node absolute acceleration

[Marco Faggella 2]

frank, silvia

was wondering if there was a way to record nodes absolute acceleration

thanks

[fmk]

if you want to record the absolute you have to do a multisupport exciataion. the nodes have no idea if they are rel or abs.

[laura83]

ciao marco
ho dei dubbi su opensees è possibile contattarti ad un indirizzo email?

[Gang Wang]

So if use uniform excitation, the absolute acceleration = base input in uniform excitation + recorded acceleration, right? I want to clearify it is a plus not a minus.

[Marco Faggella 2]

frank,

could you give me some more details about this trick that you suggested? you mean that I should "fake" a regular time history doing a multiple support analysis with the same motion at the different supports ?

thanks

[silvia]

don't do what frank suggested, it is too hard. you'd have to convert your input acceleration into displacements at those nodes, then the output would all be in absolute values.
if you want to get the absolute acceleration at the nodes, just take the relative acceleration you get from the recorders and add to it the acceleration you input into the supports -- of course this requires post-processing.
The only think you need to make sure is to match the time. In the analysis you may be reducing time steps... but just interpolate the acceleration linearly -- that's what you'd do if you were to do anything else anyways.