Please refer to sample model frame_sway1.dat.
Abstract
How to model a wind sway displacement of the pipe rack that moves the structure, while the pipes on the rack are free to slide across the steelwork on pipe shoes or trunnions.
Method
Since the pipe is large i.e. 24", suggest model an offset support (i.e. trunnion) to capture the additional bending and torsion which the support will impose on the pipe.
This involves modeling the trunnion as a Rigid beam (or equivalent beam with same stiffness as the trunnion e.g. If 10" std wt pipe trunnion select table type = "P", Section ID = 10.) from the pipe centerline to shoe contact point on the piperack. Insert a second beam with an anchor at the far end and disconnect the 1st and 2nd beam by renaming point I on beam 2. Then connect beam 1 to 2 using a V-stop with friction and large gap above to capture lift-off.
Insert the pipe rack wind sway displacement at the 2nd beam anchor in the wind case W1.
Note: This wind case will also include the wind applied to the piping segments on the rack.
Note: If effects of shielding are experience by pipes in the center of the rack then ensure these pipes are different segments so they can be "turned off" in the wind loading.
Note: When one trunnion support has been modeled completed with pipe rack sway displacement this can be copied and paste quickly to other support locations on the pipe rack as shown below.
AutoPIPE Modeling
- Place cursor on AG02.
- Insert/Frame, I-point = AG02, Point J = 1, Dy = -600 (mm i.e. distance from pipe centerline to the base of the support shoe), table name = rigid (or equiv. Beam), click ok.
- Place cursor on new frame point 1, Insert/Frame, I-point = 1, enter Point J = 2, Dy = -300 (nominal distance of 1ft or 300mm), table name = rigid, click ok.
- Cursor at point 2, Insert/Anchor. (Rigid)
- Double click to modify frame M2 and change point I from 1 to 1a. This will disconnect the two beams.
- Place cursor on point 1, insert support/ select V-stop and enter gap above pipe = 1000 (i.e. very large gap) and friction = 0.3, change connected to point from Ground to point 1a, click ok. Notice the v-stop looks upside down on the plot but is not a concern.
- Model should be as shown below:
 |
 |
Review Results
Observe the W1 results at Anchor 2 when run static analysis as linear or non-linear, if analyze linear (i.e. no friction or gap effects at the V-stop between beams 1 & 2) the W1 results are zero at anchor 2. Also perform an interactive scan / displacement for the W1 case to observe this total disconnect behavior between beams 1 & 2 for the linear analysis. Whereas in the non-linear case the friction between beams 1 & 2 shows connectivity between them and therefore interaction not seen in the linear case.