FROM 40 HOURS TO 15 MINUTES, 15 YEARS OF EXPERIENCE WITH CFD

Transcription

1 FROM 40 HOURS TO 15 MINUTES, 15 YEARS OF EXPERIENCE WITH CFD NORBERT BULTEN WARTSILA PROPULSION THE NETHERLANDS 1 / 15 Wärtsilä 23 March 2011 Doc.ID: Revision: Status:

2 Topics Introduction Wärtsilä Propulsion Products History of CFD within Wärtsilä First wave: Implementation of CFD method Get the CFD tool at similar level as existing research methods Main focus: validation Second wave: Utilizing CFD method stand-alone Use CFD to understand the physical phenomena Main focus: interpretation Summary and outlook 2 / 15 Wärtsilä 23 March 2011

3 Introduction Wärtsilä Propulsion Wärtsilä is one of the key suppliers in the world of marine Diesel engines and propulsion systems. Within Wärtsilä Propulsion a wide range of propulsion products is being developed, manufactured and maintained. Examples are: Controllable and Fixed Pitch propellers Azimuthing thrusters Tunnel thrusters Waterjets 3 / 15 Wärtsilä 23 March 2011

4 Introduction Wärtsilä Propulsion Propulsion products on wide range of vessels 4 / 15 Wärtsilä 23 March 2011

5 Introduction: history of CFD From Star-CD version 2.1 in 1996 to Star-CCM in 2011 Scope of applications has grown extremely Target is Full-scale numerical towing tank 2011: virtual towing tank 2000: CFD propeller analysis 2005: CFD thruster analysis 2008: Propellerhull interaction 1996: Star-CD at Lips -waterjet department 5 / 15 Wärtsilä 23 March 2011

6 Introduction: history of CFD 1999 Star user conference presentation on automated meshing tool for waterjet inlets 2002 Star user conference presentation on applications of waterjet inlet design tool. Calculation time 40 hours 2011 calculation time 15 minutes 6 / 15 Wärtsilä 23 March 2011

7 First wave: Implementation of CFD Key factors for successful implementation Quick pre-processing Robust geometry handling Automated meshing process Standardized boundary conditions Commercially acceptable calculation times Well-defined qualitative post-processing Standardized output format of results Method validated with aid of experimental results 7 / 15 Wärtsilä 23 March 2011

8 First wave: Implementation of CFD Second application: propellers Automated meshing tool for open and ducted propellers Validation carried out for numerous propellers Streamlined processes to ensure Quality Assurance 8 / 15 Wärtsilä 23 March 2011

9 The results of implementation in daily work Automated process should give user-independent results and should be easy in use 9 / 15 Wärtsilä 23 March 2011

10 Second wave: Utilization of CFD Creation of large data set Geometry variation Complete range of operating conditions Understanding physical phenomena Utilizing post-processing options Qualitative review: contour and vector plots Quantitative review: integrals and force summation Scaling effects on ducted propellers Validation of results Development of alternative post-processing methodology Testing of concept Interpretation of results on model scale and full scale 10 / 15 Wärtsilä 23 March 2011

11 Performance prediction of ducted propellers In marine industry a debate is ongoing about the existence of Reynolds scaling effects on the performance of ducted propellers. Comparison CFD with model scale experiments Good match How to explain these results? Comparison CFD model scale and full scale Clear difference 11 / 15 Wärtsilä 23 March 2011

12 Alternative post-processing New post-processing methodology is based on viscous losses and axial pump efficiency p in v in v out v 1 v 2 p out p 1 v p p 2 in pump = rotor + rotation out,all 12 / 15 Wärtsilä 23 March 2011

13 Verification of concept Ducted propeller performance is split into different components: Open water performance Pump efficiency (Viscous) inlet loss Exit loss Rotational loss Total efficiency adds up to 100% 13 / 15 Wärtsilä 23 March 2011

14 Comparison of results on model and full scale Ducted propeller performance for different nozzles (HR & 19A) Pump efficiency (Viscous) losses Clear trends found in pump efficiency and viscous losses. Results are in line with expectations and literature Physical explanation of scaling phenomenon has been revealed. 14 / 15 Wärtsilä 23 March 2011

15 Summary First wave: from Colors For Directors to CFD Get the CFD tool at similar level as existing research methods Main focus: validation Second wave: from CFD to Confident Foundation for Design Use CFD to understand the physical phenomena Main focus: interpretation 15 / 15 Wärtsilä 23 March 2011

16 Outlook Third wave: towards Computer-driven F(v)ariations in Design Design based on optimization methods Main focus: product performance And this might bring the calculation time back to 40 hours 16 / 15 Wärtsilä 23 March 2011

17 Questions? 17 / 15 Wärtsilä 23 March 2011