Study Support - vsb.czkatedry.fmmi.vsb.cz/Opory_FMMI_ENG/MMT... · Thermo-mechanical forming...

Thermo-mechanical forming processes

Study Support

doc. Ing. Radim Kocich, Ph.D.

Ostrava 2015

VYSOKÁ ŠKOLA BÁŇSKÁ – TECHNICKÁ UNIVERZITA OSTRAVA

FAKULTA METALURGIE A MATERIÁLOVÉHO INŽENÝRSTVÍ


2

Title: Thermo-mechanical forming processes

Code:

Author: doc. Ing. Radim Kocich, Ph.D.

Edition: first, 2015

Number of pages:58

Academic materials for the Progressive technical materials study programme at

the Faculty of Metallurgy and Materials Engineering.

Proofreading has not been performed.

Execution: VŠB - Technical University of Ostrava


3

1. Thermo-mechanical forming processes (controlled forming)

Thermo-mechanical forming controls all the manufacturing conditions with the aim to

achieve the required state of structure. It primarily includes composition of steel and

thermo-mechanical and time parameters of forming and possible cooling. Suitable

structural states enable to achieve higher mechanical properties while maintaining

favourable brittleness, fatigue characteristics and weldability. From the structural point

of view, the most significant factors are grain and sub-grain size, content of the

individual structural components, amount and dispersity of precipitates, dislocations

density and structure of solid solution.

The basic presupposition for achieving high strength properties is an increased

dislocations density; movement of dislocations during an external loading is slowed by

internal barriers. The character of these barriers and their distribution is especially

important for toughness properties, since these enable to relax peak tension stresses by

dislocations movement before the stresses exceed the strength of the material. Barriers

with a low permeability include high-angle grain boundaries, non-plastic structural

particles and non-coherent matrices including Lomer-Cottrel barriers. The opposite

properties have low- or mid-angle grain boundaries, plastic particles and coherent

matrices.

Time to study: 20 hours

Aims After study of this chapter you will be able to

define current methods of thermo-mechanical forming and its basic

mechanisms and division

describe low and high temperature thermo-mechanical processing,

isoforming, deformation aging, forming at low temperatures, SHT –

process, direct rolling, rolling with hot batch


4

describe basic types of controlled rolling, define controlled cooling

determine mechanical properties of control-formed material

define basic physical metallurgical processes in controlled rolling

Lecture

1.1. Methods of thermo-mechanical processing of steel

The amount of steel processed thermo-mechanically per year is increasing. It can be

assumed that a percentage increase of only 10% (approx. 500kt/year) in production

with the use of thermo-mechanical methods would create a financial profit of approx.

200 million CZK for the Czech Republic. (Short note: Modern production lines

technically enable application of direct thermo-mechanical processing. However, long-

term conceptual directed research would be needed so that modern literature-based

knowledge, experience from conferences and plastometric and semi-operational

(laboratory) experimental results would be fully applicable.) Other advantages of

thermo-mechanical forming (TF) will be further mentioned also for aluminium and

other non-ferrous metals.

In the future, it seems essential to expand these methods of processing. There are

several factors that promote such methods. First is the increase in capability and

decrease in cost of computers of all kinds, which means simulations can be performed

on standard PCs. Second factor is ever deeper understanding of specific phenomena of

TF which enables not only a quantitative description of a process, but also to simulate

forming operations on common computers. The third factor is a gradual application of

TF principles (and also processes which lead to a controlled microstructure

development) in an ever increasing number of commercially made products. A directed

algorithm is finally becoming a successful process for highly specific products and is

expanding its application to more common materials.

As has been stated earlier, TF enables to achieve a specific pre-defined microstructure

that imparts specific mechanical and physical properties. This path is different from


5

traditional forming processes. TF requires control and interaction of at least these

mechanisms (among others):

dislocation yield and creep

recrystallization

grain growth

phase transformations

precipitation

particle thickening

Most of these structural changes appear dynamically, i.e. during the course of

deformation or statically, i.e. after deformation.

Each of the above mentioned parameters can be studied individually even if the basic

“rules” are not entirely clear. Furthermore, during TF, synthesis of these basic

parameters occurs as well as their mutual interaction. Especially in the area of large

deformations, relatively high temperatures and high deformation speeds, these

phenomena truly are a great unknown so far.

TF can be divided into several individual methods. One of the options could be the

following:

Grain refinement for increase of strength and toughness without any other

further thermal processing.

Grain refinement for achieving superplastic behaviour.

Texture control (minimization of peaks, improved ductility, magnetic

properties …).

Inside each of these categories are processes, which must be solved concurrently. It

requires laboratory simulation as well as computer modelling.

For TMP, two types of steel are usually used:

steels with phase changes (including martensitic),

steels with precipitation of precipitates from oversaturated solid solution


6

(dispersion strengthening)

1.1.1. Current TMP methods

a) forming before austenite phase change – low- and high-temperature TMP

b) forming during austenite phase change – isoforming

c) forming after austenite phase change – deformation aging, low-temperature forming

Low Temperature TMP (LTMP)

LTMP is based on austenitization with rapid cooling into the range of high stability of

metastable austenite, forming with high degree of deformation ( over 0.6) and

quenching into martensite. Quenching is usually followed by tempering at a lower

temperature. In steels for LTMP, sufficient stability of overcooled austenite is required.

This makes chrome and molybdenum alloyed steels suitable candidates. The main

advantage of LTMP is an increased strength at sufficient plasticity.

High-temperature TMP (HTMP)

The principle is austenitization, forming just over Ar3 with subsequent quenching and

tempering at lower temperatures. It is important that large deformations of austenite

and subsequent quenching occur without recrystallization of austenite. For HTMP, steels

with slow progress of recrystallization are suitable.

In comparison with LTMP, HTMP is technologically simpler, deformation can be

performed in the temperature range between 1 000 to 800 °C. HTMP does not allow

such increase in strength as LTMP, but is does increase plasticity and fatigue resistance.

Isoforming (TBP)

Is based on austenitization and cooling down to the temperature of the perlite nose in

the IRA diagram. Steel is deformed during the entire transformation and subsequently

cooled on air. Tempering is not required. The resulting structure is characterized by

formation of fine sub-grains of ferrite and fine globular carbides. Deformation

temperatures are usually somewhere between 600-700°C, this is sufficient for the


7

process of polygonization. The deformation is over 0.6. TBP does not significantly

increase the strength properties. However, its main advantage is an increase in

toughness and decrease in transit temperatures.

Deformation aging (DA)

It is processing of either conventionally treated or LTMP processed steel, which is then

formed at a temperature at which aging occurs. For example, a processing with = 0.02

deformation and temperature between 150 and 200 °C. It increases strength, but at the

same time plasticity and impact strength are decreased.

Low temperature forming (LTF)

The initial state is either a state after quenching or a state after annealing. Deformations

after quenching help to achieve strength of up to 3130 MPa.

The aforementioned TMP methods are mostly used for the common types of steel. These

methods do not sufficiently utilize control of structure development through slowed

recrystallization and precipitation kinetics. Nevertheless, newly developed and

introduced methods of controlled rolling of micro-alloyed steels are utilized to make

ferrite grains finer through a transformation from deformed austenite. Slowed

recrystallization of austenite is achieved through the combined effect of micro-alloying

elements in a solid solution and the interaction of controlled precipitation.

1.1.2. Basic types of controlled rolling (CR)

rolling in the lower region of austenite (in two or three stages)

rolling in the region of two-phase structure

method a) complemented by controlled cooling

controlled rolling with quenching and tempering

special methods of CR - for example the SHT (Sumitomo High Toughness) method


8

A schematic development of structures during controlled rolling methods can be seen in

figure 1.

Figure 1: A schematic of microstructure development during rolling procedures.

1.2. Properties and characteristics of structure during thermo-

mechanical processing

The main factors that influence yield strength and transition temperature are:

strengthening of solid solution through substitution and interstitial elements

size of ferrite grain

precipitates in ferrite,

content of perlite and its inter-lamellar distance

Peierls-Nabarro stress ( 40 MPa)

During the transformation of from recrystallized austenite, it is possible to

express the yield strength K using the Hall-Petch equation (1).

2/1

0 . dkK (1)

where 0 is internal stress, d is grain size, ky is coefficient expressing the theoretical

value of barrier effect of grain boundaries against movement of dislocations

[MPa.mml/2].


9

The value of 0 consists of several components and for the transformation of from

non-recrystallized austenite can be expressed by the following relation:

PNPERPRZINS 0 (2)

If the deformation is ended in the + dual-phase area, the final structure consists of

recrystallized ferrite grains and subgrains. Since all the grain boundaries in such a

structure restrict the movement, the d value has to be replaced with the de effective

grain size value (Eq. 3).

rcrre fdfdd

1..111

(3)

where de-1 is length of boundaries for a unit length, dr-1 is length of high-angle

boundaries, dc-1 is length of low-angle boundaries, fr is volume fraction of recrystallized

ferrite.

1.2.1. Misorientation of grain boundary angle, types of grain boundaries

One of the first steps during TMP is to modify a coarse as-cast structure. One of the

aspects according to which structure – grain boundaries – can be distinguished is the

character of their rotation and/or tilt (also called misorientation). Several basic types

of grain boundaries can be observed (Figure 2).

Figure 2: Possible types of grain boundaries and their energies.

A tilted boundary with a low misorientation angle consists of a field of parallel edge

dislocations. A higher amount of dislocations causes higher misorientation angle and a

consequent increase in energy. Nevertheless, the speed of the increase decreases at

higher energy or amount of dislocations due to mutual annihilation of dislocation fields.

Besides the misorientation angle and distance between dislocations, the dependency of


10

grain boundary energy on misorientation angle θ can also be disturbed by dislocations

cores overlapping. This then creates a transfer between low-angle grain boundaries

(LAGBs) and high-angle grain boundaries (HAGBs).

Special boundaries are locations in which two crystals overlap with a relatively small

curvature of inter-atomic layers. This phenomenon is also the basis of the CSL

(coincident site lattice) model, which in principle describes overlapping of selected

locations in lattices. For example, if every 5th lattice point is overlapped, then the

boundary is denoted as Σ5 etc. Except several exceptions the predicted limits for cubic

lattices are Σ29 or Σ33. Nevertheless, the CLS index only rarely depicts the exact

boundary. The CSL model is also valid only for 2D systems, while in real conditions the

material system is 3D. Another fact is that the data were derived for cubic systems,

whereas for non-cubic systems the situation is more complicated.

1.2.2. Deformation bands

Plastic deformation is influenced by changes of the stress state imparted by interactions

of dislocations with grain boundaries and obstacles. Actually, each individual grain

within the material deforms in an entirely heterogenous way influenced by the

neighboring grains being deformed. This was proven by microhardness measurements

resulting into different values for individual grains.

The final deformation substructure is dependent not only on the temperature and way

of deformation, but also on grains orientations and local deformations. This means that

the tensors of deformation can differ inside one single grain due to accommodation of

the grain to different properties of its neighborhoods (Figure 3). With progressing

deformation, especially during deformation under cold conditions, heterogeneity of slip

in grains can increase due to misorientations, which tend to develop between individual

deformation bands. By this reason the grains subsequently divide into domains (cell

blocks) and deformation bands.


11

Figure 3: Various types of deformation inhomogeneities; a) slip bands, b) deformation

bands, c) dislocation boundaries, d) dislocations.

However, the term deformation bands is also sometimes used for coarse bands evident

in some big grains caused by divergences of lattice rotations. Neighboring bands tilt

away from each other in different directions and become subsequently divided by zones

of micrometers sizes with high orientation gradients called transition bands.

Cell bands are divided by walls, which increase misorientation only slowly. These can

be called as dislocation boundaries, geometrically necessary boundaries (GNBs) or

rotational walls.

Dislocation cells and dislocation walls are in principle transformation characteristics

which continuously appear and disappear during deformation due to random

interactions of dislocations. Such boundaries are denoted as random (incidental)

dislocation boundaries. On the other hand, walls of cell blocks and transition bands are

permanent characteristics of a deformed substructure resulting from plastic

deformation.

1.2.3. Shear bands

Bands of localized shear can develop in strongly deformed metals within several scales,

from macroscopic, i.e. in the range of centimeters, to microscopic, i.e. in the range of

micrometers. When a decrease in hardening rate during plastic flow occurs, plastic

deformation starts to localize into shear bands developing from originally wide diffusion

bands. Due to progressing deformation these bands rapidly densify into localized zones

of intensive shear, which consequently leads to failure. Before the occurrence of


12

macroscopic bands shear bands can occur within a grain and subsequently develop

through several grains. Alloys hardened with fine precipitates, fine twins or alloys with

high dislocation density are especially prone to development of this type of shear bands.

1.2.4. Structure development

Significantly different conditions for structure development of steel occur with different

deformation temperatures.

Region I – recrystallization

The original grain size is a function of temperature and material metallurgical character.

Refinement of austenite grains is given by the particular deformation-recrystallization

cycle. Nuclei for the transformation occur only on boundaries of austenite grains.

The final size of ferrite grains is limited by a certain limit value and its further decrease

beyond this limit using the deformation-recrystallization cycle within region I is not

possible. Eventually, relatively coarse ferrite grains develop after transformation.

Region II – without recrystallization

Forming within region I results into achievement of a limit grain size. Further grain

refinement can be achieved by forming in region II. Austenite grains elongate as a

consequence of restricted recrystallization and bands with higher dislocation density

and higher inner energy and instability develop inside them. Ferrite nuclei develop not

only on austenite grain boundaries, but also within deformation bands – deformation

bands have an effect similar to the effect of grain boundaries during the

transformation.

Region III – austenite–ferrite

Further grain refinement can be achieved by forming in region III (dual-phase). Ferrite

grains deformed after the occurrence of transformation cannot recrystallize

anymore and subgrains occur during subsequent recovery. A decreased solubility of Mn

and V in ferrite after the transformation fastens stress-induced precipitation,

which contributes to pile-up of dislocations and sub-boundaries. Grain growth is also

suppressed in the dual-phase region. The result of forming in region III is therefore a

mixed structure consisting of equiaxed ferrite grains, grains not deformed after the


13

transformation, grains with a lower dislocation density (soft) and subgrains with a

higher dislocation density (hard).

1.3. Structural features during controlled forming

1.3.1. Dissolution of carbides and nitrides, grain growth

One of the aims of heating to a forming temperature is to dissolve the largest possible

amount of precipitates into a solid solution. At the same time, conditions for critical

grain growth must not be reached. Solubility of precipitates (nitrides and carbides) can

be determined using corresponding equations and can also be influenced by other

elements present in steels, especially Mn and Si. Si decreases solubility of Ti in steels.

Titanium bonded in stable compounds does not influence solubility of carbides. It is

therefore necessary to consider only free Ti.

Grain growth is influenced by the amount of impurities and alloying elements present in

solid solution and by second phase particles. These reduce grain areas and pin grain

boundaries. Increasing amount of Nb and Ti in steels increases critical temperature for

grain growth. Vanadium does not restrict grain growth so significantly as Ti and Nb.

Vanadium carbides are less stable and dissolve at lower temperatures.

1.3.2. Strain hardening during TMP

Strain hardening during controlled forming is controlled by the amount of energy

imposed into a material as a result of plastic deformation. It is a consequence of an

increase in stress caused by dislocation movement more or less restricted by obstacles

(precipitates, grain boundaries etc.), which they have to surpass during their movement.

Stacking fault energy of a deformed material defines the possibility of dislocations to be

dissociated into partial dislocations and stacking faults within slip planes. A low stacking

fault energy (lower than 50 mJ/m2) usually results into dissociation of full dislocations

into wide partial dislocations. This subsequently limits their movement only to slip

planes and alternative dislocation movements, such as localized cross-slip, out of these

planes become extremely difficult (i.e. improbable). At lower strains, these dislocations

usually accumulate in slip planes, interact with dislocations in different slip systems and


14

form planar dislocation structures, which consequently lead to a development of high

localized stress gradients. At higher strains, materials with a low stacking fault energy

usually tend to form microtwins (or possibly localized martensitic transformations) and

shear bands. They are therefore prone to form evident planar faults of crystal lattice or

local orientations. Depending on the stacking fault energy it is possible to divide the

development of deformation stress into the four following stages.

Stage 1:

Movement of dislocations within their slip planes is usually restricted and no mutual

interactions occur. By this reason, the velocity of strain hardening is very low. Plastic

deformation results into rotation of crystals, which leads to their reorientation and

multiple slip, which results into a stronger interaction of dislocations in stage 2. Stage 1

is neglectable for commercial polycrystalline materials since the movement of the first

dislocations is limited by grain boundaries, at which dislocations usually pile-up. The

resistance to overcome these boundaries can be expressed using a modified Hall-Petch

relation (4).

2

1

1

Dk)D( (4)

where is deformation resistance for a very large grain, coefficient k1 is usually 0.7 for

carbon steel, 0.2 to 0.4 for HCP lattice metals, 0.07 to 0.1 for Cu and Al, respectively.

Presence of carbon in a solid solution significantly increases the k coefficient value.

Stage 2:

Mutual interactions of dislocations in various slip systems cause fast multiplication of

dislocations and a subsequent high and approximately constant hardening rate.

Development of 3D fields consisting of dislocation multipoles (Taylor networks) occur in

metals with low stacking fault energies, while dislocation tangles develop and often form

into cellular patterns in metals with high stacking fault energies. Typical strain values

characterizing this stage are 0.05 – 0.2.

Stage 3:

At strain values approximately up to 1 the deformation curve becomes parabolic. Strain

hardening rate decreases progressively down to values approximately an order of


15

magnitude lower than at stage 2. In this stage, the effect of the dislocation multiplication

process is reduced by dislocation annihilation (dynamic recovery due to localized cross-

slip, climb or mutual elimination of segments of opposite signs). Microstructure

development is towards a clearly defined cell substructure consisting of dislocation cell

walls. However, this results into a significant decrease in dislocation density inside cells.

The individual cell walls are at first created by a complex of dislocation tangles, which

subsequently reduce their thicknesses with progressing deformation (individual

dislocations move closer to each other). Cells dimensions reduce during deformation

from several micrometers to tenths of micrometers. At the same time misorientation of

individual cells increases from the original value of 1° up to 3° to 4° and even more.

Stage 4:

At strains higher than 1 many grains disintegrate into bands of various orientations

divided by transition zones and grain boundaries. A formation of lamellar structure

consisting of disoriented microbands parallel to the rolling direction occurs at very high

strains. Strain hardening rate in this stage is not very high, although it remains almost

constant. Therefore, increase in deformation stress can occur at high strains.

Crystal defects, especially dislocations generated by plastic deformation, have high

elastic energies, which are stored in the deformed material, especially in stress fields

surrounding dislocations. Dislocations mobility can among others be influenced by the

amount of dissolved atoms. At moderately elevated temperatures enabling diffusion,

dissolved atoms can start to move and then pin on slowly moving dislocations, by which

very strong interactions become enabled (usually between room temperature and

300°C). Reduced dislocations mobility caused by segregations of dissolved atoms leads

to heterogeneous deformation with the occurrence of Lüders bands and Portevein-

LeChatelier (PLC) effects (serrations on stress-strain curves) (Figure 4).

Except small and medium strains, the deformation is unlikely to be homogenous and

usually leads to localization of strain, typically to the form of bands. This is caused by

stacking fault energy, deformation temperature, method of deformation and others.

Nevertheless, it is necessary to distinguish two groups of deformation inhomogeneities.

The first group consists of deformation bands. These are relatively homogenously

deformed regions inside a grain characterized by a different slip system than the


16

neighboring regions within the grain. The second group is shear bands. These are areas

of very strong localized slip within a grain, as well as through many grains. Both the

groups lead to dissociation of a gran into subgrains, although by different mechanisms.

Figure 4: Stress-strain curve with evident PLC effect (Al-Mg alloy).

1.3.3. Precipitation

Defects caused by plastic deformation support diffusion of micro-alloying elements and

nucleation of precipitates. As a consequence, a quick strain-induced type of precipitation

occurs. Moreover, plastic deformation decreases solubility of micro-alloying elements,

which also fastens precipitation. The influence of plastic deformation (ε = 0.09) on

solubility of Nb(CN) was expressed by Yamamoto (Eq. 5).

TNCNb /657093,114

12.log

(5)

Strain-induced precipitation occurring during controlled rolling is especially important

during softening processes. While coarse non-dissolved particles (>0.1 m) already

present in the structure before deformation concentrate strain in their vicinity and

therefore support recrystallization, strain-induced precipitates cause its significant

delay. Recrystallization is also inhibited by micro-alloying atoms present in solid

solution. The differences in size and electron structure of the micro-alloying atoms and

the matrix support their segregation in stacking faults areas, which consequently


17

changes stacking fault energies. This causes inhibition of dislocation movement and

redistribution of dislocations, which makes nucleation of recrystallization nuclei more

difficult.

The substantial delaying effect of strain-induced precipitates on recrystallization

kinetics can be primarily explained by the fact that precipitation occurs especially on

sub-boundaries of deformed austenite. The most effective inhibitors of grain boundaries

movement during recrystallization are fine strain-induced precipitates with the

diameter smaller than 6nm. Development of recrystallization processes is driven by

particles coarsening rate during deformation. The region of precipitates formation is

usually between 850 and 1000°C. The coarsening rate at the temperature of 1000°C is

significant, while at 950°C it decreases substantially and coarsening is not evident even

at longer dwelling times at the temperature of 850°C. For an effective inhibition of

recrystallization at least 0.02% of Nb in solid solution is necessary.

Precipitation kinetics in austenite and ferrite can generally be described by ARA

diagrams. The T0 temperature is the temperature at which the Nb(CN) and Fe3C

precipitation curves cross each other. Precipitation of Fe3C at temperatures below Ar1

and at the same time above T0 is improbable, since precipitation of Nb(CN) occurs faster

due to its lower free energy when compared to Fe3C. At temperatures below T0 the

diffusion ability of Nb in ferrite as a control mechanism of precipitation on the phase

boundary of γ/α is lower. Therefore, Nb(CN) precipitates non-uniformly throughout the

entire material volume. During forming in the region between the Ar3 and Ar1

temperatures precipitates segregate in parallel layers. Origination of these layers is most

probably in precipitation on γ/α phase boundaries during transformation. Such

precipitates do not significantly increase strength.

Incubation time for static precipitation is in the order of magnitude of 102 s at 950°C. For

strain-induced precipitation this time shortens by one order of magnitude, while it

shortens even by two orders of magnitude for dynamic precipitation.


18

1.3.4. Recovery

Recovery occurs when a non-equilibrium concentration of lattice defects (point and line

defects) is decreased, usually by annealing at an appropriate temperature. Point defects

are already recovered at relatively low temperatures (below 0.3 Tt), i.e. below the

temperatures of TMPs of most materials. Recrystallization dominates at higher

temperatures, although recovery also occurs since kinetics is higher at higher

temperatures.

For metals with higher stacking fault energies (e.g. Al), processes such as dislocation

cross-slip and local annihilation proceed easily, which supports recovery. On the other

hand, for metals with low stacking fault energies (e.g. austenitic steels, -brass), the

recovery process is not much evident before recrystallization. The same phenomenon

applies for alloys with a high amount of atoms dissolved in solid solution reducing

dislocation mobility. Recovery can occur immediately after plastic deformation, but also

during deformation. Nevertheless, recovery does not influence the appearance of

microstructure or crystallographic texture. It influences properties, such as hardness,

dislocation density, size and misorientation of subgrains. However, the changes can only

hardly be detected.

Mechanisms applied during recovery

- annihilation of point defects (vacancies, interstitions) using diffusion e.g. into

dislocations

- mutual annihilations of dislocations (applies for closely located dislocations of

opposite signs or dipoles, for which a low amount of dislocation climb or cross-

slip is needed)

- rearrangement of free dislocations and random dislocations into dislocation

walls or sub-boundaries (polygonization)

- coalescence of sub-boundaries walls during subgrains growth

It is very difficult to distinguish the process of diffusion of dislocation cells into clearly

defined subgrain boundaries from the process of grain growth. This is especially due to

the heterogeneity of dislocation structure in deformed polycrystalline materials. The

late stage of recovery (formation of clearly defined sub-boundaries) is often the first


19

stage of nucleation during recrystallization, which can lead to a fast recrystallization

stopping further recovery.

Structural changes during recovery

Besides the first fast decrease in volume of point defects, the main structural changes

can be divided as follows.

- rearrangement of dislocations into cell structures (for metals with high stacking

fault energies and most of metals formed under hot conditions this proceeds

together with deformation)

- elimination of free dislocations inside cells

- rearrangement of complicated dislocation structure of cell walls into arranged

subgrain boundaries – mostly by annihilation of redundant dislocations and

rearrangement of other dislocations into low-energy configurations (Figure 5)

- subgrains growth (during progressing annealing subgrain growth occurs, since

this leads to decrease in internal energy)

Figure 5: Changes of dislocation structure during recovery – from random dislocation

tangles through cell substructures to subgrains [1].

1.3.4.1 Extended recovery/Continuous recrystallization

Recrystallization is a discontinuous process, during which absorption of

deformed/recovered areas by non-deformed grains via HAGBs movement occurs. It is

possible to suppress discontinuous recrystallization in some cases, which subsequently

leads to a relatively fine and uniform grain structure. This process features homogenous

subgrains growth and is often called extended recovery. It is not clear whether this is


20

caused only by LAGBs movement, or whether HAGBs movement also contributes to this

process.

Both these processes can occur locally (inside individual grains), as well as globally.

- local extended recovery was detected e.g. for Al, low-carbon steels and Zr-based

alloys, for grains with specific orientations, which are usually not affected by local

deformation, i.e. are subjected to a homogenous deformation

- global recovery/continuous recrystallization was detected for dual-phase alloys

and for cases of extreme plastic deformations (severe plastic deformations – SPD

methods)

Secondary phase precipitates can pin and stabilize sub-grains structure. A subsequent

dissolution/coarsening of precipitates can lead to a homogenous subgrains growth or

extended recovery. This phenomenon was originally identified as “in situ” or continuous

recrystallization. SPD processes are often applied to achieve ultra-fine grained (UFG)

structure prepared using continuous recrystallization (static or dynamic). For these

processes a large deformation together with nucleation/dissolution of secondary phases

has a significant effect. The result can be extended recovery/continuous

recrystallization, but also geometrically necessary dynamic recrystallization (see

below). In this case, recovery can make localized deformation easier.

1.3.5 Recrystallization

Primary recrystallization (also called discontinuous recrystallization) is often defined as

a process of nucleation and growth. The driving force for nucleation during

recrystallization is far lower than e.g. during solidification. Every large subgrain or a

relatively organized region inside a deformed grain can be considered as a possible

recrystallization nucleus based only on the driving force or relative differences in the

accumulated energies between the possible nucleus and its surroundings. Whether the

possible nucleus is real (active) depends on its ability to grow, which depends especially

on the presence of growth-facilitating boundaries (i.e. the presence and mobility of

HAGBs, the mobility of which is far larger than the mobility of LAGBs). LAGBs created by

deformation/recovery occurring usually between subgrains feature a very low mobility,


21

whereas boundaries with higher misorientations (10-15°) feature a very high mobility.

This results in a nucleation originating in a rapid growth of a very small minority of sub-

grains, which subsequently evolve into new growing grains.

The first necessary condition is that a subgrain should have (or quickly achieve) a local

misorientation larger than 15°. A very quick growth of a very small amount of subgrains

comparing to a slow growth of the remaining subgrains imparts a heterogeneous

character to this casual type of recrystallization, which is then depicted as nucleation

and growth. Typically, a region without deformation (i.e. without grain boundaries or

misorientations) larger than a certain (critical) size is considered to be a recrystallized

grain. To achieve this critical size (usually in the order of magnitude of micrometers),

nucleation and restricted or only local growth are necessary. It is very difficult to

distinguish whether a possible nucleus is active or not. The difference between an

original subgrain (possible nucleus) and a final recrystallized grain for typical metals is

10 – 100 times the original size. The probability of finding an active nucleus in a

deformed/recrystallized metal matrix is 10-3 - 10-6.

Sources of recrystallized grains

- Deformed grains

In deformed grains, recrystallized grains with more or less similar orientations as the

original grains can develop. Recrystallized grains can however develop also from

deformation bands of similar orientations. Such bands are effective recrystallization

sources due to substantial localized deformations inside the grains. This non-uniform

localization of deformation leads to fragmentations of the bands, as well as to large

inhomogeneities in the values of imposed strain.

- Shear bands

Shear bands passing through several grains can also be sources of recrystallized grains.

This is usually related to a high imposed energy and therefore large variations in relative

misorientations and a possible presence of growth-facilitating boundaries. Shear bands

are typical especially for metals with low stacking fault energies.

- Particle stimulated nucleation (PSN)


22

Dislocations can be pinned by relatively large particles insusceptible to slip. This pinning

and subsequent growth of dislocation density leads to formations of locally deformed

regions with large developments of misorientations around secondary phase particles.

Consequently, recrystallization occurs in these locally deformed regions due to large

differences in the imposed energies. This type of recrystallization is usually depicted as

PSN (particle stimulated nucleation). Locally deformed regions around particles are then

depicted as PSN grains with random orientations.

Low-temperature annealing performed after deformation often results in a highly

random orientation, which is caused by the existence of inner and outer locally

deformed regions, or by the influence of PSN and deformation bands.

One of the mechanisms by which a recrystallized grain with a new orientation can

originate is also recrystallization twins. These occur especially in metals with low

stacking fault energies (Cu, austenitic steel).

As was already mentioned, particles are very important during recrystallization, since

they can pin grain boundaries during their movement. However, there are more types of

the pinning effect.

- Pinning by Zener drag. Low-energy boundaries, such as CSL boundaries, have low

Zener drag pinning effect.

- Pinning using elements dissolved in solid solution (solute drag). The effect is low

for CSL boundaries.

- Orientation pinning. Growing grains can be pinned by locally deformed regions of

similar orientations, i.e. by a presence of LAGBs.

Possible types of recrystallization

- Dynamic recrystallization

Under certain conditions, a structure can recrystallize during deformation, i.e.

dynamically. Occurrence of this type of recrystallization is theoretically possible also

during deformation under cold conditions, although it is practically observed only rarely

(for very pure metals). Dynamic recrystallization can either be discontinuous,

geometrical or based on a progressive rotation of subgrains. The latter two types are


23

based on the presupposition of imposed strain with a limited or restricted movement of

HAGBs.

- Discontinuous dynamic recrystallization

Figure 6 depicts typical stress-strain curves describing deformation under cold and hot

conditions. For a deformation under hot conditions, the shape of the curve can be

influenced by the strain hardening rate depending on dynamic recovery or

recrystallization (i.e. discontinuous dynamic recrystallization). The dynamic recovery

curve is typical for metals with high stacking fault energies (e.g. Al, low-carbon steel),

where a deformation resistance steady state occurs after the first stage of strain

hardening. Evolutions of microstructures during dynamic recovery and dynamic

recrystallization are schematically depicted in Figure 7.

During dynamic recovery the original grains deform, while sub-boundaries stay more or

less equiaxed. The structure is thus dynamic and adapts continuously to the increasing

strain. The recovery process is significantly slower for metals with low stacking fault

energies (e.g. austenitic steel, Cu), which can enable accumulation of imposed energy. At

a certain critical strain, dynamically recrystallized grains start to occur on the

boundaries of the original grains. As a result, a chain structure occurs at grain

boundaries. With an increasing strain more nuclei get activated and new recrystallized

grains occur. At the same time, already recrystallized grains are deformed again. After a

certain amount of imposed strain, the stress-strain curve gets saturated. The

microstructure then consists of a dynamic mixture of grains with various dislocation

densities.


24

Figure 6. Typical deformation curves describing deformation under cold and hot

conditions.

Figure 7. Evolution of microstructures during deformation under hot conditions: (a)

recovery, (b) dynamic recrystallization and discontinuous dynamic recrystallization

Subgrain size (originating from dynamic recovery), as well as grain size (originating

from dynamic recrystallization) increase with temperature and decrease with increasing

strain rate. Contrary to dynamic recovery, dynamic recrystallization also includes

metadynamic recrystallization. During this type of recrystallization, a recrystallized

nucleus develops or grows dynamically under hot conditions. Nevertheless, only growth

occurs during subsequent static annealing.

- Geometrical dynamic recrystallization

Grains with serrated boundaries developed as a result of dynamic recovery can mutually

pin when their sizes achieve their thicknesses. This process can create a microstructure

appearing as dynamically recrystallized. However, it is a result of deformation and

recovery.

- Dynamic recrystallization via progressive subgrains rotation

During deformation, subgrains in the vicinities of already existing boundaries can be

subjected to more substantial rotations comparing to central grains areas. Development

of HAGBs can occur at high strains. Although the exact mechanism is not clear, it is

supposed that this is due to a combination of inhomogeneous plastic deformation,


25

accelerated recovery (in regions in the vicinities of grain boundaries) and grain

boundary sliding.

1.3.6. Grain coarsening (growth)

Grain growth after recrystallization is controlled especially by surface energy or grain

boundary energy. This driving force is by two orders of magnitude lower than driving

force for recrystallization. A more correct description for this grain growth process is

grain coarsening (sometimes also secondary recrystallization).

There are two basic types of grain coarsening: normal and abnormal. During normal

coarsening, the main mechanism is elimination of finest grains, while the grain size

distribution remains almost constant. During abnormal coarsening, growth of several

grains inside a pinned structure occurs. The pinning effect is usually caused by particles

of other phases or by low misorientation angles (i.e. low grain boundaries mobility). The

grain growth driving force is substantially lowered if the grain size achieves

approximately ½ of the thickness of the deformed sample. This is due to a reduced grain

boundary radius. At the grain boundaries equilibrium state, the force has to act

perpendicular to the sample surface. Moreover, grooves develop in the intersection of

grain boundaries with the surface, which restrict grain boundaries movement.

Due to a very high number of influencing factors it is not possible to determine in

advance when and how abnormal grain growth will occur.

1.4. Alternative deformation mechanisms

Although slip is the dominant deformation mechanism for most of metal materials,

activation of different mechanisms is possible under certain conditions. Among these are

twinning, creep, grain boundary sliding (GBS), and deformation related to phase

transformations. Various deformation mechanisms can be active during deformation

under hot conditions depending on the temperature and applied stress. To define areas,

in which the individual deformation mechanisms can be expected to occur, deformation

mechanisms maps are used. The maps practically bring information about the influence


26

of the temperature and applied stress on the strain rate and dominant deformation

mechanism. An example of such a map is shown in Figure 8.

Figure 8: Deformation mechanisms map for W (grain size 10 m). [1]

The upper boundary of the diagram is created by theoretical shear stress. This is stress

that should be applied in a case of perfect deformation of crystals with no defects by

mutual collective translation of crystal planes. Deformation proceeds via dislocations

movement in their slip planes, which is supported by high temperature, i.e. by the

process of dynamic recovery caused by cross-slip and climb.

In the lower part of the map, creep region can be observed. This region can further be

divided into various sub-regions, each of which is characterized by a creep sub-

mechanism. Creep can be caused by dislocation movement or diffusion. Diffusion creep

has two types, Nabarro and Coble creeps. Twinning can also be mentioned among

deformation mechanisms occurring under these conditions. However, it is typical only

for several materials - twinning, as well as GBS, are not included in Figure 8. The reason

for neglecting GBS is since GBS in polycrystalline materials is usually enabled by

diffusion creep. Therefore GBS is already included inside the deformation maps diffusion

creep areas. GBS also occurs during superplastic forming since it is supported by very

fine grain size and elevated temperatures.

A concrete example can be tungsten, which deforms by diffusion creep and the strain

rate of 10-9 s-1 at the temperature of 1560°C and a relatively low stress of 10 MPa. This


27

means that a deformation of 10% would take 3 years, which is very important

information for certain applications. On the other hand, at the stress of 100 MPa, the

material will deform by dislocation creep and the strain rate of 10-6 s-1. At the stress of

1000 MPa, the casual dislocation slip will dominate and the strain rate will be 0.1-10 s-1.

There is a presupposition that high-temperature dislocation creep is not different from

mechanisms occurring during conventional deformation under hot conditions.

1.4.1. Creep

Creep can be characterized as a deformation response of a material to long-time loading.

It comprises continuous elongation terminated by fracture. Creep is typical for high

temperatures, it occurs at low strain rates and relatively low stresses. Creep

deformation curve can be divided into three stages:

Stage 1 – primary creep

When a load is applied, a material is deformed and new dislocations are generated. The

dislocations then make obstacles in movement for each other, which causes decrease in

creep rate (Eq. 6)

dt/d 0 (6)

Stage 2 – secondary creep (steady-state creep, creep at constant rate)

In this area, strain hardening rate and dynamic recovery rate are equal. A constant load

is applied, which results into a constant strain rate after a certain time period.

Stage 3 – tertiary creep

Voids develop in the material, which causes increase in effective stress and also creep

rate. Voids growth quickly results into material fracture.

At low temperatures, strain hardening is dominant and the steady-state stage is usually

not reached. At high temperatures or high stresses, development of first voids can occur

even during the first stage, i.e. steady-state stage is either very short, or completely

missing.

Creep mechanisms

1. Diffusion creep


28

During loading of a material (plastic deformation) the material tries to deform via

movement of vacations through a crystal along its boundaries. Therefore the

deformation proceeds by a creation of vacancies on these boundaries due to tensile

stresses. At the same time, destruction of vacancies on boundaries by the influence of

compression stresses occurs. This can be expressed by Eq. 7.

v

bv

DD

DD

kTD

114

2

0 (7)

where is atom volume, D is grain size, Db, Dv are diffusion coefficients for volume and

grain boundaries diffusion, respectively, is grain boundaries area effective for grain

boundaries diffusion.

This equation assumes that ancillary processes of nucleation and annihilation at grain

boundaries and GBS require a similar negligible energy and stress. However, if this

condition is not met and volume diffusion dominates, Nabarro or Nabarro-Herring creep

occur. Contrariwise, Coble creep occurs in cases of dominant diffusion at grain

boundaries.

2. Dislocation creep

If a stress higher than corresponding to the diffusion creep region in a deformation map

is applied, another deformation mechanism is activated. Although deformation is

realized via dislocation slip and climb, diffusion is a process controlling dislocations

recovery by climbing. Depending on the temperature and stress, four main groups of

dislocation creep mechanism can be distinguished.

1 – slip and climb controlled by volume diffusion

2 – slip and climb controlled by pipe diffusion

3 – Harper-Dorn creep

4 – power law breakdown

The first two mechanisms can be characterized using the power law relation (Eq. 8).

n

b

GkT

GbDA

0 (8)

For higher stresses, the following energetic dissolution relation is valid (Eq. 9).


29

expK0 (9)

A linear dependence between stress and strain rate - Harper-Dorn creep - applies for

some materials.

1.4.2. Grain boundary sliding and superplasticity

Superplasticity is an ability of crystalline materials to extremely plastically deform when

subjected to tensile stress. The elongation is usually on the order of several hundred

percent. Superplasticity in most cases occurs at high temperatures and low strain rates,

while the necessary stress is rather low (to 20 MPa). Although this phenomenon has

already been widely studied, its application is still limited due to the requirement of very

low strain rates. Applicability of superplastic forming is therefore especially for

fabrication of components with complex shapes and with limited production capacities,

which can be for example specific components for aerospace industry.

Several types of superplasticity can be distinguished. One of them is superplasticity of

fine-grain structures, since fine grains are among the necessary conditions for this type

(GBS is the main mechanism). Besides, there are transformation superplasticity,

superplasticity induced by inner stresses and superplasticity at high strain rates

(hyperelasticity).

During GBS, mutual sliding of two grains occurs due to shear stress. In contrast to

dislocation slip during which grains get elongated in the direction of deformation, grains

more or less maintain their dimensions during GBS. For superplasticity forming GBS is

the entirely dominant mechanism. During GBS, cavities form in the regions where grains

slide along each other. It is therefore clear that during superplastic forming a

mechanism providing time suppressing (delaying) of cavities development has to be

active. To enable continuation of sliding, the material has to be shifted (usually by

diffusion creep) from regions where grains overlap (regions subjected to compression

stress) to regions with tendencies to create voids (regions subjected to tensile stress).


30

Conditions of superplasticity

The first condition of superplastic forming for ultra-fine grained materials is the

condition of high m (strain rate sensitivity) coefficient value, which is usually defined by

Eq. 10 for a concrete deformation temperature.

),T(log/log 0 (10)

High values of m minimalize formations of necks during tensile tests (i.e. localization of

deformation), expected in the absence of strain hardening. The only possible mechanism

of strain hardening during diffusion creep GBS is grain coarsening.

The second condition is an increase in temperature to a sufficient value. The critical

value is usually a temperature higher than half of the melting temperature of the

particular material.

The third condition is the already mentioned strain rate (usually to 10-2 s-1, preferably

lower). Mathematically it can be described by Eq. 11.

m)(k 0 (11)

The presupposition is therefore the existence of two possible deformation mechanisms:

dislocation slip and GBS enabled by diffusion creep. The m coefficient is reciprocal to the

n coefficient. Since m is higher for GBS than for dislocation sliding (m>1/n), the curve for

GBS has steeper slope than for dislocation creep (Figure 9). At higher strain rates, the

required stress is lower for dislocation sliding and therefore this mechanism dominates.

Contrariwise, GBS dominates for lower strain rates. In the region of dominant GBS

enabled by diffusion creep, m is a function of strain rate.

Figure 9: GBS vs. dislocation sliding for mDS < mGBS. [1]


31

The fourth condition describes the grain size, which should be between 1 and 10 m.

However, the influence of grain size is different for both the mechanisms. Decreasing

grain size increases the size of GBS and superplasticity regions at the expense of

dislocation sliding regions. This is one of the reasons for preparation of UFG materials,

which could enable high-speed superplasticity forming.

The fifth condition is the presence of fine dispersed secondary phase particles with sizes

comparable to the grain size. This suppresses grain growth at low strain rates and high

deformation temperatures.

The sixth condition relates to the grain boundaries. These enable easy sliding during

superplastic forming and can also act as sources and storages of vacancies. However, it

is still not clear how these relate to superplasticity. It was found that HAGBs support

GBS significantly more than LAGBs. On the other hand, voids develop more often in

regions with HAGBs, than with LAGBs. Consequently, material failure develops in a case

of their coalescence into networks. However, not only the frequency of their occurrence

is of influence, but also the character of grain boundaries distribution. Moreover,

likewise for recrystallization and grain growth, certain CSL boundaries are especially

important for the probability of GBS to develop. Small voids develop on grain

boundaries (triple points, precipitates). Due to continuing deformation they

subsequently grow and eventually mutually adhere, which result into a macroscopic

failure. The character of boundaries is very important during this process.

Most of the known materials can be subjected to superplastic forming. Among the most

important are many eutectic alloys, Fe-based and Ti-based alloys and many Al-based

alloys (supral, Al-Zn-Mg, AlMgMn etc.).

1.5. Kinetics of structure-modifying mechanisms

1.5.1. Influence of microalloying elements

Microalloying elements very effectively delay softening processes not only by their

presence in solid solution, but especially if present in fine dispersed precipitates. The

most effective are precipitates smaller than 5 nm segregated mostly on subgrain

boundaries. Since precipitates usually coarsen at higher temperatures (approx. 1,000°C),


32

they lose their influence on the softening processes kinetics. The most significantly delay

softening processes the following elements (from the most to the least effective): Nb-Ti-

Al-N. Niobium is effective up to 0.04%, titanium from 0.10 to 0.15%. Higher contents of

Nb and Ti do not substantially influence their effect.

Due to the delaying effect of microalloying elements, recrystallization is suppressed and

recovery is the dominant softening mechanism at lower temperatures to strains of

approximately 0.6 - 0.7. Dynamic recovery occurs at high temperatures together with

dynamic recrystallization, which is the dominant mechanism around the temperature of

950 - 1,000°C. The influences of Nb and Ti on strain required for development of

dynamic recrystallization are shown in Figure 10. When content of Nb is approx. 0.04%,

dynamic recrystallization cannot be expected in the final rolling stage at temperatures

below 900°C (stage 2).

Critical strain value for development of dynamic recrystallization pik can be expressed

by Eq. 12.

nm

pik dZA 0.. (12)

where A, m, n are constants.

After their computation can Eq. (12) be depicted as follows (Eq. 13).

415021

0 1094 .Z.d., ,/

pik (13)

The influence of microalloying elements in solid solution on dynamic recrystallization

delay can be expressed by soluble retardation parameter (SRP) (Eq. 14).

100..

1,0.log

xatt

tSRP

ref

x (14)

where tref , tx are times required for development of dynamic recrystallization in a

reference steel, and in a steel containing x element.

With increasing strain rate, the microalloying element effect increases as well (Eq. 15).

2

121 log.

e

eeSRPeSRP

(15)


33

The influence of Z parameter on the size of dynamically recrystallized grain can be

calculated as follows (Eq. 16).

21 log16.exp10 aZaddr (16)

Mn-Si steel: a1 = 0.40 a2 = 0.22

Nb- steel: a1 = 0.456 a2 = -0.16

Figure 10: Influence of Nb and Ti on strain required for dynamic recrystallization

development.

Kinetics of dynamic recrystallization is fastened by higher temperature, higher strain

and strain rate and smaller grain size. Dynamic recrystallization causes grain refinement

and shortening of incubation times for static recrystallization. In the final rolling stage it

is necessary to eliminate a possible occurrence of dynamic recrystallization.

Static recovery is important especially during rolling at temperatures lower that 900 -

950°C. It partially lowers dislocation density and the amount of nuclei for ferrite

formation.

The general relation for determination of tx time, i.e. the time during which X portion of a

structure is recrystallized by static recrystallization, is the following (Eq. 17).

RT

QdCt x exp... 42

0

(17)


34

For Nb-V steels, the following relation counting with the influence of thermomechanical

parameters and grain size was derived (Eq. 18)

RTQdZCt Rx /exp....2

0

48/3 (18)

For < cr: C = 3.54·10-21, QR = 480 kJ·mol-1

Grain size after static recrystallization:

Mn-Si steel: 167,0

0 .. dDdrex (19)

Nb steel: 671,067,0

0

| .. dDdrex (20)

The relations are valid for < cr and for Nb-steels for temperatures above 950 °C.

After static recrystallization is finished, further grain growth can be described as follows

(Eq. 21):

RTQtAdd ggrex /exp..|1010 (21)

where A| is constant, Qgg is activation energy for grain growth.

Grain growth kinetics is expressed by Eq. 22.

ntKd . (22)

where dy is austenitic grain size, t is time after recrystallization finish, K, n are constants.

1.5.2. Kinetics of static recrystallization

The basic model is the Avrami relation (23) expressing the portion of recrystallized

structure X as a function of time t. This equation provides a successful description of

recrystallization curve on the basis of two parameters.

k

xt

tX

5,0

693,0exp1 (23)

The k exponent, which is between the values of 1 and 2, is usually not sensitive to

deformation parameters, whereas the time for 50% of recrystallized portion, depicted as

t0,5, changes in a wide range and depends on deformation conditions. It is therefore


35

favorable to express the t0,5 value quantitatively in dependence on deformation

parameters changes.

The basic physical-metallurgical approach implies that the recrystallization kinetics

depends on the nuclei density, driving force for growth and grain boundaries mobility.

On the other hand, besides temperature, the theory cannot exactly define the influence

of other parameters. Therefore, an adjusted constitutive equation is usually used instead

of its physical form. For example, Eq. 24 was derived to determine the time of static

recrystallization start for a Nb-alloyed steel with Nb dissolved in solid solution (depicted

as t0,05x).

Nb

TRTdt x 185

10.75,2exp

300000exp10.75,6

542

0

20

05,0 (24)

This equation is valid when strain ε is lower than required for dynamic recrystallization

start. Its appearance is compatible with an equation derived for a C-Mn steel with 0.04%

Nb, original grain size d0 in μm and Nb content in wt. %. If k = 2, then t0,5x is equal to 0.27

t0,05x. The subsequent size of recrystallized grains drex is described using Eq. 25.

67,067,0

09,0 ddrex (25)

These empirical equations cannot be relied on completely, although there evidently is a

relation between Eqs. 24 and 25 (if grain growth during recrystallization is eliminated).

The number of nuclei in a unit volume Nv is then (Eq. 26):

343,0

rexv dN (26)

From the above mentioned is obvious that the influence of the original grain size on

kinetics of recrystallization according to Eq. 24 is physically dependent on nucleation,

since grain boundaries provide preferential locations for nucleation. The d02 dependence

is valid only if nucleation occurs only on boundaries. However, in reality boundaries

edges and mutual interaction of grains have to be taken into account as well. Likewise,

the ε-0,67 dependence in Eq. 25 depends on nucleation, for which no physical model has

been developed so far. Due to these facts, parabolic hardening leading to linear increase

in dislocation density with increasing strain and subsequently driving force for grain

boundaries migration is really expected. Considering that grain growth rate, driving


36

force and nucleation are proportionally dependent only on grain boundaries, we get an

even more complicated dependence on time. Consequently, various researchers come to

various results of analyses even for the same types of steels. Nevertheless, all the

researchers consider the influence of strain, grain size and dwell temperature to be

substantial, although the strain rate and temperature in the Zenner-Hollomon

parameter vary (can be included due to its influence on dislocation density). As has

already been explained above, this has only a neglectable influence on t0,05x. The only

important difference between the different forms of the equations is the value of critical

strain – the limit up to which can be applied.

As was already pointed out, Eqs. 24 and 25 are valid only if strain is lower than critical

for dynamic recrystallization start. At such strains only small recrystallized grains

develop on the original grain boundaries, which are the most favorable locations for

nucleation of static recrystallization. However they provide only a small amount of

nuclei for dynamic recrystallization which occurs with continuing deformation. This

results into a decreasing limit strain for static (or metadynamic) recrystallization start,

as depicted in Figure 11.

The value of critical strain has been well investigated for C-Mn steels, but less for

microalloyed steels with Nb. This value is very important for determination of one of

three critical temperatures – temperature of zero recrystallization Tzr – at the absence of

precipitation. For example, if the dwell time between passes is 1s, then Tzr = 900°C for a

steel with 0.095% Nb. As was explained above (DRCF), there is a real possibility of

development of dynamic recrystallization by an accumulated deformation lower than εc

without a prior occurrence of static (metadynamic) recrystallization. Since εc depends

on Zenner-Hollomon parameter Z and grain size d0, it is obvious that when precipitation

do not occur the Tzr temperature depends on forming conditions, but also on chemical

composition, as depicted for a carbon steel for strain rates 1 to 100 s-1 (Figure 11).


37

Figure 11. Dependence of static recrystallization start on strain at 900°C for Nb

microalloyed steel with grain size 50 m.

1.5.3. Grain growth

In inter-pass dwells, grain growth occurs in a relatively short time after recrystallization

is finished. This is depicted in Figure 12 for a C-Mn steel. Theoretically, the time

dependence of recrystallized grain size can be derived as Eq. 27.

d d AtQ

RTn

rex

n gg

exp (27)

The theoretical value of exponent n is around 2, although the real value is usually higher.

For the example depicted in Figure 12, the data for times up to 100s is described the

best even with the exponent n = 10. The Qgg value is also usually much higher than the

expected theoretical value for diffusion controlled grain boundaries migration. The

differences are even more significant after considering that the member A, which is

theoretically constant, is in reality dependent on time and also temperature. This ensues

from segregations occurring on grain boundaries. In practice A has two values, one for

short and one for longer times. It is assumed that further research would lead to more


38

complicated equations, since grain growth retardation by niobium present in solid

solution also takes place at the same time.

Figure 12. Grain growth as a function of time after complete static recrystallization for C-

Mn steel.

1.5.4. Precipitation kinetics

Retarded or delayed recrystallization due to strain-induced precipitation is the basis of

the already mentioned conventional controlled forming (CCF) process. From the point of

view of simulations, the start of strain-induced precipitation is the main factor

influencing the forming mechanisms. Therefore to perform modelling is necessary to

determine the time in which 5% of precipitates develop.

The basic theory of nucleation takes into consideration the temperature and

oversaturation coefficient ks,, but does not define the value of critical strain and density

of heterogeneously scattered nucleation points. Integration of practice and theory

results into Eq. 28.

skTRTNbZt

ln

10.5,2exp

270000exp10.3

3

1015,016

05,0 (28)

The dependences on strain and Zenner-Hollomon parameter are completely empiric and

are valid only for strains lower than critical for dynamic recrystallization start, although


39

they are physically just. Especially implementation of parameter Z instead of simple

strain rate is a contribution to a reliable solution of the equation.

The results of integration of the equation for recrystallization start (24) and the above

mentioned relation (28) for strain-induced precipitation start are shown in Figure 13. In

the figure, the intersections of the curves define recrystallization stop temperatures

(RST). Previous analyses of recrystallization data after one pass showed a delayed

development of recrystallization but not its stopping under RST as a result of

precipitates coarsening in time. Under real forming conditions, two and more passes

under RST enable further development of fine precipitates. Therefore, from practical

reasons, recrystallization is stopped when the temperature reaches the RST value. From

previous analyses is also evident that Tzr investigated during simulations corresponds to

RST if recrystallization does not occur at higher temperatures as a result of dissolution

of niobium in solid solution and if dwell times are short. The Tzr value is therefore very

sensitive to strain and chemical composition again.

Figure 13. Interaction between recrystallization and strain-induced precipitation as a

function of Nb content in 0.07C-0.004N steel rolled with 15% reduction at strain rate of

10 s-1. RLT shows the lower limit for complete (85%) recrystallization and RST shows

the upper limit for a complete recrystallization stop.

1.5.5. Interaction of precipitation and softening processes

Mutual interaction of the processes can be observed in RPTT diagram (Figure 14). The

temperature interval can be divided into 3 stages:


40

a) In the first stage, recrystallization is finished before the start of precipitation.

Recrystallization is restricted by microalloying elements dissolved in solid solution

(temperature range between 1,000 and 1,050°C for C-Mn-Nb-V steel).

b) In the second stage, recrystallization starts before precipitation, but in the final

stage strain-induced precipitation takes place (temperature interval between 930

and 1,050°C).

c) In the third stage, precipitation develops before recrystallization starts and delays

its whole kinetics. The maximal effect is during the most significant development

of precipitation, i.e. for temperatures around 850°C for C-Mn- Nb-V steels.

The influence of microalloying elements in solid solution and precipitates on the basis

phenomena occurring during controlled forming under hot conditions is summarized in

Table 1.

Figure 14. Interaction of strain-induced precipitation (RPTT diagram).

Table 1. Influence of microalloying elements.

microalloying element

in solid solution in precipitates

influence

- lowering of diffusion

ability

- recrystallization

delaying

- grain boundaries

blocking

- lowering of transformation

temperature y

- retardation of

dislocation movement


41

structural

phenomena - grain refinement

- grain refinement

- texture

development

- grain refinement

- dislocation hardening - precipitation

hardening

- influence on sulphides

shapes

1.6. Thermomechanical processing of selected materials

1.6.1. Steel

At present, most of the produced steel is fabricated using continuous casting with typical

dimensions of 0.2-0.3 m thickness and 2-3 m width. In some cases the continuously cast

products are rolled directly after casting, but they are usually cooled and divided into

slabs. Slabs are subsequently heated to a temperature around 1,200°C and rolled under

hot conditions on a heavy section or slabbing to the thicknesses of 50-30 mm. Finishing

rolling to the final thicknesses of a few millimeters is then performed on fine sections.

Strain rates on heavy sections are approximately 10 s-1, while on fine section they can

reach 100 s-1.

For unalloyed carbon steels, rolling on heavy sections starts around 1,150°C and finishes

at temperatures around 1,000°C, followed by a final rolling between 1,000 – 910°C.

Cooling after final rolling is especially important for these steels, since the cooling rate

controls phase transformations (austenite to pearlite, bainite etc.) and therefore final

properties. For special types of steels, such as HSLA or multi-phase steels, controlling of

the austenite-ferrite transformation is necessary to achieve fine-grain steels with

superior properties.

For austenite stainless steels, slabs are heated to the temperature of 1,280°C with 3

hours dwell and subsequently rolled on a heavy section from the thickness of 150-220

mm down to 30-40 mm. Additional heating is usually performed after the first pass

(25% deformation) before deformation is continued. This approach facilitates

homogenization of cast structure. Final rolling on a fine section is performed on a 5-7


42

rolling stands tandem at temperatures between 1,000-1,100°C to the thicknesses of a

few millimeters (5-0.5), followed by annealing.

Conventional steels (Al killed)

The usual forming method for these steels consists of continuous casting of a slab, its

temperature levelling (1,200°C), hot rolling with the finishing temperature of 900°C,

coiling (under 600°C), cold rolling (reduction of 70-90%) and annealing (650-

700 °C/several hours). Rolled strips recrystallize during annealing and, at the same time,

precipitation of fine AlN particles occurs. During and after hot rolling, precipitation of

AlN must not occur since these precipitates would coarsen during the continuing

deformation and would not be further effective during annealing. Therefore, rapid

cooling down to temperatures around 600°C is advantageously performed after hot

rolling. At these temperatures, the product can be coiled since further cooling does not

result into precipitation of AlN. Contrariwise, if the material is heated too fast during

annealing, recrystallization occurs before precipitation of AlN (unwanted). The final

rolling temperature should be the lowest possible in the austenitic region (closely above

A3). This results into achievement of very fine grain sizes. Optimal reduction during cold

rolling should be selected considering chemical composition of the given steel. For most

cases it is around 70-90%. It is also necessary to enable post-recrystallization grain

growth.

Continuously annealed low-carbon steels

On continuous mills, strips are annealed at the heating rate of 10°C/s, annealing is

performed at relatively high temperatures but short times (1-5 min). Due to such a high

heating rate, recrystallization for cold rolled strips is faster than precipitation of AlN

particles. This is also the reason why a method similar to the one for conventional

annealing (i.e. filtration of recrystallization nuclei) cannot be applied and it is necessary

to apply a different mechanism.

The heating temperature before hot rolling has to be low (under 1,100°C) in order to

prevent massive dissolution of MnS and AlN. These particles can further coarsen during

hot rolling without significantly influencing material behavior during annealing.

Furthermore, coiling temperature should be sufficiently high (above 730°C) to enable

development of coarse carbides during slow cooling of the coil. On the contrary, heating


43

rate during the final annealing should be sufficiently high (approx. 10°C/s) to prevent

massive diffusion of carbon atoms into the matrix. In most cases, high-temperature

annealing (up to 850°C) is used to enable grain growth after annealing.

Austenitic steels

A large number of phases can precipitate in austenitic steels during processing,

including sigma phase – an intermetallic phase usually containing 50% Cr and 50% Fe.

Sigma phase causes embrittlement and decrease in corrosion resistance. Although it is

thermodynamically very stable, it develops very slowly, which is favorable. For this

phase to occur, temperature around 600°C and time in the order of magnitude of years

are usually required. It typically develops around other precipitates and in locations

with high energies, such as triple points and grain boundaries. The development is

accompanied by a relatively complicated mechanism.

Due to precipitation, continuous recrystallization does not proceed. Development of

recrystallization is usually initiated by strain especially on deformed grain boundaries,

but other mechanisms (such as coalescence of subgrains and their growth) are probably

also important, especially in highly deformed locations. Sigma phase particles are not

necessary for recrystallization initiation. Coarser precipitates typically occur in the

vicinities of serrated boundaries, which proves the development of both the processes,

recrystallization and precipitation, although recrystallization is delayed. At lower

annealing temperatures, oversaturation increases until the start of precipitation after a

certain annealing time. Further development depends on recrystallization activation

energy.

In practice, occurring changes of chemical composition and dislocation density imply

that mutual interactions can occur in a relatively wide temperature interval. For

example, temperatures in the range of 700-800°C mean a very large and fast

precipitation of sigma phase, which is enabled by discontinuous recrystallization.

Interstitial free (IF steels)

Deep-drawing ability of IF steels is increased when the content of interstitial elements

(carbon, nitrogen) is decreased. A common IF steel contains 20-30 pm of carbon and 30-

40 ppm of nitrogen. Other interstitial elements are Ti and Nb.


44

A problem for these pure steels is achievement of fine grains after hot rolling. A rolling

mill has to provide favorable conditions to ensure precipitation of carbonitrides during

hot rolling and prevent massive austenite growth. Final passes during hot rolling are

usually connected with large reductions at relatively low temperatures. The final rolling

temperature is low (around A3), annealing is performed at high temperatures (up to

850°C) due to low content of C.

Steels with higher strength

The requirements to decrease emissions and fuel consumption in automotive industry

make car producers to, among other, use steels with higher strengths. The main reason

for their usage is the possibility to decrease the robustness of individual components

while maintaining their properties. However, increase in strength is usually

accompanied by decrease in formability.

Steels with controlled phosphorus content

The first possibility to increase strength of an alloy is to increase the content of atoms in

solid solution. Elements such as P, Si and Mn can increase strength of steels.

Nevertheless, Si and Mn also significantly decrease formability. Regarding P, it is

necessary to maintain only a very low content (0.04-0.08 vol.%).

Microalloyed steels and HSLA (High Strength Low Alloy) steels

The second possibility of material strengthening is precipitation strengthening. This can

be achieved as a result of addition of a small amount of Nb and Ti carbides and

carbonitrides. Non-dissolved fine particles prevent austenite grains from growing

during hot rolling. In final passes, austenite does not recrystallize and grains get only

flattened as a result of continuing deformation. During subsequent cooling, many ferrite

nuclei from which fine ferrite grains subsequently develop are activated. This results

into a significant increase in strength.

BH (Bake Hardening) steels

The cooling rate after recrystallization annealing is relatively high in most rolling mills.

Therefore, carbon content after cooling is far higher than for equilibrium state.

Consequently, carbon atoms diffuse into dislocations, which subsequently blocks their


45

movement (i.e. deformation aging occurs). For conventional steels, this oversaturation

by carbon is prevented via double step annealing. In the first step, a strip is annealed for

a few minutes above the temperature of 710°C and subsequently cooled down to 400°C.

In the second annealing step, the strip is homogenized for a few minutes and

subsequently cooled down to room temperature. For BH steels, the effort is to maintain

low carbon oversaturation at room temperature. During stamping of components for car

bodies, new dislocations are generated, the movement of which is subsequently during

painting at elevated temperatures blocked by carbon atoms. This increases strength by

approx. 50 MPa.

DP (Dual Phase) steels

DP steels are low-carbon steels with carbon content around 0.1%. They contain soft

ferrite and 10-20% of hard martensite. This is achieved using heating around 800°C, i.e.

in the dual phase region, and subsequent rapid cooling. During the annealing, a portion

of austenite transforms into ferrite. From transformed ferrite grains, carbon atoms

return back to the residual austenite. During subsequent rapid undercooling, this

carbon-rich austenite transforms into martensite. DP steels therefore combine high

strength and a good formability with a significant strain hardening.

After conventional hot and cold rolling, a strip is reheated to the dual phase region

(800°C). The structure consists of ferrite with 10-20% of austenite depending on

chemical composition. After annealing, a strip is rapidly cooled and austenite gets

transformed into martensite. For some types of steels, residual austenite can be present

in the structure even at room temperature. Strength of such steels can vary between 500

and 1,000 MPa.

TRIP (Transformation Induced Plasticity) steels

A subsequent research of DP steels resulted into the development of TRIP steels

containing ferrite, bainite and residual austenite. Their initial structure is similar to DP

steels. Similarly to DP steels, annealing in the dual phase region is performed as first.

During a second annealing (300-400 °C), the residual austenite is transformed into

bainite, but a portion of the austenite (5-10%) remains maintained in the structure.

Carbon content in the residual austenite is so high that the austenite stays metastable

even during cooling to room temperature. The presence of residual austenite is very


46

important, since during subsequent cold rolling, this gradually transforms into

martensite. This is the so called TRIP effect.

In most cases TRIP steels are conventionally hot and cold rolled steels subsequently

annealed in two steps. TRIP steels feature high strengths (600-1,000 MPa) and a very

good plasticity (elongation of approx. 30%).

HSLA (High Strength Low Alloy) steels

Rolling of HSLA steels is performed in the austenitic region. If the forming temperature

stays above a certain level (temperature of zero recrystallization), austenitic grains

remain more or less equiaxed due to subsequent recrystallization periods in the dwell

times between the individual passes. During final hot rolling, transformation of austenite

to ferrite and partially pearlite occurs. Since most of the ferrite grains develop on the

original austenite grains boundaries, a finer austenite grain size results into a finer

ferrite. The initial heating temperature for a conventional rolling is between 1,200 and

1,250°C. However, it is difficult to prevent austenite grains from growing in this region.

Final rolling is performed at temperatures higher than Tzr in order to eliminate high

rolling forces due to strain hardening of the material resulting from the absence of

recrystallization.

The heating temperature for controlled rolling is limited due to the requirement on

grain growth prevention (1,100°C). Final rolling is then performed at temperatures

lower than necessary for recrystallization development. This results into development

of pancake structures with increased amounts of grain boundaries, i.e. higher amounts

of nucleation sites and consequently finer ferrite grains. Another result is also a higher

strain hardening in final passes, i.e. the necessity to apply higher deformation forces.

From this reason larger reductions are usually applied in first passes and lower ones in

final passes.

Accelerated cooling starts above the A3 temperature and continues to the temperature of

500°C. This activates another ferrite grains development inside original austenite grains

and contributes to a further ferrite grains refinement. Nevertheless, a rapid cooling can

result into a development of a certain portion of bainite. This implies a higher strength,

but lower ductility.


47

ULCB (Ultra Low Carbon Bainitic) steels

Upper bainite in the form of laths or plates is formed at temperatures around 500°C and

very slow cooling rates. Lower bainite forms in the form of lentils or laths similar to

martensite due to ultra-low carbon content. Granular bainite forms at very slow cooling

rates. Martensitic austenite forms in the form of micro-islands at medium and low

cooling rates due to microsegregations. Increased mechanical properties are caused

especially by very small subgrains of bainitic ferrite developing at low transformation

temperatures. For this type of steels, the effect of secondary hardening due to

precipitation of Cu and austenite phase transformation is usually observed.

Highly-alloyed steels

Tool steels

In a low-tempered state they feature high strength and low ductility. During processing,

they are prone to development of quenching cracks and premature and retarded

fractures. The cause of these defects is coarse plates of martensite rapidly growing

during quenching. When they stop on obstacles, their phase effects initiate development

of microcracks. If a deformation is performed before quenching, martensitic needles get

refined. This is due to a decrease in mobility of austenite-martensite interfaces due to

the presence of dislocations and increase in martensite nucleation rate. It is assumed

that annihilation of dislocations does not occur before martensitic transformation.

Stainless steels

For these steels, strain hardening is used to increase yield strength above 400 MPa while

maintaining favorable plasticity. Hardening is achieved by final rolling in the region of

restricted recrystallization, i.e. under 950°C. Due to the presence of residual stresses and

their indirect influence on corrosion resistance, steels with an increased corrosion

resistance, i.e. steels with low carbon contents and highly-stabilized steels, are especially

suitable for this type of processing.

For special purposes, e.g. materials for nuclear plants components, austenite grain

refinement and fine precipitation of titanium carbonitrides are achieved via high strains

at temperatures lower than 1,050°C and stabilization annealing.


48

1.6.2. Controlled rolling of microalloyed steels

Controlled rolling of microalloyed steels enabled to increase yield strength by 50-100%

in comparison with C-Mn St52 type steels, while maintaining ductility down to -50°C and

maximum carbon equivalent Cc 0.45. The main influencing elements are C, Mn, Si, Mo, S,

Al, Nb, and V.

Basic parameters of controlled rolling

Heating temperature of slabs is between 1,050 and 1,200°C, depending on solubility of

carbonitrides. Its selection is made also regarding the required final rolling temperature.

Decrease in the heating temperature by 100°C results into up to 8% energy savings.

Heating time of slabs should ensure sufficient dissolution of Nb(CN), i.e. min. 0.02% Nb

in solid solution and uniform not-coarsened austenitic grains (150 to 200 m). Heating

time is longer than for Mn-Si steels, usually 5 to 8 minutes for 1 cm of slab thickness.

Rolling stage I (pre-rolling)

The total strain is 0.6 to 0.8 with individual thickness reductions of min. 0.15. The

temperature of the first rolling stage end is usually 50 to 150°C lower than the heating

temperature. At temperatures higher than 1,000°C the danger of grain growth occurs,

while at temperatures lower than 950°C austenite recrystallization does not develop

entirely. Optimal are temperatures around 1,000°C and higher strains in final passes.

Cooling before rolling stage II - work-piece, the thickness of which is at least 3x higher

than its final thickness, is cooled from the temperature of approx. 1,000°C to the initial

temperature of stage II. After excluding recrystallization, this temperature should not

exceed 900°C. Accelerated cooling decreasing precipitation of Nb(CN) is advantageous.

The cooling time depends especially on the work-piece thickness.

Rolling stage II

The total strain is between 0.6 and 0.7, advantageous are strains > 0.7. The final

rolling temperature depends especially on the required final mechanical properties and

deformation resistances. After reaching the yield strengths of 450 - 500 MPa, final

rolling temperatures decrease to 700 to 750°C.


49

Recently, USA and Japan have developed dual-phase steels. Their development was

enabled especially by continuous annealing furnaces with accelerated cooling. Their

structure feature ferrite and islands of martensite or bainite. They can be produced

either by annealing after rolling in and region and subsequent cooling, or by hot

rolling of strips with subsequent rapid cooling.

SHT (Sumitomo High Toughness) process

The process is suitable for production of thick plates from constructional weldable

steels. It uses four basic principles to increase mechanical properties.

- grain size refinement, precipitation hardening, strain hardening, transformation

hardening

For the SHT process the mentioned requirements are met by a four-step mechanism:

- grain refinement by heating close above Ac3 temperature

- grain refinement by rolling in the region of austenite recrystallization

- deformation in the region of non-recrystallized austenite

- grain refinement by dynamic recovery or recrystallization of ferrite formed in the

dual-phase region

The SHT process consists of heating to the temperature of 1,150°C, rolling with

subsequent cooling of the work-piece on air under the Ar1 temperature, rolling with a

total reduction over 60%, final rolling at the temperature of approx. 700°C and

subsequent cooling on air.

The structures of plates produced by SHT feature fine and homogenous grains.

Increased ductility at lower temperatures and improved weldability imply their

application especially for lower temperatures and enables 2-3% saving of nickel.

1.6.3. Controlled rolling and controlled cooling

The disadvantage of controlled rolling is high rolling forces at lowered final rolling

temperatures and the required interpass dwells decreasing the efficiency of rolling mills.

Controlled cooling after rolling can be applied to achieve:


50

- ferrite grains refinement

- pearlite or bainite refinement

- increase in final rolling temperatures

- improvement of weldability by decreasing of carbon equivalent Ce

Since plastic deformation influences temperatures and kinetics of phase transformations

it is necessary to know the particular ARA-diagrams. For structures after controlled

cooling, the cooling rate during phase transformations and stop temperature (tstop), at

which cooling is interrupted, are critical. Accelerated cooling should not be performed

till the end of a phase transformation to prevent massive occurrence of quenched

structures. For C-Mn-Nb-V steels, the optimum stop temperature is around 550°C. The

final structure consists of fine ferrite with uniformly scattered island of bainite. Fine

ferrite structure with pearlite in texture arrangement can be achieved only by controlled

rolling. For C-Mn-Nb steels, the optimum cooling rates are 10 to 15 K/s.

At the same final rolling temperatures, the differences in Re between processing by

controlled rolling (CR) and controlled rolling with subsequent controlled cooling

(CR+CC) can reach 50 MPa. For examples, the same Re values can be achieved at final

rolling temperatures of 700°C (CR) and 850°C (CR+CC).

Hardening achieved during controlled cooling is especially due to

- ferrite grain refinement, precipitation hardening and bainitic hardening

Determination of Re increase and change in transformation temperature can be made as

follows (Eq. 29 and Eq. 30):

PRe dKR 2/1. (29)

BBPRTR fKdKT .. |2/1 (30)

where K, K are coefficients expressing grain refinement in the Hall-Petch relation, d is

ferrite grain size, KB, fB are hardening coefficients of bainite and the portion of bainite in

structure, PR is strain increment from precipitation hardening, β, α are correction

coefficients.


51

1.6.4. Aluminum and its alloys

Most of the Al products is produced by the method of direct solidification into castings

with thicknesses of 0.4-0.6 m, widths of 2 m and lengths of 9 m (i.e. weights of 20-30

tons). The castings are at first homogenized at the temperature of 500-600°C for a

relatively long time (several hours) in order to eliminate segregations and non-

equilibrium low-melting eutectics. This facilitates subsequent processing and increases

inhomogeneity. Annealing can influence the final structure in various ways, e.g.

especially for Cr, Mn, Zr containing alloys precipitation reactions during which dispersed

particles develop can occur.

After homogenization, an ingot is usually rolled on a reversible rolling stand to a plate of

30-10 mm thickness. The number of passes is 9-25. After hot rolling, i.e. before cold

rolling, the strip is coiled or, on modern mills, further hot rolled on a tandem. In practice,

a tandem mill has 2-6 stands. Cold rolling is usually performed using quarto stands with

two coil boxes, on which rolling to thicknesses of 15-20 μm is performed. For even

thinner foils (6 μm) the foils are doubled and further rolled. Interpass annealing is

necessary to enable large reductions without material failure.

At present, the technology of continuous strip casting is used for production of Al strips.

The thickness of the strips is usually 10-20 mm, due to which it is possible to directly

roll such strips in a tandem mill and therefore make the entire process more effective.

2. Controlled forming processes



describe the principle of conventional controlled forming


52

describe recrystallization controlled forming and define differences

between both the mentioned processes

describe dynamic recrystallization controlled forming

define interactions between recrystallization and stress-induced

precipitation

Lecture

2.1. Conventional controlled forming (CCF)

Stress-induced precipitation of Nb(CN) in steel requires approximately 10s. After

accumulation of a favorable amount of strain and occurrence of a certain volume

fraction of precipitates, austenitic structure of a pancake type develops. When the

stress-induced precipitation occurs faster than both the static and dynamic

recrystallizations, strengthening accumulates with every single pass and the mean

deformation resistance increases with increasing temperature. This results into high

deformation loads due to high dislocation densities. On the other hand, the resulting

microstructure consists of fine ferrite grains. This is caused by recrystallization of

austenite grains between passes, which is enabled since the flattened austenitic grains

before transformation have a high specific surface Sy.

2.2. Recrystallization controlled forming (RCF)

A carefully controlled recrystallization of austenite leads to a development of a fine

microstructure. This occurs at temperatures higher than 900°C, which is higher than for

CCF. Successful performance of this type of forming requires two basic factors. The first

one – recrystallization must not be too slow so that the required times are not too long.

This can be achieved by additions of V, or better Nb. Additions of V cause grain

refinement - decrease of grain boundaries mobility during recrystallization, although the

recrystallization is not entirely stopped. The second factor – grain growth must be

prohibited after each recrystallization cycle. Grain growth occurrence has a negative


53

effect on the recrystallization grain refinement process under lower temperatures. For

this purpose 0.01% Ti can be added to support propagation of fine TiN particles during

cooling after continuous casting. When these particles are of a favorable size and

homogenous distribution, they can completely eliminate austenitic grain growth during

each recrystallization cycle.

2.3. Dynamic recrystallization controlled forming (DRCF)

An alternative form of recrystallization is dynamic recrystallization. This occurs in

cases for which the forming temperature is low and time between the individual passes

is too short for occurrence of a conventional recrystallization between passes. The

amount of precipitated carbides and nitrides is limited as well. Dynamic recrystallization

consists of nucleation and growth of new grains during deformation (contrary to the

afore-mentioned processes, for which these occurred after a deformation pass.) This

requires accumulation of a sufficient amount of deformation, i.e. usually real

deformation larger than 1, enabling a complete recrystallization of the microstructure

throughout the cross-section of the work-piece.

2.4. Interaction between recrystallization and stress-induced

precipitation

When static recrystallization occurs before precipitation, the conditions are favorable

for RCF. When precipitation occurs before static or dynamic recrystallization, the

conditions for occurrence of pancake structure are fulfilled and CCF can be performed.

Last but not least, DRCF is performed when dynamic recrystallization occurs before

precipitation.

The mechanism of ferrite grain refinement occurring in real processes can be

determined by a comparison of processes with a continuous precipitation during cooling

(CCP = Continuous Cooling Precipitation) in a dependence on temperature and time (see

Section 1.5.5). From such a dependence is evident that dynamic recrystallization can be

active as long as the final passes are within the times of 10s or shorter, i.e. before the

significant occurrence of stress-induced precipitation. This however requires no static

recrystallization during idle times between passes, but also accumulation of a significant


54

amount of strain sufficient for a development of dynamic recrystallization, which

eventually forms the microstructure.

3. Modelling of thermo-mechanical forming



describe the pros and cons of physical modelling – define the basic

condition of the process

define modelling using high-performance computers and appropriate

softwares, define algorithms of the used processes and perform their

evaluation

Lecture

3.1. Physical modelling of TMP

TMP is performed with the aim to increase physical properties of materials. Such

properties can be characterized with numerous tests, which are always standardized for

a selected material. The mostly used tests are tensile, compression and torsional. The

tests are usually performed at high temperatures in order to achieve strains and strain

rates similar to the real industrial conditions. However, one of the main practical

problems is the cooling rate after deformation, since the microstructure changes already

within a few seconds. In order to enable comparison of results of various deformation

tests performed under hot conditions, it is advantageous to evaluate them in the form of

a dependence of real stress and real strain (equivalent stress and strain).


55

The reason for which classic tests are not used to test properties after TMP in real

conditions is above all the expenses. The casual devices are highly productive.

Therefore, modification of a production process or an implemented material

development process would be very ineffective. By this reason TMP is usually tested

under laboratory conditions.

Modelling has several advantages, among which is e.g. the possibility to investigate

tribological and geometrical deformation conditions. On the other hand, achievement of

the same conditions as in the real process is difficult. For example cooling conditions are

different for most of the processes. Moreover, high strain rates occurring in practice

cannot be achieved under laboratory conditions. By these reasons, it is advantageous to

perform simple laboratory material tests before a TMP is performed under real

production processing conditions. The result of such an approach is determination of the

influence of typical industrial conditions (temperature, deformation, strain rate, idle

times between passes, annealing times etc.). Such conditions are usually determined

using FE (finite element) simulations.

3.2 Computer modelling

The aim of computer modelling is to quantify the basic mechanisms influencing a

thermomechanical forming process, i.e. especially changes of temperature, strain, strain

rate and time, and implement them into general mathematic relations for industrial

forming processes. Modelling of microstructure is at present quite well developed in the

field of heat treatment of steels.

Considering possible examples of steel forming under hot conditions, it is evident

that the conditions for modelling are complex. A modelled forming process itself

consists of gradual deformation reductions, which are performed under defined velocity

and temperature conditions with defined dwell times between the individual reductions.

Modelling of such a process thus demands a multidisciplinary approach including

physical metallurgy, metallurgical engineering, mechanical engineering, applied

mathematics, mathematics and system engineering and so on. This is due to the fact that

a general model consists of several separate, however mutually integrated, sub-models.


56

Task

Using the given values create a dependence of stress on strain. By means of exponential

regression derive a strengthening relation describing the created dependence.

This task is performed using MS Excel software. The values to be entered are

given in two columns – values of strain ε within the range of 0.01 – 1.5 with 0.01

step and corresponding values of stress. The data were experimentally obtained

using a torsional test. An example of the given values is shown in Table 2. These

values are further used to create a 2D dependence (point with smooth lines type).

Table 2 Example of the first 10 given stress and strain values

[-] 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

σ

[MPa] 217.56 226.02 231.12 234.80 237.70 240.10 242.14 243.93 245.51 246.94

The subsequent task is to create a strengthening relation for the stress-strain

dependence, i.e. an equation for calculation of stress depending on strain. This

equation can be acquired by clicking with the right mouse button on the curve in

the dependence and selecting addition of a trend line. From the line types it is

necessary to select exponential line, since the strengthening equation is of an

exponential type. The equation can further be shown in the created dependence

together with the correlation coefficient R2. The more the R2 value is close to 1,

the more precise is the description of the curve using the equation.

According to the regression analysis of MS Excel software, the strengthening

equation describing the curve is in the following form (55).

055025280 ,, (55)

In this case, the correlation coefficient is equal to 1. Check can be performed by

backwards calculations. It is necessary to write the final equation with a variable

value for deformation according to the given values into another column. The

checking calculation is shown in Table 3.


57

Table 3 Example of the first 10 given stress and strain values

[-] 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

σ

[MPa] 271.56 226.02 231.12 234.80 237.70 240.10 242.14 243.93 245.51 246.94

The dependence created on the basis of the given stress and strain values,

together with the derived equation of exponential dependence and the

correlation coefficient, is depicted in Figure 15.

Figure 15: Dependence of stress (deformation resistance) on strain.

References

[1] HUMPHREYS, F.J., HARTLEY, M. Recrystallization and related annealing

phenomena. 2nd ed. Oxford: Pergamon; 1996, 617 p. ISBN 978-0080441641.

[2] VERLINDEN, B., DRIVER, J., SAMAJDAR, I., DOHERTY, D.R. Thermo-Mechanical

Processing of Metallic Materials, Pergamon Materials Series- series ed. R.W. Cahn,

Elsevier, Amsterdam, 2007, 332 p. ISBN 978-0-08-044497-0.


58

[3] JONAS, J.J., SELLARS, C.M. Thermomechanical Processing. In Future developments

of Metals and ceramics. (Edited by J.A.Charles, G.W. Greenwood and G.C. Smith),

London Institute of Materials, 1992, 150 p., ISBN 09017-16030.

[4] WEISS, I., SEMIATIN, S.L. Thermomechanical processing of alpha titanium alloys-

an overview, Materials Science and Engineering A263, 1999, p. 243-256.

Date post:	03-Apr-2018
Category:	Documents
Upload:	vothien
View:	216 times
Download:	2 times

Study Support - vsb.czkatedry.fmmi.vsb.cz/Opory_FMMI_ENG/MMT... · Thermo-mechanical forming...

Documents