DFDC v070-ES1a USER NOTES ---------------------------------------------------------------------------------- Esotec Developments, March 2009 philip (at) esotec (dot) org This document does not replace the DFDC docs but appends them. Only issues relevant to this release are addressed here. COMPILATION Building on a Unix/Linux/OSX system is recommended. Windows files have not been modified and compilation of ES1 on Windows has not been tested. A brief look at the .dll sources indicate that the code MIGHT compile as currently configured, but some changes introduced in ES1 will be missing. Proceed at your own risk (or fix it if you can). The Whirlwind wrapper was incomplete for 0.70, so will be more incomplete for ES1. My advice to Windows users: the easiest way to run this distribution is to get Linux. The native Unix interface is very efficient, given a bit of effort learning it. g77 seems to have problems with the file read code. Intel Fortran and gfortran both have been applied without problems. Makefiles configured for OSX-gfortran are included. AERO Aero has received a thorough cleanup. The menus should be self-explanatory. FLIP inverts 3 parameters (CLmax, CLmin, CLCDmin) for negative circulation sections. A message will appear when a section is set for negative circulation. Note that the parameter dCd/dCL**2 essentially determines how quickly drag increases with CL above and below CLCDmin. Secondary dCd/dCL**2 has the same effect, but is a contribution from the annulus. I can't help with appropriate settings for this parameter. Note that A0 (zero-lift alpha) remains unchanged for negative-circulation sections, which explains the transformation to local coordinates (see BANG, below). Plotting in Aero: Use PLR and PLM to plot a set of Re values at constant Mach, or Mach values at constant Re. SETM changes Mach constant (default is 0). SETR changes Re constant (default is REref). Input 0 for the default. Standard Mach set: 0, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 (same as 0.70). Standard Re set: REref * 0.5, 0.75, 1.0, 1.5, 2.0. Or input any other set, which is stored for that Aero session until changed. Note that blade stagger and solidity are set at zero for all AERO plots. ALPHA Reporting I find this information useful, in particular for observing the influence of cascade effects (stagger and solidity) on CL vs Alpha. Alpha can be plotted in Plot 12. BANG DFDC uses a single coordinate system for all disks, resulting in rather unintuitive blade angles for negative-circulation disks. The BANG command toggles Beta and Alpha output between the internal "global" coordinate system used by DFDC and the "local" coordinate system of the negative-circulation disk. The transformations are as follows: Beta (local) = PI - Beta (global) Alpha (local) = 2 * A0 - Alpha (global) A0 is the interpolated zero-lift Alpha, stored each time airfoil data is called. A converged case can be output in global or local coordinates without re-executing the case. Just toggle BANG before DISP or Plot 12 to display current Beta and Alpha in either coordinate system. List and plot output of negative-circulation disks include a message indicating which coordinate system is being used. The coordinate transformations are applied at output only and do not affect internal data in any way. UVC Vector arrow length is now proportional to Q/Qref (roughly 1 cm per unit). Hence, adjust Vref to adjust the length of the arrows. Miscellaneous Notes Remember that negative-circulation disks (stator or negative-rpm rotor) have BGam and CL < 0. Hence, when executing DESI a negative CL must be specified and airfoil parameters must be set for negative circulation. For negative-circulation disks, PITC works in reverse (negative input to increase pitch). Suggested rotor design procedure is to set actuator disks and adjust BGam to achieve required power input or thrust. Execute. This sets up the MIL flow field. Then execute DESI to design blades within that flow field with the specified BGam and CL. Currently, for multi-disk cases, use either actuator disks or bladed disks, not a combination of the two. The THRU command currently drives pitch (or BGam) for Disk 1 only, no matter which disk is selected. This is complicated to fix so I've left it for now. This routine converges very slowly and needs work. With practice it is possible to converge on thrust quite quickly by manually adjusting pitch (or BGam), even in multi-rotor cases. QDES and GDES are unchanged from 0.70, with the exception of the GNAM command in GDES for editing the geometry name. I suggest using XFOIL for creating the duct/centerbody geometries. Use SCAL and TRAN to put into dimensional meters and copy into the case file. The solver doesn't seem to like executing a tip gap case following a run without tip gap. Hence, set tip gap first, before executing anything. END !