Аs ASME 31E said, I can use two variants for choosing seismic load direction: one is three directional excitations such as plant east-west, north-south and vertical, with the combination of square-root sum of the squares (SRSS) or two directional design approach of based on the envelope of SRSS of east-west plus vertical and north-south plus vertical seismic loadings. According to this definitions, the former is as follows if X is east-west direction and Y is vertical with the maximum design acceleration being 0.3 in horizontal and a half in vertical: E1: -0.212 0.15 -0.212. E2: 0.212 -0.15 0.212. Also if the gap or lift-off is to be considered in piping system, you should add positive and negative direction for each X, Y and Z direction with normal operating condition such as W+T1+P1, that means W+T1+P1+E1. Do I correct understand?
Hello Mykola,
Please see the following file for recommended non-linear seismic load combinations.
communities.bentley.com/.../269425.aspx
You can modify these recommendations as needed to meet your requirements.
Regards,
Mike Dattilio
Bentley Technical Support Analyst
Dear Mike,
Thank you for your response. I have question regarding option 1 (in pdf file "Non-linear Analysis - Earthquake")
Will I get the same result in case I enter values for E1 0.6 0.4 0.6?Actually this values (E1) would be combined by SRSS method (are combined as load vectors algebraically). Also when you use combination in your example GR+P+SEIS, what would be direction of EARTHMAX load?
Hi Mike and other forum members,
B31E allows two options for applying seismic loading
(1) a three-directional excitation, east–west plus north–south plus vertical, combined by square-root sum of the squares (SRSS)
(2) a two-directional design approach based on the envelope of the SRSS of the east–west plus vertical and north–south plus vertical seismic loading
The results will not be the same for either option. (1) is effectively including consideration of a resultant N-S plus E-W earthquake (with a magnitude greater than a N-S or E-W earthquake alone). I think this is given only as an option because it is easier to set up.
Here is my suggestion to apply loading in accordance with (2), if Y is vertical axis. (Updated 19-Oct-16)
First set static earthquake loading (inertial loading):
E1 +x direction
E2 -x direction
E3 +z direction
E4 -z direction
E5 +y direction
Apply the support displacements (relative anchor motion) in load cases
U1, U2 for X-direction displacements
U3, U4 for Z-direction displacements
Run the model, then create user defined Code Combination as follows:
MAX E(X) taking the ABS MAX of E1 and E2
MAX E(Z) taking the ABS MAX of E3 and E4
MAX U(X) taking the ABS MAX of U1 and U2
MAX U(Z) taking the ABS MAX of U3 and U4
It is necessary in accordance with B31E to calculate the SRSS of the east–west plus vertical, and the SRSS north–south plus vertical seismic loading. It is also necessary to combine the forces and moments due to inertial forces and seismic anchor movement. This can be accomplished in two more user defined Code Combinations:
SRSS(XY) taking the SRSS of MAX E(X) and MAX U(X) and E5
SRSS(YZ) taking the SRSS of MAX E(Z) and MAX U(Z) and E5
Next calculate the “envelope of the SRSS of the east–west plus vertical and north–south plus vertical seismic loading”, by setting up code combination:
ENVELOPE taking the MAX of SRSS(XY) and SRSS(YZ)
Finally it is necessary to consider these in combination with sustained loading, so create a final user defined Code Combinations:
SUS + ENVELOPE taking the ABS SUM of SUS and ENVELOPE.
All these user Code Combinations are created as Occasional category. In the Load Combinations window, manually enter the appropriate Allowable Stress as defined in B31E.
It is necessary to run the model twice to find the highest stress, once with E5 defined as a positive Y direction earthquake, and then run the model with E5 defined in the negative Y direction.
I believe the above method correctly applies the loading combinations in accordance with ASME B31E, and would greatly appreciate any comments or feedback ?
Regards
Fergus