|
LDi Catalog Printable PDF 14, 15
 30 years experience studying plots of "pressure over time" is an advantage.
Whether on site services, you, or a local oscillograph user captures the shock and pulsation "signatures", it is essential that readable suction side plots be taken first and interpreted.
A combination of these suction system samples will be shown by a plot.
95% of all the problems Ldi has diagnosed in the last quarter century, were related to, too much pressure, and over sized pipe on the SUCTION side.
That is why these six samples concentrate on reading the suction pressure pulse signatures. Suction force driving the check valves open, was the worst.
There are hundreds of others that make look more similar to your problem.
  1.Suction pipe signature, where a pump momentarily lost prime, or was started too fast for the suction pipe length.
When a pipe is oversized, yet not so large that it becomes an amplifier, you will see a multiple rebound of each pressure change. The interval between each repeat of the pressure signature is related to the pipe length and pressure wave velocity, depending on "softness".
The repeats will be equivalent to 1400 meters per second + or -, and the length round trip.
Where the height of each repeat decreases, as sample shown here, the pipe is slightly dissipative. It has a little delta P (Pressure Drop). The desolution of air, the same effect as reaching vapor pressure, is characterized by a flat bottom to the pressure signature.
This indicates a pressure below, which it cannot fall. Often called "cavitation".
When pump is not Centrifugal, Vane, Lobe, Gear, Progressive Cavity, Screw, etc.
Note: If the suction supply system is designed to feed a reciprocating pump, a signature like this may be seen for every stroke. Do not increase the pipe size or the supply pressure force, put a supply close to the pump.
TOP OF PAGE |
 2.Suction pipe pressure signature, where a check valve is often before the pump to maintain the prime. A check valve or "foot" valve is never simply held open by the viscous drag of the liquid.
Check valves modulate between nearly closed and near full open. They often "chop" flow into pulses. Valve beat fluctuations are characterized, particularly when they are integral with a diaphragm or plunger pump, by a different interval between beats, a.k.a. "frequency", different pressure level, and normally with a rounded top and rounded bottom.
When pipe is long and size gives a good dP, that "scrubs" pulsation with "reynolds action" reverberations will not be seen, they will have been dissipated against the pipe wall.
The mass of liquid is always greater than the weight of a check valve moving element, and the force even of a check spring, is always less than the pressure in the liquid. The supply column to a pump always excites the movement of the suction check; it then behaves according to its own response characteristics and the pounding of the liquid.
When a supply system is broken into slugs and voids by air out of solution, check valves are hammered and the pulsation becomes as violent as a relief valve bouncing.
TOP OF PAGE |
 3. There is a tendency in the trade to ignore valve action and the chopping of flow, and to diminish the importance of correctly selecting pipe size, then to consider everything in terms of "acoustics". By acoustics the experts seem to mean only the results of pressure activity that is reflected from the closed or open end of a system, and from direction changes in the conduit.
In acoustics, all that is relevant is the speed at which pressure changes travel from place to place.
The purpose of piping systems is to convey volume or mass of liquid from A to B. The forces of inertia are important in their own right for pump efficiency. No pump can do anything unless the liquid can freely and steadily move into it.
As very few systems work simply be reservoir head, or by syphones, maintaining the prime on pumps is essential. What we learn from this precise frequency and its transient nature, is only that something sudden happened in a pipe of an exact length, the length of which is determined "acoustically" according to its softness.
Ldi recommends you consider acoustic response, but concentrate on mass transfer.
As the speed of sound depends on the "softness", or the effective modulus for the liquid and the pipe elasticity - both of which depend on temperature - it is somewhat problematic to make recommendations for system changes based on "acoustic" criteria. The temperature is always changing and so also the acoustic response.
TOP OF PAGE
Ldi Page 15 |
10 bubble implosions within one 1/100th of a second. |
4. Channel "cavitation", the use of cavitation in this phrase, is connected to the implosion of millions of bubbles. The bubbles are created by massive velocity between two very close surfaces. It is unlikely to occur unless the pressure difference is at least 4:1, from the beginning of the "two close surfaces" to the end of the two close surfaces. Do not confuse this signal with electrostatic "noise" or "dirt" on your plots.
What is shown here is from a valve allowing flow then trying to almost totally close. This is similar to a relief valve. The frequency is similar to the 1000Hz. hissing noise that can be heard when detecting a small leak. Apart from warning that valve seats will be "eaten" and need replacing, the frequency is dangerously close to that used by straight tube mass flow meters.
Bubble collapse causes a most intense heat. This is said to be close to the temperature of nuclear fission.
It is believed that when all the energy from a collapsing void reaches its final point, it is all on to a final atom.
To capture the shape of this activity, there needs to be six points plotted per line, and to collect readings 4 times faster than the data points, to prevent aliasing.
That means data capture at a 25-kilohertz minimum. To view this activity in real time, the printer will have to run out, or "stream", at feet per second.
TOP OF PAGE |
5. Vortices. There are many sources for these. The pressure differences are extremely small, however they may be important. For example: *1 There are flow meters called "Vortex Shedding Meters". These devices depend on the principle that vortices are created at a frequency that is directly related to the flow velocity passed a sharp edge. The mass wants to carry on, but that causes a low-pressure zone on the backside of the sharp edge. Some liquid is pulled into the low-pressure zone, and forms a "swirl". The frequency of these swirls or vortex swirls or pulses is sensed through holes in the back surface. The frequency is then displayed as the flow rate.
If your system plots show these, be sure that their source is not just before a "Vortex Shedding Meter". If it is, then such a choice of meter will be unsuitable for the system as piped or designed.
To visualize what is happening in your system, recall that when you row a boat, the ends of your oars make "swirls". The center of the swirl is lower than the surrounding ring. Spinning the water has made a pressure disturbance. Or standing by a river, there was a fallen branch, or a rock, the water passing the obstruction made "back water" or eddies.
In many pipe system designs, there are sudden direction changes, and devices that have abrupt, not "flow contoured" internal shapes. These all cause more or less vortex pulsation sufficient to invalidate a DP meter.
TOP OF PAGE |
6. When you see this (first right), make a note of the system designers name, the piping contractor used and all the system parameters. This HEAVEN, it doesn't get any better than what you see...
It is almost impossible to start a system without some momentary dip A. Then there is always a gradual fall during the acceleration B. This is following by a faster recovery C. And as the flow catches up with the system steady state, there is a tad of overshoot, then a settling down D.
If there had been de-solution of absorbed gas, or a fall to vapor pressure, then as the pressure would not have been able to fall any further, these would have been a flat bottom.
With a startup as smooth as this, it is probable that the discharge acceleration head pressure surge was unlikely to have caused "water hammer".
If there is complaint of pulsation, then in this system start my looking for relief valves that chatter, back-pressure valves that "hunt" for a mid position, flow controllers that modulate, and feed-back loops that repeatedly cause an "over correction".
Only if none of the above give reason for suspicion, it is worth going back to look at the suction system. |
|
|
