sp command for applying displacement time-history

[nithinvl]

Hi

I am using sp command to apply displacement timehistory to the top nodes of a single bay single storey frame (Bottom nodes (fixed) - 1 and 2 , Top nodes (free) - 3 and 4). After running the analysis, i get the displacements of nodes 3 and 4 same as that in the file "displacement.dat". But, the acceleration in x-direction for these nodes are showing zero whereas some acceleration values are showing in y-direction for these nodes. Shouldn't there be some acceleration values in x-direction ? Am I doing it correctly ?
Following are the lines of the code specifying the sp constraint and displacement time-history.

constraints Transformation
numberer RCM;
system BandGeneral;
test NormUnbalance 1.0e-6 10
algorithm Newton;
integrator Newmark 0.5 0.25;
analysis Transient

set xDamp 0.05; # 2% damping ratio
set omega [expr pow($lambda,0.5)];
set alphaM 0; # M-prop. damping; D = alphaM*M
set betaKcurr 0; # K-proportional damping; +beatKcurr*KCurrent
set betaKcomm 0.0;
set betaKinit [expr 2.*$xDamp/($omega)];
rayleigh $alphaM $betaKcurr $betaKinit $betaKcomm;

set dt1 0.02
set DtAnalysis [expr $dt1/1]
set TmaxAnalysis [expr 5000*$dt1]
timeSeries Path $accelseries_timetag -dt $dt1 -filePath "displacement.dat" -factor 1.0
pattern Plain $IDloadTag $accelseries_timetag {
sp 3 1 1.0
sp 4 1 1.0
}

Kindly help me.

Regards
Nithin.

[selimgunay]

To develop accelerations you need to either apply an acceleration history from the bottom or force history from the top. Otherwise, it is static.

[nithinvl]

Thanks for your reply.

I have used "analysis Transient" and displacement as a function of time. So shouldn't the analysis be dynamic and hence the nodes show some acceleration ?
Also, if it is static, then why the top nodes are showing acceleration in y-direction ?
Could you explain.

Regards
Nithin.

[selimgunay]

I cannot think of a proper way to define the right hand side of the equation of motion when the external excitation is a displacement history at a node. Why don't you differentiate the displacements and either apply the acceleration history to the nodes or multiply the accelerations with the mass and apply the force history?

[Jhno]

Use Multi-Support Excitation Pattern instead (http://opensees.berkeley.edu/wiki/index ... on_Pattern).
You could also consider to apply the SP command to the node 1 and 2 (first remove the fix constraint in DOF 1 for the node).
Make sure the top node have mass also.

[fmk]

there is no acceleration at the nodes as the constraint handler just sets the displacement values .. as a consequence theses dof do not appear as unknowns to be solved for .. if you actually need accel at these nodes (in presence of damping only the vel terms will effect the response) use the Penalty method.

[margo]

fmk wrote:
> there is no acceleration at the nodes as the constraint handler just sets
> the displacement values .. as a consequence theses dof do not appear as
> unknowns to be solved for .. if you actually need accel at these nodes (in
> presence of damping only the vel terms will effect the response) use the
> Penalty method.

Dear Frank
I am trying to apply a dynamic displacement time history to a node (which is not constrained), I came accross this comment of yours and it would really help me if you would elaborate it as to how we can get acceleration.
Thank you for your time.
Kind regards,
Margo

[fmk]

use the penalty method .. the dof's will appear in the equations of motion and the integrator will update vl and accel given the change in displacements.