Which roughness value (Rz, Ra, ) to use in pipe flow equations

[WouterWouter]

Hello,

I'm currently doing a project were I have to design a tube through which a fluid will flow at > 10 m/s. I am validating my calculations and want to use the correct roughness values. The material is a plastic deposited using rapid prototyping.


I have measured:
Ra = ~7.5 - 8
Rz = ~1.5 - 2


Clearly, using Ra or Rz gives different result. Which one should I use? Or should I use maximum roughness measures such as Rt ?

Regards,
Wouter

 

[Ranger Mike]

chances are the pipe is made by extrusion which means that to correctly measure any surface texture parameter you must measure the pipe circumferiencially not linearly which most surface texture instruments do..pilotor moves stylus linearly in and out. see my post ref: Ra measurement last month.

 

[WouterWouter]

chances are the pipe is made by extrusion which means that to correctly measure any surface texture parameter you must measure the pipe circumferiencially not linearly which most surface texture instruments do..pilotor moves stylus linearly in and out. see my post ref: Ra measurement last month.

The pipe is actually made using stereolithography (3D printing) and you are absolutely right, the roughness changes along the section and the length. For flow, it is the roughness along the length which will be the most important.
However, I still have no idea which roughness measurement value to plug into the equations ...

 

[Ranger Mike]

remember that Surface Texture parameters should be measured perpendicular to the lay or in direction of maximum roughness reading..since you are measuring with the lay ( or grain) of the material Ra is worthless...if there is a way to secure a specimen of pipe that you know works correctly and if you can secure a specimen of pipe that is known to not function properly, we can not measure both and see which parameter is significantly different ..my guess is that Rz ( ten point average of 10 largest peak to valley excursions) will be best as you can get an idea out the surface details better than Ra...

 

[jamespb]

Hi,

I am struggling with the same issue as original poster WouterWouter,

I have used a surface profilometer to measure the roughness of a range of materials, but now have no idea which measurable to use (I know Ra Rz Rmax Rq Rm S Sm R3z Wt and a bunch of other variables - but no idea which one to use as the "absolute roughness" to calculate my "relative roughness" of any given pipe)

Thanks in advance for any help

 

[Ranger Mike]

there are over 140 different surface texture PARAMETERS..Ra being the most popular. the reason there are so many is that one particular application required one specific parameter. The profilometer was invented in 1938 and there were no computers back then so the only was to scrutinze a surface was to custom build an electronic amplifier to process the data and give a number. PeaK Count worked for the steel mill production, Rz for electric connectors, etc..what you really want to see and the profilometer can not measure this, is to look at the Total profile of the surface and analyze the best surface for each application.

but..total profile is made up from 4 components. waviness, roughness, flaws and lay...the profilometer can measure roughness parameters over short segments or cutoff lengths but can not look at longer surface segments like waviness..a more expensive lab grade instrument is required. i recommend you find a surface that is suitable for your application and a surface that is NOT suitable and use the profilometer to find parameters that really show the difference between the two.

 

[Bandit127]

I read a bit on this today and I couldn't seem to find an authoritative answer.

Having done some electronics in my time so Rq would be my choice since it is the RMS of the roughness. For an electronic waveform it is the power of the signal. For pipe roughness i think this equates to the power of the restriction of flow.

Rq will come up larger than Ra so even if it is wrong you will have a conservative answer.

Rt will not work as it will include any form error or waviness - this is unlikely to influence flow but could make the number bigger by orders of magnitude. .

Roughness in the line of flow will have much less effect than across it. While roughness should always be measured with regard to the direction of lay, because this is a flow problem I think the lay is less relevant.