adjusted K- and Cv values for glycol solution

[EnergyProfessional]

I'm writing a program to calculate pressure drop in piping systems. It takes into account % of glycol, and temperature. this all works for straight pipe.

Since I don't want to use equivalent length for elbows, valves etc. I want to sue the K- and Cv values. But those are for standard water at 60F. how do I adjust the K or Cv value for different fluids?

One (bad) idea i had is to use the ratio of how the straight pipe differed from water. for example, if in straight pipe i have 1.2 times the pressure drop compared to standard water, then I use that 1.2 factor for valves etc. but this doesn't really take into account how the fluid behaves int eh valve (which may be thinner, hence turbulent vs. laminar etc.)

Any idea? It should be some equation I can use in a software. Not just some rule of thumb number.

 

[quark]

...for example, if in straight pipe i have 1.2 times the pressure drop compared to standard water

That is possible only when you are calculating pressured drop in pressure terms (for ex. bar) and not head terms (for ex. meters)

Any idea? It should be some equation I can use in a software.

Did you check the Cv (or K) equation?

Cv = Q[×](SG/[Δ]P)^1/2

 

[quark]

Just ignore my first comment. That is plain stupidity.

 

[katmar]

This topic is not well covered in the general literature. It has come up here before and a search for terms like "Cv" "Re" and "Reynolds" will get you a few hits. See for example thread378-258214


Katmar Software - Engineering & Risk Analysis Software

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[EnergyProfessional]

I followed that other thread and the best I came up with was using this procedure:

http://www.irmdiesel.com/AmotDocsFR/Vanne thermostatique 3 voies R en acierRvalve 01.pdf

On the last page is the viscosity correction graph. Is there an equation for that graph, assuming my software is not supposed to look at a graph? Ans i assume once I have the corrected Cv value, I still use it with the density of the actual fluid (which also changes with temperature, and type of fluid)

Many of the recommendations in the thread (and elsewhere) don't take into account, that we really may have completely different flow, for example even in straight pipe i easily go to turbulent flow by adjusting temperature or pipe diameter just a bit. A valve has some thinner, curved etc. parts that we don't know (hence the manufacturer gives us the Cv value).

thanks for the help so far. this was much more than years of google

 

[stanier]

Download this freeware and do the calcualtions in that for your fluid properties, valve dP etc.

http://www.pumpsystemsmatter.org/content_detail.aspx?id=110

Go to

http://www.aft.com

and download the Fathom or Impulse Viewer and look into the help files for the design basis.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/

 

[MortenA]

You dont adjust the K and Cv values you adjust the friction factor based on Reynold no.

Best regards

Morten

 

[EnergyProfessional]

Morten: i don't understand why i would adjust friction factor. with a valve that has a Cv value, I don't really have a friction factor. since the geometry inside the valve is unknown, I also don't know the Reynolds number. Reynolds number and friction factor I only have in relatively straight pipe (at least in a way that i can calculate it)

I know how to have my software calculate Reynolds number and friction factor in straight pies for all different fluids, if that is what you meant.

 

[MortenA]

Because thats how Darcys (or Fanning) law works - K is a dimensionless number where:

dH=f*L/D*U^2/2g (variables assumed self explanetory even though various letters are used in various text book)

and

K=fL/D

f is a funtion of abs roughness and Reynold no (Re)

Re= density*V*D/vicopise (its dimensioless)

There are numerous correlations between f and Re, Abs roughness f.eks. Churchill, S.W., 1977, "Friction factor equations spans all fluid-flow ranges.", Chem. Eng.

So instead of changing your K - that should be fixed - use the existing correlations to change your f.

Its true for a valve you dont (normally) have a friction factor - but you do have a K

Your calculation may need an adjustment of the K values for low values of Re (this is true for any type of components, valves, bends, orifices ect. One reference for these correction is "Internal flow systems# by DS Miller.

Best regards

Morten

 

[EnergyProfessional]

MortenA:
it looks like what you are suggesting is to determine an equivalent length and a corresponding friction factor? I'm a bit concerned since the equivalent length isn't so good - but I have to think more about if mathematically thsi isn't the same. But my fear is, your method would assume the diameter is the same as in the pipe, which isn't the case for fittings.

I'm also not sure how this method (page) 7:

http://www.irmdiesel.com/AmotDocsFR/Vanne thermostatique 3 voies R en acierRvalve 01.pdf

would be redundant or contradicting to what you say.

 

[btrueblood]

A valve is not a pipe. Viscosity has little influence on the flow of what is, essentially, a variable-area orifice plate (at least, within the range of variation of water and glycol mixes at temperature up to near boiling). Specific gravity, or density, however, directly affects the flow rate of fluid through an orifice plate. Katmar pointed the OP to another recent thread, where all of this was hashed out in excruciating detail. Adjust the specific gravity for temperature and glycol percentage, and the Cv law will compute the correct flow for the valve (well, assuming you have a valid Cv number in the first place).

 

[EnergyProfessional]

btrueblood: I think you are right. but besides density viscosity also needs to be accounted for. This document on page 7

http://www.irmdiesel.com/AmotDocsFR/Vanne thermostatique 3 voies R en acierRvalve 01.pdf

shows that the head (of that specific fluid - to get the pressure we need the density) depends on K, which has a correction factor depending on viscosity. The chart on page 7 shows that. I entered plenty of the values in Excel and tried some curve-fitting. It seems the logarithmic fit looks the best: "K-correctionfactor" = 0.075 ln(viscosity in cp)+ 1.0471. I know, it is a bit off at standard water. So we need to:
- find the correction factor for K (with the chart, or my equation)
- calculate head with K
- calculate pressure drop using density of the fluid. So we have 2 parts where the fluid properties other than water make a difference.

Does that sound reasonable?

 

[btrueblood]

Certainly. Though I'd be a bit nervous that the document you show has little information on correcting for sp. gravity, and instead ONLY publishes a correction curve based on viscosity. Hopefully they did their maths correctly.

 

[EnergyProfessional]

actually my equation should be viscosity in centistokes...

the document shows the head calculation.... which is independent of the fluid. The actual pressure will be calculated separately.

So, i calculate 100 ft. If that is water, it is 100 ft water pressure. If that is glycol, it is 100 ft glycol pressure... actual pressure (in psi or Pa) will be calculated by head times gravity of the fluid. so they don't neglect the topic, they jsut fucus on head only (as opposed to pressure)