ASM Metals HandBook Vol. 17 - Nondestructive Evaluation and Quality Control

Подождите немного. Документ загружается.

Fig. 5 Surface crack in a boiler tube. Comparison of the (a) actual microstructure and (b) the repl

ica of the

crack

Creep Damage. Creep defects cause the majority of failures in power plant components operating under stress and

thermal load, and the replica method is especially suitable for the detection of these defects. Therefore, the replica method

has become an especially important tool in the determination of remaining life in such components as boiler tubes, steam

piping, and turbine components. The replica method reveals defects due to creep at a much earlier stage than other NDE

techniques. Creep defects begin as small holes or cavities at grain boundaries or second phases. With time and stress,

these holes or cavities can link up and form cracks that eventually lead to failure of the component (Fig. 6). Creep cracks

are usually very localized, and they form in welds, bends, or other highly stressed regions. Determining the remaining life

of components normally depends on assessments of regular inspections, as indicated in Table 2. Figure 7 shows a

comparison of creep voids in a surface replica and the corresponding bulk microstructure.

Table 2 Creep damage classification

Class

Nature

Action

1 No creep defects

None

2 A few cavities

Reinspection after 20,000 h of service

3 Coalescent cavities

Reinspection after 15,000 h of service

4 Microscopic creep cracks

Reinspection after 10,000 h of service

5 Macroscopic creep cracks

Management must be informed immediately

Source: Ref 11

Fig. 6

Schematic of creep crack formation. Small cavities (a) link up over time (b) and form intergranular

cracks (c) and eventually macrocracks (d)

Fig. 7 Comparison of creep voids in (a) a replica and (b) the actual microstructure

Precipitate Analysis. The detection of various deleterious precipitates in components subjected to high temperature

and stress can lead to improved life assessment analysis of these components. The extraction replication technique is an

excellent nondestructive method of detecting these precipitates.

Sigma phase is a deleterious FeCr compound that can form in some stainless steels, and its presence can severely limit

remaining life. Extraction replicas have been used to determine the amount of σ phase in the microstructure (Ref 12), and

the amount of phase has been directly related to the creep rate (Ref 13). Figure 8 shows an example of phase in an

extraction replica.

Fig. 8 Comparison of σ-phase formation as seen in (a) a replica and (b) the actual microstructure

The composition of carbides, and their stability with time and temperature of exposure, can indicate the remaining life of

a component. Extraction replicas have been used to evaluate carbides, and it has been suggested that changes in

morphology and chemistry can be used to assist the estimation of effective exposure temperature for use in determining

the remaining life of components (Ref 14). Figure 9 shows an example of precipitates extracted from a 200,000-h exposed

sample, together with the accompanying chemical analysis.

Fig. 9 Ex

traction replica of the microstructure (a) and the precipitate microchemical analysis (b) from an

extraction replica

References cited in this section

10.

P.B. Ludwigsen, Non-Destructive Examination, Structure, Sept 1987, p 3

11.

B. Neubauer and U. Wedel

, NDT: Replication Avoids Unnecessary Replacement of Power Plant

Components, Power Eng., May 1984, p 44

12.

F. Masuyama, K. Setoguchi, H. Haneda, and F. Nanjo, Findings on Creep-

Fatigue Damage in Pressure Parts

of Long-Term Service-Exposed Thermal Power Plants, in

Residual Life Assessment Nondestructive

Examination and Nuclear Heat Exchanger Materials, PVP-Vol 98-

1, Proceedings of the Pressure Vessels

and Piping Conference, American Society of Mechanical Engineers, 1985, p 79

13.

T. Fushimi, "Life Evaluat

ion of Long Term Used Boiler Tubes," Paper presented at the Conference on

Boiler Tube Failures in Fossil Plants (Atlanta), Electrical Power Research Institute, Nov 1987

14.

A. Afrouz, M.J. Collins, and R. Pilkington, Microstructural Examination of 1Cr-0.5

Mo Steel During Creep,

Met. Technol., Vol 10, 1983, p 461

Replication Microscopy Techniques for NDE

A.R. Marder, Energy Research Center, Lehigh University

References

1. "Standard Practice for Production and Evaluation of Field Metallographic Replicas," E 512-87,

Annual

Book of ASTM Standards, American Society for Testing and Materials

A.M. Bissel, B.J. Cane, and J.F. DeLong, "Remanent Life Assessment of Seam Welded Pipework," Paper

presented at the ASME Pressure Vessel and Piping Conference, American S

ociety of Mechanical

Engineers, June 1988

3. G.F. Vander Voort, Metallography: Principles and Practice, McGraw-Hill, 1984

4. T.S. Hahn and A.R. Marder, Effect of Electropolishing Variables on the Current Density--

Voltage

Relationship, Metallography, Vol 21, 1988, p 365

5. M. Clark and A. Cervoni, "In Situ Metallographic Examination of Ferrous and Non-

Ferrous Components,"

Canadian Electrical Association, Nov 1985

6. J.F. DeLong, private communication

7. D. Kay, Ed., Techniques for Electron Microscopy, Blackwell Scientific Publications, 1965

8. J.W. Edington, Practical Electron Microscopy in Materials Science, Van Nostrand Rheinhold, 1976

9. G.N. Maniar and A. Szirmae, in Manual of Electron Metallography Techniques,

STP 547, American

Society for Testing and Materials, 1973

10. P.B. Ludwigsen, Non-Destructive Examination, Structure, Sept 1987, p 3

11.

B. Neubauer and U. Wedel, NDT: Replication Avoids Unnecessary Replacement of Power Plant

Components, Power Eng., May 1984, p 44

12. F. Masuyama, K. Setoguchi, H. Haneda, and F. Nanjo, Findings on Creep-

Fatigue Damage in Pressure

Parts of Long-Term Service-Exposed Thermal Power Plants, in

Residual Life Assessment Nondestructive

Examination and Nuclear Heat Exchanger Materials, PVP-Vol 98-1, Proceedings of t

he Pressure Vessels

and Piping Conference, American Society of Mechanical Engineers, 1985, p 79

13.

T. Fushimi, "Life Evaluation of Long Term Used Boiler Tubes," Paper presented at the Conference on

Boiler Tube Failures in Fossil Plants (Atlanta), Electrical Power Research Institute, Nov 1987

14. A. Afrouz, M.J. Collins, and R. Pilkington, Microstructural Examination of 1Cr-

0.5Mo Steel During

Creep, Met. Technol., Vol 10, 1983, p 461

Leak Testing

Revised by Gerald L. Anderson, American Gas and Chemical Company

Introduction

LEAK TESTING is the branch of nondestructive testing that concerns the escape or entry of liquids or gases from

pressurized or into evacuated components or systems intended to hold these liquids. Leaking fluids (liquid or gas) can

penetrate from inside a component or assembly to the outside, or vice versa, as a result of a pressure differential between

the two regions or as a result of permeation through a somewhat extended barrier. Leak testing encompasses procedures

for one or a combination of the following:

• Locating (detecting and pinpointing) leaks

• Determining the rate of leakage from one leak or from a system

• Monitoring for leakage

Leak testing is increasing in importance because of the rising value of, and warranties on, manufactured products and

because of the constantly increasing sensitivity of components and systems to external contaminants. Environmental

concerns are causing additional emphasis on leak testing and its conduct.

Leak Testing

Revised by Gerald L. Anderson, American Gas and Chemical Company

Leak Testing Objectives

Like other forms of nondestructive testing, leak testing has a great impact on the safety and performance of a product.

Reliable leak testing decreases costs by reducing the number of reworked products, warranty repairs, and liability claims.

The most common reasons for performing a leak test are:

• To prevent the loss of costly materials or energy

• To prevent contamination of the environment

• To ensure component or system reliability

Terminology. The following terms must be understood in their strict definitions within the field of leak testing:

• Leak: An actual through-

wall discontinuity or passage through which a fluid flows or permeates; a leak

is simply a special type of flaw

• Leakage: The fluid that has flowed through a leak

• Leak rate:

The amount of fluid passing through the leak per unit of time under a given set of conditions;

properly expressed in units of quantity or mass per unit of time

• Minimum detectable leak: The smallest hole or discrete passage that can be detected

• Minimum detectable leak rate: The smallest detectable fluid-flow rate

The amount of leakage required for a leak testing instrument to produce a minimum detectable signal can be determined.

This amount is generally used to indicate the sensitivity of the instrument. Instrument sensitivity is independent of test

conditions, but when an instrument is used in a test, the sensitivity of the test depends on existing conditions of pressure,

temperature, and fluid flow.

Measurement of Leakage. A leak is measured by how much leakage it will pass under a given set of conditions.

Because leakage will vary with conditions, it is necessary to state both the leak rate and the prevailing conditions to define

a leak properly. At a given temperature, the product of the pressure and the volume of a given quantity of gas is

proportional to its mass. Therefore, leak rate is often expressed as the product of some measure of pressure and volume

per unit of time--for example, torr liters per second (torr · L/s), micron liters per second ( L/s), and atmosphere cubic

centimeters per second (atm cm

/s).

The two most commonly used units of leakage rate with pressure systems are standard cubic centimeters per second (std

/s) and its equivalent, standard atmosphere cubic centimeters per second (atm cm

/s). The most frequently used unit in

vacuum leak testing is torr liters per second. The recent adoption of the Système International d'Unités (SI) system of

units has resulted in a new measure of leakage, pascal cubic meters per second (Pa · m

/s). Another new unit is moles per

second (mol/s), which has the advantage that information concerning temperature implied in other units is automatically

included in this unit. The term used for leak rate in this article is standard atmosphere cubic centimeter per second, where

standard conditions are defined as 1 atm = 101.325 kPa and standard temperature is 273.15 K (0 °C). The conversion

factor is std cm

/s = 4.46 × 10

-5

mol/s. Another unit of leak rate generally accepted is the gas-flow rate that causes a

pressure rise per unit of time of 1 m Hg/s (40 in. Hg/s) in a volume of 1 L (0.26 gal.). This is termed the lusec (liter

microns per second). The clusec (centiliter microns per second) unit is equal to 0.01 lusec. Boyle's law (PV = K, where P

is pressure, V is volume, and K is a constant) enables one to convert these units into more meaningful volumetric terms;

these conversions are listed in Table 1, along with conversions for pressures and volumes.

Table 1 Conversion factors for quantities related to leak testing

std cm

Pa · m

/s,

at 0 °C (32 °F)

torr · L/s,

at 0 °C (32 °F)

1 0.1

7.6 × 10

-1

1 × 10

-1

1 × 10

-2

7.6 × 10

-2

5 × 10

-2

5 × 10

-3

3.8 × 10

-2

1 × 10

-2

1 × 10

-3

7.6 × 10

-3

5 × 10

-3

5 × 10

-4

3.8 × 10

-3

1 × 10

-3

1 × 10

-4

7.6 × 10

-4

5 × 10

-4

5 × 10

-5

3.8 × 10

-4

1 × 10

-4

1 × 10

-5

7.6 × 10

-5

5 × 10

-5

5 × 10

-6

3.8 × 10

-5

1 × 10

-5

1 × 10

-6

7.6 × 10

-6

5 × 10

-6

5 × 10

-7

3.8 × 10

-6

1 × 10

-6

1 × 10

-7

7.6 × 10

-7

5 × 10

-7

5 × 10

-8

3.8 × 10

-7

1 × 10

-7

1 × 10

-8

7.6 × 10

-8

1 × 10

-8

1 × 10

-9

7.6 × 10

-9

1 × 10

-9

1 × 10

-10

7.6 × 10

-10

1 × 10

-10

1 × 10

-11

7.6 × 10

-11

1 × 10

-11

1 × 10

-12

7.6 × 10

-12

1 × 10

-12

1 × 10

-13

7.6 × 10

-13

1 × 10

-13

1 × 10

-14

7.6 × 10

-14

Types of Leaks. There are two basic types of leaks: real leaks and virtual leaks.

A real leak is an essentially localized leak, that is, a discrete passage through which fluid may flow (crudely, a hole).

Such a leak may take the form of a tube, a crack, an orifice, or the like. As with flaws, all leaks are not the same. Leaks

tend to grow over time, and they tend to operate differently under different conditions of pressure and temperature. A

system may also leak because of permeation of a somewhat extended barrier; this type of real leak is called a distributed

leak. A gas may flow through a solid having no holes large enough to permit more than a small fraction of the gas to flow

through any one hole. This process involves diffusion through the solid and may also involve various surface phenomena,

such as absorption, dissociation, migration, and desorption of gas molecules.

Virtual leaks involve the gradual desorption of gases from surfaces or escape of gases from nearly sealed components

within a vacuum system. It is not uncommon for a vacuum system to have both real leaks and virtual leaks at the same

time.

Leak Testing

Revised by Gerald L. Anderson, American Gas and Chemical Company

Types of Flow in Leaks

Types of flow in leaks include permeation, molecular flow, transitional flow, viscous flow, laminar flow, turbulent flow,

and choked flow. The type of flow corresponding to a specific instance of leakage is a function of pressure differential,

type of gas, and size and shape of the leak. Some of these types of flow (which are described in the following sections)

also occur at locations such as piping connections and have characteristics similar to leaks.

Permeation is the passage of a fluid into, through, and out of a solid barrier having no holes large enough to permit

more than a small fraction of the total leakage to pass through any one hole. This process involves diffusion through a

solid and may involve other phenomena, such as absorption, dissociation, migration, and desorption.

Molecular Flow occurs when the mean free path of the gas is greater than the longest cross-sectional dimension of the

leak. The mean free path is the average distance that a molecule will travel before colliding with another molecule. It is an

inverse linear function of the pressure. In molecular flow, the leakage (flow) is proportional to the difference in the

pressures. Molecular flow is a frequent occurrence in vacuum testing.

Transitional flow occurs when the mean free path of the gas is approximately equal to the cross-sectional dimension of

the leak. The conditions for transitional flow are between those for laminar and molecular flow (see the section "The

Equation of the State of an Ideal Gas" in this article).

Viscous flow occurs when the mean free path of the gas is smaller than the cross-sectional dimension of the leak. In

viscous flow, the leakage (flow) is proportional to the difference of the squares of the pressures. Viscous flow occurs in

high-pressure systems such as those encountered in detector probing applications. Above a critical value of the Reynolds

number, N

(about 2100 for circular-pipe flow), flow becomes unstable, resulting in countless eddies or vortices.

Laminar and Turbulent Flow. Laminar flow occurs when the velocity distribution of the fluid in the cross section of

the passage or orifice is parabolic. Laminar flow is one of two classes of viscous flow; the other is turbulent flow.

Particles in turbulent flow follow very erratic paths, but particles in laminar flow follow straight lines. The term viscous

flow is sometimes incorrectly used to describe laminar flow.

Choked flow, or sonic flow, occurs under certain conditions of configuration and pressure. Assume there exists a

passage in the form of an orifice or a venturi, and assume that the pressure upstream is kept constant. If the pressure

downstream is gradually lowered, the velocity of the fluid through the throat or orifice will increase until it reaches the

speed of sound. The downstream pressure at the time the orifice velocity reaches the speed of sound is called the critical

pressure. If the downstream pressure is lowered below this critical pressure, no further increase in orifice velocity can

occur, the consequence being that the mass-flow rate has reached its maximum.

Leak Testing

Revised by Gerald L. Anderson, American Gas and Chemical Company

Principles of Fluid Dynamics

Leakage generally falls within the discipline of fluid dynamics. However, only the elementary principles of fluid

dynamics are needed for most of the practical requirements of leak testing. Mathematical models are very helpful in

testing gas systems, and the following discussion is mainly concerned with such systems.

The Equation of the State of an Ideal Gas (that is, a gas made up of perfectly elastic point particles) is:

PV = NRT

(Eq 1)

where P is the gas absolute pressure, V is the volume of the vessel containing the gas, T is the absolute temperature of the

gas, N is the number of moles of gas, and R is the universal gas constant. The constant, R, is the same for all (ideal) gases;

the numerical value of R depends on the system of units in which P, V, and T are measured.

The number of moles of gas, N, is equal to the mass of the gas in grams divided by the gram-molecular weight of the gas.

The gram-molecular weight is simply the number of grams numerically equal to the molecular weight of the gas.

Regardless of the molecular weight, 1 mol of a gas always contains 6.022 × 10

(Avogadro's number) molecules.

In leak testing, pressures are measured in terms of atmospheric pressure. By international agreement, the pressure of the

standard atmosphere is 1,013,250 dyne/cm

. This is equivalent to the pressure exerted by a 760 mm (30 in.) column of

mercury, at 0 °C (32 °F), under a standard acceleration of gravity of 980.665 cm/s

(32.1740 ft/s

). Another unit of

pressure commonly used, especially in vacuum technology, is the torr, which is 1/760 of a standard atmosphere,

equivalent to 1 mm (0.04 in.) of mercury. (Conversion factors are given in Table 1.)

The numerical value of the gas constant, R, in two useful sets of units is:

(Eq 2)

Equations of state, in conjunction with measurements of pressure, volume, and temperature, are used to determine the

total quantity of gas in a closed system. The quantity of gas of a given composition can be expressed in terms of the total

number of molecules, the total mass of gas, or any quantity proportional to these. For example, suppose a pressure vessel

of volume V

is pressurized to P

(greater than atmospheric pressure) at temperature T

and allowed to stand for a period

of time. At the end of this period, the pressure is found to have decreased to P

. If the temperature of the system is

unchanged and the gas in the vessel has not changed in volume, the decrease in pressure must be the result of a loss of

gas, and a leak must be presumed to be present. However, if temperature T

is less than T

, the decrease in pressure may

be caused by a decrease in the gas volume as the result of cooling. If the quantity of gas has not changed, it follows from

Eq 1 that P

, V

, T

, V

, and T

must satisfy the relation:

(Eq 3)

If V

= V

, as assumed, then the following relation exists:

(Eq 4)

Thus, if P

has decreased in proportion to the decrease in temperature, the system can be presumed not to have leaked at a

rate discernible by the pressure gage and thermometer used to monitor the system. If the precision and drift of these

instruments are known, a maximum leak rate consistent with their indication of no leakage can be computed.

If the reduction in pressure cannot be accounted for by the reduction in temperature (that is, if Eq 4 is not satisfied), a leak

or leaks may be presumed present, and the leak rate can be determined by computing the loss in quantity of gas. Referring

to Eq 1, the loss of gas (expressed in number of moles) is given by:

(Eq 5)

where V is presumed constant.

If desired, ΔN can be converted to mass by multiplying by the mean gram-molecular weight of the gas. Dividing ΔN by

the time span between measurements gives the average leak rate.

For most systems, the volume of the system is assumed to be constant. However, all pressure vessels expanded by some

amount when they are pressurized; moreover, their volume may vary with thermal expansion or contraction. When very

precise estimates of leak rate are required, explicit corrections for these effects must be made.

A given leak may exhibit several types of flow, depending on the pressure gradient, temperature, and fluid composition.

Therefore, it is important to identify the type of flow that exists in order to predict the effect of changing these variables.

Conversely, with constant temperature, pressure, and fluid composition, the configuration of the leak will determine the

type of flow. Because the leak may be a crack, a hole, permeation, or a combination of these, its configuration is often

impossible to ascertain, but general empirical guidelines can establish the type of flow for gases. If the leak rate is:

• Less than 10

-6

atm cm

/s, the flow is usually molecular

• 10

-4

to 10

-6

atm cm

/s, the flow is usually transitional

• 10

-2

to 10

-6

atm cm

/s, the flow is usually laminar

• Greater than 10

-2

atm cm

/s, the flow is usually turbulent

These types of gas flow are more accurately described by the Knudsen number:

(Eq 6)

where N

is the Knudsen number, λ is the mean free path of the gas, and d is the diameter of the leak. The relationship

between, N

and flow regime is:

< 0.01 (laminar, or a higher regime)

0.01 N

1.00 (transitional)

> 1.00 (molecular)

Mean Free Path. Values of the mean free path for various gases and pressures are listed in Table 2. In a vacuum

system, the mean free path will vary from inches to many feet; when the mean free path is very long, collisions with the

chamber surfaces occur more frequently than collisions between molecules. This, in part, is the reason that neither a gas

nor leakage diffuses evenly throughout a vacuum system at a rapid rate. The flow of gases in a vacuum is analogous to the

flow of current in an electrical system. Every baffle or restriction in the system impedes gas diffusion. Consequently, if

there are many impedances between the leak and the leak detector that reflect or absorb the molecules, movement of the

gas molecules to the leak detector does not follow theoretical diffusion rates; instead, it is greatly dependent on the

configuration of the vacuum system.

Table 2 Mean free path lengths of various atmospheric gases at 20 °C (68 °F) and various vacuum

pressures

Mean free path length at

indicated absolute pressure

1 μPa

(7.5 × 10

-9

torr)

1 mPa

(7.5 × 10

-6

torr)

1 Pa

(7.5 × 10

-3

torr)

1 kPa

(7.5 torr)

100 kPa

(750 torr)

(a)

Gas

km ft × 10

m ft mm in. μ μin.

(b)

Air 6.8 2.2 6.8 22 6.8 0.27 6.8 272

68 680

Argon 7.2 2.4 7.2 24 7.2 0.28 7.2 290

72 720

Carbon dioxide

4.5 1.5 4.5 15 4.5 0.18 4.5 180

45 450

Hydrogen 12.5 4.10 12.5 41.0 12.5 0.492 12.5

500

125 1250

Water 4.2 1.4 4.2 14 4.2 0.16 4.2 170

42 420

Helium 19.6 6.43 19.6 64.3 19.6 0.772 19.6

784

196 1960

Nitrogen 6.7 2.2 6.7 22 6.7 0.26 6.7 268

67 670

Neon 14.0 4.59 14.0 45.9 14.0 0.55 14.0

560

140 1400

Oxygen 7.2 2.4 7.2 24 7.2 0.28 7.2 290

72 720

Source: Ref 1