Up until the 1970s, rails for passenger and freight trains were regarded as relatively simple undemanding products and the specifications had changed very little for decades. However, investments in railway systems, the advent of high-speed passenger trains and the requirement for longer life track imposed a demand for rails of high quality, greater strength and tighter geometric tolerances. Therefore there have been major innovations in the past 20 years in terms of the method of manufacture, degree of inspection and range of products.

Figure 1: Evolution of Sections

Rail steel is extremely tough. As figure 2, rail steel resists breakage even after the yield point is exceeded. In addition, rail steel has satisfactory amount of ductility and after re-heating, can be used to complete most forming operations.

Their average yield point is greater than 60,000 PSI, while actual tensile strength normally ranges from 100,000 PSI to 130,000 PSI. This high yield point means rail steel provides ample stiffness, enduring heaviest demands with little deformation.

Figure 2: Stress-Strain values different grades of steels

Even after years of service and high stress, there is no difference between the grain structure of a used rail and a new rail. Age, traffic and weather do not change its basic properties. All stresses are relieved through heating prior to being re-rolled. This re-rolling, in accordance with ASTM-A-499, decreases the rails’ grain size, and that means improved resiliency. The additional working of the steel actually makes it better than when it was a rail!

Properties of Quality

So what makes rail steel superior to other steels? The simplest answer is its unique composition.

As we mentioned above, the rails are subject to heavy contact cyclic loading that accompanies increased car size and loading, to 100 and 125 ton capacity, increased train size, and increased train speeds used to transport bulk products over the last several decades. These increasing demands require manufacturing and metallurgical approaches that offset wear and other types of failure that limit rail life.

An early type of rail failure was associated with entrapped hydrogen that produced shatter crack or flakes in heavy rail sections, but that difficulty has been effectively controlled cooling and by vacuum degassing of liquid steel.

Typically, rail steels are produced in large BOS vessels and are vacuum degassed prior to being continuously cast into large blooms. Vacuum degassing, coupled with ladle trimming facilities, permits very tight control over chemical composition. After casting, the blooms are placed in insulated boxes, whilst still at a temperature of about 600°C, and are cooled at a rate of 1°C per hour for a period of three to five days. This treatment, coupled with prior vacuum degassing, reduces the hydrogen level in the finished rail to about 0.5 ppm, thereby reducing substantially the susceptibility to hydrogen cracking.

The blooms are then reheated and rolled directly to the finished rail profile. The rail produced from each bloom is hot sawn to specific lengths prior to passage through a rotary stamping machine en route to the cooling areas.

Depending upon the properties required, the rails are either cooled normally in air or subjected to enhanced cooling for the development of high strength. On cooling to room temperature, the rails are passed through a roller-straightener machine which subjects the section to a number of severe bending reversals and emerge with a very high degree of straightness. Finally, the rails pass through a series of ultrasonic, eddy current and laser inspection stations which monitor non-metallic inclusions, external defects and the flatness of the running surface.

Fracture mechanics can be used in three major areas: (i) design; (ii) material selection and alloy development; and (iii) determining the significance of defects. Ancillary areas are (iv) monitoring and control, and failure analysis. Each of these areas will be expanded upon throughout this article.

Conventional design procedures are based upon the yield strength or ultimate tensile strength. This approach was considered to be relatively safe when appropriate safety factors were used. Instances where unstable fracture occurred at stresses below the yield stress however necessitated making provision for such circumstances. Fracture mechanics provides this alternative in terms of the KIC value, but cyclic stressing and fatigue crack growth rate also have to be taken into account.

Factors of safety can then be used on the initial defect size, the working stress, and/or the anticipated number of loading cycles. An example of the practical application to design may be given in the main structural framework. Another example is the design against brittle fracture of a main coolant pump in a nuclear reactor primary circuit. The housing was constructed from three C-1/2%Mo steel castings which were welded together to form a casing having wall thickness from 4 in. to 22 in. and weighing about 32 tons, after internal surfaces had been clad with stainless steel. Because of the high toughness of the steel, an LEFM approach was not valid and COD and J-integral tests were carried out over a range of temperatures (-50 to +70°C) (-58 to +158°F).

A good correlation between these parameters was found. The finite element method was used to determine the stress fields for proof test loading and three service conditions, start, running and stop. The analysis carried out provided valuable information for design engineers, on the stress distribution and critical defect sizes in the castings. Charpy V-notch and drop-weight tests were also made and will be used in the future for quality control purposes.

Since the 1940’s, the problem of brittle fracture has been extensively studied. It has been found that with such low-stress (compared to the yield stress of the material) fractures always originate at flaws or cracks of various types. The fracture-mechanics approach to residual static strength in the presence of a crack makes use of the stress intensity factor KI concept to describe the stress field at a crack tip; when KI reaches a critical value KC the crack extends, usually catastrophically. Values of KI are known for a wide range of crack configurations, and the fracture-mechanics approach has proved useful in problems of material development, design, and failure analysis. In view of its success in dealing with static fracture problems, it is logical to use a similar general approach to analyze fatigue crack growth data.

In the mid 20th century, many researchers stated how early in the fatigue life they could observe micro cracks. Since then it has been clear that the fatigue life under cyclic loading consists of two phases: the crack initiation life, followed by the crack growth period until failure. This can be demonstrated in a block diagram, see Figure 1. The crack initiation period may cover a large percentage of the fatigue life under high-cycle fatigue, i.e. under stress amplitudes just above the fatigue limit. However, for larger stress amplitudes the crack growth period can be a substantial part of the fatigue life.

Figure 1: Various periods of fatigue life and applicable considerations

For the range of the elastic stress intensity factor, ΔK, alternative parameters were developed to correlate crack propagation rates under conditions of elastoplastic crack growth, as follows: (i) crack tip plastic range, (ii) change in crack tip opening displacement, and (iii) cyclic J-integral.

As early as the 1960s it was known that the correlation of da/dN and ΔK depends on the stress ratio R. This was to be expected, because an increased mean stress for constant loading ΔS should give faster crack growth while the R-value is also increased. Furthermore, the results of crack growth tests indicated systematic deviations of the Paris equation at relatively high and low ΔK values.

This led to the definition of three regions in da/dN-ΔK diagrams, namely zones I, II, and III, see Figure 2. Evident questions are connected with the vertical asymptotes at the lower ΔK boundary of zone I and the upper ΔK boundary of zone III. The latter boundary seems to be reasoned, since if Kmax exceeds the fracture toughness (either KC or KIC), a quasi-static failure will occur and fatigue crack growth is no longer feasible. Further, it should be identified that the Kmax value causing specimen failure in the last cycle of a fatigue crack growth test may well very from KC or KIC measured in a fracture toughness experiment.

From the standpoint of fracture mechanics, the incidence of a lower boundary in region I is not so obvious. If a K-value can be defined for the tip of a crack, a singular stress field should be on the scene and micro-plasticity at the tip of the crack should abound. Thus, why should the crack not propagate any more; for which physical reason should there be a threshold ΔK-value (ΔKth). New inspirations on ΔKth were connected with observations on so-called small cracks. These cracks occur as micro-cracks at the beginning of the fatigue life starting at the material surface or, more exactly, in the subsurface.

Figure 2: Typical Crack Growth Rate vs. Stress Intensity Range

An example of the practical application to design may be given in the main structural framework of the Beaubourg Centre, Paris, France. The design of the Beaubourg Centre was the subject of an international architectural competition, launched in 1971. Of the 681 schemes submitted, the winning architects were Piano and Rogers, and the winning engineering design was that of Ove Arup and Partners. The latter were inspired to use steel castings in their design following a visit to Japan, where they saw spherical cast steel nodes in the three dimensional structures at “Osaka 1970″ which possessed a clean simplicity rarely found elsewhere.

The Beaubourg building has six floors and is divided lengthwise into 13 bays, each measuring 12.8 x 48 m (42 x 157.5 ft.) and are uninterrupted by internal load-carrying columns. Because cast steel was a new material to the design engineers, it was necessary to devise, in collaboration with the constructional engineers, new methods for testing and inspection. Ultrasonic testing of the nodes for the main girders is shown in progress in the production stage. It was appreciated that some defects were inevitable in cast and welded components and that non-destructive tests could not be relied upon to reveal every shortcoming.

The fracture toughness tests showed the cast steel exhibited the required standards of quality. After preliminary welding tests had indicated the correct welding procedures, ultimate tests to collapse proved that it was possible to attain satisfactory and, in some cases, excellent results.

Another example is the design against brittle fracture of a main coolant pump in a nuclear reactor primary circuit. The housing was constructed from three C-1/2%Mo steel castings which were welded together to form a casing having a wall thickness from 4 in. to 22 in. and weighing about 32 tons, after internal surfaces had been clad with stainless steel.

Precipitation hardening, or age hardening, provides one of the most widely used mechanisms for the strengthening of metal alloys. The fundamental understanding and basis for this technique was established in early work at the U. S. Bureau of Standards on Duralumin.

The importance of theoretical suggestion for the development of new alloys is clear from the historical record. At the end of the 19th century, cast iron was the only important commercial alloy not already known to western technology at the time of the Romans. When age hardening of aluminum was discovered accidentally by Wilm, during the years 1903 -1911, it quickly became an important commercial alloy under the trade name Duralumin.

The strength and hardness of some metal alloys may be enhanced by the formation of extremely small uniformly dispersed second-phase particles within the original phase matrix in a process known as precipitation or age hardening. The precipitate particles act as obstacles to dislocation movement and thereby strengthen the heat-treated alloys. Many aluminum based alloys, copper-tin, certain steels, nickel based super-alloys and titanium alloys can be strengthened by age hardening processes.

In order for an alloy system to be able to be precipitation-strengthened, there must be a terminal solid solution that has a decreasing solid solubility as the temperature decreases. The Al-Cu (Duralumin is an aluminum alloy of 2XXX group) phase diagram shown in Figure 1 shows this type of decrease along the solvus between the α and α+θ regions. Consider a 96wt%Al – 4wt%Cu alloy which is chosen since there is a large degrease in the solid solubility of solid solution α in decreasing the temperature from 550°C to 75°C.

Figure 1: The aluminum rich end of the Al-Cu phase diagram showing the three steps in the age-hardening heat treatment and the microstructures that are produced.

In an attempt to understand the dramatic strengthening of this alloy, Paul D. Merica and his coworkers studied both the effect of various heat treatments on the hardness of the alloy and the influence of chemical composition on the hardness. Among the most significant of their findings was the observation that the solubility of CuAl2 in aluminum increased with increasing temperature.

Although the specific phases responsible for the hardening turned out to be too small to be observed directly, optical examination of the microstructures provided an identification of several of the other phases that were present. The authors proceeded to develop an insightful explanation for the hardening behavior of Duralumin which rapidly became the model on which innumerable modern high-strength alloys have been developed.

They summarized the four principal features of the original Duralumin theory:

- age-hardening is possible because of the solubility-temperature relation of the hardening constituent in aluminum,
- the hardening constituent is CuAl2,
- hardening is caused by precipitation of the constituent in some form other than that of atomic dispersion, and probably in fine molecular, colloidal or crystalline form, and
- the hardening effect of CuAl2 in aluminum was deemed to be related to its particle size.

The precipitation-hardening process involves three basic steps:

1) Solution Treatment, or Solutionizing, is the first step in the precipitation-hardening process where the alloy is heated above the solvus temperature and soaked there until a homogeneous solid solution (α) is produced. The θ precipitates are dissolved in this step and any segregation present in the original alloy is reduced.

2) Quenching is the second step where the solid α is rapidly cooled forming a supersaturated solid solution of αSS which contains excess copper and is not an equilibrium structure. The atoms do not have time to diffuse to potential nucleation sites and thus θ precipitates do not form.

3) Aging is the third step where the supersaturated α, αSS, is heated below the solvus temperature to produce a finely dispersed precipitate. Atoms diffuse only short distances at this aging temperature. Because the supersaturated α is not stable, the extra copper atoms diffuse to numerous nucleation sites and precipitates grow. The formation of a finely dispersed precipitate in the alloy is the objective of the precipitation-hardening process. The fine precipitates in the alloy impede dislocation movement by forcing the dislocations to either cut through the precipitated particles or go around them. By restricting dislocation movement during deformation, the alloy is strengthened.

Age Hardening – Precipitation. The strongest aluminum alloys (2xxx, 6xxx and 7xxx) are produced by age hardening. A fine dispersion of precipitates can be formed by appropriate heat treatment.

A general model for decomposition is given, followed by details of the precipitation sequences in 4 specific alloy systems: Al-Cu, Al-Cu-Mg, Al-Mg-Si and Al-Zn-Mg. The Al-Cu system is used as the main example of decomposition, i.e.

a0 (SSSS) → GP zones → θ” → →θ’ → θ or, more fully:

a0 (SSSS) → α1 + GP zones → α2 + θ” → α3 + θ’ → α4 + θ

Age Hardening – Strengthening. The 3 main mechanisms are:

- Coherency strain hardening;
- Chemical hardening;
- Dispersion hardening

Coherency strain hardening results from the interaction between dislocations and the strain fields surrounding GP zones and/or coherent precipitates. Chemical hardening results from the increase in applied stress required for a dislocation to cut through a coherent (or semi-coherent) precipitate. This in turn depends on a number of factors, including:

the extra interfacial area – and hence energy – between precipitate and matrix;
the possible creation of an anti-phase boundary (APB) within an ordered precipitate and
the change in separation distance between dissociated dislocations due to different stacking fault energies of matrix and precipitate.

Dispersion hardening occurs in alloys containing incoherent precipitates or particles – i.e. typically those that have been overaged. This hardening results from the increased shear stress required for dislocations to by-pass these obstacles.

As mentioned above, the precipitation reactions in Al-Cu are quite complex. The equilibrium phase CuAl2 is difficult to nucleate so its formation is preceded by a series of metastable precipitates. Guinier and Preston first discovered many of the age hardening phenomena. The first two precipitates to form in the sequence are, therefore, known as GP zones. GP1 consists of 10 nm diameter copper-rich plates on {100}Al planes. These develop into GP2 zones which are also coherent plates 10 nm thick and 150 nm diameter. These lead to maximum hardening. Theta’ /θ’/ precipitates then replace the GP zones as semi-coherent particles, a stage known as over-aging because the hardness begins to decrease. The equilibrium phase CuAl2 has a tetragonal crystal structure and contributes little to hardness.

In the field of 6000 series precipitation hardening aluminum alloys, for instance, process models have been able to describe the effect of quench-induced precipitation on structural defects on the hardening potential during isothermal low-temperature aging.

The fracture toughness of 7000 series alloys has been related to some elements of the microstructure resulting from the thermo-mechanical treatment in phenomenological models. The general strategy of process modeling is to use individual equations which have been developed for well defined experiments and try to integrate them in an integrated manner for the more complex practical situations where coupled effects operate.

However, a good description is still lacking when several of these phenomena are simultaneously operative. The understanding of competitive precipitation of several phases (metastable and stable) on several nucleation sites (e.g. homogeneous and on structural defects) is very limited, as well as the understanding of the shearing/by-passing transition leading to the maximum strength for precipitation hardening materials. The strain hardening behavior of materials containing precipitates (and thus necessarily a solid solution) is poorly understood, and predicting the fracture toughness in cases where several fracture modes are simultaneously operating is not possible in the present state of the art.

The discovery of nanocrystalline Fe-based soft magnetic materials is less than ten years old. The first class of such materials was the melt-spun Fe-Si-B alloys containing small amounts of Nb and Cu. The Fe-Si-B-Nb-Cu amorphous phase transforms to a body-centered cubic (bcc) Fe-Si solid solution with grain sizes of about 10 nm during annealing at temperatures above the crystallization temperature.

The presence of small amounts of Cu helps increase the nucleation rate of the bcc phase while Nb retards the grain growth. These “Finemet” alloys provide low core losses (even lower than amorphous soft magnetic alloys such as Co-Fe-Si-B), exhibit saturation induction of about 1.2 T, and exhibit very good properties at high frequencies, comparable to the best Co-based amorphous alloys.

These were first developed in Japan and have stimulated a large amount of research and development worldwide to optimize the magnetic properties. There has been relatively little work in the United States in this area, however.

Amorphous and nanocrystalline materials were recently investigated for applications in magnetic devices such as transformers, inductive devices, etc, which use soft-magnetic materials. The interest in developing nanocrystalline soft-magnetic alloys has dramatically increased during the past few years. Bulk soft-magnetic materials need to have both high induction and permeability, as well as many other non-magnetic features such as mechanical properties, corrosion resistance, etc.

Nanocrystalline soft magnetic alloys can be described in general as TL(1-x)(TE, M, NM)x, where TL denotes a late (ferromagnetic) transition metal element, TE denote an early transition metal element , M is a metalloid, and NM is a noble metal. This composition usually has x < 0.20 i.e. with as much late ferromagnetic transition metals (TL = Co, Ni or Fe) as possible. The early transition metals (TE = Zr, Nb, Hf, Ta, etc.) and metalloids (M = B, P, Si, etc.) are added to promote glass formation in the precursors. The noble metal elements (NM = Cu, Ag, Au, etc.) serve as nucleating agents for the ferromagnetic nanocrystalline phase.

Nanocrystalline alloys incorporating α-Fe nanocrystalline phases are being explored as alternative soft magnetic materials with improved properties. Excellent soft magnetic properties result due to a fine distribution of crystalline grains within an amorphous matrix.

NANOPERM-type alloys are based on the Fe-TM-B system; they contain larger concentrations of Fe (83-89 %) and have higher values of saturation induction (~ 1.6 – 1.7 Tesla). The NANOPERM nanocrystalline alloys have very low energy losses at power frequencies (60 Hz), making them potentially interesting for electrical power distribution transformers.

Over the last decade high induction bulk materials, including NANOPERM-type alloys, were investigated, however less attention has been paid to the stability of nanostructured materials under extreme conditions of pressure and temperature. It is known that high pressure has a great impact upon phase transition, at high values for applied pressure metastable structures can be obtained, which can lead to new possible application for such materials.

Table 1: Properties of NANOPERM material

Figure 1: Comparison NANOPERM-Ferrite

Striations are thought to be steps in crack propagation, were the distance depends on the stress range. Beachmarks on the other hand may contain thousands of striations.

One can determine that a material failed by fatigue by examining the fracture sight. A fatigue fracture will have two distinct regions; One being smooth or burnished as a result of the rubbing of the bottom and top of the crack (steps 1 & 2); The second is granular, due to the rapid failure of the material.

Other features of a fatigue fracture are Beachmarks and Striations. Beachmarks, or clamshell marks, may be seen in fatigue failures of materials that are used for over a period of time, allowed to rest for an equivalent time period and then loaded again as in factory usage. Striations are thought to be steps in crack propagation, were the distance depends on the stress range. Beachmarks on the other hand may contain thousands of striations.

A very useful way to visualize time to failure for a specific material is with the S-N curve. The “S-N” means stress verse cycles to failure, which when plotted uses the stress amplitude, σa plotted on the vertical axis and the logarithm of the number of cycles to failure. An important characteristic of this model as seen in Figure 3 is the fatigue limit.

Figure 3: A S-N Plot for an aluminum alloy

The significance of the fatigue limit is that if the material is loaded below this stress, then it will not fail, regardless of the number of times it is loaded. Material such as aluminum, copper and magnesium do not show a fatigue limit, therefore they will fail at any stress and at any number of cycles. Other important terms are fatigue strength and fatigue life. The stress at which failure occurs for a given number of cycles is the fatigue strength. The number of cycles required for a material to fail at a certain stress in fatigue life.

The rate of fatigue crack propagation is determined by subjecting fatigue-cracked specimens, like the compact specimen used in fracture toughness testing, to a constant-amplitude and cyclic loading. The incremental increase in crack length is recorded along with the corresponding number of elapsed load cycles to acquire the stress intensity (K), crack length (a), and cycle count (N) data during the test. The data is presented in an “a versus N” curve as shown in the image to the right. Various a versus N curves can be generated by varying the magnitude of the cyclic loading and/or the size of the initial crack.

Dependable design against fatigue-failure requires thorough education and supervised experience in structural engineering, mechanical engineering, or materials science. There are three principal approaches to life assurance for mechanical parts that display increasing degrees of sophistication:

- Design to keep stress below threshold of fatigue limit (infinite lifetime concept);
- Design (conservatively) for a fixed life after which the user is instructed to replace the part with a new one (a so-called lifed part, finite lifetime concept, or “safe-life” design practice);
- Instruct the user to inspect the part periodically for cracks and to replace the part once a crack exceeds a critical length. This approach usually uses the technologies of nondestructive testing and requires an accurate prediction of the rate of crack-growth between inspections. This is often referred to as damage tolerant design or “retirement-for-cause”.

Fatigue cracks that have begun to propagate can sometimes be stopped by drilling holes, called drill stops, in the path of the fatigue crack. This is not recommended as a general practice because the hole represents a stress concentration factor which depends on the size of the hole and geometry. There is thus the possibility of a new crack starting in the side of the hole. It is therefore always far better to replace the cracked part entirely.

Hydrogen has been and always will be a source of various problems within steel production because of its generally detrimental effects on processing characteristics and service performance of steel products. If the hydrogen content of the molten steel exceeds the solubility limit of hydrogen in solid iron, the hydrogen will be rejected during solidification, and this leads to pinhole formation and porosity in steel. Just a few parts per million of hydrogen dissolved in steel can cause hairline cracks (flakes), hydrogen embrittlement, hydrogen blistering and loss of tensile ductility, particularly in large steel castings ingots, blooms and slabs.

Thermodynamic considerations

According to Sievert’s law, a di-atomic gas reacts with a metal and is dissolved into atomic form. In the case of hydrogen solubility, data are summarized by the following equations and are shown graphically in Fig 1.

The equilibrium constant of reaction (1) is:

Figure 1: Solubility of Hydrogen in pure iron or low-alloy steel at 1 atm pressure of H2

The temperature dependence of K in iron in equilibrium with pH2 = 1 atm is as follows for α, δ(bcc) iron, γ(fcc) iron and liquid iron (I) are given by the following equations:

where the temperature T is in degrees Kelvin.

During rapid cooling of a heavy-section steel casting, e.g. thick slab or bloom, there will be little diffusion of H out of casting. Because of the hydrogen solubility decreases with a decreasing temperature, there will be a build up of H2 pressure in the steel matrix during rapid cooling. For the limiting case of no hydrogen diffusion, the H2 pressures will be as shown below in the steel containing 2, 4 and 8 ppm H.

Table 1: Build up of H2 pressure in the steel matrix during rapid cooling

Hydrogen pickup

Numerous sources of hydrogen exist during the melting, the ladle processing and the casting operations. Hydrogen pickup in the steel is primarily due to the water associated with the slagmaking materials and as an impurity in the alloy additions and the carburizers. Hydrogen is generally not a problem in the BOF steelmaking except in the bottom blown converters (Q-BOP) where natural gas (CH4), used as a tuyere coolant, is the major source of hydrogen.

The lime, e.g. calcium oxide (CaO), is an important addition during steelmaking operations, especially during ladle metallurgy to adjust slag chemistry, to facilitate inclusion removal and for desulphurization of steels. Due to the moist atmospheric conditions, lime can become hydrated to form calcium hydroxide. This hydrated lime when added to the liquid melt decomposes according to the reaction

The water vapor formed dissociates on the liquid steel surface causing hydrogen pickup by the following reaction

Figure 2, 3 and 4 show results of experiments performed in laboratory and on the industrial scale to investigate the effect of both calcium hydroxide and coke addition on the hydrogen content in carbon steels.

All figures show the variation of hydrogen pickup as a function of additional Ca(OH)2 and coke. In Figure 2 showed results where is the Ca(OH)2 was added on top of the liquid metal while in Figure 3 showed results of case that the calcium hydroxide was added along with the CaO-Al2O3 slag. Figure 4 shows the effect of metallurgical coke addition on the hydrogen content of low carbon steel.

Figure 2: Hydrogen pickup due to Ca(OH)2 addition on the top of the slag

Figure 3: Hydrogen pickup due to Ca(OH)2 addition along with the CaO-Al2O3 slag

Figure 4: Change in the hydrogen content of the liquid metal as a function of coke addition for 90 tone heat

As a follow up of these trials it can be seen that the hydrogen pickup increases in steel with both addition of lime e.g. Ca(OH)2 and coke. The capacity of the melt to absorb the hydrogen decreases as the hydrogen content in the melt approaches the equilibrium value. The capacity to absorb hydrogen increases as the steel is deoxidized.

The control of hydrogen content in steels is an important task of steelmakers during BOF and EAF steelmaking. The control of the pickup of hydrogen must include the use of inputs with low moisture or hydrogen content, avoidance of late addition of lime during smelting process, minimization of carry over slag, efficient degassing at deep vacuum and intense purging, recarburization with low hydrogen pet coke, control of ladle slag basicity etc.

The fatigue life of a component can be expressed as the number of loading cycles required to initiate a fatigue crack and to propagate the crack to its critical size.

Fatigue cracking is one of the primary damage mechanisms of structural components. Fatigue cracking results from cyclic stresses that are below the ultimate tensile stress, or even the yield stress of the material. The name “fatigue” is based on the concept that a material becomes “tired” and fails at a stress level below the nominal strength of the material. The fact that the original bulk design strengths are not exceeded and the only warning sign of an impending fracture is often just a tiny crack, makes fatigue damage especially dangerous.

In materials science, fatigue is the progressive and localized structural damage that occurs when a material is subjected to cyclic loading. The maximum stress values are less than the ultimate tensile stress limit, and may be below the yield stress limit of the material.

Fatigue occurs when a material is subject to alternating stresses, over a long period of time. Examples of where fatigue may occur are: springs, turbine blades, airplane wings, bridges and bones.

There are three common ways in which stresses may be applied: axial, torsional, and flexural. Examples of these are seen in Figure 1. There are also three stress cycles by which loads may be applied to the sample, the simplest being the reversed stress cycle. This is merely a sine wave where the maximum stress and minimum stress differ by a negative sign. An example of this type of stress cycle would be in an axle, where every half turn or half period as in the case of the sine wave, the stress on a point would be reversed.

The most common type of cycle found in engineering applications is where the maximum stress (σmax) and minimum stress (σmin) are asymmetric (the curve is a sine wave) not equal and opposite. This type of stress cycle is called the repeated stress cycle. The final type of cycle mode is where stress and frequency vary randomly. An example of this would be automobile shocks, where the frequency and magnitude of imperfections in the road will produce varying minimum and maximum stresses.

For some components the crack propagation life is neglected in design because stress levels are high, and/or the critical flaw size small. For other components the crack growth life might be a substantial portion of the total life of the assembly. Moreover, preexisting flaws or sharp design features may significantly reduce or nearly eliminate the crack initiation portion of the fatigue life of a component. The useful life of these components therefore may be governed by the rate of subcritical crack propagation.

Figure 1: Visual examples of axial stress, torsional stress, and flexural stress.

Aircraft fuselage structure is a good example of a structure that is based largely on a slow crack growth rate design. Many years ago, the USAF reviewed a great number of malfunction reports from a variety of aircraft. The reports showed that the preponderance of structural failures occurred from 1) built-in preload stresses, 2) material flaws and 3) flaw caused by in-service usage. These facts led to a design approach that required the damage tolerance analysis to assume a material flaw exists in the worst orientation and at the most undesirable location. The analysis helps to ensure that structures are designed that will support slow stable crack growth until the crack reaches a length where it can be reliably detected using NDT methods.

The fatigue life of a component can be expressed as the number of loading cycles required to initiate a fatigue crack and to propagate the crack to its critical size. Therefore, it can be said that fatigue failure occurs in three stages – crack initiation; slow, stable crack growth; and rapid fracture.

As discussed previously, dislocations play a major role in the fatigue crack initiation phase. In the first stage, dislocations accumulate near surface stress concentrations and form structures called persistent slip bands (PSB) after a large number of loading cycles. PSBs are areas that rise above (extrusion) or fall below (intrusion) the surface of the component due to movement of material along slip planes. This leaves tiny steps in the surface that serve as stress risers where tiny cracks can initiate. These tiny cracks (called microcracks) nucleate along planes of high shear stress which is often 45° to the loading direction.

In the second stage of fatigue, some of the tiny microcracks join together and begin to propagate through the material in a direction that is perpendicular to the maximum tensile stress. Eventually, the growth of one or a few of the larger cracks will dominate over the rest of the cracks. With continued cyclic loading, the growth of the dominate crack or cracks will continue until the remaining uncracked section of the component can no longer support the load. At this point, the fracture toughness is exceeded and the remaining cross-section of the material experiences rapid fracture. This rapid overload fracture is the third stage of fatigue failure.

Failure of a material due to fatigue may be viewed on a microscopic level in three steps: Crack Initiation, Crack Initiation and Failure, as shown in Figure 2 A, B.

Figure 2: A diagram showing location of the three steps in a fatigue fracture under axial stress.

The figure above illustrates the various ways in which cracks are initiated and the stages that occur after they start. This is extremely important since these cracks will ultimately lead to failure of the material if not detected and recognized. The material shown is pulled in tension with a cyclic stress in the y, or horizontal, direction. Cracks can be initiated by several different causes. The three that will be discussed here are nucleating slip planes, notches and internal flaws.

Medium and high carbon steels are widely used in many common applications. Increasing carbon as the primary alloy for the higher strength and hardness of steels is usually the most economical approach to improved performance. However, some of the effects of elevated carbon levels include reduced weldability, ductility and impact toughness. When these reduced properties can be tolerated, the increased strength and hardness of the higher carbon materials can be used to a significant advantage. Common applications of higher carbon steels include forging grades, rail steels, spring steels (both flat rolled and round), pre-stressed concrete, wire rope, tire reinforcement, wear resistant steels (plates and forgings), and high strength bars.

To increase the performance of steels in these applications, it is common to maximize strength and hardness by raising the carbon level to the highest practical level. The limiting factor to carbon additions will vary depending on the type of applications. For forging steels and bar products, it may be toughness or weldability. For high strength wire, the limiting factor for carbon addition is generally the eutectoid carbon level, above which the presence of grain boundary carbides will dramatically reduce drawability.

Generally, the high carbon steels contain from 0.60 to 1.00% C with manganese contents ranging from 0.30 to 0.90%. High carbon steels are used for spring materials and high-strength wires. Ultrahigh carbon steels are experimental alloys containing approximately 1.25 to 2.0% C. These steels are thermomechanically processed to produce microstructures that consist of ultrafine, equiaxed grains of ferrite and a uniform distribution of fine, spherical, discontinuous proeutectoid carbide particles. Such microstructures in these steels have led to superplastic behavior.

Figure 1: Historical Description of the Fe-C phase diagram

Figure 2: Microstructure of high carbon steel

Figure 2 shows the microstructure of high carbon steel with about 0.8% C by weight, alloyed with iron. The steel has one major constituent, which is pearlite. It is made up from a fine mixture of ferrite and iron carbide, which can be seen as a “wormy” texture. The pearlite has a very fine structure, which makes the steel very hard. Unfortunately this also makes the steel quite brittle and much less ductile than mild steel. The high carbon steel has good wear resistance, and until recently was used for railways. It is also used for cutting tools, such as chisels and high strength wires. These applications require a much finer microstructure, which improves the toughness.

The common standards which define high carbon spring steel wire are:

ASTM A-227 Class I/II
ASTM A-417 Class I/II
ASTM A-407 Type A/B/C/D/E/F/G/H
ASTM A-228 Music Wire

Application

High carbon steel is used for applications in which high strength, hardness and wear resistance are necessary, such as wear parts, knives, saw blades, springs, gear wheels, chains, brackets etc. cold chisels, wrenches, Jaws for vices, pneumatic drill bits, wheels for railways service, wire for structural work, shear blades, hacksaws.

Strain ageing has been found to cause a detrimental effect in low carbon structural steels. A lot of studies have been made on the effect of different parameters on strain ageing characteristics of theses steels.

In the work of S. Gündüz, the ageing behaviour of a carbon steel with 40% martensite volume fractions and a microalloyed steel with 20% martensite volume fractions were studied. The variation of mechanical properties, especially the increase in YS was measured by tensile tests. The specimens were pre-strained in tension by 2, 4 and 6%, aged at 25, 100, 150, 200 and 250°C for 30 min followed by restraining.

The steels used in this investigation are commercially produced a carbon steel (without alloying elements) and a microalloyed steel with a chemical composition shown in Table 1. All the specimens were first subjected to an annealing treatment at 900 °C for 30 min followed by air cooling to homogenize the micro-structure.

Samples used for the annealing treatment are of dimensions approximately 170mm × 30mm × 4.5 mm. A Carbolit furnace capable of operating up to 1200°C was used. The temperature in the heat treatment furnace was measured using a K-type thermocouple and temperature variation during heat treatment did not exceed ±3°C.

Table 1: Chemical composition of the investigated steels

As mentioned above, the specimens were pre-strained in tension by 2, 4 or 6%. After this, they were unloaded and aged at 25, 100, 150, 200 and 250°C for 30 min. After ageing of the specimens, they were subjected to a tensile test at ambient temperature at a crosshead speed of 2 mm/min. At least three specimens were tensile tested for each ageing temperature and average values were calculated.

The increase in flow stress as a result of restraining was taken as the strain ageing, ΔY2, which is illustrated in Figure 1. For samples, pre-strained in tension, ΔY2 was determined with single specimen by the difference between lower yield stress after ageing and the flow stress at the end of the pre-straining.

Figure 1: Stress–strain curve for low carbon steel strained to point A, unloaded, and then restrained immediately (curve a) and after ageing (curve b).

Figures 2 and 3 show the stress–strain diagrams of the dual phase carbon steel and the microalloyed steel pre-strained in tension by 2, 4 or 6%, aged at different temperatures, and restrained.

As shown, the dual phase carbon steel and the microalloyed steel, prior to any ageing, exhibits continuous yielding which has been commonly attributed to mobile dislocations introduced during cooling from the intercritical annealing temperature. Many dislocation sources come into action at low strain and plastic flow begins simultaneously through the specimen, thereby suppressing discontinuous yielding.

Figure 2: Variation of stress–strain curves of the dual phase carbon steel at different ageing temperatures for the pre-strains of 2% (a), 4% (b) and 6% (c).

Figure 3: Variation of stress–strain curves of the dual phase microalloyed steels at different ageing temperatures for the pre-strains of 2% (a), 4% (b) and 6% (c).

The main conclusions from this study are as follows:
1. Both steels displayed significant changes in appearance as the ageing temperature was increased for the pre-strain in the range studied. This indicated that static strain ageing takes place in both dual phase carbon steel and microalloyed steel.
2. In contrast to the negative effect of pre-strains on the change in ΔY2 produced by subsequent ageing, it was found that increasing pre-strain markedly increased the change of UTS of both dual phase carbon steel and microalloyed steel. This indicated that ΔY2 value is insensitive to dislocation density and is principally dependent on the solute segregation per dislocation.
3. Both dual phase carbon steel and microalloyed steel showed significant increases in YS, UTS and ΔY2, however the percentage elongation to fracture decreased as the ageing temperature was raised from 25 to 100°C for the pre-strain in the range studied. This is due to atmosphere formation at dislocation and precipitation of carbonitride on dislocations during strain ageing.
4. Further increase in ageing temperature to 150, 200 and 250°C caused a reduction in YS, but an increase in percentage elongation. These are signs of overageing probably due to tempering effect of martensite and coarsening of the precipitates on the dislocations.
5. The ageing in the dual phase microalloyed steel occurred more slowly than the dual phase carbon steel. This was associated with the chemical composition of dual phase microalloyed steel which, in addition to carbon atoms, contained nitrogen and carbide forming elements such as titanium, vanadium and aluminum.

Failure of engineering components due to fatigue is a common occurrence in the aerospace and automotive industries. It is necessary to characterize the fatigue behavior of materials so that the engineering components can be operated safely and reliably.

The fatigue behavior of a material is usually determined by conducting axial (tension and compression) and sometimes torsional fatigue experiments at the service temperature. In reality, however, engineering components are subjected to loads in multiple directions, and these loads produce complex states of stress and strain within the component.

L. P. Borrego et al. investigated the influence of stress ratio and thickness on the fatigue crack growth rate. The research was conducted using AlCuMgSi (2017) aluminum alloy with a T4 heat treatment as well as AlMgSi1 (6082) aluminum alloys with a T6 heat treatment. The T4 and the T6 treatments are full heat treatment processes comprising of the operations of solution treatment, quenching and age-hardening. The age-hardening of the T4 heat treatment is performed at room temperature (naturally ageing) whilst in T6, this is performed by artificially ageing (160°C over 10 hours for 6082 aluminum alloy).

Fatigue tests were undertaken, in agreement with the ASTM E647 standard, using middle-tension, M (T), specimens.

The influence of stress ratio and thickness on the fatigue crack growth rate for aluminum alloy 2017-T4 can be seen in Figure 1. The data was obtained using a specimen thickness of 1.8 and 10 mm tested at stress ratios of R=0 and R=0.4. The fatigued fracture surface of 10 mm thickness specimen tested under R=0 is superimposed in the figure for comparison.

A strong R-ratio effect on the fatigue crack growth rate was observed for 1.8 mm thickness specimens. The same behavior was also observed for 10 mm thickness specimens, but only for ΔK values lower than 9 MPa m1/2. The crack growth rate da/dN generally increases with the stress ratio R.

Figure 1: Effect of stress ratio and specimen thickness on fatigue crack growth rate for alloy 2017-T4

Figure 2 shows that specimen thickness has no significant influence in the crack growth behavior of this alloy except for R=0 and for ΔK values above 9 MPa m1/2.

In the work done by M.B.Grieb et al., the monotonic tensile tests of the aluminum cast alloys were carried out at room temperature after the standardized specimens had been pre-aged at defined temperatures for 500 hours in order to determine the effect of aging on the tensile strength properties of the aluminum alloys examined.

500 hours were used to establish a completely over-aged material condition at least at the highest temperatures applied. The presented values of the ultimate tensile strength in Figure 2 are the mean values from five performed tensile tests.

Between the temperatures of 150°C and 250°C the material strength decreases rapidly for almost all of the cast aluminium alloys as expected. Only in the case of AlMg3Si1(Sc,Zr)-T5, the annealing is very much delayed, even at the highest temperature of 350°C.

AlSi7Mg-T7 shows the highest reduction of the tensile strength with temperature. After 500 hours at an aging temperature of 200°C or higher, the tensile strength has reached a minimum, and there with a stable condition. At temperatures of 150°C and higher, more and more incoherent Mg2Si particles are formed from coarsening, explaining the strength decreases of the AlSiMg alloys.

At 200°C and higher incoherent Al2Cu appears and reduces the tensile strength in the case of AlSiCu. Al3Sc is responsible for precipitation-strengthening in AlMg3Si(Sc,Zr)-T5 and is stable up to 300°C. Hence, it can be summarized that strength reduction due to aging is slow for this alloy.

Figure 2: Ultimate tensile strength of aluminum castings at room temperature as a function of preaging temperature after 500h aging

Various hardenable cast aluminium alloys (AlSi7Mg-T6, AlSi5Cu3-T7, AlSi5Cu1-T7, AlMg3Si1-T6, AlMg3Si1(Cu)-T6 and AlMg3Si1(Sc,Zr)-T5) have been studied and compared regarding their applicability as diesel engine cylinder heads.

The focus of the study is put on the resistance against thermomechanical fatigue (TMF) loading, which is typical of the cylinder head application, and the development of an accurate TMF life prediction model, which takes the change of the mechanical properties during TMF into account.

A near-component-shaped specimen, the so called “valve bridge sample” was developed and TMF tested in a special testing system which closely matches the real component loading situation. Furthermore the thermal and mechanical properties of the alloys were determined as a function of time, temperature and number of TMF loading cycles.

The mechanical properties obtained in tensile tests on standardized specimens after 500 hours of annealing at various aging temperatures indicate that the influence of the aging temperature is high for AlSi7Mg-T6 and low for AlMg3Si1(Sc,Zr)-T5 compared to the other aluminum alloys studied.

Furthermore, the TMF tests on the valve bridge samples showed that the AlSiCu alloys possess a higher resistance against TMF crack initiation in terms of the number of TMF cycles necessary to form a detectable fatigue crack as compared to AlMgSi and the AlSiMg alloys. The aging of the microstructure, due to the combined thermal exposure and cyclic plastic deformation occurring in the TMF tests, was proven macroscopically by 2-dimensional hardness measurements on the surface which are heated in the TMF loading system, and microstructurally by TEM examinations of samples taken from the critical location of the valve bridge samples.

It could be shown that a simplified aging model, which reproduces the decrease of the mechanical strength during TMF loading, can improve the thermomechanical fatigue life time prediction. For this purpose, two life time prediction models (IWM Freiburg and a simplified Chaboche model) were used to assess the TMF life time until crack initiation of the valve bridge samples.

Several numerical simulations with different assumptions for the mechanical properties, i.e. constant values related to the starting condition, constant values from the final condition and changing values according to the aging model, were run for TMF loading of near-component-shaped specimens made of the aluminum alloys AlSi7Mg and AlSi5Cu3.

The resulting life times were compared with the experimental results of the TMF tests. The comparison showed that the incorporation of the aging model improves the predictive capability and accuracy of both damage evolution models. This holds true in particular for AlSi7Mg, since the mechanical properties of this alloy change rapidly within the temperature range of the TMF loading.