Acceleration as recorded by the uniformExcitation pattern

[mlw]

Hi all,

I have a query about the uniformExcitation pattern. I am modelling a pile surrounded by soil and running a dynamic analysis with the uniformExcitation pattern. The command is implemented by the following:
pattern UniformExcitation 1 1 -accel "Series -fileTime acc_time.dat -filePath acc_value.dat -factor [expr 1.0*$g]"
The base of the model may be partially fixed by dashpots (with the soil nodes as masters when fixed to the dashpot nodes using the equalDOF command) or completely fixed using the 'fix' command.

What I am finding is that the accelerations within the model are the magnitude that would be expected, but the accelerations at the base of the model are negligible (10^-13) where soil nodes are attached to dashpots or fixed. The uniformExcitation pattern should give the input acceleration at the base of the model.
I note on the Wiki it mentions that when no timeseries is provided as an input to the recorder then accelerations are relative values.
The recorder is created by the following command and is created in the prior analysis (i.e. the recorder command is followed by the wipeAnalysis and loadConst -time 0 commands before the uniformExcitation pattern is created)
eval "recorder Node -file accelX.out -time -node $nodeList -dof 1 accel"

1) Am I correct that the uniformExcitation pattern should give the input acceleration when recorded at the base of the model (regardless of the base fixity)? And is there anything I'm not doing/should be doing differently to get this?
2) Is my recorder recording relative values in that I should add the input acceleration for each respective time step to get the true acceleration?

Thank you for your time

mlw

[vesna]

Yes, with the command you use you record the relative acceleration at a node. If you want to get absolute acceleration look at the last example on this page:
http://opensees.berkeley.edu/wiki/index ... e_Recorder

[mlw]

So can I just clarify. Say if I have an input motion with a max acceleration of 0.3g, then this 0.3g would still be applied to the nodes at the base of the model even though the recorder is not recording it?

Thank you for your time

mlw

[vesna]

Yes, 0.3g will be applied to the nodes at the base. Depending on which recorder you use you will see either zero acceleration at the fixed nodes or ground motion acceleration at the same nodes.