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

Подождите немного. Документ загружается.

This page is intentionally left blank

Chapter 5

Erosion-Resistant Materials

M. Matsumura and B. E. Chen

5.1 Selection of Erosion-Resistant Materials

M. Matsumura

For hydraulic machine (turbine and pump) working in sand particles laden

water, the selection of abrasive erosion resistant structural material or of

protective surface coating is one of the most important issues in their design.

For this the following points, although which make the selection quite

difficult, should be considered:

(1) A material may show different sand erosion resistance for different

erosion conditions.

(2) Test results obtained under accelerated erosion conditions in

laboratory cannot be applied extended to real machines.

The erosion resistance of material will be discussed in the following

sections successively.

5.1.1 Multi-posion Test of Erosion Resistance in a Real Water

Turbine

A systemic research on the selection of erosion-resistant materials was made

Prof.

Duan Chang Guo and his collaborators. They conducted a multi-

position scheme of erosion testing to demonstrate that one material may show

different silt erosion resistance at different positions in a water turbine

because the components of the hydraulic machinery have different erosion

235

236

Abrasive Erosion and Corrosion of Hydraulic Machinery

condition parameters such as the velocity and turbulence of flow. The

material specimens were installed at all main components of flow passages of

the water turbine, which used to be suffered from silt erosion, as follows:

Outlet edge of runner blades;

Outlet edge on water turbine's guide vanes;

Plane of guide vane;

Labyrinth ring (wearing ring);

Inlet wall of draft tube;

6. Wall of spiral case;

as shown in Figure 5.1.

Figure 5.1 Positions where test specimens were installed

The parameters of the test turbine are shown in Table 5.1. 15 kinds of

metallic and non-metallic materials shown in Table 5.2 were tested.

The testing materials include conventional cast stainless steels (SCSI,

SCS5 and SCS13); low carbon cast steel (SC46); high nickel, high

manganese chrome stainless steels (NMC and NMW); steel plates (SM4 and

SM58Q); high manganese carbon steel (DFME); plasma spray coatings of

ceramics of aluminum oxide (AL-203) and tungsten carbide (WC);

Synthetic rubber coatings (S-61 and L-2710) and protective resin paint

coatings (EPX and PUT).

Erosion - Resistant Materials

237

Table 5.1 Parameters of the test turbine

Type

Maximum water head

Desing water head

Turbine output

Flow discharge

Rotating speed

Draft head

Dia. of runner

Vertical shaft Francis turbine

71.1m

61.0m

8,800

17.1 m

428.6 rpm

1.0m

1,400 mm

Table 5.2 Materials tested

No.

Material

SCSI

SCS5

SCS13

SC46

NMC

SM41

SM58Q

NMW

DFME

AL-203

S-61

L-2710

EPX

PUT

0.10

0.035

0.040

0.19

0.042

0.17

0.10

0.07

0.14

0.39

0.47

0.99

0.40

0.43

0.20

0.24

0.35

0.56

Chemical composition (%)

Mn P S Ni

0.72

0.68

1.00

0.78

2.54

0.85

1.34

5.60

17.05

0.022

0.020

0.023

0.020

0.023

0.014

0.010

0.017

(Ceramic of aluminum oxide)

(Ceramic of tungsten carbide)

(Synthetic rubber)

(Epoxy resin paint)

(Polyurethane resin paint)

0.013

0.012

0.010

0.009

0.013

0.006

0.003

0.001

0.89

4.01

9.54

5.56

0.20

6.50

2.16

12.24

12.08

19.52

13.27

0.07

12.40

15.19

0.34

0.29

0.60

0.12

1.24

The operating period of the test turbine was divided into three stages.

Total testing duration was 1779 hours. To obtain consecutive data of the

238

Abrasive Erosion and Corrosion of Hydraulic Machinery

variation of the silt content, water was sampled daily from the inlet of the test

turbine. The average silt content was 1-3 kg/m

; however, it reached 18.95

kg/m

maximum in a flood season. Grain size classification and mineral

composition analysis were conducted. Though the results showed some scatter

depending on the sampling date, the average values are shown in Table 5.3.

Table 5.3 Properties of silt

Grain size (|xm)

Range <53

Average 27

53-100 100-208 208-295 295-351

Mineral composition [relative content (%)]

Quartz 15

Feldspar 20

Carbonate 40

Augite & hornblende 5

Chlorite 10

154

252

323

>351

426

Mica remainder remainder remainder remainder remainder remainder

Weight loss of each specimen was measured by a precision balance scale

at every inspection, and the average depth of erosion damage was calculated

for each specimen by the following formula:

h =

10AG/yS

(5.1)

where,

average eroded depth [mm ].

AG: weight loss of specimen at each inspection [g].

specific weight of the test material [g/cm

S: surface area of the test specimen subjected to silt erosion

[cm

Relative index of abrasive erosion resistance is defined by the following

formula, in which abrasive erosion resistance of cast carbon steel SC46 is

taken as a reference value, since it is most commonly used for water turbine

components.

Erosion - Resistant Materials

239

/h (5.2)

where,

average eroded depth as calculated by equation (5.1) for each

specimen at different position of the turbine [mm].

the same for SC46 specimen at each position of the turbine [mm].

e. relative index of abrasive erosion resistance for each material at

each position of the turbine.

An s larger than 1.0 means that the material is more resistant against silt

erosion than the SC46. Table 5.4 shows the value of relative erosion index of

final measurement for each specimen at each test position

Table 5.4 Relative abrasive erosion indices

^\Position

Material s

SCSI

SCS5

SCS13

SC46

NMC

SM41

SM58Q

NMW

DFME

AL-203

S-61

L-2710

EPX

PUT

Runner

blade

1.91

1.09

1.00

1.42

1.44

0.93

Guide

vane

1.97

1.01

1.17

1.00

1.24

1.27

0.92

3.06

2.59

Bottom

ring

0.46

0.53

0.40

1.00

0.88

1.42

1.43

1.18

2.42

Wearing

ring

1.24

0.76

0.93

1.00

2.44

Draft

tube

1.00

1.43

1.45

6.61

= 0

Spiral

case

1.00

1.03

0.68

7.29

8.67

= 0

0.30

According to Table 5.4, a set of specimens of the same material does not

show same degree of relative erosion resistance at different test positions. For

240

Abrasive Erosion and Corrosion of Hydraulic Machinery

example, all the relative erosion resistance indices for cast stainless steels of

SCS5 and of SCS13 on both runner blades and guide vane exceeds 1.0. This

means they are both more resistant to silt erosion than SC46. However, those

of the specimens on wearing ring and on bottom ring are far below 1.0

showing that is their erosion-resistance is poorer than SC46 at these

positions. On the other hand, the relative erosion resistance indices of DFME

for runner and guide vane positions are less than 1.0, while that for bottom

ring position exceeds 1.0.

Prof.

Duan and his collaborators attributed the aforementioned results to

differences in erosion type with different hydroabrasive erosion conditions.

Due to the difference in flow velocity and in impinging angle of sand particles

the erosion conditions were classified as shown in Table 5.5.

Table 5.5 Classification of silt erosion

Test position

Runner blade

Guide vanes

Wearing ring

Cheek crevice

plate

Spiral case

Draft tube

Flow velocity

High

(22 ~ 27 m/s)

High

(18 m/s)

very high

low

Impinging angle

Small

Larger in cavitation area

Small

Larger near bottom end

Larger, due to vortex &

turbulent

Larger, due to vortex &

turbulent (*)

Small

Larger near inlet

Type of erosion

III

(*): Test specimens were placed under the bottom end of the guide vanes

at their open position, where the classification is as the type III.

Figure 5.2 shows the relative erosion resistance indices of six different

metallic materials for two different test positions with different erosion

types,

the type II and the type III. From this figure, it can be seen that the

Erosion - Resistant Materials

241

relative erosion resistance indices of these materials fall on the opposite

sides of the line of 1.0 for the two different testing positions. It is

concluded that the order of the materials in silt erosion resistance varies

depending on the condition and the type of erosion attack.

SCSI

SCS13

SC46

NMC

DFME

NMW

0 1 2

COEFFICIENT

Figure 5.2 Relative erosion indices for two different positions

5.1.2 Laboratory Erosion Tests

The objects of laboratory erosion tests are generally

(1) Ranking the materials according to their resistance to erosion, or

(2) Estimation of the service life of a material, which is used for the

construction of a hydraulic machine.

Measurement of a relative index proportional to the resistance, s cited in

the previous section for example, will accomplish the first object. An absolute

numerical measurement, mm/year for example, will be required for the second

object. In other words, a laboratory test with the second object is of a higher

grade than the one with the first object. However, it has often been

experienced by designers and engineers that any experimental result obtained

by these laboratory erosion tests cannot be applied to the real machine. Even

the ranking of materials hardly coincides with the order of their service life in

the field. Reasons for this discrepancy are discussed below.

The discrepancy between a laboratory test result and the performance of

the material in the field is totally attributed to the fact that the extent and

Bottom Ring/

r\""

*""■>„,

\Runncr Blade

242

Abrasive Erosion and Corrosion of Hydraulic Machinery

nature of the damage which is caused by the impingement of solid particles on

a surface strongly depends on the impact angle, on the shape of the particle,

on the impact velocity and on the material of the surface.

Figure 5.3 shows typical dependencies of erosion on impact angle for a

ductile material (mild steel, aluminum etc.) and a brittle material (quenched

steel, ceramic etc.). Here erosion is defined as the mass of material removed

per unit mass of impinging particle. It should be noted that the ductile

material exhibits the most severe damage at impact angles lying typically in

the range 20° to 30°. In contrast with this, the brittle material is damaged

most rapidly under conditions of large impingement angle, and minimally at

shallow impact angles. Suppose particles were directed at a right angle

against a test specimen in a laboratory test, a brittle material would suffer

heavily and a ductile material slightly. This test result will nominate the

ductile one for the construction material of a real machine. However if it were

used for a pneumatic powder transport pipe wall, where solid particles

impinge at shallow angles, a larger damage will result on the ductile material

than on the brittle one. Thus, from this example it can be seen that the

performance of the materials in the field is often clearly opposite to the

indication of the laboratory test.

O 30 60 90

Impingement angle a (deg.)

Figure 5.3 Schematic representation of erosion rate on impingement angle

Erosion - Resistant Materials

243

Then, how about making the impact angle of particles in tests coincides

with that in real machines, or conducting the erosion test at every impact

angle? This would be successful provided that the impact angle in a real

machine could be precisely predicted. The problem with this is that the

particle impact angle is not clear in nearly all of laboratory testing machines.

It will be surely much harder to predict it in a real machine.

Even if the prediction and the measurement of impact angle were

successful in both the machines, the test results will still be far from

predicting the real damage so long as exactly the same particles with those in

the field were not used in the test.

Figure 5.4 shows the results of two series of erosion tests which were

conducted by Sheldon and Finnie 1966, under the same impact velocity on the

same group of materials but using two different grit mesh sizes of 120

(particle size of 127 urn in diameter) and 1000 (9 urn). If we make a

comparison between the hardened steel and the plate glass in their resistance

to erosion caused by the impact of 127 um particles (Figure 5.4 (a)), the

former will be naturally judged more resistant to erosion than the latter

because it sustains a smaller erosion at every impact angle. In contrast with

this,

in the erosion test of 9 um particle resulted the typical variation for the

same materials now behaving in a ductile or more ductile manner in which

erosion reached a maximum at an angle considerably less than 90 deg, and the

hardened steel sustained a far larger erosion over all impact angles as

compared with the plate glass. Thus, the conclusion from the result of

laboratory test using 127 urn particles, would be the very reverse of the

performance of the materials in a real machine if it were engaged with 9 um

particles. Some larger particles than those used in field are often adopted for

laboratory tests in order to shorten the test duration through accelerating the

erosion rate of the test specimen. The test results obtained under such

accelerated test conditions cannot be extended to real machines even if the

measurements were conducted at every impact angle.

How about now using the particles obtained in field for laboratory tests,

and measuring the erosion rate at every impact angle? Under such test

conditions the erosion rate of test specimens will be reduced to nearly the

same level as those in field, and far longer test duration, say weeks or months

possibly, will be required to cause some measurable weight losses to the

specimen. This is not an exaggeration of the problem; the reduction of the

particle diameter from 127 to 9 urn resulted in the damage rate being reduced