Description and validation of the rotordisksource class for propeller performance estimation

Transcription

1 Description and validation of the rotordisksource class for propeller performance estimation Alexandre Capitao Patrao Department of Mechanics and Maritime Sciences Division of Fluid Dynamics Chalmers University of Technology Gothenburg, Sweden Alexandre Capitao Patrao / 48

2 Contents Contents Introduction Theory Implementation Modifications Case setup Results Epilogue Alexandre Capitao Patrao / 48

3 Background Background Propeller performance prediction methods range in order of complexity and fidelity from analytical 1D momentum theory, BEM, to 3D CFD. The actuator disk methodology lies between the BEM and 3D CFD in terms of complexity, fidelity, and computational cost. The low computational cost of the actuator disk methodology makes is suitable for studying rotor-airframe interaction. Alexandre Capitao Patrao / 48

4 Background Alexandre Capitao Patrao / 48

5 Objectives Objectives How to use it: How to setup a simplified simulation of the flow around a propeller using the rotordisksource class and the simplefoam solver. How to define the propeller geometry in the correct way. How to setup a mesh in the correct way for best results. The theory of it: A brief description of the theory behind the actuator disk will be provided. The Blade element theory will be reviewed. How it is implemented: Description of how the rotordisksource class is implemented in OpenFOAM. How to modify it: It will be shown how to fix a critical bug in the rotordisksource class. It will be shown how to implement an improved tip correction factor and various propeller performance parameters. Alexandre Capitao Patrao / 48

6 Actuator disk Theory Alexandre Capitao Patrao / 48

7 Actuator disk Actuator disk fundamentals u i x i = 0 (1) u j u i x j R ij x j = p x i + S i (2) Alexandre Capitao Patrao / 48

8 Blade Element Theory Blade Element Theory Alexandre Capitao Patrao / 48

9 Blade Element Theory Blade Element Theory Alexandre Capitao Patrao / 48

10 Blade Element Theory Blade Element Theory Alexandre Capitao Patrao / 48

11 Blade Element Theory Blade Element Theory Alexandre Capitao Patrao / 48

12 rotordisksource class Implementation Alexandre Capitao Patrao / 48

13 rotordisksource class rotordisksource class implementation Originally developed by Stefano Wahono (Development of Virtual Blade Model for Modelling Helicopter Rotor Downwash in OpenFOAM) The rotordisksource class is activated at runtime if the solver finds a fvoptions file containing a rotordisk type entry. The rotordisksouce class source code can be found by typing: OF1706 cd $FOAM_SRC/fvOptions/sources/derived/rotorDiskSource tree../rotordisksource Alexandre Capitao Patrao / 48

14 rotordisksource class rotordisksource class files Alexandre Capitao Patrao / 48

15 rotordisksource class rotordisksource class files Alexandre Capitao Patrao / 48

16 rotordisksourcetemplates.c Modifications Alexandre Capitao Patrao / 48

17 rotordisksourcetemplates.c rotordisksourcetemplates.c Three main modifications are going to be made in the code: 1 Fixing an error in the calculation of the local forces. 2 Implementing a more general blade tip correction. 3 Implementing propeller performance parameters. Start by copying the entire fvoptions folder and make sure that the future compiled binary files end up in the user directory: foam cp -r --parents src/fvoptions $WM_PROJECT_USER_DIR cd $WM_PROJECT_USER_DIR/src/fvOptions sed -i s/foam_libbin/foam_user_libbin/g Make/files cd sources/derived/rotordisksource/ Alexandre Capitao Patrao / 48

18 rotordisksourcetemplates.c rotordisksourcetemplates.c Modify the code in rotordisksourcetemplates.c as shown below: // Logging info scalar drageff = 0.0; scalar lifteff = 0.0; scalar thrusteff = 0.0; //NEW LINE scalar torqueeff = 0.0; //NEW LINE scalar powereff = 0.0; //NEW LINE scalar AOAmin = GREAT; scalar AOAmax = -GREAT; scalar epsmin = GREAT; //NEW LINE scalar epsmax = -GREAT; //NEW LINE Mark and comment the lines shown below: // NEW LINE - HERE WE COMMENT THE TWO LINES BELOW: // Apply tip effect for blade lift //scalar tipfactor = neg(radius/rmax_ - tipeffect_); Alexandre Capitao Patrao / 48

19 rotordisksourcetemplates.c rotordisksourcetemplates.c scalar f = pdyn*chord*nblades_*area_[i]/radius/mathematical::twopi; /*=======================NEW LINES========================================*/ // Flow angle scalar eps = atan2(-uc.z(), Uc.y()); if (eps < -mathematical::pi) { eps = (2.0*mathematical::pi + eps); } if (eps > mathematical::pi) { eps = (eps - 2.0*mathematical::pi); } epsmin = min(epsmin, eps); epsmax = max(epsmax, eps); //Drela tip factor: scalar lambdaval = radius/(0.5*diameterref_)*tan(eps); scalar tipfactor_f = (nblades_/2)*(1-radius/(0.5*diameterref_))*(1/lambdaval); scalar tipfactor = 2/mathematical::pi*acos(exp(-tipFactor_f)); // Tangential and axial forces scalar ftang = (f*cd*cos(eps) + tipfactor*f*cl*sin(eps)); scalar faxial = (-f*cd*sin(eps) + tipfactor*f*cl*cos(eps)); vector localforce = vector(0.0,-ftang,faxial); /*========================================================================*/ // NEW LINE - HERE WE COMMENT THE LINE BELOW: // vector localforce = vector(0.0, -f*cd, tipfactor*f*cl); Alexandre Capitao Patrao / 48

20 rotordisksourcetemplates.c rotordisksourcetemplates.c Now we add expressions for calculating the overall rotor thrust, torque and power (NEW LINE): // Accumulate forces drageff += rhoref_*localforce.y(); lifteff += rhoref_*localforce.z(); thrusteff += rhoref_*faxial; //NEW LINE torqueeff += rhoref_*ftang*radius; //NEW LINE powereff += rhoref_*ftang*radius*omega_;//new LINE // Transform force from local coning system into rotor cylindrical localforce = invr_[i] & localforce; The propeller efficiency can now be calculated outside the loop, just before the code that outputs data to the terminal: scalar etaprop = thrusteff*refveleta_/powereff; //NEW LINE if (output) { reduce(aoamin, minop<scalar>()); Alexandre Capitao Patrao / 48

21 rotordisksourcetemplates.c rotordisksourcetemplates.c The rotor thrust, torque, power, and efficiency should be output into the terminal windows, and this is done by inserting the lines with //NEW LINE below: Info<< type() << " output:" << nl << " min/max(aoa) = " << radtodeg(aoamin) << ", " << radtodeg(aoamax) << nl << " min/max(eps) = " << radtodeg(epsmin) << ", " //NEW LINE << radtodeg(epsmax) << nl //NEW LINE << " Rotor thrust = " << thrusteff << nl //NEW LINE << " Rotor torque = " << torqueeff << nl //NEW LINE << " Rotor power = " << powereff << nl //NEW LINE << " Rotor propeller efficiency = " << etaprop << nl //NEW LINE << " Effective drag = " << drageff << nl << " Effective lift = " << lifteff << endl; Alexandre Capitao Patrao / 48

22 rotordisksource.h rotordisksource.h The changes that have been made utilize some new variables (diameterref_ and refveleta_) that have not yet been defined in the rest of the code. The next step is to declare and define these variables, starting in the file rotordisksource.h: //- Reference density for incompressible case scalar rhoref_; //- Reference diameter for calculating prandtl tip loss factor scalar diameterref_; //NEW LINE //- Reference velocity for calculating propeller efficiency scalar refveleta_; //NEW LINE //- Rotational speed [rad/s] // Positive anti-clockwise when looking along -ve lift direction scalar omega_; Alexandre Capitao Patrao / 48

23 rotordisksource.c rotordisksource.c Now modify the constructor functions in the file rotordisksource.c: Foam::fv::rotorDiskSource::rotorDiskSource ( const word& name, const word& modeltype, const dictionary& dict, const fvmesh& mesh ) : cellsetoption(name, modeltype, dict, mesh), rhoref_(1.0), diameterref_(1.0), //NEW LINE refveleta_(0.0),//new LINE omega_(0.0), Alexandre Capitao Patrao / 48

24 rotordisksource.c rotordisksource.c And one last change so that the propeller diameter and flight velocity can be read from the fvoptions dictionary: coeffs_.lookup("tipeffect") >> tipeffect_; // Reference diameter for calculating prandtl tip loss factor coeffs_.lookup("diameterref") >> diameterref_; //NEW LINE // Reference velocity for calculating propeller efficiency coeffs_.lookup("refveleta") >> refveleta_; //NEW LINE const dictionary& flapcoeffs(coeffs_.subdict("flapcoeffs")); All that remains now is to compile the code: cd $WM_PROJECT_USER_DIR/src/fvOptions wmake Alexandre Capitao Patrao / 48

25 Mesh requirements Case setup Alexandre Capitao Patrao / 48

26 Mesh requirements Surface patches Alexandre Capitao Patrao / 48

27 Mesh requirements Mesh Alexandre Capitao Patrao / 48

28 Mesh requirements ROTORDISK mesh Alexandre Capitao Patrao / 48

29 Propeller specification Propeller specification Alexandre Capitao Patrao / 48

30 simplefoam setup simplefoam setup The case setup is started in the easiest way by copying the tutorial case from the OpenFOAM installation directory. OF1706+ cp -r $FOAM_TUTORIALS/incompressible/simpleFoam/rotorDisk $FOAM_RUN/ cd $FOAM_RUN/rotorDisk Alexandre Capitao Patrao / 48

31 simplefoam setup Initial and boundary conditions The patch names of the 0/k, 0/nut, and 0/omega files need to be changed as is shown below. Change the names of the patches for all three files. INLET { type value } fixedvalue; uniform $kinlet; OUTLET { type inletvalue value } DISK_MANTLE { type } inletoutlet; uniform $kinlet; uniform $kinlet; slip; Alexandre Capitao Patrao / 48

32 simplefoam setup Initial and boundary conditions The patch names of the 0/p file also needs changing. Additionally, the DISK_MANTLE boundary condition should be changed to zerogradient. internalfield uniform 0; boundaryfield { INLET { type } zerogradient; OUTLET { type fixedvalue; value uniform 0; } DISK_MANTLE { type } zerogradient; Alexandre Capitao Patrao / 48

33 simplefoam setup Initial and boundary conditions Finally, the patch names of the 0/U file also needs changing. The OUTLET and DISK_MANTLE patches should be changed to zerogradient and the INLET patch boundary condition should have a velocity of m/s. Uinlet ( ); dimensions [ ]; internalfield boundaryfield { INLET { type value } OUTLET { type } DISK_MANTLE { type } uniform $Uinlet; fixedvalue; uniform $Uinlet; zerogradient; zerogradient; Alexandre Capitao Patrao / 48

34 simplefoam setup Import mesh The existing mesh needs to be deleted and the new one extracted from the accompanying files. The mesh is in ANSYS Fluent format, and needs to be converted using the fluent3dmeshtofoam utility: cd $FOAM_RUN/rotorDisk rm -r constant/trisurface/ tar -xvzf prop_incomp_refcase.tar.gz cp prop_incomp_refcase/fluent1.msh. fluent3dmeshtofoam fluent1.msh Alexandre Capitao Patrao / 48

35 simplefoam setup Setting up the fvoptions file (1/4) The properties needed for the rotordisksource class need to be included in the fvoptions file (it will be split into several slides): disk { type rotordisk; // Specifying that the rotordisksource class // is to be used for sources in the domain selectionmode cellzone cellzone; ROTORDISK; // how to choose which cellzone to use as the // cylindrical actuator disk volume fields (U); // Names of fields on which to apply source nblades 6; // Number of blades tipeffect 1.00; // Normalised radius above which lift = 0, // should be set to 1 if using a Prandtl tip corr. diameterref 1.0; // Propeller reference diameter (used to calculate // prandtl tip loss) refveleta ;// Reference velocity for calculating // propeller efficiency inletflowtype local; // Inlet flow type specification inletvelocity (0 1 0); Alexandre Capitao Patrao / 48

36 simplefoam setup Setting up the fvoptions file (2/4) geometrymode origin axis refdirection rpm specified; (0 0 0); //used for constructing the rotor coordinate system (0-1 0);//used for constructing the rotor coordinate system (0 0 1);// Reference direction // Used as reference for psi angle (ie radial dir) ;//rotational velocity of the rotor trimmodel fixedtrim; // fixed targetforce // fixedtrim if the rotor blade angle is fixed rhoref 1.225; // reference density rhoinf 1.225; Alexandre Capitao Patrao / 48

37 simplefoam setup Setting up the fvoptions file (3/4) // BLADE SPECIFICATION blade // This entry lists a number of sections which together define a // piece-wise linear propeller. { data ( (sect0 ( )) // radius [m], blade angle [deg], chord [m] (sect1 ( )) (sect2 ( )) (sect3 ( )) (sect4 ( )) (sect5 ( )) (sect6 ( )) (sect7 ( )) (sect8 ( )) (sect9 ( )) (sect10 ( )) (sect11 ( )) (sect12 ( )) (sect13 ( )) (sect14 ( )) (sect15 ( )) (sect16 ( )) (sect17 ( )) (sect18 ( )) (sect19 ( )) ); } Alexandre Capitao Patrao / 48

38 simplefoam setup Setting up the fvoptions file (4/4) profiles { sect0 // This entry gives cl and cd as function of angle-of-attack { type lookup; data ( ( ) //angle-of-attack [deg], cd, cl ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ); } Alexandre Capitao // HERE Patrao ARE SECTIONS 1-18, SEE APPENDIX / 48

39 simplefoam setup fvoptions and fvsolution files The complete fvoptions is very long (in this case), and is easily copied from the accompanying files: cp prop_incomp_refcase/system/fvoptions system/ The relaxation factors in the fvsolution file should be changed to 0.5: relaxationfactors { equations { U 0.5; "(k omega epsilon)" 0.5; } } The case can then be run with the command: simplefoam &> log & Alexandre Capitao Patrao / 48

40 Propeller performance Results Alexandre Capitao Patrao / 48

41 Propeller performance Performance results Performance values for the reference propeller design from design program, OpenFOAM simulation, and CFX full blade simulations: T [N] P [kw ] η Design target: % OpenFOAM: % CFX: % Alexandre Capitao Patrao / 48

42 Wake velocities Wake velocities - axial velocity Axial velocity contours for the reference propeller simulated with ANSYS CFX (left) and OpenFOAM (right). Plane located one D downstream of propeller. Alexandre Capitao Patrao / 48

43 Wake velocities Wake velocities - swirl velocity Swirl velocity contours for the reference propeller simulated with ANSYS CFX (left) and OpenFOAM (right). Plane located one D downstream of propeller. Alexandre Capitao Patrao / 48

44 Wake velocities Before and after bug fix Swirl velocity contours for the reference propeller simulated with the original (left) and modified (right) rotordisksource class. Alexandre Capitao Patrao / 48

45 Advance ratio sweep Performance for different advance ratios Performance values from OpenFoam (OF) and ANSYS CFX simulations for different advance ratios J = V 0 /nd. Alexandre Capitao Patrao / 48

46 Conclusions Epilogue Alexandre Capitao Patrao / 48

47 Conclusions Conclusions The rotordisksource class provides a very fast way of simulating the time-averaged performance of a propeller (3.5 minutes per case in this report). The rotordisksource class has a critical bug which results in greatly under-predicted rotor torque and downstream swirl velocities, but this bug can be fixed relatively easy. A more general tip correction factor than the existing one has been implemented, and results in a decrease in axial and swirl velocities at the rotor tip which are similar to what is found in CFD simulations of the entire blade geometry. The values of thrust and torque are under-predicted for the OpenFOAM simulations relative to the CFX full blade simulations, and future work should investigate the effect of different boundary conditions and turbulence models on the results. The variation in thrust and efficiency with respect to advance ratio is relatively similar for the OpenFOAM and CFX simulations, but the OpenFOAM values start to differ when the propeller starts stalling at lower advance ratios. Alexandre Capitao Patrao / 48

48 Study questions Study questions How to use it: 1 What is the rotordisksource class useful for? 2 What kind cells are best suited for meshing the actuator disk volume? 3 What OpenFOAM utility can be used to convert fluent meshes so that they can be used by OpenFOAM? 4 Why does the propeller efficiency differ significantly between the OpenFOAM and CFX simulations for lower advance ratios? The theory of it: 5 What is the main benefit of using an actuator disk over a full 3D blade simulation? 6 Does the Blade Element Theory account for 3D flow effects? Name some of these 3D flow phenomena. 7 Which flow velocities does the BET take into consideration when calculating a blade section angle-of-attack? How it is implemented: 8 How did the main error in the rotordisksource class manifest itself in the flow? How to modify it: 9 In which file is the main calculation for calculating the rotor forces located? Provide a full path using OpenFOAM environment variables. 10 How was the main error in the rotordisksource class fixed? Alexandre Capitao Patrao / 48