about the UniformExcitation command

[baozai]

Hello, doesn't the UniformExcitation command need the $nodetag? In the official guide it is written in the form:

pattern UniformExcitation $patternTag $dir -accel $tsTag <-vel0 $ver0>

Thank u for your help

[bcoryell]

Uniform Excitation applies the excitation to all nodes in your model. Therefore no node tag is required. See imposed motion / ground motion for excitation on individual nodes.

[baozai]

Thank you. If I want to simulate a shaking table test, uniformexcitation isn't adaptable? Am I right?

[oleviuqserh]

More precisely... to all dofs having mass.

[baozai]

oleviuqserh ,thank you. The most problem is that the acc curve isn't right, it looks like a compressed sin-curve, but the vel and the disp curve is better. I don't know why...

[oleviuqserh]

Would you explain me what are you trying to model?

In my last post, I forgot to say that the uniform excitation is applied to all dofs that have mass and point to $dir.

[baozai]

I'm sorry I was out just now. En, I want to model a 3-dimensional prestressed frame. Which is 5m long and has 2 bays at that direction, 2m high which includes 3 stories , 900cm in the other horizontal direction. And I want to simulate the reaponses of the model in its long direction(5m direction).

After the calculation, the disp response and vel response looks better than its accel response and I don't know why....

the frequencies of the sturcture is also close to the experiment.

[baozai]

oleviuqserh, you mean that if I give the mass to all the nodes of the structure, all the nodes of the structure will be stimulated? I think it didn't look like an structure dynamic response during an earthquake simulation and what does it used for? First of all, I want to find the reason for my model, the accel of one node is :
279.452
-550.68
1144.55
-1692.32
1201.32
-2072.26
2139.55
-1941.01
2503.59
-2241.72
2655.19
-2541.74
2521.09
-2780.89
2867.39
-2498.97
2941.69
-2897.83
2578.5
-3101.91
2759.1
-2884.01
2918.8
-2869.95
2931.76
-2923.93
2833.78
-3016.88
2960.71
-2709.08
3191.92
-2826.57
2844.13
-3065.62
2880.82
-2913.79
2972.19
-2925.7
2921.52
-2992.72
2828.35
-3070.38
2875.08

and this is only the first 0.34s, the following response is just similiar to it.

[bcoryell]

My accel results have issues too... But the rms accel for 1 sec was pretty close to the experiment, so I chose to focus on that. Checked with STAAD, RISA, and ANSYS, and displacement / velocity graphs were close but again accel way off. Again the rms for 1 sec was similar tho..... Weird stuff.

[baozai]

bcoryell, Thank you for your advice, but how did you calculate the rms accel? is it the square root of the accel in the perpendicular direction (x and z)?

[oleviuqserh]

1. Yes. 2. ---. Would you post now the script and the basic units you are using on the model to have a better look? Most likely, you are getting relative accelerations from the outputs. That would explain the discrepancies.

[baozai]

Hello,oleviuqserh. Glad to see you again. The unit of the accel is cm/s/s. If you draw the data I put on the board, you can obviously see that it doesn't look like an earthquake stimulation at all. And I just did a similar stimulation to a 2-dimentional model, and the accel is better. Is there any differences in the dynamic input between the 2-dimentional and the 3-dimentional model in OPENSEES?

[baozai]

bcoryell wrote:
> My accel results have issues too...

Hello, was your model in OPENSEES is 3-dimensional? I found that if I did the stimulation to a 2-dimentional model, the accel was well, but if I did it to a 3-dimentional model, the results was like what I pasted above.