15 replies to this topic

[Jiten_process]

Dear all,

I need some information

how to calculate pressure drop for vacuum lines, for pressurised gas/liquid line we use darcy's equation, but can it be used for vacuum lines also?

can i use established 'K' values and equivanlant length values for calculation of pressure drop of fittings under vacuum?

Does pressure have any relation in pressure drop in line directly, i guess there is no relation of fluid pressure which is there in line on the pressure drop. (only inlet and outlet back pressure have relation)?

Awaiting reply.....

[latexman]

As vacuum gets higher and higher (absolute pressure gets lower and lower and the presence of matter gets less and less), the fluid flow models progress from laminar flow to "slip flow" to "molecular flow". Also, in-leakage of air becomes important for earth bound applications and has to be estimated.

[djack77494]

I have used the Darcy Equation with established methods for pipe fitting losses in vacuum service. I don't think it wrong to have done so (and if it is, I'm in some good company). I would think that the very high velocities but fairly typical viscosities that I have seen would result in turbulent flow. I suppose that under very extreme conditions more exotic phenomena may become important, but, at least as low as 25 mmHG (A) I have not worried about them. (If I'm in the dark, someone please turn on a light.)

[Ali66]

Dear all,

I need some information

how to calculate pressure drop for vacuum lines, for pressurised gas/liquid line we use darcy's equation, but can it be used for vacuum lines also?

can i use established 'K' values and equivanlant length values for calculation of pressure drop of fittings under vacuum?

Does pressure have any relation in pressure drop in line directly, i guess there is no relation of fluid pressure which is there in line on the pressure drop. (only inlet and outlet back pressure have relation)?

Awaiting reply.....

Hi Jiten_process,

Important aspects that you need to be sure about are whether or not density changes are significant as well as sonic velocity considerations.

If so, you instead should use methods applicable to compressible fluids. Adiabatic Frictional Flow and Isotherma Frictional Flow are titles that cover this subject. I used such methods at absolute pressures of 5 to 25 torr. I'd appreciate commensts here if anyone sees a limitation or a concern.

Some people suggest that it is ok (for approximation) to use equations of incompressible flow as long as pressure drops are below about 10% of the applicable pressure value.

In very high vacuum applications, where main free path between molecules is significant relevant to pipe/equipment size, a whole-other approach is likely needed.

Ali

[latexman]

The methods I described are for pressures much less than 1 torr. With the OP not stating the conditions, I answered from my experience.

[Jiten_process]

Thanks to all,

I think we generally say that in case of liquid being pumped in vacuum lines we can go ahead the same way what we do for pressurised lines. Where the density change is not anticipated.

When we talk about gases, generally they are treated as non compressible for for whatsoever pressure above atmospheric and pressure drop can be calculated in the same manner too, with darcy equation. But i feel in case of vacuum lines we have to go to reverse, we must have backpressure as well as delivery pressure and then we have to find out velocity.

Further question is, density being more sensitive to vacuum (for instance water vapor, e.g. figures 5.33m3/kg at 0.30bara and 2.08m3/kg at 0.8bara), what value of density should i take in calculating pressure drop. should i take average density??? subsequently if density change is significant the vol. flow and velocity also i feel not constant all along the line. And this case even becomes so rigoreous when it comes to two phase flow (e.g. mixture of air and water vapor).

Djcak :
what is your experience in this? do you completely treat them as non compressible? if so then of which condition's density value you take for calculating reynold's no.?

ALI66
Will you please give some resource to learn the methods which you are talking about (Adiabatic Frictional Flow and Isotherma Frictional Flow), frankly speaking i have never come across nor i have ever used, this methods? can you give me some literature on this topic...? i will be highly obliged....

Do you have any firm figures as to when exactly we should use above methods and when to use normal dacrcy equation?

And thorugh these replies, can somebody confirm my understanding that pressure certainly does not have direct relation to pressure drop (except in case when inlet and outlet pressure conditions are already established)?

to be continued......

[latexman]

With liquids under vacuum conditions, care must be taken that the liquid does not flash (cavitate, boil, etc.) significantly if pressure drop is calculated as in a pressurized line. Flashing flow is usually handled as two phase flow and significant errors in pressure drop may occur if it is teated as incompressible.

I recommend Crane's Technical Paper 410 titled "Flow of Fluids through Valves, Fittings, and Pipe". It can be purchased on the internet at http://www.flowoffluids.com/tp410.htm

There are some restrictions that should be observed when applying the Darcy formula to compressible flow:

1. If dP < 10% of inlet P, use either upstream or downstream conditions.
2. If 10% < dP < 40% of inlet P, use the average of upstream and downstream conditions.
3. If dP > 40% of inlet P, use compressible flow methods.

Most thermodynamic and fluid flow textbooks cover compressible flow. My favorite reference is Shapiro's "Dynamics and Thermodynamics of Compressible Fluid Flow". Darby's "Chemical Engineering Fluid Mechanics" is pretty good.

[Ali66]

Another textbook that has brief sections on Adiabatic Frictional Flow and Isothermal Frictional Flow is Unit Operations of Chemical Engineering, McCabe, Smith and Harriott, 4th Edition, Chapter 6.
Ali

[djack77494]

Djcak :
what is your experience in this? do you completely treat them as non compressible? if so then of which condition's density value you take for calculating reynold's no.?

My most recent experience was in dealing with the overhead system of a (crude) Vacuum Column. In a past life, I also dealt with specialty chemical processes that involved vacuum services. Perhaps none of these would be clasified as DEEP vacuum. I do not treat the vapors as non-compressible but rather use my "normal" line sizing spreadsheet which deals with not-insignificant changes in pressure.

[Jiten_process]

With liquids under vacuum conditions, care must be taken that the liquid does not flash (cavitate, boil, etc.) significantly if pressure drop is calculated as in a pressurized line. Flashing flow is usually handled as two phase flow and significant errors in pressure drop may occur if it is teated as incompressible.

There are some restrictions that should be observed when applying the Darcy formula to compressible flow:

1. If dP < 10% of inlet P, use either upstream or downstream conditions.
2. If 10% < dP < 40% of inlet P, use the average of upstream and downstream conditions.
1. If dP > 40% of inlet P, use compressible flow methods.

Hi thanks to all,

latexman has given i feel good thumb rules. But please lastly just confirm me one thing, in my application i have very less static height. So apparantly the inlet pressure would be the total dp plus the back pressure at delivery plus the small static i have. so i found my dp value would be significant in percentage to inlet pressure i required. It is like below

Inlet P = Static (2m) + dp(approx.1000mmwc) + back pressure(2.8mwc=0.28bara)

so as you have said, in my case dp is coming around 18% of inlet pre. required, right? so i need to use average value of density, am i correct?

Reply....

[Jiten_process]

and ya, i forgot to say
i got good reference in mccabe smith....thanks once again to all

[latexman]

Is your fluid a liquid or gas? Have you applied the density of the fluid to the static height to convert to pressure? Changes in elevation can usually be ignored in low pressure gas or vapor applications. Since the units are different (m vs. mwc or mmwc) and you are not ignoring elevation changes, I have some doubt.

[Jiten_process]

Is your fluid a liquid or gas? Have you applied the density of the fluid to the static height to convert to pressure? Changes in elevation can usually be ignored in low pressure gas or vapor applications. Since the units are different (m vs. mwc or mmwc) and you are not ignoring elevation changes, I have some doubt.

Thanks for paying attention. Let me tell you my case, we are supplying evaporation plant where water vapor being sucked by vacuum pump from seperators. I wanted to size the line, and i did considering the high velocity as volume of vapors are large being in vacuum so in order to optimise line size, i used velocity in order of 40-50m/s.

Now i wanted to set the vacuum pump pressure, so i was calculating pressure drop. I know the back pressure in seperator as 0.287bara. i know the equipments and instruments pressure drops in a unit of mmwc. I have static height of 2m, but i used it as it is (2m = 2000mmwc, am i correct?). For line pressure drop i used density value at 0.287bara(0.287bara = 2870mmwc, right?) (i.e. delivery pressure) am i correct, or i should use the average vaue of delivery and source pressure? I want to keep consistant unit of mmwc.

is it ok, for your comments?

to be continued....

[latexman]

First, you need to get all the terms in your equation for inlet P in consistent pressure units, or you need to use the Bernoulli equation. You need the density of your water vapor to convert the 2 m elevation change to a differential pressure:

dP/rho = dZ.g/g_c

I suspect you'll see that 2 m of elevation change for your water vapor is much, much less than 2 mwc.

I don't know what metallurgy you have, but you may want to discuss the target velocity range of the water vapor with a materials engineer, especially if there is any oxygen present and/or entrainment. "Erosion corrosion" can be quite aggressive.

[Jiten_process]

ya, you are right, i goofed actually, 2m height will be 2m gas columb, when we convert it into water column it will be too less to consider.

just lastly i need one thing to know is, now density value and all which i need to consider in above case shall be of inlet condition (in my case seperator) only or average of inlet and outlet condition (apparantly vacuum pump source pressure). I guess, it should be an average?

thanks...

[latexman]

For density, use the restictions I gave above for guideance.