Chapter 17: Rolling 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.

Welcome back to the Deep Dive, where we take complex technical sources, strip away the noise, and hand you the pure, practical knowledge.

Today we are undertaking a deep dive into, well, what is arguably the absolute foundation of modern material manufacturing.

That's right, we're jumping into chapter 17, the rolling of metals, straight from the mechanical metallurgy playbook.

And, you know, this chapter is foundational not just for metallurgists, but really for any engineer who specifies or designs parts made from sheet, strip, or structural shapes.

Right, because rolling is, I mean, it's pretty much everywhere.

It is, without question, the most widely adopted metalworking process globally.

And when we say widely adopted, we mean it's the default mechanism for transforming these huge cast ingots into usable, uniform products.

That's it.

So why does it hold this dominant position?

What's the core appeal that outcompetes something like forging or extrusion for sheer volume?

It really boils down to two critical factors.

First, incredibly high productivity.

I mean, continuous rolling mills run astonishingly fast.

And second, you get the highest level of precise dimensional control you can possibly achieve in bulk manufacturing.

The process is inherently self -correcting when it's managed properly.

Okay, let's unpack the core physics driving this then, because we're not talking about a gentle squeeze here.

This is precision metal shaping at a scale of many mega Newtons of force.

Oh, absolutely.

So what is the fundamental mechanical action that makes rolling work so efficiently?

It relies on a very carefully balanced combination of forces.

So you have the massive vertical compressive stresses exerted by the rolls, which squeeze the material and cause that bulk plastic deformation.

Right, the squishing part.

The squishing part.

But the absolutely essential component is the horizontal force.

The surface shear stresses or friction generated between the roll surface and the metal.

That friction is the engine.

It's what grips the metal and actually draws it into the roll gap, making the whole continuous operation possible.

So compression is the shaper, but friction is the dynamic mechanism.

It's the hook that pulls the material in.

If you lost friction,

you'd just have two rapidly spinning rolls and a stationary slab of metal sitting there.

You've got it.

It would do nothing.

And for you, the engineer or designer, understanding the physics behind these forces all the way down to the mathematical relationships is, well, it's non -negotiable.

Why is that level of detail so critical for someone who

you know, might simply need to order some steel sheet or aluminum plate?

Because this chapter defines the manufacturing limits.

It dictates the maximum thickness reduction you can achieve in a single pass before the mill just fails to grip the metal.

It defines the maximum load the mill must bear.

And because of that, the minimum gauge you can roll.

If you don't understand the constraints imposed by friction and roll diameter and mill stiffness,

you might specify a material or a dimension that is either impossible or just prohibitively expensive to produce.

So it helps you design with the manufacturing process, not against it.

That's the perfect way to put it.

Let's start with the specific language used in the mill, because I get the sense the terminology isn't interchangeable.

Not at all.

We define rolling as the process of achieving plastic deformation by passing metal between rotating rolls, which reduces its thickness.

But the products themselves have a very strict hierarchy, especially for steel.

That's right.

We start with the massive initial casting, which is the ingot.

But the terminology gets highly specific as we move through the early stages of production, what we call the semi finished products.

So walk us through that chain, starting with the first product of the ingot breakdown.

Okay, so the first major product after the ingot is the bloom.

A bloom is defined geometrically by its really large cross sectional area, which has to be greater than 230 square centimeters.

It's the beefiest intermediate product.

And from there?

If we take that bloom and we reduce it further, we produce a billet.

A billet is generally smaller,

often around 40 by 40 millimeters in cross section.

So the bloom is the parent of the billet.

Precisely.

Now, separate from those two shapes is the slab.

A slab is also a semi finished product, and it requires a cross sectional area greater than 100 square centimeters.

The critical distinguishing feature of a slab, though, is its geometry.

Its width has to be at least twice its thickness.

Ah, so it's wide and flat.

Exactly.

Slabs are typically reserved for rolling into sheet and strip, while billets are what you use for bars and structural shapes.

Got it.

So once we've processed these semi finished shapes, we get to the final products, and those are categorized just by thickness.

That's a key line in the sand.

Six millimeters.

Anything with a thickness greater than six millimeters is classified as plate.

Okay.

If the thickness is less than six millimeters, it becomes either sheet or strip.

And how do we differentiate between sheet and strip if they share the same thickness definition?

That differentiation is typically based on width.

So strip refers to rolled products, generally less than 600 millimeters wide.

Right.

Whereas sheet refers to the wider products.

And this distinction is really vital for logistics and handling and how the material is stored.

Strip is usually coiled up.

Sheet might be processed into flat packs.

We've been discussing traditional rolling, which starts with a big cast ingot.

But the source material touches on some highly specialized methods that use alternative inputs.

Yes, this is powder rolling.

Instead of starting with a massive cast slab, you begin with dense metal powder, which you feed between the rolls to create what's called a green strip.

A green strip.

Yeah, it's not very strong yet.

That strip is then centered and further rolled.

So what's the advantage of doing it that way?

It's all about minimizing contamination.

You bypass the massive melting and casting stages where impurities can get in.

Plus you gain microstructural benefits like a very fine grain size and sometimes a preferred crystalline orientation, which enhances certain mechanical properties.

That sounds like a powerful way to make specialty alloys.

And finally, we have specialized rolling that changes the shape dramatically, not just the thickness.

We call that roll forming.

This is a cold rolling process where you take a metal strip and continuously bend it into complex cross sections.

You know, think of channels, U shapes, or complex window frames.

It passes through a series of contoured rolls.

The thickness reduction might be minimal, but the change in shame is dramatic and continuous.

And thread rolling.

That's another one.

Thread rolling is where threads are formed on a blank piece by rolling it between two groove dies.

So you're creating threads through plastic deformation rather than by cutting metal away.

Let's look at the infrastructure necessary to handle these forces.

A rolling mill has to be designed for rigidity.

When we consider the massive forces reaching multi -mega newton levels, what are the basic components that make up a mill stand?

At its heart, a mill stand really consists of four things.

You have the rolls themselves, which are the working surfaces.

These are supported by high precision bearings.

Then you have the massive rigid housing or frame.

That's the backbone that contains those massive separating forces.

And the drive, of course.

And finally, the powerful drive mechanism that controls the speed and torque.

And because forces can easily exceed say 10 or 20 mega newtons, that's the weight of several thousand tons, the construction has to be immensely rigid to prevent excessive elastic distortion.

Which has a name, right?

We call it mill spring.

Okay, so we classify these mills primarily by how the rolls are arranged in the housing, which is what we see in figure 17 to 1 in the text.

The simplest is the two high mill.

A standard two high mill has two rolls rotating in one direction.

Pretty simple.

But if you need multiple passes to get your reduction, you have a couple options.

Right.

One is the two high pullover, where the metal is literally manually returned over the rolls for the next pass.

Sounds slow.

It is.

The second and far more common is the two high reversing mill, where the rolls stop, they reverse direction, and the workpiece is passed back through.

But reversing those massive drive motors seems complex, slow, and probably power intensive.

Is there a more elegant solution for back and forth rolling?

There is, yes.

The three high mill.

This is a brilliant configuration that avoids reversing the drive.

How does that work?

It uses a setup of three rolls.

Upper, middle, and lower.

The upper and lower rolls are driven, while the middle roll is often idle.

So the metal can pass forward between the lower and middle rolls, and then, using a mechanical lift table, it can return backward through the middle and upper rolls for the next reduction pass.

No reversing needed.

That is clever.

Now we move to the configurations needed for precision and for thin products.

Why do we need to introduce backup rolls like we see in the four high and cluster mills?

It's a classic engineering trade -off.

For higher efficiency and lower friction, you actually want to use smaller diameter work rolls.

Smaller rolls.

Why is that?

A smaller radius means a smaller arc of contact with the metal, which reduces the overall force required for the same reduction.

But small rolls can't be very strong.

Exactly.

Small diameter rolls are naturally weak.

Under the colossal rolling load, they would deflect and bend like a bow, causing the product to be thinner at the edges and thicker in the middle.

So the solution is just structural support.

That's it.

The four high mill is the answer.

It features two small diameter work rolls, the ones touching the metal, supported by two much larger diameter backup rolls.

These backup rolls add the necessary strength and rigidity, preventing the work rolls from bending.

And this is standard for thin sheet.

The standard for rolling thin sheet.

Yes.

And then the cluster mill just takes that principle to the absolute extreme.

It does.

It's a modification of the four high mill.

And you use it when the working rolls must be extremely small, like for high strength alloys or very thin foil.

It uses two tiny working rolls supported by multiple, sometimes up to 20, backing rolls in a configuration designed to maintain absolutely uniform contact pressure.

Is there a common name for that type of mill?

The Senzimir mill is the most famous example of a cluster mill.

It's capable of achieving micron level tolerances on specialty metals.

Beyond single stands, continuous production really defines modern metallurgy.

Tell us about the continuous mill shown in figure 17 -2.

A continuous mill is a synchronized series of rolling stands, all operating in tandem.

The strip moves continuously from the first stand to the last, undergoing reduction at each stage without ever stopping.

The technical challenge there must be immense, especially with the need for synchronization.

Oh, it's crucial.

We rely on the principle of volume continuity.

The volume of metal passing any point per second has to be constant.

BHV must be constant across all stands.

Since the thickness, HA is continuously decreasing at each stand.

The velocity V must continuously increase.

Exactly.

Therefore, the peripheral speed of the rolls in stand 2 must be faster than stand 1, and stand 3 faster than stand 2, and so on.

And if the speeds are not perfectly synchronized?

The strip will either bunch up between stands if the next roll is too slow, or it will be stretched if the next roll is too fast.

Either way, you get gauge defects or even breakage.

And finally, the exotic planetary mill from figure 17 -3.

You mentioned its action is closer to forging than to rolling.

That's because of its unique arrangement.

You have a heavy backing roll surrounded by numerous small planetary rolls.

So instead of a single, continuous reduction, the planetary rolls rapidly orbit the backing roll, and each one provides a tiny, instantaneous reduction as it passes over the slab.

So the entire slab goes from its initial thickness to the final strip gauge in a single pass through a sequence of extremely rapid, small, compressive events.

It's highly effective for significant single pass reductions.

Let's bring in a major variable now.

Temperature.

Hot rolling occurs above the Mel's recrystallization temperature.

What are the key objectives of doing this high temperature process?

The primary objective is really bulk transformation.

We're breaking down the coarse, brittle, as -cast structure of that ingot into those primary semi -finished products, the blooms, billets, or slabs.

So it's about refining the internal structure.

Yes.

Hot rolling refines the microstructure, it closes up internal porosity and voids, and it ensures the material is homogeneous enough for any subsequent working.

This often starts in what's called the roughing mill.

Correct.

The first few passes in the roughing mill are designed to handle that massive ingot.

And the reductions in these initial passes are often relatively small.

Do they serve any other purpose?

They also have a mechanical cleaning function.

The heavy scale, that's the oxidized layer that forms on the surface of hot metal, is often broken off and removed during these early passes.

I found the discussion of ingot manipulation fascinating,

specifically turning the ingots 90 degrees between passes.

Why is turning so important in hot rolling?

It's primarily about managing the flow of the metal and achieving the necessary reduction in all dimensions.

During rolling, metal tends to spread laterally, so it increases its width.

If we only rolled in one direction, the slab would become excessively wide.

So you turn it 90 degrees and roll it back to the right width.

Exactly.

By turning the ingot, we roll in the new, wider dimension, which helps maintain the desired width profile.

Plus, we often use vertical edge rolls, we call them edging rolls, to control the width precisely during the pass.

Moving to modern production, these continuous hot strip mills are just engineering marvels.

What kind of temperatures are we talking about here?

Slabs are typically reheated to between 1100 and 1300 degrees Celsius.

Wow.

They pass through a roughing train, often four high mills, and then a finishing train, sometimes six or more, four high finishing stands in tandem.

And the control here is incredibly tight.

I bet.

The temperature of the strip as it leaves the last finishing stand is precisely maintained, often between 700 and 900 degrees Celsius.

And why is that final temperature so critical?

Because that temperature dictates the final metallurgical structure, the grain size, and the phase formation.

For instance, in steel, cooling too fast or too slow affects the formation of desirable equiaxed ferret grains, which impacts the final strength and formability of the strip.

Do non -ferrous metals like aluminum or copper require the same scale and specialization?

Generally less so.

Non -ferrous metals usually start as smaller ingots, and critically they have lower flow stresses.

They are just easier to deform.

So simpler equipment works?

Yes.

Simpler equipment like two and three high mills are often sufficient for hot rolling them.

Although four high mills are still standard for high volume aluminum alloys, where precision is still required.

So if hot rolling is about bulk breakdown and structural refinement, cold rolling done below the recrystallization temperature is the process of precision finishing.

What are the three primary reasons to subject metal to cold rolling?

The goals shift entirely from bulk refinement to surface and strength.

First, it produces a superior surface finish that hot rolling simply cannot achieve.

Second, it yields much tighter dimensional tolerances and flatter material.

And third, and this is metallurgically vital, it dramatically increases the metal strength through strain hardening or work hardening.

This superior finish must demand a very clean starting material.

Absolutely.

The input material must be meticulously clean.

Usually it's hot rolled strip that has been surface treated, often by pickling, which is an acid cleaning process, to remove any residual scale and oxidation.

And this is also done on tandem mills.

Yes, cold rolling is typically done on high speed, four high tandem mills, often three to five stands running continuously, which apply the necessary front and back tension that we'll talk more about later.

How much reduction can be achieved in cold rolling?

The total reduction typically ranges from 50 to 90 percent.

But engineers follow a really crucial rule about how that reduction is distributed.

And what's that?

The lowest percentage reduction usually occurs in the final pass.

That seems counterintuitive if you want a small final thickness.

Wouldn't you want to hit it hard on the last pass?

You might think so, but trying to achieve a massive reduction in the final pass makes controlling the flatness, the gauge, and the surface finish extremely difficult.

Ah, because of the massive forces involved.

Exactly, and the resulting mill spring.

By saving the smallest reduction for the final pass, you ensure the tightest possible control over the most important characteristics.

That brings us to a specialized final step for steel.

Temper rolling, which is also known as the skin pass.

This is a tiny reduction, right?

Often less than one percent.

What is the highly specific mechanical purpose of this final touch?

It's a trick to fix a very specific metallurgical problem.

When annealed steel cools, it can exhibit a phenomenon called the yield point.

When this steel is subsequently stamped or deep drawn, that yield point causes discontinuous yielding, which manifests as visible surface defects known as stretchered strains.

They look like long, unsightly surface markings.

So the skin pass prevents this?

Yes.

That small cold reduction introduces just enough strain hardening to permanently eliminate that yield point phenomenon.

This ensures the steel yields continuously and uniformly during subsequent forming operations.

It also improves the general flatness and surface smoothness slightly.

And if the sheet still has some mild flatness issues after all that, what's the fix?

We use stretcher leveling.

This involves gripping the edges of the sheet with jaws and literally stretching the material with a pure tensile force.

That pulls it flat.

That stretching permanently corrects mild waviness or buckle problems by achieving uniform plastic strain across the entire width of the sheet.

We've focused heavily on flat products.

But what about structural shapes like I -beams or channels or railroad rails?

How does the process change when the goal isn't just sickness reduction, but forming a complex profile?

The key differentiating technology is the use of grooved rolls, which we see in figure 17 to 4.

Unlike flat rolling where the reduction is uniform across the width, here the metal cross section is reduced in one direction only per pass as it's constrained by the shape of the groove.

And since you can't achieve the final shape in one go, how does the material move through the mill?

It requires a sequence of carefully designed passes.

And the workpiece is almost always rotated 90 degrees between successive passes.

Why the rotation?

It ensures that the reduction is distributed equally around the cross section, and it prevents the buildup of stresses in a single direction that could lead to cracking or uneven deformation.

And spread control must be a nightmare in this process.

The metal constantly wants to expand laterally and just fill those grooves.

It's the critical design challenge of shape rolling.

The roll pass sequence has to be intricately designed to manage this lateral spread.

A very common method involves alternating passes between simple shapes, typically oval and square shaped grooves.

An oval and then a square.

How does that help?

The square pass limits the lateral spread and ensures the section remains compact.

Then the oval pass begins forming the contour and prepares the material for the next restrictive square pass.

It's a sequence of shaping and containing.

So designing a rail mill requires significantly more experience and calculation than designing a standard sheet mill.

Oh, absolutely.

The design of these grooved roll passes is highly empirical and specialized.

The same set of roll parameters won't work for different alloys or different temperatures or even slightly different dimensions.

So it's an art as much as a science.

It is an area where engineering experience, combined with computational analysis of metal flow, is absolutely paramount.

All right, let's delve into the mechanics of the roll gap itself.

Before we can even calculate the loads, we have to understand the fundamental geometric relationships that are governing the flow.

We begin with the law of conservation of mass or volume continuity, which is represented by equation 17 to 1.

Okay, let's hear it.

It's b h zero v zero equals b h v equals b h f v f.

Let's define those terms clearly for everyone listening.

Sure.

b is the width of the material, h is the thickness, and v is the velocity.

The subscript zero or not denotes the material entering the roll gap, so that's the initial thickness.

And the subscript f denotes the material exiting, so that's the final thickness.

And the physical takeaway, what does this mean for the sheet itself as it goes through?

Since, particularly in sheet and strip rolling, the width b is approximately constant.

Any reduction in thickness from h zero to h f must be compensated by an increase in velocity.

The material is accelerated.

The exit velocity v f has to be significantly greater than the entrance velocity v zero.

Next up, let's define the geometry of the contact area itself, the arc of contact, which is l p.

L p, from equation 17 to 2, is the projected length of the contact area between the roll and the metal.

It's calculated based on the roll radius, r, and the reduction in thickness, or what we call the draft, delta h.

So, eight zero h f.

Exactly.

The full equation is a bit long, but since the draft is typically small compared to the roll radius, we almost always use a simplified, easily memorable approximation.

That is.

L p is approximately the square root of r times delta h, so r h zero h f to the one half power.

That l p is then used directly and calculated the total contact area, and I assume the total rolling load.

That's right.

It's a foundational geometric value.

Now that we have the geometry, let's look at the forces acting on the material within that roll gap.

The text shows this in figure 17 -5.

We have the vertical radial pressure, p r, and the horizontal tangential frictional force, f.

These forces create a dynamic situation, which is defined by the location of the neutral point, or the no -slip point.

And you said this is arguably the most critical location in the entire rolling process.

It is.

So why is the neutral point so important?

Because it is the only point on the entire arc of contact where the velocity of the sheet, v, exactly equals the surface velocity of the roll, roll.

Since the velocity of the sheet increases continuously from v over v f, the neutral point has to lie somewhere along that path.

And how does that location define the frictional forces?

It dictates the direction of friction.

On the entrance side of the neutral point, the sheet is moving slower than the roll, so friction acts to accelerate the metal, pulling it into the gap.

Okay.

But on the exit side, the sheet is now moving faster than the roll surface, so friction reverses and acts to decelerate the metal.

The net effect of that friction determines the total load and the required torque.

And the ultimate output for mill design is the rolling load, p.

That's the big one.

p is the total vertical component of that radial pressure, p r.

This force is the entire load trying to push the rolls apart.

And it's the force the massive mill housing must physically withstand.

And we also define a specific roll pressure, p.

Right.

From equation 17 .3, that's just the rolling load distributed over the contact area.

So p equals p divided by b times l p.

Let's try to visualize those forces.

If we plot the pressure along that arc of contact, as in figure 17 to 6, we get the famous friction hill.

Can you describe the shape of this curve for us?

Sure.

So imagine a graph.

The horizontal axis is the length of contact from the entrance to the exit.

The vertical axis is the pressure.

Got it.

The pressure starts low at the entrance, it rises gradually, and then it peaks exactly at the neutral point.

After that peak, it falls off again toward the exit.

It's not a uniform pressure, this hill in the middle.

Exactly.

And the total area under this curve multiplied by the width gives you the total rolling load, p.

What's the profound significance of that friction hill?

Why does it form?

It means the compressive force you have to apply must be significantly higher than the theoretical yield stress of the metal itself.

Friction essentially traps the metal in the center, and the required pressure increases exponentially because the frictional forces restrict the material's flow.

And the peak location?

The location of the peak is critical because the center of gravity of that pressure distribution determines where the resultant force acts, and that feeds directly into our power and torque calculations later on.

Before any of this can happen, the metal has to actually enter the roll gap.

This relies on what's called the angle of bite, or alpha.

The angle of bite, alpha, is the angle between the roll center line and the point where the metal first contacts the roll surface.

And for the metal to enter on its own without being pushed.

The horizontal component of the tangential frictional force, which pulls it in, must be sufficient to overcome the horizontal component of the normal force, which is trying to push the metal back out.

And that gives us the limiting condition for unaided entry, which is equation 17 to 4.

Knowing that the frictional force is must times PR, we can simplify that whole condition down to a really elegant, powerful relationship.

Which is?

Mo must be greater than or equal to the tangent of alpha.

So, EOT.

That's it.

The coefficient of friction, mu, has to be big enough.

If this condition isn't met, the rolls will just slip and the material will not be drawn in.

So, if an engineer is designing a mill for maximum reduction, what does this tell them about their choice of rolls?

It tells them that for large reductions, the angle of bite, alpha, will be large, which requires high friction mu.

Okay.

On the other hand, if they're trying to roll a very thin slab, or they're using a small draft, they want to reduce alpha.

Since alpha depends on both the draft and the roll radius R, using smaller diameter rolls dramatically reduces the angle of bite for the same draft.

Making it easier to meet that mu is greater than tan alpha condition.

Exactly.

Let's quantify that relationship by calculating the theoretical maximum possible reduction in a single pass.

That's the draft, delta H, in equation 17 to 5.

We use the geometric approximation for the tangent of alpha, which is roughly the square root of delta H over R.

Then we set the entry condition to its limit, mu tan alpha, square both sides, and solve for delta H.

And we arrive at?

We arrive at the maximum delta H is approximately mu squared times R.

Okay.

R is linear, but mu is squared.

That squared relationship on friction seems like the absolute core takeaway for limits of reduction.

It defines everything.

If you double the coefficient of friction, you quadruple the maximum draft you can take in one pass.

This explains the entire difference between hot rolling and cold rolling capability.

Let's run the numbers from the source material to really crystallize that point.

Let's assume a standard roll radius of, say, 300 millimeters.

Okay.

For cold rolling, where we use lubrication to keep the surface finished pristine, friction is low, maybe mu is 0 .08.

Right.

The maximum draft is then 0 .08, two times 300 millimeters, which gives us a maximum delta H of only 1 .92 millimeters.

A relatively small bite.

Now, switch to hot rolling where the material is soft and the friction is high, maybe mu is 0 .5.

Now, the maximum delta H is 0 .52 times 300 millimeters, which gives us 75 millimeters.

75 millimeters versus less than 2.

Exactly.

This comparison powerfully illustrates why large diameter rolls are used for initial heavy breakdown in hot mills.

They maximize R, but the high mu is the real force multiplier for reduction.

That's the theoretical limit.

Assuming the rolls are perfectly rigid, but we know they are not.

They are elastic structures under enormous load.

This brings us to the necessity of roll flattening and using Hitchcock's correction.

Roll flattening is the elastic deformation of the rolls under that high rolling load.

This deformation makes the roll's contact profile flatter, which effectively increases its radius from the nominal radius R to a deformed radius R.

So a bigger R increases the arc of contact, LP.

Yes, meaning the load is spread over a wider area, which slightly lowers the specific pressure P.

How does Hitchcock's correction, equation 17 to 6, incorporate this?

The corrected effective radius R is given by R equal R1 plus CPE.

Okay, let's break that down.

R is the original radius, B is the width, delta H is the draft, and P is the rolling load calculated using the original R.

Critically, C is the elastic constant, which is a known value related to the Young's modulus and Poisson's ratio of the roll material.

It tells us precisely how much the roll will deflect elastically under a given load.

Since P depends on R and R depends on P, this immediately screams iteration.

You have to do it more than once.

That's the key procedural requirement.

You must start by estimating the load P using the nominal radius R.

Then you plug that load P into Hitchcock's equation to calculate the corrected radius R.

And then you do it all again with a new R.

Exactly.

You use this new R to recalculate the rolling load P.

You repeat this iterative procedure until the load values converge.

It's a necessary complication for accurate load prediction, especially in cold rolling where loads are massive.

We've established the geometry and the critical role of friction.

Now let's turn to calculating the actual force required to operate the mill.

We know the load P depends on four main factors.

Roll size, the metal's resistance to deformation, friction, and tension.

That's right.

Let's start with a simplified case just to build the concept.

What if we completely ignored friction?

As a conceptual starting point, if we neglect friction, the process simplifies greatly.

The specific pressure P would simply be equal to the plane strain yield stress, sigma bar naught prime.

So the load P would just be that stress times the contact area.

Essentially, yes, P would be approximated by equation 17 to 7, which is Pb psi.

But this is only a theoretical minimum.

We have to account for that friction hill.

To incorporate the friction hill, we look at the mean deformation pressure, P bar, relative to the mean flow stress, sigma bar naught.

This brings in the crucial friction hill factor, Q.

Yes, equation 17 to 8 shows the core mechanical principle of the friction hill.

The ratio of the pressures is equal to 1Q eq1.

Okay, and Q itself is calculated as eLp.

Yes.

Correct, where h bar is the mean thickness of the material in the gap.

That exponential term, e to the power of Q, is what we really need to focus on.

What is the physical meaning of that exponential factor in terms of the load?

The term eq1 is essentially a friction multiplier.

It shows that as friction, mu, and the ratio of contact length to mean thickness, Lp over h bar, increase, the pressure required to deform the metal increases exponentially.

So it's not a linear increase.

Not at all.

A small increase in friction or contact length results in a rapid, much larger increase in the load P.

When we combine the geometry, the flow stress, and this friction hill, we get the full simplified equation for the total rolling load P, which is equation 17 to 9.

Yes, POU 3 1Q 1Q, play Steve.

What's that 2 over root 3 factor for?

That's the necessary correction required to convert the simple flow stress into the effective plane strain flow stress, based on the Von Mises yield criterion.

This total load P is what the mill stands must support.

Let's apply this massive equation using the cold rolling example provided in the text.

We're rolling a 40 millimeter slab down to 28 millimeters.

The roll diameter is 900 millimeters, so R is 450.

Friction mu is 0 .30,

and the mean flow stress is 170 MPa.

Okay,

step one, geometry.

The draft, delta H, is 12 millimeters.

The mean thickness, h bar, is 34 millimeters.

Step two, arc of contact, Lp.

That's the square root of R delta H, which comes out to 73 .48 millimeters.

Step three, friction hill factor Q.

That's Vlph, which is about 0 .65.

And now step four, we calculate the initial rolling load P.

Using the full equation and substituting all those values, the initial load calculation gives us about 13 .4 mega Newtons.

13 .4 mm, that is an incomprehensibly large number.

To give you some context, that is the weight equivalent of over 1300 metric tons.

Wow.

Or roughly the takeoff rate of a fully loaded small jumbo jet.

That is the force being exerted on those rolls.

That is a staggering force.

And that load must inherently cause the rolls to flatten.

So this is where we bring back that necessary iterative check for roll flattening.

Correct.

The mill stand cannot ignore that 13 .4 mega Newtons.

We have to now use that calculated load P to find the new effective deform radius R using Hitchcock's correction.

And the example shows that when that load is plugged in, the radius of 450 millimeters increases slightly to R equals 464 millimeters.

We then repeat the whole process.

Using this new larger R, we calculate a new LP, a new Q, and a new load P.

The revised load calculated in the text is 13 .7 mega Newtons.

That's pretty close.

Very close.

Since 13 .7 is very close to the assumed load of 13 .4, the calculation is assumed to have converged.

And 13 .7 mega Newtons is the force the mill stand has to be engineered to withstand.

The key takeaway is procedural.

Accurate load calculation is almost never a single step process.

Never.

We've established that friction makes the load skyrocket.

But modern tandem mills use external forces front tension pulling the material out and back tension pulling the material in.

How does applying these horizontal tensile forces affect that massive vertical rolling load P?

It is the engineer's best trick for reducing the required load.

Applying horizontal tension front tension sigma and back tension sigma B effectively reduces the metal's internal resistance to deformation.

So the material yields easier.

Exactly.

Equation 17 .13 shows that the required specific roll pressure P is directly reduced by the average horizontal strip tension sigma H dot P equals two three.

So that's a massive economic benefit.

Huge.

Because lowering P means less stress on the mill components, lower power consumption, and less mill spring.

If we look at the pressure distribution curve again, figure 17 to seven,

what does adding tension do to the shape of that friction hill?

Without tension, you have the standard high friction hill curve.

When you introduce balanced tension, the entire curve shrinks dramatically.

The peak pressure decreases and the total load, the area under the curve, is much much smaller.

So tension is like partially pre -stressing the material to help it yield.

That's a great way to think about it.

Does tension also affect the location of the neutral point?

Yes.

And this is where control gets really interesting.

Applying significant back tension is highly effective at reducing the load and it shifts the neutral point toward the exit plane.

If back tension is too high, the neutral point can be pushed right to the exit, leading to instability or sliding.

Conversely, high front tension shifts the neutral point toward the entrance plane.

So by precisely controlling both, engineers can manipulate the friction behavior within the gap.

Yes, and this precise control allows them to measure friction using forward slip or SF.

How's that?

Forward slip from equation 1711 is the difference between the sheet's exit velocity vf and the roll surface velocity vr, expressed as a fraction of vr.

So vf, vr, vr.

Right.

And since vf is always greater than vr, forward slip is always positive.

The magnitude of the slip is directly related to the position of the neutral point and thus to the coefficient of friction value.

It's a key measurable parameter used in these automatic control systems.

Let's discuss the common problems that result from these massive forces.

We mentioned roll flattening, but the entire mill stand also distorts elastically.

That's mill spring.

Mill spring is the cumulative elastic distortion of the bearings, the screws, and the housing under load.

This means that the actual gap between the rolls when the metal is being rolled is always greater than the mechanical setting under no -load conditions.

And this is what ultimately limits how thin we can roll a material.

Yes, this leads to the concept of minimum thickness, or Fmin.

Why can't we just roll infinitely thin foil?

Because eventually the elastic distortion of the mill, the mill spring, becomes greater than the plastic deformation you are trying to induce in the sheet.

You reach a point where increasing the load simply increases the mill spring without achieving any further reduction in thickness.

And the calculation for Fmin, equation 1714, shows this strong dependence.

It does.

Etchmin is proportional to mu, r, and the flow stress squared.

And it's inversely proportional to the elastic modulus of the rolls.

So to roll thinner sheets, you need a stiffer mill, smaller rolls, and low flow stress material.

That's the formula.

This is why rolling high -strength steel requires much stiffer, more sophisticated mills than rolling aluminum.

Moving to surface issues, flatness defects, shown in figure 17 .9, are a huge industrial problem.

If the roll bends, the center of the strip gets thinner.

Leading to waviness.

Yes, this is roll -bending deflection.

If the center of the strip elongates more than the edges, the center buckles, creating wavy edges or a center buckle.

So how do you counteract this?

To counteract this deflection, engineers often grind the rolls with a slight curvature called a camber or a crown.

This means the center of the roll is thicker than the edges when it's unloaded.

When the load is applied, the roll deflects, and that crown shape ideally flattens out to maintain a uniform gap.

The ideal camber needed is dependent on the rolling load, which is constantly changing.

A fixed camber won't work for all products.

Exactly.

This requires dynamic control.

Modern mills use complex systems like hydraulic roll bending, where hydraulic jacks are used to apply bending moments to the roll ends during the rolling process.

So you can dynamically control the roll's deflection.

And maintain perfect flatness despite changes in load or material properties.

Finally, what causes edge cracking and alligatoring, which we see in figures 1710 and 1711.

This ties directly back to the friction hill.

When the material is compressed, it tries to spread laterally.

Friction strongly resists this spread, particularly at the edges.

So you get inhomogeneous deformation.

That's it.

The metal in the center flows easily in the rolling direction, but the edges are retarded by friction.

This creates secondary tensile stresses at the edges.

And those stresses accumulate over multiple passes.

Yes.

They lead to longitudinal splits or edge cracking.

If the deformation is too severe, the center and edges can separate entirely, leading to alligatoring, where the sheet splits like an open crocodile mouth.

And the prevention.

To prevent this cumulative stress buildup, engineers use vertical edging rolls during the initial breakdown stages to actively restrict the width and force the material to flow in the desired rolling direction.

The dynamic nature of all these factors, load, friction, defects, necessitate sophisticated control systems.

Primarily automatic gauge control, or AGC.

How does an engineer analyze and control the final thickness?

They rely entirely on understanding the mill's characteristic curves, which we see in figure 1712.

We plot the rolling load, P, on the y -axis, versus the final sheet thickness, HFF, on the x -axis.

There are two defining curves on this plot.

First, the elastic curve.

The elastic curve represents the mill stiffness.

It shows the relationship between the final thickness HFF and the load P necessary to achieve that thickness, based on the elastic constant of the mill housing and the screw setting.

The steeper the slope, the stiffer the mill.

Exactly.

And second, the plastic curve.

The plastic curve represents the metal's resistance to deformation.

It shows the load P required to plastically deform the metal from 8 -0 down to HFF.

This curve is highly sensitive to the material's flow stress, temperature, lubrication, and any strip tension.

And the intersection of these two curves is the single crucial operating point.

It is.

The intersection determines the actual final thickness HFF and the actual rolling load, P, for that specific pass.

The beauty of this diagram is that it shows immediately how any parameter change will shift that operating point.

So let's consider a failure scenario, like in figure 1713.

Suppose the lubrication fails, increasing the friction.

How does the AGC respond?

Increased friction or a drop in temperature causes the flow stress to increase.

This shifts the plastic curve up and slightly to the left, meaning it requires a higher load P to achieve the same reduction.

Since the elastic curve, the mill stiffness, is constant, the intersection point shifts, the load P increases dramatically, and the final thickness HFF also increases.

The sheet gets thicker.

So the AGC mechanism is all about shifting those curves back into alignment.

Yes.

The mill measures the error signal.

The difference between the measured HF and the desired HF, using X -ray or isotope gauges, this signal feeds back to the actuators.

And what does it do?

To correct that sudden thickening, the AGC might immediately adjust the roll gap, which shifts the elastic curve to the left to increase the load and drive HF back down.

Or it might increase strip tension, which shifts the plastic curve down and right, achieving the target thickness faster.

Is one method preferred?

Control via strip tension is often preferred because it's mechanically quicker than adjusting massive roll screws.

Now we enter the theoretical framework developed to accurately model these loads.

The goal is complex.

Express the rolling load and torque solely in terms of material properties and geometry.

This started with von Karman's equation.

Von Karman's work is the fundamental differential equation that describes the equilibrium of an elemental strip of material within the roll gap.

It's the mathematical starting point for all modern rolling models.

To solve such a complicated differential equation,

significant assumptions must be made, which are the seven most critical assumptions for the classical cold rolling theory.

They are simplifications designed to make the math tractable.

First, the arc of contact is perfectly circular, so we ignore roll flattening initially.

Second, the coefficient of friction, mu, is constant along the entire arc.

Third, we assume no lateral spread, making it a simplified plane strain problem.

Fourth,

plane vertical sections remain plane -assuming homogeneous deformation.

Fifth, the roll peripheral velocity is constant.

Sixth, the elastic deformation of the sheet itself is negligible.

And seventh, the material follows the von Mises distortion energy criterion for yielding.

These assumptions allow for a complex integration.

Later work by theorists like Bland and Ford simplified the final calculation by integrating von Karman's equation, yielding pressure relationships that factor in front and back tension.

Right.

The ultimate result is equation 1721, where the total rolling load p is found by integrating that calculated specific roll pressure p over the entire arc of contact multiplied by the width.

This is the standard procedure for high accuracy load predictions in computationally aided engineering.

Hot rolling analysis is even more complex, I assume, because the material's properties are changing dynamically during the roll.

Hot rolling is complicated by the high potential for sticking friction, where the material actually adheres to the roll surface and slides internally.

And critically, the fact that the material's flow stress is highly dependent on both temperature and strain rate.

So the material is softening as it deforms, but also hardening due to the speed.

Exactly.

It's a constant battle.

So we need to calculate the strain rate, epsilon dot, to find the correct flow stress.

We use the mean strain rate from equation 1723 for analytical purposes.

It's Vr h ln h Euro h.

Where Vr is the roll velocity.

Right.

This epsilon dot allows us to determine the appropriate flow stress for the material at that speed and temperature, which is essential input for the load calculation.

For practical engineering, the text relies on Sim's rolling load equation, which is equation 1727, a simplification of Oroland's more rigorous analysis.

Sim's equation provides a vital shortcut for hot rolling.

It's p h a q p.

Okay.

So sigma dot prime is the mean flow stress and the complex geometry and friction factors are all bundled into that single parameter q p.

That's it.

q p is the key engineering factor that translates the complexity of hot metal behavior into a usable number.

So how do engineers determine q p?

They use graphical methods like figure 1715.

This is a family of curves plotted with the reduction percentage r on the x -axis and the values of q p on the y -axis.

And the different curves.

The curves are parameterized by the geometric ratio r h f, which is the roll radius divided by the final thickness.

Can you walk us through the process of using this chart?

Sure.

First, you have to calculate two inputs,

your required percentage reduction r and that geometric ratio r h f.

Okay.

You locate the point on the x -axis corresponding to your reduction.

You then find the curve that matches your calculated r h f ratio.

You trace horizontally from that curve over to the y -axis to read off the required value for q p.

And what does the shape of those curves tell us about hot rolling load?

They show that q p increases rapidly with both higher reduction and higher r h f ratios.

Since p is directly proportional to q p, this just confirms that trying to achieve massive reductions, especially with large diameter rolls rolling thin strip, creates astronomically high forces.

So it quantifies that friction hill effect in hot rolling.

We've established the raw force needed, p.

Now let's calculate the energy required to drive these rolls against that force.

Torque and power.

Where is the massive energy expenditure in a mill?

Power is expended primarily in two areas.

One,

the necessary energy to plastically deform the metal.

And two, the energy lost overcoming friction in the rolls, the bearings, the drive systems, and of course electrical losses.

And the biggest chunks are deformation and friction.

By far.

To calculate the torque, we need to know where that total rolling load p actually acts.

We simplify the complex pressure distribution by assuming the total rolling load p is concentrated at a single point located by the distance a, the moment arm, from the line of the roll center.

And we normalize this position using the moment arm ratio lambda.

Yes, lambda equals a divided by l p, the projected arc of contact.

This ratio lambda is typically approximated as 0 .5 for hot rolling and about 0 .45 for cold rolling.

Why the difference?

It reflects the shift in the pressure distribution center of gravity due to lubrication and cold rolling.

With the moment arm a, the total torque applied to the two rolls, m t, is straightforward.

Torque, from equation 1730, is just the total force times the moment arm.

And since the force p acts on both rolls, the total torque is m t equals 2 p a.

And power, w, is the rate of doing work so, torque times rotational speed.

The total power required, from equation 1732, is calculated as w is 4 a p n.

Where p is in newtons, a is the moment arm in meters, and n is the rotational speed in revolutions per second, or hertz.

And this calculation directly informs the size and capacity of the drive motors required for the mill stand.

Let's apply sims in the power equation to a hot rolling problem.

We have a 300 millimeter wide aluminum strip reduced from 20 millimeters down to 15 millimeters.

The roll diameter is one meter, so r is 500 millimeters, and it's running at 100 rpm.

We're also given the flow stress relationship.

It can cause 140 .2 mpa.

Step one, calculate the true strain epsilon one.

That's natural log of 20 over 15, which is about 0 .288.

Step two, determine the QP chart inputs.

The reduction r is 5 over 20, so that's 0 .25, or 25 percent.

And the geometric ratio rhf is 500 divided by 15, which is about 33 .3.

Step three, we calculate the mean flow stress sigma not prime.

This involves integrating the given flow stress function over that calculated strain range.

The result is about 91 mpa.

Step four, we go to the QP chart, figure 1715.

With r at 25 percent and rhf at 33 .3, we interpolate and read QP is approximately 1 .5.

Step five, calculate the rolling load, p using sims equation.

Plugging in all our values, the total load comes out to p is approximately 2 .36 meganewtons.

That's 2 .36 meganewtons, a significant force, roughly 240 tons of pushing power required for that single stand.

Step six, power calculation.

We need the moment arm a and the rotational speed n.

For the moment arm a, we use the hot rolling moment arm ratio, lambda equals 0 .5.

The arc of contact Lp is about 0 .05 meters.

That's a is half of that, so about 0 .025 meters.

And the speed?

The rotational speed n is 100 rpm divided by 60, which is about 1 .67 revolutions per second.

Now, power.

W is 4 apn.

W equals 4 p, room 0 .025, 2 .36106, 1 .67.

The required power is 1 .24 megawatts.

Wow.

1 .24 megawatts.

To put that into perspective, that is the equivalent electrical power required to simultaneously run approximately 800 average North American homes.

That is the sheer energy density required to plastically deform metal at a continuous rate.

We have covered a tremendous amount of ground here, from initial casting terminology, all the way to the advanced computational analysis required to size a motor.

We have, and for any engineer engaging with rolling processes, I think three mechanical metallurgy takeaways are really non -negotiable.

Okay, let's hear them.

First, the physical limits of reduction are defined by that frictional condition for entry.

The maximum draft is approximately mus squared times r.

So the maximum reduction is exponentially dependent on friction and linearly on roll size.

If you want more draft, you need more friction or bigger rolls.

Second, calculating the total rolling load, p, is inherently complex and it's iterative.

You have to account for the friction hill factor, q or qp, that exponentially increases the pressure.

And you must perform the iterative check for roll flattening, that r correction, to ensure your load estimate is accurate.

And finally, control is dynamic.

Mill operation depends entirely on understanding the characteristic curves.

That intersection of the stiff, fixed elastic curve, the mill spring and the flexible plastic curve, the metal's resistance.

That's right.

Automatic gauge control is just a mechanism for shifting those two curves to maintain the target intersection point.

It's easy to be overwhelmed by the complexity of Sims or von Karman.

But consider this.

The first continuous tandem rolling mills, which needed flawless synchronization, iterative load management and dynamic AGC, were conceived and engineered using these very principles.

All calculated painstakingly, long before computers could solve these exponential equations in real time.

It's a remarkable testament to the power of applied physics and the depth of understanding required to manipulate materials on an industrial scale.

Indeed.

Mastering these core concepts is what transitions you from simply reading a gauge to truly understanding the physical limits and economic drivers of metal production.

Thank you for diving deep with us into the demanding physics of metal rolling.

We hope this deep dive provides the clarity you need to master this topic quickly and thoroughly.

We'll catch you next time.