12.16 SURFACE FATIGUE FAILURE MODELS—DYNAMIC
is still some disagreement among experts as to the actual mechanism of failure
that results in pitting and spalling of surfaces. The possibility of having a
maximum shear stress at a subsurface location (in pure rolling) has led some to
conclude that pits begin at or near that location. Others have concluded that
pitting begins at the surface. It is possible that both mechanisms are at work
in these cases, since failure initiation usually begins at an imperfection,
which may be on or below the surface. Figure 12-24 shows both surface and
subsurface cracks in a case-hardened steel roll subjected to heavy rolling
extensive experimental study of pitting under rolling contact was done by
Way in 1935. Over 80 tests were made with contacting, pure rolling,
parallel rollers of different materials, lubricants, and loads, run for up to
18 million cycles, though most samples failed between 0.5
6 cycles. The samples were monitored for the
appearance of minute surface cracks, which inevitably presaged a pitting
failure within less than about 100 000 additional cycles in the presence of a
and smoother surfaces better resisted pitting failure. Highly polished samples
did not fail in over 12
6 cycles. Nitrided rolls with very hard cases
on a soft core
Photomicrograph (100x) of
surface and subsurface cracks in a carburized and hardened roll (HRC 52-58)
subjected to a heavy rolling load (Source: J. D. Graham, Pitting of Gear Teeth,
in C. Lipson, Handbook of Mechanical Wear, U. Mich. Press, 1961, p.137.)
longer-lived than other materials tested.
No pitting occurred on any samples in the absence of a lubricant
even though dry-running produced surface
cracks. The cracked parts would continue to run dry with no failure for as many
6 cycles until some lubricant was added. Then
the surface cracks would rapidly enlarge and turn to pits of a characteristic
arrowhead shape within 100 000 additional cycles.
suggested explanation for the deleterious effect of the lubricant was that once
suitably oriented surface cracks form, they are pumped full of oil on
approaching the rollnip, and then are pressed closed within the roll-nip,
pressurizing the fluid trapped in the crack. The fluid pressure creates tensile
stress at the crack tip, causing rapid crack growth and subsequent break-out of
a pit. Higher-viscosity lubricants did not eliminate metal-to-metal contact but
did delay the pitting failure, indicating that the fluid must be able to
readily enter the crack to do the damage.
reached a number of conclusions regarding how to design rollers to delay
surface fatigue failure.
Use no oil (though he was quick to point out that this is not a practical
solution, as it promotes other types of wear as discussed in previous
Increase the viscosity of the lubricant.
Polish the surfaces (though this is expensive to do).
Increase the surface hardness (preferably on a softer, tough core).
conclusions were drawn with respect to the reasons for the initiation of the
initial cracks on the surface. Though, with pure rolling, the shear stresses
are not maximal at the surface, they
nonzero there at some
locations (see Figures 12-12, p. 354 and 12-17, p. 361).
TABLE 12-3 Modes of Surface
Failure and Their Causes
Source: W. E. Littmann and R. L. Widner, Propagation
of Contact Fatigue from Surface and Subsurface Origins
, J. Basic Eng. Trans
, vol. 88, pp. 624–636, 1966.
and Widner performed an extensive analytical and experimental study on
contact fatigue in 1966 and describe five different modes of failure in rolling
contact. These are listed in Table 12-3 along with some factors that promote
their occurrence. Some of these modes address the crack initiation issue and
others the crack propagation issue. We will briefly discuss each in the order
This describes a
mechanism for crack initiation. It is assumed that the crack originates in a
shear-stress field at a subsurface or surface location containing a small
inclusion of “foreign” matter. The most commonly identified inclusions are
oxides of the material that formed during processing and were captured within
it. These are typically hard and irregular in shape and create stress
concentration. Several researchers,, have published
photomicrographs of (or otherwise identified) subsurface cracks starting at
oxide inclusions. “These oxide inclusions are often present as stringers or elongated
aggregates of particles . . ., which provide a much greater chance for a point
of high stress concentration to be in an unfavorable position with respect to
the applied stress.” The propagation of a crack from the inclusion may
remain subsurface, or break out to the surface. In the latter case it provides
a site for hydraulic pressure propagation, as described above. In either case,
it ultimately results in pitting or spalling.