1

Design of Forming Processes: Bulk Forming

Chester J. Van Tyne

Colorado School of Mines, Golden, Colorado, U.S.A.

I. BULK DEFORMATION atures relative to the melting point of the metal. Hot
working occurs at temperatures above tJllerecrystalliza-
Bulk defonnation is a metal-fonning process where the tion temperature of the metal. There is a third temper-
defonnation is three-dimensional in nature. The pri- ature range, warm working, which is being critically
mary use of the tenn bulk deformation is to distinguish it examined due to energy savings and is, in some cases,
from sheet-fonning processes. In sheet-forming opera- used by industries.
tions, the defonnation stressesare usually in the plane of
the sheet metal, whereas in bulk defonnation, the 1. Cold Working Temperatures
defonnation stresses possess components in all three
Cold working usually refers to metal deformation that is
coordinate directions. Bulk defonnation includes metal
carried out at room temperature. Th,~ phenomenon
working processes such as forging, extrusion, rolling,
associated with cold work occurs wht:n the metal is
and drawing.
deformed at temperatures that are about 30% or less of
its melting temperature on an absolute temperature
scale. During cold work, the metal ,~xperiences an
II. CLASSIFICATION OF DEFORMATION
increased number of dislocations and elltanglement of
PROCESSES
these dislocations, causing strain hardening. With strain
hardening, the strength of the metal increases with
The classification of deformation processescan be done
deformation. To recrystallize the metal, ;i thermal treat-
in one of several ways. The more common classification
ment, called an anneal, is often needed. During anneal-
schemesare based on temperature, flow behavior, and
ing, the strength of the metal can be drastically reduced
stressstate. The temperature of the deformation process
with a significant increase in ductility. The ductility
is under direct control of the operator and has a
increase often allows further deformation to occur
profound effect on the viability of the process and the
before fracture. The final surface finish alld dimensional
resulting shape and microstructure of the finished prod-
tolerances can be well controlled in a cold work process.
uct. The flow behavior and the stress state differ from
temperature in that they are a result of the actual
2. Hot Working Temperatures
deformation process that one chooses.
Hot working l;)CCursat temperatures of 61>%or above of
A. Temperature Classification the melting temperature of the metal on an absolute
scale. At elevated temperatures, the metal has decreased
The temperature classification scheme is normally di- strength, hence the forces needed for deformation are
vided into two primary regions-cold working and hot reduced. Recrystallization occurs readily, causing new
working. Cold working occurs at relatively low temper- grains to continually form during deformation. The

j
2 Van Tyne

continual formation of new grains causesthe ductility of analysis can be applied. For a continuous-flow process,
the metal to remain high, allowing large amounts of a more complex analysis needsto be used to simulate the
deformation to be imparted without fracture. Control process accurately. The complex analysis needs to ac-
of final dimensions is more difficult in a hot-worked count for the continually changing shape of the defor-
metal due to scale formation and volumetrical changes mation region.
in the part during subsequent cooling.

C. Stress State Classification

3. Warm Working Temperatures
In all bulk deformation processes, the primary defor-
Warm working occurs between hot working and cold
mation stress is compressive in nature. This is in con-
working. It occurs in the approximate temperature
trast to sheet metal forming where 1:ensilestresses are
range of 30-60% of the melting temperature of the metal
often used. Stress state classification consists of two
on an absolute scale. The forces required to deform
categories for bulk deformation-direct compression
metal in the warm working regime are higher than dur-
and indirect compression. In direct compression, the
ing hot working. The final finish and dimensional toler-
tools or dies directly squeeze the workpiece. Forging,
ancesare better than hot working but not nearly as good
extrusion, and rolling are examples of direct compres-
as a cold working process. Although warm work seems
sion processes. In indirect compression, the deforma-
to have drawbacks, the primary driver for warm work-
tion region of the workpiece is in a \;ompressive stress
ing is economic. There is significant cost in heating a
state but the application of these compressive stresses
metal up to hot working temperatures. If the working
occurs by indirect means. Wire drawing is an example of
temperature is lowered, there can be major cost savings
an indirect compression process, where the wire is pulled
in the process.
through a die. The workpiece conta(;ts the converging
surfaces of the dies, creating high forces normal to the
B. Flow Behavior Classification die surface. The dies react to these J[orcesby pushing
back on the workpiece, causing a compressive stress
The flow behavior of a metal or alloy during bulk defor- state to exist in the deforming region of the metal. Thus
mation processes falls into one of two categories- although the equipment action is of a tensile (pulling)
continuous flow or quasi-static. The easiest way to dis- nature, the plastic deforming region is being squeezed.
tinguish between these two types of flow is to imagine a It should be noted that although the stress state for
movie being made of the deformation region during bulk deformation is compressive, there are situations
processing. If the shape of the deformation region where tensile stress components may be present within
changes during each frame of the movie, the process is the workpiece and fracture may o.ccur. The metal-
a continuous-flow process. If in each frame of the movie forming engineer needs to be aware of these types of
the shape of the deformation region remains the same, situations and to properly design the' process to avoid
even though a different material is in the region, it is a the potential fracturing that can occur on the workpiece
quasi-static-flow process. The bulk deformation process due to the tensile stresscomponents. For example, in the
of forging is an example of a continuous-flow process. forging of a right circular cylinder between two flat dies
As the metal is being shaped in the forging die cavity, the in the axial direction, if friction on the top and bottom
deforming region, which is often the entire amount of surfaces is high, the sides of the cylinder will bulge and
metal, is continuously undergoing change. Processes some tensile hoop stress may occur on the outside
such as rolling, wire drawing, and extrusion are exam- surface of the workpiece. A more insidious example is
ples of quasi-static flow. For example, in rolling, the an extrusion process where a small reduction is per-
deformation region is the metal being squeezedbetween formed through a die with a high diie angle. For this
two rolls. The shape of the deformation region does not situation, the deformation region ma~' be limited to the
vary , aside from initial startup and final finish, although surface region of the workpiece, causing some internal
different material flows into and out of the region. tensile stress components along the centerline of the
The classification based on flow is useful in deter- workpiece. If the internal tensile stress components
mining what type of modeling scheme can be used to become excessively high, they can (;ause an internal
simulate the bulk deformation process. For a quasi-stat- fracture in the workpiece. This fracture is referred to
ic-flow process, the deformation region can often be as central burst. The worst aspect of central burst is that
handled as a single region and a steady-state type of it cannot be detected via visual methods.
Design of Forming Processes: Bulk Forming 3

III. TYPES OF BULK DEFORMATION a) b) ,:)

PROCESSES

A. Forging

Forging is a metalworking process where a workpiece is ~

shaped by compressive forces using various dies and
tools. The forging process produces discrete parts. Some Figure 2 Open die forging tools: (a) fiat dies; (b) U-shaped
finishing operations are usually required. Similarly dies; and (c) V-shaped dies. (From Ref. 2.)
shaped parts can often be produced by casting or pow-
der metallurgy operations, but the mechanical prop-
erties of a forged component are usually superior
compared to other processing methods. Forging can growing in popularity due primarily to thermal energy
be done hot or cold. Warm forging is a process that is costs. Typical forged parts are shown in Fig. I.
Open die forging consists of dies ,vith very simple
geometry. The dies are usually flat, U-shaped, or v-
shaped, as seen in Fig. 2. The shaping of the metal
a) e) occurs through manipulation of the workpiece and skill
of the operator. Itisa process that is usi:ful in producing
a small number of pieces. It is difficult to hold to close
tolerance in this type of forging. Open die forging be-
tween two flat dies is often called upsetting. Cogging or
b) drawing out is an open die forging process where the
thickness of the workpiece is reduced by successive
small strikes along the length of the :metal. Open die
forging is closely related to blacksmithing.
Closed die or impression die forging consists of a die
I) set with a machined impression, as shown in Fig. 3.
c)
There is good use of metal in this operation as compared
to open die forging. Excessmetal beyond the size needed
for forging is used and flows into the ~:utter portion of
the die set to produce flash. The excess metal helps to
insure that the cavities are completely filled at the end of
the press stroke. Good tolerances and accuracy of the
final forging are attainable. The die CO!;tsfor closed die
g) forging are fairly high due to their property require-
ments and machining costs.
Closed die forging often occurs in a s(~quenceof steps.
d) Each step of the operation usually has its own impres-
sion in the die block. The first step distributes metals
into regions where extra volume is required in the final
component. This step often involves ed!:ing, where extra
metal is gathered, or fullering, where metal is moved
h)
away from the local region. In hot forgiDg, the first step
is r~ferred to as busting because the scale on the surface
of the workpiece is busted off. The second step is
blocking, where the part is formed into a rough shape.
The third step is finishing, where the final shape of the
component is imparted to the workpiece. The fourth
Figure 1 Examples of forged parts: (a) automotive crank-
step is trimming, where the excess m(~tal in the flash
shaft; (b) truck axle; (c) truck bracket; (d) universal joint; (e)
region is trimmed from the componenl:. Figure 3 illus-
automotive gears; (f) truck assembly part; (g) coupling
fittings; and (h) hammer head. (From Ref. I.) trates these various steps.
4 Van Tyne

Mechanical presses are stroke-limited equipment with

a large flywheel powered by an electril;al motor. The up-
and-down motion of the ram is handled via a connecting
rod attached to a crank shaft. The tra,'el distance during
each press stroke is controlled by machine design and
operation. Hydraulic presses are load-limited equip-
ment where the press will stop once ilts load capacity is
reached. The power comes from pressurized hydraulic
fluid. Screw presses, similar to haIDJners, are energy-
limited equipment. A large flywheel transmits power
through a vertical screw, which causesthe ram to move.
The ram movement stops when all tht~energy from the
flywheel has dissipated. Table 2 provides some numer-
ical details about hydraulic pressesto produce the same
gear blank as in Table 1.
In order to be successful in forging a metal, the
formability of the metal needs to be understood, espe-
cially with regard to temperature and speed. The im-
pression die shape needs to be carefully designed and
Finished Product
machined to allow a good flow of metal without seams
or laps developing. The die material needs to be care-
Figure 3 Impression forging dies with forging sequence.
fully chosen to match the metal bein~: shaped and the
(From Ref. 3.)
temperature of the operation.

B. Extrusion
The machines used for forging are hammers and
presses. Hammers are energy-Iimited equipment and Extrusion is a bulk deformation proce~s where a billet,
can be a simple gravity drop machine where a free- generally cylindrical, is placed in a chamber and forced
falling ram strikes the workpiece. Augmentation of the through a die. The die opening can be round to produce
energy supplied to the hammer can be done in the form a cylindrical product, or the opening ~Ln have a variety
of pressured air, steam or hydraulic fluid. In a hot closed of shapes. Typical extrusion products are shown in
die operation, multiple blows are usually needed during Fig. 4. Because of the large reductions imparted during
each step, especially the blocking and finishing steps, the extrusion process, most extrusioll processes are
when using a hammer to forge metal. Table I provides performed hot in order to reduce the flow strength of
some numerical details about hammers for a typical the metal. Cold extrusion can occur but it is usually one
gear blank forging. step in a multistep cold forging operation.
For forging, there are three types of presses used- Forward or direct extrusion is where the billet is
mechanical press, hydraulic press, and screw press. pushed from the backside and the front side flows

Table 1 Characteristics of Hammers for Forging a 4.45-lb Steel Gear Blank

Design of Forming Processes: Bulk Forming 5

Two-hundred-fifty-ton press stalled and left underfilled on outer diameter.

Fast 2000-ton press is similar to mechanical or screw press.
Smaller presses resulted in increased die temperature.
Source: Ref. 4.

through the die. Indirect or inverse or backward extru- formability of the metal being deformed. Excessive
sion is where the die, which imparts shape, moves into temperature, speed, or friction can cause surface cracks
the billet. The equipment used to perform an indirect to propagate along grain boundaries, wJlich are referred
extrusion is more complex than for a forward extrusion. to as fir tree cracking, due to hot shortness of the metal.
To overcome the significant friction resistance between Improper geometrical configuration of the tooling can
the billet and the chamber in a forward extrusion, cause central bursts if the angle of the die opening is too
hydrostatic extrusion has been developed. In hydrostat- large, or the reduction is too small. Piping or cavitation
ic extrusion, the billet is smaller than the chamber and is at the end of the extrusion can be minimized by reducing
surrounded by hydraulic fluid. The hydraulic fluid is the severity of the distortion in the product, or by
pressurized, which squeezes the billet through the die reducing friction.
opening. Caution with both the sealing of the fluid and
at the end of the process, where the final part of the billet
could become a high-velocity projectile, needs to be c. Rolling
exerted. Impact extrusion is similar to indirect extrusion
and is often performed cold. The tooling, usually a solid Rolling is a direct compression defonnation process,
punch, moves rapidly into the workpiece, causing it to which reduces the thickness or changes the cross section
flow backward and around the face of the punch. This of a long workpiece. The process occurs through a set of
produces a tubular-shaped type of product. These types rolls, which supply the compressive forces needed to
of extrusions are schematically shown in Fig. 5. plastically defonn the metal. Flat rolle:d products are
The equipment for extrusion is normally a horizontal classified as plate, sheet, or foil, dependillg on the thick-
hydraulic press. A large shape change is imparted to the ness of the product. A plate has thickness greater than 6
billet during a single stroke of the press. The sha~ mm, whereas a foil has thickness less than 0.1 mm. A
change causes significant distortion in the metal during sheet has thickness between that of the plate and the foil.
the deformation. Rolling can be done hot or cold. In Ioany products,
For successin extrusion, the temperature and speed initial reductions are perfonned hot, ~'here the metal
of the process need to be determined based on the can experience large shape changes without fracturing,
and the final reductions are perfonned cold, so that
better surface finish and tolerances can be achieved.
Flat rolling reduces the thickness of the metal, pro-
ducing a product with flat upper and lower surfaces.
Shape rolling can also reduce the thickn~ss of the metal
but, more importantly, it imparts a more complex cross-
section shape. Shape rolling can be used to produce
bars, rods, I -beams, channels, rails, etc. lUng rolling can
be used to produce a seamless produ,[;t by reducing
the wall thickness of a ring through thl: action of two
rolls. Seamless pipes can be produced and sized by
Figure 4 Examples of extruded parts. (From Ref. 5. specialized rolling operations such as rotary tube pierc-
6 Van Tyne

b)

Figure 5 Schematics of extrusion processes: (a) direct or forward extrusion; (b) indirect or reverse extrusion; (c) impact extrusion;
and (d) hydrostatic extrusion. (From Ref. 6.)

ing, tube rolling, and pilgering. A variety of rolling pro- is not uniform across the width of tht~metal. Crowned
cessesfor steel are schematically shown in Fig. 6. rolls, six-high mills, and sleeved rolls can be used to
Although large, the rolling mill equipment is relative- correct these types of defects by properly controlling the
ly simple. A two-high mill consists of two rolls, and a amount of roll bending that occurs. Small amount of
three-high mill consists of three rolls, which also allows waviness in a sheet product can be eliminated by a
reduction to occur on reverse directional flow of the postdeformation leveling operation, where the sheet
metal. A four-high mill consists of two work rolls in passes over a series of rollers whil,~ under tension;
contact with the metal and two back up rolls. A six-high Alligatoring or fish tails can occur at the front end or
mill is like a four-high mill, but has two additional rolls back end of the workpiece. Proper alignment of the feed
between the work roll and backup roll called intermedi- stock into roll gap, proper balancing of the friction
ate rolls, which allow in essencesome control over the between the top and bottom rolls, and proper choice of
crown and camber of the work rolls. Cluster mills exist roll size for reduction can be used to minimize or to
usually for the production of thin foil products. A cluster eliminate these two types of defects.
mill will have a pair of small-diameter work rolls and
a series of intermediate and backup rolls to support D. Drawing
the work rolls. A tandem rolling mill will have a series
of rolling stands where each stand imparts a specific Drawing of a round rod or wire is an indirect compres-
amount of reduction. The operation of a tandem mill is sion process where the cross-sectional ;lrea of the metal
challenging due to coupling effects between the stands. is reduced by pulling it though a converging die. A
Defects can be present in sheet and plate products if schematic illustration of wire drawing is seen in Fig. 7.
the rolling operation is not performed correctly. Wavy The process is normally done at ambient temperatures.
edges, waves along the centerline, zipper cracks along The major factors that need to be cotltrolled include:
the centerline, or edge cracks can occur if the reduction reduction, die angle, friction at the dje-workpiece in-
 Design of Forming Processes: Bulk Forming
7

Hot strip Pickling and Cold strip

~ ~~

~te Steel plates

Cold-drawn
Continuous castin~
Hot-rolled bars
~

Billet ~ -~ WIre and wire

~Rods -~ products

-L~~~ ~I ~ Seamless pipe

Bloom ~

Structural shapes

-
Rails

Figure 6 Schematics of various rolling processes for steel. (From Ref. 7.)

terface, and drawing speed. Tubes can also be drawn in become excessive, then the wire would fracture in a
a similar process. To control the interior diameter of a mode similar to a tensile test. The limit on the value of
tube, a mandrel, which can be fixed, moving, or floating, the tensile stress that can be supported limits the amount
is used. Because the metal is pulled through the die, the of reduction that can be achieved in one pass. Multiple
final product, which has the reduced cross section, is reduction passeswith multiple dies are nC(:dedto achieve
subjected to tensile stresses. If these tensile stresses large reductions in cross-sectional areas. The approach
is analogous to a tandem rolling mill with multiple
stands. The theoretical maximum reduction for a fric-
tionless, perfectly plastic material is 63~/o. In produc-
tion processing, the reduction that is used is often
limited to 35% or 40%. The ironing pr(~ss, which is
used to reduce the wall thickness of a sheet metal, is also
a drawing-type operation.
The configuration of the opening in the final die will
control the configuration of the product produced.
Although a cylindrical shape is the most common, other
shapes can be imparted to the wire in the process.
The metal is cold-worked during the wire drawing
process and intermediate anneals may be needed to
increase its ductility to sufficient levels in ,~rder to reach
Figure 7 Schematic of a wire drawing process. the final reduction desired. Internal fractures, called
 8
Van Tyne

central busts, can occur if the die angle is too large, or Plastic defonnation is often measured by the engineer-
the reduction is too small. For rods, tubular products, ing strain:
or high-strength wires, postdeformation straightening
may be required.
(3)

IV. PROCESSING ASPECTS

A. Temperature 6=ln (~1=ln
L.~ ( -AO'
01 A., (4)
In bulk working operations, thermal energy is often Often the measure of deformation for bulk deformation
supplied to the workpiece to increase its temperature. processesis expressed by the reduction in area:
There are a number of methods used to heat up metal
workpieces. Heating in a gas-fired furnace, induction R -Ao -A.
---5 ( )

heating, and electrical resistance heating are the most Ao

common methods that are used in industries. The op- For forging, the equations will be similar:
eration and control of the heating process are critical
Aoho = A1h1
features in controlling the deformation process. The (6)
workpiece needs to be at the proper working tempera- ho -h1 At -Ao
e= --=
ture in order to achieve the desired shape change and to (7)
have the proper microstructure for deformation. (
ho
ho '
In -=In
Ai
(~--.!.
A] ) = In(e+ 1)
The deformation in the workpiece is produced by 8
(8)
hi/ Aol
mechanical work. Most of the mechanical work im-
parted into the workpiece during deformation is con- It should be noted that these equatjons are simplified
verted into heat. The heat causes the workpiece to measures for strain during the prOCt:ss.In bulk defor-
increase in temperature. The maximum possible in- mation, the strain in the workpiece will usually vary
crease in temperature is often referred to as adiabatic from point to point, and for a continuous-flow process,
heating and is calculated by assuming that the entire the strain will also vary at each time instant in the
amount of mechanical work is converted in the temper- process. In its true form, strain is a se<:ond-order tensor,
ature rise. The adiabatic temperature rise for a bulk which, during deformation, has si:t unique compo-
deformation process can be calculated by: nents-three normal components and three shear com-
ponents. In deformation operations, strain is often
w expressed by its three principal comIlOnents 8i, 82, and
~T= (I)
pCp 83. For deformation processes, which have undergone
proportional loading, the effective strain at a point in
where Wis the mechanical work per unit volume for the the workpiece is often given by the Mises equivalent
deformation process, p is the density of the workpiece, strain:
and Cp is the heat capacity for the workpiece.

B. Strain 8=~ (9)

During bulk plastic deformation.. a shape change is c. Strain Rate

imposed on the workpiece. Strain is the normal measure
to quantify the amount of deformation. In operations
During defonnation processes, the speed of the opera-
such as rolling, extrusion, and wire drawing, the cross- tion is usually measured by strain rat~. Strain rate i is
sectional area A of the workpiece normally decreasesas the time rate of the change of strain:
the length L increases. In forging, the opposite usually
occurs where the cross-sectional area increases and the
i=~=!~=~ (10)
height h of the workpiece decreases. dl L dl L
In most forming operations, the volume of the work- where v is the velocity.
piece remains constant. The constancy of volume is Strain rate is an important variable because the
expressed as; strength and microstructural respons(: of many metals
is dependent on the strain rate. Like strain, strain rate in
AoLo = AiL
(2) its true fonn is also a second-order tensor. The effective
 Design of Forming Processes: Bulk Forming
9

strain rate at a point in the workpiece can be expressed The effective stress at a point within the workpiece is
as: given by:

where il, i2, and i3 are the principal strain rate com-
(14)
ponents of the strain rate tensor .
If the effective stress at a point within the workpiece has
reached the value of the flow strength of the material at
D. Stress
that point, then plastic flow will occur.
If the effective stress and effective str;lin are known
In bulk deformation operations, stress has two mean-
for the deformation process, then the "work per unit
ings. The first meaning of stress is related to the equip-
volume of material for deformation W can be deter-
ment used to deform the workpiece. It is a measure of
mined by
the load requirements necessary to get the workpiece to
plastically deform. This is an important aspect that w= Jade (IS)
needs to be considered because the sizing of the equip-
ment for bulk deformation is fundamentally dependent Another important stress measure is the mean stress
on the load requirements for plastic flow. component or hydrostatic stress component:
The second meaning of stress is related to the work-
piece. During deformation, each point in the workpiece (1M= i (11+ (12+ (13) (16)
has a stress state, which is a measure of the ma:terials'
internal resistance to the externally supplied forces. For deformation processes, the stress components
These two meanings are interrelated. must be of a sufficient deviation from tIle hydrostatic
In bulk metalworking operations, the external loads stress to cause plastic flow to occur. A pure hydrostatic
supplied are often compressive in nature. Wire drawing stress cannot cause plastic flow to occur within a nor-
is an exception, where the supplied load is a tensile force. mal material.
For compressive deformation processes, the pressure
required for deformation usually describes the external E. Friction
stress. The pressure can vary from point to point along
the tool-workpiece interface, often due to the friction During bulk defoffi1ation processes, frictional resist-
resistance present. An average pressure for deformation ance to sliding occurs at the interface between the
to occur is: workpiece and the tooling. The frictional resistance is
due to the surface asperities that are present at the
F
PAVG= A (12) microscale on both the tools and the workpiece. These
asperities impede the sliding motion tblat can occur
where Fis the force or load supplied by the equipment, during contact under pressure. Figure 8 schematically
and A is the area over which the load is being supplied. shows how the asperities interact to impede motion.
For wire drawing, a similar equation can be used, but it
determines the average drawing stress on the wire being
pulled through the die:
F
(1AVG= A (13)

The internal resistance within the workpiece to these

external loads varies from point to point. The measure
of this resistance is the internal stress that exists in the
workpiece. If the specific point in the workpiece under- b)
-
goesplastic deformation, then the internal stress is equal
to the flow strength of the material at that point. Plastic Zone
Internal stress, such as strain and strain rate, is a sec- (Microweld)
ond-order tensor. This second-order tensor has six com-
ponents-three normal components and three shear Figure 8 Schematic of frictional resistance and wear on
component. The stress tensor is often expressedin terms sliding metal surfaces: (a) interactions of asperities; and (b)
of the three principal components (11,(12,and (13. localized plastic deformation. (From Ref. 8.)
10 Van Tyne

Friction causes the required deformation loads to will occur when the uniaxial tensile stress reaches the
increase. Friction causes the flow of the material to be metals' yield strength. For bulk deformation processes,
less homogeneous. High levels of friction can result in the stress state is not a simple uni~LXialstate, hence the
surface damage to the workpiece, or seizing of the criteria for yielding are more complex relationships.
workpiece to the tooling. The Tresca yield criterion or maximum shear stress
Frictional resistance is usually described by a shear criterion indicates that plastic flow will occur when:
stress component 'CF.There are two basic models that
are used describe the frictional stress component that Tmax= ~(O". -0"3) = 0"0 (19)
occurs during metalworking operations. Both of these
where O".is the largest principal component of the stress
models are highly simplified and only capture the major
state, 0"3is the smallest principal cot(lponent of the stress
aspect of the very complex interaction that occurs at the
state, and 0"0is the flow strength of the metal. IfEq. (19)
tool-workpiece interface.
is satisfied, then plastic deformatio:(l will occur .
The first model is referred to as Coulomb's law. The
A more generally applicable criterion is the Mises
frictional stress component is directly proportional to
criterion or maximum distortion eDt~rgycriterion, which
the pressure that exists between the tool and the work-
IS:
piece at the point of interest, or:
tF = IJ.p (17)
=qo
where p.is the coefficient of friction. The value of p. can ~
vary from to 1//3 (i.e., 0.577). At low-pressure levels,
this equation is a good description of the frictional Other criteria for the relationship t>etweenthe applied
stress component. stress state and the flow strength of the metal, which can
The second model is a better description at higher cause plastic deformation, do exist, but the two equa-
pressures at the interface. It is referred to as the constant tions given here are the ones most often used to describe
friction factor equation. It assumes that the frictional bulk deformation processes.
stresscomponent is some fraction of the flow strength Uo In three-dimensional principal stress space, both
of the workpiece: yield criteria will plot as surfaces. Thus the yield criteria
are often called the yield surface j:Or the metal. The
tF = m~Uo (18) surface for the Tresca yield criterion is a hexagonal.
shaped prism, whereas the surface ;[or the Mises yield
where m is the constant friction factor. The value of m criterion is cylindrical. If 0"3= 0, then the yield surface
can vary from 0.0 for an ideal frictionless interface to 1.0 reduces to yield loci curves in thl: two-dimensional
for an interface where full sticking between the work- 0"1-0"2space. Figure 9 shows the relationship between
piece and tool occurs. the Tresca and Mises yield criteria in this reduced two-
Friction is controlled through lubrication. The role dimensional space.
of the lubricant in metalworking is important in reduc-
ing frictional resistance. Lubrication can also playa G. Hardening
vital role in cooling the tooling, preventing heat flow
from a hot workpiece into the tooling and protecting the During cold work, the metal increases in strength with
new surfaces created during the deformation from increased deformation. This phenomenon is referred to
oxidation or chemical reactions.

/ Mises Yield C:riterion

F. Yield Criteria 0"1

The ease with which a metal flows plastically is an

important factor in deformation processes. The domi-
nant factors that influence the flow (or yield) strength of
a metal are the temperature and the amount of prior
cold work. Yield criterion is the relationship between
the stress state and the strength of the metal. When the
criterion is met, then plastic deformation occurs. In Figure 9 Comparison of Tresca and Mises yield criteria in
uniaxial tensile tests, the yield criteria predict that flow reduced principal component stress space.
Design of Forming Processes: Bulk Forming 11

as hardening. Plastic hardening in metals is often rea- tion, or lead to process-induced defe<:ts in the final
sonably well characterized by a power law equation, product being produced.
where the strength is dependent on the amount of plastic In extrusion, rolling and drawing the size and shape
strain imposed: of the deformation zone have a strong influence on a
variety of forming parameters, such as friction work,
0"0= Ken (21)
redundant work, and deformation loads, as well as
where K is a strength coefficient for the hardening properties in the formed part, such as internal porosity,
behavior and n is the strain hardening exponent. These internal cracking, distortion, homogeneity of strength,
two material parameters are usually obtained via a and residual stresses.A common single parameter mea-
tensile or a compression test. Equation (21) indicates sure of the deformation zone geometry is the L\.param-
that the metal strengthens as the strain increases, which eter. The L\. parameter is defined as the ratio of the
is isotropic hardening. In isotropic hardening, the yield average thickn~ss or diameter h of the deformation
surface is continually expanding with strain. If the strain region to the contact length L between l:he tooling and
path imposed on the metal during deformation is the workpiece, or:
changed (e.g., if it is reversed), the yield strength on
reversal may be different than expected for the strain L\.= ~ (22)
imposed before the change. This difference is a mani-
festation of kinematical hardening, where the center It has been found that deformation unde:r conditions of
point of the yield surface moves with strain. Figure 10 high L\.parameters can lead to microporosity along the
shows the difference between the yield surface changes center line of the workpiece, or, in extreme cases, can
that occur for isotropic hardening as compared to lead to internal cracks. Caution needs to be used when
kinematical hardening. L\.> 2 because it is this condition that can lead to prob-
lems. Figure II shows data from an extrusion process
that exhibit both sound flow behavior and central burst.
V. DESIGN ISSUES TO PREVENT FAILURES Flow localization can occur in the workpiece during
A. Geometrical and Mechanics Issues deformation. The common cause of flo~' localization is
a dead metal zone between the workpiece and the
The shape of the tooling and the initial shape of the tooling. Poor lubrication in forging can cause sticking
workpiece are important geometrical factors for bulk friction between the die and the workpiece, and in the
deformation processes. Incorrect choices of these geo-
metrical factors can lead to problems during deforma-

0"1

Die Angle (a in degrees)

Figure 10 Comparison of isotropic hardening to kine-
matical hardening for a Mises material in reduced principal Figure 11 Criteria of the prevention of central burst in
component stress space. extrusions. (From Ref. 9.)
12
Van Tyne

sticking region, a dead metal zone can occur. Forging be locally chilled during hot working processes.Chilling
dies, which are cooler than the workpiece, can extract can lead to strength variations in thl~metal and cause the
heat from the metal, causing localized cooling. The promotion of shear banding.
metal at a lower temperature has higher flow strength Cold working causesthe strength of the workpiece to
and is more resistant to plastic deformation, which can increase during deformation. Thu:; regions where sig-
lead to a dead region in the workpiece. In extrusion, nificant cold work has been imparted to the metal are
dead metal zones can occur due to very large die angles regions of higher strength. These strength variations can
and the metal will shear over itself, leaving a dead metal lead to internal shear banding. TIle grain size of the
region adjacent to the die. workpiece also can have an influeDl;e on the final prod-
In closed die forging, the width and thickness of the uct produced. Working of large graJinedmetals can lead
land region are very important parameters. The land to a surface roughening phenomt:non called orange
region is the choke point for metal flowing into the flash peel, which is usually undesirable.
region of the forging. As multiple parts are forged, the
land will wear away. The small thickness and large
width of the land opening provide restrictive flow into VI. WORKABILITY AND TESTlf\IG METHODS
the flash and cause increased pressure to occur in the die
A. Definition
cavity. The increased pressure in the cavity allows for
better filling of the impression, but at the cost of higher
Workability is a characteristic that is usually attributed
load requirements. If the flow of the metal inside of a
to the metal or alloy, It is a relati"e measure of how
cavity during forging is not properly controlled, a lap,
a flow-through defect, or a suck-in defect may occur. A easily the metal can be plastically deformed without
fracture. It should be noted that workability depends
lap is where the metal folds back on itself. A flow-
not only on the metal itself, but also on other external
through defect occurs when the metal is forced to flow
across a recessin the die that is already filled. A suck-in processing factors. The temperature and stress state
imposed by the processing conditions will strongly
defect occurs when there is too much metal flow into a
influence workability. Most metals have high higher
centrally located rib region. These types of defects can
workabilities at higher temperatur,~s, Workability is
be avoided or minimized by proper redesign of the
die cavity. usually higher under compressive :)tates of stress as
compared to tensile states. Terms such as formability,
B. Metallurgical and Microstructure Issues forgeability, extrudability, and dra'wability are often
used to describe the workability within a specific
The common failure modes that occur in cold work metal-forming process,
deformation processes include: free surface cracking,
B. Tests
shear bands, shear cracks, central bursts, and galling. In
hot work processes,the common failures are hot short-
A number of different mechanical tests are used to assess
ness,central bursts, triple-point cracks, grain boundary
the workability of a metal or alloy. l:ne best test is the
cavities, and shear bands. Metallurgical aspects and
one that most closely mimics the actual stress state that
microstructure features can have a strong influence on
would exist in the metal during the bulk deformation
the tendency of the workpiece to experience one of these
operation. Unfortunately, the optimum is often not the
failure modes.
easiest one to perform on the amount of material
Becauseof the segregation and cast microstructure in
available, or is constrained by the type of laboratory
ingots, these types ofworkpieces need to be hot-worked.
testing equipment available for use.
Due to chemical segregation and microstructural inho-
mogeneities, the properties of an ingot are not constant
1. Tensile Tests
from one location to another. Care must be taken to
provide enough deformation to break down the cast The tensile test is the most common t~:t used to evaluate
structure. Low melting point phases may also be present the mechanical properties of a metal or alloy. The tensile
and can lead to hot shortness if the temperature during test can be set at a variety of speedsto study strain rate
deformation is not carefully controlled. effects and a variety of temperatures to study the prop-
Hot working can lead to creep-type fractures, espe- erties of the metal as a function of temperature.
cially at slower working speeds in metals with low In a tensile test, a specimen of know Il initial geometry
workability. It is also important not to let the workpiece is placed in testing apparatus and pulled until fracture.
 Design of Forming Processes: Bulk Forming
13

The pulling load and the tensile elongation are mea- pure shear and large strains can be achieved without
sured throughout the test with a strip chart or com- flow localization and necking, which Occurs in a tensile
puterized data acquisition. Load and elongation are test, or barreling, which occurs during a compression
converted into engineering stress-strain data. From test. The test is suitable in providing flow stress and
the engineering stress-strain curve elastic modulus, the ductility data for materials as a function of strain, strain
yield strength, ultimate tensile strength, fracture stress, rate, temperature, and prior processing. The test is
and tensile elongation can be determined. Figure 12 frequently used to determine these maLterialproperties
shows a typical engineering stress-strain curve for a under hot working conditions. Because the strain rate
metal. After the test specimen is removed from the test- imposed on the material is proportional to the rotation-
ing apparatus, the final cross-sectional area in the frac- al speed of the test, high strain rates (up to 103sec-l) are
ture region can be measured and the reduction in area obtainable in a torsion test.
can be calculated. The reduction in area and the tensile Because a torque is being applied to the specimen
elongation are the two primary measures for the ductil- during the torsion test, the stress statt: in the material
ity of the metal. The ductility determined from a tensile will vary from the centerline to the surface of the
test is for the tensile stress state, temperature, and strain specimen. The variation in stressstate in a torsion-tested
rate imposed on the specimen during the testing. specimen is in contrast to the tensile and compression
The engineering stress-strain curve can be trans- tests where the stressstate in the deforming region of the
formed into a true stress-true strain curve for the metal. specimen is relatively uniform. The analysis of the
The transformation is valid between the yield point and torque twist data to produce stress-strain curves for
the ultimate point, where uniaxial plastic deformation the material needs to be done carefully. with an under-
occurs and localized necking has not occurred. The data standing of the test itself.
from a true stress-true strain curve can be plotted on a
log-log scale. From such a plot, the slope is the strain
3. Compression Tests
hardening exponent n and the intercept is the logarithm
of the strength coefficient K. Becau.semost bulk deformation proces1;esinvolve com-
pressive states of stress, a compression t~st is often more
2. Torsion Tests desirable in assessingthe workability of a metal that will
be deformed by such a process. In theol:y, the compres-
The torsion test is a fairly straightforward process. The
sive force imposed on the metal durin!~ a compression
specimen is held fixed on one end and the other end
test creates a uniaxial stress state within the metal. If this
is twisted at a constant angular velocity. The torque
were the case, then the analykis of the experiment would
needed to twist the sample and the angle of twist are the
be handled in a manner similar to the dal~aacquired via a
measured parameters. The deformation is caused by
tensile test. Unfortunately, the existent;e of a uniaxial
stress state in a compression sample is not achieved
because the specimen is compressed between two flat
platens. The compression causes the cross-sectional
60 area to increase and the friction that exists at the top
and bottom surfaces, where the specimen is in contact
.ii
c with the platens, causes nonuniform flow. The uncon-
.,
.,
~ 40 strained sides of the sample will show l;he nonuniform
(j)
0) flow by bulging. A bulged sample is a ,clear indication
"
."
4> that the stress state was not uniaxial.
.,
"
.6> To overcome this difficulty with friction, a variety of
" 20
W
specimen geometries have been used, as shown in Fig.
13. Each specimen is compressed and the compressive
strain in the axial direction and the diametrical strain
0 are measured. Measurement is usually performed by
imposing a grid onto the side surface of the specimen
and periodically stopping the test to measure the change
in dimensions of the grid pattern. Whc~na cylindrical
Figure 12 .Engineering stress-strain curve from a uniaxial specimen is compressed, the strain path that it follows
tensile test with material properties indicated. can be different, as shown in Fig. 14. The specimens are
 14 Van Tyne

a) a)
.I.
~ 020 Steel I. I. I. I 0.8
/ ,
c:
Top VIeW.

" )
m
303 Sta- Steel 06 :.
U)
.
"ij
"' ~ c:
~
...
-4 ~
E
e

F~ V.-; ill ,! .I. I. I. I.

~

.'.1
-..I

00
~
E
~
u

.1.0 .&.8 006 004 002 0.0

AxIal Compressive Slrain

b)
I. ~202~-T351. I. I. 10.4
Figure 13 Schematics of compression test specimen geo-
metries: (a) cylindrical sample; (b) tapered sample; and (c) c
flanged sample. (From Ref. 10.) ~ ~- 0.3 ~
(/)
.!!
~~I!. . "0;
c
..
1-
compressed until fracture occurs to assessthe metals' r02 ~
E
~~~~~ ~~ -
workability during compression and produce a forming I ;
O Cytk1dfical 5pec8nen ~ ~ E
limit curve. Typical fracture curves (or forming limit -01
, -
G

I!. T~ Sp.!(:in1en ~ U
diagrams) for 1020 steel, 303 stainless steel, and 2024-
..I.
.D Flanged ~ I. I. I. 1-0.0
T351 aluminum are shown in Fig. 15.
.0.5 .0.4 .03 .02 .0.1 00
4. Friction Tests Axial Compressive Strain

The most common method used to detennine the fric-

lion factor for a forging process is the ring compression Figure 15 Cold upset compression failure criteria: (a) 1020
test. The test can be conducted at varying temperature steel and 303 stainless steel; and (b) 2024-T351 aluminum.
and speed, and with the lubricant and workpiece mate- (From Ref. 11.)

rial of interest. The workpiece material is machined into

a ring with dimensions usually in a 6:3:2 ratio of the
outer diameter to the inner diameter to the thickness.
The ring is compressed in the thickness direction to a
given level of deformation and the new inside diameter
is measured. Friction calibration cuf1rescan be used to
determine the friction factor from the amount of defor-
mation imparted to the ring and thl~ change in inner
diameter (Fig. 16). Rings of other dimensions can be
used but the appropriate calibration curves must be
used for the specific starting geomet~r.

VII. DEFORMATION MODELING METHODS

A diagram illustrating the input and output as well as

Figure 14 Strain paths for compression tests of cylindrical the constraints, which must be considered when trying
specimens with various height (h)-to-diameter (d) ratios and to model a bulk deformation process, is shown in Fig.
various lubrication conditions. (From Ref. 11.) 17. The input parameters fall into three major catego-

~
 Design of Forming Processes: Bulk Forming
15

The stressesthat are generated within the workpiece

and the tooling must satisfy the equilibrium equations,
yield criteria, metal fiow properties, and stressboundary
conditions. Likewise, the strains geneJratedfrom these
stressesmust satisfy compatibility equations as well as
incompressibility requirements and aJtly imposed dis-
placement boundary conditions.
For a model to be exact and complete, all of the
requirements in Fig. 18 must be met f,or a given set of
input parameters. The complete and exact solution,
except in very simple cases, cannot be obtained. Often
it is necessary to simplify the model by allowing some of
the deformation mechanics requiremeDts to be relaxed.
Although this simplification does not give an exact
solutioti, the solution obtained is often quite reliable
for many processing situations. Simplifications are of-
ten necessary to obtain solutions. The amount of time
and effort one is willing to invest is of1:endirectly pro-
portional to the closeness of the solution to the exact
solution. To get extremely close, a lar~:e investment of
time, personnel, and funds is often neetied.
To describe each of the individual techniques, a
specific example will be used. The sample problem will
be the open die compression forging of a right circular
Figure 16 Ring test calibration curve for the determination cylinder between two fiat parallel platens (Fig. 19). This
of constant friction factor for rings with a 6:3:2 geometrical simple example is used primarily for illustrative purpo-
ratio of outer diameter/inner diameter/height. (From Ref. 12.) ses. It is equivalent to the initial breakdown (or pan-
caking) of an ingot or bar in an open die press or forge.
This problem will be examined via the slab equilibrium,
slip line, upper bound, and finite el:ement method
ries-geometrical parameters, process parameters, and (FEM) techniques. The methods describe herein can
material parameters. Constraints imposed by either the be applied to other bulk deformation processes.
product requirements or by the equipment should also
be considered and incorporated into the model. Often A. Slab Equilibrium
models flag situations where one of the constraints is
exceeded, rather than directly imposing the constraints. In the slab equilibrium technique, a small element (or
The result of the modeling effort is the determination slab) is extracted from the deforming 'workpiece (Fig.
of process geometry and process performance condi- 20). A force balance is performed on this small slab. This
tions. Models, especially if they are complex and ac- balance of forces leads to a differential t:quation, which
count for the fine details of the process, can take a long relates the stressesin the workpiece to the geometrical
time to run and often the results cannot be determined in variables of the process. With the use of a yield criterion,
"real time. " The models are normally used to provide a
an assumption of the principal stress directions, and
more detailed understanding of the process, rather than some knowledge of the boundary conditions, a solution
in a control scheme. For control of a specific bulk de- to the differential equation can be obtained. For simple
formation process, empirical models based on historical geometrical shapes, an analytical solution is often
operating data are often best suited for the task. achieved. For more complex shapes, the solution can
What occurs within the core of a model is shown in only be obtained by numerically solvinE: the differential
Fig. 18. In essence,the model must adhere to the laws of equation. The solution relates the actu.ll values of the
deformation mechanics. The relationships between pressure needed for deformation to the geometry, fric-
stress and strain both within the deforming metal as tion, and material properties.
well as within the tooling and at the interface between For the forging of a cylindrical dis}~, an analytical
the workpiece and the tooling must be obeyed. solution can be obtained for pressure a.sa function of
16 Van Tyne

~~
Analysis and Optimization j Determination of process
geometry and J:'focess
Loads, energies, stresses,
performance cc,ndition
strains, temperature,
metal flow (geometrical
change)
j
I

Primary Product Requirements

Forming Limits

Secondary Product Requirements

Tolarences
Surface properties, etc.

Figure 17 Factors involved in modeling of bulk deformation processes. (From Ref. 13.:1

,- 1 J 1

I STRESS DEFORMATION

REQUIREMENTS I REQUIREMENTS

Figure 18 Fundamental mechanics involved in the core of the modeling of metalworking processes. (From Ref. 14.)
 Design of Forming Processes: Bulk Forming 17

23- Table 3 Properties and Dimensions for Open Die Disk

Forging Example

I
l1orm h
1

Figure 19 Schematic of open die disk forging process.

the radial position along the disk. The solution is as

follows: pressure distribution across the top of the disk can also
be obtained from this method by using Eq. (23). Figure
p = t10e<a-r) (23) 22 illustrates this distribution for three different reduc-
tions-25%, 50%, and 75%. The large increase in the
PAVG =~(;a)2qo[e--1] (24) center of the disk is due to friction and this shape is
usually called the friction hill.
F=PAVG7ta2 (25)
where p is the pressure at any point, qo is the material B. Slip Line Method
flow strength, Jl is the coefficient of friction, a is the
radius of the disk, r is the radial position, h is the The slip line method is a classical approach to the anal-
thickness of the disk, PAVG is the average pressure, ysis of deforming bodies. The term slip line is misleading
and F is the load. to many metallurgists because they have a specific de-
The slab equilibrium provides a solution at a discrete finition for the term. In mechanics, the slip line method
point in time. To determine how the load varies with probably should be called "maximum shear stress
displacement, an assumption of how the metal changes plane" technique.
shape as a function of time must be used. If a uniform In slip line method, a network of maximum shear
shape change is assumed (i.e., the disk remains as a stress planes is superimposed onto the deforming body.
right circular cylinder during the deformation-no There are a variety of restrictions on the generation of
bulge or foldover), then a load-displacement curved such a network. The network must adhere to specific
can be determined. shape requirements and boundary conditions, and
For an initial disk with the values for the parameters provide a realistic flow field for the deforming material.
listed in Table 3, the load-displacement curve, up to a The method is only valid for plane-strain conditions.
75% reduction in thickness, is shown in Fig. 21. The Because the open die compression of a right circular

a) b)
~

r...

~"'(
/
/

Figure 20 Schematic of slab equilibrium analysis for disk forging: (a) general geometry; and (b) slab element used for analysis.

~
18 Van Tyne

-b
-

~
0.

Q)
5
U>
U>
~
~
Q)
.?:
ro
""Qj
Q:

Figure 21 Comparison of load vs. reduction curves for the

modeling of disk forging via several methods. Figure 23 Comparison of the plane-strain forging analysis
by slab equilibrium method and slip line field method. (From
Ref. 15.)

cylinder is axisymrnetrical and not plane strain, the

analysis of this problem cannot be performed with the which is lower than the load predicted by the slab
slip line technique. equilibrium method.
Figure 23 shows a plane-strain open die forging, The slip line technique imposes ~, velocity field on
which has been solved by the slip line method. The tig- the deforming material through th(: positioning and
ure also contains the relative averaged pressure for the orientation of the maximum shear stress network.
deformation as predicted by the slab equilibrium tech- Hence the velocity field is an implicit assumption within
nique. The plane-strain flow strength of the metal uo' is the method.
2/.;3 times greater than the uniaxial flow strength uo.
The inserted diagrams show the network of maximum c. Upper-Bound Models
shear stress planes, which is used for each point in the
solution. The slip line method predicts a forging load, The upper-bound technique is an energy method where
the energy per unit time needed by the workpiece to
undergo deformation is set equal to the externally sup-
plied energy per unit time. The primary power (energy
per time) terms that must be calculated for the work-
piece include: the internal power of deformation, the
power to overcome friction, and the ~:hearpower. The
internal power is determined from the assumed velocity
field and is calculated from the strain rate field. The
frictional power term is the power needed to overcome
any tool-workpiece frictional interaction. The constant
friction factor model is usually assumed for this type of
analysis. The shear power is determiDl~ by calculating
the energy per unit time associated \vith the internal
shear that occurs over any assumed internal surfaces of
velocity discontinuity.
0.0 0.2 0.4 0.6 0.8 1.0 1.2 For the open die forging of a right <:ircular cylinder,
Distance from Center (in.) the upper-bound solution is given as:

Figure 22 Comparison of pressure distribution over the top

(
PAVG = 0"0 1 +3
2 ma
""J3h ) (26)
of the disk during forging via two different modeling methods.
Design of Forming Processes: Bulk Forming 19

where m is the constant friction factor. The first term In a FEM analysis, the workpiece and tools are dis-
inside the parentheses in Eq. (26) is the internal power cretized into a number of points, called nod~s. The more
term and the second term is the frictional term. For the points in the model present, the mort: accurate is the
simple forging process being examined here, there are solution, but the more time it takes for the computer to
no shear power losses. calculate a solution. The nodes are linked to one another
The upper bound, such as the slab equilibrium meth- by elements, which obey specific deformation laws. The
od, only determines a solution at a discrete instance in workpiece is given specific constraints, loads, and dis-
time. Because the velocity is assumed, the solution at placements, and an equilibrium solution is sought. If the
other time increments is readily available as long as the displacements and loads are given as a function of time,
flow does not change the shape of the workpiece to one the solution can be obtained as a function of time. The
for which the solution is invalid. Figure 21 shows the solution consists of the stressesand strains that exist at
load-displacement curve for the forging of a right cir- every node within the body and the tooling. Various
cular cylinder with the same properties assumed for the interpolation methods are used to c:alculate values
slab equilibrium solution. A constant friction factor of between the nodes. The solution to metal deformation
0.50 was assumed, rather than the value for a coefficient problems requires the use of a computer and a skilled
of friction. operator to interpret the results properly.
The upper-bound solution does not provide a stress For the forging of a right circular cylinder with the
field, hence a plot similar to Fig. 22 for the upper-bound properties given in Table 3, the load-displacement curve
approach cannot be determined. is shown in Fig. 21. The pressure across the top surface
One of the advantages of the upper-bound technique of the disk at reductions of 25%, 50'Jjo, and 75% is
is that it determines a value for the deformation load, shown in Fig. 22. In both of these fi!:ures, the FEM
which is greater or equal to the actual load. Hence with solution is compared to other solutions. A mesh for this
the use of this method, there is a built-in safety factor for quarter disk was a grid of20x20 square elements with a
specifying the size of the equipment to be used. width of 0.025 in. The tooling was meshed with 16x7
A major use of the upper-bound method is to predict rectangular elements 0.0714xO.0875 in. The original
conditions where a process-induced defect may form mesh and the deformed mesh at 750;,~reduction are
within the workpiece. Becauseit is an energy technique, given in Fig. 24.
a comparison between the energy needed for sound flow In contrast to the other techniques, the velocity field
can be made to the energy needed for defect flow. The is not assumed by the FEM analysis but is generated
flow field, which requires the least amount of energy, is within the analysis itself. This forging oj:'a right circular
the one most likely to occur. For example, this method cylinder at 75% reduction exhibits both foldover and
has been successful in developing criteria for the pre- bulge (Fig. 24). Foldover is when the side surface of the
vention of central bursts in wire drawing and extrusions, disk comes in contact with the tooling surface. Bulge is
central bursts in double hub forging, central bursts in when the center region of the free surface moves out-
rolling, side surface cracking in forging with double ward at a greater rate than the regions closer to the
action presses, cavitation in impact extrusion, fishskin platens. Becausethe FEM is a numerical method, which
defects in impact extrusion, and the beginning of the produces a solution at a discrete numbc:r of points, the
piping defect in extrusions. curves shown in Figs. 21 and 22 for tht~ FEM analysis
are not smooth.
Finite element method analysis can provide a large
D. Finite Element Analysis amount of information about the process. For example,
the effective strain contours that exist within the forging
The finite element method (FEM) is the technique that at 75% reduction are shown in Fig. 25. The maximum
has received the most research effort during the last strains occur in the center of the disk and at the original
several decades. It is the one that produces an over- comers of th~ disk. The material directly beneath the
whelming amount of information about the process that platens in the center of the disk unde:rgoes the least
is being modeled. The technique was developed in the amount of strain. This type of information is useful for
1960s for the analysis of elastic deformation in large the prediction of possible shear bandinE. In addition, if
complex structures (e.g., aircraft, bridges, buildings, the final properties of the product are dt:pendent on the
etc.), which have a variety of constraints and loading amount of strain, an indication of property gradients
conditions. The technique was extended in the 1970s within the workpiece might be obtained from such
and 1980s to the plastic deformation of metals. a figure.
20 Van Tyne

a) 1.5

1.0

~ 0.5

c
o
~ 0.0
o
Q.
"'iU

~ -0.5

.1.0

.1.5 I I I I I I I
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

Radial Position (in.)

Figure 24 Finite element mesh for open die disk forging: (a) before deformation; and (b) after 75% reduction in height. (From
Ref. 16.)

One of the advantages of the FEM technique is that then FEM analysis could use it. Unfortunately, these
realistic material properties can be assumed for the descriptions, even for common metals and alloys, are
deforming workpiece and the tooling. All the other not often available.
analysis methods normally are perfQrmed with idealized The second limitation for all of tht:se methods is in
mechanical properties for the workpiece and the tools. the modeling of the frictional interfaces between the
tooling and the workpiece. The two friction models,
E. Modeling Limitations which are used in these modeling methods, are simpli-
fications for the complex interactions that occur at the
Although modeling of bulk deformation processes is a tool-workpiece interface.
very powerful and useful tool, there are several limi- A third limitation is the specification of boundary
tations that exist in all of the techniques. The first is an conditions. The boundary conditions \]lsedfor the anal-
adequate description of the constitutive behavior of the ysis have a direct and profound effect on the results that
deforming workpiece. In almost all cases, some simpli- are calculated. Poor choice of the boundary conditions,
fication of the actual material flow behavior is as- or choosing conditions that make thl~ analysis easier
sumed. To be accurate, the flow behavior should be rather than reflective of the real operation can result in
known and mathematically characterized as a function misleading or erroneous results. The boundary condi-
of strain, strain rate, and temperature. If a good mathe- tions must be chosen with caution and care to ensure
matical description for the material behavior exists, that the results validly reflect the realit:f of the process.

0.5 1.0 1.5

,~ \ ~ '--=-- ~~ ~

1:0 1.5 C- -=>2.0 1.5 1..0

..-I~ ,--=---~-=---, \
0.5 1.0 1.5

Figure 25 True strain contours predicted by FEM for the open die disk forging after 75% reduction in height. (From Ref. 16.)
Design of Forming Processes: Bulk Forming 21

FURTHER READING Kobayashi, S.; Oh, S.I.; Altan, T. Metal Forming and the
Finite-Element Method; Oxford University Press: Ox-
Altan, T.; Boulger, F.W.; Becker,J.R.; Akgennan, N.; Hen- ford, England, 1989.
rung, H.J. Forging Equipment, Materials and Prac- Larke, E.C. The Rolling of Strip, Sheet anti Plate; 2nd Ed.;
tices; MCIC-HB-O3: Battelle, Columbus, OH, U8A, Chapman and Hall: London, Englancl, 1963.
1973. Laue, K.; Stenger, H. Extrusion-Processes, ,\lachinery, Tool-
Altan, T.; Gegel, H.L.; Oh, 8.1. Metal Forming-Fundamen- ing; ASM: Metals Park, OH, USA, 1981.
tals and Applications; A8M: Metals Park, OH, U8A, Metals Handbook: Forming and Forging; 9th Ed.; ASM In-
1983. ternational: Metals Park: OH, USA, 1988; Vol. 14.
A vitzur, B. Metal Forming: Processes and Analysis;McGraw- Mie1nik, E.M. Metalworking Science and l~gineering; Mc-
Hill: New York, NY, U8A, 1968. araw-Hill: New York, NY, USA, 1991.
Avitzur, B. Metal Forming: TheApplicationof Limit Analysis; Open Die Forging Technology; Forging Industry Association:
Marcel Dekker: New York, NY, U8A, 1980. Cleveland, OH, USA, 1993.
Avitzur, B. Handbook of Metal Forming Processes;John Roberts, W .L. Cold Rolling of Steel; Marcel Dekker: New
Wiley: New York, NY, USA, 1983. York, NY, USA, 1978.
Schey, J.A. Tribology in Metalworking-Friction, Lubrication
Beddoes,J.; Bibby, M.J. Principlesof Metal Manufacturing
and Wear; ASM: Metals Park, OH, USA, 1983.
Processes; Arnold: London, England, 1999.
Schrader, a.F.; Elshennawy, A.K. Manufa(:turing-Processes
Blazynski, T.Z., Ed.; Plasticity and Modern Meta/-Forming
and Materials; SME: Dearborn, MI, lJSA, 2000.
Technology;Elsevier:London, England, 1989.
Slater, R.A.C. Engineering Plasticity-Theor.v and Application
Boer, C.R.; Rebelo, N.; Rystad, H.; Schroder, G. Process
to Metal Forming Processes; John Wiley: New York,
Modelling of Metal Forming and Thermomechanical
NY, USA, 1977.
Treatments;8pringer-Verlag:Berlin, Gennany, 1986.
Spencer, a.c. Introduction to Plasticity; Chapman and Hall:
Byrer, T.G., Semiatin, S.L., Vollmer, D.C., Eds.; Forging
London, England, 1968.
Handbook; Forging Industry Association: Cleveland,
Talbert, S.H.; Avitzur, B. Elemen!ary Mec,ianics of Plastic
OH, USA, 1985.
Flow in Metal Forming; John Wiley: New York, NY,
Calladine, C.R. Plasticity for Engineers- Theory and Appli-
USA, 1996.
cations;Horwood: Chichester,England, 2000.
Thomsen, E.a.; Yang, C.T.; Kobayashi, :). Mechanics of
DeGanno, E.P.; Black, J.T.; Kohser, R.A. Materials andPro- Deformation in Metal Processing; Macmillan: New
cessesin Manufacturing; 8th Ed.; Prentice-Hall: Upper
York, NY, USA, 1965.
SaddleRiver, NJ, USA, 1997. T1usty, a. Manufacturing Processes and Equ.pment; Prentice-
Dieter, G.E., Ed.; Workability Testing Techniques;ASM: Hall: Upper Saddle River, NJ, USA, 2000.
Metals Park, OH, USA, 1984. Wagoner, R.H.; Chenot, J.L. Fundamentals o.(MetaIForming;
Dieter, G.E. MechanicalMetallurgy: 3rd Ed.; McGraw-Hill: John Wiley: New York, NY, USA, 1~~7.
New York, NY, USA, 1986. Wagoner, R.H.; Chenot, J.L. Metal Formin~' Analysis; Cam-
Everhart, J.L. Impact and ColdExtrusionof Metals; Chemical bridge University Press: Cambridge, England, 2001.
Publishing:New York, NY , 1964.
Ginzburg, V .B.; Ballas, R. Flat Rolling Fundamentals;Marcel
Dekker: New York, NY , U8A, 2000.
Hartley, P., Pillinger, I., Sturgess,C., Eds. Numerical Mod- REFERENCES
elling of Material DeformationProcess-Research,De-
velopmentand Applications; Springer-Verlag: Berlin, Byrer, T.G.,Semiatin,S.L., Vollmer,D.C.,Eds.;Forging
Gennany, 1992. Handbook; Forging Industry Association: Cleveland,
Hill, R. The Mathematical Theoryof Plasticity; Oxford uru- OH, USA, 1985; pp. 16-18.
versity Press:Oxford, England, 1950. 2. Metals Handbook: Forming and Forging; 9th Ed.; Vol.
Hosford, W.F.; Caddell, R.M. Metal Forming-Mechanics 14. ASM International: Metals Park, OH, 1988; pp. 44.
and Metallurgy; 2nd Ed.; Prentice-Hall: Englewood 3. DeGarmo, E.P.; Black, J.T.; Kohser, R.A. Materials
Cliffs, NJ, USA, 1993. and Processes in Manufacturing; 8th Ed.; Prentice-Hall:
Johnson, W.; MelIor, P.B. EngineeringPlasticity; Van Nos- Upper Saddle River, NJ, USA, 1997; po476.
trand: London, England, 1973. 4. WaIters, J. Scientific Forming Technologies; Columbus,
Johnson, W.; Sowerby, R.; Haddow, J.B. Plane-Strain Slip OH, USA, 2000, personal communication
Line Fields: TheoryandBibliography;Elsevier:London, 5. DeGarmo, E.P .; Black, J .T .; Kohser, R.A. M aterials and
England, 1970. Processes in Manufacturing; 8th Ed.; Prentice-Hall:
Kalpakjian, S.; Schmid, S.R. ManufacturingEngineeringand Upper Saddle River, NJ, USA, 1997; p. 486.
Technology; 4th Ed.; Prentice-Hall: Upper Saddle 6. Avitzur, B. Handbook of Metal Forming Processes;John
River, NJ, USA, 2001. Wiley:NewYork,NY,USA, 1983;p. 150.
22 Van Tyne

7. Kalpakjian, S.; Schmid, S.R. M anufacturing Engineering 12. Schey, J.A. Tribology in Metalworking-Friction. Lubri-
and Technology; 4th Ed.; Prentice-Hall: Upper Saddle cation and Wear; ASM: Metals Park, OH, USA, 1983;
River, NJ, USA, 2001; p. 321. p.451.
8. Kalpakjian, S.; Schmid, S.R. Manufacturing Engineering 13. Kobayashi,S. Metalworking Pr~: Modelling and the
and Technology; 4th Ed.; Prentice-Hall: Upper Saddle Finite Element Method. Proceeding~ of NAMRC IX;
River, NJ, USA, 2001; p. 888. pp.I6-21.
9. Zimerman, Z.; Avitzur, B. Analysis of the effect of strain 14. Boer, C.R.; Rebelo, N.; Rystad, H.; S.;hroder, G. Process
hardening on central bursting of strain hardening in Modelling of Metal Forming and Thermomechanical
drawing and extrusion. Trans. ASME J. Eng. Ind. 1970, Treatments; Springer-Verlag: Berlin, Germany, 1986;
92, 135--145. p.20.
10. Lee, P.W.; Kuhn, H.A. Cold upset testing. In Work- 15. Hosford, W.F.; Caddell, R.M. Metal Forming-Me-
ability Testing Techniques; Dieter, G.E., Ed.; ASM: Met- chanics and Metallurgy; 2nd Ed.; Prentice-Hall: Engle-
als Park, OH, USA, 1984; pp. 37-50. wood Cliffs, NJ, USA, 1993; p. 203.
11. Kuhn, H.A.; Lee, P .W .; Ertuk, T. A fracture criteria for 16. Van Tyne, C.J. Modeling of Open Die Forging Pro-
cold forging. Trans. ASME J. Eng. Mater. Technol. cesses.Mechanical Working and Steel Processing Confer-
1973,95, 213-218. ence Proce~dings; 1989; pp. 209-218.
 14
Van Tyne

a) a)

~
Top VIeW

"'-- )

Front View:
rT\

b)
Figure 13 Schematics of compression test specimen geo-
metries: (a) cylindrical sample; (b) tapered sample; and (c)
flanged sample. (From Ref. 10.)

compressed until fracture occurs to assessthe metals'

workability during compression and produce a fonning
limit curve. Typical fracture curves (or fonning limit
diagrams) for 1020 steel, 303 stainless steel, and 2024-
T351 aluminum are shown in Fig. 15.

4. Friction Tests
The most common method used to determine the fric-
tion factor for a forging process is the ring compression Figure 15 Cold upset compression failure criteria: (a) 1020
test. The test can be conducted at varying temperature steel and 303 stainless steel; and (b) 2024-T351 aluminum.
and speed, and with the lubricant and workpiece mate- (From Ref. 11.)

rial of interest. The workpiece material is machined into

a ring with dimensions usually in a 6:3:2 ratio of the
outer diameter to the inner diameter to the thickness.
The ring is compressed in the thickness direction to a
given level of deformation and the new inside diameter
is measured. Friction calibration curves can be used to
determine the friction factor from the amount of defor-
mation imparted to the ring and the change in inner
diameter (Fig. 16). Rings of other dimensions can be
used but the appropriate calibration curves must be
used for the specific starting geometry.

VII. DEFORMATION MODELING METHODS

A diagram illustrating the input and output as well as

Figure 14 Strain paths for compression tests of cylindrical the constraints, which must be considered when trying
specimens with various height (h)-to-diameter (d) ratios and to model a bulk deformation process, is shown in Fig.
various lubrication conditions. (From Ref. II.) IJ. The input parameters fall into three major catego-