You are on page 1of 56

UNIVERSITY OF TECHNOLOGY MATERIALS ENGINEERING DEPARTMENT UNDERGRADUATE STUDIES

By

DR. ALAA ABIDULHASAN ATIYAH ASSISTANT PROFESSOR MECHANICAL ENGINEERING (METALLURGICAL ENGINEERING)

BAGHDAD 2010-2011
1

PREFACE Machinery and their components are required to perform continuously without interruptions to get desired output and achieve higher productivity, whatever is the end product of a manufacturing industry. However, it is also known that as soon as a component or a system is put on use its degradation start immediately owing to wear and tear. Unplanned and catastrophic breakdown/failure leads to huge losses to the industry in terms of production and use of resources. Therefore, efforts are always made to avoid failures of key components (gears, springs, lifting equipments, shafts, pressure vessels etc.) and system whose failure can lead to loss of life and property apart from production. It is not uncommon to have such kind of failures of engineering components and systems which lead to stoppage of entire production especially industries like cement, steel, food processing etc. Presently, the most of maintenance related jobs in industries are taken care of mechanical engineers who are generally not very well equipped systematic procedure to be used for failure analysis and identify the root cause of failure and so as to take corrective action to avoid further failures in future. It is, thus, important that materials engineers/personnel are trained and equipped with necessary technological skill and the systematic procedure to be used for failure analysis so as to identify the root cause of failure and take corrective actions. Based on inputs and requests received from practicing engineers of different Industries, proposed short course program entitled Failure has been planned. It is therefore high time for the students in materials engineering department to take the benefits of recent technological developments in failure analysis in the interest of achieving better skills to support the National economy of our country.

Dr. Alaa Baghdad-Iraq 2011

COURSE OBJECTIVES It is hoped to get the following objectives: 1. To provide basic knowledge about the nature of failure of mechanical components 2. To introduce methods of identification of mode and causes of failures 3. To update the information on recent advances in area of characterization techniques for failure analysis. 4. To provide an understanding on systematic procedure for failure analysis. 5. Application of fracture mechanics concepts in failure analysis 6. Multi case studies in failures analysis.

COURSE COORDINATOR (S) Dr. Alaa Abdulhasan Atiyah Assistant Professor PhD in Mechanical engineering/Metallurgical Engineering Department Of Materials Engineering University of Technology Baghdad-Iraq E-mail:dr.alaaatiyah@uotechnology.edu.iq

CHAPTER 1 GENERAL INTRODUCTION


1.1 Introduction When one considers the many millions of metallic parts that are fabricated and placed in service, it is not unusual that some will fail prematurely. Simply from a statistical viewpoint it is not reasonable, with present engineering practices, to expect no failure. However, even though the number of failures of a particular component may be small, they are important because they may affect the manufacturers reputation for reliability. In some cases, particularly when the failure results in personal injury or death, it will lead to expensive lawsuits. It is not unusual for automotive manufacturers under prodding and publicity from consumer watchdogs to recall millions of cars to correct a design or heat treating defect even though the actual number of failures was very small. The purpose of this chapter is to briefly explain the basic causes for metal failure and to illustrate some of the failures by case histories. Most of the illustrations in this chapter were taken from real failures accidents. 1.2 Procedure In many failure analyses it is important to get as much information as possible from the failed part itself along with an investigation of conditions at the time of failure. Some of the questions to be asked are: 1. How long was the part in service? 2. What was the nature of the stresses at the time of failure? 3. Was the part subjected to an overload? 4. Was the part properly installed? 5. Was it subjected to service abuse? 6. Were there any changes in the environment? 7. Was the part properly maintained?
4

A study of the fractured surface should answer the following questions: 1. Was the fracture ductile, brittle, or a combination of the two? 2. Did failure start at or below the surface? 3. Did the failure start at one point, or did it originate at several points? 4. Did the crack start recently or had it been growing for a long time? 1.3 Modes of Failures The fundamental requirement of any structure is that it should be designed to resist mechanical failure through any (or a combination of) the following modes: 1. Elastic instability (buckling) 2. Large elastic deformation (jamming) 3. Gross plastic deformation (yielding) 4. Tensile instability (necking) 5. Fracture Most of these failure modes are relatively well understood, and proper design procedures have been developed to resist them. However, fractures occurring after earthquakes constitute the major source of structural damage. In fact, fracture often has been overlooked as a potential mode of failure at the expense of an overemphasis on strength. Such a simplification is not new, and finds a very similar analogy in the critical load of a column. If column strength is based entirely on a strength criterion, an unsafe design may result as instability (or buckling) is overlooked for slender members. Thus failure curves for columns show a smooth transition in the failure mode from columns based on gross section yielding to columns based on instability.

1.4 Important definitions In order to starting over, some definitions must be remembered: 1. Stress - Force or load per unit area of cross-section over which the force or load is acting. 2. Strain - Elongation change in dimension per unit length. 3. Youngs modulus - The slope of the linear part of the stress-strain curve in the elastic region, same as modulus of elasticity. 4. Shear modulus (G) - The slope of the linear part of the shear stress-shear strain curve.
5

5. Viscosity ( ) - Measure of resistance to flow, defined as the ratio of shear stress to shear strain rate (units Poise or Pa-s). 6. Thixotropic behavior - Materials that show shear thinning and also an apparent viscosity that at a constant rate of shear decreases with time. 7. Load - The force applied to a material during testing. 8. Strain gage or Extensometer - A device used for measuring change in length and hence strain. 9. Glass temperature (Tg ) - A temperature below which an otherwise ductile material behaves as if it is brittle. 10.Engineering stress - The applied load, or force, divided by the original crosssectional area of the material. 11.Engineering strain - The amount that a material deforms per unit length in a tensile test. 1.5 Mechanical Loading Types knowing the types of mechanical loading is highly important as shown below:

Figure (1.1) (a) Tensile, compressive, shear and bending stresses. (b) Illustration showing how Youngs modulus is defined for elastic material. (c) For nonlinear materials, we use the slope of a tangent as a variable quantity that replaces the Youngs modulus constant

Figure (1.2): a unidirectional force is applied to a specimen in the tensile test by means of the moveable crosshead. The cross-head movement can be performed using screws or a hydraulic mechanism.

Figure (1.3): Tensile stress-strain curves for different materials. Note that these are qualitative.
7

Figure (1.4): The stress-strain curve for an aluminum alloy from Table 6-1

Some useful conversion data:

Example 1.1 Design of a Suspension Rod, An aluminum rod is to withstand an applied force of 45,000 pounds. To assure a sufficient safety, the maximum allowable stress on the rod is limited to 25,000 psi. The rod must be at least 150 in. long but must deform elastically no more than 0.25 in. when the force is applied. Design an appropriate rod.
8

1.6 Properties Obtained from the Tensile Test 1. Elastic limit 2. Tensile strength, Necking 3. Hookes law 4. Poissons ratio 5. Modulus of resilience (Er) 6. Tensile toughness 7. Ductility

Figure (1.5): Localized deformation of a ductile material during a tensile test produces a necked region. The micrograph shows necked region in a fractured sample

1.7 True Stress and True Strain 1. True stress: The load divided by the actual cross-sectional area of the specimen at that load. 2. True strain: The strain calculated using actual and not original dimensions, given by t ln(l/l0).

10

Figure (1.6) The relation between the true stress-true strain diagram and engineering stress-engineering strain diagram. The curves are identical to the yield point Example 1.2 True Stress and True Strain Calculation, Compare engineering stress and strain with true stress and strain for the aluminum alloy in Example 6.1 at (a) the maximum load and (b) fracture. The diameter at maximum load is 0.497 in. and at fracture is 0.398 in.

11

1.8 Strains An infinitesimal normal strain is defined by the change of length, L, of a line: d = dL/L.. (1.1) Integrating from the initial length, Lo, to the current length, L, =dL/L = ln(L/Lo).(1.2) This finite form is called true strain (or natural strain, logarithmic strain). Alternatively, engineering or nominal strain, e, is defined as e = L/Lo.. (1.3) If the strains are small, then engineering and true strains are nearly equal. Expressing = ln(L/Lo) = ln(1 + e) as a series expansion, = e e2/2 + e3/3! , so as e0, e. This is illustrated in the following example.

Example 1.3 Calculate the ratio e/ for several values of e. Solution: e/ = e/ln(1 + e). Evaluating: for e = 0.001, e/ = 1.0005;
12

for e = 0.01, e/ = 1.005; for e = 0.02, e/ = 1.010; for e = 0.05, e/ = 1.025; for e = 0.10, e/ = 1.049; for e = 0.20, e/ = 1.097; for e = 0.50, e/ = 1.233. Note that the difference between e and is less than 1% for e < 0.02. There are several reasons that true strains are more convenient than engineering strains. 1. True strains for equivalent amounts of deformation in tension and compression are equal except for sign. 1. True strains are additive. For a deformation consisting of several steps, the overall strain is the sum of the strains in each step. 2. The volume change is related to the sum of the three normal strains. For constant volume, x + y + z = 0. These statements are not true for engineering strains, as illustrated in the following examples. Example 1.4 An element 1 cm long is extended to twice its initial length (2 cm) and then compressed to its initial length (1 cm). 1. Find true strains for the extension and compression. 2. Find engineering strains for the extension and compression. Solution: 1. During the extension, = ln(L/Lo) = ln2 = 0.693, and during the compression, = ln(L/Lo) = ln(1/2) = 0.693. 2. During the extension, e = _L/Lo = 1/1 = 1.0, and during the compression, e = L/Lo = 1/2 = 0.5. Note that with engineering strains, the magnitude of strain to reverse the shape change is different.

13

Example 1.5 A bar 10 cm long is elongated by (1) drawing it to 15 cm, and then (2) drawing it to 20 cm. 1. Calculate the engineering strains for the two steps and compare the sum of these with the engineering strain calculated for the overall deformation. 1. Repeat the calculation with true strains. Solution: 1. For step 1, e1 = 5/10 = 0.5; for step 2, e2 = 5/15 = 0.333. The sum of these is 0.833, which is less than the overall strain etot = 10/10 = 1.00. 2. For step 1, 1=ln(15/10)=0.4055; for step 2, 1=ln(20/15)=0.2877. The sum is 0.6931 and the overall strain is tot =ln(15/10)+ln(20/15)=ln(20/10)=0.6931. Example 1.6 A block of initial dimensions Lxo, Lyo, Lzo is deformed so that the new dimensions are Lx, Ly, Lz. Express the volume strain, ln(V/Vo), in terms of the three true strains, x, y, z. Solution: V/Vo = LxLyLz/(LxoLyoLzo), so ln (V/Vo) = ln (Lx/Lxo) + ln (Ly/Lyo) + ln (Lz/Lzo) = x + y + z . Note that if there is no volume change (ln(V/Vo) = 0), the sum of the normal strains x + y + z = 0.

1.9 Principal stresses. It is always possible to find a set of axes (1, 2, 3) along which the shear stress components vanish. In this case the normal stresses, 1, 2, and 3, are called principal stresses and the 1, 2, and 3 axes are the principal stress axes. The magnitudes of the principal stresses, p, are the three roots of: p3 I1 p2 I2 p I3 = 0 (1.1) Where,

14

I1 = xx + yy + zz I2= yz2 + zx2 + xy2- yyzz - zzxx- xxyy I3 = xxyy zz + 2 yz zx xy- xx yz2 - yyzx2 - zzxy2 The first invariant I1 =p/3, where p is the pressure. I1, I2, and I3 are independent of the orientation of the axes and are therefore called stress invariants. In terms of the principal stresses, the invariants are I1 = 1 + 2 + 3, I2 = 2233 3311 1122 (1.2) And I3 = 112233. Example 1.2 Find the principal stresses in a body under the stress state x = 10, y = 8, z = 5, yz = zy = 5, zx = xz = 4, and xy = yx = 8, where all stresses are in MPa. Solution: Using Equation (1.13), I1 = 10 + 8 5 = 13, I2 = 52 + (4)2 + (8)2 8(5) (5)10 108 = 115, and I3 = 108(5) + 25(4)(8) 1052 8(4)2 (5)(8)2 = 138. Solving Equation (1.1) gives p3-13p2-115p + 138=0, p =1.079, 18.72, 6.82. 1.10 Elastic/Plastic Deformation When a sufficient load is applied to a metal or other structural material, it will cause the material to change shape. This change in shape is called deformation. A temporary shape change that is self-reversing after the force is removed, so that the object returns to its original shape, is called elastic deformation. In other words, elastic deformation is a change in shape of a material at low stress that is recoverable after the stress is removed. This type of deformation involves stretching of the bonds, but the atoms do not slip past each other. When the stress is sufficient to permanently deform the metal, it is called plastic deformation. As discussed in the section on crystal defects, plastic deformation involves the breaking of a limited number of atomic bonds by the movement of dislocations. Recall that the force needed to break the bonds of all the
15

atoms in a crystal plane all at once is very great. However, the movement of dislocations allows atoms in crystal planes to slip past one another at a much lower stress levels. Since the energy required to move is lowest along the densest planes of atoms, dislocations have a preferred direction of travel within a grain of the material. This results in slip that occurs along parallel planes within the grain. These parallel slip planes group together to form slip bands, which can be seen with an optical microscope. A slip band appears as a single line under the microscope, but it is in fact made up of closely spaced parallel slip planes as shown in the image.

Example 1.3: In a copper wiring drawing process, a 150mm diameter billet of Cu is drawn down to a final wire diameter of 2mm. a. b. c. d. What is the ductility measured in terms of area reduction? If the original billet is 2 meters long, how long is the final wire? What is is final engineering strain of the wire? What is the final true strain of the wire?

16

Answer: ) In a copper wiring drawing process... a) The area reduction ductility is given by equation 6.12:

b) Because volume is conserved during plastic deformation,

c) Use the definition of engineering strain...

d) ...and the definition of true strain:

Now, you have studied the above topics and of course you have some questions? If you have this the time to ask every think related? If you (not), the take the following assignments:

Assignments (1-1) 1. The results of a tensile test on a steel test bar are given below. The initial gauge length was 25.0 mm and the initial diameter was 5.00 mm. The diameter at the fracture was 2.6 mm. The engineering strain and engineering stress in MPa are as follows;

17

Strain 0.0 0.0002 0.0004 0.0006 0.0015 0.005 0.02 0.03 0.04 0.05

Stress 0.0 42 83 125 155 185 249.7 274.9 293.5 308.0

Strain 460.06 0.08 0.10 0.15 0.20 0.22 0.24 0.26 0.28 0.30

Stress 319.8 337.9 351.1 371.7 382.2 384.7 386.4 387.6 388.3 388.5

Strain 0.32 0.34 0.38 0.40 0.42 0.44 0.46 0.47

Stress 388.4 388.0 386.5 384.5 382.5 378.6 362.0 250

A. Plot the engineering stressstrain curve. B. Determine: a. Youngs modulus. b. The 0.2% offset yield strength. c. The tensile strength. d. The percent elongation, and e. The percent reduction of area. 2. Construct the true stresstrue strain curve for the material in Problem 1. Note that necking starts at maximum load, so the construction should be stopped at this point. 3. The tensile stress-strain curve for a cylindrical specimen of brass is reproduced in Figure 1. Find the following: a. The work done to fracture per unit volume: Answer: The work done to fracture per unit volume is the area underneath the Nominal Stress vs. Nominal Strain curve. If we had an equation to describe this curve, we could integrate the equation from zero to the strain at fracture. Since we dont have this equation available, estimate the work done by estimating the area

18

beneath the curve. Work done to fracture per unit volume ~ 20.5 boxes * 10 J/m^3/box = 205 J/m^3. b. The Young's modulus of brass Answer: The Youngs modulus is found from the slope of the Nominal Stress vs. Nominal Strain curve in in the linear regime where nominal strain is very small. In this case you should estimate the slope of the curve. E ~100 MPa/.0025 = 40,000 MPa. c. The ultimate tensile stress Answer: The ultimate tensile stress (UTS) occurs where the Nominal Stress vs. Nominal Strain curve shows a maximum. In this case the UTS ~530 MPa. d. The 0.2% offset yield stress Answer: To find the 0.2% offset yield stress, extrapolate a line with the same slope as the linear regime (which is equal to the Youngs modulus) from the x axis of Nominal Stress vs. Nominal Strain curve upwards until it intersects with the curve. The stress at this intersection is the 0.2% offset yield stress. Here it is ~ 150 MPa. e. The proportional limit Answer: The proportional limit occurs where the stress and strain stop being linear or proportional. This is sometimes used as a measure the yield stress (but not in this class) and is ~ 95 MPa. 4. Suppose a similar specimen is given a tensile prestrain of 0.2. On subsequent testing this specimen you would find an approximate yield stress of 492 MPa ? Answer: When the specimen is loaded again, the stress would be proportional to the strain, starting at the 0.20 strain, until it hits the normal stress-strain curve. Applying this to the first n n plot given, a nominal yield stress based on the original cross-sectional area Ao would be 410 MPa. However if as usual the crosssectional area Ao1 is measured before the second test we need to find this area in order to get the correct stress from the value we read off the original stress-strain curve. We know this stress (the true stress) = 410 Ao/Ao1 , where Ao/Ao1 = Lo1/Lo = 1+n = 1.20. The yield stress will be thus 492 MPa. Another way to look at this problem is as follows: The sample will not start to yield again (undergo permanent plastic deformation) until it reaches a stress equivalent to that stress it was experiencing the last time it was yielding. In this case yielding last occurred at a pre-strain of 0.2. We know that permanent plastic

19

deformation occurred during the pre-strain, so we know that the area of the sample changed from Ao to Ao1. Yield stress (y) is the point where elastic deformation stops and permanent plastic deformation begins. You can calculate the yield stress after the pre-strain, by converting the nominal stress of the pre-strain of 0.2 (this is the strain at which the sample last yielded) to what it would be with the new area Ao1. (As I stated earlier, the yield stress will occur at the a stress equivalent to the stress value it last yielded at. This calculation finds that equivalent stress.) However, as stated earlier, the sample has a new area, Ao1. So y = F/Ao*( Ao/Ao1)= n*( Ao/Ao1). Note, this is the same thing as calculating the true stress at a nominal strain of 0.2. 5. What is the true strain and true stress at a point of a curve in Figure 1 that corresponds to 0.35 nominal strain? Answer: At the nominal strain of 0.35 we can find the true strain by using the fact that true strain = ln(L/Lo) = ln(1+ n ) = 0.30. To find the true stress you need to multiply the nominal stress 520 MPa by Ao/A = 1+ n = 1.35. This gives a true stress of 702 MPa. Why can't you compute the real and at the nominal strain of 0.45 from this curve? Answer: This is because at nominal strain of 0.40,where the nominal stress-nominal strain curve is a maximum, necking starts. Hence at 0.45 nominal strain, necking has already occurred. Because the curve is obtained by monitoring the variation of nominal strain from changes in the length, beyond 0.40 strains the data cannot be used to calculate the true strain. 6. An annealed-steel specimen has a 0.505 in. diameter and a 2.000 in. gage length. Maximum load is reached at 15,000 lbs and fracture occurs at 10,000 lbs. For this steel, E = 30 x 106 psi. a. What is the ultimate tensile strength (UTS)?. b. Why does fracture occur at a lower load than the maximum load?. c. What is the elongation when a tensile stress of 15,000 psi is applied? 7. A 200-mm-long rod with a diameter of 2.5 mm is loaded with a 2000-N weight. If the diameter decreases to 2.2 mm, compute the: a. Final length of the rod.
20

b. True stress and strain at this load. c. Engineering stress and strain at this load. 8. A 5-cm-long circular rod (with diameter of 1.28 cm) of Ti-8Al-1Mo-1V Duplex annealed is loaded to failure in tension. (From Hertzberg, Sys = 950 MPa, SUTS= 1000 MPa and, for the specified gage length, there is a 15% elongation for a 5 cm gage.) a. What is load necessary to break the sample? b. If 75% of the total elongation was uniform in character prior to onset of localized deformation, compute the true stress at the point of incipient necking. c. What is the reduction in area at necking? d. Recall that uniform elongation, eu, is the elongation found at the UTS (with no load). 9. A cylinder of grey cast iron (i.e. it is brittle) 0.5 in in diameter by 2 in. long is tested in compression. Failure occurs at an axial load of 50,000 lbs on a plane inclined at 400 to the cylinder axis. a. Calculate the maximum shearing stress on the plane of failure. b. A cylinder of the same grey cast iron is now pulled in tension and fails in a brittle manner. What is the plane of failure? Should the failure load be greater or less than in part (a)? Can you estimate by how much less? 10. The stress at a point is defined by the components xx = 0 MPa, yy = 100 MPa,xy = -40 MPa. Find the principal stresses s1 and s2 and the inclination of the plane on which the maximum principal stress acts to the x plane. Use both the Mohr circle approach and the matrix method approach. 11. What deformation mechanisms are involved during the elastic and plastic deformation of thermoplastics? Answer: During elastic deformation, the covalent bonds of the molecular chains stretch. Subsequently, the chains uncoil in a process that involves both elastic and plastic deformation. Finally, plastic deformation occurs as secondary dipole bonds are broken, allowing the chains to slide relative to each other and establish new dipole bonding forces.

Note: As you see some of assignments are solved, others not..So you can take the unsolved as home works??

21

CHAPTER 2 FRACTURE ANALYSIS



2.1 Fracture Definition Fracture can be defined as the separation or fragmentation of a solid body into two or more parts under the action of stress. The process of fracture can be considered to be made up of two components: 1. Crack initiation. 2. Crack propagation. The fracture can be classified into two general categories: 1. Brittle fracture. 2. Ductile fracture. Fracture occurs in characteristics ways, depending on the following parameters: 1. State of stress. 2. Rate of stress applications. 3. Temperature. Note: Try to understanding what is the mean of parameters above? The objective of this chapter was to present a precise picture of the fundamentals of fracture of metals. Since most of the research has been concentrated on the problem of brittle fracture. The engineering aspects of brittle fracture will be considered in greater detail in this lecture.

22

2.2 Types of Fracture in Metals Metals can exhibit many different types of fracture, depending on the following parameters: 1. Material type. 2. State of stress. 3. Rate of loading. The two broad categories of ductile and brittle fracture have already been considered. The following figure is schematically illustrates some of the types of tensile fractures which can occur in metals.

Figure (2.1): types of fracture observed in metals subjected to uniaxial tension. (a) Brittle fracture of single crystal and polycrystalline; (b) shearing fracture in ductile single crystals; (c) completely ductile fracture in polycrystals; (d) ductile fracture in polycrystals.

Question: What are the characteristics of brittle fracture? The characteristics of brittle fracture (fig.2.1-a) can be summarized as follows: 1. Separation accurse normal to the tensile stress. 2. Outwardly there is no evidence of deformation, although with x-ray diffraction it is possible to detect thin layer of deformed metal at the fracture surface.

23

Brittle fracture have been observed in bcc and hcp metals, but not in fcc metals unless there are factors contributing to grain boundary embrittlement.? Can you answer such question?? Ductile fractures can take several forms as shown above: 1. Single crystals of hcp metals may slip on successive basal planes (define this term) until finally the crystal separates by shear (fig. 2.1-b). 2. Polycrystalline specimens of very ductile metals, like gold or lead, may actually be drawn down to a point before they rupture (fig. 2.1 c). 3. In the tensile fracture of moderately ductile metals the plastic deformation eventually produces a necked region (fig. 2.1 d). 4. Fracture begins at the center of the specimen and extends by a shear separation along the dashed lines in figure (2.1 d). This results in the familiar cup-and-cone fracture.

15th century- DaVinci conducted experiments on iron wire, longer wires failed at lower loads than shorter wires, any ideas why?

Fractures are classified with respect to several characteristics, such as: 1. Strain to fracture. 2. Crystallographic mode of fracture. 3. Appearance of the fracture. Gensamer has summarized the terms commonly used to describe fractures as follows: Behavior described Crystallographic mode Appearance of fracture Strain to fracture Terms used Shear Fibrous ductile Cleavage Granular brittle

So, how can we understand the table above?

24

A shear fracture occurs as the result of extensive slip on the active slip plane. This type of fracture is promoted by shear stresses. The cleavage mode of fracture is controlled by tensile stresses acting normal to a crystallographic cleavage plan. A fracture surface which is caused by shear appears at low magnification to be gray and fibrous, while a cleavage fracture appears bright or granular, owing to reflection of light from the flat cleavage surfaces. Fracture surfaces frequently consist of a mixture of fibrous and granular fracture, and it is customary to report the percentage of the surface area represented by one of these categories. Based on metallographic examination, fracture in polycrystalline samples is classified as either transgrannular (the crack propagates through the grains) or intergrannular (the crack propagates along the grain boundaries). A ductile fracture is one which exhibits a considerable degree of deformation. The boundary between a ductile and brittle fracture is arbitrary and depends on the situation being considered. For example, nodular cast iron is ductile when compared with ordinary gray iron; yet it would be considered brittle when compared with mild steel. As a further example, a deeply notched tensile specimen will exhibit little gross deformation; yet the fracture could occur by a shear mode.

2.3 Fracture Mechanics Why Study Fracture Mechanics? 1. The presence of cracks or crack-like defects in engineering structures cannot be precluded Current energy and material conservation considerations require that structures be designed with less safety margin, and hence greater tolerance to defects 2. Requirement to determine residual strength and life of cracked structure. Fracture mechanics measure a material constant which is directly applicable in design. In low carbon structural steel, when the crack propagates slowly, there is an evidence of plastic flow near the fracture, the new surface is fibrous and the material at least under the existing conditions of: 1. Temperature. 2. Strain rate. 3. Initial crack sizeis considered to be tough.

25

Under other conditions (e.g., a larger initial crack or a lower test temperature) the crack will grow catastrophically with little or no evidence of deformation. It will have a cleavage surface appearance, and the material will behave brittly. In dead, under the latter case, crack growth is usually an instability phenomenon, and propagation continues even with a decreasing load. In dealing with the Critical Stress Intensity factor usually expressed as K Ic, and it is a true material property. There is a minimum thickness in order to be certain that plane strain conditions exist, and there are several modes of deformation that could be applied to the crack. These have been standardized as shown in the figure shown below:

Or see the following:

Figure (2.2): Crack modes

26

In the above figure which represent the crack modes: 1. Mode I or the so called crack-opening mode and refers to tensile stress applied in y-direction normal to the crack faces. This is the usual mode for fracture toughness tests and a critical value of stress intensity determined for this mode would be designated as KIc. 2. Mode II, the foreword shear mode, refers to a shear stress applied normal to the leading edge of the crack in the plane of the crack. 3. Mode III, the parallel shear mode, is for shearing stresses applied parallel to the leading edge of the crack. Mode I loading is the most important situation. There are two extreme cases for mode I loading. With thin plate type specimens the stress state is plane stress, while with thick specimens there is a plane-strain condition. Stress Intensity Factors It is instructive to examine the concentrated stresses at the edge of an elliptical hole in a thin plate, when the plate is subjected to tensile stresses

Figure (2.3): Effect of specimen thickness on the stress and mode of fracture.

Plane-strain conditions represent the more severe stress state and the value of KIc is lower than for plane-stress specimens. Plane-strain values of critical intensity factor KIc are valid material properties independent of specimen thickness to

27

describe the fracture toughness of strong materials like heat-treated state, aluminum and titanium alloys.

2.4 Brittle Fracture Brittle fracture in metals is characterized by the following characterizations: 1. Rapid rate of crack propagation with no gross deformation and very little microdeformation. 2. It is akin to cleavage in ionic crystals. The tendency for brittle fracture is increased with the following: 1. Decreasing of temperature. 2. Increasing of strain rate and triaxial stress conditions Brittle fracture must be avoided at all conditions and cost because it occurs without warning and usually produces disastrous consequences.

2.4.1 Griffith Theory of Brittle Fracture Griffith approached the subject of fracture by assuming that materials always have preexisting cracks. He considered a large plate with a central crack under a remote stress, , and calculated the change of energy, U, with crack size (Figure 14.3). There are two terms: One is the surface energy associated with the crack,

28

Figure (2.4): The effect of a cracks length on its energy. The elastic energy decreases and the surface energy increases. There is a critical crack length at which growth of the crack lowers the total energy.

Where t is the plate thickness and 2a is the length of an internal crack. The other term is the decrease of stored elastic energy due to the presence of the crack

Combining Equations above:

This equation predicts that the energy of the system first increases with crack length and then decreases, as shown in Figure (14.3). Under a fixed stress, there is a critical crack size above which crack growth lowers the energy. This critical crack size can be found by differentiating Equation (14.15) with respect to a and setting to zero:

29

This is known as the Griffith criterion. A preexisting crack of length greater than 2a will grow spontaneously when Equation (2.16) is satisfied. Griffith found reasonable agreement between this theory and experimental results on glass. However, the predicted stresses were very much too low for metals. In this development, the plate is assumed to be thick enough relative to the crack length so that there is no strain relaxation in the thickness direction; that is, plane-strain conditions prevail. If, on the other hand, the plate is very thin, so that there is full relaxation in the thickness direction (plane-stress conditions), Equation () should be modified to

Modification of Griffith Theory Orowan theory Orowan proposed that the reason that the predictions of Equations () and were too low for metals is that the energy expended in producing a new surface by fracture is not just the true surface energy. There is a thin layer of plastically deformed material at the fracture surface and the energy to cause this plastic deformation is much greater than . To account for this, Equation () is modified to:

Where Gc replaces 2 and includes the plastic work of generating the fracture surface.

2.5 Ductile Fracture Failure in a tensile test of a ductile material occurs well after the maximum load is reached and a neck has formed. In this case, fracture usually starts by nucleation of voids in the center of the neck, where the hydrostatic tension is the greatest. As deformation continues, these internal voids grow and eventually link
30

up by necking of the ligaments between them (Figures 2.3, 2.4, and 2.5). Such a fracture starts in the center of the bar where the hydrostatic tension is greatest. With continued elongation, this internal fracture grows outward until the outer rim can no longer support the load and the edges fail by sudden shear. The final shear failure at the outside also occurs by void formation and growth (Figures 2.6 and 2.7). This overall failure is often called a cup and cone fracture. If the entire shear lip is on the same broken piece, it forms a cup. The other piece is the cone (Figure 2.8). More often, however, part of the shear lip is on one half of the specimen and part on the other half. In ductile fractures, voids form at inclusions because either the inclusion-matrix interface or the inclusion itself is weak. Figure 2.9 shows the fracture surface formed by coalescence of voids. The inclusions can be seen in some of the voids. Ductility is strongly dependent on the inclusion content of the material. With increasing numbers of inclusions, the distance between the voids decreases, so it is easier for them to link together and lower the ductility. Figure 2.10 shows the decrease of ductility of copper with volume fraction inclusions. Ductile fracture by void coalescence can occur in shear as well as in tension testing.

Figure 2.5 Section through a necked tensile specimen of copper, showing an internal crack formed by linking voids.

31

Figure (2.6): Schematic drawing showing the formation and growth of voids during tension and their linking up by necking of the ligaments between them.

Figure(2.7). Schematic drawing illustrating the formation and growth of voids during shear, and their growth and linking up by necking of the ligaments between them. Example: Estimate the theoretical fracture strength of a brittle material if it is known that fracture occurs by the propagation of an elliptically-shaped surface crack of length 0.5 mm and having a tip radius of curvature of 5 x 10 -3 mm when a stress of 1035 MPa is applied. Solution: This uses the formula

32

am= 2 0(a/rhot)0.5 where sigma0, the applied stress, is 1035 MPa, a is 0.0005 m, and rhot is 0.000005 m. Evaluating the formula then gives an effective stress of 20.7 GPa. Example: A specimen of 4340 steel alloy with a plane strain fracture toughness of 54.8 MPa m0.5 is exposed to a stress of 1030 MPa. Will this specimen experience fracture if it is known that the largest surface crack is 0.5 mm long? Why, or why not? (Assume that the parameter Y has a value of 1.0.) Solution: The critical stress, sigma, can be calculated from the equation KIC = Y(a)0.5 where KIC=54.8 * 106, Y=1 and a = 0.5 * 10-3. Using these figures, sigma evaluates to 1382 MPa. The applied stress, 1030 MPa, is less than this, so fracture will not occur.

Assignments (2-2) 1. What are the characteristics of the surface of a ductile fracture of a metal? Answer A ductile fracture surface has a dull, fibrous appearance and often resembles a cup-andcone configuration. 2. What are the characteristics of the surface of a brittle fracture of metal? Answer A brittle fracture surface typically appears shiny with flat facets, which are created during cleavage fracture. 3. How does the carbon content of a plain-carbon steel affect the ductile brittle transition temperature range? 4. Answer

33

As the carbon content of the steel increases, the ductile-brittle transition temperature range increases in terms of both the width of the range and the temperature values. 5. Draw a typical creep curve for a metal under constant load and at a relatively high temperature, and indicate on it all three stages of creep. 6. Using the equation KIC= f (a)-0.5 plot the fracture stress (MPa) for aluminum alloy 7075-T651 versus surface crack size a (mm) for a values from 0.2 mm to 2.0 mm. What is the minimum size surface crack that will cause catastrophic failure? Answer 7.25. 7. Compute the stresses (normalized by KI /(2a)0.5 around a crack loading in plane strain and mode I at: =0, 45, 90, 135, and 180o . Draw the principal directions of stress in each location. 8. 2. Find two examples of inclusions with significant differences of strength than the surrounding matrix (one harder inclusion and one softer than the matrix). Provide data (with references of the strength of each phase, i.e., inclusion and matrix). Photos are welcome. 9. 3. Find as many as possible online, fully readable textbooks and handbooks on fracture and fatigue that exist in our library (In about 10 minutes I found 8; see if you can find more.

34

CHAPTER 3 DISLOCATION THEORY


3.1 Introduction It was well known in the late 19th century that crystals deformed by slip. In the early 20th century, the stresses required to cause slip were measured by tension tests of single crystals. Dislocations were not considered until after it was realized that the measured stresses were far lower than those calculated from a simple model of slip were. In the mid-1930s G. I. Taylor, M. Polanyi, and E. Orowan independently postulated that pre-existing crystal defects (dislocations) were responsible for the discrepancy between measured and calculated strengths. It took another two decades and the development of the electron microscope for dislocations to be observed directly. Slip occurs by the motion of dislocations. Dislocations can explain many aspects of the plastic behavior of crystalline materials. Among these are how crystals can undergo slip, why visible slip lines appear on the surfaces of deformed crystals, why crystalline materials become harder after deformation, and how solute elements affect slip. 3.2 Theoretical strength of crystals Once it was established that crystals deformed by slip on specific crystallographic systems, physicists tried to calculate the strength of crystals. However, the agreement between their calculated strengths and experimental measurements was very poor. The predicted strengths were orders of magnitude too high, as indicated in Table 9.1. The basis for the theoretical calculations is illustrated in Figure 9.1. Each plane of atoms nestles in pockets formed by the plane below (Figure 9.1a). As a shear stress is applied to a crystal, the atoms above the plane of shear must rise out of the stable pockets in the plane below and slide over it until they reach the unstable position shown in Figure 9.1b. From this point they will spontaneously continue to shear to the right until they reach a new stable position (Figure 9.1c).

35

For simplicity, consider a simple cubic crystal. An applied shear stress, , will displace one plane relative to the next plane as shown in Figure 9.2. When the shear displacement, x, is 0, d, 2d, or nd (i.e., = 0, 1, 2. . . n), the lattice is restored, so should be zero. The shear stress, , is also zero when the displacement is x = (1/2) d, (3/2) d, etc. ( = 1/2, 3/2 . . .). A sinusoidal variation of with as shown in Figure 9.3 seems reasonable, so:

Here max is the theoretical shear stress required for slip. If the stress is less than max, the shear strain is elastic and will disappear when the stress is released. For very low values of (Figure 3.4), Hookes law should apply:

36

A somewhat more sophisticated analysis for real crystal structures predicts something close to:

In Table 3.1, the theoretical strength values predicted by Equation (3.5) are orders of magnitude higher than experimental measurements. The poor agreement between experimental and theoretical strengths indicated that there was something wrong with the theory. The problem with the theoretical calculations is that it was assumed that slip occurs by one entire plane of atoms sliding over another at the same time. Taylor, Orowan, and Polanyi realized that it is not necessary for a whole plane to slip at the same time. They postulated that crystals have preexisting
37

defects that are boundaries between regions that are already displaced relative to one another by a unit of slip. These boundaries are called dislocations. Movement of a dislocation allows slip to occur. The critical stress for slip is the stress required to move a dislocation. At any instant, slip need only occur at the dislocation rather than over the entire slip plane.

3.3 The nature of dislocations One special form of dislocation is an edge dislocation, sketched in Figure 9.5. The geometry of an edge dislocation can be visualized as having cut partway into a perfect crystal and then inserted an extra half plane of atoms. The dislocation is the bottom edge of this extra half plane. The screw dislocation (Figure 3.6) is another special form. It can be visualized as a spiral-ramp parking structure. One circuit around the axis leads one plane up or down. Planes are connected in a manner similar to the levels of a spiral-parking ramp. An alternate way of visualizing dislocations is illustrated in Figure 9.7. An edge dislocation is created by shearing the top half of the crystal by one atomic distance perpendicular to the end of the cut (Figure 3.7b). This produces an extra half plane of atoms, the edge of which is the center of the dislocation. The other extreme form of dislocation is the screw dislocation. A screw dislocation is generated by cutting into a perfect crystal and then shearing half of it by one atomic distance in a direction parallel to the end of the cut (Figure 3.7c). The end of the cut is the dislocation. In both cases the dislocation is a boundary between regions that have and have not slipped. When an edge dislocation moves, the direction of slip is perpendicular to the dislocation. In contrast, movement of a screw dislocation causes slip in the direction parallel to itself. The edge and screw are extreme cases. A dislocation may be neither parallel nor perpendicular to n the slip direction.

38

3.4 Burgers vectors Dislocations are characterized by Burgers vectors. Consider an atom-to-atom circuit in Figure 9.8 that would close on itself if made in a perfect crystal. If this same circuit is constructed so that it goes around a dislocation, it will not close. The closure failure is the Burgers vector, denoted by b. The Burgers vector can be considered a slip vector a. b. c.

39

direction and its magnitude is the magnitude of the slip displacement caused by its movement of the dislocation. A dislocation may wander through a crystal with its orientation changing from place to place, but its Burgers vector is the same everywhere. See Figure 3.9. If the dislocation branches into two dislocations, the sum of the Burgers vectors of the branches equals its

3.5 Burgers vector. There is a simple notation system for describing the magnitude and direction of the Burgers vector of a dislocation. The direction is indicated by direction indices and because its direction is the slip direction. For example, b = (a/3)[211] in a cubic crystal means that the Burgers vector has components of 2a/3, a/3, and a/3 along the [100], [010], and [001] directions, respectively, where a is the lattice parameter. Its magnitude is |b| = [(2a/3)2 + (a/3)2 + (a/3)2]1/2 = a6/3).

3.6 Energy of a screw dislocation The energy associated with a screw dislocation is the energy required to distort the lattice surrounding the dislocation elastically. The distortion is severe near the dislocation but decreases with distance from it. Consider an element of length L and thickness dr at a distance r from the center of a screw dislocation (Figure 3.10a). The volume of the element is 2rLdr. Imagine unwrapping this element, so that it lies flat (Figure 3.10b). Unwrapping causes no strain because the element is differentially thin. The shear strain associated with this element is = b/(2r ), where b is the Burgers vector of the screw dislocation. The energy/volume associated with an elastic distortion is Uv = (1/2), where is the shear stress necessary to cause the shear strain . According to Hookes law, = G, where G

40

is the shear modulus. Combining Equations (9.6), (9.7), and (9.8), the energy/ volume is Uv = G 2/2 = (1/2)Gb2/(2r )2. The elastic energy, dU, associated with the element is its energy/volume times its volume, dU = [(1/2)Gb2/(2r )2](2rLdr ) = Gb2L/4)dr/r.

The total energy of the dislocation per length, L, is obtained by integrating Equation (3.10): U/L = Gb2/(4) = Gb2/(4) ln(r1/r0).

3.7 Dislocation pile-ups When dislocations from a FrankRead source come to an obstacle such as a grain boundary or hard particle, they tend to form a pile-up (Figure 3.7). Because they are of like sign they repel one another. The total repulsion of a dislocation by a pile-up is the sum of the repulsions of each dislocation in the pile-up. With n dislocations in the pile-up, the stress on the leading dislocation, n, will be n = n. It takes a relatively small number in a pile-up to effectively stop further dislocation movement. A pile-up creates a back stress on the source so that the stress to continue to operate the source must rise. This higher stress allows other sources of slightly smaller spacing, d, to operate.

3.8 Cross-slip
41

A dislocation cannot move on a plane unless the plane contains both the dislocation and its Burgers vector. This requirement uniquely determines the slip plane, except in the case of screw dislocations. Screw dislocations are parallel to their Burgers vectors, so

They can slip on any plane in which they lie. When a screw dislocation gliding on one plane changes to another, it is said to undergo cross-slip. In principle, screw dislocations should be able to cross-slip with ease to avoid obstacles (Figure 10.8). If, however, screw dislocations are separated into partials connected by a stacking fault, both partials cannot be screws. They must recombine to form a screw dislocation before they can dissociate on a second plane (Figure 3.9). Such recombination increases the total energy and this energy must be supplied by the applied stresses, aided by thermal activation. How much energy is required depends on the degree of separation of the partials and therefore on the stacking fault energy. If the stacking fault energy is high, the separation of partial dislocations is small, so the force required to cause recombination is low. In metals of high stacking fault energy (e.g., aluminum) cross-slip occurs frequently. In crystals of low stacking fault energy (e.g., brass) the separation of partials is large. A high force is required to bring them together, so cross-slip is rare. This is in accord with observation of slip traces on polished surfaces. Slip traces are very wavy in aluminum and very straight in brass (Figure 3.10). Double cross-slip has been proposed as a variant of the original FrankRead source. In this case, a segment of dislocation, which has undergone double crossslip, can act as a FrankRead source on its new plane, as shown in Figure 3.11.
42

3.9 Dislocation intersections The number of dislocations increases as deformation proceeds and the increased number make their movement becomes more difficult. This increased difficulty of movement is caused by intersection of dislocations moving on different planes. During easy glide the rate of work-hardening is low because slip occurs on parallel planes and there are few intersections. As soon as slip occurs on more than one set of slip planes, dislocations on different planes will intersect and these intersections impede further motion, causing rapid work-hardening. The nature of dislocation intersections can be understood by considering several types of intersections in simple cubic crystals, as illustrated in Figure 10.12. When two dislocations intersect, a jog is created in each dislocation. The direction of the jog is parallel to the Burgers vector of the intersecting dislocation and the length of the jog equals the magnitude of the Burgers vector of the intersecting dislocation. For dislocations a and b in Figure 10.12, the jogs create no problem. If the upper part of dislocation a and the right side of dislocation b moves slightly faster, the jogs will disappear. The same is true for dislocation c if its left side moves faster. The jog in dislocation d simply represents a ledge in the extra half plane, and it can move with the rest of the

43

dislocation. However, the jogs in dislocations e and f cannot move conservatively. The jogs have an edge character and the direction of motion is not in their slip plane. Figure 3.13 is an enlarged view of dislocations e and f in Figure 3.12. Figure 3.14a shows that continued motion of these jogged dislocations would force atoms into interstitial positions. Figure 3.14b shows that with jogs of the opposite sense, vacancies would be produced. If the intersection had been in the opposite direction, movement of the jogs would create a row of vacancies. Figure 3.15 shows rows of vacancies being created by moving jogs.

3.10 Climb The movement of an edge dislocation out of its glide plane can occur only if there is a net diffusional flux of atoms away from or toward the dislocation. Such motion is called climb. Removal of atoms from a dislocation causes upward climb and addition of atoms to the dislocation causes downward climb (Figure 3.17). Because climb requires diffusion, it is important only at elevated temperatures during creep. Diffusion-controlled climb allows dislocations to avoid obstacles that impede their glide and may become a controlling mechanism during creep (Chapter 16).

44

45

CHAPTER 4 FATIGUE FAILURE


3.1 Introduction: Fatigue is a form of failure that occurs in structures subjected to dynamic and fluctuating stresses (e.g., bridges, aircraft, and machine components). Under these circumstances it is possible for failure to occur at a stress level considerably lower than the tensile or yield strength for a static load. The term fatigue is used because this type of failure normally occurs after a lengthy period of repeated stress or strain cycling. Fatigue is important inasmuch as it is the single largest cause of failure in metals, estimated to comprise approximately 90% of all metallic failures; Polymers and ceramics (except for glasses) are also susceptible to this type of failure. Furthermore, it is catastrophic and insidious, occurring very suddenly and without warning. Fatigue failure is brittle like in nature even in normally ductile metals, in that there is very little, if any, gross plastic deformation associated with failure. The process occurs by the initiation and propagation of cracks, and ordinarily the fracture surface is perpendicular to the direction of an applied tensile stress. 3.2 Cyclic Stress The applied stress may be axial (tension-compression), flexural (bending), or torsional (twisting) in nature. In general, three different fluctuating stresstime modes are possible. One is represented schematically by a regular and sinusoidal time dependence in Figure 3.1a, wherein the amplitude is symmetrical about a mean zero stress level, for example, alternating from a maximum tensile stress (max) to a minimum compressive stress (min) of equal magnitude; this is referred to as a reversed stress cycle. Another type, termed repeated stress cycle, is illustrated in Figure 3.1b; the maxima and minima are asymmetrical relative to the zero stress level. Finally, the stress level may vary randomly in amplitude and frequency, as Notch Toughness of Pearlitic Steels,

46

The process of fatigue failure is characterized by three distinct steps: (1) crack initiation, wherein a small crack forms at some point of high stress concentration; (2) crack propagation, during which this crack advances incrementally with each stress cycle; and (3) final failure, which occurs very rapidly once the advancing crack has reached a critical size. The fatigue life Nf , the total number of cycles to
47

failure, therefore can be taken as the sum of the number of cycles for crack initiation Ni and crack propagation Np : The contribution of the final failure step to the total fatigue life is insignificant since it occurs so rapidly. Relative proportions to the total life of Ni and Np depend on the particular material and test conditions. At low stress levels (i.e., for highcycle fatigue), a large fraction of the fatigue life is utilized in crack initiation. With increasing stress level, Ni decreases and the cracks form more rapidly. Thus, for low-cycle fatigue (high stress levels), the propagation step predominates (i.e., Np _ Ni). Cracks associated with fatigue failure almost always initiate (or nucleate) on the surface of a component at some point of stress concentration. Crack nucleation sites include surface scratches, sharp fillets, keyways, threads, dents, and the like. In addition, cyclic loading can produce microscopic surface discontinuities resulting from dislocation slip steps which may also act as stress raisers, and therefore as crack initiation sites. Once a stable crack has nucleated, it then initially propagates very slowly and, in polycrystalline metals, along crystallographic planes of high shear stress; this is sometimes termed stage I propagation (Figure 9.28). This stage may constitute a large or small fraction of the total fatigue life depending on stress level and the nature of the test specimen; high stresses and the presence of notches favor a shortlived stage I. In polycrystalline metals, cracks normally extend through only several grains during this propagation stage. The fatigue surface that is formed during stage I propagation has a flat and featureless appearance.

48

Eventually, a second propagation stage (stage II ) takes over, wherein the crack extension rate increases dramatically. Furthermore, at this point there is also a change in propagation direction to one that is roughly perpendicular to the applied tensile stress (see Figure 9.28). During this stage of propagation, crack growth proceeds by a repetitive plastic blunting and sharpening process at the crack tip, a mechanism illustrated in Figure 9.29. At the beginning of the stress cycle (zero or
49

maximum compressive load), the crack tip has the shape of a sharp double-notch (Figure 9.29a). As the tensile stress is applied (Figure 9.29b), localized deformation occurs at each of these tip notches along slip planes that are oriented at 45_ angles relative to the plane of the crack. With increased crack widening, the tip advances by continued shear deformation and the assumption of a blunted configuration (Figure 9.29c). During compression, the directions of shear deformation at the crack tip are reversed (Figure 9.29d) until, at the culmination of the cycle, a new sharp double-notch tip has formed (Figure 9.29e). Thus, the crack tip has advanced a one-notch distance during the course of a complete cycle. This process is repeated with each subsequent cycle until eventually some critical crack dimension is achieved that precipitates the final failure step and catastrophic failure ensues. The region of a fracture surface that formed during stage II propagation may be characterized by two types of markings termed beachmarks and striations. Both of these features indicate the position of the crack tip at some point in time and appear as concentric ridges that expand away from the crack initiation site(s), frequently in a circular or semicircular pattern. Beachmarks (sometimes also called clamshell marks) are of macroscopic dimensions (Figure 9.30), and may be observed with the unaided eye. These markings are found for components that experienced interruptions during stage II propagationfor example, a machine that operated only during normal work-shift hours. Each beachmark band represents a period of time over which crack growth occurred. On the other hand, fatigue striations are microscopic in size and subject to observation with the electron microscope (either TEM or SEM). Figure 9.31 is an electron fractograph which shows this feature. Each striation is thought to represent the advance distance of the crack front during a single load cycle. Striation width depends on, and increases with, increasing stress range.

50

At this point it should be emphasized that although both beachmarks and striations are fatigue fracture surface features having similar appearances, they are nevertheless different, both in origin and size. There may be literally thousands of striations within a single beach mark. Often the cause of failure may be deduced after examination of the failure surfaces. The presence of beachmarks and/or striations on a fracture surface confirms that the cause of failure was fatigue. Nevertheless, the absence of either or both does not exclude fatigue as the cause of failure.

51

CHAPTER 5 CREEP FAILURE


What is creep Creep is time-dependent plasticity occurring at stresses below the flow stress and at temperatures in excess of Tm/3. Creep occurs by a number of mechanisms, some well characterized others not, but all of which have an element of thermal activation which enables plastic deformation at stresses below those needed to deform the lattice without thermal activation. Creep is critical in a number of applications: steels used in power and chemical plants where the service temperature is high, turbine components, ice and lead at ambient temperatures.

Primary creepdecreasing creep rate as dislocation microstructure develops to reduce strain rate Secondary creep-equilibrium is established between deformation and recovery mechanisms to maintain a steady state strain rate. Tertiary creep increasing creep rate as the effective cross section reduces leading to failure typically logarithmic curve.

52

CREEP DATA Creep strength:-stress to produce nominal strain (normally 1.0 %) in given time typically 10,000 h (1 year), for aeronautical industry and 100,000 h (11years), for power plant. Ideally timescale of tests should be comparable with the timescale of service exposure. Creep life:-Time to nominal 1% or 2% stress at given stress and temperature. Minimum creep rate:measured as a function of stress and temperature. Tests expensive focus on accurate value for the minimum creep rate need to maintain constant stress and temperature over very long periods of time. Tests can last 100,000 h. Stress Rupture life:the time to break the specimen at a particular stress and temperature. Use of the word rupture implies tests done at high stress these are quick and cheap but the strain is not measured accurately. MONKMAN-GRANT RELATIONSHIPF or many materials the steady state regime dominates the creep life and the creep life is thus mainly dependent on the minimum creep rate, . This relationship is expressed as the MonkmanGrant relationship:Eq. 3.1where C is a constant and tRis the life to rupture. TEMPERATURE AND STRESS DEPENDENCE Empirically the minimum creep rate is well described by the expression:Eq. 3.2where o and , are creep constants, n takes values from 3-8 at high stress and 1 at low stress, and EA is the activation energy for creep which is usually the same as the diffusive process by which the creep occurs, e.g. self diffusion.

53

LARSON-MILLER PARAMETER:A creep test can be accelerated by raising the temperature T and the time t is adjusted using the activation energy, Q, assuming this to be the same as at the service temperature. Where t is the time to rupture (or to some defined strain e.g. 1%) Q is the activation energy for creep, R is the gas constant and C is a constant for the material and has a value between 30-65, but 46 is typical for most metals. (a value of 20 will be quoted if log10 are used.) CREEP MECHANISMS:

54

Strategies for creep resistance: 1. Removal of grain boundaries to prevent diffusion creep. 2. Addition of refractory elements as solid solution strengtheners and in reducing diffusion coefficients. 3. The effect of ordered precipitates in strengthening the materials and maintaining very high yield stresses at high temperatures. 4. At lower stresses these precipitates force dislocation activity into the narrow gamma channels. 5. By manipulating the misfit rafting is caused to prevent dislocation climb. 6. Dislocation networks are created at the interface further impeding dislocation movement. 7. Avoid formation of TCP phases limit Re, Cr, W and Mo

Example: A cylindrical 1045 steel bar is subjected to repeated compression-tension stress cycling along its axis. If the load amplitude is 66,700 N, compute the minimum allowable bar diameter to ensure that fatigue failure will not occur. Assume a safety factor of 2.0. Solution:

55

This question relies on the fact that steels have a fatigue limit, that is, a limiting stress below which fatigue failure will not occur. From Figure 8.44 (Callister, 6th Edition), this limit is about 315 MPa (anything from 310 to 320 MPa is acceptable here). So the stress amplitude must be less than 315/2 MPa, given the safety factor of 2. So 315,000,000/2 = 66,700/(pi d2/4) from which d = 23.2 mm. (Anything in the range 22-24 mm is acceptable.)

Question: How does the SN curve of a carbon steel differ from that of a high strength aluminum alloy? Answer A fatigue test SN curve is a plot of the fatigue stress to which a specimen is subjected versus the corresponding cycles, or stress reversals, up to and including the point of failure. The number of cycles is plotted on a logarithmic scale while the fatigue strength is plotted on either a linear or logarithmic scale, depending on the data. The SN data is typically obtained by repeatedly subjecting a rotating specimen to reverse or fluctuating bending while counting the cycles until destruction occurs. However, specimens may also be subjected to reversed or fluctuating axial stresses, torsional stresses, or combined stresses in testing. Question: Where do fatigue failures usually originate on a metal selection? Answer Fatigue failures typically originate at a point of high stress concentration, such as a sharp corner or a notch.

56

You might also like