4 replies to this topic

[Michael from US Navy]

Dear All,

I have recently inherited a compressible flow simualtion that make use of Fanno Flow to determine the mass flow rates and pressures, etc. in a system of pipes and valves. It is quite old and there is no documentation. I have been asked to make predictions on system performance if various new valves are inserted to the system.

My problem lies in obtaining a good equivalent length over diameter (Leq/D) values for these valves when fully open, since that is what the code requires, thus I know . The vendors seem ill equiped to supply anything but valve coefficients (Cv), Equivalent Sharp Edge Orifices, or Flow coefficients. About the only thing I have found is equation 2-5 in Crane's Technicl Paper # 410, but this relationship seems only to apply to water. I have also found relations for calculating a Cv for a critical or non-critical gas, thus I know Cv's exist for a valve were gas is the medium.

What I really need is an algebraic way of going from the given valve parameters above, like Cv, to and Leq/D. If there is no method for determining such a relationship, I sure would appreciate an explination of the physics.

Thank You,

Michael

[breizh]

Hi Michael ,

This link may support your query : from Cv you can get access to K then to Leq/D
http://www.engineeri...ents-d_277.html

additional info attached ( masoneilan handbook)


Hope this helps
Breizh

[katmar]

Michael, there are many references to this problem scattered around the internet (for example http://www.eng-tips....d=249555&page=1), and a search right here in this forum would probably find several, but seeing that you have Crane 410 let's use that.

See equation 3-16 in Crane. The intermediate form contains almost everything you need.

From the relationship Cv = (29.9 x d²) / √(ƒ x L/D) we can get

L/D = (894 x d⁴) / (ƒ x Cv²)

In Crane's terminology the ƒ is ƒ_T, the friction factor for fully turbulent flow in commercial steel pipe - See Crane pg A-26 (top), but if you want an equivalent length of some other pipe you could substitute the ƒ. Cv's are only applicable to turbulent flow, so don't try to calculate laminar ƒ's for this formula.

[Michael from US Navy]

Hi Michael ,

This link may support your query : from Cv you can get access to K then to Leq/D
http://www.engineeri...ents-d_277.html

additional info attached ( masoneilan handbook)


Hope this helps
Breizh

Hi Breizh,

Thank you for the quick response. I too had reviewed the web site you sent me during my research, but when I looked at my case of non-critical gas flow, I only found a relationship of Cv as a function of the flow and pressure drop across the valve. No where could I find a further relationship to an equivalent length of pipe of a given diameter. I also reviewed the catalogue you sent, but again could find no relationship. If I have missed anything please let me know.

Thank you for your time,

Michael

[Michael from US Navy]

Michael, there are many references to this problem scattered around the internet (for example http://www.eng-tips....d=249555&page=1), and a search right here in this forum would probably find several, but seeing that you have Crane 410 let's use that.

See equation 3-16 in Crane. The intermediate form contains almost everything you need.

From the relationship Cv = (29.9 x d²) / √(ƒ x L/D) we can get

L/D = (894 x d⁴) / (ƒ x Cv²)

In Crane's terminology the ƒ is ƒ_T, the friction factor for fully turbulent flow in commercial steel pipe - See Crane pg A-26 (top), but if you want an equivalent length of some other pipe you could substitute the ƒ. Cv's are only applicable to turbulent flow, so don't try to calculate laminar ƒ's for this formula.

Hello Katmar,

Thank you for your response. My concern with using the above relationships is that Crane's Handbook states that equation 3-16 is for liquids of similar viscosity to water at 60 degrees F, while my fluid medium is air. However, while looking at that page more closely, I see a relationship in the form of equation 3-20 which has K, flow, and pressure drop in it. If I take one of those relationships, and another for my Cv relationship with flow and pressure drop, I should be able to back out a K, and then a Leq/d.

Thanks again,

Michael