what is the meaning of this warning

Forum for OpenSees users to post questions, comments, etc. on the use of the OpenSees interpreter, OpenSees.exe

Moderators: silvia, selimgunay, Moderators

Post Reply
lkang6
Posts: 36
Joined: Mon Nov 15, 2010 7:09 pm
Location: GEORGIA INSTITUTE OF TECHNOLOGY

what is the meaning of this warning

Post by lkang6 »

Hi,
I am doing analysis of a connection. I use several zero length spring to simulate the connection between beam/column and shear panel.

when I run the model, the warning appears as following:
WARNING BandGenLinLapackSolver::solve() -LAPACK routine returned 86
WARNING NewtonRaphson::solveCurrentStep() -the LinearSysofEqn failed in solve()
StaticAnalysis::analyze() - the Algorithm failed at iteration: 0 with domain at load factor 0

I checked my model, I can not find problem. Is the WARNING has some relation to the springs, which I give a very large value of stiffness. Or the number of spring is too many?(I use 23 springs in this connection)

Could you please tell me what dose the WARNING mean? What kind of problem I may have in the analysis of spring model?

Thank you so much!
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Re: what is the meaning of this warning

Post by vesna »

This warning usually means that you do not have enough constraints in your model so your system of equations can not be solved.
lkang6
Posts: 36
Joined: Mon Nov 15, 2010 7:09 pm
Location: GEORGIA INSTITUTE OF TECHNOLOGY

Re: what is the meaning of this warning

Post by lkang6 »

Thank you so much!

I tried to add more constraints on the nodes, and it woked.

But another problem appeared. In the result file, there are all the files I want, but all the result is zero. I checked my recorder, and checked my loading, but I can not find any problems. Could you please tell me what the problem it might be?

Thank you!
lkang6
Posts: 36
Joined: Mon Nov 15, 2010 7:09 pm
Location: GEORGIA INSTITUTE OF TECHNOLOGY

Re: what is the meaning of this warning

Post by lkang6 »

Hi,
When I run the gravity analysis in OpenSEES,all the results are zero. And then I run the lateral force analysis,there are some warnings as follows:

WARNING-Force BeamColumn2d::update-failed to get compatible element force &deformations for element: 5<dw:<<1.14142>
WARNING-Force BeamColumn2d::update-failed to get compatible element force &deformations for element: 5<dw:<<213933>
Domain::update-domain fialed in update
DisplacementControl::update_model failed to update for new dU
WARNING NewtonRaphson::solveCurrentStep<>-the Integrator failed in update<>
StaticAnalysis::analyze<>-the Algorithm failed at iteration:0 with domain at load factor 0.0373494

Could you please tell me what is the problem?

Thank you so much!
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Re: what is the meaning of this warning

Post by vesna »

If all your results are zero that means that no load or zero load was applied in your model. If you can not solve the problem post the load pattern definition and all analysis objects and I will take a look.

In case of lateral force analysis your element 5 is having problems. It usually means the sections are trashed and the flexibility based element just cannot figure out forces that agree with end deformations. You can try several things:
1) try the -iter option on the element
2) tweak the section properties
3) reduce number of integration points.
4) use a different element type.
5) consider that the element has failed and remove this element from the analysis at this point (you may need to look at ele end forces to see if element is carrying any load here .. if it goes away there will be unbalanced forces which other ele meeting at node are going to have to deal with at next step.)
lkang6
Posts: 36
Joined: Mon Nov 15, 2010 7:09 pm
Location: GEORGIA INSTITUTE OF TECHNOLOGY

Re: what is the meaning of this warning

Post by lkang6 »

Thank you so much!

My loading and analysis input is as follows:

pattern Plain 1 Linear {
load 101 0.0 -10000 0.0
load 102 0.0 -10000 0.0
load 44 0.0 -10000 0.0
}

system BandGeneral
#constraints Plain # Plain constraints can not be used if there is MP-constraint in the model
constraints Transformation
set NstepGravity 10
integrator LoadControl [expr 1/$NstepGravity]
test NormDispIncr 1e-5 50
algorithm Newton
numberer RCM
analysis Static
analyze $NstepGravity
loadConst -time 0.0
puts "Gravity analysis succeed"

#lateral load
set increments 50;
set nodeTag 44;
set dofTag 1;
set fir 0.01;
pattern Plain 2 Linear {
load 44 [expr $p/5] 0.0 0.0
}
integrator DisplacementControl $nodeTag $dofTag $fir
analyze $increments

In my model, I used lots of springs to simulate the connection. I don't know if this is a reason that causes analysis failure. Element 5 is a nonlinearbeamcolumn element as a column, which is connected by several springs to the shear panel.
lkang6
Posts: 36
Joined: Mon Nov 15, 2010 7:09 pm
Location: GEORGIA INSTITUTE OF TECHNOLOGY

Re: what is the meaning of this warning

Post by lkang6 »

Hi Vesna,

I checkde my model and compare it to the examples, and I found that:
if I define step like this:
set NstepGravity 10
integrator LoadControl [expr 1/$NstepGravity]
analyze $NstepGravity
then, all the results for gravity are zeros.

But if I define step like this:
integrator LoadControl 0.1
analyze 10
then, I can get the results. But in the results there are some errors:
0.1 -0.000119495 6.33312 2.31801e-015....
0.2 1.#QNAN 1.#QNAN 1.#QNAN ...
0.3 1.#QNAN 1.#QNAN 1.#QNAN ...
I ran the analysis again, but this problem remained. What should I do to avoid this?

Another problem is that after I change the input as above(so I can get the results of gravity analysis), the WARNING changed to:
WARNING DisplacementControl::newStep()dUahat is zero--zero reference displacement at control node DOF
StaticAnalysis::analyze()-the Integratoe failed at iteration: 0 with domain at load factor 0
What is this mean?

I tried to use "-iter 10 10e-10" in "element", but it didn't work. The integration point of each element is 3, so I think this is not a problem.

Thanks!!
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Re: what is the meaning of this warning

Post by vesna »

If you make this change:

set NstepGravity 10
integrator LoadControl [expr 1.0/$NstepGravity]
analyze $NstepGravity

it will work. The reason for this is that for tcl 1/10 is 0 but 1.0/10 is 0.1. You have to be careful when using integers.

QNAN means "not a number". Most probably you have an error somewhere in your model. This usually happens if something is divided by zero. Look at the nodes, elements and materials and see where it happens.

The new message you get refers to your DisplacementControl integrator. It looks like you are defining your displacement increment to be zero but it has to be non-zero.
lkang6
Posts: 36
Joined: Mon Nov 15, 2010 7:09 pm
Location: GEORGIA INSTITUTE OF TECHNOLOGY

Re: what is the meaning of this warning

Post by lkang6 »

Hi,

In the connection model, I use "nonlinearBeamColumn" type for the column and beam, the warning is "LAPACK routine returned 38", which means "not enought constraints in the model". Then I try "elasticBeamColumn" type for the column and beam, and the warning is"failed to converge".

So my question is: are there not enought constraints in the model? Or there are some problems about my element (material) definition?
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Re: what is the meaning of this warning

Post by vesna »

Your model may have multiple problems. From the information you gave me I can not identify the problem.

It is always better to start with the simpler model and to gradually add complexity to it. This way it is easier to test for errors.
lkang6
Posts: 36
Joined: Mon Nov 15, 2010 7:09 pm
Location: GEORGIA INSTITUTE OF TECHNOLOGY

Re: what is the meaning of this warning

Post by lkang6 »

Hi Vesna,

I made a simpler connection model, with two beams and two columns and one shear panel, the configeration is like in the Example BeamColumnJoint Element. I use one set of springs(two translational, one shear ) to connect the beam to shear panel, and another set for column and shear panel. Shear panel is composed by 8 elasticBeamcolumn element, with one rotational spring in one corner ,and three pins in other three corners. In order to check the model, I use elasticBeamcolumn element for the beam, column and give the spring a very large stiffness.
But the warning is LAPACK routine returned....I checked the modal for many times, and I can not find the problem.
Could you please take a look at my model and tell me where is the problem? Thank you so much!!


#create the ModelBuilder object
model BasicBuilder -ndm 2 -ndf 3
set dataDir Data_no_1; # set up name of data directory
file mkdir $dataDir; # create data directory

#dimension set, in SI unit::m,kg,second
#column
set lcol 1.050; #the length of the column
set bc 0.120 ; #the width of the column
set dc 0.120 ; #the depth of the column
set tf 0.00346 ; #the thickness of the column flange
set tw 0.00346; #the thickness of the column web
puts "set column finished"
#beam
set lbeam 1.500; #the length of the beam
set db 0.160; #the depth of the H-beam
set bf 0.080; #the width of the flange
set tbf 0.00353; #the thickness of the H-beam flange
set tbw 0.00353; #the thickness of the H-beam web
set density 7800 ; #the density of the beam
set gravity [expr ($db*$bf-($bf-$tbw)*($db-2*$tbf))*$lbeam*$density*9.8/2] ;
puts "set beam finished"
#shear panel zone
set hz $db;
set bz $dc;

#nodes
node 1 [expr -$lbeam/2] 0
node 2 [expr -$lbeam/4] 0
node 3 [expr -$bz/2] 0
node 4 [expr $bz/2] 0
node 5 [expr $lbeam/4] 0
node 6 [expr $lbeam/2] 0
node 7 0 [expr -$lcol/2]
node 8 0 [expr -$hz/2]
node 9 0 [expr $hz/2]
node 10 0 [expr $lcol/2]
node 101 [expr -$bz/2] [expr -$hz/2]
node 102 [expr -$bz/2] [expr -$hz/2]
node 103 [expr -$bz/2] [expr $hz/2]
node 104 [expr -$bz/2] [expr $hz/2]
node 105 [expr -$bz/2] [expr $hz/2]
node 106 [expr -$bz/2] [expr $hz/2]
node 107 [expr $bz/2] [expr $hz/2]
node 108 [expr $bz/2] [expr $hz/2]
node 109 [expr $bz/2] [expr $hz/2]
node 110 [expr $bz/2] [expr $hz/2]
node 111 [expr $bz/2] [expr -$hz/2]
node 112 [expr $bz/2] [expr -$hz/2]
node 113 [expr $bz/2] [expr -$hz/2]
node 114 [expr $bz/2] [expr -$hz/2]
node 115 [expr -$bz/2] [expr -$hz/2]
node 116 [expr -$bz/2] [expr -$hz/2]
node 117 [expr -$bz/2] 0
node 118 0 [expr $hz/2]
node 119 [expr $bz/2] 0
node 120 0 [expr -$hz/2]

#element
#beam
geomTransf Linear 1
element elasticBeamColumn 1 1 2 1 1.0e12 1.0e15 1
element elasticBeamColumn 2 2 3 1 1.0e12 1.0e15 1
element elasticBeamColumn 3 4 5 1 1.0e12 1.0e15 1
element elasticBeamColumn 4 5 6 1 1.0e12 1.0e15 1
#column
geomTransf Linear 2
element elasticBeamColumn 5 7 8 1 1.0e12 1.0e15 2
element elasticBeamColumn 6 9 10 1 1.0e12 1.0e15 2
#rigid link in the connection
geomTransf Linear 3
element elasticBeamColumn 101 101 3 1 1.0e13 1.0e16 3
element elasticBeamColumn 102 102 117 1 1.0e13 1.0e16 3
element elasticBeamColumn 103 103 3 1 1.0e13 1.0e16 3
element elasticBeamColumn 104 104 117 1 1.0e13 1.0e16 3
element elasticBeamColumn 105 105 9 1 1.0e13 1.0e16 3
element elasticBeamColumn 106 106 118 1 1.0e13 1.0e16 3
element elasticBeamColumn 107 107 9 1 1.0e13 1.0e16 3
element elasticBeamColumn 108 108 118 1 1.0e13 1.0e16 3
element elasticBeamColumn 109 109 4 1 1.0e13 1.0e16 3
element elasticBeamColumn 110 110 119 1 1.0e13 1.0e16 3
element elasticBeamColumn 111 111 4 1 1.0e13 1.0e16 3
element elasticBeamColumn 112 112 119 1 1.0e13 1.0e16 3
element elasticBeamColumn 113 113 8 1 1.0e13 1.0e16 3
element elasticBeamColumn 114 114 120 1 1.0e13 1.0e16 3
element elasticBeamColumn 115 115 8 1 1.0e13 1.0e16 3
element elasticBeamColumn 116 116 120 1 1.0e13 1.0e16 3
#spring
set Es200 1.0e12;
set epsP200 0.5;
set epsN200 0.5;
uniaxialMaterial ElasticPP 200 $Es200 $epsP200 $epsN200
element zeroLength 201 101 102 -mat 200 -dir 1
element zeroLength 202 3 117 -mat 200 -dir 2
element zeroLength 203 103 104 -mat 200 -dir 1
element zeroLength 205 105 106 -mat 200 -dir 2
element zeroLength 206 9 118 -mat 200 -dir 1
element zeroLength 207 107 108 -mat 200 -dir 2
element zeroLength 209 109 110 -mat 200 -dir 1
element zeroLength 210 4 119 -mat 200 -dir 2
element zeroLength 211 111 112 -mat 200 -dir 1
element zeroLength 213 113 114 -mat 200 -dir 2
element zeroLength 214 8 120 -mat 200 -dir 1
element zeroLength 215 115 116 -mat 200 -dir 2
equalDOF 101 102 2 3
equalDOF 3 117 1 3
equalDOF 103 104 2 3
equalDOF 105 106 1 3
equalDOF 9 118 2 3
equalDOF 107 108 1 3
equalDOF 109 110 2 3
equalDOF 4 119 1 3
equalDOF 111 112 2 3
equalDOF 113 114 1 3
equalDOF 8 120 2 3
equalDOF 115 116 1 3

#shear panel
element zeroLength 300 108 110 -mat 200 -dir 6
equalDOF 102 116 1 2
equalDOF 104 106 1 2
equalDOF 108 110 1 2
equalDOF 112 114 1 2

#boundary conditions
fix 1 0 1 0
fix 6 0 1 0
fix 7 1 1 0

recorder Element -file $dataDir/GF.out -time -ele 1 2 3 4 5 6 globalForce;
recorder Element -file $dataDir/E300F.out -time -ele 300 force;
recorder Element -file $dataDir/E300D.out -time -ele 300 deformation;
recorder Element -file $dataDir/E201F.out -time -ele 201 force;
recorder Element -file $dataDir/E201D.out -time -ele 201 deformation;

#load condition
pattern Plain 1 Linear {
load 2 0.0 -1000 0.0
load 5 0.0 -1000 0.0
load 10 0.0 -2000 0.0
}

system BandGeneral
#constraints Plain # Plain constraints can not be used if there is MP constraint in the model
constraints Transformation
integrator LoadControl 0.1
test NormDispIncr 1e-5 5
algorithm Newton
numberer RCM
analysis Static
analyze 10
loadConst -time 0.0
puts "Vertical force done"

#lateral load
pattern Plain 2 Linear {
load 10 1.0 0.0 0.0
}
integrator DisplacementControl 10 1 0.01
analyze 5
puts "Lateral load done"
Post Reply