Disk Friction Losses

In document Volk, Michael W-Pump Characteristics and Applications-CRC Press (2014) (Page 97-112)

Hydraulics, Selection, and Curves

D. Disk Friction Losses

If the pump impeller is thought of as a rotating disk, rotating in very close proximity to a fixed disk (the casing), there is a frictional resistance to this rotation known as disk friction.

The pump efficiency is expressed as a decimal number less than 1, for exam-ple, 0.75 for 75% efficiency. The relative importance of the above four losses varies from one pump type to another. Actual efficiencies for various types of centrifugal pumps can vary widely, over a range from less than 30% to over 90%, for reasons that are explained in more detail in Section XIII.

Comparing Equations 2.12 for WHP and Equation 2.16 for BHP, the only difference between the two is the pump efficiency term. Therefore, the pump efficiency is equal to the ratio of the two:

η = WHP

BHP  = Q   H   SG 3960   BHP

× ×

× (2.17)

The pump manufacturer uses Equation 2.17 to determine the pump effi-ciency at the time the factory pump performance test is done, as described

below. This same testing procedure can be done in the field as well to verify pump performance and compare efficiency with the as new condition.

When a new pump is being designed by a pump manufacturer, there is usually a predetermined objective for the pump’s flow and head at the best efficiency point, as well as an expected maximum efficiency. However, it is not until the performance test is run on the prototype that the performance that the manufacturer lists in the catalog for that pump is finally determined.

In the early phases of the pump hydraulic design, the designer does calcula-tions to determine the parameters of the design of the impeller and volute or diffuser. These include selection of vane inlet angle, radius of curvature, number of vanes, exit angle, etc.

Note that impeller and volute hydraulic design is beyond the scope of this book. Readers interested in learning more about pump hydraulic design are referred to Refs. [3] through [6] at the end of this book.

With the design parameters selected, the designer can then complete the layout of the impeller and volute or diffuser. This allows creation of pattern and machine drawings for these components. With the completion of the design of the other mechanical components of the pump such as the stuffing box and bearing assembly, a prototype pump can be built and made ready for the performance test.

The manufacturer’s pump performance test is usually conducted with the pump mounted on the floor above a large pit or sump filled with water. The pump takes suction from the sump, the flow passes through instrumentation that can measure flow Q and total head TH, and then the flow is returned back to the sump. (Refer to Chapter 3, Section VI, for a more detailed discus-sion on how total head and flow are measured in the pump test.) The test loop has a throttling valve to allow for variation of the flow and total head so that the pump can be run over its full performance range.

Finally, the manufacturer’s laboratory test facility has the capability to measure BHP, the power required by the pump. This is done in the labora-tory in one of several ways. One common method measures the torque on the shaft between the pump and motor, and converts this to horsepower by the formula:

BHP   = rpm   T×

5250 (2.18)

where T is the torque (in ft-lb).

Dynamometers are also used to measure torque. A more common approach uses electrical instrumentation to measure the input power drawn by the motor at a given flow rate, the wire-to-water horsepower previously dis-cussed. This is then multiplied by the motor efficiency. The motor efficiency is a value that is available from the motor manufacturer. (For most AC elec-tric motors, the efficiency remains unchanged from full load to nearly 50% of full load.) Motor input power times motor efficiency equals motor output

power, and the motor output power thus measured is the pump BHP. This is also the approach that would be used to measure BHP in a field test of a pump, as described in Chapter 3, Section VI.D.

The pump is turned on in the test loop and the throttle valve is set at an arbitrary position. Then, using the laboratory instrumentation, the values of Q and TH are measured, as is BHP using one of the above-described meth-ods. Then, using Equation 2.17 (with SG = 1.0 because the test loop contains water), the value of pump efficiency η is determined for that particular point on the pump curve. The data obtained from this test point (Q, TH, BHP, η) are recorded, and then the throttle valve is repositioned and a new set of data points is taken. This procedure is repeated over the full range of performance of the pump. Usually, a minimum of five to seven points on the pump curve are measured. The data can then be plotted to create the head– capacity, BHP, and efficiency curves for the pump, using a full-diameter impeller.

Figure 2.10 illustrates typical performance curves generated by the perfor-mance test just described. In a typical centrifugal pump, the head–capacity

76

0 200 400 600 800 1000 1200 1400 1600

0 200 400 600 800 1000 1200 1400 1600

0 200 400 600 800 1000 1200 1400 1600

50

H-Q, BHP, efficiency, and NPSHr curves for a pump with a given speed and impeller diameter.

(H–Q) curve (blue curve in Figure 2.10) typically rises toward shutoff, with the pump developing lower flows at higher heads, and vice versa. The horse-power curve (red curve in Figure 2.10) typically is rising as flow increases, although Section VII to follow illustrates that this is not always the case.

Finally, the pump efficiency curve (green curve in Figure 2.10) shows that the efficiency varies with flow, rising to a peak value known as the best efficiency point (BEP).

Also shown in Figure 2.10 are other key landmarks that include design point, shutoff head, runout flow, minimum flow, preferred operating region, and allowable operating region. These terms will be more thoroughly discussed in later sections of this book. Figure 2.10 includes the pump NPSHr curve (brown curve in Figure 2.10). The significance of this curve will be discussed in Section VI to follow.

Note that where color is shown on pump curves throughout this book, the color scheme as shown in Figure 2.10 will be used (blue for H–Q curves, red for BHP curves, green for efficiency curves, and brown for NPSHr curves.

System head curves, which are introduced later in this chapter, will be shown in purple.

It is only after the performance test described above has been completed that the manufacturer finally knows whether the performance objectives for the pump have been reached (i.e., what is the best efficiency flow, head, horsepower, and pump efficiency, and how these values vary over the full range of performance of the pump).

The next step for the manufacturer is to perform additional performance tests with reduced impeller diameters. The pump is disassembled, the impel-ler is machined to a smalimpel-ler diameter, and then the test described above is repeated. Trim increments may be as little as 1/16 in and as much as several inches, depending on the size of the impeller.

After several complete performance tests at different impeller diameters have been conducted by the procedure just described, the manufacturer is able to finally generate the family of curves for the full performance enve-lope of the pump, which is then published in the manufacturer’s catalog. This cataloged family of composite curves for a pump over its range of offered impeller diameters is illustrated in Figure 2.11.

As exemplified in Figure 2.11, some pump manufacturers display the information on BHP and efficiency on their cataloged curves in a differ-ent format from the way Figure 2.10 displays it for a single impeller diam-eter. In Figure 2.11, the BHP and efficiency data are plotted using iso curves (lines of constant BHP and constant efficiency). Other manufacturers simply show a separate efficiency and BHP curve for each impeller trim offered. Either method of displaying BHP and efficiency as a function of pump flow and impeller diameter allows the pump selector to determine the required impeller diameter and motor size for a particular application.

Iso-horsepower lines, when they are used, normally show only the commer-cially available electric motor sizes.

As an example, using the Figure 2.11 pump curve, if the pump design rating is 1800 gpm and 175 ft, the Figure 2.11 curve shows that the required impeller diameter would be just under 14 inches. The motor size can then be chosen using the BHP curves in Figure 2.11. If the pump flow is never expected to exceed the design flow rate of 1800 gpm, a motor size of 100 HP can be chosen.

Note however that this assumes a specific gravity of 1.0. Remember that if a liquid other than water is being pumped, the BHP curve must be adjusted up or down by the specific gravity of the liquid to be pumped.

As Section IX illustrates, in many cases the pump system allows a pump to operate over a wide range on its H–Q curve. Often, particularly in industrial applications, a motor size is chosen so that the pump can operate over the full range of performance at a given diameter, that is, to the end of the curve.

For the example above, this would lead to a selected motor horsepower of 125 HP. This selection of a motor size to allow operation at any point on the pump curve for a given impeller diameter is known as a nonoverloading motor selection, and is considered a good selection criterion by most industrial users. A less conservative approach that is acceptable in many lighter-duty applications selects a motor size that is adequate for the design point, and

Capacity

1600 2000 2400 2800 3200 gpm

1200

Eye area 50 sq. in. Steel 256–116 55437

Typical manufacturer’s published performance curve family for a centrifugal pump operating at a fixed speed and with a range of impeller diameters. (Courtesy of Goulds Pumps, Inc., a subsidiary of ITT Corporation.)

relies on controls or system limitations to keep the pump flow from going beyond that which would overload the motor.

Another approach allows the motor to make use of its service factor. The motor service factor is a design margin used in the design of motors, essen-tially putting more copper in the motor windings to allow the motor to gen-erate more horsepower than the motor is rated for without causing the motor to run excessively hot. Typical service factors for industrial motors are 1.10, 1.15, or 1.2. A 100-HP motor with a 1.15 service factor is actually capable of delivering 115 HP without running so hot that the motor insulation would be harmed or the motor would fail because of excessive heat.

Most conservative industrial users of pumps select motor sizes so that the motor does not make use of the service factor at all (i.e., the motor is chosen to be “nonoverloading” over the entire pump performance range, without mak-ing use of the service factor.) This is especially recommended if the pump is to run continuously. This simply means that the motor service factor lets the motor run cooler than it otherwise would. Many lighter-duty fractional horsepower motors have quite high service factors (e.g., 1.5), and it is quite common with residential pumps and other intermittent service or light-duty commercial and industrial applications for the pump to make use of the motor service factor at some points of normal operation on the pump curve.

It is recommended, when sizing and selecting centrifugal pumps, to choose a pump such that the design duty point (head and capacity) is a small amount to the left of the BEP on the pump curve. The reason for this is that the vast majority of pumps are oversized, due to the conservatism used by the pump selector in arriving at estimates for total head in the system. Because the actual resulting system head is typically less than that predicted by the engineer at the time of the pump selection process, the pump will tend to move to the right on its performance curve, to a point where the total head requirement is less. If the original selection were made to the right of the BEP, the lower than predicted pump total head would tend to move the operating point still further to the right on the performance curve. So, it is preferred to make the initial selection to the left of the BEP, so that, if the actual head is less, the pump will move closer to its BEP as it moves to the right on the curve. This is not a hard and fast rule, but the engineer should bear in mind that if a pump is selected at a point well to the right of the BEP on the perfor-mance curve, and if the actual pump total head is less than predicted, this will allow the pump to move even further to the right, which could lead to problems with overloading the motor or cavitation in the pump.

Engineers who are sizing pumps often ask what is the maximum amount away from the best efficiency point on the pump curve that they should choose a pump to operate. Refer to Figure 2.10, which defines the preferred operating region and allowable operating region. One rule of thumb puts the preferred operating region at between 70% and 120% of the BEP flow (which varies by impeller diameter) for continuous operation. This is a tighter range around BEP than the allowable operating region, which is generally given as

the range defined by the envelope of performance shown in Figures 2.7 and 2.8. The preferred operating range is even tighter than this for larger, high energy pumps, and for pumps with higher suction specific speed, which is dis-cussed in Section VII to follow.

The one set of curves on Figure 2.11 not yet discussed are the NPSHr curves. These are discussed in Section VI below.

VI. NPSH and Cavitation A. Cavitation and NPSH Defined

As stated in the overview of this chapter, NPSH or net positive suction head is probably the most misunderstood aspect of pump hydraulics. It is very important to understand this concept because NPSH problems are among the most common causes of pump failures, and are often mistakenly blamed for failures that are completely unrelated.

NPSH must be examined when using centrifugal pumps to predict the pos-sibility of cavitation, a phenomenon that has both hydraulic and sometimes destructive mechanical effects on pumps. Cavitation, illustrated in Figure 2.12, is a phenomenon that occurs when vapor bubbles form and move along the vane of an impeller. (What causes the vapor bubbles to form in the first place is discussed shortly.) As these vapor bubbles move along the impel-ler vane, the pressure around the bubbles begins to increase. (Figure 1.5 in Chapter 1 shows that the local pressure increases as the flow moves along

Rotation

Collapsing bubbles Vapor bubbles

FIGURE 2.12

Cavitation occurs when vapor bubbles form and then subsequently collapse as they move along the flow path on an impeller.

the path of the impeller vane.) When a point is reached where the pressure on the outside of the bubble is greater than the pressure inside the bubble, the bubble collapses. It does not explode, it implodes. This collapsing bubble is not alone, but is surrounded by hundreds of other bubbles collapsing at approximately the same point on each impeller vane.

The phenomenon of the formation and subsequent collapse of these vapor bubbles, known as cavitation, has several effects on a centrifugal pump. First, the collapsing bubbles make a distinctive noise that has been described as a cracking or popping or rattling sound, or a sound like the pump is pump-ing gravel. This can be a nuisance in an extreme situation where a cavitatpump-ing pump is operating where people are working. This physical symptom is usu-ally the area of least concern with cavitation, however. Of far greater concern is the effect of cavitation on the hydraulic performance and the mechanical integrity of the pump.

The hydraulic effect of a cavitating pump is that the pump performance drops off of its expected performance curve, referred to as break away, as illustrated by Figure 2.13, producing a lower than expected head and flow.

An even more serious effect of cavitation is the mechanical damage that can occur due to excessive vibration in the pump. This vibration is due to the uneven loading of the impeller as the mixture of vapor and liquid passes through it, and to the local shock wave that occurs as each bubble collapses.

The shock waves can physically damage the impeller, causing the removal of material from the surface of the impeller. The amount of material removed varies, depending on the extent of the cavitation and the impeller material.

If the impeller is made of ferrous-based material such as ductile iron, mate-rial is removed from the impeller due to a combination of corrosion of the ferrous material from the water being pumped and the erosive effect of the cavitation shock waves. If the impeller material is more corrosion resistant

TH(ft)

Q (gpm) Break away

FIGURE 2.13

Material loss from impeller vane due to cavitation.

but softer, ordinary bronze, for example, the damage that cavitation causes is similar to a peening operation, in which a piece of relatively soft bronze is repeatedly struck with a small ball peen hammer. Materials such as 316 stainless steel, with superior corrosion resistance and ability to work harden under the peening action, have a better ability to resist the metal loss associ-ated with cavitation.

In any case, the removal of material, if it occurs at all, proceeds as long as the pump is cavitating. Pits can be formed gradually on the impeller vanes and, in the extreme, the removal of material can actually cause a hole to be eaten clear through an impeller vane, as Figure 2.14 illustrates. This removal of material from the impeller has the obvious effect of upsetting the dynamic balance of the rotating component. The result is similar to what happens if an automobile tire is not properly dynamically balanced, or if it loses one of the balance weights, causing excessive vibration.

It is very important to remember that excessive vibration from cavitation can occur even without the material loss from the impeller described above.

This is true because the vibration from cavitation is caused by the uneven loading of the impeller and the local shock wave, as mentioned previously, as well as by the removal of material.

Often, the excessive vibration caused by cavitation subsequently causes a failure of the pump’s seal and/or bearings. This is the most likely failure mode of a cavitating pump and the reason why NPSH and cavitation must be properly understood by the system designer and pump user.

What causes the formation of the vapor bubbles in the first place, without which the cavitation would not have a chance to occur? To a person who

FIGURE 2.14

Material loss from impeller vane due to cavitation.

has never studied thermodynamics, the most obvious way to create vapor

has never studied thermodynamics, the most obvious way to create vapor

In document Volk, Michael W-Pump Characteristics and Applications-CRC Press (2014) (Page 97-112)