There are several strengthening mechanisms, in particular, in ferritic steels, which we discussed in some more detail. What you'll see, however, in many steels is that very often they have their multi-compliment,

multi-phase microstructures. And so there are additional aspects to the strength of these microstructures. So one of the important microstructures

is traditionally, I have to say, is the constituent called perlite, which you probably know from your undergraduate years already, which consists of this lamellar microstructure of alternating ferrite and perlite.

And there are two, in the strength aspects of this microstructure, there are two important things. That is the interlamellar spacing, and then what we call the size of the perlite colonies. You can see here that you've got parallel lats, lamellas,

and that there are boundaries between different regions. Well, these regions we call perlite colonies. And the most important aspect of the strength

of this perlite appears to be related to the interlamellar spacing. That's the interlamellar spacing. And you can see that if you plot the yield

strength of perlite steels, various perlite steels, you find a whole patch type of relationship between the strength and the inverse of the square root. Yes? And that's very clear here. And it's for different pearlitic steels, some

that are fully pearlitic. For instance here, this is a real pearlitic steel, which would look like this, basically. And then we have these hyper-eutectoid, really high carbon steels with high levels of, much higher levels

of carbon. So you get an increase in the strength due to the fact that you have a lot more cementite in the microstructure. That's the reason, cementite in the microstructure, which

is a very hard carbide, if you didn't know. But still, you see approximately the same slope of about 250 megapascals per times square root of microns.

Increase in the strength as you reduce the size. So that is important. Now, do we actually have steels that look like this? Well, yes. There are many, many steels are used today

which use this microstructure, pearlitic microstructures. A lot of wire rod, cables have this microstructure.

Tire rod has this structure and makes use of this refinement to get these very high strengths. You can see the strengths here are very high. And so where does this strength come from? Well, you have two phases here.

And so you basically have a composite. And it's a composite where you have a very high, if you look at the cementite, you have a compound, a carbide, which has extremely high strength. And up to now, people have tried

to measure what is the strength of cementite. And for instance, if you pull or you compress it, and what you usually find if you take cementite, you make bulk cementite, and you try to deform it, it usually breaks. You cannot deform it. So we don't really know exactly what the yield

strength is of cementite as a bulk material. It tends to break, fracture before it plastically deforms. So we kind of think it's probably around 3,000 megapascal, 3 gigapascal, or perhaps more.

And the other phase is ferrite. And we know very well what the stress strain curve looks like. It's here, yes? And it's a relatively low strength phase, yes? So you have a hugely strong material

and a much softer material. And if you combine them both in this, well, you can get stress strain curves. And there's no fracture of the cementite in this perlite. Even if you give very large deformations, very

large strains, for instance, associated with making very thin wires, you get plastic deformation of the cementite in this case. And the stress strain curves for a typical lamellar perlite, like I showed you in the picture, is shown here. You have a yield strength around 500.

And tensile strengths are typically around 1,000 megapascal. So let me say a few more words here. So what we often do in, for instance,

constructional steels, we will not use like we do for wires, for instance. But in constructional steel, we'll use more or less cementite, excuse me, more or less perlite in the microstructure to change the strength of our steel.

So we'll talk about this in a moment. But that's basically what you do. You add carbon. And the carbon, it doesn't work in solution or anything. It works to make more perlite. And the more perlite you have, the stronger it is. And that's a traditional way of strengthening structural steels

is by, in fact, adding carbon to make more perlite in the microstructure. And you can see, of course, if you have 100% of perlite, well, you're kind of stuck around the gigapascal strength.

So nowadays, we work very differently. And the reason is, when you add carbon to make more perlite, to make your steel stronger, you basically, there are other properties that are not so good. Toughness becomes less because the cementite in the perlite

is not very tough. So it will break. It will easily fracture. So toughness is not so good. The other thing, which is very important technically, is joining.

When you have to weld, the presence of high amounts of carbon is not a good thing. So as a consequence, people have to use other solutions, non-carbon solutions to strengthen steels. And so we'll talk about those as we see examples of steels.

OK, so we have, of course, talked about martensite. We know that in martensite, we can have lots of carbon in the martensite because it's a supersaturated solution of carbon.

And there are equations available that will allow you to calculate the yield strength of martensite if you know the carbon content. And there are different formulas. And there may be different dependencies on the carbon content. Again, as I've told you, it's not surprising.

In fact, if you plot some of these equations, this is an equation where the carbon, the strength is proportional to the carbon content. Here, it's proportional to the square root of the carbon content. So it's a formula for the yield strength as a function of the carbon content.

And you can see there's not really much difference. So you can't really tell whether you have a square root of carbon or proportional to the carbon content relation between yield strength and carbon content.

So these are these equations. And note, please, that when the concentration of carbon becomes very low in martensite, you can make martensite. And it will be harder than ferrite.

But it will be relatively soft. So what makes martensite strong and hard is not the fact that it's martensitically transformed. What makes it hard and strong and brittle, et cetera, is the fact that you have carbon in solid solution.

So very often, you hear people say, well, martensite is brittle. Martensite is not brittle. There's no reason why martensite should be brittle. Martensite is the result of a martensitic transformation. And of course, in certain cases, such as in the case of steels, you will have a finer microstructure.

And you will have dislocations density, which is relatively high, because you need transformation dislocations. But it's not very hard. You can see here. And it's certainly not necessarily brittle.

What makes it brittle and hard is the fact that you have carbon in solid solution. And you know that this carbon in solid solution is very unhappy, because the solubility of carbon in ferrite

is zero at room temperature. So very quickly, there will be a tendency for the carbon to move out of the lattice. And that is the reason why at low carbon levels, we have so little strength. The reason is because the microstructure of martensite

contains what we call lats, so microstructural units with boundaries. And these lats contain large densities of dislocations.

The carbon will readily go to the strain field of the dislocations. And as a consequence, it will stop distorting. Remember this tetragonal distortion I talked about? This tetragonal distortion disappears.

And so the hardening effect of the carbon is limited.

Now, what happens to the martensite also is that, in particular, in engineering steels, where

we have relatively large carbon contents, for instance, 0.4% of carbon, we will temper this microstructure. So if the carbon content, say, is 0.4%,

we will temper the microstructure. And we will obtain carbide particles in the microstructure, fine carbide particles. So again, if your carbon content is low, less than 0.2%

or even lower, the carbon will go to your dislocations. And that gives you soft martensite. If it's higher, for instance, when you're making engineering steels, what's

an engineering steel? Steels, for instance, that you use to make crankshafts for motors, for instance, or you use to make transmissions, gears, et cetera. So this is the kind of carbon levels these steels have,

0.4. And so here, you find out that, in these steels, there is also an impact from the distance between the carbide particles in tempered martensite.

And of course, where does this come from, this relation? Well, that basically comes from the precipitation strengthening. So and again, you can see here, and it's both for tempered martensite and for bainite.

You can see that the relation, there is a one over the distance relation. And that, of course, what you want to have is a high density of very small particles.

We already know this from the discussion of precipitation strengthening we gave. Carbides are impenetrable particles. They cannot be cut by dislocation. And so you need to have high density and very small radii. And you have a strong increase in the strength.

So that's another parameter. The lat size also plays a role, lat size being this here, the size of these microstructural units.

And this is some data here showing that the lat size here, the increment, the strength increment due to lat size is proportional to one over distance, the lat width,

excuse me. So it's a kind of a, so the width of these lat here. I do have to say, however, let's go back, that there is an alternative idea which questions this idea

is the following. So the idea that the lats are somehow structurally determining the strength of the martensite is based on the idea that the lat boundaries

act as very efficient obstacle to dislocations. And so if you look at the microscope picture of lat martensite, that's indeed what you would think.

Because you see all these small boundaries, all these narrow lats. And you can clearly see these lat boundaries. The problem is that the misorientation

between these lats is actually very, very, very small. If you do, for instance, a diffraction analysis of lat boundaries, many lats, it looks like a single crystal. So these lat boundaries are not very good boundaries.

They're not really as efficient as real grain boundaries to stop dislocations from moving. Because there is not much, very, very little misorientation between them.

So the alternative theory is that it's the packet size which determines

the strength of the martensite, which is a structural. And so what's a packet size? So when you have original austenite grain, the austenite grain can transfer to ferrite.

And it can be, as I told you, 24 different variants. Well, groups of variants will appear together when they share the same, what we call the same, 111 plane.

So you can readily visualize them. They all share their different variants. But they all share the same 111 planes.

So they form these blocks that you can see, which we call packets. And that's clearly the better unit, structural unit, to specify what the influences of microstructure

on the strength, so the packet size. And I'm going to skip things about the stress strain curves, which you are familiar with.

I just want to stop here, because this is a very important graph here. And I just want to make sure we understand each other when

it comes to strength. There is, of course, from your undergraduate education, you know that when things get stronger, they tend to be less plastically formable.

Something that's hard will be, or strong will be, brittle. Or you won't be able to deform it. That's not really true. Of course, in general, if you compare ceramics to metals, yes, it's true.

But it's not right. You're not comparing the same things. So if you work in one specific area of materials, like steels, stronger doesn't mean less plastically formable. And the parameter that plays the important role here

is the strain hardening.

And this is illustrated here. If you plot the stress strain curve for what's called a high strength IF steel, you find this black line here.

And if you plot also the strain hardening, and that's very simply just the slope of this line

in a true stress, true strain diagram, at one point they will intersect. And that point is the point of instability. Beyond that point, you don't have uniform deformation anymore.

And that's a very good measure for plasticity of a steel. So if you make this material stronger without changing the slope of this material,

if I have a stress strain curve and derivative, and I make exactly the same, say, a material which just has a higher yield strength with exactly the same slope, you are right.

Now the intersection is here. So I will have a stronger material with less formability, with less ability to deform plastically in a uniform matter. But you can tweak the strain hardening.

You don't have to keep the strain hardening the same. You can change it by working on the microstructure. And we already told you that the way you do this is by storing dislocations and finding

ways to store dislocations in the microstructure. And an example here is, for instance, this is a trip steel. It's much stronger. You can see here this one is about 500 megapascal. This one is double the strength, double the strength.

And is it less formable? No, it's not less formable. It's more formable. So by working on the microstructure, you can independently change strength and formability. And higher strength does not mean less formability.

I don't know whether it means less formability. It can mean less formability, because it also means more formability. As long as I don't know what the strain hardening is, you can't tell. But in steels, so you have to be careful about this.

And yes, you can get larger uniform elongation and higher strength, no problems, if you tweak the microstructure of your steel and you understand the mechanical properties.

Now in your notes, you will see that there's actually a lot you can already do with available knowledge. For instance, this is an empirical stress strain

equation. Let's have a look a few slides back here, which one it is. It's this one. It's the Swift equation. It's very similar to what you're

used to the Holloman or the Ludwig equation, but it's the Swift equation. So very careful when you give data of strain hardening or yield strain or whatever to say what empirical equation did you fit your data, which

is almost never done. So you always kind of have to guess. So say you have this Swift equation, empirical equation. So you have parameters A should be capital A and capital B. If you

could correct this. There are empirical equations based on a large amount of data for ferrite, perlite, and martensite, which will allow you to actually plot stress strain

curves for ferrite, perlite, bainite, and martensite, because the parameter A, the parameter B, and the parameter N have been determined as a function of the most important alloying elements, silicon, manganese,

phosphorus, and the grain size. There's nothing theoretical about this approach. It is very engineering approach. It works well enough if you need to do some research,

certainly exploring the effect of compositions, et cetera. It's a good way to work, because it gives you a stress strain curve, rather than just a yield strength. And if you have a stress strain curve, for instance,

if you say the plasticity is, the plastic strain is zero, that means you can determine the yield strength, yes? OK. Right, so it's a very nice thing to use.

And again, the things you see are what you expect. For instance, for martensite, the main effect of the main effect is due to carbon, yes? Whatever the composition is of martensite, the impact of the carbon overwhelms everything else.

OK? So you can work with this. And if you do this, you will be able to plot stress strain curves for various typical compositions.

For instance, this would be for interstitial free steel, steel which basically contains a few tenths of a percent of manganese and silicon, and no free carbon. So this would be the stress strain curve. This would be for bainite, yes?

And there is martensite. And you can see the variety of strengths you have available, yes, when you work with steels, and also the kind of elongations that are available for these microstructures.

And this kind of gives you a range, or a range for ferrite, perlite, and martensite as a function of the carbon content. And what is interesting to see is that in fact this overlap

of properties. For instance, you can see here that there is a lot of overlap between martensite and bainite in terms of strength, yes?

And also with bainite and perlite on the lower end of the strength scale, yes? And so it also means that look at the bainite,

for instance, the property range goes from 100 ppm to 0.8. And the one for perlite goes from 0.4 to 0.8. So the fact that you can make the same mechanical properties

with a much lower carbon content using bainite, bainitic microstructure, is actually actively used to make steels that are high strength, but also weldable, with lower carbon content.

And for instance, in one field that's very important lately of line pipe steel development, that those are steels used to make pipes to transport gases or petroleum products,

there is a tendency to use bainite instead of perlite in the microstructures. That's one thing. Another example is the evolution of rail steels. Rail steels are traditionally, just like wire steels,

are traditionally pearlitic, yes? You can make bainitic rails, yes? And that's also an evolution that's happening in certain areas where we see people using low carbon bainite instead

of pearlitic microstructures. And that's because you can get the same type of properties, and actually a wider range, than perlite.

The reason why we don't see these microstructures as much as we could in current technology is because making bainite requires a different processing, yes? More, a little bit more difficult processing.

You need to alloy much less, but you need to compensate for the loss, the reduction of carbon. So that means that lots of, and we don't know much about the behavior of bainite in applications. And that makes it difficult to introduce some bainitic grades

in many applications where they would, in principle, be more useful. But anyway, it's an evolution that's coming. So when we have multi-phases, more phases in steels, we can have two extremes of equal strain deformation

or equal stress behavior. And in practice, we have situations somewhere in between. And this means that the harder of the two phases

will take on the larger stresses. The softer will take on the higher strains, yes? And this is an example here of what happens, for instance, in a prolific microstructure where

you have ferrite. Actually, no, this is a steel that contains ferrite and cementite, but not perlite.

The perlite is formed if you cool down quickly. If you cool down a steel with a low carbon content slowly, you will form ferrite grains and cementite particles. This is the stress drinker for such a steel,

where you don't have perlite, but you have cementite. Anyway, the cementite has this extremely hard phase in the microstructure. The ferrite is this one here, is this one. The behavior of the steel is somewhere,

the actual steel when you do this, is actually somewhere in between. And if you look at every point in the stress strain curve of the steel, the actual stress and strain in the ferrite is here, and the actual stress and strain

in the cementite is here. So you have a huge, what we say, partitioning of stress and strain in the microstructure. And you can see the cementite takes a very much larger load,

but the deformation, the strain on the cementite is very small. And so that means that even though the cementite is a brittle compound, because most of the deformation is done by the ferrite matrix, it takes quite some deformation, quite a straining,

before the cementite particles actually break. Although, if you take the cementite separately and you would try to deform it, it would almost break instantly. It couldn't plastically deform.

In the composite, it's different. So that's why, if you wonder why in perlite, you're pulling perlite, why doesn't it instantly break? Because the cementite is so brittle. That's because of this phenomenon. You get partitioning.

Another nice example of steel with the measurements of partitioning has been done. So it's a duplex stainless steel. It contains 50% austenite and 50% ferrite. It's highly alloyed steel. Again, you can see that the actual stress

strain for a certain point on the stress strain curve of the steel, the macroscopic stress strain curve, there is a partitioning between the stress and the strain, between the harder ferrite and the softer

austenite. And you see, the ferrite will have a higher stress and less strain. And the austenite will have a higher strain and less stress.

Some more on the multiphase steel behavior. So this is martensite stress strain curve.

And this is ferrite stress strain curve. And you can make steels, which we call DP steels, where you have microstructures which contain ferrite and then some martensite particles also in the microstructure.

And it's possible, as we've discussed I think already, to introduce martensite in the microstructure and change the amount of martensite. So if I start with a microstruct that's

fully ferritic, the stress strain curve will be this one. And if it's fully martensitic, it will be here. But by changing the volume fraction of martensite, I can basically vary, have different stress strain curves.

So if I measure, for instance, the tensile strength, I'm sorry that this is the tensile strength here, I will see that the more I add martensite,

the higher it becomes. And in practice, this relation is almost linear. So the amount of martensite, if you plot the amount of martensite, volume fraction of martensite, in the x-axis for DP steel, and you measure the tensile

strength, you'll see that there is a linear relation between the tensile strength and the volume fraction of martensite. And so you can make basically a DP steel with, say, a tensile strength of 500.

And you can go up to 1,000. So that's quite a variety, quite a range of strengths that you have there to play with.

So I'm not going to discuss the formability issues. You can just read it. And some of the aspects we've already discussed.

There is this graph here that I think is important, which I want you to definitely have a look at. It's that we were talking, remember,

about reduced grain sizes, and reduction of grain sizes in steels, and how when you reduced the grain size, you would also decrease formability.

And this is an example of the way you can look to this. So you know that if you plot the stress strain curve of a steel, and you plot the derivative, where the derivative is equal to the stress,

that is the point of uniform elongation. And that's a measure for plastic formability. If you increase, excuse me, if you decrease the grain size, you basically move this stress strain curve upward.

Yes. And so you end up with a smaller uniform elongation. And of course, if you continue doing this,

you can see that, yes, you will be able to achieve a lot of strength, but you will not be able to achieve a lot of formability. So you remember that if you plot the strength,

or the tensile strength as a function of 1 over the root, it goes like this. And actually, when you plot the tensile strength, the slope is slightly lower. Yes.

So what I show here is actually correct. You cannot increase, so the reduction of the grain size doesn't allow you, doesn't give you possibilities to increase the strain hardening. So the strain hardening is actually,

as you increase the strength through to reduction of the grain size, it actually gets lower. So it's actually worse than here.

But that's an important point. In general, steels, which tend to be single phase, or mostly single phase ferritic steels,

we will indeed see that when you increase the tensile strength, you will see a reduction in the strain hardening. Yes. And you can measure this. The engineering approach to measuring strain hardening is looking at the n value. Yes.

The n value from the Hollomon equation. That's what happens. That's because an interstitial free steel, a low carbon steel, a solution strengthen steel, or a structural steel, or an acyclic steel, they're basically mainly ferritic steels.

Yes. So if you want to increase the strain hardening, you have to work on the microstructure. So we won't discuss hardness here and fracture.

Yes. Perhaps a word about fracture here, just so we understand each other, if there is a,

when you have a sample, a steel sample, that you test, so what happens beyond the UTS is not unimportant. In certain application, it is totally unimportant

to a certain extent. For instance, there are many applications where you never strain the material beyond 10%, 20%, or something like that. So you're never close to necking.

For instance, when you make car bodies, you never deform the material so that you get non-uniform necking, because obviously you don't want the part to look like this. So in many applications, the only forming that you do

is within the uniform straining range. So anyway, but you test the material. But having said this, there are many steels that you can strain even beyond the UTS a lot.

For instance, IF steels, you can still strain 15% or more beyond the uniform necking. And so you get a lot of the necking region is very high. So you get the diffuse necking.

This diffuse necking is related to the creation of internal defects. Yes? So why do you get suddenly a diffuse necking, yes? And so this diffuse neck, this shape change here,

that's not the same as this localized neck, which occurs because of mechanical reasons. This diffuse necking is related to the fact that you've created internal voids in your sample.

And at a certain place, you get an increase in the strain rate. And then all the deformation is localized here. So as soon as this occurs, by the way, it means that all this, the rest of the sample here,

stops deforming. Stops. No more deformation. So that means that when you're doing research and you have a broken sample, this area here and this area is actually very different.

So if you look at the stress-strain curve, elastic part, and this, and this is the UTS. This sample here has the microstructure at the UTS. And this microstructure here is the sample here.

So it's a very different microstructure. And I often see this. And then also, if you're doing your samples carefully, here the microstructure is here. So a single specimen can tell you a lot of things.

And very often I see students not making use of this fact that you have a lot of information on one sample in terms of microstructure. In particular, because a broken sample actually

has a lot of information of the sample at different stages. And so if you were to look at your material here,

just beyond the UTS, this is what you would see. You'd see grain boundaries where you have formed holes. Or here an inclusion, and around the inclusion

is a void. Or here, this is a perlite in ferrite. You see here voids at the perlite-ferrite interface. Yes? So voids are, and the ability to create voids

in the microstructure, actually determine what my uniform elongation is, to a large extent from a microstructural point of view. And actually, if you plot the, so if you make bars,

tensile bars, and you look at the area reduction in the tensile test, and you do this as a function of the percentage of cementite

in your microstructure, you see a very strong relation. The more you have particles that will give you, that will allow you to nucleate voids in the microstructure, the lower your formability will be. The quicker you will start having

diffuse necking in your material. So these are called cavity nucleation second phases, have a very strong impact on ductility, and in particular, the necking.

So you can improve things. That's why it's important when you design steels, or in practice, to make sure that internal cleanliness of your steel is very high, so that you

don't have unnecessary sulfide or non-metallic inclusions in your steel. Because they will also impact your properties. Because all these non-metallic inclusions, these carbides, et cetera, that are not necessary,

will very often act as cavity nucleating effects, inclusions during the straining. And the same holds with very hard phases,

such as martensite. This is, for instance, martensite in an austenite. This is an austenite matrix, and here we have some martensite in it. You can see that it's a very high carbon martensite, by the way, so it's very brittle. You can see that in this case, you

don't have cavity nucleation. But when the martensite fractures, it creates a lot of cracking that goes into the austenite also. And you're familiar with the Sharpie test?

Again, I want to stress to you, because we'll come back to that when we discuss some structural steels, is the ferritic steels. For instance, this is ferritic steel 1% of manganese, a binary steel.

You've got this incredibly sharp drop in the absorbed energy. And it's usually associated with brittle transgranular fracture,

this type of behavior. That is due to the dislocations, the screw dislocations which have such low mobility. Because you don't have this effect in the austenite.

In austenite, you have much lower pearls stresses. So the screw dislocations are not subject to these very strong pearls valleys. There is a decrease, but it's never that strong.

Good. So we have said that in relation to this ductile to brittle transformation,

that grain size reduction is very positive. So if we plot the yield strength as a function of the inverse square root of the grain size,

we see that on top there. We see an increase in our yield strength. So the material becomes strong. We know this. If I plot the ductile to brittle transition temperature,

so where there is a sharp decline in the absorbed energy in the Sharpe test, I see that when I reduce the grain size, I have a drop in this transition

temperature. And what is very important here is that if you use a conventional way,

a conventional steel, a perlite ferrite with a perlite ferrite microstructure, you can only go up to here in terms of properties. You can reduce the grain size, but you cannot reduce it as

much as in an HSLA steel. So you remember these are steels where you add niobium, and it allows you to reduce the grain size considerably. And so this is what the niobium additions and the processing of these niobium steels

allows you to do is a reduction of the grain size from 10 to about five or six, and you get an additional increase in strength and reduction in ductile to brittle transformation temperature.

You, of course, do lose elongations. However, in the applications where we use TMCP niobium alloyed steels, we don't need very high elongations.

When you make a line pipe, you don't have huge elongations. When you make steel plates for offshore, same thing. You basically do not deform them very much. If you use these plates for shipbuilding, same thing.

So these are not. So you can live with the loss of, and it's not a very huge loss, but you can live with the loss the slight loss in uniform elongation when you reduce the grain size. Because you have three big advantages, more strength,

more toughness, and better weldability. And weldability is something that's very essential for offshore shipbuilding and line pipes.

So again, with these transition temperatures, we know we can improve things with.

with reductions of the grain size, there are certain additions which will work the other way. And in particular, phosphorus. Phosphorus will increase the transition,

ductile brittle transition temperature, a lot. So this is pure iron. Yes? Ductile brittle transition temperature is around 40, 50. If you add phosphorus, it increases.

0.2, 0.6. So people will not use phosphorus. And that's why in steel making, iron and steel making, phosphorus removal is so important.

Sulfur removal is important, but phosphorus is very important because it's difficult. It's difficult to go lower than 100 ppm. But it's very important to get as low as possible.

Right, so phosphorus, in many applications, you will not use it because the ductile brittle temperature goes up. And you remember, that was our best solid solution hardener in ferritic steels.

So it's very sad, but you can't use it. For silicon, that was the second best one. So the trouble is this. You get the same effect as you increase the amount of silicon. You increase the ductile to brittle transition temperature,

and that's, of course, a big problem. However, you will note that at lower amounts of silicon, there is actually a decrease. So people will add up to half a percent of silicon, also in constructional steels, knowing

that it has no negative impact. And that you can use, of course, even a half a percent of silicon will give you 50, 60 megapascal of strength. So it's something that is nice to have.

But there are maybe some people here that are interested in developing steels which contain larger amounts of silicon. Because as you know, silicon is a very interesting element

in connection to bainitic steels to suppress carbide formation. And so there, you want to have silicon contents which are in the range of around 1.5 to have this silicon work

effectively as suppressant. So that's a little bit of a delicate choice to make. And very little is known about this level of silicon in steels for structural applications.

But it's a silicon level that is acceptable in certain cases, such as in automotive applications.

You know that, and the reason is the following. When you test for ductile to brittle transitions with a Sharpie test, you're thinking of thick gauge materials. You're thinking about millimeters thick plate,

for instance. And so when you test them with the Sharpie test, you test at very high strain rates in plain strain conditions, as we say. And so those are very hard mechanical conditions

and high strain rate and low temperatures. In sheet material, because it's much thinner, it's more difficult to obtain plain strain conditions, testing conditions.

So even if the material is more brittle, it will not actually behave so badly as if it was thick material. So this is also something you have to realize. But anyway, for many applications or structural applications in gas and petroleum industry,

silicon not higher than half a percent, typically. And what happens if you have 3% of silicon, or even more? Are there situations where we have this much? Yes, electrical steels.

Electrical steels have these very high silicon levels. And it turns out that this material, these materials, are indeed very brittle. And there are cases where, at this level of silicon, the slabs, the material as-caste, will just fracture in a brittle fashion,

just under the influence of its own weight. So silicon is very sensitive to people in iron and steel making, because they have to clean up afterwards.

Manganese, I said. So manganese here, zero manganese, half a percent manganese. You get a decrease in the ductile to brittle transition temperature, and an increase in the absorbed energy. So manganese is always good to use. What about carbon?

Well, carbon, it's a mixed message here. What you see, in general, with carbon is as you go from very low carbon, so 10 to the minus 3 means 0.003.

So that is 30 ppm of carbon. So 30 ppm of carbon, I see that if I increase the carbon a little bit, if I increase the carbon, I see a decrease in the ductile

to brittle transition temperature. Yes? And this continues up to around 10 to the minus 2. So that's 0.01. Yes? After that, I see an increase.

By the way, these are different curves, because they're different phosphorus contents. And so you can see here, again, that the more phosphorus you add, the worse it gets. But this is important. Why does it increase here, and why does it decrease here? Yes?

Well, it's very simple. As long as carbon is in solid solution in the ferrite, or in solution in the ferrite, it will have a positive effect on the ductile to brittle transition temperature.

As soon as you form cementite particles in the microstructure, so that's at around this maximum solubility of carbon in ferrite, so that's about 200 ppm, so around 0.01, 0.02 mass percent,

you form cementite in the microstructure, and you get worsening of the increase of the ductile to brittle transition temperature. So why does the carbon in solution

have a positive impact? Well, at these very low contents of carbon in ferrite, the carbon tends to go to the few dislocations that are available and the grain boundaries. And in the grain boundaries, it's a well-documented fact

that carbon increases the grain boundary cohesion. And that's how we explain the reason why the carbon improves the ductile to brittle transition temperature.

So carbon and manganese always improve things, and carbon also improves things as long as carbon is in solution.

And you can learn a lot from looking at the fracture surfaces in ductile to brittle transformation, whether it's some cleavage type of fracture or whether you have intergranular fracture or whether you have what we describe as dimples.

And the dimples are nothing else than these voids that you form, the inner voids that you form that fracture. And then usually these dimples will contain the nucleation site, a little particle that's

inside the matrix and which has caused the void formation and then finally the breakage of this void.

So OK, and I won't be talking about fatigue. There is a separate course on this. And I will stop now anyway. And then we'll