Integrating Design Optimization and AI-Driven Maintenance in Centrifugal Pumps for Sustainable Industrial Operations

Abstract

Centrifugal pumps are crucial in a variety of industrial sectors, but their performance is often affected by energy inefficiencies, wear-related failures, and suboptimal hydraulic characteristics. This research presents a comprehensive experimental study aimed at optimizing the hydraulic performance, energy consumption, and reliability of centrifugal pumps through design innovations, advanced manufacturing techniques, and predictive maintenance. Computational Fluid Dynamics (CFD) simulations, additive manufacturing, and Long Short-Term Memory (LSTM)-based predictive algorithms were employed. The findings revealed up to a 7.3% increase in hydraulic efficiency, a 75% reduction in failure incidents through predictive maintenance, and 8.2% savings in energy usage. The proposed framework offers a scalable, data-driven approach to improve centrifugal pump operations in modern industrial applications.

Introduction

Centrifugal pumps are crucial in various industries but often face efficiency losses due to hydraulic design flaws, cavitation, and maintenance issues. This study integrates computational modeling (CFD), experimental validation, additive manufacturing, and machine learning-based predictive maintenance to optimize pump performance.

Key points include:

• Performance Parameters: Efficiency, flow rate, head, NPSH, and power consumption are critical metrics, with CFD used to analyze internal fluid dynamics and energy losses.

• Design Innovations: Advanced impeller designs (curved blades, biomimicry-inspired shapes) and optimized volute casings reduce losses and improve pressure distribution.

• Materials and Manufacturing: New materials like carbon fiber composites and 3D printing (DMLS) enable complex, precise, and durable pump components.

• Predictive Maintenance: Machine learning models, especially LSTM networks, analyze sensor data to predict failures early, significantly reducing downtime.

• IoT and Digital Twins: Real-time monitoring via IoT and digital twins allows better operational control and maintenance planning.

• Energy Efficiency: Use of variable speed drives and optimized designs reduce energy consumption by 8–30%, supporting sustainability goals.

• Cavitation Mitigation: Improved designs and anti-cavitation features reduce damage, noise, and performance loss.

• Experimental and Simulation Results: Optimized impeller designs showed improved hydraulic efficiency (74.7% vs. 67.4%), higher pressure head, reduced power consumption, better cavitation resistance, and extended pump life through predictive maintenance.

Despite advancements, the integration of these approaches into unified frameworks and long-term industrial validations remains limited, highlighting areas for future research.

Conclusion

This study demonstrated how design optimization, advanced manufacturing, and predictive analytics can significantly improve centrifugal pump performance. Key outcomes included: • A 7.3% increase in hydraulic efficiency. • Reduced cavitation and noise levels. • 75% fewer failures with ML-driven predictive maintenance. • 8.2% reduction in energy consumption. The integration of simulation, fabrication, and AI models creates a holistic solution for sustainable and cost-effective pump management in industrial environments.

References

[1] Wang, L. et al., “Performance evaluation of centrifugal pumps using CFD analysis,” Journal of Fluids Engineering, vol. 140, no. 3, pp. 1–8, 2018. [2] Zhang, H., & Li, Y., “Flow simulation and performance prediction of a mixed-flow pump,” Energy Conversion and Management, vol. 168, pp. 340–348, 2018. [3] Bhosale, A. et al., “CFD analysis of volute casing to improve hydraulic efficiency,” Engineering Reports, vol. 2, no. 3, e12064, 2020. [4] Hosseinipour, S. J., “A study on impeller design for centrifugal pumps,” Mechanical Systems and Signal Processing, vol. 136, pp. 106547, 2020. [5] Xu, Z. et al., “Design optimization of impellers using AI algorithms,” Journal of Mechanical Design, vol. 142, pp. 110501, 2020. [6] Kim, M. & Park, J., “Bio-inspired blade designs for improved flow characteristics,” Applied Sciences, vol. 11, no. 1, pp. 115, 2021. [7] Al-Bahadly, I. H., “Volute design influence on pump performance,” Journal of Energy Engineering, vol. 147, no. 4, pp. 04021045, 2021. [8] Chen, L. et al., “Advanced composite materials in pump manufacturing,” Composites Part B: Engineering, vol. 179, pp. 107514, 2019. [9] Boparai, K. S., et al., “Application of additive manufacturing in pump component fabrication,” Materials Today: Proceedings, vol. 45, pp. 3611–3617, 2021. [10] Arunkumar, N. et al., “Additive manufacturing for centrifugal pump impellers,” Materials Research Express, vol. 7, no. 11, 2020. [11] Dey, R. et al., “Predictive maintenance for rotating machines,” Journal of Intelligent Manufacturing, vol. 31, pp. 1065–1083, 2019. [12] Jain, A. & Singh, D., “Machine learning in condition monitoring,” IEEE Access, vol. 8, pp. 145695–145706, 2020. [13] Hassan, M. et al., “LSTM networks for anomaly detection in pumps,” Sensors, vol. 21, no. 5, pp. 1654, 2021. [14] Kumar, A., “Smart maintenance models using IoT and ML,” Procedia Computer Science, vol. 170, pp. 365–372, 2020. [15] Zhao, L., “IoT-based condition monitoring systems,” Automation in Construction, vol. 96, pp. 1–10, 2018. [16] Wang, J. et al., “Digital twin-driven smart pump maintenance,” Journal of Cleaner Production, vol. 237, pp. 117755, 2019. [17] Bureau of Energy Efficiency (BEE), “Energy efficiency in pump systems,” India, 2020. [18] Madhu, R. et al., “Impact of VFD on pump energy consumption,” Renewable Energy, vol. 146, pp. 2035–2043, 2020. [19] Shukla, A. & Varma, R., “Energy optimization in centrifugal pumps,” Energy Reports, vol. 6, pp. 1113–1120, 2020. [20] Li, H. et al., “Numerical investigation of cavitation in pumps,” Journal of Mechanical Science and Technology, vol. 35, no. 6, pp. 2413–2422, 2021. [21] Yao, S., “Anti-cavitation strategies in centrifugal pumps,” International Journal of Fluid Machinery and Systems, vol. 12, no. 4, pp. 329–338, 2019. [22] Feng, Y. et al., “Bubble dynamics and suppression techniques,” Experimental Thermal and Fluid Science, vol. 107, pp. 107569, 2019. [23] S. M. Al-Shammaa, et al., “Experimental analysis of a closed-loop centrifugal pump system under variable speed conditions,” Applied Thermal Engineering, vol. 145, pp. 697–706, 2018. https://doi.org/10.1016/j.applthermaleng.2018.10.075 [24] M. Elbatran, M. Ismail, and A. H. Hassan, “Performance enhancement of centrifugal pumps using computational fluid dynamics,” Journal of Fluids Engineering, vol. 141, no. 3, pp. 031102, 2019.https://doi.org/10.1115/1.4041730 [25] S. J. Yoon, D. Lee, and J. H. Park, “Optimization of impeller blade geometry for centrifugal pumps using CFD simulations,” International Journal of Heat and Fluid Flow, vol. 78, pp. 108450, 2019.https://doi.org/10.1016/j.ijheatfluidflow.2019.108450 [26] H. G. Kim and C. K. Lee, “CFD-based design improvement of industrial pump impellers for energy efficiency,” Mechanical Systems and Signal Processing, vol. 135, pp. 106378, 2020. https://doi.org/10.1016/j.ymssp.2019.106378 [27] T. R. Faramarzi, M. Zolfaghari, and M. Jafari, “Additive manufacturing of pump components using DMLS: A case study on impellers,” Additive Manufacturing, vol. 36, pp. 101500, 2020. https://doi.org/10.1016/j.addma.2020.101500 [28] M. T. Islam, S. K. Saha, and M. M. Rahman, “Experimental validation of 3D printed stainless steel centrifugal impellers,” Materials Today: Proceedings, vol. 47, pp. 1414–1420, 2021. https://doi.org/10.1016/j.matpr.2021.02.214 [29] Roy, S. Gupta, and P. Nath, “Predictive maintenance in pumps using deep learning LSTM networks,” Journal of Intelligent Manufacturing, vol. 33, pp. 1951–1965, 2022. https://doi.org/10.1007/s10845-022-01962-3 [30] N. Patel, H. Zhou, and Y. Zhang, “Implementation of LSTM and IoT for centrifugal pump fault prediction,” IEEE Access, vol. 10, pp. 54321–54330, 2022. https://doi.org/10.1109/ACCESS.2022.3178675

Copyright

Copyright © 2025 Kavar Niravkumar Savjibhai, Dr. Khubilal Khatri, Dr. Vijay Kumar Gadhavi, Patel Sanketkumar Ketanbhai, Vaghela Suraj Kumar Khodidas . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.