eigen analysis for mass spring system

[huzb]

hi~
I encounter the following problem during eigen analysis

ArpackSolver::Error with _saupd info = -3
NCV must be greater than NEV and less than or equal to N.
WARNING DirectIntegrationAnalysis::eigen() - EigenSOE failed in solve()

code
wipe all;
model BasicBuilder -ndm 2 -ndf 3
node 1 0. 0.
node 2 0. 6.
node 3 0. 10.
node 4 0. 14.
node 5 0. 18.

mass 2 383576 383576 383576
mass 3 376438 376438 376438
mass 4 374062 374062 374062
mass 5 499070 499070 499070

fix 1 1 1 1
fix 2 1 0 1
fix 3 1 0 1
fix 4 1 0 1
fix 5 1 0 1

uniaxialMaterial MultiLinear 1 0 0 0.037 2071617 0.073 2071617
uniaxialMaterial MultiLinear 2 0 0 0.024 2262560 0.049 2262560
uniaxialMaterial MultiLinear 3 0 0 0.024 1810243 0.049 1810243
uniaxialMaterial MultiLinear 4 0 0 0.024 1560201 0.049 1560201

element twoNodeLink 1 1 2 -mat 1 -dir 2
element twoNodeLink 2 2 3 -mat 2 -dir 2
element twoNodeLink 3 3 4 -mat 3 -dir 2
element twoNodeLink 4 4 5 -mat 4 -dir 2

puts "model analysis"
set numModes 4
set lambda [eigen $numModes]
set period "Periods.txt"
set Periods [open $period "w"]
puts $Periods " $lambda"
close $Periods
record

[Jhno]

I tried your code. I went for a single analysis steps and it failed.
So the model is currently ill defined and this is why your eigen analysis also fails.

I replaced your element with elasticBeamColumn and it worked.
So you have a problem with either your material or your elements.

I replaced your material only (I used eleastic uniaxial material) and kept the element and it failed again.
This make me think your elements are illdefined thought I am not familiar (never used) the twoNodeLink.
Hope it helps.

[huzb]

Dear, Jhno
Thank you for your inspiring answer.
I replaced uniaxialMaterial MultiLinear with uniaxialMaterial Steel01 and it worked.
But I can only obtaine the first three natural periods of the structure. Actually it should have 4 periods.

code:
wipe all
model BasicBuilder -ndm 2 -ndf 3
node 1 0. 0.
node 2 0. 6.
node 3 0. 10.
node 4 0. 14.
node 5 0. 18.

mass 2 383576 383576 383576
mass 3 376438 376438 376438
mass 4 374062 374062 374062
mass 5 499070 499070 499070

fix 1 1 1 1
fix 2 1 0 1
fix 3 1 0 1
fix 4 1 0 1
fix 5 1 0 1

uniaxialMaterial Steel01 1 2071617 56670000 0.001
uniaxialMaterial Steel01 2 2262560 92840000 0.001
uniaxialMaterial Steel01 3 1810243 74280000 0.001
uniaxialMaterial Steel01 4 1560201 64020000 0.001

element twoNodeLink 1 1 2 -mat 1 1 1 -dir 1 2 3
element twoNodeLink 2 2 3 -mat 2 2 2 -dir 1 2 3
element twoNodeLink 3 3 4 -mat 3 3 3 -dir 1 2 3
element twoNodeLink 4 4 5 -mat 4 4 4 -dir 1 2 3

puts "model analysis"
set numModes 3
set lambda [eigen $numModes]
set period "Periods.txt"
set Periods [open $period "w"]
puts $Periods " $lambda"
close $Periods
record

[Jhno]

Use the following command :
eigen fullGenLapack 4

[selimgunay]

Or you can change the parameter numNodes to 4 in your script.

set numModes 4

[huzb]

Dear, Jhno
Thank you very much.

[huzb]

What is the difference between 'localDisplacement' and 'basicDisplacement'? Does the former one mean the relative displacement between the two nodes of the link element?

[Jhno]

My guess would be that localdisplacement are the displacements related to the local axes whilst basicdisplacement are those in the general axes. You could easily try to figure it out by doing a little code with a single element you know the answer then check the results you obtain whit either of them.
Good luck.

[huzb]

Thank you very much Jhno.
Do you know how to set the P-Delta for Two Node Link Element or is there any document such as user's manual to introduce it. I do not understand clealy the instruction on the website “http://opensees.berkeley.edu/wiki/index ... nk_Element”.

[selimgunay]

Please contact the developer of the element indicated in the bottom of the wiki page.

[huzb]

ok, thank you