pattern UniformExcitation 1 1 -disp

[mohsenVazirizade]

I am trying to use UniformExcitation command and use displacement record rather than acceleration record.
There are two ways for implementing that:

1
timeSeries Path 1 -dt .1 -filePath record/recdis.txt -factor 1
pattern UniformExcitation 1 1 -disp 1

2
set Disp "Series Path 1 -dt .1 -filePath record/recdis.txt -factor 1"
pattern UniformExcitation 1 1 -disp $Disp


However, it turns out that this command just works with -accel. Actually, I did not encounter any errors; nevertheless the results are not valid.
In this regard, is my conclusion correct? or is there something that I neglect in my command and model?
_____________________________________________________________________________
another question is that when I use UniformExcitation with dt=0.1 and the following values as acceleration (a=20t):
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00

the final displacement should be 3.33 (20t^3/3) but I get 3.35. Actually, it turns out that OpenSees uses Trapezoidal (I am not sure about Trapezoidal correct me if I am wrong*) method. In this regard, is there any way to change the default integration type in this command (UniformExcitation )?

* It is stated that MultipleSupport uses Trapezoidal as default, and if MultipleSupport is used rather than UniformExcitation, the final displacement is 3.33. Thus, it seems one of these commands do not use Trapezoidal as default.

[fmk]

it is not an option to uniform excitation ..

http://opensees.berkeley.edu/wiki/index ... on_Pattern

if you don't like Trapezoidal in matlab or something you can do your own integration of the accel record and just provide the result to OpenSees in the form of a displacement record.

[mohsenVazirizade]

Thanks for your consideration.
In this regard, the only option for UniformExcitation and default option for MultipleSupport is Trapezoidal. With this in mind, why the final results, as mentioned earlier (3.35 and 3.33), are not the same?

[fmk]

it is called 'NUMERICAL' integration for a reason .. here is some matlab code to run:

Aend=20;
tEnd=1.
nDiv=[2 4 10 20 100 1000]

for j=1:6,
dx = Aend/nDiv(j);
dt = tEnd/nDiv(j)
a=[0:dx:20]
v(1) = 0;
u(1) = 0.;
for i=2:nDiv(j)+1,
v(i)=v(i-1)+0.5*(a(i-1)+a(i))*dt;
u(i)=u(i-1)+0.5*(v(i-1)+v(i))*dt;
end
dispEnd(j)=u(nDiv(j)+1);
timeStep(j) = dt;
end

timeStep
dispEnd



and here is the result:
timeStep =

0.5000 0.2500 0.1000 0.0500 0.0100 0.0010


dispEnd =

3.7500 3.4375 3.3500 3.3375 3.3335 3.3333

[mohsenVazirizade]

Let me ask you my question in other way.
I understand that UniformExcitation uses Trapezoidal as the only option and MultipleSupport as the default option.
I expected their results to be the same. Therefore, I provided an example. According to the acceleration values and Trapezoidal method I expected 3.35 as the final displacement for both of them; however, this value for UniformExcitation is 3.35 while this value for MultipleSupport is 3.333. Where did this difference come from?