Tutorial - Water Hammer (Time History) Example - Modeling Approach in AutoPIPE

Applies To
Area: Modeling
Original Author:Bentley Technical Support Group

Comments, Questions, and Answers related to "Water Hammer (Time History) Example" in AutoPIPE


1. There is a complete Fluid transient (Water Hammer) tutorial and model example with complete details on the calculations performed by the program, etc.., please see the following AutoPIPE help section for details: 
Help > Contents> Contents Tab> Modeling Approaches> Water Hammer (Time History) Example

2. A good video and reference material can be found at the following website: www.drbratland.com  

Item #1:

On Step 2, why in this example have you decided to use 0.005 sec. as a start time and not zero as recommended in other parts of the help ?


This is just an arbitrary short time to establish the Fluid flow in the example. See the "Calculation" section of the example above. note that the Start Time of the Shock Wave (ts) is not located at 0.00 sec. but some small value.
Suggest using the same in your modeling.

Item #2:

On step 3, where does 0.0425 secs come from?


As mentioned in the online help: 

Problem Definition:

A valve at the end of the discharge line is closed suddenly which creates a pressure shock wave traveling at the speed of sound back down the line. A similar shock wave is assumed to travel down the suction line.

Executing Water Hammer Example (apham1)

3. Create a Water hammer load case W2 for the suction line from point A06 to A00 as shown below.

Note, 0.0425 sec is the Start time of the 2nd shock wave upstream of the pump. Another words the first shock wave ends and the 2nd one begins, See "W1.THL" file

 The equation to calculate this value is mentioned in the same step of this example, refer to the step in the example for details.

 Item #3:

Where does the values for checking the maximum suge pressure (362+228=590 psi). come from in the example steps #7 and #8?


228.75 psi = See *.TIH files, the highest Joukowski pressure in the transient (ex. open W1C12.TIH).

362 psi = Pressure Load case. See P1 load case setting on the Press/Temp/Pipe Id tab of the input grids.

Note: Joukowski pressure is the change in pressure from 

Item #4:

On step 9, what is the 150hz selected, ie why 150, why not 200, where did you get 150 from. You state for larger systems, larger frequencies are needed, how do you know what is required for a 1.5km 30" line?


I apologize for the inconvenience, however I would suggest that you read the following WIKI on What is Bentley TSG? here

With that said, this is just an example that used arbitrary values from the typical range. Overall the user is responsible for selecting the actual values used in their analysis. As a recommendation, see the following WIKI page for more information on MODAL Analysis

To answer your question directly, how do you know what is required for a 1.5km 30" line, Start with 150 hz and keep increasing the value until the higher frequency contributes nothing to the analysis results or until your "Total captured modal mass (%)" on your frequency sub-report has reached acceptable levels and again higher frequency contributes very little to the results. Overall, this value should be based on sound engineering judgement where typical values range from 150 - 200 hz.

Item #5:

On step 11, time history... 0.443 in duration field, what exactly is this? and where did you get the 0.384 secs for first field?


When running a Dynamic Analysis> Time History, one of the key pieces of information required on that dialog is the "Duration" time, 

From the online help:

Duration = Enter duration of time history event.

Another words how long did it take for the Water hammer / Fluid Transient to run from start to finish. 

As mentioned in the online help, 

Next enter 0.443 in the "Duration" field (First period, 0.384 sec, plus water hammer duration 0.059 sec),

The first period is time required for the first mode, see the "Frequency" sub-report.

For a more easier explanation of this value, please see the following WIKI here: Item #4. 

Item #6:

The trapezoid in the example help file doesn't compare with the time history profile in the model


See Help/Contents/ Reference Information/ Fluid transient/ Fluid Transient Method.  In this case the lag time is less than the rise time.

Item #7:

The calculation for speed of sound for fluid appears to be wrong. 


See AutoPIPE help ribbon> Contents> Contents Tab> Modeling Approaches> Example Systems>

The most import issue is that you use the correct values and all of the conversion factors:

Due to unit conversion add in the following factors appropriately = 144 in^2/Ft^2, and 32.2 Ft / sec^2.

Speed of sound in a fluid for this example should calculate out to 4163 ft / sec (same as mentioned in help example).

See Also

Dynamic "Time History" Analysis

Bentley AutoPIPE