Revisions for ASWING v 4.3 19 Nov 1996 -------------------------- * Made the definitions of the flaps/ailerons more flexible. Up to three flaps can now be defined. Antisymmetric (aileron) deflection is set up by specifying antisymmetric dCL/dF and dCM/dF distributions on both left and right wings. * Angular accelerations in x and z can now be specified in addition to the rates. * Full flexibility is now provided for specifying moments in lieu of rates or accelerations. * Redundant or nonexistent constraints are detected before a solution is attempted, thus avoiding an arithmetic crash. * The variables plotted in the point-table above the loading plot can be toggled on/off with the O,V menus from .OPER, and the selection is saved to the operating-point save file (PSAV). * A few bug fixes were made to the input-file generating routine invoked with SAVE. Revisions for ASWING v 5.0 30 Nov 1997 -------------------------- * Multiple beam can now be defined, which allows modeling of an entire configuration. Both surface (lifting) and fuselage (non-lifting) beams can be defined. The beams can be either structurally isolated or joined in various ways. * The locations of point masses and strut/wire attach points are now specified directly via their cartesian x,y,z locations in the unloaded state. The previous arcane specification via the local c,s,n offsets is no longer used. Point masses now have a drag area CDA associated with them. * All moment and rotation components (pitching has been added) can now be imposed. Airfoil-section pitching moment of inertia is now an additional beam-distribution variable. Unsteady contributions and compressibility corrections have been added to the section Cm. * A general sensitivity or derivative calculation routine has been added. All the stability derivatives can be coughed up for the current operating state. * The format of the input file has changed slightly to allow the additional entities to be specified. Upgrading "old" input files should be relatively simple. * The theory document has been expanded, with more details on the implementation of point masses, bracing struts or wires, More details are also provided on the various boundary conditions required to close the elasticity equations. Revisions for ASWING v 5.1 2 Jan 1998 -------------------------- * Full unsteady treatment is now implemented. Solutions can be obtained in the time domain (time marching), or in the frequency domain (modal analysis). * The unsteady formulation is documented in the theory addendum. * Real-time movie of time-marched solution or eignemode time trace can be displayed. Revisions for ASWING v 5.2 20 Feb 1998 -------------------------- * The BODE frequency-analysis facitility is now implemented. Requires the ARPACK public-domain sparse eigenvalue package. * Flight conditions are now specified via a visual toggle table, rather than the old hard-to-follow prompt chains. * A user-defined control law can be imposed in lieu of explicit control deflection settings. All the facitilies can be used with this autopilot on or off. Revisions for ASWING v 5.30 2 April 1998 --------------------------- * Generalized the reference-location V,alpha,beta, accelerations, etc. * Switched from V,alpha,beta to Ux,Uy,Uz as primary variables. The former were causing major complication and confusion with the "reference" V,alpha,beta, which are different quantities. * Added a greatly-improved lift and drag post-stall model into the section characteristics. Convergence is now reliable all the way up to 90 degrees local incidence. Post-stall drag behavior is now rather realistic. * Added climb angle to list of specified constraints. Typically, this would be imposed to implicitly determine power setting. * Added matrix balancing to eigenmode analysis module, to reduce sensitivity to machine roundoff. * Added more smarts to eigenmode scaling. Airplane will no longer appear as a pixel speck a mile away during a phugoid excursion. * Made the flight-condition printout more complete and organized. Various toggle menus have been streamlined. * Xplot 4.38 now has screen buffering (courtesy of H.Youngren) to eliminate annoying flicker in the movie animation. May cause pixel blackout on window resizing -- use previous Xplot 4.37 if this is a problem. Revisions for ASWING v 5.31 27 April 1998 --------------------------- * Added an "All.Pnt" button to the constraint-selection table. This makes all points have the same settings as the ones shown, and saves a lot of laborious toggling and clicking to set up multi-point cases. * Added printout of inertias about CG in addition to the ones about 0,0,0. * Operating points can be changed inside Geometry Viewer via the + and - keys. Allows single-frame stepping through all points present. * Structural Euler angles are now listed in degrees as advertised. They were previously displayed in radians. * The gust parameters can now be read in from a file when a gust index of -1 is specified with the GUST command. Revisions for ASWING v 5.32 31 June 1998 --------------------------- * Pixel blackout with the new "flicker-free" Xplot11 was fixed by H.Youngren. Revisions for ASWING v 5.33 8 August 1998 --------------------------- * Fixed a bug in the node-allocation logic. Previously it would fail if the t parameter was equal and opposite at the two ends of a fuselage beam, for example t = -1.0 ... +1.0 . It would think the fuselage beam is "symmetric", and assign an incorrect zero-length interval at the middle node. * The new VGET command now reads a gust file. Before this was done via the arcane "-1" gust type option. * Almost all of the code is free of dependence on the number of hard-wired flap and engine indices KPFLP1,KPFLP2.. KPENG1... etc. This simplifies changing the maximum flap and engine dimensions NFLPX,NENGX in DIMEN.INC. The aswing.doc file has a list of what must be done for the change. * Added the E option to the MODE menu. This outputs the current eXamine'd eigenvector to a file (the W option just dumps all the eigenvalues). The output is done by SUBR.EVCOUT (in mode.f), which can be customized as needed -- one generally doesn't want the *entire* eigenvector. Revisions for ASWING v 5.34 8 September 1998 --------------------------- * Fixed a bug in some of the local-acceleration terms. The error showed up only in unsteady flight situations with extreme aircraft path curvature. The Udot vector now denotes the absolute rather than the relative acceleration. This is now more consistent with the Wdot vector, which denotes absolute rotational acceleration. * Added the "=" option to most of the menus. This simply displays the current solution parameters. Previously, there was no way to display this information without executing a calculation first. * Added grid-overlay enable/disable toggle to "O"ption sub-menu. Dynamic eigenmodes and response mode viewing is MUCH better without the grid overlay -- doesn't blink on/off. * Starting a new time-marching sequence from the current point now does NOT reset the time to zero. The starting time is taken directly from the starting point. This allows an arbitrarily long time sequence to be run in OPER (see aswing.doc). Revisions for ASWING v 5.35 15 January 1999 --------------------------- * Vector quantities can now be optionally specified in Earth axes. This is more convenient for specifing many non-level flight conditions, such as steady circling flight with the bank unknown. The Earth/Body axis toggle is a new button on the visual parameter-setting table, invoked with "T" in OPER and elsewhere. * The Earth-axes components of various vectors can now be listed. Useful for seeing things like ground speed in maneuvers, sink rate in turning flight, etc. * Single-key time stepping has been added to the OPER plots. A movie of the stresses (S), strains (C), and loading (L) can be displayed in the unsteady mode by holding down the point-advance ">" key. Revisions for ASWING v 5.4 28 March 1999 --------------------------- * A clear distinction is now made between the tension and elastic axes. Previously, the two were assumed to nearly cooincide. The computed moments are post- processed to translate them to these locations, making them consistent with the moment definitions in standard beam theory. The translated bending moments Mc and Mn are defined about the tension axis, and the translated torsion moment Ms is defined about the elastic axis. The translated moments are what's now displayed. The theory document has the translation definitions. There should be little change from the previous display, unless the s axis is located very far from the elastic and tension axes. * The "P" menu in OPER now allows plotting of selected quantities versus operating-point index. Previously, only unsteady point sequences could be plotted versus time. * The xxx.set file format has been changed to include the few new toggles. The 5.3-format xxx.set will still be read in properly, though. Revisions for ASWING v 5.41 4 June 1999 ---------------------------- * Added point parameter list at the top of the eigenmode root map. * All stability derivatives are now printed in OPER (with "M"). Previously, only the standard derivatives were printed. The non-standard derivatives, such as Clq, Cmr, etc, are nonzero for non-symmetric aircraft, or for any aircraft in non-symmetric flight conditions. * Each line in the "Engine" block of the input file now begins with KPeng, which is the index of the engine-power parameter controlling that engine. Previously, KPeng was assumed to be incremented sequentially for each line. Specifying KPeng explicitly allows... - one parameter to influence more than one engine - one physical engine to be influenced by more than one parameter In the second case, the physical engine is modeled as two superimposed point engines at the same location. Each point engine is controlled by one parameter, so the effect of the two parameters is in effect superimposed. * The F/Peng and M/Peng parameters specified in the "Engine" block are now scaled with the specified F and F*L units. This was not done in ASWING 5.40 (a bug). If the specified units were unity, e.g. L 1.0 m F 1.0 N then the results were correct (which is why I didn't see the bug). * For time-marching solutions, the "X" command in OPER can now accept a specified-parameter filename argument. The file contains a time trace of parameter values which are to be imposed during the calculation. This allows easy specification of closed-loop or open-loop forcing, such as a commanded bank angle or an explicit rudder doublet input. * The CONTROL routine argument list has been extended to include the current commanded state values and specified control-deflection values. These can be ignored or used to drive the control law or to explicitly bias the control output (e.g. to simulate a pilot biasing the control law with stick inputs). Previously the commanded state and specified deflection were added on externally to CONTROL's input and output. This "second guessing" is no longer done. CONTROL receives the actual state and commanded-state values, and whatever control equation residual CONTROL puts out is imposed as-is. * Fixed a few minor logic bugs in the NODES menu. The "F" and "P" commands were not processed properly. * Fixed minor bug in the preprocessing info dump. The individual-beam jig inertias for all beams except the first were not being displayed properly. There was no effect on the overall inertias or the solution. Revisions for ASWING v 5.42 7 July 1999 ---------------------------- * Fixed bug in xxx.asw file output routine. * Added a second argument to the aswing run command: % aswing xxx yyy This will try to read files xxx.asw , yyy.set , yyy.pnt , so that one xxx.asw geometry file can be easily run with a variety of yyy.set and yyy.pnt files. Previously these yyy files had to be manually read with SGET and PGET which got tiresome. If the yyy argument is missing, it is taken to be the same as xxx, thus defaulting to the previous convention. Revisions for ASWING v 5.43 15 March 01 ---------------------------- * Added circling flight relation to constraint table: Udot_Y - U_X Omega_Z = Ar (Ar = 0 in steady circle) * Added the Sensor entity, which is specified in the xxx.asw input file. This delivers a pile of quantities at a specified location. These quantities can be listed or displayed in time-trace, Bode, or eigenvector plots like all the other state parameters. * Added the sensor information to the call list of the CONLAW routine. The sensor information can therefore be used to drive control laws. * Made all the CONLAW global and sensor parameters indexable by the indices declared in the INDEXC.INC include file. * Created a more general engine formulation, implemented in the ENGINE routine. This allows the thrust and moment to be some user-specified functions of the power setting and the local flow velocities. * The parameter-scheduling file for time-marching calculations in OPER can now optionally have incremental values for any parameter. These are added to the parameter's initial value at the start of the time-march calculation. Revisions for ASWING v 5.44 02 Feb 03 --------------------------- * Added Makefile options for Intel Fortran Compiler (for Pentium 4, etc). Added GETARG0 function which doesn't barf if no Unix argument is provided. Revisions for ASWING v 5.45 27 Jan 04 HHY --------------------------- * Changed plotting code in io.f, xyproj.f to give more realistic lifting surface plots for swept wings. Projects chords "streamwise" rather than normal to the spar line. * Added option to dump "movie" as sequence of postscript frames. Added code to ASWING.INC, aswing.f, bode.f, init.f, mode.f, oper.f, option.f Revisions for ASWING v 5.55 27 Jan 05 --------------------------- * Fixed "freezing" of geometry for rigid-body calculations. * Added sinusoidal gust excitation for Bode response. * Added structural damping. New beam properties "tdeps" and "tdgam" have been added. * Added induced velocity effect on sensors. * Improved engine modeling treatment. Changed format of Engine data block. * Added propeller plume model (doesn't quite work yet) * Added input-error info printing to CONINI routine. * More detailed decriptions added to theory documents: - Engine models (asw.ps) - Propeller plume models (asw.ps) - Structural damping (aswu.ps) - Sinusoidal gust response (aswu.ps) Revisions for ASWING v 5.56 5 Mar 05 --------------------------- * Added the capability to declare two beams to be parts of the same physical surface. This turns off the finite-core model in their mutual vortex lattice induced velocities, giving more accurate results. Examples: - A winglet defined as a separate beam is declared to be physically part of the wing. This ensures a proper circulation continuation from wing to winglet. - A stab of a T-tail is declared to be the same physical surface as the vertical tail. The vertical tail then gets the proper endplating effect, which otherwise would be weakened by the finite core model. * Added spanwise-mode treatment of frequency gust * BODE now can have multiple forcing parameters active. Intended for multiple spanwise-mode gust components. Revisions for ASWING v 5.57 8 Apr 05 --------------------------- * Engine plume jet model implemented. Seems to work OK. Engine plume is visible to surface and fuselage beams, point masses, sensors. * Put limiters in propeller model type 1 to handle non-physical specified shaft power without crashing * Added propeller force listing output * Fixed Bode plot scaling for frequency gusts * Added dCD/dFlap surface beam variables. Intended for modeling a spoiler, which has a nearly linear cd response to deflection. This is not well suited to drag of conventional flaps, which have a more quadratic cd variation with deflection. Revisions for ASWING v 5.58 --------------------------- * Engine forces and moments are now taken to be part of the overall aero forces and moments in the stability derivative calculation. e.g. CXdP1 includes the direct contribution of engine thrust, in additional to secondary effects from propwash, etc. * Fixed sensor output bugs in BODE * Added adjustable reference length for spanwise gust wavenumber * If the engine thrust-axis vector Tx,Ty,Tz=0,0,0, it will default to -1,0,0 Revisions for ASWING v 5.60 27 Jul 05 --------------------------- * Added Reduced Order Modeling (ROM) capability to the MODE menu. * Added E command to MODE menu. This is identical to R command in BODE. The same flags are used in both BODE and MODE. i.e. when you select response variables in BODE, you will also be setting the variables which are output by the ROM generation command. * Added % command to OPER, BODE, MODE. This is the same as the T command, except that with % you have to give keyboard inputs rather than cursor click inputs. This allows running ASWING in batch mode with a keystroke stript file. * Unsteady Extension docuument (aswu.ps) has been significantly reorganized and expanded. Notation in sections 11,12,13 has been changed to match standard control notation. New material on Reduced Order Model theory has been added to section 13. * Changed the example control-function residual in SUBR.CONTROL from R = f(U,Q) - Delta to R = f(U,Q) + Delta in order to match the new theory writeup in the Unsteady Extension document, section 11.2.3 . For a linear controller, this will simply require changing the signs of all the gains (or change CONTROL). * Fixed default settings for circulation modes when a new geometry is LOADed. * session*.txt files have been updated * Added closed-loop control Bode analysis example to session2.txt file Revisions for ASWING v 5.61 1 Nov 05 --------------------------- * Added printname definitions of recently-added parameters * Added "Y" command in MODE (not in menu), which dumps eigenvectors via the user-customizable SUBROUTINE EVCOUT (in src/evcout.f). Partial eigen output is also available via the W,G commands. * Added the option to disable unsteady aero-lag effects (of shed vorticity). Done with new L command in the "K ontrol" submenu in OPER,MODE,BODE. Revisions for ASWING v 5.62 10 Nov 05 --------------------------- * Redid some internal procedures for calculation of control derivatives. Extended the SUBROUTINE CONTROL call list to return derivatives w.r.t. specified control deflection. * ROM model now has sensor output sensitivities w.r.t. control deflections. These are output in the new xxx_00.rom file. * Added related theory derivation to aswu.ps document. * Improved the labeling in xxx_nn.rom files. Revisions for ASWING v 5.63 5 Jan 06 --------------------------- * Fixed problem with gust-mode input prompt in BODE Revisions for ASWING v 5.64 20 Jul 06 --------------------------- * Blank lines can now contain tab characters (before this caused a read error) * Added = command to display current operating point which is the baseline state. * Fixed calculation of some derived variables in eigenvector, and added proper scaling to eigenvector outputs. Some of the angle variables in the eigenvector were output as radians, even though the unit was printed as "deg". Now there are actually printed in degrees. * Added gust-selection command I to MODE menu. Previous I command to change Arnoldi iteration limit has been removed (deemed unnecessary anyway). * Extended the ROM control-influence matrix w*B to include gust modes * BODE can now display the response of either the full system or the ROM system. The ROM system must have been previously generated in MODE. These compare sorta OK, but some strange results are present. Needs more testing. Revisions for ASWING v 5.65 20 Sep 06 --------------------------- * MODE will now only store and display eigenvalues with positive imaginary parts. There's less clutter on the root plot that way. The ROM generator will still properly use the complex conjugates of the complex modes. * Added "-" command to MODE to display current stored eigenvalues. * ROM matrices are now all written to one file, rather individual files. * In BODE, a near-zero response amplitude is now stored as 1.0e-10, rather than 1. * Fixed plotting selection of sensor vector parameters (couldn't select them before) * Updated the gust mode derivations in Section 9 of the Unsteady Extension document. Revisions for ASWING v 5.66 22 Dec 06 --------------------------- * A ground can now be pinned, or sliding, or both (rigid). Previously, only the rigid ground worked properly. * A joint endpoint can now be pinned, or sliding, or both (rigid). Previously only rigid joints were permitted. The joint types are specified by new 5th and 6th numbers on the Joint data line. If these are absent, the joint types default to the old rigid type. Hence, old .asw files will work as before. * A beam can now have multiple #2 joint endpoints without causing the dreaded "? No end kinematic constraint slot" error message. However, such overconstrained beams must have sufficient compliance in the appropriate directions to prevent matrix ill-conditioning problems. See the user guide. * Added acceleration-integrators to the state and sensor variables. * The sensor linear acceleration output now includes gravity, so that the sensor will now properly mimic an actual accelerometer. A sensor in level flight will now report an upward acceleration of g, whereas previously it would have reported zero. * The KU* and KQ* indices used to index variables in the control routines have been reordered to make the now rather extensive list better organized. This may require corresponding changes in your .con files. See the INDEXC.INC file where the indices are declared. Revisions for ASWING v 5.67 22 Jan 07 --------------------------- * Cleaned up node distribution and assignment logic. In the NODE menu, an insufficient number of nodes is now detected, and a new minimum value is automatically set, with warning message. So a minimum number of nodes (one per breakpoint), can be found simply by specifying a very small number with the i command. * Also fixed up the node display plot in NODE. The legend shows only those nodes which are present. It now also shows sensors and engines, which weren't displayed before. * Added the true velocity, Mach, and Mach_PG to main operating-point listing. * All the ROM info now goes into one .rom file Revisions for ASWING v 5.68 18 Feb 07 --------------------------- * Cleaned up the new joint treatment first added in v5.66 . The two optional KJtype flags were redundant, since there's only one physical joint (duh). Now, only one KJtype flag (0,1, or 2) is expected for each joint definition line. Absence of this flag defaults to KJtype=0 as before. * Fixed a small error in one of the terms contributing to the sensor acceleration and rotation rate, due to sensor displacement from beam deformation. * Added gust-mode response info to the .rom file Revisions for ASWING v 5.69 1 Jun 07 --------------------------- * Fixed SUBROUTINE CONINI so that it properly reads sensor-variable gains. * Added the ability to specify commanded sensor variables in the specified-parameter file given to the X command in OPER. Revisions for ASWING v 5.70 6 Jul 07 --------------------------- * Fixed treatment of sensor error-integrator variables. They weren't being updated properly during time-marching. Revisions for ASWING v 5.71 1 Sep 07 --------------------------- * Fixed uninitialized variable problem in BODE, which failed with g77. * Made two separate SUBROUTINE SECONDS versions: src/second_ifc.f for Intel fortan compiler src/second_g77.f for g77 compiler Set SECOND variable in bin/Makefile to appropriate version. Revisions for ASWING v 5.72 9 Sep 07 --------------------------- * Added ground effect. Both positive images (solid image surface) and negative images (constant-pressure images surface) can be simulated. Theoretical description is now in the steady theory document. A description has also been added to the aswing_doc.txt manual. * New G command now toggles ground effect for one or all operating points. It also sets the ground normal vector. * The old G command in OPER has thus been superseded. No great loss, since it was a leftover from versions 4.x, and in 5.x it was mostly redundant anyway: - Bank and elevation angles can still be set in the constraint table or with the ! prefix like any other variable. - Gravity magnitude can now be set with the A command, along with the air density. * Added PDFs of constraint-table snapshots for various trim setups. These are for the hawk case, but will apply to most other cases. * Fixed the plot library to properly make hardcopy Postscripts of solid-fill polygons, such as the square markers in the constraint-selection table. Previously these showed up as unfilled squares. There's no change in on-screen graphics. Revisions for ASWING v 5.73 --------------------------- * Added graphics enable/disable toggle to .PLPA menu from top level. This allows turning off all graphics operations when running ASWING in batch mode, to maximize speed. * If a bad variable keyword is found in the input file, ASWING will now stop after printing the usual error message. Previously, ASWING kept running, and the error message was likely to quickly scroll off the screen undetected. * Fixed an error in the apparent pitch inertia that's printed out in the info dump during startup. This had no effect on any calculations. Revisions for ASWING v 5.74 --------------------------- * The small green xyz axes in the perspective plots are now "3D". This makes is easier to see the orientation of the airplane. * Fixed a degree/radian conversion bug in SUBR.CONTROL, for Flap5 and above. * Added strain and shear stress to the sensor output quantities Revisions for ASWING v 5.75 --------------------------- * Added a new joint type, specified by KJtype=3 in the "Joint" block. This is the same as the rigid KJtype=0, except that it allows an angle change about one specified axis, dependent on the moment about that same axis. Hence, it simulates a sprung hinge. The angle vs moment relation is arbitrary, i.e. it can be nonlinear. * The distribution functions along a beam can now be specified in reverse order, with the t values either decreasing or increasing. For example, this makes it more convenient to specify a left image of a beam --- just copy the beam data, and negate all the t and y values using the multipliers. Data which is input in reverse order will be reversed upon input to make t strictly increasing internally. * Added an chord-unsweep flag, which allows selecting of how chord lines are to be plotted. F: Chord lines are normal to beam axis T: Chord lines are projected to lie parallel to x axis. The flag is toggled in the Options submenu, command U. * Beams which have the same physical index and adjacent endpoints are now plotted with a common coordinate in the Loading plot. So if a wing is defined via several separate beams (but with a common physical index), the entire spanwise loading will now be displayed. Previously, each of the beam pieces were displayed separately, which was kinda annoying. * Reduced the under-relaxation threshholds for the beam Euler angles, i.e. the max allowable Euler angle changes between each Newton iteration are now smaller (about 5 degrees). This discourages limit cycling with strongly nonlinear sprung hinges for KJtype=3 joints. Revisions for ASWING v 5.76 31 Jan 08 --------------------------- * Reformatted the ASWING.INC file so that all the variable descriptions are conveniently in-line, rather than at the end of the file. * Fixed the fuselage-beam loading plot (L command in OPER,BODE,MODE) to not produce glitches at zero-length intervals. This is strictly post-processing plot change --- there is no change in the solution calculations. Revisions for ASWING v 5.77 31 Jan 08 --------------------------- * Eliminated yellow from the color-index arrays. It was difficult to see on white paper hardcopy. * Fixed the sensor alpha,beta axes. They weren't being set up correctly whenever the A axis was "leaned" left or right, away from the x,z plane. I didn't see this bug before because the A axis is normally not leaned in any typical application. * Sensor V,alpha,beta can now be calculated from either 1) kinematic+induced velocities (i.e. everything), or... 2) kinematic velocities only Option 1) is what a real sensor would report. Option 2) is useful for gauging the kinematic alpha seen by a tail or wingtip. The options are toggled in the K menu, item K. * The aero-lag disable toggle (K menu, item L) now also applies to the _plotted_ f_aero load. Previously it was applied only to the actual f_aero load that's used in the calculations. The discrepancy was unnoticable in most cases, except maybe in fast modes or extremely violent maneuvers. * Added the inertial reaction load f_reac to the stuff which can be listed. The listing is toggled on/off in the K menu, item V. Also added to aswing_doc.txt a description of all the variables which can be selected for listing. * Changed # of sensors to 12 in DIMEN.INC * Added new gust types type=4 gust with specified x (streamwise) and z (vertical) velocity same as type 2 gust with x,z velocity specified along x-y line type=5 cosine vertical gust (z velocity) per FAR 25.341 specified by gust Vz and gradient distance H this gust can be slanted along an x-y line (similar to type 2,4) * Fixed the QVW*B control-rate-influence matrix that's output for the ROM model. * Fixed the broken-loop gain calculation in BODE. Revisions for ASWING v 5.78 12 Apr 08 --------------------------- * Fixed sensor overlay data reading and plotting, in the P submenu in OPER. Any sensor overlay data past the 8th one wasn't being read correctly. * Fixed ROM Bode calculation -- conjugate modes weren't being included in the Bode sum. * In the E submenu in MODE and R submenu in BODE... Added 0,-1 special cases to toggle all parameters or turn off all parameters. Reduces typing when lots of variables most be toggled. Revisions for ASWING v 5.79 22 May 08 --------------------------- * Fixed output parameter toggling Revisions for ASWING v 5.80 1 Jul 08 --------------------------- * Added phase step and mode-factor step to O menu and .set file * In O menu, changed "F rame-number limit" to "N umber-of-frames limit" The F item is now for changing the mode-factor step. * In BODE, fixed the ROM response calculation of the following quantities: - linear accelerations dUx,dUy,dUz - angular accelerations dWx,dWy,dWz - Earth position RX,RY,RZ - Heading angle Psi The problems were due to the fact that the ROM-basis eigenvectors do not properly capture the accelerations and the zero-eigenmode variables. These are now computed explicitly from the rotation rates and velocities: (dUx,dUy,dUz) = i omega (Ux,Uy,Uz) + Omega X (Ux,Uy,Uz) + (Wx,Wy,Wz) X U (dWx,dWy,dWz) = i omega (Wx,Wy,Wz) (RX,RY,RZ) = T (Ux,Uy,Uz) / (i omega) Psi = R3 (Wx,Wy,Wz) / (i omega) Revisions for ASWING v 5.82 2 Sep 08 --------------------------- * In BODE, fixed remaining problems with some of the sensor ROM responses. All the responses now appear to match well between the exact and ROM options. * Added a short document examining the ROM Bode calculation issues which have been giving headaches recently, and now appear resolved (knock on wood). rom_bode_test.pdf * Modified SUBROUTINE ENGINE to accept motor rotation rates, to allow computation of all "P-factor" prop force and moment components. This in principle allows prediction of Whirl Flutter for the case where the motor mount has elastic compliance. Test case whirl.* demonstrates the capability. * Added engine prop circle to the geometry plot. * Added sensor-axis triple to the geometry plot. * Fixed the unswept-chordline plotting. Now it looks reasonable even for very large deformations. * Updated the Unsteady Extension document. Added description of the new ROM response calculations for accelerations, Earth position, and heading angle. Added section describing the unsteady prop model needed for whirl flutter predictions. Added a table of contents. * Updated the Steady document. Added description of the new compliant-joint formulation, which is enabled with KJtype=3 (see aswing_doc.txt). Added a bunch of minor clarifications. Added a table of contents. Revisions for ASWING v 5.83 6 Sep 08 --------------------------- * Fixed engine plume modeling. Only the first beam (typically the wing) was correctly seeing the engine plumes. * Fixed the time-march plotting in OPER (V command in unsteady mode). This got broken when the new plot features were added in 5.82. Revisions for ASWING v 5.84 2 Oct 08 --------------------------- * Fixed the sensor's M and F outputs. These were returned as their xyz components. Now they are returned in their csn components which matches the documentation. * Fixed the apparent-mass inertias about the CG that are printed out on the initial info dump. These weren't quite right when the airplane was displaced significantly from the origin. Note that this does not affect the simulation calculations, since these do not depend on the apparent-inertia concept. Revisions for ASWING v 5.85 9 Dec 08 --------------------------- * Re-fixed the apparent-mass inertia output. * Fixed the stability-derivative output (M command in OPER). A few of the rotation-rate derivatives were missing a contribution from the origin-based velocity derivatives, for cases where the velocity reference point which isn't at the origin. Typically this was noticable only for cases where the reference point was relatively far from the origin. Revisions for ASWING v 5.86 15 Nov 09 --------------------------- * Added a least-squares constraint option to the constraint selection table. This is useful for trimming cases with underdetermined (redundant) controls. * Added least-squares weights for flaps and engines to F and E submenus in OPER * In the F submenu in OPER, a flap index of 0 specifies that all the flaps are to receive the flap value. This is useful for setting them all to zero. * Added two user-defined constraints, labeled C_user1 and C_user2. These are intended to be custom-defined, via hard coding in the SUBROUTINE GCON. Current constraints drive some sensor quantities to specified values. * The input file parser now allows a "!" character in the block-variable line, with everything after the "!" ignored. * Added a check and warning message for successive zero-length intervals. The check is invoked at startup and after a LOAD. * Added ability to list total alpha change at each spanwise location, due to all flap variables F1,F2... a_flap = Sum_k [ F_k (dcl/dF_k) / (dcl/da) ] Revisions for ASWING v 5.90 24 Nov 09 --------------------------- * Constraint table now shows only up to the max number of flaps and engines which are actually defined in the .asw file, and not up the the NFLPX and NENGX array limits as before. This allows making these array limits quite large, without making the constraint table unnecessarily large for simple cases. * In the interest of code simplification and long-term ease of maintenance, I got rid of the separate scalar and vector parameter arrays PARAMS,PARAMV, and put everything into one array PARAM. Nearly a hundred do-loops got eliminated as a result, which produced a significant shrinking and simplification of the source code. No difference is apparent to the user, except in the layout of some of the output-selection toggle menus. Revisions for ASWING v 5.91 10 Feb 10 --------------------------- * Set KBREAK=1 for Cm variable, so if it has any discontinuities they will be properly interpolated onto zero-length intervals. * Everything after a "!", "#", or "%" character in the data lines in distribution blocks is now guaranteed to be ignored. Previously, if the comment contained a valid keyword Aswing would recognize it and try to act on it, giving an error message. Revisions for ASWING v 5.92 20 Feb 10 (from HYY) --------------------------- * Fixed wake-filament plotting Revisions for ASWING v 5.93 18 Mar 10 --------------------------- * ASWING will now terminate if the sensor or engine array limits are exceeded when the .asw file is processed during startup or LOAD. Before there was only a warning message which was easy to miss. Revisions for ASWING v 5.94 2 Feb 11 --------------------------- * Added wing-flapping setup code in q0flap.f, which was used to create the flapping bird movie on the Aswing web page. The setup is hard-wired, and it's all commented-out in the standard code version. * Added additional file-save options to the Y command in OPER. * Fixed the control-derivative WRITE statements in io.f, so that a string-overflow error no longer occurs if a very large number of flaps or engines is used. Instead, the string is simply truncated, and can be increased in size via the NLDIM parameter in io.f. Revisions for ASWING v 5.95 27 Jul 11 --------------------------- * Added state-history and state-snapshot input/output capability, courtesy of Harold Youngren Revisions for ASWING v 5.96 15 Mar 12 --------------------------- * Added CY,Cl,Cn,sideforce to point arrays and point printout ( "=" command in OPER). These can also be displayed in the parameter list at the top of the loading plot ("L" in OPER), after being selected in the K menu, item P. Revisions for ASWING v 5.97 20 Jul 12 --------------------------- * Point masses can now have inertias Ixx,Iyy,Izz,Ixy,Ixz,Iyz. This eliminates the need to use mass dumbells to represent concentrated masses with non-negligible inertias. Revisions for ASWING v 5.98 15 Mar 14 --------------------------- * Fixed problems with calculations in graphics-disabled mode. The constraint setup routine (% command in OPER) wasn't parsing the inputs properly. Revisions for ASWING v 5.99 13 Dec 14 --------------------------- * Eliminated the limit on the number of pylon objects (masses, engines, etc.) which can be attached at any one node. Previously this was limited to 1000 objects, or parameter NPX (in DIMEN.INC), whichever was smaller. Now, the limit is set only by NPX. * Fixed some problems with the inertias associated with point masses, which were added in version 5.97. The products Ixy,Ixz,Iyz were not being handled properly. The inertias added to a point mass now appear to be working correctly -- a cloud of point masses each with zero inertia, if attached at the same beam node, can be replaced with one mass with the equivalent inertias, to give exactly the same solution. * Added CG_Calc.pdf to the documentation which describes the calculation of the total inertia of any collection of masses. * Added file session5.txt, which describes the calculation of aileron reversal speed. Addendum for v 5.99 8 Mar 15 - - - - - - - - - - - * Fixed filename handling bug in M command in OPER The "Y" response set "Y" to be the matrix output filename. It now takes the default filename as it should. Revisions for ASWING v 6.00 22 Aug 15 --------------------------- * Improved the calculation of very low frequency gust response in BODE. This now correctly produces the expected quasi-steady solution response, at least within the linearized approximations implicitly assumed by any Bode analysis. * Added gust-type 8, which is a sinusoidal vertical gust of indefinite duration. See SUBROUTINE VGUSTE in src/vgust.f Revisions for ASWING v 6.01 22 Mar 16 --------------------------- * Added two special cases to the beam display selection sub-submenu K, B: 0 (toggle all) -1 (turn off all) Revisions for ASWING v 6.02 8 Oct 16 --------------------------- * Eliminated annoying "insufficient nodes" messages on startup, for large cases which have the proper number of nodes set up in the .set file. * Added session6.txt file demonstrating Reduced Order Model generation, and comparison of its Bode response with the full-state response.