Unequal axial forces in Section 1 and 2 of dispBeamcolumn el

[Katrin]

Hi,

I am carrying out transient analysis of a wall where I have modelled the plastic hinge zone with a dispBeamColumn element with 2 integration points. The element has a fibre section of concrete and reinforcement fibres which has been tested with moment-curvature analysis. The analysis is basically a bi-directional pushover analysis. I am trying to model a quasi-static cyclic test of a wall which was pushed in both horizontal directions.
My problem is that the axial forces I get in in Section 1 and 2 are very unequal (about +/-10%) and therefore the moment capacity at the wall base is incorrect. The average of the axial forces at Section 1 and 2 is correct. I am not an expert in beam element formulation but I understood that for the displacement-formulated beam element only global equilibrium is checked while the local equilibrium at the integration points is not checked. Hence, I was wondering whether this is an intrinsic problem I have to live with or whether there are any ways to improve the performance? I enclose a part of my tcl-file which shows the load application.

Many thanks for any help!!
Katrin



set factor 1000
set dt_input [expr $factor]
set nodeTag 9; # h=2.95m, Rotation
set nodeTagx 10; # h=3.35m, EW
set nodeTagy 9; # h=2.95m, NS

# define Motion histories
set rampX "Series -dt $dt_input -filePath $FolderID/zero_hist.txt"; # EW direction
set rampY "Series -dt $dt_input -filePath $FolderID/hist.txt -factor -1.0";
set disp_zeros "Series -dt $dt_input -filePath $FolderID/zero_hist.txt";

pattern MultipleSupport 2 {
groundMotion 101 Plain -disp $rampX
groundMotion 102 Plain -disp $rampY
groundMotion 100 Plain -disp $disp_zeros
imposedMotion $nodeTagx 1 101
imposedMotion $nodeTagy 2 102
imposedMotion $nodeTag 6 100
}


puts "test Energy Increment"
test EnergyIncr 0.1e-2 10000

algorithm Newton
system SparseGeneral -piv
constraints Transformation
integrator Newmark 0.5 0.25
numberer RCM
analysis VariableTransient

set dt_analysis [expr 1]
set Nsteps [expr $factor]
analyze $Nsteps $dt_analysis

[silvia]

with the displacement beam-column element, you need to define a larger number of elements, right?

[Katrin]

Hi Silvia,

I am modelling the wall with 9 elements along the height. The lowest I gave the length of the plastic hinge zone because I wanted a linear curvature profile in that zone. I also tried to vary the element length but it had no effect on the axial load.
I am very grateful for any ideas what else I could try!!

Katrin

[silvia]

i don't know, just make sure you recheck the entire model.
maybe test it under simple loading, like just axial load.