Chapter 7: Fracture of Metals

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

If you've ever watched a piece of critical machinery fail, it is usually less a slow surrender and more a sudden, dramatic catastrophe.

Exactly.

A bridge section snaps, a pressure vessel ruptures, or an aircraft component breaks without any of the expected sagging or stretching.

And these are the catastrophic consequences of, well,

of inadequate fracture mechanics.

That abrupt moment of separation is what mechanical metallurgy defines as fracture.

It's the fragmentation of a solid body into two or more parts under the application of stress.

It's the end of the line.

It really is.

Fracture represents the ultimate, absolute limit for nearly every structural component designed by engineers.

Okay, let's unpack this and set our mission.

We are diving deep into the mathematical and physical backbone of this field, focusing exclusively on chapter seven of the mechanical metallurgy text.

Our mission is to master the fundamentals of metal fracture, the math, the models, and the key diagrams so you can quantify the process of crack initiation and propagation.

This deep dive is your analytical shortcut.

And we are targeting this chapter specifically because engineering research overwhelmingly focuses on brittle fracture.

Why is that?

Why the focus on brittle?

Because brittle fracture occurs suddenly.

It happens often without any visible warning signs like nicking or elongation.

It tends to initiate around geometric discontinuities like notches.

And that leads directly to catastrophic structural failure.

So it's the one you don't see coming.

Precisely.

Understanding how to model and prevent that rapid unworn failure is just paramount for safe design.

We should be clear upfront about the scope of this chapter then.

We are starting with the foundational models for fracture under a single static application of uniaxial tensile stress.

We are purposefully leaving the highly complex time -dependent failures, the fatigue, the creep, the embrittlement for later chapters.

We have to.

We're building the pure single load model first.

You have to walk before you can run.

Right.

Let's start walking.

Where do we begin?

We have to begin by establishing the core behavioral difference in metals.

We do that by defining the two general categories of failure,

ductile fracture versus brittle fracture.

The distinction here is fundamentally about energy absorption and plasticity, right?

Exactly.

Ductile fracture is characterized by appreciable plastic deformation that happens both before and importantly during the crack propagation stage.

So the material visibly gives, it stretches, it fights the failure.

It does.

And when you look at the fracture surface, it reflects this struggle.

It looks rough, fibrous, and gray due to that massive plastic deformation.

Crucially, in ductile failure, you do not observe rapid runaway crack propagation.

It's a slower,

more graceful failure, if you can call it that.

In stark contrast, you have brittle fracture.

Yeah.

And this is characterized by very little, if any, micro deformation.

It occurs suddenly and involves rapid high -velocity crack propagation.

I always think of the difference between tearing a piece of tough leather, which is ductile, and snapping a piece of cold glass.

That's brittle.

That's a perfect analogy.

The surfaces of brittle fractures appear bright, often shiny or granular.

This behavior is similar to cleavage in ionic crystals, a clean separation that requires minimal energy absorption from the structure.

And this is the failure mode we worry about in certain types of metals.

Yes, often seen in body -centered cubic, or BCC,

and hexagonal, close -packed HCP metals,

particularly at lower temperatures.

We can visualize the spectrum of behavior clearly.

If you look at the schematics in Figure 7 -1 describing tensile fracture, you see the journey from extreme brittleness to high ductility.

Let's walk through them.

Stematic A is the ideal brittle fracture.

Straight separation, perpendicular to the stress axis, with virtually no reduction in area or change in shape, just a clean break.

Then you have Schematic B.

This shows the shearing fracture, often seen in highly oriented materials or single crystals.

Here, the failure happens along the preferred slip planes, and the specimen slips until the separation occurs at an angle.

Then we move into the highly ductile regime.

Schematic C represents a completely ductile fracture.

This is typical of ultra -pure metals like gold or lead.

You see extreme necking here.

The specimen draws down to a point or a thin thread before it finally ruptures, and often twists significantly.

But the most common ductile failure in the materials we actually build with, the polycrystalline metals, is represented by Schematic D.

This is the classic cup and cone fracture.

Yes.

The specimen first develops a necked region.

The fracture then initiates in the center of the specimen, where the stress state is most triaxial.

We'll definitely come back to that idea of triaxiality.

Oh, absolutely.

So it propagates outward perpendicular to the load.

The final remaining rim of material separates by shear, forming that characteristic cone on one side and a cup on the other.

And that separation happens at a 45 degree angle, which is the signature of separation governed by maximum shear stress.

Exactly.

It tells you the mechanism right there on the surface.

So to speak the language of the failure analyst, we use standardized terminology summarized in the text.

We classify fracture by its crystallographic mode.

Is it happening by shear or by cleavage?

By its appearance.

Is the surface fibrous and dull or granular and bright?

And by the observable strain to fracture?

Is it truly ductile or dangerously brittle?

These are the key questions.

And you also have to specify the direction of propagation relative to the microstructure.

Right.

When the crack propagates through the body of the individual grains, cutting across the crystal lattice, we call it transgranular fracture.

That's the most common path.

It is.

And it often results in a mixed appearance of fibrous and granular features.

However, if the crack finds an easier, lower energy path by propagating along the boundaries between the grains, we call it intergranular fracture.

And that's a red flag for an investigator.

A huge red flag.

This intergranular failure mode is highly diagnostic because it's often associated with specific types of environmental attack or embrittlement processes where impurities or precipitates have segregated the grain boundaries and weakened them drastically.

Okay, so we've set the stage with the qualitative descriptions.

Now let's transition to the quantitative analysis, starting with the theoretical limit of strength.

Why do metals hold together so well?

It's because of the immensely high cohesive forces between their atoms.

These atomic forces dictate the ultimate strength, which is directly related to properties like a high elastic modulus and a high melting point.

To visualize this ultimate limit, we can look at figure 7 -2.

This plot shows the cohesive force, sigma, against the atomic separation, x.

So on the y -axis, you have the cohesive stress, sigma,

and the x -axis tracks the displacement x, which is measured relative to the atom's equilibrium spacing, which we call a sub -zero.

The shape of this curve is crucial for understanding atomic bonds.

It begins at zero force, when the atoms are perfectly spaced, at x equals zero.

And as you pull them slightly apart, the attractive forces rapidly increase until the curve reaches its maximum point.

And that peak, that sigma max, that is the theoretical cohesive strength.

It is.

Pull the atoms any further, and the force rapidly decays towards zero as the bonds break.

This sigma max is the highest stress a perfectly flawless, pristine material could ever sustain.

And to approximate this value mathematically, we simplify the complex atomic force relationship.

We use a sine wave model.

It's an approximation, but a very powerful one.

This gives us equation 7 -1, the cohesive stress model.

It approximates this relationship as sigma equals sigma max times the sine of 2 pi x over lambda.

Right, so sigma is the stress, sigma max is that theoretical peak strength, and x is the relative displacement.

Lambda is the atomic force wavelength, and we usually take that as twice the equilibrium spacing or two times the sub -zero.

So how do we find the value of that theoretical peak, sigma max?

Well, we rely on two key relationships.

First, for very small displacements, we can approximate the sine function.

And second, we know that for small elastic displacements, Hooke's law has to hold.

Right, stress equals modulus times strain, or equation 7 -3, sigma equals e times x over a sub -zero.

By combining these two approximations, we arrive at the simplified theoretical strength estimate.

That's equation 7 -5.

And it's incredibly simple and powerful.

Sigma max is approximately equal to e divided by pi.

This is a profound theoretical result.

It suggests that a perfect material strength should be about one -third of its stiffness.

It's Young's modulus, e.

This is the absolute physics -driven upper limit of material performance.

But the theory goes deeper by incorporating energy.

Fracture fundamentally creates two new surfaces, and creating surface area costs energy.

That's the key.

The work you do to produce the fracture, let's call it u sub -zero, must be equal to the energy required to create those two new surfaces, which is 2 times gammas.

Where gammas is the surface energy.

Correct.

By integrating that force curve, equation 7 -1, to find the work done, and then setting that work equal to 2 gammas, we can derive a more fundamental expression for the maximum theoretical strength.

And that gives us equation 7 -8.

Sigma max equals the square root of e times gammas divided by a sub -zero.

Let's talk about what that physically means.

It's a powerful equation.

It tells us that the ideal material strength is proportional to the square root of its stiffness, e, and its surface energy, gammas.

And it's inversely related to the initial atomic spacing, a knot.

So strong materials are stiff materials that require a lot of energy to separate their atoms.

Let's run the numbers to see how this plays out.

The text provides a calculation for a silica fiber, which is one of the few materials that actually gets close to this ideal.

Okay, so we're given e as 95 gigapascals, a surface energy gammas of 1 joule per meter squared, and an atomic spacing and knot of 0 .16 nanometers.

You plug those in, and the calculated theoretical strength, sigma max, is approximately 24 .4 gigapascals.

24.

Gigapascals.

That is a truly huge number.

And here is the crucial engineering insight, the point where theory crashes into reality.

But why?

Most engineering steels, the ones we build bridges and cars with, they only achieve fracture strength in the range of 1 to 2 GPA.

So the theoretical maximum strength is anywhere from 10 to 1 ,000 times higher than what we actually see in the structures we build.

And the reason for this massive critical gap is simple.

Flaws.

Flaws.

No material is perfect.

Every metal component contains microscopic imperfections, microcracks, voids, or grain boundary weaknesses that act as stress amplifiers.

And to quantify the impact of these unavoidable flaws, we move to the concept of stress concentration using Orowind's model of an elliptical crack.

You can picture this in figure 7 -3.

The key idea here is that the nominal stress, sigma, that's applied far away from the crack gets magnified significantly at the tiny sharp crack tip.

This local stress, which we can call sigma max again, can be exponentially larger.

The factor of magnification is determined entirely by the geometry of the flaw.

Which brings us to the stress concentration equation, equation 7 -9.

It tells us that the maximum stress at the tip is roughly proportional to the nominal stress, multiplied by the square root of the crack length, 2C, divided by the radius of curvature at the tip, rote.

And the physical meaning here is profound.

The longer the crack, and critically the tighter, the sharper the tip radius rote, the higher the local stress concentration.

If that tip radius rote approaches the atomic spacing a knot, then the stress magnification approaches infinity, theoretically.

So by setting that localized crack tip stress equal to the ultimate theoretical strength we just derived, we can solve for the nominal stress, sigma, required to actually cause the material to fail.

And this is the stress the engineer actually cares about, because it's the stress the component sees.

If we make the worst case assumption that the crack is atomically sharp, meaning the tip radius rote is equal to the atomic spacing a knot, the expression simplifies.

That's equation 7 -11.

It shows that the nominal fracture stress, sigma, is inversely proportional to the square root of the crack length, C.

Let's revisit our numerical example, but this time with a flaw.

We take that theoretical material, which could sustain 24 .4 GPA of stress.

And we introduce a tiny yet finite crack length, C, of 2 .5 micrometers.

That's just .000025 millimeters.

It's invisible to the naked eye.

With that one tiny flaw, the calculated nominal fracture stress, sigma -faff, drops from 24 .4

all the way down to a mere 100 megapascals.

100 MPa.

It's an incredible drop.

That single microscopic flaw reduced the strength of the material by a factor of over 240.

And this is the biggest engineering lesson of this section.

Because real -world components contain flaws, we can never ever rely on the theoretical strength E divided by pi.

We must always calculate the failure stress based on the size and sharpness of the largest possible existing flaw.

And this massive sensitivity is exactly why non -destructive testing, or NDT, like ultrasonic inspection, is absolutely essential in structural engineering.

We are literally looking for those tiny 2 .5 micrometer cracks that can mean the difference between 24 GPA strength and 100 MPa strength.

This realization that flaws govern strength was formalized elegantly by A.

A.

Griffith back in the 1920s.

Yes.

He shifted the focus away from just calculating stress concentrations and introduced the energy criterion for fracture.

He started with perfectly brittle materials, like glass.

Griffith proposed that a crack will only propagate if the total potential energy of the system does not increase.

It's a thermodynamic competition.

Right.

The system is always trying to get to a lower energy state, so it balances two opposing energy components.

The first component is the surface energy, U sub S.

This is the energy required to create the two new crack surfaces.

This is an energy cost, always positive.

It's the energy the material has to absorb to fracture.

The second component is the elastic strain energy, U sub E.

This is energy that is released as the crack spreads and relieves the immense elastic stress that was stored in the material surrounding the crack tip.

So this is an energy return to the system, which means it's always negative.

Correct.

Equation 712 provides the elastic strain energy released, U sub E, per unit thickness.

It's minus pi times C squared, sigma squared, all over E.

And notice the relationship there.

The energy released is proportional to the square of the crack, half length C squared, and the square of the applied stress, sigma squared.

This means if the crack grows, the energy driving it to grow further increases exponentially faster.

It's a runaway process.

Meanwhile, the surface energy required, U sub S, is simpler.

That's equation 713.

U sub S equals 4 times C times gammas.

So this energy cost is linearly related to the crack length C.

You double the crack length, you just double the energy cost.

So the central conflict is between an energy cost that grows linearly with crack length and an energy return that grows quadratically with crack length.

Precisely.

This means that past a certain critical size, the energy released will rapidly overwhelm the energy required, leading to unstable catastrophic propagation.

The critical criterion for failure is when the rate of change of total energy with respect to crack length is zero or negative.

So d delta U over dC is less than or equal to zero.

And setting that derivative to zero yields the critical stress required to make the crack run.

This gives us the famous Griffith equation for plane stress.

Equation 715.

Sigma equals the square root of 2E gammas over pi C.

This equation is a cornerstone of fracture mechanics.

It provides the mathematical proof that the fracture stress is inversely proportional to the square root of the crack length.

Which confirms the result we found with Orwin's stress concentration model, but from a totally different energy -based perspective.

That's right.

If you make the crack four times longer, you only need half the stress to cause failure.

Now I want to take a moment to clarify the difference between plane stress and plane strain.

The Griffith equation we just cited is for plane stress.

Which generally applies to very thin plates.

In thin plates, the material is free to contract in the thickness direction, so the stress component perpendicular to the plate, sigma z, is essentially zero.

But for thick plates, or deeply notched components, the material in the center is highly constrained by the unyielded material surrounding it.

It cannot contract in the thickness direction.

And that condition is called plane strain.

This geometric constraint fundamentally changes the stress balance and requires a modification to the Griffith equation, which you see in equation 716.

It involves Poisson's ratio.

But the overall inverse square root dependency on crack length remains the same.

It does.

Now here is where the story gets complex.

Griffith's theory worked perfectly for glass and ceramics, but when engineers applied it to metals, it grossly underestimated the strength.

So the math was right, but the physical assumption was incomplete.

Exactly.

This is the metallic problem.

Metals are not perfectly brittle, they deform plastically.

And it was Oroen who brilliantly modified Griffith's model.

He suggested that for metals, the energy balance must include the massive amount of plastic work, gamma p, required to extend the crack wall.

This led to the Oroen modification, equation 717, which simply adds that gamma p term to the surface energy term inside the square root.

So it's sigma f equals the square root of two e times the quantity, gamma thess plus gamma p, all over pi c.

This seemingly small addition fundamentally changes the application of fracture mechanics to metals.

The plastic work term, gamma p, is colossal.

It's typically a hundred to a thousand times larger than the thermodynamic surface energy, gammas.

Wait, if the plastic work is hundreds or even a thousand times larger, does that mean that the original surface energy term, gammas, is almost purely academic for structural steels?

Precisely.

For most structural metallic materials, the surface energy term becomes completely negligible.

The resistance to crack propagation is overwhelmingly dominated by the energy required for plastic deformation near the crack tip.

So to design a fracture resistant metal, you need a material that can locally absorb huge amounts of energy through yielding and stretching.

You need a high gamma p.

That's it.

This is why steel is so much tougher than glass, even if they have similar theoretical cohesive strengths.

It's all about the plastic work.

Okay, let's zoom back down to the atomic and microscopic level.

Before we look at polycrystals, how does brittle fracture occur in something like a single crystal?

Brittle failure in a single crystal, when it cleaves, is determined by the critical normal stress, sigma car, acting specifically on the crystallographic cleavage plane.

This is similar to the shear stress analysis for slip, but we're looking at the normal force component instead.

Equation 718 models this critical normal stress.

Sigma c equals the gross applied stress p over a times cosine squared phi.

Where phi is the angle between the tensile axis and the normal to that specific cleavage plane.

So fracture occurs when the component of the applied stress acting purely normal to that plane reaches the critical stress, sigma c.

This is why the orientation of a single crystal relative to the load is so important.

As noted in the data, the critical normal stress for cleavage in a material like iron or zinc can vary dramatically depending on crystal orientation.

Sometimes by a factor of 10 or more.

The crystal is at its weakest when the cleavage plane is oriented perfectly perpendicular to the applied tensile axis.

Now, moving to the larger scale, the metallographic evidence in sections 7 to 6 really challenged Griffith's initial assumptions.

Engineers spent years looking for pre -existing Griffith cracks in unstressed metals.

And they are rarely found reliably.

What the evidence suggests instead is that microcracks are actively produced by plastic deformation.

They often nucleate at points of high stress concentration created by dislocation pileups.

Or very commonly around hard secondary phase particles like carbides or oxides.

Exactly.

These particles are weak points in the matrix.

Figure 7 to 6 is extremely informative in this context.

It plots true strain to fracture, which is our measure of ductility, against the volume of second phase particles present in the material.

And the interpretation is clear.

Ductility plunges dramatically as the volume fraction of these secondary particles increases.

But look closer.

The particle geometry is just as critical as the volume.

Materials containing spherical particles maintain much higher ductility than those with elongated or plate -like particles like sulfides.

Why does the geometry matter so much here?

Because of localized stress concentration,

a sharp elongated particle acts as a much more efficient stress concentrator than a round spherical one.

So the stress field around that sharp particle accelerates the nucleation of voids and microcracks at the particle matrix interface, leading to premature failure and lower overall fracture strain.

Okay, let's talk about postmortem analysis.

When a structure fails, engineers rely heavily on fractography.

The study of the fracture surface, usually using a scanning electron microscope, or FMCM, the fracture surface is literally the historical record of the failure event.

And we can distinguish three main types of fracture surface appearances based on the energy absorption mechanism.

First up is cleavage fracture.

This is characteristic of highly brittle failure, and you can see it in Figure 7 -7.

It shows flat facets that follow the crystallographic planes.

The most distinguishing feature here is something called river marking.

Yes, these are steps or plateaus that trace the exact path and direction of the crack propagation.

These rivers always flow in the direction of the failure.

It's a key diagnostic tool.

Second, we have quasi -cleavage, which is in Figure 7 -8.

This is related to cleavage, but it's less strictly planar.

You see larger cleavage facets interspersed with areas of localized plastic tearing.

These often appear as small dimples around the periphery of the facets.

And you'd see this in certain types of steels.

Yes.

This type of failure often occurs in specific heat -treated steels that are fractured at low temperatures.

And finally, the ultimate signature of ductile failure.

Dimpled rupture.

You can see a great example in Figure 7 -9.

This is the result of microvoid coalescence.

The surface is covered entirely in cup -like depressions, or dimples.

And these are created when microvoids, which nucleated around particles, grow large enough to join together and cause the final failure.

Exactly.

And the shape of the dimple itself, whether it's equiaxed, sheared, or elongated, is highly diagnostic of a localized stress state at the very moment of failure.

Since we've established that microcracks are created during plastic deformation, not just waiting there preformed, we need to transition into the dislocation theories that explain brittle fracture initiation.

Right.

This moves us away from purely elastic models and into plastic instability models.

The process of microcrack formation, particularly in BCC metals, is generally broken into three specific stages based on dislocation movement.

First, plastic deformation occurs, causing a massive pileup of dislocations along their slip planes.

This pileup is then halted by a strong barrier, like a grain boundary or a brittle precipitate.

Second, the enormous buildup of shear stress at the head of this locked pileup reaches a critical level, and that forces the formation of a wedge -shaped microcrack.

And third,

that newly formed microcrack, now subject to the concentrated tensile stress, propagates rapidly via cleavage.

It's driven by the stored elastic strain energy being released, just as the Griffith and Orwin criterion predicted.

The seminal model for that second stage, the nucleation stage, is the Zenner -Strow model.

It's shown schematically in Figure 7 -10.

This depicts a stack of edge dislocations pressed against a grain -bound barrier by an applied shear stress, which we call tau.

The cumulative stress concentration at the tip of that pileup is what creates the microcrack.

And this model allows us to quantify the shear stress required for that to happen.

It does.

Equation 7 -19 defines the necessary shear stress, taus.

The equation is a bit complex, but the physical meaning is crucial.

Okay, what's the takeaway?

The shear stress required to start the crack, taus, is the sum of the frictional stress,

which resists dislocation movement, plus a term that is inversely related to the square root of the slipband length, L.

Okay, let's think about that dependency.

The required stress, taus, increases as the slipband length, L, decreases.

Right, so if the slipband is short, like in a fine -grain material, the pileup is restricted.

You need a higher stress to force the crack to nucleate.

Conversely, a long slipband in a coarse -grain material allows a much larger, more powerful pileup to form at a lower applied stress.

Which means the crack nucleates more easily.

This concept was then refined by Cottrell, who elegantly connected the stress required for nucleation to the energy required for propagation, making it much more physically meaningful.

Cottrell's key insight, that's equation 7 -22, is an energy balance focused on the initiation phase.

It states that the work done by the released shear stress in pushing the opening displacement of the crack tip must equal the work done in creating the two new surfaces.

In equation form, that's tau s minus tau i times n times b equals two gammas, where n is the number of dislocations and b is the burgers vector.

This is powerful because it ties a microscopic parameter dislocation motion directly to a macroscopic criterion, surface energy.

It is, and when you substitute in the expressions for the number of dislocations in a pileup, This theoretical framework leads directly to an equation for fracture stress that looks remarkably familiar.

That's equation 7 -23, which is analogous to the famous Petsch Relution for yield strength.

Sigma f equals sigma i plus kf times d to the minus one half.

The implication here is enormous.

The brittle fracture stress, sigma f, isn't just sensitive to cracks, it's also dependent on grain size, d.

Just as fine grains increase the resistance to yielding, they also increase the resistance to brittle fracture.

Smaller grains restrict the slip band length L, preventing large catastrophic dislocation pileups from forming easily.

Then there is Smith's model, in figure 7 -11.

This addresses brittle fracture that occurs specifically in materials containing brittle grain boundary films.

Like cementite carbides in mild steel, this is a very common failure mechanism in structural steels.

Smith assumes the crack is initiated within the brittle carbide, where the energy required for cleavage is much lower than in the surrounding ferret matrix.

And he derived equation 7 -25 for the stress required to propagate such a microcrack, showing again an inverse relationship to the square root of the grain diameter, d.

The key insight is that propagation is easier in larger grain sizes.

Smith then formulated a complete criterion for fracture, which is equation 7 -26.

It's expressed as an inequality.

Fracture occurs if the effective driving force, the left side of the inequality, is greater than the material's resistance, the right side.

And that resistance includes the plastic work, gamma p, required to propagate the crack into the tough matrix.

Right.

This inequality gives the engineer a detailed microscale tool for predicting when a carbide -initiated crack will run uncontrollably.

We have to circle back to ductile failure, which, while less catastrophic, is the primary goal for structural design because it provides warning.

The classic ductile failure mechanism is the cup and cone fracture, and it's driven by the growth and coalescence of microvoids.

Figure 7 -13 illustrates the five stages of this under uniaxial tension.

Stage B is critical.

Voids nucleate at hard second phase particles, carbides, oxides, inclusions, in the highly stressed center of the specimen just as necking begins.

These microvoids then grow via extensive plastic deformation in stage C.

They stretch, elongate, and eventually coalesce into a central planar crack that is perpendicular to the load axis.

And in stage E, you get the final separation.

The remaining material on the outer circumference shears rapidly under maximum shear stress, creating that characteristic 45 -degree angled cone.

And the resulting fracture surface, remember, is entirely covered in dimples, confirming that microvoid coalescence mechanism.

The McClintock model tries to rigorously quantify the relationship between the stress state and this ductile limit.

Equation 7 -28 is analytically complex.

It involves parameters like the strain hardening exponent, M, the flow stress, sigma, and the specific stresses parallel and perpendicular to the voids.

But the physical takeaway is essential.

The fracture strain, epsilon f, is drastically reduced as the void fraction increases.

And it is most critically influenced by the triaxiality of the stress state.

Hydrostatic tensile stresses promote void growth and dramatically accelerate the reduction of ductility.

So you want a material that avoids high triaxial tension if you want maximum ductility.

Absolutely.

This brings us to perhaps the most important applied engineering concept in the entire chapter, the ductile to brittle transition, or DBT.

This is primarily observed in BCC metals like carbon steel, and it dictates whether a component will yield safely or snap brittily.

The DB transition occurs because plastic yielding, which leads to ductile failure, is governed by shear stress, and shear stress is heavily dependent on temperature.

Whereas cleavage, which is brittle failure, is governed by normal stress, which is relatively temperature insensitive.

Cottrell reformulated the transition condition in Equation 7 -27 as an equality that compares the conditions for yielding versus the conditions for cleavage.

The left side is fundamentally related to the shear stress needed for yielding.

The right side is related to the stress needed for brittle fracture.

So the conceptual meaning is key here.

If the stress required for yielding is less than the stress required for cleavage, the material yields plastically, it's ductile.

But if, due to changing conditions, the stress required for cleavage becomes less than the stress required for yielding, the material cleaves before it ever yields.

And that results in catastrophic brittle failure.

Let's discuss the critical variables that influence this balance.

The frictional resistance to dislocation motion, tau e, increases rapidly as temperature drops.

This raises the required yield stress, sigma e, which in turn promotes brittleness.

And grain size, d, also plays a dual role.

Fine grains raise the yield stress via the petrulation, making it harder to initiate the ductile yield condition.

Which subtly promotes brittleness in the context of the DBT, unless the temperature is extremely low.

However, fine grains significantly raise the brittle fracture strength, sigma, making it a more devirable trade overall.

This balance is beautifully visualized in Figure 7 -17, the transition temperature schematic.

This is one of the most important diagrams for applied engineering design.

It really is.

On the y -axis, we plot strength, and on the x -axis, we plot temperature.

And you observe two distinct curves.

First, the cleavage strength, sigma curve, is relatively flat and constant as temperature changes.

This is the normal stress failure limit.

Second, you have the yield stress, sigma, and curve.

This drops rapidly as temperature increases.

This is because dislocation motion becomes much easier at higher temperatures, reducing that frictional stress host tau.

The ductile to brittle transition temperature, or TT, occurs exactly where these two curves intersect.

It's where sigma feq is sigma e.

Below this point, the material fractures before it can yield, leading to brittle failure.

Above this point, it yields before it fractures, leading to safe, ductile behavior.

This schematic helps explain historic failures, like the catastrophic brittle failure of the Liberty ships in the freezing North Atlantic during World War II.

That's a classic example.

The steel used was perfectly ductile at shipyard temperatures,

but the sub -zero operational temperatures pushed the steel below its transition temperature, making it fail brittlely when subject to high loads.

But look at the critical shift caused by a notch.

A notch significantly raises the local yield stress due to stress concentration and triaxiality.

On the schematic, this means the entire sigmi curve shifts upward dramatically.

And that action raises the transition temperature from the simple tension TT to the much higher notch TT.

This explains why a component that is perfectly safe and ductile when it's smooth can snap brittily the moment a flaw or a sharp corner is introduced, even at room temperature.

The profound influence of geometric features, specifically notches, requires us to spend a bit more time on these complex stress states.

Notches are engineering nightmares.

They really are.

They increase the tendency toward brittle fracture in four primary interrelated ways.

First, the obvious one, stress concentration, leading to extremely high local stresses.

Second and most insidious, the introduced triaxial tensile stress state at the root of the notch.

Third, they introduce high local strain rates.

And fourth, they produce rapid work hardening locally, which increases the yield stress.

All four of these factors push the component toward brittle cleavage.

Let's visualize that critical second factor triaxiality using figure 715.

This shows the stress distribution beneath a notch in a thick plate under plain strain conditions.

In a simple tension test, you have only one major stress component, sigmi.

But under a deep notch, two additional tensile stresses develop dramatically at the notch root.

Sigma x, which is the lateral stress parallel to the notch root, and sigma's the transverse stress into the thickness of the plate.

And this happens because the surrounding, unyielded material constrains the plastic flow in the notch root, literally forcing the development of these transverse tensile stresses.

The presence of this triaxial tensile stress state, three tensile stresses acting simultaneously drastically increases the difficulty of initiating localized plastic flow.

We call this phenomenon constraint.

And this inability to yield locally promotes fracture, because the material reaches its cleavage strength before it reaches its constrained yield strength.

We quantify this constraint using the plastic constraint factor, k -thigma.

This is the ratio of the notched flow stress to the unnotched flow stress.

This factor can reach values above 2 .57.

Meaning the material under the notch requires more than two and a half times the stress to yield than the material in a simple tension test.

It's a huge fact.

And to estimate the magnitude of the strain concentration within this localized plastic zone, we often use Neuber's relation, which is equation 729.

K -sigma times k -epsilon equals k -t squared.

Right.

This approximation relates the plastic stress concentration factor, k -thigma, and the plastic strain concentration factor, p -epsilon, to the square of the purely elastic stress concentration factor, k -t.

It's a useful tool for connecting the geometry of the notch to the actual localized damage.

When analyzing failure under combined stresses, section 712, it's crucial to remember the core difference between failure criteria for ductile versus brittle materials.

Ductile materials, like aluminum, yield based on shear stress criteria.

They typically follow von Mises or maximum shear stress, or Tresca, rules.

Their failure is governed by the shear component.

But brittle materials, like cast iron, follow the maximum normal stress criterion.

Fracture in these materials is governed only by the magnitude of the largest principal stress component, irrespective of the shear components.

So an engineer needs to know not only the material properties, but also the specific failure rule that governs that material when assessing complex stress states.

Finally, let's look at the fascinating effect of applying high hydrostatic pressure, which is section 713.

Since we know that triaxial tensile stress suppresses plastic flow and promotes brittle fracture.

Then applying external compressive hydrostatic pressure should have the exact opposite effect.

It should suppress fracture and dramatically increase ductility.

And it does.

This principle is highly utilized in metalworking processes, like high pressure extrusion and wire drawing, where materials must withstand immense plastic deformation without fracturing.

Figure 720 is a great visualization of this.

It plots true strain to fracture versus the applied hydrostatic pressure.

The interpretation is unambiguous and it applies across the spectrum of materials, from normally ductile to low ductility and even truly brittle materials.

As hydrostatic pressure rises, the true strain to fracture increases significantly.

Sometimes by over 100%.

The external pressure effectively counteracts the internal tensile stresses that drive void nucleation and growth,

allowing the material to deform far beyond its normal limit.

We've completed an intensive deep dive into the analytic core of fracture mechanics.

If we distill this down, we realize that fracture in metals is a constant, highly sensitive energy battle.

It is.

The material is fighting to minimize the crack driving energy released by the stored elastic strain while maximizing the energy absorbed via plastic work, that gamma p term.

That's the high level takeaway.

But to truly master this chapter, you must internalize three specific sets of tools and concepts.

First, master the fundamental equations.

Understand the conceptual framework of theoretical strength.

Sigma max is approximately the square root of e gammas over a knot.

And use that context to appreciate the nominal fracture stress based on Oroen's modification.

Sigma f is proportional to the square root of e times the sum of gammas in gamma p, all over c.

And never forget that for any structural metal, the plastic work, gamma p, is the dominant term dictating resistance to failure.

Second, you must internalize the lessons from the key curves.

Know the cohesive force curve from figure 7 -2 and how the atoms fall apart at sigma max.

And understand the significance of the ductility versus second phase plot, figure 7 -6.

You have to recognize that inclusion geometry is often more dangerous than inclusion volume.

And most critically, fully understand the dB transition temperature schematic, figure 7 -17.

Use this schematic to explain why brittle failures happen at lower temperatures and why a notch or flaw is so dangerous.

Because it shifts the yield stress curve higher, raising the transition temperature, and making the material fail brittily at temperatures where it should be safe and ductile.

And third, remember the crucial role of context.

Stress concentration and triaxiality caused by notches are not just factors of safety.

They fundamentally change the failure mode from ductile to brittle.

Conversely, applying high hydrostatic pressure is a powerful engineering technique to suppress brittle failure and unlock latent ductility in difficult materials.

This deep dive into dislocation theories and micromechanics showed us exactly how microstructure dictates catastrophic failure.

The stress required for microprack nucleation depends heavily on factors like the frictional resistance to slip, tau i, and the grain size, d, to the minus one half.

So here's a final provocative thought for you to carry into your engineering practice.

Let's hear it.

If a material's resistance to brittle failure is so sensitive to tau i and d, consider this.

The engineer designing the specific heat treatment or purification process for a component is often far more responsible for preventing catastrophic brittle failure than the person calculating the nominal stress in the final structure.

That's a great point.

Modifying the microstructure is often the ultimate act of controlling safety.

That brings this deep dive to a close.

Thank you for joining us on this exploration of fracture mechanics.

Continue your journey into materials failure with the subsequent chapters on fatigue and creep.

Until next time, stay curious and keep learning.

This has been the deep dive.