Topic Options
#70744 - 01/11/18 08:19 AM "Time History Time Step" parameter - Water Hammer
RodrigoVicente Offline
Member

Registered: 10/12/17
Posts: 14
Loc: Argentina
Hello,
We are simulating a case of water hammer, and we have the Time History with a time step of 1s, since it is a long event. In the control parameter "Time History Time Step (ms)" we set 1000 ms, but the results are 6000% above the limit. Being that the event already happened several times without consequences. By reducing the "Time History Time Step" parameter to the order of 20 ms, the results give tensions well below the admissible ones. The question is: Is it possible that the parameter "Time History Time Step" is smaller than the time step of the "Time History"? What is the criterion to select this parameter?
Thank you very much.

Top
#70745 - 01/11/18 09:05 AM Re: "Time History Time Step" parameter - Water Hammer [Re: RodrigoVicente]
Dave Diehl Offline
Member

Registered: 12/14/99
Posts: 2382
Loc: Houston, TX, USA
Do not confuse the two time history Control Parameters: "Time History Time Step" and "Load Duration" Load Duration is the overall time simulated in the analysis - start to finish. Time History Time Step is the interval between each analysis performed.
I would set my time step based on the highest system natural frequency used in the time history analysis - I wish to have ten time steps (analyses) through this highest mode. In other words - the time step used should be less than [1/(10*highest frequency)].
ALSO - you want a time step that can capture the actual event timing. If my event has changes every 10 ms, my time step should be ten ms. A time step of 1000 ms would not have sufficient "resolution" to show this detail.
My guess is you are running your hammer through several elbows in a rather large system (based on speed of sound and your long time specified). I think a more dependable approach would be to get a good model for a single event (one elbow) first. Once that one is working well for you, then build up the other hits.
_________________________
Dave Diehl

Top
#70748 - 01/11/18 09:58 AM Re: "Time History Time Step" parameter - Water Hammer [Re: Dave Diehl]
RodrigoVicente Offline
Member

Registered: 10/12/17
Posts: 14
Loc: Argentina
Thank you.
Antother question.
Can the parameter "Time History Time Step" be smaller than the time between Time History data?
Thanks

Top
#70749 - 01/11/18 10:39 AM Re: "Time History Time Step" parameter - Water Hammer [Re: RodrigoVicente]
Dave Diehl Offline
Member

Registered: 12/14/99
Posts: 2382
Loc: Houston, TX, USA
Yes, when a time step does not line up with a specific data point, the program will set the load magnitude by interpolation.
_________________________
Dave Diehl

Top
#70750 - 01/11/18 11:10 AM Re: "Time History Time Step" parameter - Water Hammer [Re: Dave Diehl]
RodrigoVicente Offline
Member

Registered: 10/12/17
Posts: 14
Loc: Argentina
Thanks a lot.
Then, I may select this parameter as small as mi computer can process it.

Top
#70752 - 01/11/18 12:15 PM Re: "Time History Time Step" parameter - Water Hammer [Re: RodrigoVicente]
Dave Diehl Offline
Member

Registered: 12/14/99
Posts: 2382
Loc: Houston, TX, USA
Yes, but not too small. You may have to boost your requested mamory.
_________________________
Dave Diehl

Top



Moderator:  Denny_Thomas, uribejl 
Who's Online
0 registered (), 38 Guests and 1 Spider 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)