Topic Options
#70903 - 01/29/18 03:10 AM DLF calculation vs Dynamic results
pooria1978 Offline
Member

Registered: 02/25/08
Posts: 264
Loc: Netherlands

I have made a "test" model where there is only a vertical pipe anchored in the bottom. the pipe column is properly "meshed" (it is a 10inch pipe and every 2m there is a node break)
In the dynamic input, I applied 15000N in one direction horizontally at the free end of the pipe with the time duration of 1ms only (1ms rise time, 1ms duration in the peak and 1ms drop time)
in the dynamic output, I see that the reaction force on the anchor is 5123N which if I divide it by the 15000 (original force) it will give me a DLF of 0.3415.
this is the part that baffles me; reading from CII output, wn (the radial frequency in which the system gets the highest contribution from the force) is 283.5 rad/sec. from the spectrum that I made; radial frequency would be 2094 rad/sec. if I input these values into the DLF calculator formula as per the attachment, the DLF would be 0.019! this means the reaction force shall be only 0.019x15000=285N
I wonder if anyone could explain this difference.


Attachments
Capture.JPG



Top
#70907 - 01/29/18 09:46 AM Re: DLF calculation vs Dynamic results [Re: pooria1978]
Bob Zimmerman Offline
Member

Registered: 12/29/99
Posts: 197
Loc: Houston,TX,USA
option 1- Try a quick check by ensuring the "MISSING MASS" option is activated. This accounts for model frequencies beyond your "cutoff frequency".

Option 2- The wf is 2094 rad/s which is 36.5 hz which is greater than the out of the box 33 hz cutoff frequency. Increase the cutoff freq to at least 40 hz or preferably > 2*wf or ~80 hz to capture the applied frequency data.
_________________________
Bob Zimmerman, P.E.
Vice President of The Piping Stress International Association (The PSI)

Top
#70908 - 01/29/18 09:47 AM Re: DLF calculation vs Dynamic results [Re: pooria1978]
Dave Diehl Offline
Member

Registered: 12/14/99
Posts: 2382
Loc: Houston, TX, USA
The equation you have selected is used to estimate the amplification associated with a harmonic load (omega f) in a single degree of freedom system. Your load appears to be a one-time hit; this would not be harmonic - it is an impulse.
I pushed your event (1ms rise, 1ms flat, 1ms fall) through the program's spectrum generator where a time history is broken down into a frequency response spectrum. The resulting DLF at 45 Hz (283 rad/sec) is about 0.56; not the 0.019 produced by your harmonic load equation.
So why doesn't your calculated "DLF" of 0.3415 line up with this 0.56? Keep in mind the point about the "single degree of freedom system". A single degree of freedom system would have all the mass of the system participating in this single (and only) response. Your piping system is not a single degree of freedom system. Your system has several modes of vibration - the textbook modes of a simple cantilever. The great thing about the response spectrum method is that each mode of vibration is independent of (orthogonal to) all the other modes of vibration and each mode can be evaluated alone. (The total system response being the sum of these modal contributions.) But you must keep in mind that only a portion of each node's mass is assigned to each mode of vibration. If you wish to see the TOTAL response to the impulse, you would have to include ALL modes of vibration in the analysis. Fortunately, high frequency modes respond in a rigid manner and these higher responses can be estimated with a static load - BUT only for the remaining nodal masses (the node's mass not associated with the lower modes). This is that missing mass term in the dynamic Control Parameters.
_________________________
Dave Diehl

Top



Moderator:  Denny_Thomas, uribejl 
Who's Online
0 registered (), 40 Guests and 3 Spiders online.
Key: Admin, Global Mod, Mod
May
Su M Tu W Th F Sa
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31
Forum Stats
12065 Members
14 Forums
16973 Topics
75151 Posts

Max Online: 303 @ 01/28/20 11:58 PM
Top Posters (30 Days)