Craziest APDL mistery ever
 

Hello everybody,

I have to say that I started using ANSYS APDL only a few months ago but this just doesn't make any sense to me...

So I start using ANSYS 15.0 from the APDL Product Launcher...
Simulation environment: ANSYS
License: ANSYS Mechanical
Add-on Modules: None

And then I wrote this code to obtain the first five modal frequencies of a horizontal beam fixed at both ends:

FINISH ! Exit from processor
/CLEAR,NOSTART ! Clear database (not reading start145.ans)
/FILNAME,AxialLoadEstimation,ON ! Change Jobname for the analysis (start new files)
/Title,Axial Load Estimation for a steel bar with circular section

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

/PREP7 ! Enters mode creation preprocessor

MatDens=1101 ! Material's Density (kg/m3)
MatE=1.95e11 ! Material's Elastic Moduli (N/m2)
MatPoisson=0.3 ! Material's Major Poisson's Ratio
!
BeamLength=2.501999201 ! Length of the beam (m)
SecRad=0.0075 ! Radius (m)
!
AxForce=22400 ! Axial Force applied (N)


ET,1,beam188 ! Define local element > type number 1 > beam188

! Defines linear material properties
MP,DENS,1,MatDens ! Density = 1101 kg/m3
MP,EX,1,MatE ! Elastic moduli = 1.95x10E11 N/m2
MP,PRXY,1,MatPoisson ! Major Poisson's ratio = 0.3

! Create keypoints and line
K,1,0,0,0 ! Keypoint 1 (0,0,0)
K,2,BeamLength,0,0 ! Keypoint 2 (BeamLength,0,0)
L,1,2,(BeamLength*100) ! Line between K1 and K2 with divisions by cm
k,3,0,1,0 ! Keypoint 3 (0,1,0) [Orientation]
SECTYPE,1,BEAM,CSOLID ! Circular solid beam section
SECDATA,SecRad ! Radius=18.525mm (Diam=37.05mm)
! 8 div around circumference + 2 div through radius

LATT,1,0,1,,3,,1 ! Associates element attributes with unmeshed lines
! MAT=1(beam188),0,TYPE=1(steel)
! KB=K3 KE=
! SECNUM=1(circular)

LMESH,1 ! Generates nodes and line elements along Line 1

! Definition of DoF constraints at nodes
D,ALL,UZ,0 ! Neglect movement in z-axis
D,ALL,ROTX,0 ! Neglect rotation in x-axis
D,ALL,ROTY,0 ! Neglect rotation in y-axis

D,1,UY,0 ! Neglect movement in y-axis
D,2,UY,0
D,1,ROTZ,0 ! Neglect rotation in z-axis
D,2,ROTZ,0

F,1,fx,-AxForce ! Specifies force load at node 1
F,2,fx,AxForce ! Specifies force load at node 2

ALLSEL

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

/SOLU ! Enter solution processor

ANTYPE,STATIC ! Static analysis

PSTRES,ON ! Specifies that prestress effects are calculated
EMATWRITE,YES ! Forces the writing of all the element matrices to File.EMAT

SOLVE ! Starts solution

FINISH

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

/SOLU ! Enter solution processor

ANTYPE,MODAL ! Modal analysis

MODOPT,LANB,5 ! Extraction of 5 modes by Block Lanczos

PSTRES,ON
D,1,UX,0 ! Neglect movement in x-axis
D,2,UX,0

SOLVE ! Starts solution
FINISH

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

/POST1 ! Enter database results postprocessor

SET,LIST ! Defines data set by scanning + listing each load step
! Extract modal frequencies

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

I copy-paste this script and ANSYS gives me a first modal frequency of 78.140, but if I change the parameter BeamLength from 2.501999201 to 2.5019992, the frequency falls to 28.365!! Which is the frequency of the beam but unloaded...

I don't know why by making the beam 0.000001 mm shorter the results change like this... Does anybody know what am I missing??