2 replies to this topic

[ryn376]

I was trying to calculate the f_T values for small pipe or copper tubing and I noticed an apparent discrepancy between the Crane f_T values and the friction factor for straight clean commercial schedule 40 steel pipe (See attached Excel) at fully turbulent flow. Is this just a calculation difference or does anybody have a different explanation? 

Additionally, since the values don't seem to match, I am not sure if I can use the calculated fully turbulent friction factor for smaller pipe sizes such as 1/4", 3/8", or for tubing. How do I approach this using Crane's method?

 

Attached Files

•  Comparing f_T.xlsx   14.93KB   15 downloads

[breizh]

ryn376 ,

Probably Harvey will reply to you in a "scientific" way . Using the search function in this forum you will find a very nice paper from Katmar on similar subject.
What is your expectation ? Seems to me that the figures are matching well .

Good luck.

Breizh

[katmar]

The differences between your calculated friction factors and those given by Crane are too small to worry about. The Moody chart looks nice and smooth, but those beautiful curves will probably only fit the experimental data to within about 10-15%.

You have not explained what you are using the ƒ_T values for, but I would imagine that it is for calculating pipe fitting resistance coefficients (K-values) because you could never achieve the Reynolds numbers you have quoted in calculating the friction losses for the straight pipe itself. The Crane method works well for calculating K-values for turbulent flow in pipe fittings of different sizes with the same geometry. As fittings get smaller it becomes more difficult to achieve geometric similarity - see Figure 2-13 in Crane. I suspect that most experimental data is generated in pipe sizes of 1" and larger, so I would be a bit cautious of the results for very small pipes. The expected accuracy for K-values is in the region of 25-30% and perhaps a bit worse for these small sizes. Use pressure drop data from the fitting suppliers if it is available.

If you want a bit more detail on applying the Crane method have a look at this article (which I think is the one breizh was referring to).