Hi,
- 2 node sets : Node set 1 & Node set 2
- 1 frequency index : 1- Freq Index 3 : Applied 10 Hz
- 2 node sets : Node set 1 & Node set 2
- 2 frequency index : 1- Freq Index 3 & 2-Freq Index 2 : But I applied same frequency for comparing.
Hi Duckjae,
I checked your model and agree with what you found.
Basically, in frequency response, if only one frequency index is defined, the software will calculate three portions of response: 1. in-phase response; 2. out-of-phase response (related to damping); and 3. SRSS of these two responses. You can check them under "Results Options" -> "Response Type".
However, if more than one frequency index are defined, the software will assume no pure in-phase or out-of-phase response are available since the two exciting frequencies may not be the same. Although another SRSS procedure will be performed to generate the final result, it will be operated on different operands and it will yield a little bit different results.
I will let you know more details shortly...
Hi Duckjae,
To continie with the previous reply. In Autodesk Simulation Freqeuncy Response Analysis, we use a decoupling method to solve structure response. Basically, this method first decouples the system based on modal results from a non-damping modal analysis. After the decoupling, we can solve the response based on each decoupled single DOF system, and then combined these individual response to get the response for the real system.
Assume the real system is consisted by DOFs x1,x2,.....xN, and the decoupled system is consisted by DOFs y1,y2,.....yN.
=======
If we have two exciting frequency index (f1 and f2), no matter the input values are same or not, such as your scenario 6, we will first calculate the response of y. Take y1 as an example, we name the contribution from f1 as y1_f1, and from f2 as y1_f2. Then the response of y1, name it R_y1 will be:
R_y1=sqrt((y1_f1)^2+y1_f2)^2). This is a first SRSS we performed.
Then the real response on x1 will be:
R_x1=sqrt((c1*R_y1)^2+(c2*R_y2)^2+...+((cN*R_yN)^2). This is the second SRSS we performed.
where c1,c2...cN are the cofficients calculated based on the non-damping modal analysis.
=======
However, if we have only one input frequency index, such as your senario 7, the above two SRSS are not needed since the response is composed by the in-phase response and out-of-phase response, each of them can be calculated individually. We can further calculate the SRSS based on these two responses.
R_x1_SRSS=sqrt((R_x1_in-phase)^2+(R_x1_out-of-phase)^2)
I agree that when two input exciting frequncy indexes are the same in the first senario, we should use the second senario to solve it. I will create an issue to log it. Hope the above explanation helps you to understand the whole idea we implemented.
Best regards!