M. K., & Kruger, A. A. (2020). Modeling batch melting: Roles of heat transfer and reaction kinetics. Journal of the American Ceramic Society, 103(2), 701–718. https://doi.org/10.1111/jace.16820
40. Pokorny, R., Pierce, D. A., & Hrma, P. (2012). Melting of glass batch: model for multiple overlapping gas-evolving reactions. Thermochimica Acta, 541, 8–14. https://doi.org/10.1016/j.tca.2012.04.020
41. Raič, J., Gaber, C., Wachter, P., Demuth, M., Gerhardter, H., Knoll, M., Prieler, R., & Hochenauer, C. (2021). Validation of a coupled 3D CFD simulation model for an oxy-fuel cross-fired glass melting furnace with electric boosting. Applied Thermal Engineering, 195, 117166. https://doi.org/10.1016/j.applthermaleng.2021.117166
42. Rowley, C. W., & Dawson, S. T. M. (2017). Model reduction for flow analysis and control. Annual Review of Fluid Mechanics, 49(1), 387–417. https://doi.org/10.1146/annurev-fluid-010816-060042
43. Schill, P., & Chmelar, J. (2004). Use of computer flow dynamics in glass technology. Journal of Non-Crystalline Solids, 345, 771–776. https://doi.org/10.1016/j.jnoncrysol.2004.08.199
44. Simcik, M., & Ruzicka, M. C. (2015). CFD model for pneumatic mixing with bubble chains: Application to glass melts. Chemical Engineering Science, 127, 344–361. https://doi.org/10.1016/j.ces.2015.01.052
45. Simons, P., Jochem, K., & Aiuchi, K. (2008). A power consistent mathematical formulation for Joulean heat release. Glass Technology-European Journal of Glass Science and Technology Part A, 49(3), 109–118.
46. Soubeih, S., Luedtke, U., & Halbedel, B. (2015). Improving residence time distribution in glass melting tanks using additionally generated Lorentz forces. J. Chem. Chem. Eng, 9, 203–210. https://doi.org/10.17265/1934-7375/2015.03.006
47. Staněk, J. (1990). Problems in electric melting of glass. Journal of Non-Crystalline Solids, 123(1), 400–414. https://doi.org/10.1016/0022-3093(90)90812-Z
48. Taşkesen, E., Dirik, M., Tekir, M., & Pazarlioğlu, H. K. (2023). Predicting heat transfer performance of Fe3O4-Cu/water hybrid nanofluid under constant magnetic field using ANN. Journal of Thermal Engineering, 9(3), 811–822. https://doi.org/10.18186/thermal.000000
49. Thuerey, N., Weißenow, K., Prantl, L., & Hu, X. (2020). Deep learning methods for Reynolds-averaged Navier–Stokes simulations of airfoil flows. AIAA Journal, 58(1), 25–36. https://doi.org/10.2514/1.J058291
50. Viskanta, R. (1994). Review of three-dimensional mathematical modeling of glass melting. Journal of Non-Crystalline Solids (Vol. 177). https://doi.org/10.1016/0022-3093(94)90549-5
51. Wiewel, S., Becher, M., & Thuerey, N. (2019). Latent space physics: Towards learning the temporal evolution of fluid flow. Computer Graphics Forum, 38(2), 71–82. https://doi.org/10.48550/arXiv.1802.10123
52. Zhang, R., Liu, Y., & Sun, H. (2020). Physics-guided convolutional neural network (PhyCNN) for data-driven seismic response modeling. Engineering Structures, 215, 110704. https://doi.org/10.1016/j.engstruct.2020.110704
53. Zier, M., Stenzel, P., Kotzur, L., & Stolten, D. (2021). A review of decarbonization options for the glass industry. Energy Conversion and Management: X, 10(February), 100083. https://doi.org/10.1016/j.ecmx.2021.100083