×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Fluid dynamic problem

Fluid dynamic problem

Fluid dynamic problem

(OP)
Gentlmen,

I have a problem with a pump circuit.
A centrifugal pump has sends a neutralization effluent with density 960 kg/m3 from ground to the top of an atmospheric drum (Bottom T.L. = 6 m).
Before the drum there is a filter with pressure drop of 4 bar that is in elevation, at 21 m.
The drum is exactly below the filter. The filter is complitely full of liquid.

I don't understand what kind of flow is in the pipe going from the filter to the drum.

If the pipe if full of liquid, the liquid flows for pressure difference, but in this case the filter should operate under vacuum. The pump should be calculated considering the atmospheric drum as arriving point.
In this case I will calculate a low static head, because the drum upper tangent line is at 10 m.

If the pipe is not full of liquid, I must consider the filter atmospheric. In this case I will obtain an higher static head for the pump (since the arriving point is the filter at 21 m).
At the following link you'll find a sketch of the circuit:
http://img822.imageshack.us/img822/4707/eqo.pdf

Thank's a lot for your replies.

Sergio

RE: Fluid dynamic problem

(OP)
No, I'm a process engineer and I work for an engineering company.

RE: Fluid dynamic problem

If you keep a vapor space in the tank at atmospheric pressure as maintained by the vent looking thing, then you will most likely have "cascade flow" between the filter and tank, which will happen if the pump's discharge head - pressure drop from friction of your flowrate reaches atmospheric pressure at any point between the pump and tank. You must have at least enough head from the pump to reach the 22(?) meter inlet/outlet filter elevation, after subtracting the friction loss between pump and filter and the filter's 4 bar dP. Let's say you have a 2 meter friction loss in the pipe. That means the head at the pump's discharge (not pump's differrential head) must be at least 21 m + 1m + 40 m (using 10m/bar dP of filter) = 62 m, so you can arrive at the filter outlet at 0 bar. If you only provide that 62 m, you will have no more pressure to drive flow, so gravity will take over. That is what we pipeliner's call "cascade flow", which means that the pipe is flowing full until reaching the high point where the fluid then starts spilling over the edge and running inside the downstream pipeline only as fast as gravity will carry it; usually in a partially full condition. If you provide slightly more pump discharge head, the pipe will flow full a little longer and the cascade point will effectively move closer and closer to the vessel before gravity takes over. If you provide a great deal more of pump discharge pressure, you will move the cascade point farther and farther downstream until it perhaps reaches the point where the vapor pocket starts at the vessel. If you provide even still more pump discharge head, you will pressurize the vessel, fill it entirely with liquid and then start blowing out the vent.... or some other place.

Independent events are seldomly independent.

RE: Fluid dynamic problem

(OP)
Thank you BigInch.
I think that the cascade flow is more probable solution.
However, I've calculated Froude number for a vertical downflow (fluid velocity is 1 m/s and line size is 1.5 in with 80S schedule).
The result is Fr = 2.7 .
I know that for Fr > 2 (if liquid doesn't seal the bottom of the pipe), the total cross section of the pipe should be water filled. So, we don't have a cascade flow but
a liquid moving for pressure difference. As result, the filter will be under vacuum.
What is my mistake?

RE: Fluid dynamic problem

I don't know a lot about froude numbers, but I suspect your error is the 1m/sec liquid velocity. the liquid will increase in veleocity until the friction losses match the gravtiational force on the liquid.

BI has it exactly right in his description of events. The only thing you can really do is create more frictional losses in the section from the filter to the tanks such that the pressure required at the top of the pipe is greater than 0 barg. for a flowing system this could be either a smaller pipe than you currently have, an orifice plate at the bottom or a control valve maintaining a set back pressure in the incoming pipe of around 1 barg. The disadvantage of the first two is that they will drain down when flow stops.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

RE: Fluid dynamic problem

I don't see how you can have it both ways. If the pipe run to the tank is full, that means that it has a truckload of resistance that keeps water in it against gravity, which you have to pump against, which increases your head requirement anyway. Your drawing shows "atm" in the top of the tank, so as with faucets, the pipe between filter and the tank will almost always empty by gravity and there is no realistic way that you can take advantage of the potential energy of the water in the downleg to the tank.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies

Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers

RE: Fluid dynamic problem

Actually there is no way to avoid the potential energy of gravity in the downcomer. Gravitational energy is always there. To keep that downcomer full, he has to pump more flow than gravity can keep drained out, or he has to put a partially closed valve there to keep gravity from doing what it likes to do... drain... drain... drain. But even if he puts a valve there, gravity is still there with its potential energy and ready to start draining as soon as the valve is opened again. When the valve is opened, he is taking advantage of gravity drainage. He will get all gravity has to offer in downcomer flow, just by pumping to 62 meters. The flowrate that gravity will power in the downcomer is always free. If he wants to flow faster than that, then he'll have to pay, but only for the extra flowrate that gravity won't evacuate alone.

Your F number calc isn't correct, because I believe that you've assumed a steady state flow of 1 meter/sec throughout the system, however gravity will accelerate that flowrate in the downcomer, hence at a faster velocity there, that pipe will not flow full, it will cascade down increasing velocity at 9.81 m/sec2 x time it takes to get to the bottom, while it continuously reduces its cross-sectional area of flow. As the fluid flows downward, the pipe will flow less and less full to make up for the steady mass flowrate, but at faster and faster velocities.

Independent events are seldomly independent.

RE: Fluid dynamic problem

This is a very real and common problem in plant design. Especially in cooling water circuits. It is very possible to have pressures below atmospheric at the high point of the piping. This is a "bad thing" (TM). It results in localized boiling, cavitation and vibration. It is very hard to design the vertical downflow section (from filter to tank) to give sufficient back pressure to prevent a vacuum unless your flow rate is very steady and consistent. I haven't seen that situation in the plants I have worked in. If you want the pipe to always run full (without vibration) you will have to install a control valve at the inlet to the tank to keep the pressure at the high point above the boiling pressure.

Luckily our predecessors solved this problem for us long ago. At the high point after the filter you must install a vent to ensure that the pressure there is always atmospheric. This removes the problem of the pressure at that point varying with flow rate. The vertical downflow section from the vented point down to the tank should be designed for self venting flow. This is achieved by keeping the Froude Number below 0.3. By doing this you ensure what BigInch has called "Cascade Flow" occurs all the way down. It also make the sizing of the pump a whole lot easier because the outlet pressure of the filter is fixed at atmospheric.

It is unusual for me to disagree with either of the Inch Brothers, but in this case I have to query LittleInch's assertion that the velocity will increase in the downflow section. If the pipe is full, by continuity the velocity has to be equal all the way down. If the pipe is not full the friction losses are less than the change in static head and become irrelevant.

Katmar Software - AioFlo Pipe Hydraulics
http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

RE: Fluid dynamic problem

I think the OP was hoping that the falling water would provide "siphoning" that was equivalent to the change in head, hence his question re. 10-m head. If the flow rate is slow, then the cascade flow will not support any such siphoning, so head is 21-m. If the flow rate is faster than what gravity would accelerate the water to, head is still greater than 10-m, because you now must pump against the friction in the down leg as well. So, unless the tank is filled, and there is no air anywhere in the system, the OP will never be able to only see a 10-m head and take advantage of siphoning.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies

Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers

RE: Fluid dynamic problem

Katmar,
Does the continuity equation really apply here? If I dump a bucket of water from 20 m, the velocity of that lump of water will increase continuously until it hits the pavement. I think of cascade flow the same way--each lump ("control volume" if you want to be less colloquial) of water that overtops the hill in a line that is not full will fall under the influence of gravity disconnected from the last lump or the next lump. "Under the influence of gravity" means that it is accelerating at 9.81 m/s2. It really can't accelerate downwards at the same time it is "maintaining a constant velocity". I know that the continuity equation does not apply to channel flow where fluid can "stack up or starve" the flow. Any time you have a partially full line you have that situation.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.
The plural of anecdote is not "data"

RE: Fluid dynamic problem

It was the BigInch, but I can see why it's confusing.

And that's exactly why I'm correct; continuity DOES apply here (nobody's adding mass flow anywhere) and, just as it does to all steady state mass flow, including steady open channel flow, to maintain the same mass flow under accelerating velocities in a VSLC (very slightly compressible liquid) the cross-sectional area of flow must reduce to compensate for the increased velocity. Otherwise A x V x DENS = mass flow would not be constant, and nobody was adding any mass flow anywhere in that system.

In open channel flow, steady state flow at constant depth occurs where the down-slope of the channel imparts just enough energy to the flow to counter the tendency for the fluid to accelerate as it decreases its elevation. A shallower slope will cause the velocity to slow down, due to lack of sufficient gravitational energy to keep up the speed, but if steady mass flow is maintained, the depth must then increase. Likewise, a steeper slope will add more gravitational energy to the fluid than what was previously being expended, and again, if no mass is added to the stream (it's not raining right), the constant mass flow will be maintained during the increase in velocity by a reduction in the depth of flow.

Independent events are seldomly independent.

RE: Fluid dynamic problem

David, yes continuity has to be maintained in the context used by LittleInch. I agree with what you describe as happening if you dump a bucket of water in free air. No problem with that. But the context described by LittleInch where the water accelerates until the friction losses match the gravitational losses is very different from what you describe. If the water is constrained by friction against the wall then the pipe has to be full and continuity applies. If the pipe is not full there will of course be some friction against the wall, but some of it will also be in free fall (BigInch's cascade).

It's really splitting hairs though. In the calculation of the Froude Number we are not interested in whether the pipe is full or not. The velocity in the determination of the Froude Number is the superficial velocity, which assumes that the pipe is full of the phase we are considering (i.e. liquid in this case). We happily use the Fr < 0.3 criterion for self venting flow - where the pipe is most definitely not full of liquid because the vapor is flowing countercurrent up the pipe.

Katmar Software - AioFlo Pipe Hydraulics
http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

RE: Fluid dynamic problem

The only hydraulic reason I can see to have a vent at the filter is to let out any air that might collect at the high point and prevent a potential vapor lock at slow fluid velocities. If flow was fast enough to keep full area flow in the downcomer, and keep air from backing into the downcomer from the vessel vent and bubbling up to the filter, no vent would be needed up there, but its probably better with it, especially if the flow is not fast enough to clear out any of those potentially trapped bubbles.

Independent events are seldomly independent.

RE: Fluid dynamic problem

It is coming back to me now. The mass flow rate is constant. For an approximately incompressible fluid the volume flow rate is constant. The product of density, velocity and cross-sectional area is constant (so in a pipe full of incompressible liquid the velocity is constant, in a partially full pipe the area of the flow is not constant so the velocity is not constant. It is amazing what you can mangle in your mind in 20 years.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.
The plural of anecdote is not "data"

RE: Fluid dynamic problem

I had no doubt that you would get there. Rotate hat 180{&deg;] and it's easier. Thinking in terms of gas flows through constant Ax (if pipe diam is constant) changing their volume, velocity and density to hold constant mass flows, whereas liquids don't change volume, or density, but do change Ax ... steady mass flow in both cases.

Independent events are seldomly independent.

RE: Fluid dynamic problem

(OP)
BigInch, I agree with you about continuity equation but it dosen't solve my problem with Froude number.
Since we assume that the pipe is not full. The fluid cross section decrease when we approach near to the tank, so fluid velocity has to increase.
I've considered 1 m/s to calculate Froude (this is the result that we obtain for pipe full of liquid).
For this reason calculated Froude number is the lowest that we can obtain, and it is yet higher than 0.31 (or 1... I've not understood if there is a difference in Froude limit if the liquid seals the bottom of the pipe). It means liquid full pipe... How validate the cascade flow?

RE: Fluid dynamic problem

Perhaps you should refer to this thread: thread378-86817: Air entrainment / Vortice formation

As I understand it, Froude number will simply tell you whether the flow is self-venting. However, for the purposes of calculating head, the Froude number at the value you are calculating is irrelevant, since all the Froude number is telling you is that the opening of the downpipe is fully covered by water, and does not tell you whether the pipe itself is filled.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies

Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers

RE: Fluid dynamic problem

first of all, froude number is not appropriate for very steep slopes such as vertical as shown in the sketch. Without venting you will not get uniform flow conditions. You will get slug flow. the flow will vary from full flow to cascading because of the vacuum created in the pipe due to the lack of a vent. Alternatively, you will need to increase your pump pressure or use a throttling valve to prevent the vacuum from occurring in which case, who cares about Froude?
second, if you are going to calculate froude, you need to know what the normal depth is. you can't just assume a velocity based on full pipe flow if the pipe is not actually flowing full - note that the pipe will probably never flow full without venting, the vacuum will prevent it from happening. Ever heard your bathroom sink drain make funny burping noises? that's because the vent is not working properly creating a vacuum... third, a vent or combination air/vac valve is very commonly used in force mains where pumping over a hill and would be recommended for this application in order to reduce the pumping pressure and prevent vacuum.

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources