Pressure in Hammer transient simulation

Hi,

During hammer simulation two issues occured:

1- the pressure , the first 400 sec  in simulation HGL fluctuate near the upper value - max pressure (No.1), after that HGL moves to position No.2 for little time then it moves to position No.3 till the end of the simulation (1200 sec).

2- when I reduce time interval of simulation from 0.05 sec to 0.01 sec the vapor volume disappears. 

Kindly I need illustration for possible causes for these events.

Parents
  • Abdallah, you had posted in the Communities Feedback forum which is intended for questions about the site itself. I have moved your post to the OpenFlows Forum where you can get help from OpenFlows specialists.

    1- the pressure , the first 400 sec  in simulation HGL fluctuate near the upper value - max pressure (No.1), after that HGL moves to position No.2 for little time then it moves to position No.3 till the end of the simulation (1200 sec).

    It's hard to say what specifically is causing the "fluctuations" without seeing the model. What type of transient event are you modeling? It could be due to interaction with vapor and/or air pockets, it could be related to stability issues with a certain element configuration (such as a valve, hydropneumatic tank or pump specific speed), an issue with the selected transient friction method, or something else. If you have not done so already, be sure to animate the profile path to get a better understanding of how things are interacting during the transient simulation. If this does not help, provide a copy of the model and steps to reproduce: Sharing Hydraulic Model Files on the OpenFlows Forum

    2- when I reduce time interval of simulation from 0.05 sec to 0.01 sec the vapor volume disappears.

    This may suggest that the 0.05 seconds is not small enough to accurately capture fast changes occurring during the transient simulation, or possibly an indication of some other instability with the model (which could be "real" due to the complexity of wave interactions, or unrealistic due to some other model instability). Again without a deeper understanding of the transient event and model configuration it is impossible to tell for sure. We may need to see a copy of the model.


    Regards,

    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.

  • Hi Abdallah, all I see in your reply is "Transient". Did you mean to say something else? If you still need help after reviewing my previous answer, please explain in more detail and we'll be glad to help.


    Regards,

    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.

  • Hello Scott,

    I re-uploaded the model after proper orientation of the scenario i want to understand the behaviour of hgl due to pump shut down failure.

    Regards,

  • Hello Abdallah,

    In the new model, I see vapor forming In the profile for transient time steps of 0.05 seconds and 0.01 seconds. This is occurring because the hydraulic grade after the pump shut down is lower than the elevation at the local high point. Putting some sort of protection at that high point may help mitigate the vapor formation. 

    Regards,

    Scott

  • Hello Scott,

    Many thanks for your efforts.

    I would also like to understand why HGL is gradually falling after pump failure, shoudn't it be sudden falling in HGL.

  • Scott and I looked at your model.

    I have taken a look at the model and I am seeing a drop in HGL immediately after the pump shuts down – it doesn’t look like much of a drop but I think that’s because of the Y axis scale – the HGL is about 1000 meters above the physical elevation. This is because the pump is adding almost 1000 m of head, which seems very excessive. Is this correct? Do you really have a pump that can add this much head? (the pipes are only about 500 mm). If so, is your pump's Specific Speed correct? (see this) I do see there’s a bit of a high point in the middle of the profile but still the pump head seems very excessive.

    If you look at the graph of head downstream of the pump, the drop after the shutdown is actually around 32 m – from 1340 m to 1308 m. See below.

    The HGL eventually drops even lower as the system settles down from the reduced headloss through the downstream FCV, but the profile is over 17 km so it takes a while for that to happen. The wave needs to travel to the end of the system and back.

    I also noticed there is a flow control valve near the downstream end of the system, which needs to close almost all the way (a calculated headloss of 770 m). It seems highly wasteful to add over 1000 m of head at the pump and then bring it back down to 560 m (reservoir R-2) by closing an FCV. The midway high point does not seem to be nearly high enough to warrant this. Is this a real FCV or are you trying to use it to model something else? See tips and points here: Using a Flow Control Valve (FCV)


    Regards,

    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.

Reply
  • Scott and I looked at your model.

    I have taken a look at the model and I am seeing a drop in HGL immediately after the pump shuts down – it doesn’t look like much of a drop but I think that’s because of the Y axis scale – the HGL is about 1000 meters above the physical elevation. This is because the pump is adding almost 1000 m of head, which seems very excessive. Is this correct? Do you really have a pump that can add this much head? (the pipes are only about 500 mm). If so, is your pump's Specific Speed correct? (see this) I do see there’s a bit of a high point in the middle of the profile but still the pump head seems very excessive.

    If you look at the graph of head downstream of the pump, the drop after the shutdown is actually around 32 m – from 1340 m to 1308 m. See below.

    The HGL eventually drops even lower as the system settles down from the reduced headloss through the downstream FCV, but the profile is over 17 km so it takes a while for that to happen. The wave needs to travel to the end of the system and back.

    I also noticed there is a flow control valve near the downstream end of the system, which needs to close almost all the way (a calculated headloss of 770 m). It seems highly wasteful to add over 1000 m of head at the pump and then bring it back down to 560 m (reservoir R-2) by closing an FCV. The midway high point does not seem to be nearly high enough to warrant this. Is this a real FCV or are you trying to use it to model something else? See tips and points here: Using a Flow Control Valve (FCV)


    Regards,

    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.

Children
No Data