ACADEMIC STUDY

Dr. David Kirk | Coventry University

Back to Basics:
Dent Formation and Coverage
INTRODUCTION energy loss is now a mixture of plastic and elastic energy
The aim of this mini-series is to cover the basic scientific transfer. The higher the ratio of plastic to elastic energy
principles of shot peening. Fundamental principles are transfer the greater is the efficiency of kinetic energy usage.
presented together with relevant theoretical explanations.
You do not need to understand the mathematics—they are DENT DIAMETER
only needed to justify the ways in which quantification and The controllable variables that influence indent diameter are
prediction can be achieved. well-known to shot peeners. For a given type of shot they are
A necessary feature of shot peening is that dents are shot diameter, shot velocity and component hardness. These
produced on the surface of the component. It is these dents variables influence three interrelated factors: the volume of
that induce the beneficial effects of surface work-hardening the indent, V, the amount of work done by the shot particle,
and compressive residual stress to form the “magic skin.” This W, and the amount of work, B, that has to be done to create
article deals only with dent formation and coverage, leaving each unit of indent volume. A simple equation connects the
these beneficial effects of dents for later articles. three variables:
Dent formation requires work being done. A very
important scientific principle is that energy is indestructible— V = W/B (1)
it can only be transferred. In our case we have the kinetic
energy of a flying shot particle available to carry out the work As a “hole digging” example for equation (1), if W represents
needed to create a dent. During dent formation two types of 80 man-hours of digging work and B represents a situation
kinetic energy transfer are involved: plastic and elastic. where 10 man-hours of work are needed to create 1 cubic
Kinetic energy and work energy are identical in metre of hole, then 8 cubic metres of hole are created.
terms of units. Every impacting shot particle has a kinetic
energy—½Mv2 where M is mass and v is velocity. Work done DENT VOLUME, V
is force times distance so that its units are Nm where N is Shot particles are almost spherical so that dent shape is close
force and m is distance. Mass can also be expressed as kg and to what mathematicians call a “spherical cap”. The volume of a
v as ms-1 so that Mv2 becomes kgm2s-2. One Newton, N, can spherical cap, see fig. 1 on page 30, can be represented by the
also be expressed as kgms-2 so that Nm becomes kgm2s-2— following equation:
the same as Mv2. V = πd4/32D (2)
Shot peening induces vast numbers of dents. These dents
give us progressive coverage. The greater the number of WORK, W, DONE IN CREATING DENT
dents per unit area the greater will be the coverage. Because A flying shot particle has a kinetic energy, E, given by the
coverage is a specified requirement it has been thoroughly expression E = ½Mv2 where M is the mass of the particle
analyzed. This article includes a summarized version of the and v is its velocity. The mass of a sphere is its volume, D2π/6
relevant theoretical explanations of coverage evolution. multiplied by its density, ρ. Hence:

DENT FORMATION M = D2 .π. ρ./6 (3)

Therefore:
PLASTIC AND ELASTIC ENERGY TRANSFER E = D2 .π.ρ.v2/12 (4)
Imagine dropping a tennis ball onto a steel plate. The ball will
rebound but not to the same height indicating a loss of kinetic After the shot particle has struck the component it bounces
energy. No dent is formed so that all of the kinetic energy loss off at a lower velocity thereby losing some of its kinetic
has been elastic. A steel ball bearing dropped from a height of energy. The proportion, P, of energy lost is the work, W, done
several meters will also rebound but will form a dent. Kinetic in creating the dent. W is therefore given by W = P.E.

28 The Shot Peener | Fall 2020

 ACADEMIC STUDY Continued

Fig. 2. Ball dropped

from height h1 and
then rebounding to
height h2.

Fig. 1. Indent of diameter d, created by sphere of diameter D. Equation (7) is absolutely fundamental to shot peening
control. Experienced shot peeners already know that the
The proportion of energy lost varies with the hardness factors in the equation are important. Indent diameter does
of the component. If the component was made of modelling increase with shot diameter, shot density and particularly
clay it would not rebound at all! A very simple experiment shot velocity but decreases as the hardness of the component
to find a value for P is illustrated in fig. 2. A ball bearing is increases—other factors being kept constant.
dropped from a known height, h1, onto a metal plate. The Science is based on a combination of theory and exper-
rebound height, h2, is measured using a rule held vertically imental verification. Experimental verification of equation
and monitored using the video function of a smartphone. P is (7) was presented in a previous article (TSP, Summer, 2004,
then given by: “Actual and Predicted Shot Peening Indentations”). Fig. 3
illustrates one factor, velocity, that was investigated. A 2 mm
P = (1 – h2/h1) (5)
weighted ball bearing was dropped from different heights
If h2 = h1 then h2/h1 =1 so that P = 0. This means that onto mild steel. The diameters of the indentations produced
no kinetic energy has been lost at all—perfect elasticity. More were measured optically and then plotted as a function of
realistically if h2/h1 = ½ then P = 0.5. drop height. Indentation diameter was found to be propor-
Knowing, or assuming, a value for P we can incorporate tional to the fourth root of the drop height. Impact velocity, v,
it into equation (4) to give that: is proportional to the square of the drop height. This means
that the data proves that indentation diameter is a function of
W = P. D2 .π.ρ.v2/12 (6) v0.5 as given in equation (7).

WORK DONE PER UNIT VOLUME OF INDENT, B

The indent strength, B, is equivalent to the work done during
a Brinell hardness test. Brinell hardness values are normally
quoted in kgf/mm2 but can be converted into MPa by
multiplying by 9.8. A Brinell hardness value for mild steel of
200 kgf/mm2 is equal to 1,960 MPa. The theoretical basis for
assuming indent strength to be equal to the Brinell hardness
value is described in a previous article (TSP, Spring, 2004,
“Prediction and Control of Indent Diameter”).

EQUATION FOR PREDICTION OF INDENT DIAMETER

Combining equations (1), (2) and (6) gives that:

d = 1.278D.P0.25.ρ0.25.v0.5/B0.25 (7) Fig. 3. Effect of drop height on indent diameter using

2 mm diameter weighted ball.

30 The Shot Peener | Fall 2020

 ACADEMIC STUDY Continued

COVERAGE

INDENT RATE
A key factor in coverage control is the number of dents that
are being produced per unit area per unit time, i.e., rate of
denting. The rate of denting is controlled by the shot feed rate
and the average size of the dents. As an example, assume that
an average of 50 dents, with each dent having an area of 1mm2,
are being produced every second for each area of 100mm2.
The rate of denting is therefore 50 x 1mm2 x s-1/100mm2 or
0.5s-1. Note that the units for area cancel each other out. Fig. 4. Seven and forty-two “craters”
distributed randomly.
The rate of denting can easily be measured. A polished
strip of the same hardness as the component can be peened much greater chance of tiny uncovered areas being present.
for a short time so that individual dents can be counted. For a These features can easily be seen using this type of model but
known area of strip, the number and average size of dents can difficult to identify for real peened components.
therefore be measured and converted into a denting rate.
COVERAGE ESTIMATION
COVERAGE In the author’s opinion, coverage can only be estimated. It
Coverage is generally defined as “The percentage area of cannot be measured with a high level of accuracy. Estimation
the peened surface that has been dented.” This sounds very techniques fall into three categories: manual comparison,
simple but “the devil is in the detail.” As a definition, it is manual measurement and computer-based image analysis.
incomplete! On a microscopic scale, coverage is a mixture Manual comparison involves simply comparing an
of 100% and 0%—either dent or not a dent. A more precise image of a peened area with reference images that span a
definition would therefore be: “For a specified area of a range of coverage levels. Fig. 5, copied from J2277, is a typical
peened surface, coverage is the percentage of that area that example of this useful, quick, but rough method. There is,
is comprised of dents.” If the defined area is reasonably large, obviously, a subjective element to this technique. An image
statistical variation of denting will then be averaged out. of a selected peened area can be photographed, via say a
Estimation of high levels of coverage is so tricky that the smartphone, and then downloaded to a computer for side-
term “Full Coverage” has been included in specifications as by-side comparison with stored reference images.
corresponding to 98%.
A major practical problem is to define the precise
area of each dent. Life becomes much simpler when using
models with clear defined edges. Coverage is then commonly
explained using a model based on the random distribution Fig. 5. Manual comparison
of identical circular dents. Fig. 4 is a typical example. Seven images.
and forty-two circular dents have been distributed randomly
with their centers all inside the outer square. In order for the
model to be accurate, coverage has to be measured using the
amount of “greying” inside the specified red square. Coverage
then becomes:

% Coverage = 100 x total “greyed” area/specified “red” area

The larger the number of dents within the specified area

the greater will be the coverage. This will correspond to either
an increase in the length of the peening time or an increase in
shot flow rate. Manual Estimation
Fig. 4 illustrates the characteristic features of increasing Every shot peener should be capable of carrying out a manual
coverage. At a low coverage level, individual dents can be estimation of actual coverage for a shot-peened component.
identified and there is only a small proportion of dent overlap. Fig. 6 illustrates the procedure, using identical spherical dents
As coverage level increases there is a much greater chance of whose centers lie within the defined greyed rectangle. To
dent overlap and also of multiple overlapping. There is also a show how easy it is: photograph fig. 6, download the image to

32 The Shot Peener | Fall 2020

 ACADEMIC STUDY Continued

Fig. 6. Manual technique for estimating coverage.

Fig. 7. Lineal coverage analysis of a shot-peened specimen.
a computer, adjust the image to a width of say 200 mm, print
in landscape and then measure and add the green distances Drawbacks are that the identification of artefacts becomes
between the black arrows using a metric office ruler. I found difficult and dent edge location can be imprecise. Image
lengths of 158, 158, 136 and 178 mm for lines 1 to 4. Dividing analysis programs are readily available but require training
by two (to allow for 200 mm as the reference length) gives 79, and experience to be effectively employed. At least one
79, 68 and 89% for coverage. The average is 79%. Variation company provides a complete unit for coverage estimation.
between test lines inevitably occurs over micro distances.
The lineal manual technique just described does contain COVERAGE PREDICTION
a small element of subjectivity. This is mainly due to having Coverage prediction is a jewel in the crown of shot peening
to decide where a line intersects a dent. It is, however, much science.
more precise than the manual comparison technique. As peening progresses, coverage increases but becomes
Applying the lineal manual technique to actual shot less and less effective as more and more of the surface is
peened components requires a slightly different approach. already dented. Fig. 8 on page 36 (the figure is copied from a
First photograph a test area of the peened component previous article) is a graphical representation of the changes
and paste it into, say, Word. Crop a suitable rectangle after in coverage rate that occur for a constant indentation rate.
appropriate image magnification. In landscape mode, As is universally recognized, the coverage rate decreases
magnify the cropped area to give a convenient reference with increase in shot peening time (or number of passes).
width (say 200 mm). Horizontal lines can then be drawn The mathematical shape of the curve is called “inverse
on a printout at convenient equal intervals. Measure (a) the exponential”. This shape can be expressed as:
length of the reference line and (b) the intersections of each
Ct % = 100[1 – exp(-A*t)] (8)
line with dents. Sum the intersections for each line and divide
by the length of the reference line. Multiply by 100 to obtain where Ct % is the coverage after a peening time t and A is the
percentage coverages. indent rate.
Fig. 7 illustrates the use of manual lineal analysis for A single measurement of coverage can be used to predict
an actual peened specimen. For the lines L1 to L4 the the amount of peening (time or passes) needed to achieve
coverages were 31, 30, 34 and 42%, giving an average of 34%. a required level of coverage. An Excel-based program is
Nine dents appear on the photograph, together with two available free from Shotpeener.com as “Coverage_Predictor.
artefacts—marked X. The artefacts should be ignored as not xls”. Fig. 9 on page 36 illustrates how the program can be
corresponding to actual dents. applied. Entering the measured value of coverage after one
pass, say 42%, automatically predicts coverage after different
Computer-Based Image Analysis numbers of passes. It also calculates the indent rate, A.
Computer-based image analysis is very similar in principle
to the manual technique just described. The main difference DISCUSSION
is that thousands of lines can be measured very quickly. Experienced shot peeners are well aware of the effects of shot

34 The Shot Peener | Fall 2020

 ACADEMIC STUDY Continued

Shot Peening
NDT
Mass Media Finishing
Aerospace, Military &
Commercial Approvals
FAA Repair Stations
KJ1R272K (CT) &
G89R878X (GA)
On-site Capabilities

Fig. 8. Variation of coverage rate with peening time.

Surface Enhancement (CT & GA)

Nondestructive Testing (GA)

Fig. 9. Example showing application of Coverage

Prediction program.
size, velocity, density and component hardness parameters on
the dent size that is induced. Semi-quantification is achievable
through experience and, importantly, the ability to refer back
to stored data. Hopefully, the relationships presented in this
article will increase understanding of quantitative parameter
effects.
Coverage has a long history of published explanations
and has its own standard specification, J2277. One significant
feature in this article is the challenge to readers to actually
carry out quantitative measurements. The lineal method
proposed has been used by the author ever since his
undergraduate days. A simple explanation of the method
has been included based on that long experience. Somewhat
peentech.com surprisingly, a Google search revealed only very complicated
accounts. l
8 Eastern Park Road 3117 Emery Circle
East Hartford, CT 06108 Austell, GA 30168
860-289-4328 770-941-9573
Established 1966

36 The Shot Peener | Fall 2020