Dear Jhon,

The roughness is a function of the pipe type only. It is effective in turbulent flow only, where the friction factor is a function of Reynolds number and the pipe roughness. Moody diagram is a general chart and it is built from a reliable empirical relation, Halaand relation. Therefore, I think you have a mistake in finding out variable roughness with Reynolds number.

Muneer, Thank you for replying. I agree that relative roughness should not vary with Reynolds number. I have checked the data and can see no obvious error hence my question: should friction values from the Moody diagram match all experimental data? The pipes in question are seam-welded steel: could this influence the values?

John Wheeldon

The Moody chart was constructed in 1944, based on data from Nikuradse (if I remember well), so if friction factors from the Moody chart are not universally applicable in pipe fluid flow, I suppose it would have been discovered by now.

Hi,

Please pay attention to this fact that Moody diagram only could be applicable for straight pipes. So, if your pipe is not straight, you have errors.

You can see the suitable formulation for helical coil heat exchanger, and some other equation for friction factor, download the attachments.

Best Regards.

Jan, Thank you for your comment. The Moody Diagram and the principles supporting it have been available for some time, and I agree that if there were problems they surely would have been identified by now. Consequently, when I discovered the apparent anomaly my first reaction was, “What am I doing wrong?” I thoroughly checked the calculation procedure and could find no error. The data were consistent and reproducible so seem acceptable. There could have been errors in the transducers used, but all of them would have to have similar errors rather than just one being awry.

I then decided to check line pressure drop data collected by others. I was able to access five sets and to my surprise, all five had the same anomaly, i.e. rather than following the lines of relative pipe roughness the friction factors cross them. Three of the sets presented pressure drop data without gas density data, so the anomaly could have arisen from an inaccurate determination of the friction factor, although the extent of the departure appeared greater than could be attributed to the missing information.

Two of the data sets presented friction factors plotted against Reynolds number. One data set was presented in the following reference for 0.05, 0.10, 0.20, and 0.40 m diameter pipes

W. Siegel, Experimental Investigation into the Pneumatic Conveying of Granular Material in Horizontal Pipes and a Review of the Similarity Laws (in German). VDI-Forschungsheft 538, Dusseldorf, 1970.

I have scaled the values from Siegel’s Figure and presented them on the attached Moody Diagram. Clearly, the data do not follow the classic trend. I assume Siegel incorporated gas density in his calculations. His research was on dilute-phase pneumatic conveying and this presented gas density and gas velocity in the pressure drop measurement section.

As I said, I find it hard to believe that if there is an anomaly that it has not been discovered previously. That said, five other data sets besides mine appear to contain the anomaly. Nikuradse and Darcy both used water in their experiments. Does a gas behave differently? Has anyone rigorously compared measured and predicted friction factors for gas systems? If not, seems like a good research project.

Mohammad, Thank you for taking the time to forward the three references. The pipes referred to in my question are straight sections and air is the fluid flowing in them. Please see my reply to Jan’s comment. If the data are valid, I am wondering if using a gas rather than a liquid has an effect that is unaccounted for in the Moody diagram.

Moody diagram is for steady incompressile flow fully developed in commercial smooth clean straight pipe of constant diameter. Roughness is also assumed constant in every sense within the pipe. (nothing likely to be found in used pies in industrial applications). Error is always present and also needs to be worked out and estimated. A good reference is Streeter & Wiley, Fluid Mechanics, 7th edition, Mc Graw Hill.

John, I am sorry that I did not see your second question/comment earlier And it might be that in my first answer there was a sense of "I cannot believe your results". But now I think there could be something interesting in your findings. Since air flow in pipes, compressible and incompressible, is one of my areas of interest, I agree it could be a nice research project. I would like to take a look at your results. In your attached Moody diagram the red curve runs to Re = 6.E6 and pipe diameters are 50, 100 and 200 mm. When I try to calculate the corresponding velocities I get unrealistic high values. What could be wrong?

Can we apply Moody chart to compressible flows??

Someone can do research on it and develop it.

Else can add conventional compressible flow corrections.

I think moody chart is a non dimensional graph and related to Darcy factor and Reynolds number use to predict pressure drop. As pressure drop could be happen in compressible flows, therefore it could be used to calculate pressure drop and apply correction number to compensate.