Problem with rigid Diaphragm + eigenvectors

[r0d4n]

Dear Silvia and Frank,

I'm modeling a 3D frame, 2x2 bays, 4 storeys and processing TH analyses.
Columns and beams are fiber elements.

I was using elastic shells (IsotropicElastic NDmat=> PlateFiber NDmat => PlatFiber Sec => ShellMITC4) to model the floors (R.C. thickness 15 cm/5.9 in), but I decided to change them with rigid diaphragms in order to keep the same mutual distance between nodes of the same levels during the analysis.

The modal analysis worked pretty well before the change but now, when I try to carry out the eigenvectors I obtain weird modal shapes. The diaphragms seem to rotate on the dummy node.

The dynamic analysis seems to work almost correctly, even if the damping is messed up depending on the weird modal analysis.

Do you have any ideas about how to solve this problem?

Let me know if you need more detailed info.

-------------------------------------------------------------------------------------

typical floor node disposition (1st level):

11.....6m.........12...4m....13
x------------------x------------x
|.......................|...............|
|.......................|...............|.5m
|.......................|...............|
x14--------10@-x15---------x16....--------> x (direction of earthquake)
|.......................|...............|
|.......................|...............|.5m
|.......................|...............|
x------------------x------------x
17..................18.............19



# The sides of the frame are 10mx10m

set L 10.
set Lhalf [expr $L/2]

node 10 $Lhalf $y1 $Lhalf
node 20 $Lhalf $y2 $Lhalf
node 30 $Lhalf $y3 $Lhalf
node 40 $Lhalf $y4 $Lhalf


# MASTER NODES
################################

# MNode Dx Dy Dz Rx Ry Rz
fix 10 0 1 0 1 0 1
fix 20 0 1 0 1 0 1
fix 30 0 1 0 1 0 1
fix 40 0 1 0 1 0 1

set RigidDiaphragm ON

set perpDirn 2


rigidDiaphragm $perpDirn 10 11 12 13 14 15 16 17 18 19
rigidDiaphragm $perpDirn 20 21 22 23 24 25 26 27 28 29
rigidDiaphragm $perpDirn 30 31 32 33 34 35 36 37 38 39
rigidDiaphragm $perpDirn 40 41 42 43 44 45 46 47 48 49


Other info:

I'm using rigid links for rigid offsets. It doesn't seem to give any kind of problem

I apply nodal masses in x and z direction. (applied to the nodes showed above, dummy nodes excluded)

Dynamic Analysis parameters:

constraints Lagrange
numberer RCM
system UmfPack
test EnergyIncr 1.e-8 10 0
algorithm Newton
integrator Newmark 0.5 0.25
analysis Transient