Hi,
I am modelling a simple system to compare the results to some codes that I have. But, despite having all the configurations similar to codes, I cannot get similar pressure fluctuation. Could you please have a look and let me know the reason?
Thanks in advance.
Regards,
Example 1.xlsx8233.Example 1.zip
As Scott mentioned, the calculation methodology used to generate your "code" may be different from HAMMER. HAMMER uses the Method of Characteristics, which you can read about in the Help and in the below articles. I suspect that a simplistic assumption was made in the results you are comparing to.
Numerical method used for transient simulations in HAMMER
Method of Characteristics vs. Wave Plan Method
Jesse DringoliTechnical Support Manager, OpenFlowsBentley Communities Site AdministratorBentley Systems, Inc.
Thanks for your response. I am using the MOC and I wrote the code in Matlab. I compare my code to some results from other researchers that are very close as you see in the spreadsheet. I want to develop my code more for some research purposes and compare them to results from Hammer. I tried to use the same valve pattern and other specifications. But still, the Hammer result is far from other cases (please see the attached graph too). I am trying to understand how to calibrate the Hammer model with some real or reliable data. I appreciate more comments on this issue.
HAMMER models cavitation (vapor pocket formation) when the pressure reaches vapor pressure (set to -10m by default). That's why your HAMMER pressures stop going down at -10m and instead start to 'flat line'. Then later when the vapor pocket collapses you get a pressure spike.
I assume the other calculations you are using don't attempt to model cavitation, which may explain the difference. As a quick workaround you can change the vapor pressure value in the calculation options to some large value (like -100m).
Answer Verified By: Hiwa
Here is some information from our Wiki to supplement Mal's response about cavitation:
What standard method does HAMMER use to simulate vapor pockets? (column separation / Cavitation)
Assumptions and limitations of tracking air or vapor pockets in HAMMER
Dear Mal,
Very good point. I appreciate it. Yes, you are right please see the attachment. I am not using a cavitation model.
But for the other one, I didn't have the negative pressure (as you see below). Do you have any comment on improving accuracy?
Hiwa
Hiwa said:But for the other one, I didn't have the negative pressure (as you see below). Do you have any comment on improving accuracy?
It could be due to a difference in how the valve characteristics are modeled, or with the discharge-to-atmosphere (D2A) pressure vs. flow characteristics. What assumptions are you making in your program for these elements? This article explains the Ball valve characteristics and this article explains how the D2A works. If you change the valve type, you will see a large difference in the pressure time history. For example if you try the Globe valve type instead of Ball, the results will be quite close.
Thanks Jesse,
Yes, it worked and now the results are quite similar! Could you please introduce me a reference on the difference between different valve actuation in Hammer?