ISimT-26-Web-banner-Banner-v1

Catch Our Technical Presentation on Tuesday, 21 April 2026 at 17:30

Preliminary Sizing and Inverse Design Optimization of High Temperature Heat Pump Centrifugal Compressors through Integrated CFD and Machine Learning

Will you be attending the ISimT-26 Symposium in Germany? We would love to connect with you to discuss how our turbomachinery solutions can help your organization eliminate time-consuming design iterations.

Please do not hesitate to reach out and let us know your availability so we can reserve a dedicated time slot to meet.

We are also excited to announce that Prof. Mehrdad Zangeneh will be presenting our latest paper, "Preliminary Sizing and Inverse Design Optimization of High Temperature Heat Pump Centrifugal Compressors through Integrated CFD and Machine Learning," on Tuesday, 21st April at 17:30.

The paper presents an enhanced methodology for the design of centrifugal compressors used in high-temperature heat pump (HTHP) applications.

Building on previous inverse design work, an expanded cycle analysis is introduced to enhance flexibility in early-stage sizing, including RPM optimization and off-design performance evaluation. The workflow couples heat pump cycle modelling with meanline analysis to guide refrigerant selection and initial impeller sizing, ensuring that thermodynamic requirements for real gases are well integrated into the aerodynamic design process, by using REFPROP look up tables. By considering three different refrigerants (R1233zd, R1224yd and R601), a baseline impeller for the most promising refrigerant is then generated by using a 3D inverse design method.

A novel machine learning approach is applied, combining TURBOdesign’s key design parameters with high-fidelity CFD simulations to optimize aerodynamic performance and flow characteristics at multiple operating points. Results demonstrate how machine learning enhanced inverse design optimization can contribute to efficient and robust compressor designs, adaptable to varying refrigerants and operating conditions. This ensures high COP for the high temperature heat pump over its most important operating conditions.

Authors: Prof. Mehrdad Zangeneh, Oskar Freytag, Eric Hsieh and Mihai Bleiziffer.

 Complete the form to:

✅ Schedule a Meeting:
Secure a time to speak with Prof. Zangeneh or Markus Gfrörer at the event.

✅ Get the Research:
Receive a digital copy of the paper after the conference concludes.
 
✅ See the Software:
Learn how TURBOdesign Suite generates breakthrough designs without the manual guesswork.
  

 

We look forward to meeting you! Please complete the above form.

Pareto front obtained from Reactive Response Surface model using small data sets which integrates 3D inverse design
Pareto front obtained from Reactive Response Surface model using small data sets which integrates 3D inverse design method in TURBOdesign1 with ANSYS CFX to evaluate performance parameters of the compressor at various operating points 
Exploration design space at cycle / meanline level
Legend: Exploration design space at cycle / meanline level to impact of different parameters such as RPM, Tip diameter, exit width, axial length on heat pump COP
Comparison of improvement in COP achieved by the new cycle/meanline design exploration system for heat pumps
 Comparison of improvement in COP achieved by the new cycle/meanline design exploration system for heat pumps versus previous manual design for different working substances 
Comparison of 3D geometry of baseline versus optimized impeller
 Comparison of 3D geometry of baseline versus optimized impeller

 

 We look forward to meeting you! Please complete the form above.