1. OPTIMISATION OF THE SHOT PEENING PARAMETERS
A research project sponsored by USF Vacublast, in co-operation with Design Unit
16/11/1998-15/11/1999
By
Franck PETIT-RENAUD
University of Newcastle-upon-Tyne
Department of Mechanical, Materials and Manufacturing Engineering
M. Phil Thesis
September 2000

2. The University of Newcastle-upon-Tyne
Department of Mechanical, Materials and Manufacturing Engineering
M. Phil Thesis
Academic Year 1998-1999
Franck PETIT-RENAUD
OPTIMISATION OF THE SHOT PEENING PARAMETERS
Supervisors: Dr. B. SHAW and Dr. J. T. EVANS, University of Newcastle
Mr R. DICKINSON, USF Vacublast Ltd
September 2000
This thesis is submitted in partial submission for the degree of Master of
Philosophy in Mechanical, Materials and Manufacturing Engineering

3. To Seana, for her support, love and patience,
And to my parents for their encouragement.
With special thanks to Françoise Parrain who made this possible.

4. I
ACKNOWLEDGEMENTS
The author is extremely grateful that USF Vacu-Blast agreed to sponsor this one-
year project. As his industrial supervisor, Mr R. Dickinson was at all times very
supportive and interested in the progress made during the investigation. This
commitment from a leading industry was very helpful and motivating.
The support and advice from his academic tutors, Dr. B. A. Shaw and Dr. J. T.
Evans, was also very valuable as their interest of the subject led to an
enthusiastic attitude.
Special thanks are due to Dr. A. V. Metcalfe for his help on the statistics,
providing resources and advice in the design of the experiment and analysis of
the results.
Others such as Phil Wilson and technicians in the workshop have been also very
kind in helping me, answering my questions and providing technical support with
the equipment.

5. II
ABSTRACT
The shot peening process is a complicated process in the materials science,
which is still not fully understood. Indeed, despite a long existence and many
investigations into this metal finishing process, it still exhibits many areas of
uncertainty.
Until recently, shot peening, also often described as a “Black Art”, was not even
regulated by any international standards and most of the process knowledge was
based on the experience of the few industries involved in applications where shot
peening was required. Nevertheless, for years, the aerospace and automotive
industries have considered shot peening as a state-of-the-art process for
cleaning, forming and improving the lifetime of many parts.
The majority of the information and data available on this subject are the result of
many years of experience and in-house research carried out by the main
industries making use of shot peening. It is therefore still difficult to get detailed
practical information other than general theories, which are published openly.
The work described in this thesis is a study of the effect of a range of process
parameters on the residual stress profiles produced by shot peening bars of case
carburised 17CrNiMo6 steel.
Shot peening was achieved using a commercial shot peening unit supplied by
USF Vacublast, using only one size of shot (0.6mm diameter). The process
parameters investigated were the air pressure, the mass flow, the impact angle,
the distance nozzle-specimen, the exposure time and the nozzle size. Using the
appropriate software (Minitab v12), regression analysis were performed on the
results obtained from statistically designed experiments, presented below. It was
found that the important (most significant) parameters were the air pressure, the
mass flow, the impact angle and the exposure time. Further significant
interactions were also detected between the following parameters: exposure time
and air pressure, nozzle size and mass flow, air pressure and impact angle,
nozzle size and air pressure.

6. 0-1
INTRODUCTION
The study of the different parameters involved in shot peening applications is
important in order to have better understanding and control of such process. The
significance and influence of these parameters are not yet clearly established
and most of the knowledge is based on practical experience rather than detailed
research. There are only limited methods of assessing the results obtained from
peening (e.g. Almen strips) and prediction of final properties is not possible yet.
The investigation presented in this thesis was aimed at designing and carrying
out experimental procedures in order to understand the effects of shot peening
on components by analysing the changes occurred during the process.
Therefore, determining the parameters involved to carry out the process and
measuring residual stress in peened specimen were the two objectives of the
investigation.
Being able to relate the shot peening parameters directly to the result produced
by the process would indeed be of great advantage as it could lead to a better
and more accurate control of shot peening. It would mean predicting the result
induced by peening a component and increase the reliability of such process [40-
60].
Even if the variables necessary to the peening process have been known for a
long time, it is still a difficult task to control effectively most of them and the actual
effects of these parameters combined together can not be predicted accurately
and reliably. Usually, from an industrial point view, the process will rely on one,
two, sometimes three parameters (air pressure, impact angle, exposure time) to
set and control, keeping all the others (shot size and hardness, nozzle size…) as
constant as possible [44][48]; the process is qualified by two measures: coverage
and Almen intensity.

7. 0-2
This is enough to meet requirements set by the standards. However, it would be
a breakthrough to be able to set up any peening equipment, knowing that a
particular combination of the parameters will give a unique result.
Starting by describing the peening process, as it is known, this thesis will be
organised around two main topics: parameters and residual stress. Focusing on
the variables that were thought to have an effect on the process, experiments
have been designed and carried out with the only objective of measuring the
variation caused on the residual stress within the material.
A literature review was also carried out, investigating current and past research
programs. This step was necessary, as valuable information (e.g. results
obtained, findings/progress and experiments described) was needed to set up
the investigation from a different angle. It was therefore important to keep this
investigation as general as possible in order to have the widest range of results,
even though the main objective was to understand shot peening of hardened
steel components (e.g. gears) [50][53]57].
Full descriptions of the equipment used to carry out this investigation and
preliminary work (e.g. set up of the equipment, calibrations of some of its
elements) are also presented, leading to the two main experimental programs
and the results obtained.
It was not intended to generalise the whole process but, focusing on what
seemed to be the heart of shot peening and limiting the investigation to one type
of material. Measuring the change in residual stress within the specimens used
was thought to be the most interesting and useful way of understanding the
process.

8. 0-3
A statistical approach to the problem was used to design all the experiments and
specific tools for the analysis of the results considered.
It is believed that this investigation was one of the most complete in terms of
relating different set of parameters to different responses; the considerable
number of residual stress measurements carried out could be used as a good
basis for an even wider research program into the process, aiming at building
software and database, dedicated to produce known effects on components and
increasing the reliability of the shot peening process.

9. 1-1
CHAPTER 1: THE PROCESS
1.1. INTRODUCTION
In this chapter, the shot peening principles will be presented. The known effects
of the process will also be described and illustrated.
Some common industrial applications will be used as examples to give a general
view of the process and show how widely shot peening is used in the aerospace
(e.g. structure components), medical (e.g. titanium replacement limbs), car (e.g.
gears, shafts) industries.
1.2. THEORY AND FUNDAMENTALS OF THE SHOT PEENING PROCESS
1.2.1. Process description
Shot peening is a cold surface working process [1] in which a stream of small
spherical shots, propelled at high velocity and under controlled conditions, are
bombarded onto a metallic component or target causing a thin layer of the
exposed surface to deform plastically (Figure 1-1).
Figure 1-1: The shot peening process.

10. 1-2
Many types of shot can be used for this purpose; however, the most commonly
used ones are the steel shot and glass beads (Appendix 1).
In order to accelerate the shot, special equipment is required. There are
essentially two major types of peening machines in use:
The air-blast peening machine;
The centrifugal blast peening machine.
The air-blast peening machines use compressed air in conjunction with a
convergent nozzle to propel the shot, whilst the centrifugal-blast peening
machines use the centrifugal action of one or more high speed rotating wheels
for the same purpose [1].
Other types of peening or blasting application are using other means than air or
centrifugal force to accelerate the shot [1][2]. Indeed, water peening/blasting or
ultrasonic peening uses respectively highly pressurised water or ultrasonic waves
to accelerate and propel shot towards the components to be treated. Laser
peening is also a process that is being developed [55]. The beam of dye lasers
are used to excite the atoms at the surface of the target, generating heat and
producing a high amplitude pressure at the surface of the material.
1.2.2. Process effects
The immediate effect of bombarding high velocity shots onto a metallic target is
the creation of a thin layer of high magnitude compressive residual stress at or
near the metal surface, which is balanced by a small tensile stress in the deeper
core (Figure 1-2).
The magnitude of this compressive residual stress is a function of the mechanical
properties of the target material and may reach values as high as 50 to 60% of
the material’s ultimate tensile strength [1][3].

11. 1-3
Figure 1-2: Effects of shot peening.
Its depth is largely dependent on the peening intensity and the relative hardness
of the impinging shot and target material. For a relatively soft target material
(230-300 HV), it is feasible to produce a compressive layer of 800 to 1000µm
deep, whilst for a harder material (700 HV), it can be difficult to produce a
compressive layer of much more than 200 to 250µm [1][2].
The introduction of this compressive residual stress at the metal surface layer
brings one major benefit: it reduces and can negate any residual or subsequently
imposed tensile stress at the metal surface [26][28]. As it is well known, most
fatigue failures and stress corrosion failures normally start at or near the surface
stressed in tension [2][3]. Therefore, by reducing the net tensile stresses at and
near the surface of the component, fatigue crack initiation and stress corrosion
can be delayed, improving the fatigue life of the component
treated[27][29][30][31].

12. 1-4
If the resultant surface stress can be made compressive enough, cracks could
virtually be prevented from opening up at the component surface resulting in a
much enhanced fatigue life [3][5][6]. This is generally true for shot-peened
components subjected to low stress amplitudes.
Another process effect of shot peening is the tendency of a thin metal work-piece
to curve up towards the peened surface [14]. This phenomenon confirms that
internal stresses have been produced within the treated work-piece [1][2][3]. If
the shot-peened compressed layer is carefully removed, the strip will theoretically
return to its original flat condition, demonstrating that the internal forces causing
the curvature are confined to a thin layer at and near the surface. This tendency
for the work-piece to curve up towards the shot-peened surface can be explained
by the fact that plastic strain has taken place at the top peened layers, thus
requiring to occupy a greater space; but as this is opposed by the elastic layer
underneath and causes the thin plate to curve. In the resulting equilibrium, the
upper surface layers are subjected to compressive residual stress while the inner
layers are subjected to tensile residual stresses. Of course, the net force on the
material arising from the residual stresses is zero. This phenomenon has also
been widely exploited in industry in the precision forming of thin metal parts such
as aircraft wings.
1.2.3. Process parameters
The shot peening process has to be a precisely controlled and repeatable
process for optimum benefit. To achieve this, all its process variables must be
identified and controlled [7][8]. There are many fundamental parameters affecting
the shot peening process (Appendix 2).

13. 1-5
The most common are as follows:
Shot density;
Hardness and size of the shot;
Nozzle characteristics (diameter, deflection angle, length);
Air pressure:
Impact angle;
Distance from nozzle to work-piece;
Exposure time, number of passes;
Linear and rotational speed of work-piece relative to nozzle.
To specify all these variables every shot peening job would require time
consuming investigations and industrially impractical procedures. To overcome
this problem, J. O. Almen [4] [14] [15] introduced the concept of peening intensity
measurement based on curvature induced in a thin test strip, by which most of
the previously listed process parameters can automatically be incorporated into
one process variable called the Almen peening intensity [2][3][4]. With peening
intensity known, one has only to define the shot type and size and peening
coverage desired to fully define the peening process.
1.2.4. Industrial applications of the shot peening process
The applications [3][37] of shot peening are consequences the basic effects
described previously (The introduction of high magnitude compressive residual
stress of some finite depth to the material surface, a controlled forming of thin
metal sections and a controlled surface deformation or dimpling). Invariably, all
these effects take place simultaneously in any shot peening operation. However,
for a particular application, only one or two of these effects are being exploited.

14. 1-6
Listed below are some of the most common shot peening applications:
Surface compressive residual stress effect: improving fatigue strength of
dynamic components, resistance to stress corrosion cracking, reducing or
eliminating residual tensile stress introduced by heavy grinding, electro-
discharge machining and welding, treatment before plating [40].
Controlled forming of thin metal sections: peen forming and straightening of
distorted thin sections [1].
Controlled surface dimpling or deformation effect: improving parts lubrication
and lowering of machinery noise level, favouring resistance to inter-granular
corrosion, increasing coefficient of friction on brake disc and wedges,
detecting and improving decarburised surfaces, adhesion testing of silver
plate, reducing seal leakage, anti-glare treatment of reflective surfaces,
decorative texturing of surfaces [3][35][43].
Increase of the surface hardness of treated parts can also be observed. This
fourth effect of the shot peening process would improve resistance to fretting,
post electrochemical treatment and fatigue [1].
As experience and various studies have demonstrated the improvements
induced by the peening process, it is widely used to enhance the life of
components operating in highly stressed environment and other critical parts
such as in Formula 1 motor racing, aeroengines and aerostructures [32][33][34].
Despite important progress in understanding the process, some areas are not
totally mastered yet and difficulties are still hard to avoid. Being able to predict
the effect of the process in set conditions is indeed the key to gain complete
control over the process and to make it much more reliable.

15. 1-7
1.3. OBJECTIVES OF THE INVESTIGATION
This global investigation is aimed at widening the understanding we have of the
shot peening process by trying to relate the main parameters to the actual result.
Indeed, the change in the residual stress within the components should
somehow be linked to the process conditions.
As the only standardised means to qualify the peening process consists of
measuring the curvature induced on flat strips (Almen intensity), the main
objective was to find an alternative route as it was thought that this was not
representative of the different components peened. Indeed, components made
from different materials (Aluminium, titanium, and steel…) will not “respond” the
same way to the process. The ideal case would be to actually get strips made of
the same material than of the components to be processed to actually qualify the
Almen intensity as relevant to the component.
First of all, a better understanding of the process was thought to be necessary.
By analysing residual stress profiles obtained from peened specimens, the
objective was to be able to evaluate the importance and significance of the
parameters as a whole and individually. Different settings of parameters were
then carried out and their direct effect on the residual stress investigated.
Using statistical tools, the significance of a set of parameters could be evaluated
and the parameters were individually assessed. Finding out which parameter
was more or less significant could then lead to a pattern and it was hoped that
the results obtained would help in optimising the process.

16. 2-1
CHAPTER 2: SHOT PEENING: IMPACT, COVERAGE, INTENSITY
AND SATURATION
2.1. INTRODUCTION
Focusing the different database searches between 1990 and 1999, about 300
articles were directly relevant to the shot peening process, covering general
applications of the process and some aspect of its mechanisms. As the main
objective of this work was investigating the relationship(s) between the shot
peening parameters and the compressive residual stress introduced, about 65
articles, only focusing on the effects of the process on fatigue life of components
and the residual stress introduced have been selected and will be discussed in
this section as well as referred through this thesis.
A search for projects on shot velocity was also carried out with less success. The
very few investigations achieved in this area were mostly focused on the
following parameters: shot diameter, shot density and air pressure. Indeed, it was
considered that the actual shot velocity was directly dependant on these
particular parameters and the experimental results obtained seemed to match
the theory quite well [9].
However, a common technique to investigate the process effects was to
establish relationships between residual stress and Almen intensities in different
types of conditions but not necessarily as the consequence of the influence of
one particular parameter. In all the cases described, the process and its effects
were studied from a general point of view and the real impact of each parameter
was not the main objective.

17. 2-2
This section will aim at presenting the mechanisms of the shot peening process
and will discuss different aspects of the process the most often investigated.
2.2. THE IMPACT THEORY
The shot peening is a cold working process that involves hitting the surface of
critical components (e.g. gears) with spherical parts projected by compressed air
at velocities between 70 and 350km/h (20 to 100m/s). It is widely used to
improve fatigue behaviour of metal components, as the main effect is to
introduce compressive residual stress as well as increasing the surface
hardness.
Such process can be divided into two separate stage [8][20]. As each particle
acts as "hammer" (Figure 2-1) when hitting the surface of the component, this
undergoes plastic deformations at the impact location. As the shot rebounds, the
elastically deformed layer expands, pushing the plastically deformed parts,
resulting in the creation of compressive residual stresses.
Figure 2-1: Particle indentation at the surface of the work piece.

18. 2-3
As shown by previous research [8][18][20], the residual stress state of the work
piece can be described by two distinctive phenomena (Figure 2-2). First of all,
the first stage, also known as the Hertzian pressure, is responsible for
introducing compressive contact stress as the particle hits and moves into the
treated component. These residual stresses vary over depth and a maximum is
normally achieved near the surface. Then, in a second stage, as the particle
rebounds, plastic stretching of the surface layer creates residual stresses with a
maximum at the surface. As a result of both phenomena, and in the case of a
totally covered area (overlaying impacts), the residual stress distribution will vary
in depth and be uniform in parallel planes to the surface treated [8][20].
Figure 2-2: A single impact- Description of the two stages.
These phenomena mainly depend on the following parameters:
The radius of the shot;
The initial velocity of the shot;
Hardness of the spherical particle and hardness of the work piece;
Thickness of the work piece;
Exposure time or coverage (See also section 2.3.).

19. 2-4
Figure 2-3: Relationship between indentation and shot size.
Assuming that the particles are perfect spheres, with a constant velocity and that
the area density of impacts is uniform it is possible to relate the shot size and the
shot velocity to the diameter of the indentation [8].
It has been previously shown [62] that plastic deformation has to occur as a
result of the impact if residual stresses are to be generated. Thus the yield
strength σy of the specimen must be an important factor. The velocity V of the
shot must also be important. Jonhson [62] has also showed that a critical velocity
Vc exists for the onset of plastic deformation. For a steel sphere impacting on a
flat steel surface, the critical velocity is described by the following relationship:
2
*5 





=
E
Vc
yy σ
ρ
σ
(Eq. 2-1)
where ρ is the density and E is the Young’s modulus. For instance , if σy =
2500MPa (hardened steel), the critical velocity is 1.4m/s. At smaller velocities,
elastic deformation alone occurs.
It is also possible to estimate the depth of the region plastically deformed by the
impact of a single shot of radius R.

20. 2-5
Indeed, the radius of the plastic zone is approximately proportional to the radius
of the indent and is given by:
3
1
)1(*6
)21(*4
*












−
−+
=
ν
ν
σ R
aE
ac
y
(Eq. 2-2) [62]
where c is the depth of the indentation, a is the radius of the indentation at the
surface and ν is Poisson’s ratio.
This radius can also be calculated with the next relationship:
RVa
y
** 2
14
1








=
σ
ρ
(Eq. 2-3)
Thus a is directly proportional to the radius of the shot R. a simple estimate can
be made by considering realistic estimates of the process parameters. Taking R
= 0.3mm, ρ = 7.8.103
kg/m3
, V = 10m/s, σy = 2400MPa, ν = 0.3 and E = 200GPa,
we obtain a plastic zone radius c = 60µm.
Inspection of the above equations shows that the depth of plastic deformation
from a single impact is only approximately proportional to the radius of the shot
and proportional to the square root of its velocity. However, as explained
previously, the shot peening process consists of many impacts over the surface
and the interactions between individual particles of shot and the surface, it is
much more complicated than that for a single shot. The effect of the shot peening
process will consist of the superposition of the effects of individual impacts since
deformation is non-linear, superposition is not simply linear.

21. 2-6
Previous research work in this area, also showed that it was possible to calculate
models and simulate shot peening impact. However, as the theoretical models
obtained were based on the analysis of a single impact it is not fully
representative of the process. Indeed, interactions between the particles before
hitting the surface and at the moment of impact should be considered as they
might alter the actual work produce by the impacting particles.
Considering that the effect of the shot peening process is the combination of
multiple and random impacts, the study of single impact models provided useful
insight but was not fully relevant to the present work.
2.3. COVERAGE, INTENSITY AND SATURATION
The coverage is defined the percentage of area that has been exposed to the
shot peening process over a certain period of time. The following diagram (Figure
2-4) shows different stages, from isolated to overlapping impacts.
Figure 2-4: Different stages towards full coverage.

22. 2-7
At the beginning of the process, it is most likely to observe many isolated indents,
leading to a near linear increase of coverage with time [2]. As the exposure time
increases, the area to be covered decreases, leading to more and more
overlapping impacts and to decreasing coverage rate. Over a long period of time,
a much smaller amount of the initial area remains free of impacts and the
probability of it to be covered as well becomes smaller.
The general theoretical approach of coverage can be assumed as being
exponential, meaning that 100% can always be approached but theoretically
never equalled [2][12].
An equation of the following type can then be assumed being a good
representation of coverage:
−= 1(*100C e-[f(x)]
) (Eq. 2-4)
where C is the percentage of the area that has been covered and f(x) is function
that is dependent on the shot diameter, the diameter of the impression, shot
velocity, impact angle, and properties of the shot and work piece. This type of
equation is also known as an Avrami equation [12], a representation of which can
be seen in Figure 2-5.
Figure 2-5: Mathematical representation of coverage.

23. 2-8
Assuming that the parameters listed above provide a uniform rate of coverage
(all constant), the actual Avrami equation is then written as follows:
)]exp(1[*100 2
RtrC π−−= (Eq. 2-5)
Where r is the radius of the impressions (all assumed constant), R is the rate
(uniform) of creation of impressions (also called coverage rate) and t is the time
during which the indentations are created (also called exposure time).
From the industrial point of view, the evaluation of coverage is left to the
operator's judgement and experience, who inspects the work piece using a light
and a magnifier. Other methods consist of spraying the components with special
paints of dye tracers prior to the process [16]. Once shot peening has been
carried out, coverage can be checked with a magnifier and fluorescent lights. The
better the coverage, the less paint or dye should be "visible".
Another practical way would be, when possible, pre-peen the component with
some Aluminium Oxide grit at low intensity. Indeed, the surface of the workpiece
will have a particular colour (mat grey) and, once traditionally, shot peened at a
satisfactory coverage, this coloration should be gone.
A practical example of partial and "full" coverage is shown in the following
photograph (Figure 2-6).
Figure 2-6: Coverage-Partial coverage (left) and full coverage (at least 100%; right)

24. 2-9
Coverage is not enough to evaluate and control the effect of the process on a
work piece. A standardised method to carry out the control of the process is
known as the Almen strip technique [14][15].
From an engineering point of view, it is important to be able to quantify the
effects of the different parameters involved in any process. In this case, the use
of Almen strips is a recognised method.
The figure 2-7 describes the technique accurately.
Figure 2-7: The Almen strip principle.
A thin strip of steel (SAE 1070), of known dimensions (Length, Width and
Thickness) is secured on a suitable holder and shot peened with the desired
parameters set and to be controlled, for a known period of time. It is then
removed from the holder and the curvature, a consequence of the introduction of
compressive residual stress at the surface, is measured with a calibrated gauge.
The value measured corresponds to the shot peening intensity, for the
parameters set and the exposure time.

25. 2-10
These operations are repeated for different period of time, each time with a new
Almen strip.
The next stage is to plot the arc height as a function of the peening time (or
exposure time). The curve obtained is called the saturation curve (Figure 2-8).
Figure 2-8: A typical Almen curve.
If the arc height or Almen intensity does not vary by more than 10% when
doubling the exposure time (T2), it is assumed that saturation has been reached.
T1 is then called the saturation peening time and the corresponding Almen height
is the shot peening intensity for the set of parameters chosen.
Combined with the visual control of coverage, this technique is the only one
recognised as the mean of controlling the process.

26. 3-1
CHAPTER 3: EXPERIMENTAL METHODS
3.1. INTRODUCTION
This chapter gives the details of all the experimental aspects of this investigation.
Starting with a detailed description of the shot peening equipment and the
principle of its main elements, the selection and design of the specimen and the
masking technique will be presented.
The experimental procedures followed and the different techniques, such as X-
ray diffraction and hardness measurement, used to establish and evaluate the
effects of the shot process are also included in this chapter. The chapter will then
end with a short overview of the statistical treatment of the results.
3.2. EQUIPMENT SET UP
This section will deal with all the aspects related to the shot peening equipment
used to carry out this investigation. The different components of the machine,
their principle and function will be fully described for a better understanding of the
process investigated.

27. 3-2
3.2.1. The shot peening equipment: the Ventus VB 11432-75
The following picture gives a general view of the shot peening equipment.
Figure 3-1: The shot peening equipment.
It is possible to distinguish four major elements:
The cabinet;
The Storage hopper or reclaimer;
The pressure vessel, also called generator;
The dust collector.
Using such equipment to carry out shot blasting or shot peening can be simply
described as follows (Appendix 3).
The components to be treated are placed inside the cabinet, on a perforated
workplate. An operator, located outside the blast chamber, manipulates the
pressure-fed nozzle and the part, wearing the full-length rubber gloves fitted and
sealed in the cabinet front panel.

28. 3-3
A viewing window allows him to observe the workpiece as the shots hit its surface
at high velocities.
The process can be stopped at any time by releasing the foot valve placed on
floor.
The reclaimer, the pressure vessel and the dust collector are part of the system
that will provide the shot and supply the compressed air to the nozzle. It is also
used to recover the used media, recycling it before sending it back to the storage
hopper. This side of the equipment also carries out ventilation and vacuum of the
cabinet.
3.2.2.1. The cabinet
Figure 3-2 shows the cabinet used during this investigation.
Figure 3-2: The peening cabinet.
The cabinet, central component of the peening equipment, is a structure made of
welded steel plates.

29. 3-4
The operator, standing in front of the cabinet, is able to shot peen the parts
introduced inside the cabinet by holding the nozzle (1) through the rubber
gauntlets (2) sealed onto the cabinet. The control of the peening pressure is
achieved by using the foot valve (16). For safety reasons, this valve operates with
a blast cut off system, ensuring that the process cannot take place until the door
is closed, masking an air bleed (11). In any case, releasing the foot valve will stop
the process immediately.
The whole operation can be safely carried out and observed through the viewing
window (3), made of safety glass, protected by an outer screen of clear acrylic
and mounted in a rubber seal. Lighting inside the cabinet is provided by a fully
protected fluorescent unit and can be easily switched on and off by operating the
switch place at the front of the cabinet.
Air entry holes (8), located at the top of the cabinet, allow an air flow to carry on
full ventilation of the cabinet, carrying used abrasive, dust and debris generated
by the process, through the work plate (4) and back into the hopper below (5).
Inside this hopper, abrasive, dust and debris fall by gravity to the bottom; as air
flows through a take off tube (9) (Venturi), they are then conveyed through a slot
(10) and a flexible hose (11) straight to the reclaimer to be separated, cleaned
and sent back in the circuit.

30. 3-5
3.2.2.2. The generator-reclaimer assembly
The following diagram gives an overview of the heart of the equipment.
Figure 3-3: The generator-reclaimer assembly.
Sitting next to the cabinet, the generator-reclaimer assembly is linked to the
cabinet by the blast hose (18) and recovery hose. The dust collector (15) is
mounted on the same frame (10) and is connected to the reclaimer by a spiral
ducting.

31. 3-6
As the media, dust and debris are conveyed from the bottom of the cabinet, the
reclaimer will recover and re-circulate the re-usable media back to the process.
The reclaimer will then store this re-usable media that will fall by gravity through
an internally mounted dump valve (5) into the generator (9). An externally fitted
feed valve (8) will ensure the control of the flow of the media and mixing with the
air, when shot peening.
The reclaimer is made of two compartments, separated by a vibrating screen, an
air-wash device and the actual storage hopper (6). Media and dust accumulated
at the bottom of cabinet are conveyed through the flexible hose and reach the
reclaimer at the top inlet (12); the reclaimer the dust collector are connected by a
spiral ducting, from the reclaimer outlet (14).
A. The reclaimer
Figure 3-4: The reclaimer.

32. 3-7
The previous figure (Figure 3-4) shows the reclaimer, part of the peening
equipment that is designed to recover the used media, recycle it and send it back
to the process.
As the recovery airflow carries the media, the dust and the debris from the
cabinet, the reclaimer will carry out a separation process, aiming at cleaning and
recycling the media for a longer use and better performance.
Air, media, dust and debris reach the reclaimer through the inlet (1), which is
offset near the top of the body of the reclaimer. As a consequence of this
location, the airflow is subjected to a cyclonic influence, spiralling around the
upper section (2) in a descending movement and with the heaviest particles
outermost.
The lighter dust and air reach the annulus formed by the upper cylinder (3) and
the cone tube (4) and are drawn inwards and up into the annulus, passing then
vertically through the top of the upper cylinder and out through the outlet (6).
The remaining bigger and heavier debris, with the media continue spiralling
downwards, passing through the annulus (7) where the smaller particles are
washed away from the media by air flowing through the annulus (8) and sent
through the cone tube. It is possible to adjust the airflow induced inside the
annulus by moving up or down the cone (5), located at the top of the cone tube.
This smaller debris are then carried away the same as the dust.
Once separated from the debris and cleaned from the dust, the remaining clean
media falls into the storage hopper (11), where a vibrating separator screen is
located. The mesh of the separator retains oversize debris and let the good
reusable media go through into the hopper.

33. 3-8
B. The dump valve
Figure 3-5: The dump valve.
The dump valve, as shown in Figure 3-5 is located between the reclaimer and the
generator and forms a seal to retain the media in the storage hopper and to allow
pressurisation of the pressure vessel.
During the process, the operator activates the foot valve, releasing compressed
air, which flows to the inlet (1), through the hose (10), underneath (9) the
diaphragm (8). This action lifts the diaphragm and raises the plunger (7), sealing
the cone (2) into the aperture (3). The return spring (6) is then compressed and
the air above the diaphragm is expelled through the hose (5) towards the outlet
(4).
Once the dump valve is closed, compressed air flows into the vessel,
pressurising it and forcing the air to flow through the feed valve, mixing with the
media accumulated at the bottom of the generator and expelling the mixture
through the blasting hose towards the nozzle.

34. 3-9
If the operator releases the foot valve, the dump valve return spring pushes the
plunger and the diaphragm down, opening the cone, depressurising the vessel
and allowing cleaned media to fall by gravity.
C. The generator (Pressure vessel)
Compressed air is a continuous requirement of the process. So a pressure vessel
is built-in to ensure storage and supply of compressed air (Figure 3-6).
Figure 3-6: The pressure vessel or generator.
The pressure vessel fitted on the peening equipment has a volume of 60 litres.
This vessel, located at the bottom of the reclaimer, is bolted with the dump valve
positioned inside its top aperture.
During the process, the vessel or generator is pressurised with compressed air
through the hose (4). The air mixes with the media at the bottom of the pressure
vessel and the mixture is expelled from the generator through an aperture, filling
the body of the feed valve (8), metering the mixture flow through the blast hose
(10).

35. 3-10
As described previously, the generator pressurises to deliver the air and media
mixture to the blast hose for transmission to the nozzle and depressurises on
completion of the process.
D. The feed valve
Figure 3-7 illustrates the device responsible for metering the amount of steel shot
and to ensure a good mix with the compressed air.
Figure 3-7: The feed valve.
This device aims at metering the quantity of media mixing with the compressed
air before being expelled into the blast hose.
During the process the body of the feed valve body (11) is always full of media.
Air from the generator enters the feed shaft through the orifice (1) and media
drops from the body (11) through the orifice (5) and the slot (3) to enter the shaft
and being mixed with the compressed air. The mixture is then expelled through
the shaft towards the outlet (2) of the feed valve, directly into the blast hose.

36. 3-11
The lever (8), when operated, rotates the shaft (4), closing and/or opening the
slot (3).
Adjusting the position of the lever increases or decreases the effective area of the
orifice, allowing more or less media to mix with the compressed air.
The bottom plate of the feed valve is fed with air to fluidise the surrounding media
and ensuring free flowing through the slot (3). The tee bar (9) is used to tighten
the bottom plate onto the feed valve body and can be loosened to allow drainage
of the media during cleaning and maintenance operation of the equipment.
3.2.2.3. The dust collector
Figure 3-8: The dust collector.
The dust collector (Figure 3-8) is a box built with welded steel sheets and can be
divided into three distinctive sections.

37. 3-12
Each of these sections, listed below, enables the dust collector to reclaim the air
full of dust, clean it and to separate other debris.
The lower section (1), a hopper with a 20 litres discharge bin (9) clamped to
its bottom flange;
The centre section (2), a filter chamber that houses 6 cylindrical filter
elements (3) held in an inclined position.
The upper section (4), the fan chamber.
The dust and fine debris released in the cabinet during the process are conveyed
to the filter section by the airflow generated by the centrifugal fan (5). This dust
enters the dust collector inlet (7) and is deflected by the plate (8). At this stage,
the air is subjected to an initial separation due to the sudden slowing of the
airflow as it changes direction and passes from the small section of the inlet into
the larger area of the chamber.
As a consequence, the heavier dust particles are released and fall by gravitation
through the hopper section (1) into the clamped discharge bin (9). The finest dust
particles and debris are drawn through the filter section where the air flows
through the material of the filter elements, leaving the dust suspended onto their
outside surfaces. Clean air passes through their centre and is exhausted to the
atmosphere.
3.2.3. Holding and positioning the nozzle
It was been necessary to modify the machine to adapt it to the purposes of the
present investigation.
As a result of some of the experimental requirements, positioning and holding the
nozzle inside the cabinet was one of the most important aspects of the set-up.

38. 3-13
The design of such a system had to meet the following requirements:
Allowing the set-up of different nozzle-specimen distances;
Allowing the set-up of different impact angles;
Nozzle efficiently tightened to avoid excessive displacement during the
process;
Simple and easy use.
Figure 3-9: Holding and positioning the nozzle.
As shown on the previous picture, a hard steel bar was fitted between the floor
and the top of the cabinet. Using adjustable clamps, a second bar was fitted on it.
The nozzle was also clamped and tightened by similar means on this second bar.
A 3-D movement of the nozzle was then possible and tests have shown reliability,
meeting the stated specifications needed.

39. 3-14
3.2.4. The test specimens
The dimensions of the test specimen are 10*10*100 mm. The material used was
the steel 17CrNiMo6, chosen because of its interest for gear manufacturing. The
manufactured bars were carburised, quenched, tempered and lightly ground
before being peened to maximise the effects of the process. At the end of heat
treatment, the bars were expected to exhibit a Vickers surface hardness of
approximately 700 kgf.mm-2
, which is typical of many case hardened gears.
Figure 3-10: The specimen dimensions.
In order to limit the number of specimens to be manufactured, a masking
technique was devised so that a number of different peening operations could be
carried out on one specimen.
3.2.4.1. The masking technique
It was necessary to design and manufacture a reliable masking technique, taking
account of the hostile working environment and harsh experimental conditions.
Figure 3-11: Masking the specimen.
10 mm
10 mm
100 mm

40. 3-15
The specimen was enclosed within a steel box, exhibiting a single round slot in its
centre, protected by a reinforcing washer. The dimensions of the box matched
accurately with the specimen’s dimensions, thus avoiding peening of the other
sites.
The positioning of the specimen inside the masking box was achieved by the
mean of two calibrated screws, allowing a translating displacement of the
specimen inside the box and maintaining it at a set position during the process.
One screw was marked at regular intervals; the displacement from one mark to
the next one ensured that 6 sites of equal size were regularly spaced. The
second screw was used to tighten the specimen at the desired location.
A removable plate was located on one side of the box allowing an easy release
and change of specimen when required. This allowed circular patches on the
specimen to be peened independently.
A total of 24 experimental sites could be obtained on each test specimen leading
to the requirement of 10 bars to carry out the investigation. In order to
differentiate each experimental site, the following code was used.
Each bar had one of its square sections marked. This mark was used as a
reference.
Figure 3-12: The experimental sites.
For instance, the site position S on the face F on the bar B had the code BFS.
Thus, the code number 524 referred to the bar 5, face 2, site 4.
Face 1
Face 2
Face 3
Face 4
Site 1Site2Site3Site4Site5Site6

41. 3-16
3.2.4.2. Holding and positioning the specimen during the process
Figure 3-13 shows a photograph of the system that was used to hold and position
the specimen in front of the nozzle during the peening process.
The fixture, fabricated with thick steel plates, was located inside the cabinet. The
specimen was kept straight and clamped in front of the nozzle, at a fixed position.
Figure 3-13: Holding and positioning the specimen inside the cabinet.
Such a set-up was suggested to improve the drainage of media during the
process, avoiding accumulation of steel shot on the area peened.

42. 3-17
3.2.4.3. Setting up a test
The alterations described previously were necessary to ease the set up of the
nozzle and the specimen inside the cabinet. As illustrated by the following
photograph, adjusting the nozzle to the required distance and angle as for
moving from one experimental site to the next were simple operations and a
considerable amount of saving in time was gained.
Figure 3-14: Setting up a test.
Further details of the shot peening experiments are given below, in the section
3.5.

43. 3-18
3.3. THE RESIDUAL STRESS DETERMINATION: X-RAY DIFFRACTION
X-ray diffraction is an accurate method of measuring the actual residual stress at
the surface of a specimen and provided a reliable method for quantifying the
effects of the peening process [49].
Metals are made of atoms, arranged in an ordered sequence. The structures
created are called crystalline nets or crystal lattice; they define particular planes
that will generate the diffraction of X-ray beams directed to the material studied
(Figure 3-15).
Figure 3-15: Material crystal lattice.
The crystal lattice planes are regularly spaced from each other by a distance d
called the inter-planar spacing.
Figure 3-16: Grain.

44. 3-19
The whole group of crystal lattice having their planes all orientated to the same
direction is called a grain (Figure 3-16) whereas material is made of different
identical grains orientated in different directions (Figure 3-17).
Figure 3-17: A group of grains: material.
Using the previous definitions, diffraction of X-rays to measure residual stress in
materials can be described as follows.
An X-ray beam of wavelength λ is emitted and directed to the surface of the
specimen (Figure 3-18) at an angle ψ.
Figure 3-18: The X-ray diffraction principle.
ΨΨΨΨ

45. 3-20
These X-rays will then be diffracted by the atomic planes of the crystal lattice.
The wavelength λ, the angle of the diffracted beam θ and the inter-planar
distance d are related by the Bragg’s law and describe the following relationship:
λθ =sin2d (Eq. 3-1)
Residual stress measurements are achieved by measuring the angle of reflection
θ in relation to the angle of incidence ψ, which varies dependently of the material
composition and the residual stresses present. A peak of diffraction is obtained
by exposing the surface of a specimen with the incident X-ray beam at different
angles ψ.
The presence of stresses in a material will change the inter-planar spacing,
generating a displacement of the peak of diffraction measured (Figure 3-19).
Figure 3-19: The displacement of the diffraction peak.
The deformation ε within the material is defined by the following relationship:
hklhkl
d
dd
d
d







 −
=




 ∆
=
0
0
0
ψ
ε (Eq. 3-2)
Differentiating Bragg’s law, the next formulae is then obtained:
θθε 2.cot
2
1
0 ∆−= (Eq. 3-3)
The determination of the stress is done using the relationship:
εσ E= (Eq. 3-4)

46. 3-21
The residual stress determinations were carried by the sin2
ψ method, using the
Cr-Kα radiation, generating the (211) diffraction peak from the austenitic steel
matrix [5][6].
The irradiated area for the X-ray measurements was of approximately 2mm in
diameter and five ψ angles were used (0, ±30°; ±45°).
ΨΨΨΨ
Figure 3-20: Principle of the X-ray diffraction equipment.
The equipment used was equipped with two detectors (Figure 3-20), provided a
better accuracy of the measurements acquired.
Plotting the representation of d as a function of sin2
ψ (Figure 3-21), calculation of
the residual stress was achieved by determining the slope p of the straight line
and combining (3-2) and (3-4) as follows:
p
E
ν
σ
+
=
1
(Eq. 3-5)

47. 3-22
Figure 3-21: Representation of d = f (sin ψ)
2
.
As the main goal of this investigation was the study of the effects of the shot
peening process on the residual stress within the material, profile measurements
of the stress were carried out by etching the specimen to remove layers of
material (Appendix 5). Following the procedure and principles previously
described, a value of the residual stress corresponding to each layer removed
was obtained. Plotting the stress measured against the depth was carried out and
residual stress profiles obtained.
These profiles could then be used to determine the stress levels within the
material and evaluate the effects of the process.
3.4. HARDNESS MEASUREMENT AND MEDIA CHARACTERISTICS
3.4.1. Surface hardness of the test specimen
To ensure that the surface hardness of the bars manufactured was of the
expected level, some tests have been carried out.
Different types of hardness can be determined (Vickers, Rockwell, Knoop) and it
is essential to remember that they are all following a particular relationship:
A
F
HARDNESS = (Eq. 3-6)

48. 3-23
Where F is the load applied and A is the impression area.
The surface hardness of a typical specimen was tested using the Vickers
technique.
This method involved the use of a pyramidal diamond indenter, exhibiting a
typical top angle of 136°. Once positioned at the surface of the specimen to be
tested, a constant load F is transmitted to the indenter for a few seconds (Figure
3-22).
Figure 3-22: Vickers hardness test principle.
After releasing the indenter, a square impression is left at the surface of the
specimen and the length of the diagonal of the impression is measured.
The VHN (Vickers Hardness Number) is then calculated using the following
relationship:
VHN
F
d
=
1854
2
.
(Eq. 3-7)
where F is the load, in Newton (N);
d is the length of the diagonal of the impression, in meter (m).
The four faces of the specimen were measured and hardness profiles were
obtained (Figure 3-23), starting from 50µm deep and up to 3000µm from the
surface. Hardness 1, 2, 3 and 4 refer respectively to the hardness for the face 1,
2, 3 and 4.

49. 3-24
Error! Not a valid link.
Figure 3-23: Specimen hardness. Surface and up to 3mm in-depth.
It can be observed that the profiles obtained are similar. The average hardness,
for each face, is as follows:
Face 1: 570.6HV (54HRC);
Face 2: 620.5HV (57HRC);
Face 3: 585.1HV (55HRC);
Face 4: 603.7HV (56HRC).
The hardness measured on each face of the specimen are quite close to each
other, varying from 54 to 57HRC. Knowing that the shot used for this
investigation has an approximate hardness of 55 to 65HRC, which is harder than
the specimen, it is expected to observe some effect from the shot peening
process.

50. 3-25
3.4.2. The media characteristics
The media used to carry out the investigation was a round steel shot of grade
S230.
This is a high carbon steel material, hardened and tempered to ensure durability
and resistance to fracture. The grade S230 corresponds to a shot diameter of
approximately 0.6mm and is a standard size commonly used for shot peening
applications.
The chemical composition and other physical characteristics can be found in
Appendix 1.
The hardness of this shot typically ranges from 55 to 65 HRC (600 to 760 HV).
However, it should be noted that the overall hardness of the shot could increase
as a result of the peening process.
Indeed, one of the well-known effects of the process is a significant increase in
the hardness of the treated components and the first to be peened is the media
itself. Therefore, in order to ensure homogeneity in the shot hardness, a phase of
conditioning of the shot was carried out. This operation consisted in running the
equipment with the new steel shot to be used, pointing the nozzle towards a steel
plate. This process resulted in breaking the most fragile shots and hardening the
rest. Once recycled, the reusable shot was in a stable condition.
Comparing the surface hardness of the tested specimen and the characteristics
of the shot used, it can be observed that one of the basic requirements was
fulfilled. Indeed, to actually shot peen a component, the shot hardness needs to
be greater than the surface hardness of the treated part.

51. 3-26
3.5. DESIGN OF THE EXPERIMENTAL PROCEDURES
As this investigation had a broad spectrum of possibilities, designing the
experiments was a necessary step in order to focus on the relevant information
and establish the effects and significance of the process from a practical point of
view.
Listing the different parameters involved in the process and evaluating their
possible importance, six were finally chosen:
The air pressure;
The mass flow;
The impact angle;
The distance nozzle-specimen;
The exposure time;
The nozzle size.
These parameters have various degrees of importance. Indeed, the air pressure
and the mass flow are involved in the part of the process known to provide
energy to the shot, whereas the exposure time, the impact angle and the distance
will qualify the notion of coverage produced by the process.
From a statistical point of view, three experimental levels (Low, medium and high)
were assigned to each of the chosen parameters and a factorial design applied to
calculate the number of tests needed to produce a statistically significant
analysis.
It is also important to add that another parameter was considered but kept
constant all along the investigation for convenience: the shot size. Indeed, the
size, the hardness and other properties of the media used to carry out shot
peening are known to have great influence over the results of the process.
However, experiments involving different types and sizes of media would require
a considerable amount of extra time. Using only one peening machine would
mean draining all the media stored inside and cleaning of all the elements of the
equipment in contact with the media if different shot sizes were investigated.

52. 3-27
The different procedures and all the experiments carried out will be described in
more details later in the following chapter.
3.6. STATISTICAL TREATMENT OF THE RESULTS
Software designed to help carry out statistical interpretations of experiments was
used to determine the significance of the results obtained during this
investigation. Using probabilities and other statistical tools, Minitab 12 has been
of great help to understand the process, evaluating the significance of the
selected parameters from the different results obtained in all the experiments
carried out.
Once all the parameters were set up according to the factorial model chosen
different types of results to be tested were stored and the software provided tools
to calculate the different regressions, relating the sets of parameters to the
responses (See Appendices 7C, 7D, 9D and 9E). The importance and influence
of the different parameters could then be calculated and a theoretical “recipe” to
control the shot peening process was established.

53. 4-1
CHAPTER 4: PRELIMINARY EXPERIMENTS AND PILOT STUDY
4.1. INTRODUCTION
In order to set the parameters for the main experimental program, it was
necessary to carry out some preliminary experiments. Tests were done in order
to check the equipment and to calibrate some of the components.
Some work prior to this investigation will also be described leading to a pilot
study carried out to check the feasibility of the main program experiments.
4.2. PRELIMINARY EXPERIMENTS
4.2.1. The air supply
The compressed air supply was made of three different devices:
A heater coupled to a fan;
A desiccant dryer;
A compressor.
This type of set up was necessary as it is necessary to control the humidity of the
compressed air supply for the peening operation.
The fan and the heater were used to fill the dryer with hot and humid air; the
desiccant dryer, filled with silica gel, aimed at retaining most of the water present
in the air, keeping the relative humidity (RH) as low as possible before entering
the compressor.
The following experiments were carried out to test and evaluate the efficiency of
the system used.

54. 4-2
4.2.1.1. Air temperature before drying
This simple measurement was obtained as follows.
The heater and fan assembly supplied the dryer with hot air. A thermocouple
probe, located between the dryer and the compressor, was used to monitor the
variations of the temperature. Values were recorded as a function of time after
switching on.
The results are shown in figure 4-1.
Hot air supply efficiency; variations of air temperature vs time
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35
Time (minutes)
Temperature(Celsius)
Room
Temperature
Figure 4-1: Variations of temperature against time.
From these results graph, it was reasonable to assume that the air supplied to
the dryer is warm after about 15-20 minutes.

55. 4-3
In practice, it was decided to switch on the heater and fan assembly a few hours
before using the equipment, storing the hot air inside the dryer, for optimum
quality of the air.
4.2.1.2. Relative humidity of compressed air; efficiency of the dryer
In the same time that the air was warming up, the relative humidity (RH) levels
were recorded over a longer period of time by a sensor located at the
compressor outlet.
The following figure is also important as it shows how fast the dryer actually
becomes efficient in lowering the relative humidity of the air.
Dry air supply; variations of humidity level RH (Relative Humidity) vs time
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160 180
Time (minutes)
RH(%)
Figure 4-2: Relative humidity variations against time.
Clearly, the dryer became efficient after about 20 minutes, as shown by the
abrupt decrease in the relative humidity levels.

56. 4-4
It is important to note that at the time the tests were carried out, the atmospheric
conditions represented “the worst” possible case as it was raining. However, as a
conclusion from this experiment, it was decided to run the heater and fan for
about 2 hours, storing the hot air in the desiccant for drying before supplying the
compressor with it. This was achieved by running the system overnight, using a
timer fitted on the heater and blocking the hot air in the dryer by closing a valve,
located between the dryer and the input of the compressor.
4.2.2. Setting up the different parameters to the required value
The following procedure describes each stage of the set up of the equipment for
all the experiments carried out and explained in this section.
The air pressure was set and measured using the calibrated adjustable pressure
gauge fitted on the cabinet. The flow rate of the media introduced into the air
stream was set using the standard feed valve fitted at the bottom of the pressure
vessel. Finally and if necessary, the distance nozzle-specimen and the impact
angle were adjusted using the holding device allowing a 3-D displacement of the
nozzle inside the cabinet. The alignment of the nozzle with the target was
ensured by using a tube inserted inside it. Once the nozzle set and aligned, this
tube was removed.
4.2.3. Feed valve calibration
The feed valve is one of the vital components of the peening equipment. Indeed,
it controls the mass flow rate of media introduced through the air stream so the
results will directly depend on its efficiency and accuracy. Therefore, testing it in
the true experimental conditions and calibrating it were important objectives to be
achieved.

57. 4-5
4.2.3.1.Effect of air pressure on shot flow rate
Air pressure is the parameter that has the most important significance in the shot
peening process. Indeed, it is through the compressed air that velocity is
transmitted to the media, giving the necessary energy to each particle to produce
plastic deformation in the peened component surface.
Clearly, for a set feed valve aperture (mass flow rate constant), the higher the air
pressure, the higher the shot flow rate at the nozzle output.
The following test was used to establish the influence of the air pressure on the
shot flow rate. The nozzle used had a ¼” bore diameter, the feed valve was set
at a constant aperture (mark 10, ~50%) and the air pressure was varied in the
range [1 bar-5 bar]. The shot flow rates were determined by making “catches” of
shot over a fixed period of time, by collecting the shot in a container.
For each air pressure, 5 measurements of the shot flow rate were carried out in
order to check the reproducibility.
For each air pressure set, the average shot flow was determined and an
estimated experimental error calculated. BP is the blasting pressure (in bar) and
SF is the shot flow rate measured (in kg/min).
As shown in Figure 4-3, the relation between shot flow rate and air pressure is
linear over the pressure range from 1 to 5 bar.

58. 4-6
Influence of air pressure on shot flow for a set feed valve aperture (Mark 10; 50%)
SF = 1.2178*BP + 2.4978
R2
= 0.9939
3.00
4.00
5.00
6.00
7.00
8.00
9.00
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Air pressure BP (bars)
ShotflowSF(kg/min)
Figure 4-3: Shot-flow rate against air pressure.
It can be seen that by increasing and/or decreasing the air pressure, the
influence on the shot flow rate can be controlled quite accurately, within the
pressure range 1 to 5 bar.
Better repeatability and accuracy could have been achieved by replacing the
standard feed valve with a magnavalve (Not available to the author). However, if
the experimental procedures are carefully followed, errors can be limited and
consistent results obtained with the present experimental set up.
4.2.3.2. Calibration
The calibration of the feed valve was carried out for the three levels of air
pressure described in the investigation plan: 1, 2.5 and 4 bars.

59. 4-7
At each pressure, 8 different valve settings were tested twice, following this
experimental procedure:
1- Disconnect the hose from the bottom of the cabinet and block the aperture left;
2- Set the air pressure at the required value;
3- Set the feed valve aperture;
4- Press the pedal and start recording the time using a stop watch;
5- Once a sufficient amount of media is trapped at the bottom of the cabinet,
release the pedal and stop recording the time;
6- Empty the media from the bottom of the cabinet into a container and weigh;
7- Knowing the mass delivered and the time, determine the shot flow rate for the
considered valve aperture set point;
8- For a new point, follow the same procedure from 2.
Figure 4-4 shows the calibration curves obtained for each air pressure. SF1
stands for the shot flow at 2.5 bars, SF2 at 4 bars and SF3 at 1 bar.
Feed valve calibration at BP = 1, 2.5 and 4 bars
SF [4bars] = 0.7465*Vap - 0.0115
R
2
= 0.996
SF [2.5bars] = 0.5758*Vap - 0.1153
R
2
= 0.9969
SF [1bar] = 0.492*Vap - 0.1404
R
2
= 0.9897
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
0 5 10 15 20 25
Valve aperture Vap (mark n)
ShotFlowSF(kg/min)
Figure 4-4: Feed valve calibration at different air pressures.

60. 4-8
These calibration curves are quite accurate and were used at all stages of the
investigation to carry out the set up of the different experiments. Regular checks
on the set up of the feed valve were also carried out in order to detect any
deviation from these measurements.
4.2.4. The nozzle efficiency: shot peening through a defined area
Considering the fact that all the experiments were carried out in such a way that
only a small area was shot peened (~ 64 mm2
), it was important to find out how
representative this was of the full shot peening process. In this section, the
experiments to obtain the nozzle efficiency will be presented.
Figure 4-5 shows the cone of media generated by the nozzle. It is also known
from previous work [2] that the higher intensity was provided by the media
situated in the middle of this cone, approximately covering the same area as that
of the nozzle bore.
Figure 4-5: The shot stream.
Specimen
Nozzle bore area-
Maximum effect
Nozzle
Hose
Media Spray cone-
Deviation angle α rad

61. 4-9
In the experiment carried out at this stage, the aim was to determine how much
media was actually used to shot peen the experimental sites, as the site area is
smaller than the area enclosed in the cone (The site area has been calculated in
such a way that the central beam was enclosed in this area, to obtain maximum
peening).
To carry out this experiment, a collecting device was designed and
manufactured. This device consisted of two elements:
A pipe, closed at one end with a plate containing a circular hole of same
diameter as the experimental sites (∅ 9 mm) and opened at the other end;
A reinforced plastic bag connected to the other end of the pipe.
The nozzle was then set in front of this system, as if shot peening the
experimental sites. The parameters were set and two catch tests were carried
out:
Mass of media in the bag;
Mass of media at the bottom of the cabinet.
The total of both masses, related to the time, gave the actual total shot flow rate.
By calculating the ratio (Mass of media in the bag)/(Total mass), it was possible
to know the actual quantity of media used or the “efficiency” of the nozzle.

62. 4-10
The following table presents the parameters set and the results obtained:
Nozzle size
(in)
Distance
(mm)
Air pressure
(bar)
Mass flow
(kg/min)
Time
(s)
Mass bag
(g)
Mass cabinet
(g)
Total mass
(g)
Ratio (%)
0.25 100.00 4.00 5.00 16.00 474.90 1078.30 1553.20 30.58
0.25 100.00 4.00 5.00 15.00 396.90 1134.00 1530.90 25.93
0.25 100.00 4.00 5.00 15.00 396.90 1134.00 1530.90 25.93
0.25 150.00 4.00 5.00 15.00 276.50 1318.30 1594.80 17.34
0.25 150.00 4.00 5.00 15.00 262.30 1219.00 1481.30 17.71
0.25 150.00 4.00 5.00 15.00 262.30 1219.00 1481.30 17.71
0.25 150.00 4.00 1.00 60.00 255.10 1162.40 1417.50 18.00
0.25 150.00 4.00 1.00 60.00 248.10 1162.40 1410.50 17.59
0.25 150.00 4.00 1.00 60.00 255.10 1134.00 1389.10 18.36
Table 4-1: The nozzle efficiency.
From Table 4-1 it can be seen that the actual quantity of media used when
peening the sites can range from 20 to 30%, for different parameters set.
A few catch tests, in similar conditions were carried out for the second size of
nozzle. The calculated values for this test were inside the range reported above.
Once the nozzle efficiency had been established, it was important to estimate
how significant the efficiency was for the residual stresses produced. The
experiment described below was aimed to achieve this.
4.2.5. The patch experiment
This experiment was complementary to the one reported above. Indeed,
depending on the results obtained in this case, a new design of the main
experimental program might have had to be considered. It was hoped that the
experimental sites would be shown to be representative of the process and that
the existing design could be retained.

63. 4-11
4.2.5.1. Parameters
The “patch experiment” was, in principle, quite simple to do. A bar, identical in
dimensions and properties to the bars used in subsequent investigations, was
placed inside the cabinet without any masking. The nozzle was then positioned
and aligned in such a way that the centre of the bore area was in front of the
centre of the bar. Parameters were then set and peening was carried out.
As a consequence, a patch in the centre of the bar was obtained, exhibiting
different levels of coverage across the diameter of the patch.
The next step in the experiment was to carry out residual stress measurements
across the peened patch to determine how the residual stresses were varying
and where the “maximum” effect was observed. The width of the area where the
maximum values were measured was important for the experimental design.
The following peening parameters were used:
Nozzle size: ¼”;
Distance: 100 mm;
Air pressure: 4 bars;
Mass flow: 5 kg.min-1
;
Exposure time: 30 s.
A width of 27.5 mm was estimated as being wide enough and, to start with, 12
positions were set across the patch at which the residual stress measurements
were made.

64. 4-12
The Patch Experiment
-1200
-1000
-800
-600
-400
-200
0
-12.5 -10 -7.5 -5 -2.5 0 2.5 5 7.5 10 12.5
Position (mm)
ResidualStress(MPa)
rs0-s
rs90-s
Figure 4-6: The patch experiment; surface residual stress measurements.
It can be seen that stress distribution is reasonably uniform within a particular
range of positions. The positions to be most considered are then positions
between –5 mm to +5 mm. The position 0 is situated right at the centre of the
bar. This range of positions was selected as the residual stresses enclosed
within this area were expected to be uniform in both measuring directions.
The residual stress profiles were obtained using the standard procedure describe
in Chapter 3. The bar was chemically etched and residual stresses measured at
the different depth, for each selected position.
Residual stresses on an un-peened area of the specimen were also measured.

65. 4-13
4.2.5.2. Discussion and follow up experiment
Residual stress measurement were carried out, at all pre-determined positions,
as a function of depth removed by etching up to a depth of approximately 500
µm. One profile represents the variations of residual stress for one position with
the depth.
Patch Experiment-Residual stress profiles for positions from -5mm to +5mm
-1600
-1400
-1200
-1000
-800
-600
-400
-200
0
0 50 100 150 200 250 300 350 400 450 500
Depth (mum)
Residualstress(MPa)
-5mm
-2.5mm
0
2.5mm
5mm
Un-peened
Figure 4-7: The patch experiment; residual stress profiles.
It can be seen in Figure 4-7 that the stresses, within the range of positions and
for all position, are varying closely to each other. The residual compressive
stress reaches a peak of more than 1400MPa at a depth between 75 and 85µm.
The residual stress profiles are remarkably consistent at depths down to the peak
position. More variation is observed at greater depths.

66. 4-14
Considering that the positions are located 2.5mm apart, the total area where
maximum effects of the shot peening process can be observed, with a uniform
distribution, for a single pass, is approximately 10 to 12mm wide with the
particular nozzle chosen. As the experimental sites to be used subsequently are
9mm diameter, their area is enclosed in the range of positions and for the set of
parameters chosen, the results obtained for the patch are close to the results
from the pilot investigation for the same set of parameters.
As a conclusion, this experiment shows that the site design is representative
enough of the process and that useful results can be obtained. Therefore, even if
only 20 to 30% of the shot stream is used, it can be considered that what is
measured and observed are the maximum effects of the process on a
component and that the size of the experimental sites is representative of the
phenomenon.
The same experiment, with the same parameters but the exposure time set at 3
seconds instead of 30 seconds, was also carried out. Then, surface residual
stress measurements were obtained over a wider area. In this test, residual
stress was measured every 5mm across an area 80mm wide. The starting point
was called position 0 and located exactly in the centre of the bar and
measurements were taken up to 40mm on each side of this reference.
The next figure shows the surface residual stress measurements across the
designated area.

67. 4-15
Patch experiment- Surface residual stress
-1200.0
-1000.0
-800.0
-600.0
-400.0
-200.0
0.0
-40 -30 -20 -10 0 10 20 30 40
Position (mm)
RS(MPa)
RSo (MPa) RS90 (MPa)
Figure 4-8: The patch experiment; surface residual stress measurements.
As for the first patch obtained, it is possible to observe a particular area where
the residual tress is maximum and uniformly distributed (From –10mm to +10mm
in this case). This zone is commonly designated as the “hot spot” and is the area
where the most intense peening occurs. This is confirmation that the design of
the experimental sites is suitable for the investigation of the process. It can also
be observed that although the residual stresses measured are very consistent in
both directions within this zone, they diverge quite quickly. This difference is
directly related to the un-peened area of the specimen and is the consequence of
the previous manufacturing processes of the specimen (grinding). This shows a
major effect of the shot peening that is introducing and uniformly distributing
compressive stresses.

68. 4-16
4.2.6. The survey
The parameters to be tested in the factorial experiments had to be selected. To
help with this selection, opinion of experienced people involved in this field was
surveyed.
The survey was carried out as follows. Fifteen people, all with different positions
within USF Vacu-Blast (From Engineer to Peening Technician) and different
experience of the shot peening process were asked, independently, to answer
only one question (Appendix 4). This question based on a diagram describing the
peening process (Appendix 2), aimed at classifying, by importance/significance,
a non-exhaustive list of parameters. These parameters were selected due to their
obvious importance in the shot peening process. Obviously, comments and
advice were also welcomed
The starting list of parameters included:
Nozzle Size;
Air Pressure;
Distance Nozzle-Specimen;
Impact Angle;
Shot Size;
Feed Valve Adjustment (i.e. shot mass, flow rate);
Exposure Time.

69. 4-17
The following table summarises the different scores, in %, obtained for the
selected parameters, for each allocated position:
Position
Parameters 1 2 3 4 5 6 7 8 Total
Shot Velocity 8 1 1 0 2 0 2 1 15
Shot Mass 1 5 1 2 2 1 1 2 15
Impact Angle 2 0 0 2 5 4 2 0 15
Nozzle Size 1 1 0 1 3 3 2 4 15
Air Pressure 5 2 3 2 0 1 2 0 15
Distance 1 2 1 1 1 2 3 4 15
Shot Flow Rate 1 2 5 3 1 1 1 1 15
Exposure Time 2 0 2 2 3 2 2 2 15
Table 4-2: The results of a survey of opinion on the importance of different parameters.
A straight and crude conclusion is to give the following classification (1=most
important):
1- Shot Velocity.
2- Air Pressure.
3- Shot Mass.
4- Shot flow Rate.
5- Impact Angle and Exposure Time.
6- Distance and Nozzle Size.
A quick statistical investigation was then carried out to check on this assumption.

70. 4-18
Using the software Minitab 12, a score was calculated for each parameter. Table
4-3 shows the results obtained:
Position Score Shot
velocity
Shot
mass
Impact
angle
Nozzle
size
Air
pressure
Distance Shot
flow-
rate
Exposure
time
1 1 8 1 2 1 5 1 1 2
2 2 1 5 0 1 2 2 2 0
3 3 1 1 0 0 3 1 5 2
4 4 0 2 2 1 2 1 3 2
5 5 2 2 5 3 0 1 1 3
6 6 0 1 4 3 1 2 1 2
7 7 2 1 2 2 2 3 1 2
8 8 1 2 0 4 0 4 1 2
Total
Score
45 61 73 86 47 82 58 73
Mean
Score
3.00 4.07 4.87 5.73 3.13 5.47 3.87 4.87
Table 4-3: The results of a survey of opinion on the importance of different parameters. A
statistical approach.
The total score was determined by:
( )∑= NScoreTotalScore * (Eq. 4-1)
where N is the number of persons.
The mean score was calculated using the following relationship:
15
eTotalsScor
MeanScore = (Eq. 4-2)
Obviously, the lower the mean score was the more important the parameter was
thought to be.
The shot velocity came first, followed by air pressure, shot flow-rate, shot mass,
impact angle and exposure time rated as having similar importance, distance and
finally nozzle size. These results were very close to the previous interpretation
and the general pattern concerning the most potentially significant parameters
was the same.

71. 4-19
It was not surprising to see that shot velocity, air pressure, shot mass and shot
flow -rate were the most highly rated parameters; there are technical and
physical reasons, including:
The 4 parameters are closely linked together. Indeed, it is the air pressure
that produces the shot velocity;
The second reason is based on the well-known relationship describing the
kinetic energy (KE) of moving solids:
2
2
1
MVKE = (Eq. 4-3)
where M is the mass in kg and V the velocity in m.s-1
. As the principle of peening
depends on the energy acquired by the shot before hitting the surface of the
treated component, the best control of this energy is required. From this, the level
of residual stress introduced and the surface finish of the peened part will depend
on the accuracy of KE control [9][10][11].
However, some comments from the individuals polled revealed other ideas. Shot
characteristics (type, size, hardness, and roundness), air quality, temperature,
etc were also suggested as having a significant influence over the result of
peening. These additional factors should be considered in future studies. Starting
with the poll of opinions, a “new list” of potential significant parameters was
established. These were:
Shot Velocity;
Shot Mass;
Impact Angle;
Nozzle Size (bore diameter);
Air Pressure;
Distance Nozzle-Specimen;
Shot Flow Rate;
Exposure Time;
Shot Size;

72. 4-20
Shot Hardness;
Air Quality.
This list was then used for designing the experimental procedures whilst making
sensible assumptions and simplifications.
4.2.7. A post-pilot investigation
Before carrying out the actual pilot investigation, it was decided to run a quick
test, focusing on two particular parameters:
Exposure time;
Distance nozzle-specimen.
Indeed, in the peening industry, they are considered as very important as they
define the notion of coverage (The longer you peened, the more shot will hit the
component, the better the coverage; the further away the nozzle is, the wider
area covered).
4.2.7.1. Plan of the investigation
One pressure was chosen to carry out this experimental investigation. Six
different experimental sites were used as follows:
Run n° Site reference Distance (mm) Exposure time (s)
1 00-01-03 100 10
2 00-01-05 100 30
3 00-03-03 150 10
4 00-03-01 150 30
Blank 1 00-02-04 - -
Blank 2 00-04-04 - -
Table 4-4: The experimental sites with the corresponding parameters used.

73. 4-21
For each run, the feed valve setting (Vap) and the air pressure (or blasting
pressure (BP)) were kept constant, at the following values:
Vap = 50% (mark 10);
BP = 2.5 bars.
4.2.7.2. Exposure time
The exposure is one of the most important parameter to be controlled during the
peening process. Indeed, as it is the actual time spent to process a component,
residual stress introduced (also measured by Almen intensities) will depend on it.
Therefore, testing two different exposure times should be enough to show the
potential influence of this parameter on the residual stress introduced.
Often, exposure time is subsumed in the concept of coverage. In theory, 100%
coverage is an impossible task as it can only be achieved over an infinite time.
To solve the problem, the industry is relying on Almen strips to define and control
the process and the quality of the results achieved. Based on the fact that hitting
one side of a flat plate with shot over a set period of time will induce plastic
deformation at its surface (i.e. introducing compressive residual stress), the
procedure is then to measure the curvature of the strips after peening. The value
is called the intensity Almen and qualifies the process for a particular set up of
selected parameters and a known time. The standards specify that when the
intensity Almen does not increase of more than 10%, saturation is then reached.
Using this, to set up equipment and specify the process when peening
components, several strips will be peened under set parameters (usually air
pressure, distance and angle) for different periods of time. A graph is then plotted
(Intensity Almen against time) and the saturation point established.

74. 4-22
The intensity and the corresponding exposure time are then used to qualify the
process and/or control that a component is being processed within the
requirements specified by a customer.
4.2.7.3. Nozzle-specimen distance
Some of the results obtained, when setting up and testing the peening
equipment, clearly showed the influence of this parameter (Figure 4-9).
10mm
Figure 4-9: Influence of the distance nozzle-specimen. Two patches on a flat steel plate, for two
different distances.
Two patches can be seen. Each patch was obtained by setting the nozzle at a
different distance from the plate, at a 90° angle and peening for a set time. Mass
flow-rate and air pressure were kept constant at all time and only the distance
was changed from one patch to another.
The left patch corresponds to a nozzle-plate distance of 150mm; the second
patch is the results for a distance equal to 50mm. Two important differences can
be observed. The first is the diameter of the mark. The further the nozzle was
from the plate or component, the larger the diameter. The second important
observation is the depth of the indentation at its centre. In this case, it appears
that for the shortest distance the mark is deeper.

75. 4-23
This post-pilot investigation was intended to clarify in a qualitative way the
influence of two variables widely considered as significant parameters of the
peening process. It has shown that distance and time could be of influence but it
will be necessary to carry out more experiments to find out if they actually are.
4.3. THE PILOT INVESTIGATION
As part of the experimental design, the research program was carried out in two
main steps:
A pilot investigation;
A main experiment program.
The pilot investigation was an important step of this investigation as the feasibility
of the experimental procedures and a wider experimental program depended on
the results obtained.
This section aims at presenting and explaining the pilot investigationt. The results
and residual stress profiles obtained will be shown and interpretation of some of
the effects of the parameters set given. The important steps in the experiment
procedure followed will also be presented.
Following the post-pilot study (See previous section), a few improvements in the
experimental set-up were made (test of the masking technique, experiment
procedure, detection of errors).
The main objectives of the pilot investigation were:
To check the suitability and feasibility of the experimental procedures
designed;
To obtain some indication about the importance of potential interactions
(importance/significance of some of the parameters);
To obtain information about the relative variability between sites, faces and
bars.

76. 4-24
In addition, some complementary experiments were carried out and some
surface measurements on un-peened and peened sites gave useful information
before running the pilot. Indeed, in order to check that the equipment was
working and that the process was actually causing measurable changes, two
specimens were used to carry out some X-ray measurements. One specimen
was kept un-peened whereas the second was shot peened. The set up was not
important to be known, as the only objective was to observe a variation in the
residual stress within the specimens studied. X-ray diffraction was carried out on
both specimen and the residual stresses were measured in two directions: along
and across the bars. Three measurements per face were acquired: one at each
extremity of the face and the third one in the middle.
For the un-peened specimen, typical results gave tensile surface residual stress
of +200MPa to +300MPa along the face and +500MPa to +600MPa across. For
the peened specimen, the compressive surface residual stress measured was of
-900MPa approximately, in both directions.
These data were reassuring as this simple experiment showed that the peening
process was actually producing quantifiable changes and that the pilot
investigation, which was a deeper and more accurate experiments, should give
much more information about these variations of the residual stress.
4.3.1. Experimental procedure
Six parameters and their significance were investigated, aiming at relating their
conjugate effects to the residual stress introduced.
Each parameter was tested at three different levels (Low, Medium and/or High).

77. 4-25
In the following table, the list of control variables is shown, with their respective
experiment levels and assigned values:
Testing levels
Parameters Low Medium High
Exposure Time (s) 10 30 (2 Levels)
Nozzle diameter (in) 1/4 5/16 (2 Levels)
Air pressure (bars) 1 2.5 4 (3 Levels)
Distance nozzle-specimen (mm) 100 125 150 (3 Levels)
Impact angle (deg) 45 67.5 90 (3 Levels)
Mass flow adjustment (kg/min) 1 3 5 (3 Levels)
Table 4-5: The control variables and testing levels.
Each parameter listed in the table above was investigated at only two distinct
levels for the pilot experiment. The aim of this was to establish the actual
influence of each of the parameters and the importance of possible interactions.
It was also an opportunity for checking the suitability of the experimental
procedure for the main programme and identifying potential variability between
sites, faces and bars.
a b c d e f abcdef
Run Exposure
time
Nozzle
size
Air
pressure
Distance Impact
angle
Mass flow
adjustment
Global set
level
1 -1 -1 -1 -1 -1 -1 +1
2 +1 -1 -1 -1 +1 -1 +1
3 -1 +1 -1 -1 +1 +1 -1
4 +1 +1 -1 -1 -1 +1 -1
5 -1 -1 +1 -1 +1 +1 -1
6 +1 -1 +1 -1 -1 +1 -1
7 -1 +1 +1 -1 -1 -1 +1
8 +1 +1 +1 -1 +1 -1 +1
9 -1 -1 -1 +1 -1 +1 +1
10 +1 -1 -1 +1 +1 +1 +1
11 -1 +1 -1 +1 +1 -1 -1
12 +1 +1 -1 +1 -1 -1 -1
13 -1 -1 +1 +1 +1 -1 -1
14 +1 -1 +1 +1 -1 -1 -1
15 -1 +1 +1 +1 -1 +1 +1
16 +1 +1 +1 +1 +1 +1 +1
Table 4-6: The pilot investigation. The parameters and corresponding set-up for the 16 runs.

78. 4-26
The “+1” stands for high level and “-1“ stands for low level. This is an arbitrary
notation necessary when carrying out the statistical analysis of the results and it
has to be remembered that they correspond to actual values (e.g. for air
pressure, +1 is equivalent to 4 bar, 0 is equivalent to 2.5 bar and –1 is equivalent
to 1 bar).
In the previous table, “the global set level” was obtained by multiplying all the
parameters levels between them and was an arbitrary way of assigning the 16
runs planned (e.g. Run 1: ”-*-*-*-*-*-=+” or Run 12:”+*+*-*+*-*-=-“). All 8 runs
exhibiting a “+” global set level will be carried out on 1 specific bar and the 8
remaining runs exhibiting a “-“ global set level will be carried out on a second bar.
The aim in splitting these 16 runs between two distinctive bars was to make sure
that there was no significant difference between bars. 4 sites per bar (1 on each
face), randomly allocated, were kept masked to evaluate possible interactions
between sites within one face, from one face to the next one and between
different bars. Obviously, it was hoped not to have any.
4.3.2. The results
Once all sixteen runs were carried out, the residual stress was measured in each
experimental site, at different depths. Graphs of the residual stress measured
versus depth were plotted (Appendices 7A and 7B). From these profiles, different
information were used to set up and carry out the statistical analysis described in
the following section

79. 4-27
4.3.3. The statistical treatment of the results and discussion
Using the results from the pilot investigation, a statistical analysis was carried out
to determine the significance of the parameters used in shot peening on the
residual stress produced in the work piece.
Four results from the residual stress profiles were taken as responses to the shot
peening process [22][23]. These were:
The maximum compressive residual stress (RS max);
The depth where the maximum compressive residual stress occurred;
The shot peened outer layer (Layer of material where the stresses are
compressive The depth of this layer defines a boundary between
compressive and tensile residual stress);
The area under the residual stress versus depth curve, minus the same area
calculated from the blank specimen.
The following diagram (Figure 4-10) illustrates these four responses.
Error! Not a valid link.
Figure 4-10: The key values explained.

80. 4-28
The case describe in previous diagram would be ideal where the blank site would
a residual stress equal to 0MPa.
Studying the impact of the process and its variables on the residual stress
introduced and the depth where the maximum is reached seemed the most
obvious responses to be studied. Shot peening is used to introduce compressive
residual stress in the surface of the component, up to a certain depth. This is this
effect that is believed to improve fatigue life of treated components. However,
attention must be drawn that large compressive residual stresses very deep
within the component are not proved to be all beneficial.
The shot peened outer layer was also important as its length may vary as
parameters change.
Concerning the area, it was important to check on the possible variations
produced by the different set of parameters.
This area has units of N.m-1
or J.m-2
and could be associated with the work or
energy per unit area required to go from one state (un-peened) to another
(peened). As the blank was representative of the component before being
processed, evaluating this area and relating it to the different set of parameters
might give helpful information. The methods used to calculate this area are
detailed in Appendix 8.
The next table (Table 4-7) summarises the experimental levels tested for each
parameter and the results obtained.

81. 4-29
C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C24 C26
Run n a b c d e f bar RS (MPa) D (mum) SPOL (mum) A (N/m)
1 -1 -1 -1 -1 -1 -1 0 -1090 35 192.3 96747 1 1 1 1 1 0 1 1
2 1 -1 -1 -1 1 -1 0 -1245 75 229.4 148337 -1 -1 -1 1 -1 0 1 1
3 -1 1 -1 -1 1 1 1 -1165 33 197.9 102611 -1 1 1 -1 -1 -1 -1 1
4 1 1 -1 -1 -1 1 1 -1240 31 195.5 104181 1 -1 -1 -1 1 1 -1 1
5 -1 -1 1 -1 1 1 1 -1200 85 368.9 214987 1 -1 1 -1 -1 -1 1 -1
6 1 -1 1 -1 -1 1 1 -1300 80 250 198653 -1 1 -1 -1 1 1 1 -1
7 -1 1 1 -1 -1 -1 0 -1285 83 350 203055 -1 -1 1 1 1 0 -1 -1
8 1 1 1 -1 1 -1 0 -1380 85 325.2 228045 1 1 -1 1 -1 0 -1 -1
9 -1 -1 -1 1 -1 1 0 -1110 22 210 85613 1 1 -1 1 -1 0 -1 -1
10 1 -1 -1 1 1 1 0 -1120 31 178.4 105025 -1 -1 1 1 1 0 -1 -1
11 -1 1 -1 1 1 -1 1 -1200 41 242.3 158148 -1 1 -1 -1 1 -1 1 -1
12 1 1 -1 1 -1 -1 1 -1200 40 224.5 153784 1 -1 1 -1 -1 1 1 -1
13 -1 -1 1 1 1 -1 1 -1300 85 328.5 231371 1 -1 -1 -1 1 -1 -1 1
14 1 -1 1 1 -1 -1 1 -1250 80 316.5 222753 -1 1 1 -1 -1 1 -1 1
15 -1 1 1 1 -1 1 0 -1295 81 333.4 167056 -1 -1 -1 1 -1 0 1 1
16 1 1 1 1 1 1 0 -1155 83 250 186384 1 1 1 1 1 0 1 1
RS: Residual Stress C12=C1*C2
D: Depth C13=C1*C3
SPOL: Shot Peened Outer Layer C14=C1*C4
A: Area C15=C1*C5
C16=C1*C5
C17=C1*C7
C24=C2*C4
C26=C2*C6
Table 4-7: The pilot investigation; statistical analysis.
In this table, the 6 parameters tested are:
a: Exposure time (Column C1);
b: Nozzle size (Column C2);
c: Air pressure (Column C3);
d: Distance nozzle-specimen (Column C4);
e: Impact angle (Column C5);
f: Mass flow (Column C6).
The seventh parameters tested (Column C7), called “bar” was included later to
observe the variation from one bar to another, as two different bars were used to
carry out the pilot experiment.

82. 4-30
The four responses or results investigated are presented as follows:
-The maximum compressive residual stress in column C8;
-The depth where this maximum was reached column C9;
-The shot peened outer layer in column C10;
-The area in column C11.
The columns C12 to 17, C24 and C26 represent the interactions. Their choice
was based on the statistical model used to carry out the pilot investigation. Other
interactions can be studied as well.
The study was carried out in different steps:
1) Each result (Residual stress, Depth, Shot peened outer layer and Area) was
tested separately. This aimed at investigating the significance of each
parameter and their influence on each result;
2) Determination of the interactions between parameters;
3) 1) and 2) were repeated, adding the “bar” parameter to observe any effect of
variations in the virgin material.
In each case, 3 values are looked at to estimate the significance of the parameter
on the result or response considered (Listings in Appendix 7C):
1) The standard deviation S (in the units of the response tested; for the
maximum residual stress, units of MPa);
2) The column P of the listing, representing the significance of the result for each
parameter;
3) The value P in the analysis of the variance, representing the significance of
the result for the set of parameters.

83. 4-31
For each response, a regression analysis was performed and an equation was
calculated. This was of the form:
GfFeEdDcCbBaAsponse ++++++=Re Eq. 4-4
where A, B, C, D, E, F and G are constants and a, b, c, d, e and f represents the
parameters.
In the case of the Maximum Residual Stress, the regression equation was:
fedcbaMPasssidualStre 8.223.02.177.491.193.151221)(Re +++−−−−= Eq. 4-5
If a=+1, b=-1, c=-1, d=+1, e= +1 and f=-1, the residual stress corresponding is
equal to –1127.2 MPa. This means that for an exposure time of 30s, with a
nozzle of ¼” bore diameter, at an air pressure of 1bar, at a distance of 150mm,
with an impact angle of 90deg and a mass flow equals to 1kg/min, the estimated
maximum residual stress obtained in these conditions would be –1227.2MPa.
The more significant a parameter is the smaller the P value is (0<P<=1) and is
related to the significance of the parameter on the response.
The value P from the variance analysis represents the global influence of a set of
parameters for the response considered. The smaller it is, the more significance
there is.
For example, the analysis for the response Maximum Residual Stress gave a P
value of 9.4%. This means that the significance of the set of parameters was not
very high. However, it could be noticed that the most significant parameter in this
case was c (Air pressure), with an individual p-value equal to 0.011 (1.1%).
Looking at the response Depth, the P value was 0%. Therefore, the set of
parameters has a clear influence on this response.

84. 4-32
Observing each parameter, two of them seemed the most significant and most
likely to have influence on the result: c (Air pressure) and f (Mass flow).
From this, it was already possible to state:
-Air pressure and mass flow are likely to have direct influence on the process;
-If parameters have influence on the depth where maximum compressive
residual stress is reached then they will have influence on the maximum
value, even if not clearly shown. The significance on maximum residual stress
might have been more obvious if the pilot investigation involved more
experimental sites.
Observing the results obtained in the case of the Shot Peened Outer Layer, the P
value was low as well (0.2%) and two parameters seemed to have a significant
effect: c (Air pressure) and a (Exposure time).
The area aspect was also interesting. A P value equal to 0% and 3 parameters
being significant: c (Air pressure), f (Mass flow) and e (Impact angle).
From these results, it was possible to identify the following parameters as having
a significant influence on the process:
Air pressure;
Mass flow;
Impact angle;
Exposure time.
Introducing the parameter “bar” in this evaluation did not change the results
significantly for the residual stress, the depth and the shot peened layer. This
was a good result as it showed that the nominally identical bars exhibited similar
characteristics and were in fact close to being identical.

85. 4-33
The analysis of the interactions between the parameters (Appendix 7C) gave the
following results:
For the Maximum Residual Stress, the Depth, the Shot Peened Outer Layer and
the Area, the p-values were respectively equal to 10.9%, 7.8%, 18.2% and 1.8%.
Globally, the p-values were quite high, which meant that the interactions were
non-existent or of little influence on the result. However, for the area, the p-value
was quite low (1.8%), which showed the possible influence of interacting
parameters on this response; observing the p-values individually, it seemed that
exposure time and impact angle were the most likely to interact with each other,
as well as the nozzle size with the mass flow, leading to a significant influence on
the response.
Introducing the “bar” parameter did not have a significant influence on these
results.
The study of the interactions was limited to the model chosen for the pilot
investigation. However, expanding the study to the other possible interactions
might show more useful information.
From a general point of view, the pilot met most of the objectives and gave a
good overview of the possible importance of some of the parameters chosen for
this investigation. Indeed, it seemed clear that the air pressure, the mass flow
rate, the exposure time and the impact angle might be the parameters to watch
as they had had low p-values (meaning a high probability of significance). This
would make sense as between the four of them, aspects such as coverage,
velocity and energy of impact of the shot on the work-piece can be defined.

86. 4-34
The interactions between parameters were not obvious, which was what was
hoped for. However, to clarify some doubts and improve the results, a broader
investigation was necessary, involving a much larger amount of experiments.
This would help to separate significant parameters from the rest without
ambiguity, provided that enough experiments are carried out.
4.4. OTHER EXPERIMENTS AND RESULTS
4.4.1. Exposure time investigation
This section was a necessary complementary investigation as it was thought that
the exposure time used was high compared to the size of the specimen shot
peened. Indeed, two different times were set up: 10s and 30s. It was suggested
to reduce these to 1s and 3s.
Following this decision, questions were raised about the exposure time. A quick
experiment was then carried out. Using one specimen, identical to the ones used
for the pilot investigation, different sites were peened for different periods of time,
keeping the other parameters constant.
The parameters set were as follows:
Nozzle size: ¼”;
Air pressure: 4bar;
Mass flow (at the nozzle output): 1kg/min; a catch test gave 1.170kg/min;
Impact angle: 90deg;
Distance nozzle-specimen: 100mm.
The different exposure times were 1s, 3s, 10s, 20s, 30s, 40s, 50s and 60s. A
value 0s was assumed to correspond to an experimental kept virgin of shot
peening.