about dirction definition

[gaohouzhi]

Recently,I construct a 10 story structure in opensees.
and when I load the Mx ,some results is abnormal .
Following is my tcl commands:

model basic -ndm 3 -ndf 6
set conBC 1
set stlBC 2
set millin 1000000
set revmilli 0.000001

##（N/m2)

uniaxialMaterial Concrete02 $conBC [expr -20.1*$millin] -0.002 [expr -10.0*$millin] -0.0033 0.5 [expr 2.01*$millin] [expr 30000*$millin]

uniaxialMaterial Steel01 $stlBC [expr 335*$millin] [expr 200000*$millin] 0.3

set secTorsioncolumn111 10

uniaxialMaterial Elastic $secTorsioncolumn111 [expr 14.4839*$millin]
##


set IDcolumnTrans 7302
geomTransf Linear $IDcolumnTrans 0 -1 0

set NIntPCol 10
#-------------basic parameters has been defined----------
set n111 15
node $n111 0 0 0 -mass 0 0 0 0 0 0
set n112 51
node $n112 0 0 4.4 -mass 3.29906 3.29906 3.29906 3.29906 3.29906 0

fix $n111 1 1 1 1 1 1

set seccolumn111 346
set Agseccolumn111 541
###-------------------section of coloumn-----------------------

section Fiber $seccolumn111 {
patch quad $conBC 16 20 -0.2 -0.25 0.2 -0.25 0.2 0.25 -0.2 0.25;
layer straight $stlBC 2 [expr 310*$revmilli] -0.17 0.22 0.17 0.22;
layer straight $stlBC 1 [expr 310*$revmilli] 0 0.22 0 0.22;
layer straight $stlBC 2 [expr 200*$revmilli] -0.17 0.073 0.17 0.073;
layer straight $stlBC 2 [expr 200*$revmilli] -0.17 -0.073 0.17 -0.073;
layer straight $stlBC 2 [expr 310*$revmilli] -0.17 -0.22 0.17 -0.22;
layer straight $stlBC 1 [expr 310*$revmilli] 0 -0.22 0 -0.22;
}

section Aggregator $Agseccolumn111 $secTorsioncolumn111 T -section $seccolumn111
set elecolumn111 736
element nonlinearBeamColumn $elecolumn111 $n111 $n112 $NIntPCol $Agseccolumn111 $IDcolumnTrans

##---------------define the load--------------------------
set N [expr -713.7*1000]
set Mx [expr 106.2*1000*$factor]
pattern Plain 20100901 Constant { load $n112 0 0 $N $Mx 0 0 }

##---------------recorder cmds------------------------------

recorder Element -file eps_y.txt -time -ele $elecolumn111 section 1 fiber 0 0.22 $stlBC stressStrain
recorder Element -file eps_cu.txt -time -ele $elecolumn111 section 1 fiber 0 -0.25 $conBC stressStrain

#---------------------------ANALYSIS----------------------------------------

system UmfPack;
constraints Plain;
test NormDispIncr 1.0e-5 50 1;
algorithm Newton ;
numberer RCM;
integrator LoadControl 0.1 0 0.1 0.1 ;
analysis Static
analyze 50

BUT ,after the analysis ,with N=-713.7kN, M=0kN*m,
eps_y: -0.000162627

eps_cu: -0.000162627

with N=-713.7kN, M=106.2kN*m,

eps_y: -0.00041403

eps_cu: 9.63523e-005

(eps_y: strain of steel under tesion;
eps_cu: strain of concrete under compresure)

The problem is the strain in the compressive zone of the section under Mx and N has a biger strain than that under N load only.
I suspect there must be wrong in the definition of Mx direction,
but after times check, I found none.
Could you give me some suggestions and help?
THANKS YOU VERY MUCH IN PRE