AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 1 -- Welcome back! The next several
sections of our failure analysis seminar will
be devoted to learning principles of
metallurgy, wear, and fractures that can be
applied to all failures of any type of
equipment under any conditions. We will
conclude this theory segment with a detailed
introduction to perhaps the most important
fact-gathering tool -- visual examination.

Applied Failure Analysis Metallurgy Slide 2 -- In this section we will review some
basic metallurgical facts and metal shaping
processes that will familiarize us with the
physical structure of components. With this
knowledge, we can better recognize flaws
caused by materials and processes. Later,
when we learn about fractures, we will see
that parts fail in specific locations if the part is
overloaded. If a part fails anywhere else we
should suspect a material or process flaw.

Applied Failure Analysis Metallurgy Slide 3 -- In this presentation, we will review

metal refining, metal forming, and heat
treatment, after that we will learn to recognize
material and process flaws that can occur at
each step. The metallurgical principles
contained in this section have been simplified
to help students understand general concepts.
For more detailed information, refer to the
course handout entitled "Materials and
processes Reference Manual".

1
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 4 -- Cast iron and steel are the two most
common metals used in Caterpillar products.
Both are produced by refining iron ore found
in nature as a stable iron oxide mixed with
impurities. Refining begins with the heating
of iron ore with coke (coal baked without
oxygen) and limestone in a blast furnace lined
with refractory brick. As these three materials
melt, two important things happen: (1) coke
combines with oxygen from the iron ore
releasing molten iron that sinks to the bottom
of the furnace, and (2) limestone combines
with impurities (dirt, sulfur, etc.) and floats to
the top forming slag. These chemical and
physical actions produce a less stable, refined
metal called pig iron that is tapped off for
further refinement into steel or cast into ingots
called pigs.
Applied Failure Analysis Metallurgy Slide 5 -- Steel is pig iron that has gone
through a refining process to lower the carbon
content, further reduce impurities, and to
adjust percentages of other elements. Today,
most steel is produced in electric arc furnaces.
The pig is charged into the brick lined
furnace, along with scrap, and carbon
electrodes are lowered to within a few inches
of the metal. The current is turned on and heat
from the metals resistance to current flow
quickly melts the charge. Refining continues
until the levels of carbon, impurities, and
certain other elements reach required
specifications. Since these processes are
imperfect, some small particles of furnace
brick and slag, as well as some dirt from the
ore still remain in the steel.

2
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 6 -- The fully refined steel, called a heat,
is poured into ingot molds and allowed to
solidify. While the steel is cooling some
trapped gas may form internal voids in the
ingot. Ingots are later reheated and taken to a
rolling mill for hot rolling into sheets, slabs,
plates or bars. During rolling most of the
internal voids weld together.

Applied Failure Analysis Metallurgy Slide 7 -- The percentage of the element

"carbon" together with the element " iron"
determines whether we have cast iron or steel,
arid also determines most of the cast iron or
steel's properties. Most cast iron is refined
until it contains between 2% and 4% carbon
making it brittle with little or no ductility.

Applied Failure Analysis Metallurgy Slide 8 -- When carbon levels in iron are
adjusted from .02% to 2.0%, steel is
produced. Steel is generally stronger, harder,
tougher, more ductile and shock resistant than
cast iron.

3
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 9 -- Steels are classified according to

their carbon content. Low carbon steels
contain from .02% to .25% carbon, are very
ductile, and are used to make non-heat treated
parts such as oil pans and fuel lines. Some
low carbon steels have additional carbon
added to the surface (called carburizing) so
they can be used for making case hardened
parts such as gears and piston pins. We will
discuss this in more detail later in this section.
Medium carbon steels contain from .25% to
.50% carbon, are less ductile, can be heat-
treated and are used to make many forged
parts such as crankshafts and connecting rods.
High carbon steels contain .50% to 2.0%
carbon, are brittle, and are used to make fuel
injection pump components and springs.

Applied Failure Analysis Metallurgy Slide 10 -- Any of these steels can be alloyed
with other metals and elements to improve
physical properties such as tensile strength,
toughness and corrosion resistance. Common
alloying agents are chrome, nickel, vanadium,
tungsten and molybdenum. These modified
steels are called alloy steels.

4
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 11 -- Since carbon determines cast iron
and steel properties, we might expect a
difference in the structure of these two very
different metals. Before looking at each
individually, let's look first at the general
physical structure of metals. All metals are
composed of grains that are crystals that have
grown from the molten metal as it solidifies
just like ice crystals form on windowpanes on
a frosty morning. When the crystals bump
into each other, they form random grain
boundaries. It is in these grain boundaries that
we will find impurities such as refractory
brick or trapped slag left from the refining
process. These impurities are called
inclusions and are usually small enough not to
weaken the metal.

Applied Failure Analysis Metallurgy Slides 12 -- Taking a look at the grain

structure of cast iron, we notice the grain
boundaries produce a random grain structure.
In addition, each grain contains carbon in the
form of graphite flakes. This free carbon
produces most of the properties of cast irons
such as brittleness, good wear resistance,
good machinability, etc.

5
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 13 -- Looking at the steel structure, we

do not find graphite flakes, but only grains
and grain boundaries. Steel's lower carbon
levels allow carbon to mix in the grains and
no free carbon exists to form flakes. This
structural difference produces steels
properties of ductility, strength and toughness.
(Note: Inclusions have been left out of the
drawing for simplicity.)

Applied Failure Analysis Metallurgy Slide 14 -- Taking a closer look at the grain
structure reveals an orderly arrangement of
"unit cells" in rows and columns much like a
stack of bricks. However, the alignment of the
rows and columns is different in each grain.
Later on when we study loading and steel
failure we will find that this different
alignment of the rows and columns explains
crack growth and appearance.

Applied Failure Analysis Metallurgy Slide 15 -- Unit cells are the smallest building
blocks of a grain. In cast iron and steel, these
cells are usually cube shaped with iron atoms
at the cube center and corners (called a body
centered cubic structure). The atoms in a unit
cell can be rearranged or different atoms can
be added by heat treating and alloying. This
gives metals different physical and chemical
properties.

6
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 16 -- Both iron and steel are made into
useful forms using the metallurgical processes
of casting and shaping. Most iron used in
Caterpillar product is formed by casting while
most steel is shaped using one of the
following high temperature, high pressure
processes:

1. Rolling
2. Forging
3. Extruding
4. Drawing

Steel reshaped by one of these processes is

referred to as "wrought". Cast and wrought
materials form two general categories of
metal with very different properties.

Applied Failure Analysis Metallurgy Slide 17 -- The casting process consists of

pouring iron into a mold containing the
desired part shape. Molds are produced by
pressing a pattern of the part into sand mixed
with a binder (glue). This sand sets up and
holds the patterns shape. Internal cavities are
produced by core molds supported on small
iron pins called chaplets. Hot air and gasses
contained in the sand and molten metal escape
through small passageways in the mold called
vents. Molten iron at 2200-2400°F (1204-
1315°C) flows into the mold passages, melts
chaplets, produces gasses that are vented out
of the mold and then begins to solidify.

7
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 18 -- The casting may contain some
trapped gas or slag inclusions, or if not
properly designed for slow cooling, it may
contract too quickly and pull away from itself
forming shrinkage cavities. The solidified
casting consists of a random grain structure
with graphite flakes. No further forming can
be done to this brittle structure, but it can be
easily machined and heat-treated.

Applied Failure Analysis Metallurgy Slide 19 -- Rolling is the most common

process used to produce wrought shapes such
as sheets, slabs, plates, and bars. Rolling is
usually done with steel heated to about
2200°F (1204°C) to make it very plastic. High
pressure is applied with large steel rollers and
the steel is squeezed into the desired shape.
These shapes can then be reformed later by
forging, extruding, drawing, or using other
metal working processes.

Applied Failure Analysis Metallurgy Slide 20 -- During rolling, the rows and
columns of unit cells in the grains Slip past
each other allowing the grain boundaries to
take on a new shape without tearing apart.
The new grain structure along with the
inclusions aligns in the rolling direction
forming grain flow lines. This makes the steel
stronger in the direction of the grain than
across it. However, cracks will grow faster
with the grain just like wood splits easier with
the grain.

8
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 21 -- Forging is a common method of

forming structural shapes by heating to about
2200°F (1204°C) and applying force with a
forging hammer. The hot steel grains are
forced into the shape of the forging die
causing grain flow to be parallel to maximum
anticipated loads. This operation sometimes
produces folds in the metal called laps, or
overheats the metal resulting in forging burns.

Applied Failure Analysis Metallurgy Slide 22 -- The key principles that we should
remember about cast and wrought materials
are that cast materials have a random grain
structure containing graphite flakes while
wrought materials have flow lines. This
accounts for most of the difference in physical
properties and behavior of these two
materials. When we look at fracture
principles, we will see it also accounts for the
different fracture appearance of cast and
wrought metals.

Applied Failure Analysis Metallurgy Slide 23 -- Once parts have been formed and
machined, they may require heat treatment for
increased strength, toughness or wear
resistance. Heat treating changes the unit cell
structure of cast iron and steel and puts
carbon atoms into the cells making the cells
harder. Heat-treating is generally a three step
process: (1) austenitizing, (2) quenching, and
(3) tempering.

9
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 24 -- Austenitizing is the first stage of

heat treatment and is done at elevated
temperatures to put carbon into the unit cell.
Carbon steels at room temperature have a
body centered cubic unit cell that does not
have room for carbon atoms in it. During
austenitizing, temperatures are increased
above 1400ºF (760ºC) to change the unit cell
from body centered cubic to face centered
cubic that has room for carbon atoms. The
face centered cubic unit cell has iron atoms
located at each cube corner and in the center
of each cube face. This leaves room for the
carbon atoms to fit on each cube edge. This
ability to add carbon to the cell structure is the
basis for heat treatment of metals.

Applied Failure Analysis Metallurgy Slide 25 -- Parts can be austenitized in several

ways, the most common being furnace
heating. Hardening furnaces are used to
through heat parts for quenching. Here we see
an orange-hot crankshaft at 1600°F (871°C)
coming from a hardening furnace.

10
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 26 -- Some parts only require a hard
surface layer and do not need to be through
heated. Induction heating is a method of
electrically heating part surfaces about .250
inches (6 mm) deep. Here we see a crankshaft
with the surface layer being austenitized by
induction heating.

Applied Failure Analysis Metallurgy Slide 27 -- Flame heating is a third method of

surface heat treatment used to austenitize
metal surfaces about .5 inches (12 mm) deep.
Quenching and tempering immediately follow
all of these methods of austenitizing.

Applied Failure Analysis Metallurgy Slide 28 -- Quenching is the second stage of

heat treatment consisting of rapidly cooling
the metal from the austenitizing temperature
to room temperature. The face centered cubic
unit cell changes to a new form called body
centered tetragonal that locks carbon into it.
This produces steel of high strength and
hardness that is also very brittle and contains
high residual stresses. A typical medium
carbon steel crankshaft will have hardness
increased from Rc 20 to Rc 60 after
quenching. (These hardness numbers will be
explained in just a few minutes.)

11
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 29 -- Quenching can be done in water,

oil or air. Water can usually be used on low
and medium carbon steels, while oil or air are
used on high carbon or alloy steels.

Applied Failure Analysis Metallurgy Slide 30 -- Tempering is the final stage of

heat treatment and relieves residual stresses
while improving toughness through controlled
escape of some carbon from unit cells.
Careful control of tempering temperatures
gives good control of the carbon escape rate
and minimizes hardness loss. Unless parts are
raised above this tempering temperature, the
final toughness and hardness will remain
stable. Tempering of the crankshaft seen
previously reduces the hardness to about Rc
52.

Applied Failure Analysis Metallurgy Slide 31 -- Another type of heat treatment

used to produce hard surface skins on parts is
called case hardening. Since medium and high
carbon steels are expensive, it is cheaper to
use low carbon steel for case hardened parts
and add carbon, nitrogen (another alloy
element like carbon) or both to the surface
layers to be hardened. Three common heat
treatments to do this are carburizing,
carbonitriding, and nitriding. Carburizing is
done at a temperature of about 1700°F
(927°C) and can add carbon to a depth of
about .125 inches (3 mm). It requires
quenching and tempering afterwards and
produces a hard, wear resistant surface with a
tough supporting core capable of carrying
high loads. Quenching from the high

12
AFA – 3 Metallurgy 4/22/2001

carburizing temperature produces a lot of

distortion in carburized parts that must be
removed by straightening or machining back
to correct specifications. Carbonitriding can
add carbon and nitrogen to a depth of about
.015 inches (.3 mm) at about 1400°F (760°C),
requires quenching and tempering afterwards,
and produces a thin, hard, wear resistant
surface. The lower temperatures for this heat
treatment reduce part distortion. Nitriding
adds nitrogen to about .015 inches (0.3 mm)
of surface metal at about l000°F (538°C) ,
requires austenitizing, quenching and
tempering before the nitriding operation and
produces a hard, wear resistant case with a
tough supporting core capable of carrying
medium loads. The very low nitriding
temperature produces no distortion and can be
applied to finished parts without changing the
dimensions.

Applied Failure Analysis Metallurgy Slide 32 -- Hardness tests are used to measure
the effectiveness of heat treatment. The
Rockwell C hardness test measures a metals
resistance to indentation by a diamond
indenter that is pushed into the metal surface
by a very accurately determined amount of
weight. The harder the metal, the shallower
the indentation. A typical unheat-treated
machinable steel part will have Rockwell C
hardness less than Rc 20. Parts harder than Rc
40 are not machinable and must be ground.
Hardened wear surfaces and heat-treated
cases may be Rc 45 - 65.

13
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 33 -- Refining, shaping and heat

treatment are all metallurgical processes that
can leave flaws or defects in materials.
Refining may not remove all the dirt from the
ore, may leave slag trapped in the melt, or
may introduce furnace brick particles into the
metal. These trapped impurities, called
defects or inclusions, are present in all
materials but are usually too small to cause
any problems.

Applied Failure Analysis Metallurgy Slide 34 -- Inclusions are usually so small

after being rolled or forged that they do not
cause failures. Occasionally one may be in a
critically loaded area, and may be large
enough to start a crack. Since these flaws are
mostly internal, we will see that cracking
starts inside the part whether started by dirt,
slag or brick.

Applied Failure Analysis Metallurgy Slide 35 -- Forming processes can introduce

surface and internal defects in both cast and
wrought metals. We need to become familiar
with their appearance so we can recognize
them on fracture faces. The key idea to
remember is that when cracking starts
internally, a flaw is usually going to be found
at the starting point. The presence of such a
flaw indicates faulty material, processing or
perhaps severe overloading. We should also
remember that internal cracks grow to the
surface while surface cracks will grow
internally. Let's take a more detailed look at
the most common defects learning what
causes them, their location in parts, and the
appearance of each.

14
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 36 -- A common internal casting flaw is

a shrinkage cavity caused by pouring iron too
hot or not providing enough metal to thick
casting sections. As metals cool, they contract
and more metal must flow into these areas to
keep the casting from pulling apart. If this is
not done, large irregular shaped cavities will
form inside the metal weakening it in that
area. If the cavity is too large, or if unusually
high loads occur during operation, a crack
may start at the cavity.

Applied Failure Analysis Metallurgy Slide 37 -- A common casting surface flaw is

handling cracks caused by physical abuse.
When opened, they have no discoloration
unless exposed to heat or corrosive
environments while in service. Fatigue cracks
that we will learn more about in the fractures
section can grow from these surface cracks if
loads are high.

Applied Failure Analysis Metallurgy Slide 38 -- The three most common internal
defects in wrought materials are pipe, flakes,
and forging hot spots. Pipe occurs when steel
ingots solidify leaving a shrinkage cavity in
the top center portion of the ingot. This part
of the ingot is usually sawed off before
rolling. Occasionally, some of the cavity
remains in the ingot and is rolled into the
finished shape. This forms an irregular hole in
the center referred to as pipe. Pipe can be a
stress raiser and generate an internal fatigue
crack in a part during service.

15
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 39 -- Flakes are caused by hydrogen gas
being dissolved in the steel during the refining
process. Hydrogen collects around inclusions,
creates high pressure and bursts the steel
internally. Flakes look like small, shiny,
rounded bright spots when present on fracture
faces.

Applied Failure Analysis Metallurgy Slide 40 -- Overheating steel during forging

causes Forging hot spots. Heat generated from
the severe mechanical working raises the
grain boundary temperature in the heaviest
worked areas to the melting point causing
weakening of the grains internally. If this
condition leads to failure, the molten areas
will look like large grains or crystals on the
fracture surface.

Applied Failure Analysis Metallurgy Slide 41 -- Seams and laps are common
surface defects having a very similar
appearance. They are caused by metal folding
in on itself during rolling or forging
operations. Surface scale gets pinched into the
seam or lap preventing it from welding to
itself at the high hot-working temperatures.
When present on fracture faces, they appear
dark black and rough due to this scale.

16
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 42 -- In addition to refining and shaping

flaws, heat treatment can produce defects
such as quench cracks, soft spots and
straightening cracks.

Applied Failure Analysis Metallurgy Slide 43 -- Quench cracks are caused by

heating parts too hot, quenching them liquid
that is too cool, or using the wrong quench
liquid. The severe thermal shock results in
metal contracting too fast and cracking at the
surface. Quench cracks usually occur in stress
raisers already present in the part such as gear
tooth fillets, keyways, splines, and thread
roots. If parts are tempered above 500°F
(260oC) in air, the quench crack may develop
some discoloration.

Applied Failure Analysis Metallurgy Slide 44 -- Heat treatment causes residual

stress in parts that usually results in distortion.
A common method for removing distortion is
straightening in hydraulic presses. If parts are
overstressed during this operation, a crack
may develop at a stress raiser. These cracks
are called straightening cracks, and they can
be the source of a fatigue crack during
service.

17
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 45 -- Nondestructive testing (NDT) is

used to monitor refining, shaping and heat
treating processes to assure finished materials
are free of harmful flaws. NDT is also used to
find defective material by 100% sorting for
flaws. Commonly used methods include
magnaglo, zyglo, ultrasonic and eddy current
testing. Magnaglo, zyglo, and eddy current
tests are used to find surface defects while
ultrasonics is used to locate subsurface
defects.

Applied Failure Analysis Metallurgy Slide 46 -- Piston manufacturers make use of

ultrasonic testing equipment to verify the
bond quality of the cast iron ring band that
carries the piston rings. Since this was begun,
the number of pistons experiencing bond
failures in service has dropped drastically.

Applied Failure Analysis Metallurgy Slide 47 -- Certain engine applications, such

as marine, require higher than usual
crankshaft quality. Ultrasonic testing of large
bore crankshafts helps certify that inclusion
levels in these forgings is well below the size
that can cause a failure.

18
AFA – 3 Metallurgy 4/22/2001

Applied Failure Analysis Metallurgy Slide 48 -- This concludes our section on

basic metallurgy. Metal refining, shaping, and
heat-treating can create flaws or defects in
parts. By knowing what these flaws look like,
we can quickly determine if any are present
on a fracture. If none are found, we can avoid
blaming the part or process for the failure and
look for other facts or road signs to lead us to
the root cause of failure.

Applied Failure Analysis Metallurgy Slide 49 – As we face the challenge of finding

the root cause of a problem and determining
proper corrective action, we need to
remember these basic metallurgical
principles. The metallurgy of Caterpillar parts
is very carefully designed and closely
controlled. We therefore need to insure we
have the facts before we accuse apart of
causing the failure. Too often we find that
carefully made steel or cast parts have been
exposed to abnormal conditions that have
caused the failure.

19