7 replies to this topic

[Fred den Boer]

I have been checking my Darby 3K spreadsheet calculation for pressure losses in fittings, and noticed that different values for the Ki constant are published in different places for flow run through a Tee (flanged/welded).  I had a value of 0.017, though have seen 0.05 elsewhere, which makes quite a bit of difference.  Can anyone advise the 'correct' value to use, and if there is a revision history on this.

[Bobby Strain]

Can you tell us what "quite a bit of difference is? I would compare results with Harvey Wilson's software. And the 2-k method.

 

Bobby

[ankur2061]

Hi,

 

Refer the link below for what Bobby has suggested regarding checking it against the 2-K method by Hooper. Harvey Wilson's valuable observations are also provided in the thread. A spreadsheet for 2-K method also appears as an attachment for 2-K method in post #4 of the thread.

 

https://www.cheresou...-pipe-fittings/

 

Regards,

Ankur.

[Fred den Boer]

Hi,

 

Thanks for the replies.  The comparison spreadsheet in that thread (K-Value calculation comparison) is one I looked at.  It uses Ki = 0.05 for run through a Tee. You can see for the example given in that spreadsheet that the resultant Le for fully turbulent flow (3.35m) is more than double for the 2K method (1.25m).  It would be much closer if Ki was 0.017.  However the Crane method gives Le larger again (6.06m).  I am a bit surprised there is so much variation in what must be a fairly standard configuration.

A ki of 0.017 is given for example in http://files.enginee...sure_Drop.xls. My original source was a 1999 paper, though this is a 2001 reference.

 

regards

Fred

[katmar]

The original articles on the 2K and 3K methods by Hooper and Darby published in Chem Eng contained several typographical mistakes. 

 

For Tees used in straight through mode, in Hooper's 1981 article the K_∞ value was given as 0.50 and was later corrected in the collection of articles published as the "Piping Design Handbook" by John J McKetta to 0.05.  In Darby's 2001 article K_i is given as 0.017 and this error was repeated in Darby's book "Chemical Engineering Fluid Mechanics" 2nd Ed (2001) on page 211.  The correction to 0.05 is given in the errata sheet at

https://www.crcpress...ata 2ndEd-2.pdf

[Fred den Boer]

Great, thanks for that, that is the confirmation I was looking for.  Even when taking values from published data, we need to watch out for the errata!

I still find the range between the 2-K method and 3-K surprising, perhaps there is another story behind that  For now I'll stick with the 3-K method.

regards, Fred

[katmar]

Fred, the pressure drop through the run of a tee is so small that I don't find it too implausible that the different methods give different results.  If you have access to Crane TP 410 check Figure 2-16 which shows that different workers have given a similarly wide range for the K value for bends.

 

Some experimental values are available at https://conservancy....13368/pr461.pdf

[breizh]

Good day All,

 

To add to Katmar's comments , an updated table from the 3rd Edition together with a link describing the method

 

https://neutrium.net...ings-3k-method/

 

Hope this is helping you .

 

Breizh

Edited by breizh, 22 July 2018 - 02:46 AM.