Hello,
I am working with a gravity fed model that has a TCV valve at its downstream end. The valve type is "defined by user" and I have set its characteristics so the relative closing / discharge rate is a linear relationship.When I start the first simulation with the valve opened (with valve closing partern), the initials conditions are what they are expected (280 cms) and the flow remains the same until the valve closes.
The second simulation starts with the valve closed and is set with a valve opening pattern. Without any change to the model geometry or setting (except of course for the initial relative closing of the valve), the flow reaches 338 cms at the end of the simulation.
How is this possible?
Thank you in adavnce for your kind support.
Dominique
Hello Dominique,
Could you please upload model files for our review?
Sharing model files.
Meanwhile here is information about modeling TCV relative closure, if you haven't gone through it yet.
Configuring TCV relative closure patterns for EPS
Regards,
Sushma Choure
Bentley Technical Suppport
Hello Sushma,
I have uploded the model files (Copy_Bentley_TCV_Valve). Here are the scenarios you can compare:
1) Modeled scenario: valve full opened throughout the simulation:
-Initial settings/Relative closure (initial) (%) = 0 ; Transient (Operational)/Operatig rules = Fix
->Result (flow): Initial condition = 284.58 cms ; steady state = ~284.6 cms (=> OK)
2) Modeled scenario: valve closing after 20 sec:
-Initial settings/Relative closure (initial) (%) = 0 ; Transient (Operational)/Operatig rules = Closing
->Result (flow): Initial condition = 284.58 cms ; steady state (begining) = ~284.6 cms ; steady state (end) = 0 cms (=> OK)
3) Modeled scenario: valve opening after 20 sec:
-Initial settings/Relative closure (initial) (%) = 100 ; Transient (Operational)/Operatig rules = Opening_Interval_10s
->Result (flow): Initial condition = 0 cms ; steady state (begining) = 0 cms ; steady state (end) = 337.9 cms (=> UNEXPLAINED)
NOTE: The flow is read in transient results / Historic / (point) P-13:TCV
Please note that I have submitted another question concerning the flow in the surge tank riser and your colleague Scott Kampa is reviewing an anterior version of this model (different type of valve). Here is the thread of this question to your colleague: https://communities.bentley.com/products/hydraulics___hydrology/f/haestad-hydraulics-and-hydrology-forum/192638/surge-tank-detailed-report---flow-in-tank-and-riser/569595#569595
Hello Dominque,
This may be related to how HAMMER is calculated the Darcy-Weisbach friction factor. When the valve is initially closed, there is no flow in your pipes. When the valves starts out open, there is flow. The discharge coefficient calculated and used in the model may be different in the two cases. See the section "For pipes that start with zero initial flow" in this link for some details.
If you want to assure that that friction factor is correctly applied, you can start with the valve open and set up an operating rule where the valve closes and then reopens. The timing of the operating rule may require some trial and error, since you will want to the system to come to a new equilibrium after the valve is closed. If you want to see results just for when the valve is being reopened, you can set your Report Times in the transient calculation options to only be reported after a given time.
Scott
Hello Scott,
Thank you for your answer. You are right, after checking the output log I have different DW friction coefficients. I will use your method.
Thank you again, your help has been truly useful.