Duan C.G., Karelin V.Y. Abrasive Erosion and Corrosion of Hydraulic machinery

Подождите немного. Документ загружается.

This page is intentionally left blank

Chapter 6

Interaction between Cavitation

and Abrasive Erosion Processes

V. Ya. Karelin, A. I. Denisov and Y.L. Wu

6.1 Effect of Suspended Particles on Incipient and

Developed of Cavitation

In the study of the joint effect exerted by suspended particles and cavitation

on working surface, there appear two problems of significant theoretical and

practical interest. The first problem refers to determination of the effect

produced by suspended matter on origin of cavity and development of

cavitation in a liquid flow. The second problem deals with defining of

cavitation-abrasive erosion intensity decided by ratio of these two

deteriorating processes, each taken individually. Both problems have not been

solved yet due to insufficiency of experimental data obtained and the

difficulties in defining the influence produced by each of these processes

separately in hydraulic machines.

The presence of sediments suspended in a flow results in alteration of the

physical-chemical properties of this flow and, specifically, of its volumetric

strength, which decides the ability of such a liquid to resist interruption of its

discontinuity. In accordance with the theory, assuming existence of something

like "weak points" or "nuclei" of cavitation in any volume of real liquid

[6.1],

the presence of solid particles in water should encourage much earlier

emergence of cavitation and rapid development of the latter. V. Iterson was

the first to confirm this phenomenon and to measure experimentally

development of accelerated cavitation of water saturated with carbon powder

315

316

Abrasive Erosion and Corrosion of Hydraulic Machinery

in a Venturi tube. With powder content of 0.5 g/1 the pressure in the tube was

15%

higher at cavitation incipience than in case of pure water. The same

effect produced by solid particles is described in

[6.2],

when the cavitation

developed was studied in the region behind a cylindrical model placed in a

water tunnel.

To investigate this interesting problem a flat-shaped Venturi nozzle was

used [6.3] which enabled the researchers to consider the basic factors

influencing the intensity of hydroabrasive, cavitation and cavitation-abrasive

erosion patterns: flow velocity, pressure values at the input and output,

expansion angle of the diffusion region, different phases in development of the

cavitation, abrasive particle sizes and their concentration, nature of the

damage on surface being worn, temperature of the mixture etc.

Model 8 was installed in the modified section of the pressure piping

constituting a part of the stand shown in Figure 6.1. The throat cross-

sectional area in the Venturi tube having dimensions 30 x 30 mm is formed by

two brass guides, the first one is placed parallel to the axis of the flow

direction and the second one is positioned at angles of 9 = 5°, 10°, 15° and

20°.

The transparent side walls allow observation of the moment of cavitation

incipience, to follow its development and to measure precisely the length of

the cavitation area. By use of a pump model 4K-8 it was possible to set the

flow

rate,

in the throat of the Venturi tube, ranging up to 32 m/s, which is

entirely in agreement with the liquid flow behaviour taking place within the

flow-passages of modern pumps.

The experiments have shown that cavitation emergence occurs somewhat

earlier, when using water with solid particles, and in doing so the advance of

this cavitation origin in hydro-mixtures, as compared with pure water,

increases with a rise of the sediment concentration

[6.3].

With the solid

particle content being p

—

g/1,

this advance within the throat of the nozzle

is equal to about 15% of the flow velocity (or flow rate). When the sediment

concentration is smaller than the indicated level, the effect of the solid

particles on incipient cavitation is less evident. As an example, Figure 6.2 (a)

illustrates the change in the cavitation area according to flow rate in the

Venturi nozzle

Q/Q

(here Q

is flow rate at the moment of cavitation origin

in pure water), at the angle of 0 = 5° input pressure being

inp

= 42 m and

sediment concentration p- 16

g/1.

This suggests that in case of hydromixture,

a smaller flow rate

Q/Q

is required for development of, the same 5m long

steady cavitation zone than in the event of pure water. In other words, the

Interaction between Cavitation and Abrasive Erosion Processes

317

presence of sediments serves to speed up cavitation emergence. In the course

of further developed cavitation its length goes on to increase. However, the

effect of

the

solid particles present in the flow gradually subsides. When the

length of this cavitation zone comes to l

= 110 mm, the both curves are

joined to form one curve (see Figure 6.2 a).

nso

nsd

1-pump,

2-intake

valve,

3-input

pipe,

4-by-pass

pipe,

5-inlet

tank,

6-level

indicator, 7-particles' input,

8-Venturi

tube,

9-outlet

valve,

10-tube

Figure 6.1 Test stand for analysis of hydroabrasive erosion

in centrifugal pumps

ass to w no Q/Q

(solid

line:

pure water; dashed

line:

mixture with sand content p

= 16

g/1)

Figure 6.2 a Alteration of cavitational zone length

versus flow rate of a Venturi tube

318

Abrasive Erosion

and

Corrosion

Hydraulic Machinery

Figure 6.2 b illustrates, for the same model, change of the cavitational

zone length versus sediment content for given the cavitation number K,

derived in accordance with the known formula (6.1), i.e.

K =

EIZ£L

(61)

1 2

1« (mm)

K=0,36

ZL0A2'

"-0AS

•-0S0

"-0J5V

K=a57

_.._. I

0 10 20 30 W 50 p(g/l)

Figure 6.2 b Effect of sediment matter concentration on cavitational zone

length at constant value of cavitation number K

It is inferred from the diagrams that with the flow being saturated up to

10 ~ 15 g/1 the cavitation zone length l„ increases from 5 to 35 mm at the

initial stages of cavitation development (K = 0.57). The further rise of the

sediment content results in certain drop of the l

value. Although these terms

correspond to concentrations which are much higher than maximum values

typical of natural water sources, it should be noted that the situation

considered changes is most likely characterized by composite physical

processes associated with changes of surface tension, viscosity, density,

frictional forces and inertia, with the hydromixture being saturated by

abrasive particles.

In transition to the more developed phases of cavitation (with decrease of

AT values) the stimulating effect of solid particles starts to reduce and when the

cavitation number becomes K

0.36 the presence of sediments in the flow

having concentration of up to 50 g/1 fails to determine the cavitational zone

sizes.

It is most likely that disturbance of the liquid continuity, under the

Interaction between Cavitation and Abrasive Erosion Processes

319

conditions of cavitation being strongly developed, is accompanied by the

strength much greater than forces governing the liquid strength at the moment

of its separation from a particle. Therefore, at this phase of cavitational

development the solid particles fail to affect the cavitation.

The cavitation performance of centrifugal pumps using hydromixtures

with solid particle density p

exceeding the liquid density p

is amenable to

theoretical treatment

[6.1,

6.4].

It is most simple to assume that the mixture under consideration, at the

leading edges of the impeller, acts as homogeneous liquid having density p

Such assumption holds true if the velocities of solid particles and those of the

liquid are the same throughout the flow-passages in a pump. Since this is not

the case, let us find out how the lag of the solid particles moving in the

accelerated flow can affect the distribution of pressures in the flow direction.

Let us consider the hydromixture flow in the region restricted by two

sections 0-0, upstream of the entrance leading to the impeller blade, and 1-1

positioned in the region of the minimum pressure. Let us also assume that,

before the entrance to the blade, the velocities of

the

solid particles and those

of the liquid are the same and equal to w

. Within section 1-1 the liquid

velocity has become equal to Wi, and that of the solid particles is equivalent to

. Using the equation for energy, as applied to the hydromixture mass

restricted between the cross-sections under review, and neglecting the losses,

we shall have

where Q

, QT, and Q

are respectively flow rate of the hydromixture, solid

particles and water.

Dividing the right and left hand sides of the equation (6.2) by the product

we shall obtain

o-P\

PTQT

IT ~

o , PoQo

f -Wo

Pr8 PrQr

PrQr

320

Abrasive Erosion and Corrosion of Hydraulic Machinery

Taking into account that Q

= Q

^ *

better to present

equation (6.3) in a form more convenient for analysis

P - P

Prg

■w:

PTQT

PrQr

■w

\w,

(6.4)

It is seen from the expression obtained that the differential pressures

(heads) emerging in-between the cross-sections under consideration are equal

to alterations of velocity heads, provided only that the velocities of solid

particles w

and those of the water wi are the same. In reality, w\

<w\

due to

the action of head resistance forces, which result in additional drop of

pressure within the cross-section 1-1 and hence constitutes a great threat of

cavitation emergence.

Even if we assume that, with small sediment concentrations and without

particles of large sizes, as typical of natural waterways, W\

w;, when the

sediment carrying stream flow s around the blade leading edge, the additional

drop of pressure based on equation (6.4), i.e. Ap = p

- p

]

will be the same

as in case of pure water motion. However, the absolute pressure decrease in

this case, in the region of the minimum pressure corresponding to the greatest

hazard possible from the incipient cavitation viewpoint, will be greater than in

the event of pure water transfer, since

Prg

2 2

**L AP

=AP

Pog Po

Hence, cavitation in pumps dealing with hydromixtures begins earlier than

it does in case of pure water.

The conclusion based on theoretical data is confirmed by the results

obtained during experimental analyses and full-scale test observations. Figure

6.3 illustrates alteration of

the

feed at the moment of incipient cavitation in a

centrifugal pump (impeller dia. is 268 mm, rotational speed is 1750 rev/min)

as a function of the hydromixture density p

. The increase in density

apparently encourages earlier emergence of cavitation and can consequently

become the cause for intensity rise of the cavitational erosion.

Interaction between Cavitation and Abrasive Erosion Processes 321

It should be noted that the influence of density upon cavitation emergence

becomes noticeable at the concentration levels of solid particles which

significantly exceed the sediment concentration existing in natural reservoirs.

Consequently, for conventional pumps, the change in cavitation incipient

conditions, due to presence of solid particles, seems hardly probable. But for

custom-designed pumps, for transferring concentrated hydromixtures, this

should be taken into account.

1300

1200

1100

1000

65 70 75 Q(l/s)

Figure 6.3 Dependence of critical (incipient cavitation) flow rate

of a pump on density of transferred hydromixture

The problem of defining the total erosion intensity, including both

cavitation and erosion effects of

sediments,

has a number of unsolved issues.

In the process of cavitation-abrasive erosion the joint hydrodynamic

forces acting on the surfaces, hydraulic machines; are from bursting of

cavitation bubbles and collision of sediment particles contained in the water.

Because of difference in behavior of these two processes, dependent on the

cavitation development degree, concentration of sediments contained in the

water and erosion resistance of the material for operating components of the

pump, such joint cavitation-abrasive erosion process will follow rules other

than those described earlier.

The data obtained in multiple experimental studies indicate that there is a

large number of alternatives in cavitational-abrasive erosion development.

322

Abrasive Erosion and Corrosion of Hydraulic Machinery

Depending on a number of factors its intensity can either rise or drop, as

compared with pure cavitational erosion.

In our researches aimed at studying the mechanism of the joint

cavitational-abrasive erosion taking place in the diffusion area of a planar

Venturi nozzle, the specimens used were fabricated from silumin An - 9B and

had the working surface of 120 x 30 mm. The erosion was evaluated by loss

in the mass of these specimens after 6 hours of testing. The tests were

performed entirely with respect to the same phase of cavitational development

defined by the cavitational zone of 110 mm in length. In this case the closure

of the cavitational zone under review took place within the limits of the

specimen. The hydroabrasive erosion conditions were created through

suppressing the cavitation by means of raising the general pressure level in the

installation system.

Figure 6.4 illustrates the effect produced by divergence angle 6 of the

nozzle diffusive area upon the erosion level in pure water and in water

containing sediment matter (p = 5 g/1, diameter of particles d

0.25 ~ 0.5

mm) with the cavitational zone length being l

= 110

mm.

The presented data,

in addition to characterizing the specific design features of the Venturi nozzle

model used in these experiments, help to form conclusions concerning the

effect of the by-passed surface shape on the erosion process. With the

increase of angle 6, a gradual transition takes place from the surface

cavitation to the hiding-off form of cavitation. While with 6 = 5° the

cavitational zone covers the whole surface of a specimen and collapses within

its boundaries, at the angle values corresponding to 6 >5 ° there begins to

emerge, at the end area of the cavity, fluctuating flares which disappear

without contacting the specimen surface. With further increase of

angle

and

the permanent length of the cavitation zone being preserved, the thickness and

volume of this zone become larger, while the intensity of

the

action produced

by the spherical impact wave in the specimen surface diminishes, which is in

agreement with principles of the cavitation erosion theory. Moreover,

enlargement of the cavitational volume corresponds, strictly speaking, to an

increase in development cavitation degree and hence, to a decrease of the

cavitation erosion intensity, which is known to have its maximum level at the

initial phases of the cavitation development.

Reduction of the cavitation-abrasive erosion, due to increase of angle 0,

causes enlargement of cavitation zone sizes. But the latter moves away from

Interaction between Cavitation and Abrasive Erosion Processes

323

the deteriorated surface and deforming action on the solid particles

penetrating through the depth of this zone.

AG(mg)

600

too

ZOO

0 5 tO 15 20 Of)

cavitation erosion; 2. cavitation-abrasive erosion

Figure 6.4 Effect of Venturi nozzle diffusion angle 0on erosion loss AG

Figure 6.5 shows specimen mass loss versus test duration at angles 0=5°

and

15°

for two modes of erosion processes. The erosion line obtained during

the cavitation process possesses several characteristic bends, which suggests

for a certain incubation period, as well as for alterations in erosion intensity.

During the combined effect produced by both cavitation and sedimentation,

the bends in the erosion curve become smoother and the relationship comes

closer to linear, the same as in the process of hydroabrasive erosion. But still,

the absolute combined erosion level is much higher than the erosion from the

cavitation effect and that of the sediments matter, exerted individually.

Therefore, the test results given in Figure 6.5 are presented in a

semilogarithmic coordinate system.

Sharp multiple amplification of the cavitation-abrasive erosion, as

compared with separate forms of failure and their algebraic sum, has been

experimentally demonstrated earlier [6.1, 6.2]. The surface reacts with the

flow due to formation of water microstreams of spherical impact waves

containing solid particles. Moreover, due to heavy velocity fluctuations

accompanying development of cavitation, the abrasive particles present in the

area of the cavitational zone acquire additional accelerations, thus enhancing