ASM Metals HandBook Vol. 8 - Mechanical Testing and Evaluation

Подождите немного. Документ загружается.

Fig. 17 High-speed (3.5 in. center distance) PC test rig

Test Procedures for Surface Durability (Pitting) and Root Strength (Bending) Testing. One or both gears in the

specimen set are coated with a thin layer of layout blue prior to installation in the test machine. After the gears

are installed, they are run briefly at low load with room-temperature lubricant to determine the contact pattern.

The low-speed test rigs allow some alignment adjustment to optimize contact pattern. The specimen mountings

in the high-speed test rigs are rigid, so the only remedy for nonuniform contact is to change specimen gear sets.

A break-in run is generally conducted before the start of the test run. On the low-speed machines this run is

accomplished by applying one-half of the test load and running for h at test speed starting with room-

temperature lubricant. The friction of the meshing gears warms the lubricant to near test temperature in this

time. On the high-speed machines, load and speed are increased in a set pattern to accomplish break-in. As

described previously, break-in is expected to increase the life of the gears. Thus, all specimens tested in a given

project should be broken in using the same method. For root strength (bending fatigue) tests, break-in may not

affect bending fatigue life but may delay the onset of pitting long enough to allow the gear to fail in bending.

In general, the main difference in surface durability (pitting) and root strength (bending) tests is the load. Most

of the standard specimen gear sets are designed to allow surface durability testing at loads high enough to

produce failure in less than 10 million cycles. Root strength tests, therefore, have to be conducted at loads high

enough to produce failure in less than 1 million cycles to avoid surface durability failures. Thus, unless

specially designed specimens are tested, the results of this test can only be compared to the lower-life portion of

the stress-cycles relationship developed in the STF test.

Surface durability gear tests have similar problems with scatter as those previously described for RCF tests.

However, each specimen gear has 18 or more teeth; therefore, the result of a single PC surface durability test

represents the lowest performance of 18 or more items tested. This statistical difference causes the results of PC

surface durability tests to exhibit less scatter than those of RCF tests, and Weibull slopes are generally three or

higher. Six tests are still generally conducted with each variant at each load, the reduced scatter being used to

allow more statistically reliable comparison of the results rather than to reduce the number of tests. A similar

statistical difference exists between the results of STF tests and PC bending tests, and this difference must be

considered when comparing STF and PC bending results.

Specimen Results. The end point of a surface-durability test may be somewhat ambiguous. The section

regarding pitting in the discussion of failure modes indicates several points in the development of surface origin

pitting that may be considered failure. Figures 3 and 4 show gear teeth with different stages of pitting, both of

which were considered failures. Test gear teeth are also subject to wear and localized scoring, both of which

can affect the appearance of the contact surface in ways similar to some of the stages of pitting development. A

composite failure criterion has to be adopted that specifies the level of damage in each of these categories that

constitutes failure. A typical failure criterion for pitting is one pit 4 mm ( in.) wide or several smaller pits

with slightly greater total area. Typical failure criteria for micropitting is micropitting over half of the contact

width of all the teeth and for wear, a loss of 12 μm (0.0005 in.) at any point on the profile.

A test program evaluating bending and surface durability performance of candidate materials would include 24

tests with each material. Six tests would be conducted at each of two loads intended to result in bending failure;

the balance of the tests would be conducted at two loads intended to result in pitting failure. The results of tests

at each of the four loads are analyzed with Weibull statistical analysis (as illustrated in Fig. 7). A load-cycles

diagram is then constructed showing loads corresponding to 10, 50, and 90% failures at various lives for both

bending and surface durability. Comparisons are made on the basis of life to 50% (or 10%) failures at a given

load, or the load corresponding to 50% failures at a given life. It is useful to construct a stress-cycles diagram

(determining contact stress as described here) to compare surface durability results obtained from specimens

having differing crown.

Test Procedures for Scoring (or Scuffing) Testing. Scoring is typically most severe in the dedendum of the

driving gear (at the start of contact) and at the tip of the driven gear. Therefore, to facilitate observation of

scoring, the gears and loading are set up so that the specimen is the driven gear. Scoring is more probable at

high speed, high load, and/or high-lubricant temperature. Because of this probability, the test machine is set up

to run at high speed and high-lubricant temperature. Resistance to scoring is then measured by the load that will

precipitate scoring. Load is increased in comparatively small steps until scoring starts. Because small variations

in lead or crown can produce as much variation in unit loading at the most heavily loaded area as one or more

load steps, specimens used for scoring should be those with the most consistent geometry that can be selected

from the lot to minimize scatter. Gear alignment should be adjusted, if possible, to obtain the best contact

pattern.

Break-in is a particularly critical part of this test. Therefore, all specimens should be broken in using an

identical process, and control specimens should be included in the project test matrix to benchmark the results

for comparison with prior tests. Typically, break-in is started at low load in room-temperature oil and run-in

one to four load steps lasting 20,000 to 50,000 loading cycles each. The lubricant temperature is allowed to

slowly rise due to the friction of the meshing gears during break-in. The condition of the lubricant also can

affect the outcome of the test; thus, lubricant should be changed prior to the start of each test.

Specimen Results. After break-in, the surface roughness (R

) is measured in the axial direction at the tips of

several teeth on the specimen gear, and the surface condition is documented. The load is increased, and the

lubricant temperature is raised to test temperature. The gear set is run 20,000 to 50,000 loading cycles for each

load step. After each load step, surface roughness and surface condition are examined and recorded. The

appearance of score marks, or an abrupt increase in surface roughness, indicates scoring failure. The gear

shown in Fig. 2 illustrates a typical failure in this test. Typically, three tests are conducted per variant.

Comparisons between variants are made on the basis of the load that precipitates scoring. Contact stress

considered alone has little influence on scoring; consequently, comparisons made on the basis of contact stress

may be misleading.

Mechanical Testing of Gears

Douglas R. McPherson and Suren B. Rao, Applied Research Laboratory, The Pennsylvania State University

References

1. E.E. Shipley, “Failure Modes in Gears,” presented at Forum on Gear Manufacture and Performance, 19

Oct 1973 (Detroit), American Society of Mechanical Engineers

2. D.W. Dudley, Gear Wear, Wear Control Handbook, M. Peterson and W. Winer, Ed., American Society

of Mechanical Engineers, 1980

3. G. Niemann and H. Winter, Getriebeschäden und Abhilfe, Entwicklungstendenzen, Maschinen-

elemente, Vol 2, 2nd ed., Springer-Verlag (Berlin), 1983

4. “Appearance of Gear Teeth—Terminology of Wear and Failure,” ANSI/AGMA 1010-E95, American

National Standards Institute/American Gear Manufacturers Association

5. R. Errichello, Friction, Lubrication, and Wear of Gears, Friction, Lubrication, and Wear Technology,

Vol 18, ASM Handbook, ASM International, 1992, p 535–545

6. L. Alban, Systematic Analysis of Gear Failures, American Society for Metals, 1985

7. A.P. Boresi and O.M. Sidebottom, Advanced Mechanics of Materials, 4th ed., John Wiley & Sons, 1985

8. “Information Sheet: Geometry Factors for Determining the Pitting Resistance and Bending Strength of

Spur, Helical, and Herringbone Gear Teeth,” AGMA 908-B89 (R1995), American Gear Manufacturers

Association, 1995

Testing of Pressure Vessels, Piping, and Tubing

E. Roos, K.-H. Herter, and F. Otremba, Staatliche Materialpruefungsanstalt (MPA), University of Stuttgart

Introduction

THE MECHANICAL EVALUATION OF PIPE FAILURES has evolved over time. An initial purpose of such

analysis was to determine the causes of large breaks occurring in oil and gas pipelines. The development of

commercial nuclear power plants initiated the need for additional tools to assess the reliability and the failure

behavior of pressurized vessels and pipes under different loading and environmental conditions. The results of

these efforts have been transferred to other relevant industrial branches as well.

The main effort in evaluating the mechanical and structural behavior of pressurized components started about

1950. Since that time, numerous investigations have been performed to assess the loading capacity and the

failure behavior of pressure vessels, piping, and tubing. Investigations have also focused on determining failure

loads and quantifying the margins of safety.

Testing of Pressure Vessels, Piping, and Tubing

E. Roos, K.-H. Herter, and F. Otremba, Staatliche Materialpruefungsanstalt (MPA), University of Stuttgart

Safety Aspects and Integrity Concepts

The license and operation of pressure vessels and piping in the different countries throughout the world are

subject to national laws, special regulations, and safety criteria. Pressure vessels and piping are designed

according to the relevant national technical codes and standards that contain detailed rules regarding materials,

construction, design and calculation, manufacture, and quality assurance. These codes require that state-of-the-

art technology is applied to the safe operation and maintenance of this equipment.

The relevant national technical codes and standards for pressure vessels and piping include the following:

• In the United States, the American Society of Mechanical Engineers (ASME) Boiler and Pressure

Vessel Code—for example, Section III for nuclear power plant components (Ref 1) and Section VIII for

the construction of pressure vessels (Ref 2)—and the ASME Code for Pressure Piping (B31) (Ref 3)

• In Germany, the Technical Rules for Steam Boilers (TRD) (Ref 4), the Technical Rules for Pressure

Vessels (AD-Rules) (Ref 5), and the Safety Standards of the Nuclear Safety Standards Commission

(e.g., KTA 3201 and KTA 3211) (Ref 6)

• In France, the code for the Construction of Pressure Vessels (CODAP) (Ref 7), the Industrial Piping

Code (CODETI) (Ref 8), and the Design and Construction Rules for Mechanical Components of

Nuclear Power Plants (RCC-M and RCC-MR) (Ref 9, 10)

• In the United Kingdom, the standards of the British Standards Institution related to Boilers and Pressure

Vessels (BS 5500) (Ref 11)

For a general or a component-specific design of pressure vessels and piping according to the technical codes

and standards, it has to be demonstrated that all stresses resulting from mechanical and/or thermal as well as

environmental loadings are within the allowable stress limits, and the usage factor developed by a fatigue

analysis is well below the limiting value.

This article describes general concepts related to the evaluation of pressurized components without or with

defects. The reliability of these concepts is demonstrated by tests of full-scale components with and without

defects under internal pressure loading and, in some cases, with an external bending load superimposed at the

relevant temperatures.

Integrity Concept

The integrity concept for pressure vessels, piping, and tubing is demonstrated in the following example for

pressurized components of nuclear power plants. Requirements for nuclear power plants take into account not

only the stresses resulting from operational loads, but also extreme loading conditions that could be caused by

accidents (for example, water hammer, aircraft crash, earthquake, and postulated pipe rupture). For such loads it

must be demonstrated that, on the one hand, catastrophic failures of the components can be excluded, or, on the

other hand, sequential damages will not lead to critical sequences for the plant.

The engineering of high-performance steam-generating plants and chemical plants (including design and

construction, selection of material, manufacture, and quality assurance of ductile steel components and systems)

has evolved over decades. The ongoing advances in science and technology have increased the standards for the

construction and operation of the plants, especially concerning the safety against catastrophic failure of

pressurized components and systems (Ref 12).

Safety principles and approaches, especially for nuclear power plants, vary somewhat from country to country.

However, a generally accepted standard requires that the possibility of catastrophic failure from double-ended

guillotine breaks (DEGB) must be completely excluded for pressure vessels and piping. A leak-before-break

(LBB) assessment is used to show the likelihood of detectable crack sizes, which allows countermeasures to be

taken before a large break occurs (Ref 13, 14, 15, 16, and 17).

Analyses of the operational behavior of high-quality pressurized components and systems as well as flaw and

damage analyses have shown that proven and well-defined principles exist to reliably attain and maintain high

quality. In Germany, these principles have been codified as the basis safety concept (BSC) and break preclusion

concept (Ref 12). These concepts require that the quality of components must be guaranteed during the

production process. Furthermore, the BSC requires the existence of redundancies to ensure that any possible

deviation from optimization is sufficiently unimportant and unlikely to cause catastrophic failure. One

redundancy of considerable importance relies on verification and validation using calculation codes and

fracture mechanics as well as nondestructive testing technologies. Testing and evaluation are indispensable to

provide the knowledge basis for the validation of codes and standard. The validation work is done mainly based

on the results of the uniquely large full-scale quasi-static and dynamic testing of components with and without

degraded sections to demonstrate their load-bearing capacity and their failure behavior, for example, Ref 18,

19, 20, 21, 22, 23, 24, 25, and 26.

References cited in this section

1. “Rules for Construction of Nuclear Power Plant Components,” ASME Boiler and Pressure Vessel Code,

Section III, American Society of Mechanical Engineers, 1998

2. “Rules for Construction of Pressure Vessels,” ASME Boiler and Pressure Vessel Code, Section VIII,

Divisions 1 and 2, American Society of Mechanical Engineers, 1998

3. ASME Code for Pressure Piping, B31, An American National Standard, The American Society of

Mechanical Engineers, 1989

4. German Technical Rules for Steam Boilers (Technische Regeln für Dampf-kessel, TRD), Carl

Heymanns Verlag KG, Cologne, 1998

5. German Technical Rules for Pressure Vessels (Arbeitsgemeinschaft Druckbehälter, AD-Merkblätter),

Carl Heymanns Verlag KG, Cologne, 1998

6. German Safety Standards of the Nuclear Safety Standards Commission (Sicherheitstechnische Regeln

des Kerntechnischen Ausschusses, KTA), KTA Rule 3201 and KTA Rule 3211, Carl Heymanns Verlag

KG, Cologne, latest edition

7. French Code for the Construction of Pressure Vessels (CODAP), Association Francaise des Ingénieurs

en Appareils à Pression (AFIAP), Paris

8. French Code for the Construction of Industrial Piping Components (CODETI)

9. French Design and Construction Rules for Mechanical Components of PWR Nuclear Islands (RCC-M),

Association Francaise pour la Construction des Ensembles Nucleaires (AFCEN)

10. French Design and Construction Rules for Mechanical Components of Fast Breeder Reactors (RCC-M),

Association Francaise pour la Construction des Ensembles Nucleaires (AFCEN)

11. “Unfired Fusion Weld Pressure Vessels,” BS 5500, British Standard Institution

12. K. Kussmaul, German Basis Safety Concept Rules Out Possibility of Catastrophic Failure, Nucl. Eng.

Int., Vol 12, 1984, p 41–46

13. R. Wellein and G. Preuβer, “State of Engineering of the Leak-Before-Break Procedure in Germany,”

Seventh German-Japanese Joint Seminar, 29–30 Sept 1997, MPA Stuttgart

14. K. Kussmaul, D. Blind, E. Roos, and D. Sturm, The Leak-Before-Break Behavior of Pipes, VGB

Kraftwerkstech., Vol 70 (No. 7), July 1990, p 465–477

15. G.M. Wilkowski, R.J. Olson, and P.M. Scott, “State -of-the-Art Report on Piping Fracture Mechanics,”

Report NUREG/CR-6540, BMI-2196, Jan 1998

16. Y. Asada, K. Takumi, H. Hata, and Y. Yamamoto, Development of Criteria for Protection against Pipe

Breaks in LWR Plants, Int. J. Pressure Vessels Piping,, Vol 43, 1990, p 95–111

17. C. Faidy, P. Jamet, and S. Bhandari, “Developments in Leak-Before-Break Approaches in France,”

NUREG/CP-0092, March 1988, p 69–82 (available from U.S. NRC)

18. D. Sturm, W. Stoppler, and J. Schiedermaier, The Behavior of Dynamically Loaded Pipes with

Circumferential Flaws under Internal Pressure and External Bending Loads, Nucl. Eng. Des., Vol 96,

1986, p 99–113

19. E. Roos, K.-H. Herter, P. Julisch, G. Bartholome, and G. Senski, Assessment of Large Scale Pipe Tests

by Fracture Mechanics Approximation Procedures with Regard to Leak-Before-Break, Nucl. Eng. Des.,

Vol 112, 1989, p 183–195

20. Y. Asada, K. Takumi, N. Gotoh, T. Umemoto, and K. Kashima, Leak-Before-Break Verification Test

and Evaluations of Crack Growth and Fracture Criterion for Carbon Steel Piping, Int. J. Pressure

Vessels Piping, Vol 43, 1990, p 379–397

21. K. Kashima, N. Miura, S. Kanno, K. Miyazaki, M. Ishiwata, and N. Gotoh: A Research Program for

Dynamic Fracture Evaluation of Japanese Carbon Steel Pipes, Nucl. Eng. Des., Vol 174, 1997, p 33–39

22. D. Moulin and P. Le Delliou, French Experimental Studies of Circumferential Through Wall Cracked

Austenitic Pipes under Static Loading, Int. J. Pressure Vessels Piping, Vol 65, 1996, p 343–352

23. P. Scott et al., “IPIRG-2 Task 1—Pipe System Experiments with Circumferential Cracks in Straight

Pipe Location,” NUREG/CR-6389, Feb 1997 (available from U.S. NRC)

24. K. Shibata, T. Isozaki, S. Ueda, R. Kurihara, K. Onziawa, and A. Kohsaka, “Results of Reliability Test

Program on Light Water Reactor Piping,” Nucl. Eng. Des., Vol 153, 1994, p 71–86

25. G.M. Wilkowski et al., “Degraded Piping Program—Phase II, Summary of Technical Results and Their

Significance to Leak-Before-Break and In-Service Flaw Acceptance Criteria,” March 1984-Jan 1989,

NUREG/CR-4082, Vol 8, March 1989 (available from U.S. NRC)

26. G. Yagawa, Y. Takahashi, N. Kato, M. Saito, K. Hasegawa, and T. Umemoto, Fracture Behavior of

Cracked Type 304 Stainless Steel Pipes under Tensile and Thermal Loadings, Int. J. Pressure Vessels

Piping, Vol 19, 1985, p 247–281

Testing of Pressure Vessels, Piping, and Tubing

E. Roos, K.-H. Herter, and F. Otremba, Staatliche Materialpruefungsanstalt (MPA), University of Stuttgart

Fracture Mechanics Analysis

Fracture mechanics approaches are used to validate components with longitudinal cracks and with

circumferential cracks and to analyze crack growth behavior under cyclic loading.

Components with Longitudinal Cracks

The well-known methods for the validation of pressurized components with longitudinal part-through or

through-wall cracks (slots) were developed on the basis of a huge number of experimental results, for example,

Ref 14, 27, and 28, 29. These semiempirical engineering methods (Fig. 1) require major plastic deformation to

occur in the region around the crack tip. In one of the methods, the behavior of a material is expressed solely by

the flow stress, σ

= (R

p0.2

+ R

)/2 (yield strength and tensile strength dependent method). In another method,

the toughness values of the material, expressed by the Charpy energy impact toughness (C

) in the upper shelf,

are also considered (toughness-dependent approach). In Fig. 2, the influence of different toughness values of

the material used is shown with respect to the critical length of through-wall cracks (for the same pipe

dimension and corresponding material strength data). In the tensile load (σ

) range of interest, σ

/σ

≤ 0.4, the

critical crack length for through-wall cracks increases with increasing Charpy energy.

Fig. 1 Methods for calculating maximum loads for piping with longitudinal cracks. Note: exp [A] = e

Fig. 2 Toughness-dependent critical crack length for longitudinal crack in a pipe with an outer diameter

of 800 mm (31.5 in.), wall thickness of 47.2 mm (1.86 in.), internal pressure of 20 MPa (2.9 ksi), and a

ratio of crack depth to wall thickness of 1.0

Components with Circumferential Cracks

For pressurized cylindrical components with circumferential part-through or through-wall cracks, there are

several methods available to determine the failure load for the degraded component. In piping systems, the

relevant loads are external bending moments in addition to the internal pressure. Consequently, the methods for

the validation of pressurized components with circumferential defects take into account not only the internal

pressure, but also, in particular, the external bending moment.

Plastic Limit Load Concept. For pipes made of ductile materials, local yielding in the region of the crack tips

can propagate until, under limit load conditions, the entire pipe cross section is subjected to plastic deformation.

If linear-elastic/ideal-plastic material behavior is assumed, the stress distribution over the cracked cross section

is as shown in Fig. 3 (Ref 30).

Circumferential part-through

crack

Circumferential through-wall

crack

Criterion Flow stress Flow stress

2 σ

t (2 sin β - f sin α) 2 σ

t (2 sin β - sin α)

Bending moment, M

Factors

Scope of application Thin-walled vessels Thin-walled vessels

Fig. 3 Procedure and designations for failure moment of pipes with circumferential cracks according to

the plastic limit load concept, p

, internal pressure; NA, neutral axis

The shift of the neutral axis within the cracked-pipe cross section can be determined according to the balance of

the forces in axial direction, with the flow stress σ

acting uniformly on the tension side of the cross section up

to the neutral axis and the compressive flow stress -σ

acting on the compression side. To allow for work

hardening of the material, the assumption made in Ref 30 is, for example, σ

= 0.5 (R

p0.2

+ R

Compared with experimental results, the calculated failure loads for degraded pipes under internal pressure and

superimposed external bending moment are conservative when using the flow stress as σ

= (R

p0.2

+ R

)/2.4. A

more conservative assessment of the failure loads or the corresponding critical crack sizes is achieved by using

flow stress of σ

= R

p0.2

Two-Criteria Approach (R6-Method). This section describes a calculation method for which the strength values

determined by tensile testing are required as material characteristics as well as fracture toughness values. This

is demonstrated using the R6-method and the R-curve method. It is indeed possible in fracture mechanics

procedures to include not only the strength but also the fracture mechanics characteristics of the material in the

failure analysis.

The R6-method is based on the idea that, depending on the material characteristics, failure of a component can

occur from brittle fracture up to full plastic deformation with the transition between these two extreme

conditions assumed to be continuous.

The limit curve is defined as:

= (1 - 0.14 ) [0.3 + 0.7 exp (−0.65 )]

and

(max) = σ

p0.2

where:

= (R

p0.2

+ R

)/2

The loading parameters of the components are given by:

= F/F

as the ratio of the actual load F to the plastic limit load F

and

= K

as the ratio of the actual stress-intensity factor K

as a function of crack size and load, and the fracture

mechanics material characteristic K

for linear-elastic and J

for elastic-plastic conditions (Ref 31).

In the above equation, the J

values are derived from the linear-elastic K

values. According to this concept,

onset of stable crack extension occurs when the load point developed by L

and K

is on the limit curve.

Instability is derived from the load path, which forms a tangent to the limit curve, taking into consideration the

stable crack extension. A schematic of the application of this procedure is shown in Fig. 4. For this purpose, the

constant load F acting in L

is normalized to the crack-length-dependent limit load F

(a + Δa) and in K

the

applied K

value, calculated with regard to the crack growth (K

= a + Δa) is converted to J and normalized with

the crack resistance curve of the material, which is usually represented by the J

-curve (J

+ (Δa)), (a = crack

depth).

Fig. 4 Application of the two-criteria approach (R6-method, Rev 3)

R-Curve Method. The R-curve method combines the crack driving force calculated for the component with the

material characteristics (evaluated from the crack resistance curve), especially the fracture mechanics

characteristics. The crack driving force is calculated in terms of the J-integral by superposition of the elastic

and the plastic part, which is, according to Ref 32, 33, and 34:

J = J

+ J

with