Since ZAERO is a higher order panel code than the Aero Options provided by NASTRAN, it is not as sensitive to panel cuts (e.g., box AR).  However, there are still guidelines that should be followed. Please see Chapter 5 of the ZAERO User's Manual for ZAERO Modeling Guidelines. A simple rule of thumb would be that a minimum of 10 chordwise boxes are required to capture a wave associated with a reduced frequency of 1.0. If k goes up to 2.0 then you will need more than 20, and so on. This assumes your reference length (REFC entry of the AEROZ bulk data card) is equal to the root chord length of the wing (or a representative chord length of the configuration).

Similar to the lifting surfaces described above, the chordwise divisions are a more important consideration when modeling bodies. Also important for the body is to capture accurate cross-sectional areas. This is the key to obtain accurate unsteady aerodynamics results.  Since unsteady aerodynamics is not like CFD, exact modeling of the fuselage (i.e., exact shape) is not critical. So try to keep the number of circumferential cuts to a minimum so that the AIC computation time doesn't take forever. In other words, you could model the configuration exactly with aero boxes, but your flutter results wont be that much different, and you may have to wait much longer for the results.

Yes, set the root chord or tip chord length (RCH or TCH entry of the CAERO7 bulk data card) to zero.

No, the chordwise divisions do not have to be equal and in alignment. Only in the streamwise direction do the spanwise divisions of in-plane (or close to planar) lifting surfaces need to be in alignment.

At this point there is no way to just dump the aero model and stop the program execution. But what you can do is set up the input deck of the aero model only and introduce a bug in a downstream module (like the spline input).  You could leave out the spline input and the program will terminate with errors but will have already punched out the aero model (assuming you have a PLTAERO bulk data card in the input deck).

Yes, the FEM and aero grid id's can overlap and do not need to be distinct.

The AEFACT must be in percentage from 0.0 to 100.0 - 0.0 to 1.0 is not allowed.

Multiple segments are used when the shape of the fuselage has drastic changes such as before the inlet section of the fuselage. However, if you can model the whole body with one segment you simplify the input setup effort. For this reason using a single segment when possible is always better.

Since your FEM model is a full model, you need to model the complete aerodynamic model to match the FEM model and use BOUNDARY=ASYM throughout the Executive Command ASSIGN statements. You must also set XZSYM entry of the AEROZ bulk data card to NO. Then model both the left and right-hand sides. See the SEGMESH bulk data card for modeling the body segments.

BIG NOTE: We strongly recommend that you cut the FEM model in half and only model half of the aero model. Otherwise, you will have a very large aero model that in the long run will only waste a lot of your time (e.g., run time, file sizes, etc.)

No, modeling the arbitrary body with flat sides will not lead to problems since the fuselage appears to be pretty smooth in the freestream direction. No, modeling the fuselage as elliptic is not good for this case since the rear portion of the fuselage is constant square shaped - stick with the true cross-sections (i.e. arbitrary input option, ITYPEi=3 in the SEGMESH bulk data card).  Note though that you must model the strake and wing with CAERO7 cards.

No, whatever you can do with DLM you can also do with ZAERO. The weighting factor can be introduced thru the CPFACT bulk data card.

Yes you can. It's okay because the body panels use source singularities. If they were vortex singularities like on the wing panels then streamwise direction cuts become important.

The wing chordwise divisions and body axial cuts do not have to line up and there wont be any singularity problems. It's just cleaner to model it all in alignment. Just remember to use the ATTCHR or ATTCHT to attach the wing to the body. This will give more accurate wing-body interference effects.

No, just use equal cuts. ZAERO utilizes high order panel methods and therefore does not require higher density spanwise cuts near the wing tip.

No, this is not true for ZAERO.  The gaps between the wing panels and fuselage will not be a problem. By using the ATTCHT condition in the CAERO7 you've effectively created a vortex-carry through wing to the centerplane (X-Z plane) of the fuselage. Without the ATTCHT turned on, the strake would be treated as a "free-edge" lifting surface. But with the ATTCHT entry of the CAERO7 bulk data card refering to the BODY7 (fuselage), the wing-body juncture interference effects will effectively be accounted for.

No, the egde of the lifting surface does not have to be tangent to the body.

You can model the airfoil by specifying the camber lines, half thicknesses and leading edge radii. Set the camber line to the wing to the mean line (dead center) for all x-coordinate locations. Your half thicknesses can then be specified for each x-location giving you the upper and lower surface points. Keep in mind that for unsteady aerodynamics we are computing Delta Cp (lower - upper surface pressure) on a flat plate - not physical steady surface pressures like for CFD. The effect of ZONA7U is like Piston Theory in that a surface slope is used to account for the thickness effect of an airfoil. So in essence it's like a delta surface slope (lower - upper) thats used - DC/DX. See the picture on page 6-11 of the User's Manual - you'll understand.

Flutter tracking will not print out if the flutter velocity or dynamic pressure are <= 0.0. The PLTFLUT bulk data card will not print out as well if the frequency is zero. Also, make sure that the flutter mode you are trying to dump exists in the g-method output tables. For example, if the g-method dumps 2 tables for the first 2 lowest flutter mode and I request the 3rd lowest flutter mode be dumped via the PLTFLUT card, then it wont print out anything.

The inclination is the slope of the individual box, not from the nose of the body. The Mach line eminates from the leading edge of each aero box.

When you say any part of the box or just the control point - it's the same thing, since the aero boxes are flat.

Nothing will happen, the program will know that the box is not really superinclined and it will treat this box as a regular panel.

Yes, it is necessary to specify super-inclined boxes for the ZONA7 + ZONA7U methods but for body boxes only. It is never necesssary to specify superinclined boxes for lifting surface boxes. Also note that superinclined boxes are not applicable to the subsonic ZONA6 method since it is Mach number dependent (occurs only for M>1).

Yes, the chordwise locations can be varied from strip-to-strip.

The effect of the mean flow condition specified in the TRIMFLT bulk data entry should have a small effect on the flutter results based on the ZONA6/ZONA7 linear aerodynamic analysis.

Setting the BOUNDARY entry on the ASSIGN executive control command to either SYM or ASYM will printout:

SYMMETRIC FINITE ELEMENT MODAL RESULTS ARE SUCCESSFULLY READ IN FROM FILE xxxx

since ASYM is treated as a form of SYM.  For BOUNDARY = ANTISYMMETRIC,

ANTISYMMETRIC FINITE ELEMENT ...

will be printed to the output.

The aerodynamic models in generated by the PLTAERO bulk data card and PLTMODE bulk data card are different because corner grid points are shared between aero boxes for the PLTAERO ouput file, while for the PLTMODE output file each aero box has its own 4 grid points.  Assigning 4 grid points to each aero box is required to generate an animated mode shape file, since the spline matrices are used to interpolate the structural mode shape onto the aero model. In order to display aero model discontinuities such as control surface deflections, each aero box must have its own 4 grid points.  With shared grid points (like in the PLTAERO output file), such discontinuities cannot be displayed.

You can use the FLUTTF bulk data card of ZAERO to alter the generalized stiffness of the structure. To do this, you must first delete all elements in the FEM model except the elements of the actuator (but leave all grid points unchanged). By running the free vibration analysis of NASTRAN with the DMAP alter statements as follows:

ASSIGN OUTPUT4='kgg.dat' STATUS=UNKNOWN UNIT=11 FORM=FORMATTED
$ Normal Modes Analysis, Database
$ COMPILER LIST $
SOL 103
COMPILE SEMODES SOUIN=MSCSOU LIST NOREF $
ALTER 95 $
MATGEN EQEXINS/INTEXT/9//LUSETS $
MPYAD INTEXT,KGG,/KGGE/1 $
MPYAD KGGE,INTEXT,/K3GG $
OUTPUT4 K3GG//-1/11/2 $
ENDALTER
CEND

NASTRAN will output the whole stiffness matrix on a file called "kgg.dat" in the OUTPUT4 format. This stiffness matrix has G-set d.o.f (same as the baseline FEM model) but has many zero elements except at d.o.f associated with the actuator. Note that NASTRAN may fail to complete the analysis due to the ill-conditioned stiffness matrix. But this does not matter so long as NASTRAN can complete the stiffness matrix assembly.

Once this stiffness is obtained in the file "kgg.dat", it can be imported to ZAERO by an "ASSIGN MATRIX=" Executive Control Command such as:

ASSIGN MATRIX='kgg.dat',MNAME='K3GG'

To perform a massive flutter analyses for different stiffness of the actuator, One can set up many SUBCASES in the Case Control Section of the ZAERO input. Each SUBCASE points to a different FLUTTER Bulk Data Card that refers a different FLUTTF Bulk Data Card. An example of such a FLUTTF Bulk Data Card is shown as follows:

$FLUTTF    SETID    GSET    C1    MASS    C2    DAMPING    C3    STIFF
FLUTTF      10       YES                                  -1.0    K3GG

In so doing, ZAERO will add the generalized stiffness of the C3*[STIFF] matrix to the generalized stiffness matrix of the baseline (imported by the "ASSIGN FEM=" Executive Control Command which contains the modal results of the baseline FEM model). In the example shown above, because C3=-1.0 , after adding the two generalized stiffness matrices the stiffness of the actuator is nearly zero. To increase the actuator stiffness, one can increase C3 incrementally. At each increment, a flutter analysis can be performed by referring to a different FLUTTF bulk data card that contains the incremental C3 value.

The matrix specified in the MNAME entry of the PZTMODE bulk data card must contain G-set number of rows and 1 column, where G-set is equal to 6 x no. of structural finite element grid points. The data contained in the matrix represents the structural deflections of the 6 degree of freedom (d.o.f.) at each structural grid point. PZTMODE can be used in many different ways. For example: