Thermal Field Mapping Technique for Friction Stir Process 
 
by 
 
Sakthivael Kandaswaamy 
 
 
 
 
A dissertation submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Doctor of Philosophy  
 
Auburn, Alabama 
Dec 18, 2009 
 
 
 
 
 
 
 
 
Copyright 2009 by Sakthivael Kandaswaamy 
 
 
 
Approved by 
 
Lewis N. Payton, Chair, Associate Research Professor of Mechanical Engineering 
Robert E. Thomas, Professor of Industrial and Systems Engineering 
Robert L. Jackson, Associate Professor of Mechanical Engineering 
 
 
 
 
 
 
 
 
 
 
 
Abstract 
 
 
Friction Stir Welding is a solid state ?green? welding method developed by The Welding 
Institute (UK). An internal thermal mapping instrument has been developed which allows for 
symmetrical mapping of the thermal fields developed by a Friction Stir Welding tool as it passes 
through the material being welded. This symmetrical mapping conclusively documents 
statistically the asymmetrical nature of the heat sources within the friction stir welding process. 
The various models in the literature are compared against these results. A model developed by 
the authors using classic metal cutting theory predicts the observed thermal fields. A successful 
predictive model will facilitate tool optimization and welding schedules, while optimizing the 
mechanical properties of the weld. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ii 
 
 
 
 
 
 
Acknowledgments 
 
 
 First of all, I would like to thank my mentor and my advisor Dr.Lewis Payton, for 
rendering tremendous support, not only in this project but throughout my course, academically 
and morally, without whom, completion of my course work and the project would not have been 
possible. Dr.Payton?s support also includes funding for this project. 
            I would also like to thank my committee members, Dr.Robert Thomas and Dr.Robert 
Jackson for their guidance in this project and Dr. Saeed Maghsoodloo, for his guidance and 
support with the statistical tools involved with this project.  I extend my thanks to Dr. W. A. 
Foster, for his valuable suggestions as an outside reader. 
           I am thankful to Mr. David Branscomb for helping me with the Lab view programming 
part of the project. 
          Last but not the least; I would like to show my gratitude to my parents, my sister, my 
brother-in-law and my friends for their support and love.   
 
 
 
 
 
 
 
 
 
 
 
 
 
 iii
 
 
 
 
 
Table of Contents 
 
 
List of Tables ...???..................................................................................................................vi 
List of Figures ?..........................................................................................................................viii 
Chapter 1?..?????????????????????????.... ??????...1 
    Introduction???.??.?????????????????.????...?.......1 
Chapter 2?..??????????????????????????. ??????..3 
  
            Scope and Objectives?.?????????????????????????. 3 
 
Chapter 3?..?????????.???????????????????????...4 
 
Literature review on thermal measurements in FSW.?.?..??.?.????????.4 
 
Chapter 4?..????????????????????????????????. 43  
 
            Material, Equipment and Software??.????.?????.??? ??????43 
 
Chapter 5?..??????????????????????????. ??????51 
 
            Construction and Methodology of the instrument....???..??..?............................ 51 
 
Chapter 6?..??????????????????????????. ??????62 
 
          Validation of the instrument??..?????..???????................................. 62 
 
Chapter 7??..?????????????????????????. ??????74 
        Statistical design of an experiment..??.????.???????? ?????...74 
Chapter 8???...????????????.??????????????????76 
Results of the experiment?.?????.??????????????????76 
Chapter 9????...???????????.??????????????????90 
 iv
Simulation of an asymmetric thermal model??????.???????????90 
Chapter 10???..?...????????????????????????????98 
Discussion??????.??????????.??????????????98 
Chapter 11????.....???????..????????????????????105 
Conclusion????????????????.?????????????..105 
References???????..?????????????????????????.107 
Appendices?..?????..???????????????????????..??112 
Appendix A?????.??...???????.??????????????113 
Appendix B?????.??...???????.??????????????139 
Appendix C?????.??...???????.??????????????165 
Appendix D?????.??...???????.??????????????173 
Appendix E?????.??...???????.??????????????181 
Appendix F?????.??...???????.??????????????201 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 v
 
 
 
 
 
List of Tables 
 
 
Table 1: Maximum temperatures recorded by various researchers ..............................................36 
Table 2: Experimental conditions for repeatability and sensitivity analysis ................................61 
Table 3: Thermocouple labeling assignments ...............................................................................62 
Table 4: Recorded Temperature for repeatability analysis-Shoulder ...........................................63 
Table 5: Recorded Temperature for repeatability analysis-Pin ....................................................64 
Table 6: Repeatability of the shoulder sample temperatures.........................................................66 
Table 7: Repeatability of the Pin Tool Temperatures ...................................................................66 
Table 8: Repeatability analysis for the underside of the pin tool temperatures.............................67 
Table 9: Peak average temperature readings for 1.0 inch shoulder tool........................................68 
Table 10: Peak average temperature readings for 0.7 inch shoulder tool......................................68 
Table 11: Sensitivity analysis ........................................................................................................69 
Table 12: Data to find the number of replicates for the experiments ............................................71 
Table 13: Raw experimental data ..................................................................................................76 
Table 14: Calculated data...............................................................................................................77 
Table 15: Ranking of Impact Factors for each shoulders thermal response..................................80 
Table 16: Ranking of Impact Factors for each pins thermal response...........................................81 
Table 17: Impact factors for temperature underneath the pin tool.................................................82 
Table 18: P-Values for the Retreating Versus Advancing Shoulders............................................83 
Table 19: P-Values for the Retreating Versus Advancing Pin Temperatures ...............................84 
 vi
Table 20: Power of the Retreating Versus Advancing Shoulders Observation.............................84 
Table 21: Power of the Retreating Versus Advancing Pin Temperature observations..................85 
Table 22: Raw experimental data: Shoulder experiment...............................................................86 
Table 23: Raw experimental data: Copper experiment..................................................................88 
Table 24: Material and machining variables..................................................................................93 
Table 25: Recorded temperatures to find the variable in heat equation ........................................93 
Table 26: Comparison of simulated results to actual results .........................................................96 
Table 27: Number of elements generated for default meshing conditions ....................................96 
Table 28: Simulated thermocouple temperature............................................................................98 
Table 29: Temperature at Group 2 thermocouples ........................................................................98 
Table 30: Temperature calculated using Payton?s equation ..........................................................99 
Table 31: Simulated temperature at tool-workpiece interface.......................................................99 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 vii
 
 
 
 
 
List of Figures 
 
 
Figure 1: Schematic of Tool and Material Process?......................................................................2 
Figure 2: Isometric nomeclature and geometry of FSW..................................................................5 
Figure 3: Side view of FSW geometry with nomenclature..............................................................5 
Figure 4: Cross section of typical friction stir weld.........................................................................6 
Figure 5: Murr?s diagram illustrating dynamic recrystallization.....................................................9 
Figure 6: C.M. Chen and R. Kovacevic?s thermocouple positions ...............................................12 
Figure 7: Comparison of Simulated and measured results of Chen...............................................13 
Figure 8: Yuh J.Chao, X.Qi and W.Tang?s Thermocouple Positions............................................14 
Figure 9: Comparison between Chao?s modeled and recorded temperature .................................15 
Figure 10: M Song and R Kovacevic?s thermocouple positions ...................................................17 
Figure 11: Comparison of Song?s Simulated and the Experimental results. .................................19 
Figure 12: J.C. McClure?s thermocouple positions .......................................................................20 
Figure 13: S. G. Lambrakos?s thermocouple positions ?.............................................................24 
Figure 14: Lambrakos?s comparison between simulated and recorded temperature.....................24 
Figure 15: Junde Xu?s thermocouple positions .............................................................................25 
Figure 16: Ulysse?s simulated model.............................................................................................27 
Figure 17: Comparison of Predicted and the measured values of Ulysse?s model........................27 
Figure 18: Liu?s thermocouple placements....................................................................................30 
Figure 19: Comparison of the predicted and experimental results of Li .......................................30 
 viii
Figure 20: Experimental set up of Schmidt ...................................................................................31 
Figure 21: Awang?s Predicted result..............................................................................................33 
Figure 22: Colegrove?s temperature field in Kelvin......................................................................34 
Figure 23: Side view of the steel holder with stainless steel wells................................................42 
Figure 24: End view of the steel holder.........................................................................................43 
Figure 25: Rough schematic of the instrument..............................................................................43 
Figure 26: Drawing of the holder...................................................................................................45 
Figure 27: Tool Dimensions. .........................................................................................................46 
Figure 28: NI 9205 modules installed in NI-CDAQ-9172 chassis................................................47 
Figure 29: Entire instrument installed in CINNINATI ARROW CNC Mill ?............................47 
Figure 30: Basic thermocouple setup LABVIEW 8.3 program.....................................................48 
Figure 31: LabVIEW 8.3 setup for multiple thermocouples .........................................................48 
Figure 32: Rough schematic of the instrument..............................................................................51 
Figure 33: Side and end view of the thermal profile instrument ...................................................52 
Figure 34: Placement of the load cell for measuring the downward force ....................................54 
Figure 35: Basic Data Collection Block Diagram .........................................................................55 
Figure 36: Physical placement of the thermocouples ....................................................................56 
Figure 37: One dimensional heat transfer......................................................................................58 
Figure 38: Cross sectional view of holder .....................................................................................59 
Figure 39: Statistical Hypothesis Diagram ....................................................................................70 
Figure 40: Statistical Power test result for a sample size of 7 .......................................................72 
Figure 41: Factor Level Combinations of the Principal Experiment.............................................73 
Figure 42: Example of average of seven replicates for a factor level combination.......................78 
 ix
 x
Figure 43: Example average of seven replicates calculated temperature. .....................................78 
Figure 44: ANOVA table comparing the impact factors...............................................................79 
Figure 45: Average of the calculated peak temperatures of the shoulder experiment ?..............87 
Figure 46: Shoulder and pin temperature of copper and aluminum ..............................................89 
Figure 47: Simulated advancing and retreating temperatures........................................................95 
Figure 48: Peak temperature Vs number of elements in meshing .................................................97 
Figure 49: Temperature distribution ............................................................................................100 
 
Figure 50: Bad Thermocouple in the experiment ........................................................................102 
 
Figure 51: Shoulder temperature with respect to the position of the tool....................................104 
 
 
 
 
 
 
 
I. INTRODUCTION 
 Friction Stir Welding (FSW) is a relatively new industrial process that was invented at 
The Welding Institute (TWI, United Kingdom) and patented in 1992 under research funded in 
part by the National Aeronautics and Space Administration (NASA). It is considered to be a 
?solid state? welding process since the base material never melts. Often quoted advantages of the 
process include good strength and ductility along with minimization of residual stress and 
distortion. It produces superior weld products in difficult-to-weld materials without producing 
any toxic fumes or solid waste that must be controlled as hazardous waste. Noise pollution in the 
work place is greatly reduced. The process is robust and environmentally friendly. Operators 
may be quickly trained in a few hours to produce highly repeatable, highly desirable weld 
structures. 
  Although there have been several flow models and heat source models proposed in the 
refereed literature of Friction Stir Welding, there is a shortage of real world thermal data against 
which to validate the proposed models. This is largely due to the contained nature of the actual 
Friction Stir Welding process. The heat sources of deformation are located within the base metal 
being welded and underneath the rotating tool shoulder.  As such, surface pyrometry measures 
are ineffective.   The limited amount of thermal data in the reported literature is usually 
coincidental to the primary goal of the experimental design and non-symmetrical. 
 
 1
An internal thermal mapping protocol/device has been developed which allows for symmetrical 
mapping of the thermal fields developed by the tool as it passes through the material being 
welded. This symmetrical mapping clearly documents the asymmetrical nature of the heat 
sources within the friction stir welding process. One side of the pin is always without fail hotter 
than the other, with similar results for the two sides of the shoulder.    
The system developed at Auburn University with NASA MSFC funding uses a 
shouldered pin tool as detailed in Figure 1 below. Each tool has a shoulder whose rotation 
against the substrate generates most of the heat required for welding. The pin on the tool is 
plunged into the substrate and helps stir the metal in the solid state.  The pin tool is made to 
rotate and is slowly plunged into a weld joint with constant pressure until the shoulder contacts 
the work surface. The pin-tool is then made to traverse the joint affecting a weld as it travels.  
 
Figure 1: Schematic of Tool and Material Process 
 2
 
 
 
 
 
II. SCOPE AND OBJECTIVES 
 
 This experiment was conducted to gain a better understanding of the thermal fields 
surrounding a Friction Stir Welding tool as it passes through a material being welded together by 
the passage of the tool.  A detailed symmetrical observation of these thermal fields is essential in 
order to understand the thermodynamically driven recrystallization of the material behind the 
tool after it has passed. 
      The objectives of the experiment included: 
(1) A comprehensive review of the available Friction Stir literature to include: 
a. Previous thermal measurements.  
b. Previous thermal models. 
(2) A comprehensive review metal cutting literature which applies to rotating tools. 
(3) Build an instrument capable of accomplishing the following goals: 
a. Rapid exchange and setup of samples. 
b. Use commercially available metal geometries to hold down cost. 
c. Validate statistical repeatability and statistical sensitivity. 
d. Achieve a statistical sampling power of 90% with 7 replicates. 
(4) Utilize analysis of variance to design an experiment with the equipment. 
(5) Compare resulting data to previous models. 
(6) Demonstrate the use of multi-physics software to simulate the most promising model 
from the literature, comparing it to the real world data obtained. 
 
 3
 
 
 
 
 
III. LITERATURE REVIEW OF THERMAL MEASUREMENTS IN FSW 
 
 This chapter presents a comprehensive overview of the thermal aspects of the Friction 
Stir Welding (FSW) process.  The basic process and nomenclature commonly found in the 
literature is reviewed and illustrated.  This is followed by a detailed description of the thermal 
zones and nomenclature associated with the Friction Stir Welding process and dynamic 
recrystallization.  A detailed review of thermal observations will then be presented.  This will 
include those observations which make no attempt at modeling the results and those observations 
made by researchers attempting to model the temperatures they observed during experiments.  
The survey will conclude with a brief review of the efforts of classical metal cutting to predict 
temperatures in the geometrically similar process of slot milling. 
Basic Review of the Friction Stir Welding Process 
 In order to visualize the fundamentals of the process, consider Figure 2.  The two 
materials to be welded are placed in contact via either an overlapping or, in this case, a butt joint 
fashion.  A broad tool with a narrower pin on the end is fabricated.  The tool is then inserted 
while rotating at a high speed into the material until the wider ?shoulder? of the tool makes 
contact with the material being welded.  At this point, the tool begins a traverse of the weld 
seam, deforming the material in its passage, leaving behind a formed weld.  The material does 
not melt during this solid-state deformation.  Personnel with shop experience will recognize 
similarities with a type of end-milling referred to as slot-milling.  For the travel direction 
depicted in Figure 2, the left hand side of the weld is referred to as the advancing side of the tool 
and the right hand side is referred to as the retreating side of the tool.  Classic milling theory 
 4
refers to the advancing side as the side undergoing ?up-milling? and the retreating side as 
undergoing ?down-milling?.  Figure 3 details a flatter 2D aspect of the process which illustrates 
several important features.  
 
 
Figure 2: Isometric nomeclature and geometry of FSW 
 
 
Figure 3: Side view of FSW geometry with nomenclature 
The base material illustrated in Figure 3 will always be contained  from above by the shoulder of 
the passing tool and from below by the relatively immobile backing plate which is commonly 
 5
referred to as the ?anvil?.  The leading edge is commonly referred to as the advancing edge and 
the back side is referred to as the trailing edge.  Pin tools may be smooth as illustrated in Figure 
3, threaded or geometrically shaped in any number of ways.  It is also common to ?incline? the 
axis of the tool so that there is a height difference between the advancing versus retreating edge 
of the tool.  The tool in Figure 3 would thus be said to have an inclination angle of 0 degrees. 
Dynamic Recrystallization and Temperature 
             FSW could be considered as a hot working process instead of a welding process.  As the 
tool passes through the material (Figure 4), material flows around the tool without melting.  This 
leaves a distinct, asymmetrical nugget within the material as shown below. 
?
?
?
?
?
?
 
Figure 4: Cross section of typical friction stir weld. Top photo shows weld cross section 
without markup. Lower photo is same photo marked up to show temperature zones.  
This process involves a large amount of deformation at high strain rates.  Such deformation leads 
to a change in the grain structure of the material in the deformed region.  
 6
 The weld usually consists of four distinct regions.  They are the base metal, the refined 
weld nugget, the Thermo Mechanically Affected Zone (TMAZ) and the Heat Affected Zone 
(HAZ).  Figure 4 depicts all of these zones.  The base metal zone is left unmarked since it 
surrounds all the other zones.  The base metal zone does not experience either the heat or the 
deformation.   Its microstructure and mechanical properties are never affected by these.   The 
HAZ comes next to the base metal zone. This region is affected by the heat generated but not by 
the plastic deformation.  The TMAZ zone experiences both plastic deformation and heat 
generated during welding. Due to this the microstructure of the region is altered. It usually has 
elongated and recovered grain structures [1] with different mechanical properties from the parent 
base metal(s) being welded together.   The weld nugget refers to the core of the weld zone where 
recrystallization occurs due to severe thermo-mechanical effect and is comprised of extremely 
fine grains [2].    This leads to a general strengthening of mechanical properties in most welds. 
             According to Jata [3] the weld nugget, whose extent is comparable to the size of the pin, 
comprises of equiaxed, fine and dynamically recrystallized grains.  When the size of these grains 
was compared to the size of the grains in the parent metal, it was found that the recrystallized 
nugget grains are smaller than that of the parent metal. 
            Jata compared his micro structural analysis study of FSW with the micro-structural 
analysis study of hot forging.  He concluded that the dynamic recrystallization in FSW had the 
same dependence as that of the dynamic recrystallization of hot forging process. 
         Jata?s study was based on research by Murr [4].   Murr also stated that the grain size in the 
weld nugget was smaller than that of the parent metal.  He noted that grain size was smaller near 
the weld bottom.  Murr detailed a significant difference in the grain size of welded region and the 
parent metal in both aluminum alloy (AA) 6061 and aluminum alloy (AA) 1100.  He suggested 
 7
that the dynamic recrystallization is the principal feature of friction stir weld development.  He 
gave out a relation for the grain growth occurring, in the form, 
   
22
0
D Dkt? =                                                       (1) 
Where - Initial recrystallized grain size 
0
D
            D - Residual FSW grain size 
And is given by an equation in the form of, k
   
exp( )
Q
kA
RT
=?
                                                    
(2) 
Where A -constant, 
            Q- Appropriate energy for the grain growth 
            R - Gas constant 
            T - Absolute temperature characterizing the grain growth over time, t . 
               Murr [5], has presented a diagrammatic system to better explain the approach of 
dynamic recrystallization in FSW as shown in the below figure.  The process starts with the 
clockwise rotation of the head pin stirring metal A into B.  
 8
 
Figure 5: Murr?s diagram illustrating dynamic recrystallization and grain growth 
associated with the FSW process. 
Murr?s approach to dynamic recrystallization has 5 steps.                      
Step 1 shows the original microstructure of the base metal in which dislocations are 
denoted by symbol ???. 
Step 2 shows the deformed grain as a result of increase in dislocation density. 
Step 3 is the result of more dislocations and more distorted making the grains saturated 
 9
Step 4 shows the nucleation of new stress free recrystallized grains which occurred due 
to the stored energy. 
Steps through 1-4 would be repeated continuously for further plastic deformation of 
recrystallized grains. 
Step 5 shows the recrystallized grains which are subjected to certain amount of grain 
growth.  This is a result of the frictional heat generation during the FSW process.    The 
amount of heat determines the amount of grain growth. 
 James [6] hypothesizes a relation between the defect formation in FSW and the grains 
structures obtained.  He classified the known defects of FSW to include a lack of penetration, 
voids and root defects, also called as kissing bonds.  James hypothesized a   region in the 
advancing side where chaotic flow occurs.  There is a location above and below this region 
where the flow tends to be in the opposite direction, thus creating a vortex which leads to the 
development of voids.  The voids developed might be a source for the crack formation.  He also 
states that larger planar facets are the possibility for larger cracks and higher crack growth rate.     
 One thing is common in all discussions:  effects of temperature upon grain recovery and 
defect generation/presence.    It is therefore important to be able to predict the temperatures 
around the rotating tool.    A review of temperature studies is indicated.  There are two main 
types of thermal observation in the literature.  Type one are those temperature observations 
conducted randomly in conjunction with an experiment whose objective lies elsewhere.   The 
type two temperature observations were typically conducted by modelers attempting to predict 
temperatures so that they can simulate weld properties.   There is a small subset of this group 
whose goal is to design better tools based upon classic metal cutting theory.    This chapter will 
 10
review the temperature observations and the temperature models, followed with a review of 
classic milling theory as it might apply to FSW.   
Temperature Measurement, Attempts to Model Temperature Fields: 
             Various researchers have documented thermal results as part of their overall report. Most 
of the experiments were carried out by recording temperatures only on one side of the weld.   
Few experiments captured the temperature on both sides of the weld. All the researchers who had 
recorded temperatures on both sides of the weld had an asymmetric placement of thermocouples 
except Xu, who had thermocouples placed on both sides of the weld symmetrically.  
Interestingly, all the researchers has treated FSW as a process with symmetrical temperature field 
in their models. This chapter covers the experiment done by various researchers and their models 
in the following order. 
? Experiments capturing temperature on one side of the weld. 
? Experiments capturing temperature on both sides of the weld asymmetrically. 
? Experiments capturing temperature on both sides of the weld symmetrically. 
Experiments capturing temperature on one side of the weld: 
             W.Tang, X.Guo, J.C.McClure, L.E.Murr, and A.Nunes [7] had a series of thermocouples 
placed in the holes made on the under-side of their 6061-T6 aluminum samples. The 
thermocouples were placed at one quarter, one half and three quarters of the plate thickness 
respectively and were at different distances from the weld seam, mainly between 5 mm to 15 mm 
from the weld seam. They recorded a maximum of 420 Celsius from a thermocouple kept at 5 
mm from the weld seam. The diameter of the shoulder was twice as much as the pin. The 
rotational speed of the tool ranged from 400 to 1200 RPM, with a traverse speed of 2 mm per 
second. 
 11
            Chen and Kovacevic [8] in 2003 made an attempt to study the thermal history and 
thermo-mechanical process of FSW using aluminum alloy 6061 T6. They measured the 
temperature with eight thermal couples embedded in the plate on the retreating side.  
 
Figure 6: C.M. Chen and R. Kovacevic?s thermocouple positions. Dimensions in mm. 
Four were located 1.6 mm below the top surface and four were located 1.6 mm from the bottom 
surface, with a distance of 10 mm, 14 mm, 22 mm and 30 mm respectively to the weld centre-
line.  The tool was made of AISI A2 steel with 24 mm shoulder diameter and 6 mm pin diameter 
and the weld was traversed at 140 mm per minute at 500 RPM. ANSYS was the finite element 
tool used for the simulation of the process.   The source of heat was assumed to be the friction 
between the work piece and the tool, considering both the shoulder and the pin.  
 Heat source due to friction between tool and the work piece was given by, 
                                             
33
2
() ()( )
3
qtptR???= r?                                            (3) 
Where, ?  -coefficient of friction. 
           ?- Angular velocity  
        - Radius of the shoulder and pin   ,Rr
The thermal and mechanical solutions were coupled in the model to increase the accuracy of the 
temperature gradient obtained.  
 12
 
Figure 7: Comparison of Simulated and measured results of Chen. 
The modeling solution was compared with the actual thermocouple measurements, and the peak 
temperature obtained from simulation was around 450C.   According to the result obtained, the 
maximum temperature was obtained just beyond the shoulder edge in both the longitudinal and 
lateral directions. 
           Chao [9], in 2003, formulated the heat transfer problem into a steady state boundary value 
problem for the tool, and a transient boundary value problem for the work piece. According to 
Chao, the heat generated due to the friction between the tool and the work piece, is a result of the 
sum of the heat lost to the backing plate, heat lost from the surface of the work piece to the 
environment and the heat content in the work piece, neglecting the radiation.  Chao claimed that 
the model, which does not consider the heat generated due to the plastic deformation of the 
material, is a reasonable approach for the study.  
 13
 
Figure 8: Yuh J.Chao, X.Qi and W.Tang?s Thermocouple Positions. 
They conducted experiments with 3 sets of thermocouples placed at 3 different positions.  
Position 1 was below the shoulder.  Position 2 was below the pin and the Position 3 was near the 
base point. First set was placed 5 mm from the center of the tool and the second 12.7 and the 
third set 25.4 mm from the center of the tool. They measured 440 Celsius for the shoulder and 
425 Celsius for the pin as their peak temperature.  The welding simulation code WELDSIM was 
used for the study. Only half of the work piece was simulated as Chao state that the other half is 
symmetrical to the simulated half. Since the model does not consider the heat generated due to 
the plastic deformation, larger heat conduction coefficient was used to compensate this. 
The heat equation used was given as, 
3
3
0
3
2
Qr
q
r?
=  For 
0
rr?                                                    (4) 
 Where - Outside radius of the shoulder. 
0
r
 14
 
Figure 9: Comparison between Chao?s modeled and recorded temperature 
The peak temperature of 420 Celsius was predicted at a region underneath the shoulder at about 
5 mm from the weld line. 
          Khandkar [10], in 2003, proposed a three dimensional thermal model to study the thermal 
mapping during the process.  Temperature measurements were carried out in aluminum alloy 
6061. A tool with 25 mm shoulder diameter, 10 mm pin diameter and 8 mm pin length was used.  
The experiment was carried out at 390 RPM with a traverse speed of 142 mm/min.  The 
thermocouples were placed at different distances from the weld line.  This model considered the 
backing plate and the heat lost to it.   The heat input in this experiment, was given, assuming a 
uniform shear stress at the work piece-tool interface.  Also the heat input was correlated with the 
torque that was measured experimentally. 
 
 Heat input equation at the interface was given as, 
int erface tot
Generated
qq
Total
?
?
=?                                               (5) 
 15
Where ? - Torque 
            - Heat input at an interface 
int erface
q
            
tot
q - Total heat input. 
The torque at the shoulder interface was found by, 
0
1
()(2 )
r
shoulder
r
M rrd??=
?
r                                                 (6) 
The torque at the pin bottom was given by, 
1
0
()(2 )
r
pinbottom
M rrd??=
?
r
                                                (7) 
The torque at the vertical pin surface was given by, 
11
()2
pinsurface
M rrh? ?=                                                    (8) 
Where, - Radial distance from the tool center r
            - Radius of the shoulder 
0
r
            - Radius of the pin 
1
r
            ? - Assumed shear stress             
A peak temperature of around 460 Celsius was obtained as a result of the simulation at the 
interface of the tool and the work piece. The limitation of this experiment was the uncertainty of 
thermal contact conductance. Khandkar also claims that the model developed was unique in the 
way the friction coefficient was back calculated.             
            A moving tool approach was adopted by Song [11] in 2003.   For the experimental setup, 
they had two sets of thermocouples in their FSW experiment with AL 6061 T6 plate.   Each set 
had 4 thermocouples and were placed at 8, 12, 16 and 25 mm from the weld line.  The first set 
was 2 mm from the top surface of the plate and the second set at 8 mm from the top surface and 
 16
recorded a peak of around 580 Celsius (850 kelvin).  The tool had a 50 mm shoulder diameter 
and 12 mm pin diameter. They used 344 RPM, 637 RPM and 914 RPM at 1.59 and 3.18 mm/sec. 
 
ALL DIMENSIONS ARE IN MM 
Figure 10:  M Song and R Kovacevic?s thermocouple positions. 
He developed a thermal model of the process in a moving coordinate. This reduced the difficulty 
of modeling the moving tool.  The model had two sources of heat. One is the heat from the 
shoulder and the other from the pin.  
The heat generated by the shoulder was given as, 
2
ni
q FRn??=
                                                             
(9) 
Where ? - Friction coefficient 
            - Downward force 
n
F
            - Distance from the calculated point to the axis of the tool 
i
R
            n- Rotational speed of the tool. 
 17
The heat input from the pin was the sum of the heat generated by the shearing action of the pin, 
heat generated by the friction of the bottom surface of the pin and the heat generated by the 
friction of the vertical surface of the pin.  
The heat generated due to pin was given as, 
2
24
2
3
3(1 )
prp pm
m
p
kY r hV F V
V
qrhkY
cos? ??
?
?
?
=+ +
+
?
                                  (10) 
Where, 
01
90 tan ( )? ??
?
=??  
            
0
sin
sin(180 )
mp
Vv
?
??
=
??
 
             
0
sin
sin(180 )
rp p
Vv
?
??
=
??
 
             
pp
vr?=  
             ? -Helix angle of the thread 
             ? -Friction coefficient 
              Y - Average shear stress 
Song claims that the simulated temperature is very accurate and the model developed could be 
easily applied to model the coupled heat transfer process in both the tool and the work piece. The 
simulated temperature results were compared with the real experimental data. Song also states 
that the maximum temperature obtained beneath the tool shoulder would be close to the melting 
temperature of the work piece.          
 
 18
 
Figure 11: Comparison of Song?s Simulated and the Experimental results. 
Experiments capturing temperature on both sides of the weld asymmetrically: 
           McClure [12], in 1998, developed a friction stir welding model with a moving point heat 
source .  He used AL 6061 for their FSW experiment and captured the temperature field on both 
sides of the weld.   They had thermocouples inserted through holes drilled at the bottom of the 
piece.   The holes were randomly placed and were not symmetrical to the weld line.   The pin 
tool used was High Carbon steel with 19 mm (0.75 inch) shoulder diameter and 5.46 mm (0.215 
inch) pin length. The weld was done at 400 RPM at 2 mm/sec. 
 19
 
 
ALL DIMENSIONS IN MM 
Figure 12: J.C. McClure?s thermocouple positions 
To model the process, McClure considered the fact that the shoulder was the major heat source 
in the process. This was concluded from the macro view of the experimental weld which clearly 
indicated that the heat affected zone (HAZ) was equal to the diameter of the shoulder and the 
observation of the temperature gradient through the work piece thickness which was indicated by 
the boundary between the work piece base metal and the HAZ.  
The heat generation equation used was, 
                                      
2
2qPdd???? ??=                                                   (11) 
Where ,? - Friction coeffiecient 
            ?- Spindle rotation speed 
 20
            P - Downward pressure. 
The downward pressure was given as, 
22
21
()
F
P
RR?
=
?
                                                        (12) 
Where, - Radii of the shoulder and the pin. 
21
,RR
This equation shows that downward pressure changes with varying radius, which in turn, varies 
the heat generated according to the diameter of the tool.  The maximum temperature obtained in 
the model was around 420
 
Celsius and McClure claims that the results obtained were similar to 
that of the experimental results. 
           Unlike McClure, Dong considered the heat generated by pin due to the plastic 
deformation of the material. Dong, Lu and Hong [13] in 2001, made an attempt to study the 
coupled friction heat, plastic flow slip zone development and the three dimensional heat flow in 
Friction stir welding using a numerical model.  According to Coulomb friction law which was 
assumed to apply to the heat generation, shear work was completely transformed into a point 
wise heat generation along the interface of the tool and the work piece. The coupled friction heat 
generation equation was given as, 
2
2()(,)q rTp Ts?? ?=
                                           
    (13) 
Where, ?- Angular velocity 
    
-Pressure distribution along the interface s (,)pT s
            
r - Radius of the tool 
           
? -Friction coefficient. 
The equation for heat generated due to plastic deformation at a given time and material point was 
given as, 
 21
 
                                                       (14) 
ij ij
qsi?=
Where, ? - Measurement of the fraction of the plastic work converted to heat flux. 
           
ij
s -deviatoric stress 
           
- Incremental plastic strain rate. 
ij
i
Conditions related to finite deformation thermoplasticity were considered for this project. With 
the results obtained, Dong found that the coupled thermomechanical friction heating was 
responsible for the heat in the upper region, and the plastic work induced heat was responsible 
for the heat in the lower region of the weld. They also suggested that the pin geometry and the 
pin height have more influence on the heat generated, and a high temperature zone would be 
created below the bottom of the pin having a temperature of about 300 Celsius. 
            Though Dong was able to include the plastic deformation of material in his model, he did 
not model a moving tool. Khandkar [14], in 2001, used a finite element analysis to develop a 3D 
thermal model of the process. The model captures the heat generated due to the moving tool and 
also the convective and diffusive heat transfer caused due to the plastic deformation of the 
material. In the model, the heat generated in the process was due to the friction between the 
shoulder and the work piece, and also the friction between the pin and the work piece. 
 Heat generated by the shoulder was given by, 
2
2
2
shoulder pin
FNr
q
rr
?
=
?
                                                      (15) 
Where, ? - Friction coefficient 
            F - Downward force 
           -Tool rotational speed N
            r - Distance from the center of the tool 
 22
Heat generated due to the pin was given as, 
                                                         
2
2
2
shoulder pin
FNr
qk
rr
?
=
?
                                                    (16) 
Where k - constant and was assumed to be 3%. 
Since the heat was assumed to be generated only by the friction between the tool and the work 
piece, Khandkar has used the same equation, for both the shoulder and the pin, except for an 
addition of a constant for the pin equation. The maximum temperature obtained in the model was 
around 500 Celsius. 
               T.J. Lienert, W.L. Stellwag.Jr and L.R.Lehman[15], measured the temperature of AA 
6082-T6 with two thermocouples on top of the piece, one on the advancing and the other on the 
retreating side of the weld at 6.35 mm (1/8 inch) from the edge of the weld.  The tool used was 
made of H13 tool steel with 19 mm (3/4 inch) shoulder diameter.  They measured a peak of 360 
degree Celsius in both the thermocouples at eleven inches per minute at 1250 RPM. This system 
failed to detect the temperature difference between the advancing and retreating side. The reason 
could be because the thermocouples were placed above the workpiece and away from the weld.    
   S. G. Lambrakos, R. W. Fonda, J. O. Milewski [16], and J. E. Mitchell also had 
thermocouples placed asymmetrically on both sides of the weld. They had 5 thermocouples 
placed in the holes drilled from the bottom side of 6.35 mm thick AL 6061-T6. Actual placement 
of the thermocouples was assumed to be at a depth of 3.8 mm from the top surface of the plate.  
They had one thermocouple at the weld center and all others around the tool on both advancing 
and retreating side of the tool, but the placement of the thermocouples were not symmetric to the 
weld line.  The tool used was made of H13 tool steel with 25.4 mm shoulder diameter and 5.72 
long pin with 6.35 mm diameter.   
 23
 
Figure 13: S. G. Lambrakos?s thermocouple positions. 
 The tool had a 6 degree lead angle.  The weld was done at 2.17 mm per second at 400 RPM.   
The model was constructed considering only the HAZ.  An arbitrary value was assigned to each 
element of the cylindrical surface corresponding to the pin of the tool where the heat is 
generated.  This arbitrary value was scaled relative to the surface elements corresponding to the 
base of the tool shoulder and the base of the pin according to their radial distance to that of the 
arbitrarily scaled cylindrical surfaces. 
 
Figure 14: S. G. Lambrakos?s comparison between simulated and recorded temperature. 
 24
Experiments capturing temperature on both sides of the weld symmetrically: 
             Interestingly, Xu had thermocouples symmetrical to the weld line. Junde Xu, Suhas 
P.Vaze, Robert J.Ritter, Kevin J.Colligan, Joseph R.Pickens[17], recorded temperatures in 
aluminum magnesium alloy 5083-H116 plate. They used an infrared camera and thermocouples 
to measure the temperature at different points of the workpiece.  
 
Figure 15: Junde Xu?s thermocouple positions. 
 They had thermocouples placed randomly in the work piece, but had two pairs of thermocouples 
on the advancing and retreating sides of the weld at equal distances from the weld center. Xu did 
not document any difference in temperature between the advancing and the retreating side 
 25
thermocouples even though the thermocouples were placed in such a way that they fall directly 
under the shoulders used.  The tool material was MP159, with 9 mm (0.35 inches) pin diameter 
and 25 mm (0.996 inches) pin length. The tool rotation speed was 250 RPM at traversing speed 
varying between 100 to 177 mm per minute (4 to 7 inches). The maximum temperature recorded 
was around 537 Celsius (1000 F).    Xu?s model was derived from Chao?s model. 
            Temperature recording was also done when two dissimilar metals were welded together. 
W H Jiang and R Kovacevic utilized FSW to join two dissimilar metals, 6061-T6 aluminum (Al) 
alloy and AISI 1018 steel. The tool used was H13 tool steel with 25 mm shoulder diameter and 
5.5 mm pin diameter.  The traverse speed was 140 mm per minute at 914 RPM.  Temperature 
was measured during the process near the nugget using K-type thermocouples.  The 
thermocouples were placed at the center thickness of the sheets about 5 mm and 2 mm from the 
original interface between the two sheets in the Al alloy side and the steel side respectively and 
measured around 490 Celsius for aluminum and 620 Celsius for steel.            
            Ulysse[18] in 2002, used a different approach for his model. He solved conductive - 
convective steady state equation to get the temperature distribution equation for the model. 
Ulysse used a three-dimensional visco-plastic modeling technique to model the Friction stir 
process.  
The equation used was given as, 
()
p
cu k Q? ??? =? ? +                                               (17) 
Where ? -density 
           -Specific heat 
p
c
            k -Conductivity 
            u -Velocity vector 
 26
. 
Figure 16: Temperature distribution of Ulysse?s simulated model 
The peak temperature obtained was in the range of 400 Celsius below the shoulder.  The 
obtained results were compared with the measured values. 
 
Figure 17: Comparison of Predicted and the measured values of Ulysse?s model. 
The behavior of the metal for a wide range of strain-rate is discussed to be the reason for the 
predicted values being greater that the measured values.  
          Chang [19] in 2003 developed a numerical heat flow model for the process and used a 
thermo-elastic-plastic finite analysis tool to simulate the process and examine the thermal 
 27
behavior. The model had an assumption that the heat was generated in two ways. One was due to 
the friction between the tool and the work piece. The other was due to the plastic deformation at 
stir zone in the rotating pin zone.  
The heat equation was assumed to be, 
23
4
3
qP??= NR
                                                  
   (18) 
Where, ?  - Friction coefficient  
            P - Pressure distribution 
           - Rotation speed N
          R - Surface radius.   
The analysis of heat distribution characteristics of the process was done using the two 
dimensional unstationary heat conduction theories. Temperature distribution in different 
segments of the weld were obtained as the result of the simulation. The stir zone had a maximum 
temperature of 530 Celsius, the thermo-mechanically affected zone had a temperature ranging 
from 480-500 Celsius and the heat affected zone had a temperature ranging from 450-480 
Celsius. For the experiment, Chang used aluminum alloy 6061-T6, 300 mm in length, 300 mm in 
width and 4 mm in thickness. The experiments were carried out with different speeds varying 
from 87 mm/min to 567 mm/min at 1600 RPM.                
                 Lawrjaniec [20], in 2003, used two different numerical heat sources for the modeling 
of the process. They were a 2D superficial source developed with SYSWELD and a 3D axi-
symmetrical source developed with MARC, where, MARC and SYSWELD are finite element 
codes. According to him, there are three heat generation regions during the process and it 
includes the shoulder contact region, pin contact region and the region where material shears. 
The equation for heat generated was given by, 
 28
23
1
4
()
3
n
ii
qPR???
?
=?
?
3
R
                                            
(19) 
Where, ? - friction coefficient 
             ?- Rotational speed 
             P - Vertical pressure 
             - Shoulder radius 
i
R
 
He states that the shoulder radius was responsible for the heat generated and the maximum 
temperature was obtained at the periphery. He also claims, there is a need to calibrate the 2D and 
the 3D source is to maximize the accuracy of the results obtained, since some of the parameters 
in the equation including the friction coefficient and the tool vertical pressure were difficult to 
obtain. The results obtained were checked against the recorded temperature results using 
thermocouples and infrared cameras. 
The results obtained by the 2D superficial source were not symmetrical to the weld line as the 
retreating side was hotter than the advancing side with a maximum temperature of about 360 
Celsius whereas the results obtained by the 3D source gave a symmetrical temperature field. 
               Liu[21] in 2004 attempted to study thermal and heat transfer phenomenon in Friction 
stir welding. Aluminum alloy 6061-T6 was used for this experiment. Heat generation during the 
process was considered to be from the friction between the tool and the workpiece and from the 
plastic deformation of the metal due to the stirring action of the pin. According to Liu, there were 
several issues that were not addressed by previous researchers. Those issues include the relative 
contribution of the heat generated at the interface of tool and the workpiece, the heat contributed 
by the plastic deformation of the material due to the stirring action, the heat conducted by the 
backing plate. WELDSIM, the welding simulation code was used to model the heat transfer 
 29
phenomenon. Liu?s model was based on Chao?s study [22],[9],[23] that 5% of the total energy 
generated during the process was conducted by the machine tool. Liu also states that the pin 
accounts for considerable amount of heat generated. According to him, a full length pin of about 
24.3 mm accounts for 28% of heat generated and a short length pin of about 6.7 mm accounts for 
6.7% of total heat generated in the workpiece.   
 
Figure 18: Liu?s thermocouple placements. 
The experiments were carried out in aluminum 6061-T6 at 350 RPM with a traverse speed of 2.5 
mm/sec. 31 thermocouples were placed symmetrically on both sides of the weld line. 
 
Figure 19: Comparison of the predicted and experimental results of Liu. 
He concluded saying that the maximum temperature, close to the melting temperature, occurs at 
the top surface of the workpiece near the weld line. The maximum temperatures obtained by 
 30
simulation and by the experiment were around 500 Celsius and 580 Celsius respectively at the 
top surface of the work piece. 
             Schmidt [24] in 2004 developed an analytical model for heat generation in friction stir 
welding. The experiment was carried out in 2024 T3 alloy at 400 RPM and 120 mm/min traverse 
rate.  
 
Figure 20: Experimental set up of Schmidt. 
The model had three sources of heat. They were the heat generated from the shoulder, heat 
generated by the plastic deformation of the material due to the stirring action of the pin and heat 
generated from the bottom surface of the pin. The special thing about this model is that the heat 
generation equation was developed to satisfy not only flat cylindrical shoulders but also concave 
or conical shoulders. Some process use conical shoulders for the escape of the material during 
the submerging phase and enhancing the extrusion process during the operation phase. 
 
The heat generated by the shoulder was given as, 
 31
2
2
0
(1 tan )
Rshoulder
contact
Rprobe
qr
?
dr???=+
??
?
dz
                                     
(20) 
 
The heat generated by a cylindrical probe was given as, 
2
2
00
Hprobe
contact probe
qR
?
??=
??
?
rd
                                            (21) 
 
The heat generated by the probe tip flat surface was given as, 
2
2
00
Rprobe
contact
q
?
??=
??
?                                              (22) 
Equations (20), (21) and (22) were added together for the total heat generated. The result 
obtained was similar to that of Khandkar[14],[10]. 
             M. Awang and V. H. Mucino [25] have done a finite element modeling of friction stir 
spot welding (FSSW) process by an explicit finite element code in 2005. The model of FSSW 
process has been done using Abacus software. They have simulated the material flow and 
temperature distribution of the process. However they claim that the lack of refinements of 
several modeling aspects provided them with less realistic results. According to this model, heat 
is generated by two ways. One is due to the friction between the rotating tool and the work piece. 
The other is due to the plastic deformation of the material.  
Frictional heat input was given by,         
2
n
q RNF??=                                                         (23) 
Where, ? ? Friction coefficient 
            
R   -  The distance of the calculated point from the tool axis  
            
  -  Rotational speed of the tool.  N
 32
           
  - Normal force applied to the work piece. 
n
F
Heat due to plastic deformation was given by,                             
pl
q ???=
                                                       (24)
 
Where, ?  - Fraction heat dissipated due to plastic straining  
            
? -  Shear stress 
          
pl
?  - The plastic straining rate. 
 
Figure 21: Awang?s Predicted result. 
Approximated friction coefficients used would be one of the reasons for the unrealistic 
temperature curves obtained from the model.            
             There were researchers who also included the heat lost to the atmosphere and the 
neighboring elements, during the process in their models. Xu [17] , in 2003, developed a model 
to find the heat generated due to the interaction of the tool with the work piece considering the 
loss of heat to the tool and the anvil. Xu claims that the result obtained accurately matches with 
the thermocouple and the infrared camera readings.  In this project, Xu had a mica frame inserted 
between the work piece and the backing plate to reduce the heat dissipated to the backing plate 
 33
while thermocouples were used to measure the temperature. The measured temperature was used 
to combine with the finite element model to find quantitatively the heat lost to anvil, heat lost to 
the tool, downward pressure, heat lost to the air and the heat generated due to the interaction of 
the tool and the work piece.  The maximum temperature obtained in the model was around 970 
F. He concluded, stating that,  both the shoulder and the pin contributes significantly for the heat 
generated in the workpiece and it might vary from 30% to 70% of the total heat generated for 
both shoulder and the pin. His statement also includes that the heat generated by the shoulder 
increases with increasing diameter, and the heat generated by the pin increases with the 
increasing traverse speed there by increasing the peak temperature in the nugget.           
             Colegrove and Shercliff [26] , used FLUENT to model FSW, in 2003. The experiment 
was done in aluminum alloy 7071-T7. The heat was assumed to be generated in the tool due to 
the friction of the tool and the material surface, and viscous dissipation within the deformed 
material .Some part of heat generated was assumed to be convected from the top surface of the 
workpiece, and some portion was assumed to be conducted by the backing plate.  
 
Figure 22: Colegrove?s temperature field in Kelvin. 
 34
 Heat generated due to the stirring of material was ignored. This condition was similar to the 
condition used by Khandkar and Khan [10]. They suggest that the peak temperature of the weld 
depends on the tool material that is been used, since different tool material would exert different 
surface shear stress, and they also suggest that the rotation speed of the tool had very little effect 
on the peak temperature obtained. The peak temperature obtained for aluminum 7075-T7 was 
very near to the solidus when Tool steel was assumed to be the pin tool.             
 The documented thermal measurements and temperature models in FSW research and 
projects were carried out to find the temperature of the tool and the work piece by placing the 
thermocouples in both of them at different positions with different parameters.  No research was 
done to find the temperature of the advancing shoulder and the retreating shoulder at the same 
moment. This is true with the pin also. There is no documented research of the simultaneous 
advancing and retreating side of the pin.   Table 1 on the following page summarizes thermal 
studies conducted in FSW to date. 
 
 
 
 
 
 
 
 
 
 
 35
 
Researcher?
Material? RPM?
Radius:???
Shoulder
Radius:
Pin?
Traverse?
speed?
Length:
Pin?
MaxTemp
:?Shoulder?
Chao? AA?2195? 240? 25.4?mm 10?mm?
2.36?
mm/sec?
?? 440?C?
Chen? AL?6061? 500? 24?mm? 6?mm?
140?
mm/min?
?? 480?C?
Song? AL?6061?
344,?
637,?
914?
50?mm? 12?mm?
1.59?
,3.18?
mm/sec?
?? 850?K?
McLure? AL?6061? 400?
0.75?
inch?
0.215?
inch?
2?
mm/sec?
?? 430?C?
Nunes? AL?6061?
400?
to?
1200
D? D/2?
2?
mm/sec?
?? 420?C?
Lienert? AA?6082? 1250
0.75?
inch?
?? ?? ?? 360?C?
Lambrakos? AL?6061? 400? 25.4?mm 6.35?
2.17?
mm/sec?
5.72? 750?C?
Xu?
AL?Mg?
alloy?
250?
1.65,?1.2?
inches?
0.35?
inch?
4?to?7?
inches/m
in?
0.996?
inch?
1000?F?
Kovacevic?
AL?
6061?
1018?
Steel?
914? 25?mm?
5.5?
mm?
140?
mm/sec?
??
490?C???AL???
630?C???
Steel?
Khandkar?? AL?6061? 390? 25?mm? 10?mm?
142?mm/?
min?
8?mm? 460?C?
Chang? AL?6061? 1600 ?? ??
87?to?567?
mm/min?
?? 530?C?
Lui? AL?6061? 350? 16.5?mm
6.3?
mm?
2.5?
mm/sec?
?? 571?C?
Schmidt? AA?2024? 400???6?mm?
120?
mm/min?
3.5?mm? 400?C?
Colegrove? AA?7075? 330??? ??
105?
mm/min?
?? 530?C?
 
Table 1: Maximum temperatures recorded by various researchers with alloys and the 
          Friction Stir Welding parameters they used in their studies. 
 
 
 36
Classic Milling Theory applied to Friction Stir Welding 
               Thermal study and thermal modeling of the process was done in metal cutting also. Tay 
[27], in 1973, used a two dimensional finite element analysis tool to study the temperature 
distribution in the workpiece, chip and tool. He obtained the distribution of heat source from 
quick stop tests and flow field measurements. 
              Klamecki [28], in 1973, developed a model for the initial stages of the chip formation 
using a three dimensional finite analysis tool. Stephenson [29], in 1983, attempted to study the 
temperature distribution in the tool and chip in continuous cutting by developing a finite element 
program. But he had a difference of 50 C between his predicted temperature and his recorded 
temperature.  
             Stephenson and Ali [29], in 1992, modeled the tool tip to study the temperature 
distribution of the tool temperature in interrupted metal cutting. 
             Boothroyd [30], in 1963, also tried to measure the radiation from the workpiece, a 
technique similar to that of Schwerd. He photographed workpieces to find the radiation from the 
work, tool and the chip after heating the work pieces for about 600 C. To find the point to point 
radiation intensity a microdensitometer was used. 
             Shore [31],  in USA, Gottwein [32], in Germany and Herbert [33], in Great Britain, 
simultaneously, found a way to measure the mean temperature, by using thermocouple 
technique, along the face of a metal cutting tool. 
             A very valuable insight to the temperature distribution on shear plane, tool and the chip 
was given by, Schwerd [34] in 1937. But his method of measuring the infrared radiation from the 
tool, work and the chip to find the temperature distribution on the outside surface of the region 
did not reveal the temperature inside the workpiece. 
 37
            Abdel-Hamid [30], in 1996, developed a thermomechanical model to better understand 
the end milling cutting process using a high speed steel tool. The temperature distribution in both 
the tool and the chip was predicted during the cutting and as well as in the non-cutting period.  
He gave the temperature rise at the shear plane by, 
(1 ) /
s ss c
TPvA? ?= ?                                                  (25) 
Where, 
s
P - Instantaneous gross rate of energy liberation of shear source. 
           c? - Volumetric specific heat 
            v  - Cutting speed 
          
s
? - Portion of the shear heat source 
          
c
A - Instantaneous chip cross section area, given as  
sin( )
ct i
A sd?=                                                      (26) 
Where 
t
s - Feed rate per tooth 
           
i
? - Instantaneous cutter rotation angle 
          d  - Chip width. 
ANSYS was used to simulate the process for the study. He predicted the peak temperature to be 
around 800 K for a feed rate of 0.1 mm/tooth and a cutting speed of 40 m/min. 
               Martelloti [35] suggests that in up milling, when the tool enters the work, the relative 
velocity along the tool path is maximum. He also states that the value of the relative velocity will 
decrease as the tooth of the tool progress along the tooth path. So, he concludes that the value of 
velocity has a maximum value when the tool enters the work and has a minimum value when the 
tool leaves the work. The equation for finding the relative velocity is given by, 
1
2
2
2[{ } 2]VnRrry?=??                                            (27) 
 38
        Where - RPM n
                    R - Radius of the tool 
                    r - Radius of Pinion 
                    y - Depth of cut. 
In equation 1.3, ? ? will be used for up milling and ?-?  will be used for down milling, thus 
giving higher value for up milling and a lower value for down milling. 
+
 Shaw [33], gave an expression to calculate the total frictional energy generated at the 
chip- tool interface per unit time per unit area, which is, 
CC
FV
q
Jab
=                                                          (28) 
Where, - Friction force along the tool face. 
C
F
             - Velocity of the chip relative to the tool. 
C
V
             a- Length of contact between chip and tool 
            b - Width of the chip. 
                Martelloti?s and Shaw?s principles were combined by Payton [36] for Friction Stir 
Welding as a specialized form of slot milling.   According to Payton, the pin tool used in the 
FSW process, works similar to a milling tool. So the numerical equations that apply for a milling 
tool also apply to the FSW tool. In Machining theory, the leading side or the up milling side of 
the workpiece would be hotter than the trailing or down milling side. This is due to the velocity 
of the tool tip would be maximum on the leading side and minimum on the trailing side. In FSW, 
the leading side is the advancing side and the trailing side is the retreating side. According to 
this, the advancing side of the piece in FSW is hotter than the retreating side. 
The heat equation was given as, 
 39
*
[16,148.58]
*( )
s
Vt
THP
kpc
=                                           (29) 
Where, 
s
HP - Specific Horsepower 
               V - Cutting speed 
               k - Coefficient of thermal conductivity 
             - Volume specific heat of the work material pc
               t  - Uncut chip thickness, given by 
 
12
*
*
cos (1 ) *
*
t
t
Fd
t
Fnd
R Dd d
RD?
?
=
?+ ?
                                   (30) 
 
Where, - Radius of the cutting edge R
            d  ? Depth of cut 
            D  ? Diameter of the cutter 
             n? Number of teeth 
            - Feet per tooth 
t
F
Payton claims that the above equation could be used to find the temperature of the workpiece 
during the experiment for a given set of parameters, which would help the welder to design a tool 
which would avoid the defects. Payton?s model forms the basis for this project 
Conclusions of the literature review       
 Although there have been several flow models and heat source models proposed in the 
refereed literature of Friction Stir Welding, there is a shortage of real world thermal data against 
which to validate the proposed models.   This is largely due to the contained nature of the actual 
 40
Friction Stir Welding process.  The heat sources of deformation are located within the base metal 
being welded and underneath the rotating tool shoulder.  As such, surface pyrometry measures 
are ineffective.   The limited amount of thermal data in the reported literature is usually 
coincidental to the primary goal of the experimental design and non-symmetrical. 
 Thermal studies are also complicated by the complex methods required to embed 
thermocouples within the work-piece.   This is further compounded by the attention to detail 
which would be required to exactly place the thermocouples in the same place each time within 
the specimen as it is being prepared. 
 None of the proposed models have been validated against more than a few data runs.   A 
quick way to collect a statistically valid body of thermal data has eluded researchers attempting 
to model the FSW process. 
 
 
 
 
 
 
 
 
 
 
 
 
 41
 
 
 
 
IV. MATERIALS, EQUIPMENT AND SOFTWARE 
 
 This chapter documents the materials, equipment and software used in the conduct of the 
experimental research.    
Overview 
 The core of the final experimental setup consists of a steel holder (Alloy 1020) which 
acts as the Friction Stir Welding anvil, work-holder and thermocouple holder (Figure 23 and 
Figure 24).   The unit allows for a rapid exchange of specimens without the need to place or 
replace thermocouples needlessly. 
 
 
Figure 23: Side view of the steel holder with stainless steel wells 
 42
 
Figure 24: End view of the steel holder.   Tool is inserted into the sample 
 The holder has two slotted provisions, one for the aluminum shoulder specimen and the 
other for the aluminum pin specimen. The specimens are clamped to the holder using screws on 
either ends of the holder and through out the length of the stock. The dimensions and the material 
of the elements in the holder are given below.   Figure 25 provides a schematic representation of 
the instrument base. 
 
Figure 25: Rough schematic of the instrument above.   The thermocouple wells must be 
drilled at convergent angles in order to place all 5 thermocouple data points in the same plane 
as the tool moves through the plane.  A load cell is located underneath the entry point on the 
left.    The advancing side of the weld is located on the side away from the operator as he 
faces the machine.    The tool enters from the left.  Normal rotation is clockwise. 
 
 43
Materials: 
 The steel base is steel alloy 1020.   The thermocouple wells are made from stainless steel 
318.   The specimens (workpiece) used in this experiment was aluminum alloy (AA) 6061 with a 
T6 temper.   Common commercial dimensions were used to create the top stock and bottom 
stock quickly using a simple cutoff saw. 
Top (Shoulder) stock dimension: 
Length: 22 inches 
Breadth: 2 inches 
Height: 0.125 inches 
Bottom (Pin) stock dimension: 
Length: 21 inches 
Breadth: 0.5 inches 
Height: 0.5 inches 
 The actual friction stir welding tools were manufactured from titanium alloy purchased 
from TIMETAL.    The alloy used was AMS 4928N / HT R6968.   The material machines easily 
and had outstanding wear characteristics. 
 OMEGATHERM was used to seal any gaps which may have existed along any 
boundaries.   High accuracy thermocouples (OMEGA HH-K-20-SLE) with an accuracy of +/- 2 
degrees were used in the construction of the base holder. 
 Closed end stainless steel wells were an OMEGA product (SS-316) with a standard 
diameter of 0.1875 inches and a uniform wall thickness of 0.025 inches. 
 
 
 44
Fabricated Devices: 
 The base holder was fabricated from steel alloy 1020.   Basic dimensions are depicted in 
Figure 26.   Overall length was 25 inches.   Widest width was 10 inches.   Maximum height was 
2 inches.   Figure 23 above illustrates it photographically.  
 
Figure 26: Drawing of the holder with dimensions in inches without the accommodations 
for steel wells; all dimensions are in inches. 
 The pin tools used in this experiment were simple pins without threading.   Three tools 
were used for the experiment with three different shoulder diameters as illustrated below.    All 
tools were produced on a computer numerically controlled (CNC) HAAS TL2 tool-room lathe 
locally.    
Shoulder: Titanium alloy (AMS 4928N by TIMETAL) 
Pin:  Titanium alloy (AMS 4928N by TIMETAL) 
 45
Shoulder diameter: 0.9 inch, 0.7 inch, 0.5 inch 
Pin diameter: 0.25 inch 
Length of the pin: 0.3 inch 
Radius between the shoulder and the tool: 0.1 inches 
The tools were made on the CNC lathe. 
 
Figure 27:  Tool dimensions; All dimensions are in inches.  
 
 
 
 
 
 46
Instrumentation: 
 National Instruments? LabVIEW Version 8.3 software was used to collect the data using 
a NI-CDAQ 9172 backplane chassis combined with NI 9205 instrumentation modules.  Figure 
28 shows the chassis with modules installed and thermocouples attached.  Figure 29 shows the 
entire final instrument from a distance in Auburn University?s Design and Manufacturing 
Laboratory.  Figure 30 and Figure 31 illustrates the software setups used to program the NI 9205 
modules for the thermocouples utilized. 
 
Figure 28:  NI 9205 modules installed in NI-CDAQ-9172 chassis. 
 
Figure 29:  Entire instrument installed in CINNINATI ARROW CNC Mill 
 47
 
Figure 30:   Basic thermocouple setup LABVIEW 8.3 program. 
 
 
Figure 31:  LabVIEW 8.3 setup for multiple thermocouples  
 48
Machinery 
 A Cincinnati ARROW 750 mill was used (see earlier Figure 28) as the friction stir 
welding platform in this experiment.   A HAAS TL-2 CNC lathe was utilized to fabricate the 
welding tools and a HAAS TM-2 CNC mill was utilized to fabricate the steel 1020 holder of 
Figures 23, 24 and 26.   
Software  
 National Instruments? LabVIEW version 8.3 was utilized to build the instrumentation 
circuitry.  A commercially available multiphysics finite element modeling software program 
developed by FEMLAB called COMSOL was utilized to simulate the results.  COMSOL was 
utilized because of its low cost and availability to the major professor and student.   Additionally, 
COMSOL works well with the popular MATLAB software used throughout academia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 49
 
 
 
 
 
V. CONSTRUCTION AND METHODOLOGY OF INSTRUMENT 
 
 The experimental objective of the instrument is to simultaneously and symmetrically 
measure the thermal field around a tool as it passes a given plane.   Ideally, this would capture 
the temperature under the shoulder at its maximum radius, just under the pin as it passes 
overhead and below the midpoint side of the pin.  This ?nominal? configuration requires a 
pairing of 5 thermocouples to completely capture the transverse plane as the tool passes by the 
observation point. Multiple pairings allow the tool to pass through the thermocouples. 
 Embedding thermocouples in such a manner for every sample is a very time consuming 
effort.    Controlling the variability of thermocouple fabrication and placement is an extremely 
daunting prospect for even a skilled technician.  Each of the earlier studies was conducted on a 
limited number of carefully prepared samples.     It became apparent that it would be best to 
construct a standard holder with embedded thermocouples.   Instead of changing the 
thermocouples, the experimentalist changes the sample.  This allows the collection of a large 
body of statistical data rapidly.   Figure 25 provides a schematic view.   Figure 23 is a 
photograph of the instrument.   The first two sets of thermocouples are separated from the final 
five sets to allow for the dwell in temperature during tool insertion.    The last five sets or 
pairings all view the approaching tool as a steady state, if transient, heat source. 
 
 50
 
Figure 32: Rough schematic of the instrument above.   The thermocouple wells must be 
drilled at convergent angles in order to place all 5 thermocouple data points in the same 
plane as the tool moves through the plane.  A load cell is located underneath the entry 
point on the left.    The advancing side of the weld is located on the side away from the 
operator as he faces the machine.    The tool enters from the left.  Normal rotation is 
clockwise. 
 51
            
Figure 33:   Side and end view of the thermal profile instrument. In the first photograph, 
the tool enters on the left hand side and traverses to the right along a 22 inch long path.   
Exit is marked in each view in white. In the first photograph the run is complete (tool up).  
A run is in progress in the second photograph (tool down, moving towards the front). 
 52
The overall length of the data run is 22 inches measured from the tool entry point until the tool 
exit point.   Seven thermocouple pairs are arranged along the length of the unit in groups of five 
for a total of 35 thermocouples.   Within each planar group, there are two shoulder 
thermocouples (advancing and retreating side), two pin sides (advancing and retreating sides), 
and one under the pin thermocouple.     The first two of seven pairs are separated by a small 
offset to the left hand side from the final 5 paired sets.  Downward thrust force is also recorded at 
this time using a small force plate underneath the entry point.  This allows for observation of the 
tool entry and initial traverse independently of the routine weld passage observed through the 5 
pairs on the right hand side. 
 The design objective was to achieve an instrument with a statistical power of 90%.    The 
unit is currently capable of capturing a +/- 2 degree Celsius temperature difference 95% of the 
time with 7 replicates of each factor level.   The long term goal is to achieve 95% confidence 
with 3 replicates of each factor level.   Many samples per hour are possible since only a slight 
cooling off period is needed between samples. 
 The holder is designed to accept commercially available stocks in order to fit the cavity 
in the steel holder. The pin stock (0.5 by 0.5 inches) was dropped in the pin cavity and the 
shoulder stock (2.0 inches by .125 inches) was dropped in the shoulder cavity. Both the stocks 
were clamped down to the steel holder with a screw on each corner of the holder and through out 
the length of the holder. Figure 33 above shows the end view of the holder with thermocouple 
placements. 
 Thermal grease (OMEGASEAL) was used to fill the air gap between the aluminum stock 
and the steel holder. The holder has provision to measure seven sets of data. At each ?set? for 
that plane of the work piece, a total of five holes were drilled at different locations to 
 53
accommodate five K-type thermocouples.  
Force Measurements 
 Traditionally, Friction Stir Welding machines are run in a method referred to as ?load 
control? where the machine maintains a constant downward force.   Auburn University does not 
have machines capable of performing ?load control? welding.  In order to ensure a consistent 
starting downward force, the downward force of the pin tool was measured using a load cell. The 
load cell was placed under the backside of the steel holder, directly under the point of entry of 
the tool.  A slot was machined in the back side of the holder to accommodate the load cell.   Data 
runs which fell outside the nominal pressure range were rejected for analysis.   Figure 34 shows 
the load cell placement. 
 
Figure 34: Placement of the load cell for measuring the downward force. 
 
 
 54
The specification of the load cell: 
Make: Omegadyne Inc 
Model: LCGD ? 30K 
Measuring Range: 0 ? 30000 lbs 
Sensitivity: 18.1112 mVdc 
Data Collection 
Lab view software and a data acquisition system were used to record the data from the 
thermocouples and the load cell.      Details of the software are presented in the section on 
equipment.     The basic flow diagram representation is depicted in Figure 35 below. 
 
Figure 35: Basic Data Collection Block Diagram 
 
The time versus temperature graph and the time versus force graph obtained has one sample 
every second. The reading obtained in a second would be the average of one thousand readings 
obtained in the same second. This was done to reduce the noise disturbances. 
Thermocouple Nomenclature and placement 
 There are seven sets of thermocouples in the apparatus.   Each set includes 5 
thermocouples.  Figure 36 shows the placement of the terminal thermocouple joints within the 
apparatus.   
 55
 
Figure 36: Physical placement of the thermocouples 
Thermocouple Designations: 
Each Thermocouple was represented by a unique alphabet and number. 
 S ? Shoulder thermocouple. (Retreating side: S
1
, S
3?
 S
13;
 Advancing side: S
2
, S
4?
 S
14
) 
P ? Pin thermocouple. (Retreating side: P
1,
 P
3?
 P
13;
 Retreating side: P
2
, P
4
? P
14
) 
U ? Under the pin thermocouple. (U
1,
 U
2?
 U
7
) 
The first two sets of thermocouples are separated from the last five sets of thermocouples.  The 
intention here is that the first two sets are observing the entry of the tool.   The last five sets all 
observe and document the approaching tool as it transits a weld position.   This is similar to a 
long production weld such as on the space shuttle external fuel tank. 
 56
Recorded and Calculated Thermodynamic Results 
            The experimental method used in this project does not directly record the temperature of 
the specimen. Since the welled thermocouples are placed in the steel holder, the recorded 
temperatures would be the steel temperature at the thermocouple tip within the well. Steady state, 
one dimensional heat transfer equations were used to find the approximate temperature in the 
aluminum specimen.   This is necessary in order to compare real world data to the predicted 
values of the thermodynamic models. 
 Steady state, one dimensional heat transfer equation is based on Fourier?s law of 
conduction. According to Fourier?s law of conduction, the rate of heat flow through a 
homogenous solid is directly proportional to the area of the section perpendicular to the direction 
of heat flow, and the temperature difference along the path of heat flow. 
                                                        )/(*)/( dldTAdtdQ ? ,                                          (28) 
where,  Q ? Heat flux, 
             A ? Area of the section, 
             T ? Temperature. 
From equation 28, the one dimensional, steady state equation is derived as, 
                                                           Q = -kA (T
2
-T
1
)/ L                                                (29) 
where,  k ? Thermal conductivity of the material 
             Q ? Heat flux, 
             A ? Area of the section, 
             T
2
- Lower temperature, 
             T
1
- Higher temperature. 
The minus sign for thermal conductivity denotes that the heat is transferred in the direction of the 
 57
decreasing temperature. 
  Equation 29 can be rewritten as, 
                                                        Q = kA (T
1
-T
2
)/ L                                                     (30)     
  This project uses equation 30 for finding the temperature of the specimen. For one dimensional 
steady state condition, heat flux, Q is the same through out the material. In order to find Q, there 
has to be two known temperatures (T
1
, T
2
). 
 
Figure 37: One dimensional heat transfer 
 Measurement of temperatures (T
1
, T
2
) in FSW 
              In this case, temperature is measured only at one single point in the steel holder using 
the welled thermocouples which provides us with only one temperature, T
1
. In order to find T
2
, 
the same set up was used without the wells. Instead of welled thermocouples recording the 
temperature from the steel holder, plain thermocouples were used to find the temperature of the 
steel holder thus finding T
2
.  With two known temperatures (T
1
, T
2
) the heat flowing through the 
stainless steel well could be found. The calculated Q could be used to find the temperature of the 
specimen using equation 30. 
 
 
 58
Application of One dimensional steady state Conduction to the instrument 
              Equation 30 was used in the project to find the temperature of the specimen. Cross 
section view of the holder is shown in figure 38. 
 
Figure 38:  Cross sectional view; Welled thermocouple in the steel Holder 
The interfaces represented at each boundary are: 
1 ? T
s
: Temperature of the specimen at the aluminum/steel interface 
2 ? T
h
: Temperature of the steel holder. (Recorded data, T
2
) 
3 ? T
w
: Temperature of the stainless steel. (Recorded data, T
1
) 
4 ? T:  Temperature at the tool. 
In order to calculate the actual temperature at the tool: 
CASE 1: To find Q: 
     Q = (T
w
-T
h
)k
1
A/L
1
  
CASE 2: To find Ts: 
    T
s
= (QL
2
/k
2
A) +T
h
 
CASE 3: To find T: 
   T= (QL
3
/k
3
A) +T
s
 
 59
Where, 
k
1
 - Thermal conductivity of steel well 
k
2
 - Thermal conductivity of steel holder 
k
3
 - Thermal conductivity of welded metal (Aluminum) 
A- Area of cross section. 
L
1
- Thickness of the well. 
L
2
- Distance between the aluminum/holder interface and the face of well.  
 A detailed, worked example is provided in the results appendices.   The author shall 
clearly document all reported results as either ?recorded?, which is the actual live data without 
manipulation, or ?calculated?.   Calculated data has been transposed as detailed above.  This is 
necessary so that actual recorded data can be compared against the predicted results of the 
thermal models in the literature. 
 
 
 
 
 
 
 
 
 
 
 
 60
 
 
 
 
 
VI. VALIDATION OF THE INSTRUMENT 
 
Construction and design of a new instrument requires validation.   One must conduct trials to 
answer three fundamental questions: 
(1) Is the instrument repeatable for constant operating conditions? 
(2) Is the instrument sensitive enough to detect small changes? 
(3) What is the standard deviation of the instrument? 
Once these questions are answered, the confidence level or statistical power of the instrument as 
an observational tool can be determined.  
Two different experimental conditions (factor levels) were chosen to check the 
repeatability and the sensitivity of the instrument. The different conditions are given in the 
following table (Table 2). 
 
Factor Level 
Combination 1 
Factor Level 
Combination 2 
Factor  
Tool Shoulder Diameter 1 inch 0.7 inch 
Tool Pin Diameter 0.25 inch 0.25 inch 
RPM 400 400 
Traverse speed 4 inches/min 4 inches/min 
 
Table 2: Experimental conditions for repeatability and sensitivity analysis 
 
The basic premise of a repeatability study is that one conducts a number of runs over several 
days using the same operating parameters.      Then the investigator compares the statistical 
 61
averages against each other using a T-test.   If the instrument is valid, then there should be no 
difference between the results.    Having established repeatability, the investigator then 
deliberately varies the operating parameters to establish that the instrument is sensitive to 
changes in parameters. 
The instrument has seven sets of thermocouples. The first two capture the entry of the 
tool and the last five capture the traverse of the tool.  Temperatures captured by all the 
thermocouples were recorded. First two sets which capture the entry of the tool are called ?Entry 
1? and ?Entry 2?.  The remaining five sets which capture the traverse of the tool are called 
?Group 1?, ?Group 2?, ?Group 3?, ?Group 4? and ?Group 5?.    The following thermocouple 
labeling system was used throughout the data capture runs. 
Entry 1 Entry 2 Group 1 Group 2 Group 3 Group 4 Group 5 
S1 
RE 
S2 
AD 
S3 
RE 
S4 
AD 
S5 
RE 
S6 
AD 
S7 
RE 
S8 
AD 
S9 
RE 
S10 
AD 
S11 
RE 
S12 
AD 
S13 
RE 
S14 
AD 
P1 
RE 
P2 
AD 
P3 
RE 
P4 
AD 
P5 
RE 
P6 
AD 
P7 
RE 
P8 
AD 
P9 
RE 
P10 
AD 
P11 
RE 
P12 
AD 
P13 
RE 
P14 
AD 
U1 U2 U3 U4 U5 U6 U7 
 
Table 3: Thermocouple labeling assignments 
 
Where RE ? Retreating side 
           AD ? Advancing side 
              S ? Under the shoulder thermocouple 
              P ? Side of the pin thermocouple 
             U ? Under the pin thermocouple 
 
 62
Repeatability test: Five experiments were done for the 1 inch shoulder tool at different times to 
validate the instrument?s repeatability.   There were five replicates within each experiment.   The 
standard deviation of the five readings obtained then calculated.  If the standard deviation found, 
lies within the standard deviation of the instrument, the data is said to be repeatable.  If the 
standard deviation found exceeds the standard deviation of the instrument, the data is considered 
not repeatable. 
The following are the peak temperatures recorded during the experimental runs at 
different thermocouples. 
Experimental condition: 1 Inch Shoulder tool, 400 RPM, 4 IPM 
 
?? Entry 1 Entry 2 Group 1 Group 2 
 S1 S2 S3 S4 S5 S6 S7 S8 
RUN 1 169.35 171.64 187.07 195.99 170.01 189.3 140.92 154.4 
RUN 2 168.74 181.77 181.91 192.58 172.76 184.41 148.84 158.9 
RUN 3 162.06 180.14 183.1 190.89 172.41 184.15 142.61 153.5 
RUN 4 170.52 179.44 190.17 192.63 174.5 184.23 142.07 154.3 
RUN 5 174.41 183.69 178.83 190.75 171.81 182.65 145.97 161.7 
  ?? ?? ?? ?? ?? ?? ?? ??
?? Group 3 Group 4 Group 5 ?? ?
 S9 S10 S11 S12 S13 S14 ?? ?
RUN 1 158.68 168.73 152.97 178.61 131.49 139.31 ?? ?
RUN 2 168.61 170.69 167.08 182.56 132.34 150.61 ?? ?
RUN 3 158.52 171.27 159.74 180.71 132.97 144.93 ?? ?
RUN 4 163.43 183.76 167.16 189.46 141.67 152.7 ?? ?
RUN 5 162.4 184.26 161.4 190.37 133.93 149.49 ?? ?
 
Table 4: Recorded Temperature for repeatability analysis for shoulder thermocouples: All values 
in Celsius 
  
 
 
 
 63
  Entry 1 Entry 2 Group 1 Group 2 
 P1 P2 P3 P4 P5 P6 P7 P8 
RUN 1 175.43 196.41 138.98 146.66 142.99 180.93 150.74 161 
RUN 2 185.07 195.78 142.99 154.14 147.94 176.77 147.23 163.7 
RUN 3 184.89 190.34 136.2 150.83 143.94 164.43 142.62 167.6 
RUN 4 186.68 191.26 137.87 147.54 149.46 177.51 152.55 170.2 
RUN 5 180.56 195.55 139.43 151.8 153.87 183.84 156.51 177.7 
  ?? ?? ?? ?? ?? ?? ?? ?
  Group 3  Group 4 Group 5   
 P9 P10 P11 P12 P13 P14   
RUN 1 149.66 163.42 128.13 170.88 126.8 139.86   
RUN 2 160.51 165.9 139.44 171.73 128.26 148.47   
RUN 3 151.28 164.68 135.09 170.13 128.37 138.96   
RUN 4 157.94 166.49 140.81 175.81 135.55 146.8   
RUN 5 159.35 171.83 142.38 183.25 128.51 147.78   
?
  
Entry 
1 
Entry 
2 
Group 
1 
Group 
2 
Group 
3 
 Group 
4 
Group 
5 ??
 U1 U2 U3 U4 U5 U6 U7 ??
RUN 1 195.8 175.86 179.65 140.48 137.14 164.07 125.92 ??
RUN 2 199.55 177.34 179.61 148.46 143.34 157.32 131.61 ??
RUN 3 210.61 167.6 170.36 142.47 137.5 164.25 124.67 ??
RUN 4 197.8 174.28 174.54 145.03 139.48 170.42 135.17 ??
RUN 5 210.24 168.34 176.76 154.08 148.56 164.16 127.9 ??
 
Table 5: Recorded Temperature for repeatability analysis for pin and under the pin 
thermocouples: All values in Celsius 
Standard Deviation: 
The standard deviation of the instrument was given by,  
Standard deviation of the instrument = Standard deviation of the thermocouples + 
                                                               Standard deviation of the module NI 9205.                                                
 64
The standard deviation of thermocouples is ?2 degree Celsius. The module has a standard 
deviation of at least ?2% of the recorded temperature. So, the module standard deviation range 
was ?2% of the average of the five peak temperature readings of the same thermocouple. 
 The standard deviation of the instrument was compared with the standard deviation of the 
readings obtained for each thermocouple. The following tables show the comparison of the 
standard deviations. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 65
 
 
Std 
deviation 
of 
readings 
(1) 
Std dev 
range of 
thermoc
ouples 
Std 
deviation 
of 
modules 
Std 
deviation 
range of 
modules 
Std 
deviation 
range of 
instrument 
(2) 
Is 
 
(1)<(2) Remarks 
S1 RE 4.47 4 3.38 6.76 10.76 Yes Repeatable 
S2 AD 4.61 4 3.59 7.17 11.17 Yes Repeatable 
S3 RE 4.45 4 3.68 7.37 11.37 Yes Repeatable 
S4 AD 2.11 4 3.85 7.70 11.70 Yes Repeatable 
S5 RE 1.62 4 3.45 6.89 10.89 Yes Repeatable 
S6 AD 2.53 4 3.70 7.40 11.40 Yes Repeatable 
S7 RE 3.26 4 2.88 5.76 9.76 Yes Repeatable 
S8 AD 3.57 4 3.13 6.26 10.26 Yes Repeatable 
S9 RE 4.14 4 3.25 6.49 10.49 Yes Repeatable 
S10 AD 7.61 4 3.51 7.03 11.03 Yes Repeatable 
S11 RE 5.89 4 3.23 6.47 10.47 Yes Repeatable 
S12 AD 5.29 4 3.69 7.37 11.37 Yes Repeatable 
S13 RE 4.12 4 2.69 5.38 9.38 Yes Repeatable 
S14 AD 5.35 4 2.95 5.90 9.90 Yes Repeatable 
 
Table 6:  Repeatability of the shoulder sample temperatures 
        
 
Std 
deviation 
of 
readings 
(1) 
Std 
deviatio
n of 
thermoc
ouples 
Std 
deviation 
of 
modules 
Std 
deviation 
range of 
modules 
Std 
deviation 
of 
instrument 
(2) 
Is 
(1)<(2) Remarks 
P1 RE 4.57 4 3.65 7.30 11.30 Yes Repeatable 
P2 AD 2.84 4 3.88 7.75 11.75 Yes Repeatable 
P3 RE 2.51 4 2.78 5.56 9.56 Yes Repeatable 
P4 AD 3.09 4 3.00 6.01 10.01 Yes Repeatable 
P5 RE 4.40 4 2.95 5.91 9.91 Yes Repeatable 
P6 AD 7.42 4 3.53 7.07 11.07 Yes Repeatable 
P7 RE 5.28 4 3.00 6.00 10.00 Yes Repeatable 
P8 AD 6.45 4 3.36 6.72 10.72 Yes Repeatable 
P9 RE 4.94 4 3.11 6.23 10.23 Yes Repeatable 
P10 AD 3.22 4 3.33 6.66 10.66 Yes Repeatable 
P11 RE 5.74 4 2.74 5.49 9.49 Yes Repeatable 
P12 AD 5.43 4 3.49 6.97 10.97 Yes Repeatable 
P13 RE 3.45 4 2.59 5.18 9.18 Yes Repeatable 
P14 AD 4.58 4 2.89 5.78 9.78 Yes Repeatable 
 
Table 7: Repeatability of the Pin Tool Temperatures 
 
 66
 
 
Std 
deviation 
of 
readings 
(1) 
Std 
deviati
on of 
thermo
couple
s 
Std 
deviation 
of 
modules 
Std 
deviation 
range of 
modules 
Std 
deviation 
of 
instrument 
(2) 
Is 
(1) < (2) Remarks 
U1 7.09 4 4.06 8.11 12.11 Yes Repeatable 
U2 4.45 4 3.45 6.91 10.91 Yes Repeatable 
U3 3.90 4 3.52 7.05 11.05 Yes Repeatable 
U4 5.37 4 2.92 5.84 9.84 Yes Repeatable 
U5 4.79 4 2.82 5.65 9.65 Yes Repeatable 
U6 4.63 4 3.28 6.56 10.56 Yes Repeatable 
U7 4.31 4 2.58 5.16 9.16 Yes Repeatable 
 
Table 8: Repeatability analysis for the underside of the pin tool temperatures 
 
The above table indicates that the instrument reliably repeats temperatures.  Additionally, 
T-Tests of between replicate samples indicate that there is no variance in the standard deviation 
of the instrument between replicate runs. 
Sensitivity test  
Five additional experiments were done for the second condition and five readings were 
obtained for all the thermocouples.  Since the shoulder diameter was the only change between 
the two conditions, the data obtained were statistically compared for a change in the shoulder 
temperature.  Since the ?Entry 1? and ?Entry 2? thermocouples were used to capture the entry 
point of the tool, these sets were not considered while comparing the two conditions. Groups 1 to 
5 were considered for the sensitivity test. Temperature recorded by the shoulder thermocouples 
from Group 1 to 5 in both the conditions were subjected to statistical T-test to find if there is a 
difference in the data obtained or not. 
 Table 8 shows the peak temperatures obtained in the shoulder thermocouples of Group 1 
to 4 with the 1 inch shoulder tool.    Table 9 shows the peak temperatures for the 0.7 inch 
shoulder tool. 
 67
Condition 1:   1.0 inch shoulder tool, 400 RPM, 4 IPM 
  Group 1 Group 2 Group 3 Group 4 
  S5 RE S6 AD S7 RE S8 AD S9 RE S10 AD S11 RE S12 AD 
RUN 1 170.01 189.30 140.92 154.38 158.68 168.73 152.97 178.61 
RUN 2 172.76 184.41 148.84 158.88 168.61 170.69 167.08 182.56 
RUN 3 172.41 184.15 142.61 153.48 158.52 171.27 159.74 180.71 
RUN 4 174.50 184.23 142.07 154.25 163.43 183.76 167.16 189.46 
RUN 5 171.81 182.65 145.97 161.67 162.40 184.26 161.40 190.37 
 
Table 9: Peak average temperature readings for 1.0 inch shoulder tool (Condition 1) 
 
Condition 2:  0.7 inch shoulder tool, 400 RPM, 4 IPM 
  Group 1 Group 2 Group 3 Group 4 
  S5 RE S6 AD S7 RE S8 AD S9 RE S10 AD S11 RE S12 AD 
RUN 1 130.55 150.87 132.09 146.29 142.54 154.48 140.19 153.06 
RUN 2 128.32 151.18 137.28 149.55 146.99 160.35 145.98 151.07 
RUN 3 133.80 156.96 133.73 150.98 147.51 163.88 148.95 155.30 
RUN 4 141.94 154.45 143.30 148.73 147.38 169.95 141.00 151.21 
RUN 5 137.41 155.54 143.03 151.21 155.85 166.05 140.34 156.19 
 
Table 10: Peak average temperature readings for 0.7 inch shoulder tool (Condition 2) 
 
The data obtained by the same thermocouples at different conditions were compared with 
each other. For example, S5 of condition 1 was compared to S5 of condition 2 and S6 of 
condition 1 was compared with S6 of condition 2 and so on. A T-test was used to determine if 
the averages were different.   If they were in fact different, then the instrument is sensitive to the 
change in variable. 
 
 
 
 
 68
Statistical Analysis: Null Hypothesis: Condition 1 equals Condition 2 
Comparison 
(1) Vs (2) 
Diff in 
Averages 
of the two 
sets 
Pooled 
STD 
DEV 
Ratio 
between 
diff and 
pooled 
STD DEV 
P 
value 
Power Comment Report 
S5(1) Vs S5(2) 37.9 4.01 9.45 0 100% (1)?(2) Sensitive 
S6(1) Vs S6(2) 31.15 2.61 11.93 0 100% (1)?(2) Sensitive 
S7(1) Vs S7(2) 6.19 4.32 1.43 0.05 51% (1)=(2) Not Sensitive 
S8(1) Vs S8(2) 7.18 2.89 2.48 0 92% (1)?(2) Sensitive 
S9(1) Vs S9(2) 14.27 4.49 3.18 0 99% (1)?(2) Sensitive 
S10(1) Vs S10(2) 12.8 6.8 1.88 0.02 74% (1)=(2) Not Sensitive 
S11(1) Vs S11(2) 18.38 5.02 3.66 0 100% (1)?(2) Sensitive 
S12(1) Vs S12(2) 30.97 4.08 7.58 0 100% (1)?(2) Sensitive 
S13(1) Vs S13(2) 21.83 5.53 3.94 0 100% (1)?(2) Sensitive 
 
Table 11:  Sensitivity analysis.  Rejecting the null hypothesis proves sensitivity. 
 
 
All the thermocouples are sensitive enough to sense the difference in the change in the 
operating condition, except S7 and S10 due to the low ratio between the difference and the 
pooled standard deviation. The system should have at least a ratio of 2.4 to sense a difference 
with a 90% Power, which implies a 90% confidence interval.  The system is sensitive enough to 
capture the difference when the groups are compared as a whole instead of the comparison made 
individually on the thermocouples.  
Power 
Statistical power is defined as the probability that a statistical analysis will detect a 
difference which actually exists between two sets of data.  In other words, power is the statistical 
term used to describe the ability to reject the null hypothesis when it is false, as indicated in the 
statistical hypothesis diagram (Figure 39). 
 69
 
Figure 39:  Statistical Hypothesis Diagram 
Statistically, power is1 ?? , where? , is the possible chance of committing a Type II error. 
 
           The power of an experiment depends on the size of the experiment and the 
variability of the response.  Factors that influence the power are the size of the effect, 
sample variance, sample size and the criteria for significance. Standard error will also play 
a vital error in calculating the power. Decreasing the standard error will increase the 
power. Standard error will be decreased by increasing the sample size or by reducing the 
sigma of the dependent variable. 
            Power analysis could be done either before the experiment or after the experiment. 
Power analysis done before the experiment helps to determine the sample size of the 
experiment when the repeatability and sensitivity of the instrument have been established.  
The maximum power a test could have is 1.0 and the minimum is zero (0).  Ideally an 
experiment which has a power close to 1 is considered to be an experiment with high 
power.   Ninety percent (0.9) is often the goal of the experimentalist. 
 
 
 70
Number of replicates 
            Number of replicates for the actual experiments are found considering S1 and S2 in 
the repeatability test.  A statistical power test is done on the following data to find 
significant differences between S1 and S2. 
? Entry 1 
 S1 S2 
RUN 1 169.35 171.64 
RUN 2 168.74 181.77 
RUN 3 162.06 180.14 
RUN 4 170.52 179.44 
RUN 5 174.41 183.69 
 
Table 12:  Data to find the number of replicates for the experiments 
The Power test results for a sample size of 5 is shown below (using MINITAB 15) is: 
 
      2-Sample t Test 
 
      Testing mean 1 = mean 2 (versus not =) 
      Calculating power for mean 1 = mean 2 + difference 
      Alpha = 0.05  Assumed standard deviation = 4.53 
 
                  Sample 
      Difference    Size     Power 
         -10.32       5  0.882346 
 
According to the analysis, S1 and S2 are significantly different stating that both data are 
not equal with a statistical power of 88%.  In statistics, an experiment is considered as a 
strong experiment if it has a power of 90%. In order to obtain a power of 90%, the sample 
size of the experiment is increased from 5 to 7.  An other power test is done for the same 
experiment with a sample size of 7. 
 
 
 
 
 71
The Power test result for a sample size of 7 is shown below. 
 
     2-Sample t Test 
 
     Testing mean 1 = mean 2 (versus not =) 
     Calculating power for mean 1 = mean 2 + difference 
     Alpha = 0.05  Assumed standard deviation = 4.53 
 
 
                 Sample 
     Difference    Size     Power 
         -10.32       7  0.973899 
 
 
 
Figure 40: Statistical Power test result for a sample size of 7 
For a sample size of 7, statistically S1 and S2 are significantly different with a power of 
97%.  Since the power of the experiment is above 90%, this is considered to be a 
statistically strong experiment.  In order to maintain a statistical power of 90% for all the 
experiments, a sample size of 7 is maintained in all the experiments done. 
 
 
 
 
 
 72
 
 
 
 
 
VII. STATISTICAL DESIGN OF THE EXPERIMENT 
 
 Upon validating the instrument and determining the number of replicates needed to 
achieve a confidence level of at least 90 percent, the factor levels of the main experiment could 
be determined.  As in all ANOVA experiments, it is desirable to separate the factor levels by as 
large a deterministic value as possible while retaining meaningful physical results.   For this 
physical experiment, it was desirable to look at the effects of tool shoulder diameter, revolutions 
per minute and traverse rate.   
 
 
 
Shoulder Diameter 
 
 
350 RPM 
 
 
450 RPM 
 
0.500 Inches 
 
Traverse 
4 Inches Per Minute 
 
Traverse 
7 Inches Per Minute 
 
0.700 Inches 
 
Traverse 
4 Inches Per Minute 
 
Traverse 
7 Inches Per Minute 
 
0.900 Inches 
 
Traverse 
4 Inches Per Minute 
 
Traverse 
7 Inches Per Minute 
 
Figure 41: Factor Level Combinations of the Principal Experiment. 
 
 
 73
 Two minor additional experiments were also conducted to investigate principal effects 
suggested by the first experiment.     In the first of these experiments, the shoulder pin size was 
held constant at 0.300 inches while the shoulder diameter was increased incrementally as 0.500 
inches, 0.600 inches, 0.700 inches, 0.800 inches, 0.900 inches and finally 1.000 inches.   
Revolutions were held constant at 400 RPM and the traverse speed was limited to 4 inches per 
minute (IPM).  Two replicates of each combination were conducted (statistical power of 80%).  
In the final experiment, the top plate of aluminum was replaced by Copper Alloy 10101 to study 
material effects.  Five replicates of this experiment were conducted at 400 RPM, 1.0 inch per 
minute traverse.   The pin tool in this case had a 0.500 inches shoulder and a 0.300 inches pin. 
 Although experiments can be conducted quite quickly with this setup, financial 
considerations limited the investigation to the parameters detailed above.    The apparatus can 
collect as much data in a week as was formerly possible in a year of research. 
 
 
 
 
 
 
 
 
 
 
 
 
 74
 
 
 
 
 
VIII. RESULTS OF THE EXPERIMENT 
 
  The experiments were conducted over a three week period using the factor levels set out 
in Chapter VII.   Exemplars of each result are discussed in this chapter.   Complete data sets are 
contained in the appropriate appendix cited below.    General observations are documented in 
this section.   Comprehensive discussions are conducted in Chapter X. 
Principal Experiment: 
 Three different shoulders, with 2 different RPMs and 2 different traverse rates were used 
in this experiment. That produced twelve (12) possible factor level combinations in a full 
factorial experiment. Seven replicates of each factor level combinations were conducted in a 
random fashion suggested by a random number generator for a total of 84 welds.   This produced 
an experiment with a confidence factor of at least 95%.     
 The instruments data was originally organized as entry data (at position E1 and E2), with 
traverse group points (G1, G2, G3, G4 and G5), corresponding to the 3
rd
 through 7
th
 set of 
thermocouples.    Post data analysis indicated that the G5 transit point was always lower than G1 
through G4 averages.    This is due to the physical fact that the pin was extracted by the 
computer exactly at position G5, which precluded the passage of the trailing edge of the tool 
over that portion of the weld.   This prevented the total heat input seen at positions G1 through 
G4.   Therefore, T5 is more properly referred to as X1, the exit point.    
 
 
 
 75
Tabular Data 
 The complete set of raw recorded data is provided in Appendix A.    This data represents 
the actual ?thermocouple data? at the tip of the thermocouple wire.   This data was then 
translated mathematically into the ideal calculated value at the positions specified at the tip 
shoulder, pin surface and under the base of the pin.  The calculated values are listed in Appendix 
B, along with a worked example of the calculations.    The calculated values will allow a direct 
comparison the simulated model values in the next chapter (IX, Simulation).    Table 12 
documents the average advancing and retreating temperatures for the advancing (AD) versus 
retreating (RE) sides of the equipment. 
Factor Levels AD Shoulder RE: Shoulder Ad: Pin Re: Pin Under Pin 
0.9"Sh/4"min/450 RPM 195.53 183.02 189.84 166.08 177.70 
0.9"Sh/4"min/350 RPM 182.33 169.32 178.21 158.14 163.98 
0.9"Sh/7"min/450 RPM 149.42 140.16 145.00 127.42 137.16 
0.9"Sh/7"min/350 RPM 145.74 132.87 140.09 123.04 129.32 
0.7"Sh/4"min/450 RPM 162.43 148.43 159.83 139.67 150.54 
0.7"Sh/4"min/350 RPM 162.99 149.84 157.18 135.72 148.92 
0.7"Sh/7"min/450 RPM 143.36 134.59 138.75 121.75 133.30 
0.7"Sh/7"min/350 RPM 141.60 132.97 137.49 121.67 131.76 
0.5"Sh/4"min/450 RPM 157.66 143.13 152.05 131.19 141.01 
0.5"Sh/4"min/350 RPM 138.10 125.21 135.61 114.73 128.93 
0.5"Sh/7"min/450 RPM 139.52 131.65 138.17 121.43 134.25 
0.5"Sh/7"min/350 RPM 116.31 103.58 116.77 98.47 106.99 
 
            Table 13: Raw experimental data: Average of peak temperatures of Entry 2,  
 
Group 1, 2, 3, and 4; All temperatures are in Celsius. 
 76
Table 13 documents the average CALCULATED advancing and retreating temperatures for the 
advancing (AD) versus retreating (RE) sides of the equipment. 
Factor Levels AD Shoulder RE Shoulder 
AD: 
Pin 
AD: Pin Under Pin 
0.9"Sh/4"min/450 RPM 379.76 367.25 377.83 354.07 370.98 
0.9"Sh/4"min/350 RPM 366.56 353.55 366.19 346.13 357.26 
0.9"Sh/7"min/450 RPM 333.65 324.39 332.99 315.41 330.44 
0.9"Sh/7"min/350 RPM 329.97 317.10 328.07 311.03 322.60 
0.7"Sh/4"min/450 RPM 346.66 332.66 347.82 327.66 343.82 
0.7"Sh/4"min/350 RPM 347.22 334.07 345.17 323.70 342.20 
0.7"Sh/7"min/450 RPM 327.59 318.82 326.74 309.74 326.58 
0.7"Sh/7"min/350 RPM 325.83 317.20 325.48 309.66 325.04 
0.5"Sh/4"min/450 RPM 341.89 327.36 340.04 319.17 334.29 
0.5"Sh/4"min/350 RPM 322.33 309.44 323.60 302.72 322.21 
0.5"Sh/7"min/450 RPM 323.75 315.88 326.16 309.42 327.53 
0.5"Sh/7"min/350 RPM 300.53 287.81 304.76 286.46 300.27 
 
Table 14:  Calculated data: Average of peak temperatures of Entry 2, Group 1, 2,  
 
3, and 4; All temperatures are in Celsius. 
 
Graphical Data 
 Appendix C provides comparitive graphs of all average values for each directly recorded 
factor level combination, in the example of Figure 42 on the next page.  Appendix D provides 
comparitive graphs of all average CALCULATED values for each factor level combination, as per 
the example in Figure 43.   This allows quick observation that in general the advancing side will 
always be hotter than the retreating side. 
 77
 
Figure 42:  Example of average of seven replicates for a factor level combination (FLC) . 
Advancing is higher than corresponding retreating side. 
 
 
Figure 43:  Example of average of seven replicates (calculated temperature) for a factor 
level combination (FLC).   Advancing is higher than corresponding retreating side. 
 
 78
Analysis of Variance (ANOVA) 
 Each response was analyzed individually using ANOVA with the STATEASE software 
program by DesignEase, Inc.    The individual responses are all detailed in Appendix E using the 
APA approved style format of Figure 44 below.   It is thus possible to rank the impact factors of 
the factor levels using the F-value as a ranking number. 
 
Figure 44: ANOVA table comparing the impact factors (F-Value) of the factor levels 
chosen for S1, the temperature on the advancing side of the shoulder. 
 
 
 
 
 
 79
Table 15 gives the first, second and third largest impact factors on the individual responses for 
the shoulders.     Table 16 gives the first, second and third largest impact factors for the pin.    
Table 17 ranks the impact factors for the underside of the pin. 
 
Responses 
1st largest 
Impact Factor 
2nd largest  
Impact Factor 
3
rd
 largest 
 Impact Factor 
S1 Shoulder Traverse speed Shoulder X Speed X RPM 
S2 Shoulder Traverse speed Shoulder X Speed X RPM 
S3 Traverse speed Shoulder Shoulder X Speed 
S4 Traverse speed Shoulder Shoulder X Speed 
S5 Traverse speed Shoulder Shoulder X Speed X RPM 
S6 Traverse speed Shoulder Shoulder X Speed X RPM 
S7 Traverse speed Shoulder RPM X Speed 
S8 Traverse speed Shoulder RPM X Speed 
S9 Traverse speed Shoulder RPM X Speed 
S10 Traverse speed Shoulder RPM X Speed 
S11 Traverse speed Shoulder RPM X Speed 
S12 Traverse speed Shoulder RPM X Speed 
S13 Traverse speed Shoulder RPM X Speed 
S14 Traverse speed Shoulder RPM X Speed 
 
Table 15: Ranking of Impact Factors for each shoulders thermal response. 
 
 
 80
Responses 
1st largest 
Impact Factor 
2nd largest  
Impact Factor 
3
rd
 largest 
 Impact Factor 
P1 Shoulder Traverse speed Shoulder X Speed X RPM 
P2 Shoulder Traverse speed Shoulder X Speed X RPM 
P3 Traverse speed Shoulder RPM X Speed 
P4 Traverse speed Shoulder Shoulder X Speed 
P5 Traverse speed Shoulder Shoulder X Speed 
P6 Traverse speed Shoulder Shoulder X Speed 
P7 Traverse speed Shoulder RPM X Speed 
P8 Traverse speed Shoulder RPM X Speed 
P9 Traverse speed Shoulder RPM X Speed 
P10 Traverse speed Shoulder RPM X Speed 
P11 Traverse speed Shoulder Shoulder X Speed 
P12 Traverse speed Shoulder Shoulder X Speed 
P13 Traverse speed Shoulder RPM X Speed 
P14 Traverse speed Shoulder RPM X Speed 
 
Table 16:  Ranking of Impact Factors for each pins thermal response. 
 
 
 
 
 
 
 81
Responses 
1st largest 
Impact Factor 
2nd largest  
Impact Factor 
3
rd
 largest 
 Impact Factor 
U1 Shoulder Traverse speed RPM X Speed 
U2 Traverse speed Shoulder Shoulder X Speed 
U3 Traverse speed Shoulder RPM X Speed 
U4 Traverse speed Shoulder RPM X Speed 
U5 Traverse speed Shoulder RPM X Speed 
U6 Traverse speed Shoulder Shoulder X Speed 
U7 Traverse speed Shoulder Shoulder X Speed 
 
Table 17: Impact factors for temperature underneath the pin tool. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 82
T-Tests: 
 T-Tests were conducted using the 7 replicates of each advancing side versus the seven 
replicates of the retreating side for each result.   The tables below provide the P value of the T-
test conducted for different factor level combination.    Table 18 documents the shoulder 
comparisons.    Table 19 documents pin comparisons.    Tables 20 and 21 document the 
power/confidence with which we can state the advancing sides are hotter than the retreating 
sides. 
Factor level 
combination 
S1 Vs 
S2 
S3 Vs 
S4 
S5 Vs 
S6 
S7 Vs 
S8 
S9 Vs 
S10 
S11 Vs 
S12 
S13 Vs 
S14 
                
0.9"Sh/4"min/450 RPM 0.001 0.000 0.001 0.001 0.002 0.000 0.000 
0.9"Sh/4"min/350 RPM 0.001 0.002 0.002 0.001 0.002 0.001 0.003 
0.9"Sh/7"min/450 RPM 0.000 0.000 0.000 0.000 0.002 0.001 0.000 
0.9"Sh/7"min/350 RPM 0.000 0.003 0.001 0.001 0.001 0.002 0.000 
0.7"Sh/4"min/450 RPM 0.004 0.000 0.000 0.000 0.000 0.000 0.002 
0.7"Sh/4"min/350 RPM 0.001 0.006 0.005 0.005 0.005 0.004 0.004 
0.7"Sh/7"min/450 RPM 0.000 0.001 0.000 0.000 0.002 0.000 0.000 
0.7"Sh/7"min/350 RPM 0.000 0.001 0.000 0.000 0.003 0.000 0.002 
0.5"Sh/4"min/450 RPM 0.001 0.001 0.002 0.002 0.002 0.002 0.000 
0.5"Sh/4"min/350 RPM 0.002 0.004 0.001 0.000 0.002 0.000 0.001 
0.5"Sh/7"min/450 RPM 0.001 0.002 0.003 0.001 0.000 0.000 0.000 
0.5"Sh/7"min/350 RPM 0.000 0.000 0.001 0.001 0.001 0.000 0.001 
 
Table 18: P-Values for the Retreating Versus Advancing Shoulders 
 83
 
Factor level 
combination 
P1 Vs 
P2 
P3 Vs 
P4 
P5 Vs 
P6 
P7 Vs 
P8 
P9 Vs 
P10 
P11 Vs 
P12 
P13 Vs 
P14 
                
0.9"Sh/4"min/450 RPM 0.000 0.001 0.000 0.001 0.000 0.000 0.000 
0.9"Sh/4"min/350 RPM 0.002 0.001 0.000 0.005 0.002 0.000 0.002 
0.9"Sh/7"min/450 RPM 0.000 0.000 0.001 0.000 0.000 0.000 0.000 
0.9"Sh/7"min/350 RPM 0.000 0.001 0.001 0.001 0.001 0.000 0.001 
0.7"Sh/4"min/450 RPM 0.000 0.001 0.000 0.000 0.000 0.000 0.000 
0.7"Sh/4"min/350 RPM 0.002 0.000 0.001 0.002 0.001 0.000 0.001 
0.7"Sh/7"min/450 RPM 0.000 0.000 0.000 0.000 0.000 0.000 0.002 
0.7"Sh/7"min/350 RPM 0.001 0.002 0.000 0.000 0.000 0.000 0.000 
0.5"Sh/4"min/450 RPM 0.002 0.004 0.001 0.001 0.001 0.000 0.001 
0.5"Sh/4"min/350 RPM 0.002 0.000 0.000 0.000 0.000 0.000 0.006 
0.5"Sh/7"min/450 RPM 0.001 0.000 0.000 0.000 0.000 0.000 0.000 
0.5"Sh/7"min/350 RPM 0.000 0.001 0.000 0.000 0.000 0.000 0.000 
 
Table 19: P-Values for the Retreating Versus Advancing Pin Temperatures 
 
 
 
Factor level combination 
S1 Vs 
S2 
S3 Vs 
S4 
S5 Vs 
S6 
S7 Vs 
S8 
S9 Vs 
S10 
S11 Vs 
S12 
S13 Vs 
S14 
                
0.9"Sh/4"min/450 RPM 98 99 98 98 96 99 99 
0.9"Sh/4"min/350 RPM 96 90 95 98 93 95 93 
0.9"Sh/7"min/450 RPM 99 99 99 99 95 98 99 
0.9"Sh/7"min/350 RPM 99 92 99 99 99 94 99 
0.7"Sh/4"min/450 RPM 97 97 96 96 95 95 99 
0.7"Sh/4"min/350 RPM 98 90 90 90 90 91 91 
0.7"Sh/7"min/450 RPM 99 97 99 99 95 99 99 
0.7"Sh/7"min/350 RPM 99 98 97 97 92 99 94 
0.5"Sh/4"min/450 RPM 90 99 99 99 99 99 95 
0.5"Sh/4"min/350 RPM 95 94 97 99 96 99 97 
0.5"Sh/7"min/450 RPM 99 99 97 96 96 99 98 
0.5"Sh/7"min/350 RPM 97 97 93 93 97 99 99 
 
Table 20: Power of the Retreating Versus Advancing Shoulders Observation 
 
 
 
 
 
 84
Factor level combination 
P1 Vs 
P2 
P3 Vs 
P4 
P5 Vs 
P6 
P7 Vs 
P8 
P9 Vs 
P10 
P11 Vs 
P12 
P13 Vs 
P14 
                
0.9"Sh/4"min/450 RPM 99 98 99 98 99 99 99 
0.9"Sh/4"min/350 RPM 95 96 90 96 90 94 95 
0.9"Sh/7"min/450 RPM 99 99 96 99 99 99 99 
0.9"Sh/7"min/350 RPM 99 97 97 97 96 99 97 
0.7"Sh/4"min/450 RPM 94 90 98 98 99 99 97 
0.7"Sh/4"min/350 RPM 99 99 98 96 98 99 98 
0.7"Sh/7"min/450 RPM 99 99 99 99 99 99 95 
0.7"Sh/7"min/350 RPM 97 95 99 99 99 99 99 
0.5"Sh/4"min/450 RPM 99 98 99 99 99 99 99 
0.5"Sh/4"min/350 RPM 99 99 99 99 99 99 90 
0.5"Sh/7"min/450 RPM 99 99 99 99 99 99 99 
0.5"Sh/7"min/350 RPM 98 99 99 99 99 99 99 
 
Table 21: Power of the Retreating Versus Advancing Pin Temperature observations 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 85
Shoulder Experiment 
 A second, smaller experiment was conducted where the shoulder was varied 
incrementally from 0.5 to 1.0 inches at a constant speed and RPM to investigate the shoulder 
radius to pin radius ratio (R/r) as a thermal factor.  Due to time and budget constraints, only two 
replicates were run for each factor level.  This results in a confidence level of 30% for the results. 
Table 22 summarizes the advancing and retreating side temperatures for the shoulder and pin for 
all different parameters used in this experiment. 
  Ad: Shoulder 
Re: 
Shoulder Ad: Pin Re: Pin 
1.0" Sh/4"min/400RPM 185.25 178.15 186.56 179.88 
0.9" Sh/4"min/400RPM 176.10 169.64 176.17 168.42 
0.8" Sh/4"min/400RPM 159.47 151.78 160.96 149.71 
0.7" Sh/4"min/400RPM 143.63 135.43 145.18 135.30 
0.6" Sh/4"min/400RPM 128.32 121.03 133.92 124.54 
0.5" Sh/4"min/400RPM 121.85 116.29 126.75 118.66 
 
Table 22: Raw experimental data: Average of peak temperatures of Entry 2,  
 
Group 1, 2, 3, and 4; All temperatures are in Celsius. 
 
 
Figure 45 shows the plot of the temperature behavior for various size shoulder diameters on the 
next page 
 86
 
Figure 45: Average of the calculated peak temperatures of the shoulder experiment 
 87
Aluminum Copper Experiment 
 The main objective of this experiment is to study the behavior of the temperature field 
when two dissimilar metals are welded together, with common tool geometry, common traverse 
and RPM.     
           For this experiment, the shoulder stock was selected to be copper 10101 and the pin stock 
was selected to be aluminum 6061-T6.  Tool shoulder diameter was 0.5 inches.  Tool pin 
diameter was 0.3 inches    Tool pin height was 0.3 inches.   RPM was chosen to be 500 with a 
traverse rate of 1 inch/min.    Five runs were conducted.   The complete data set is provided in 
Appendix F.   Average values are depicted in Table 23 below.  Figure 46 shows the difference in 
temperature between copper and aluminum at different points. 
Parameter 
Sh: 
Material 
Ad: 
Shoulder 
Re: 
Shoulder 
Ad: Pin Re:Pin 
0.5"Sh/1"min/500RPM Copper 235.67 224.63 231.71 205.35 
0.5"Sh/1"min/500RPM Aluminum 211.11 204.18 209.35 187.24 
 
Table 23: Raw experimental data: Average of peak temperatures of Entry 2,  
 
Group 1, 2, 3, and 4; All temperatures are in Celsius. 
  
 88
 
Figure 46:   Shoulder and pin temperature of copper and aluminum  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 89
 
 
 
 
 
IX. SIMULATION OF AN ASYMETRIC THERMAL MODEL 
 
The physical results of Chapter 8 clearly documents that the advancing side will always be the 
hotter side.  Only one model in the literature predicts the asymmetrical thermal field around the 
tool (Payton [36]).  This chapter documents an attempt to simulate the physical results obtained 
with a readily available, inexpensive finite element modeling software (COMSOL) using the 
metal cutting model first proposed by Payton [36].  
 Historically, simulation meant the modeling of one physical phenomenon at a time, such 
as structural integrity, electromagnetic compatibility or aerodynamic behavior. In reality, though, 
a component?s behavior can depend on a number of physical phenomena combined with each 
other. Today, thanks to the increasing computer power available for engineers and scientists, 
simulating multiphysics phenomena is easy, accessible, and affordable using personal computer 
software such as COMSOL (formerly FEMLAB).  COMSOL is based upon the widely available 
MATLAB family of computational products. It is not as powerful as traditional mainframe 
simulation software (ANSYS, NASTRAN, LS-DYNA, DEFORM), but it is easily used and 
inexpensive.  Complete screen shots of the software, including equation setups used in this 
simulation, are provided in Appendix F at the end of this dissertation.     
 90
Original Asymmetric Temperature Model of Friction Stir Welding 
 
 None of the earlier thermocouple measurements by researchers in the literature review 
observed a well behaved difference between the advancing and retreating sides.  Perhaps as a 
consequence of this, only one of the researchers has modeled the FSW process as an asymmetric 
thermal process.  Modelers need data in order to formulate realistic models. Simulators need 
large amounts of data in order to validate their simulations of the various models. 
 Only one model in the literature review predicted the asymmetric result observed during 
data collection above.  Payton [36] equates the advancing side and retreating sides of FSW with 
what classical machining theory in general refers to as up milling and down milling.    On the 
leading (up-milling side) side, as the tool enters the work, tool tip velocity (V) assumes a 
maximum value, which decreases as the tool progresses along the tool path.  On the trailing side, 
velocity (V) has a minimum value when the tooth leaves the work and a somewhat higher value 
when the tool enters the work.  The tool edge traces out a looped tracheiod path through the 
material.   Incorporating the eccentric geometry and travel features of the tool, and generating the 
proper coefficients to use the data in ASM Handbook of Metals (9th Edition), Payton generated 
the following useful expressions for the FSW tool designer in aluminum alloy 6061-T6 which 
predict a hot advancing side and a colder retreating side using the following equations. 
                                        
*
[16,148.58]
*( )
s
Vt
THP
kpc
=
                            
            (31) 
HP
s
 is the specific horsepower for the alloy or workpiece, k is the coefficient of thermal 
conductivity for the workpiece, (pc) is the volume specific heat of the work material, V is the 
cutting speed at the tip of the tool.  This is a geometric consideration which varies between left 
and right side (equation 32). 
 91
          
2
2( )2VnRrr?=??y                                                (32) 
One must use a (+) sign in equation (2) for the up-milling side and a (-) sign for the down-
milling side of the geometry.  The variable t is the uncut chip thickness, generally given by the 
following equation: 
12
*
*
cos 1 *
*
t
avg
t
Fd
t
Fnd
RD
RD?
?
=
?+ ?
??
??
??
d
                                     
(33) 
Where, R is the Radius of the cutting edge (specific to shoulder or pin in this application), d is 
the depth of the cut (specific to the design of tool), D is the diameter of the cutter, n is the 
number of teeth, and F
t
 is the feed per tooth.   These equations predict the results of 
thermocouple measurements in the laboratory to date very exactly, since the velocity term V 
captures the difference between the advancing and retreating sides.    The availability of a large 
amount of data over different machining values and modern multiphysics based finite element 
analysis using equations 31 through 33 predict simulate the results accurately.  Table 24 
summarizes the material and machining parameters associated with equations 31 through 33.  
English units are used throughout because of the predominance of those units in the applicable 
literature sources on milling and metal cutting. 
 
 
 
 
 
 
 
 92
Symbol Nomenclature Unit 
HP
s
 Specific horse power hp/in^3/min 
k thermal conductivity of W/P BTU/ft.h.F 
pc Volume specific heat of W/P BTU/lb.F 
R Radius of shoulder inch 
D Diameter of shoulder inch 
t Average chip thickness inch 
Ft Feed per tooth for shoulder inch 
y depth of cut inch 
n RPM RPM 
r Radius of pin inch 
V Cutting speed inch/sec 
 
Table 24:  Material and machining variables used in equations (1) through (3).  
 
Modification to the Model 
In equation (31), the constant 16,148.58, was derived by Payton for specific alloys in use at 
Marshall Space Flight Center where his research was conducted.  It is necessary to adjust this for 
the current instrument. As Chao points out [9], the maximum temperature increases with 
decreasing feed rate, the constant used is replaced by a variable that best matches the recorded 
temperature within the instrument.  In order to find the variable, two data runs were conducted to 
determine the endpoints.  Table 25 summarizes the parameters and results. 
Recorded Temperature, C Parameter 
Advancing side 
900 shoulder/4?min/450 RPM 378 
900 shoulder/7?min/450 RPM 330 
 
Table 25:  Recorded temperatures to find the variable in heat equation 
 
 
 93
The variable must meet three conditions: 
          Condition 1: Temperature decreases with increasing Traverse speed. 
         Condition 2: At 4 IPM traverse, 450 RPM, a 0.9 inch shoulder produces 378 C.                                            
         Condition 3: At IPM traverse, 450 RPM, a 0.9 inch shoulder produces 330 C.  
In equation (31), let?s take the constant as ?X?. 
)(*
*
][
pck
tV
HPXT
S
=                                               (36)                   
From the data obtained, feed rate (F
d
) is inversely proportional to the Temperature. 
Therefore, 
d
F
X
1
?                                                              (35)     
b
F
a
X
d
+=                                                           (36)     
For parameter 1: 4?min/450 RPM/ 900 shoulder 
)(*
*
][
pck
tV
HPb
F
a
T
S
d
+=  
                                              
)(*
*
][378
pck
tV
HPb
F
a
S
d
+=                                            (37)                               
Where, F
d
 = 4 
For parameter 2: 7?min/450 RPM/ 900 shoulder 
)(*
*
][
pck
tV
HPb
F
a
T
S
d
+=  
                                              
)(*
*
][330
pck
tV
HPb
F
a
S
d
+=                                            (38)                      
Where, F
d
 = 7 
 94
Solving equations (37) and (40), 
a = 170000; b = 10000 
Substituting a and b in equation in the heat equation, 
                                          
)(*
*
]10000
170000
[
pck
tV
HP
F
T
S
d
+=                                     (39) 
Equation (39) was used in the simulation for both shoulder heat and pin heat. Height of the pin is 
not considered for the heat generated by the pin. 
PHYSICS BASED SIMULATION 
Equation 39 provides the heat input at both the pin and shoulder surfaces.  
 
Figure 47: Advancing (top) and retreading (bottom) outer edge of the shoulder. 
Simulated advancing (385 C) and retreating (381 C) temperatures for 0.9? shoulder at 
450 RPM and 4 ipm.  
 95
As Song summarizes in his paper [11], most simulation efforts only attempt to model the heat 
input from the tool shoulder, which is a very great simplification.   Modern physics based 
software allows the application of physics based expressions such as equations 41 directly to the 
contact surfaces of the pin and tool.  This model worked well for temperatures under the 
shoulder at the shoulder thermocouple.  For example, at thermocouples S7 and S8, the agreement 
was almost within the overall error of the instrument (+/- 5.0 degrees).   It did not work well for 
the pin tool though since COMSOL does not allow the modeler to include combined thermal 
effects from the shoulder (above) and the pin (below and to the side).   Actual temperatures are a 
combined result of those two heat sources.  Also, increasing the MESH density might improve 
the results below. 
RPM 450, 4 IPM, 0.9 inch
Simulated Actual 
Shoulder Advancing 183.2 Celsius 188.2 Celsius 
Shoulder Retreating 169.3 Celsius 174.1 Celsius 
 
Table 26:  Comparison of simulated results to actual results AT THE THERMAL WELL. 
 
Meshing Conditions: 
The following meshing conditions are for the model with 0.9 inch shoulder, 450 RPM with 
4?/min traverse. 
Meshing 
Conditions # of elements 
Peak Temperature 
(Celsius) 
Coarser 11351 390.85 
Coarse 17281 387.85 
Normal 31525 385.85 
Fine 49215 380.85 
 
Table 27: Number of elements generated for default meshing conditions. 
 
The above table shows the number of elements produced for four different meshing conditions. 
 96
The table also includes the peak temperature obtained when each condition was used in the FEA 
tool. It was noted that there was no big difference between the last 3 meshing conditions, 
?Course?, ?Normal? and ?Fine?. 
 
Figure 48: Peak temperature Vs number of elements in meshing. 
The graph above shows the peak temperature obtained for different ?number of elements? 
generated during meshing process.  For this model, NORMAL meshing was used and the 
statistics of the meshing option is given below. 
Actual Meshing Values: 
A default meshing condition was used for the simulation. NORMAL mesh was used in 
the model which has 31525 elements. The elements are of tetra, hexahedral and prismatic shapes. 
Number of degrees of freedom was 45894. There were total of 6240 mesh points and 9477 
boundary elements. Increasing the mesh to ?FINE? would have improved the computational 
values of Table 27. 
 
 97
Table 28 gives the simulated temperature at the point where the thermocouples were placed for 
the experiment and table 29 gives the actual temperature recorded at Group 2 (S7, S8, P7 and P8) 
thermocouples. 
  Simulation at thermocouple 
Factor level combination
Ad: 
Sh 
Re: 
Sh 
Ad: 
Pin 
Re: 
Pin 
0.9"Sh/4"min/450 RPM 183.2 169.3 102.4 101.0 
0.9"Sh/4"min/350 RPM 167.5 160.3 96.7 93.4 
0.9"Sh/7"min/450 RPM 125.8 118.9 65.3 60.2 
0.9"Sh/7"min/350 RPM 118.6 111.1 57.8 52.3 
0.7"Sh/4"min/450 RPM 166.3 156.4 95.5 90.2 
0.7"Sh/4"min/350 RPM 143.0 131.2 72.3 64.3 
0.7"Sh/7"min/450 RPM 128.5 120.9 67.4 62.9 
0.7"Sh/7"min/350 RPM 117.8 105.2 55.3 51.1 
0.5"Sh/4"min/450 RPM 128.0 115.1 67.7 59.0 
0.5"Sh/4"min/350 RPM 105.0 95.1 48.3 46.2 
0.5"Sh/7"min/450 RPM 109.7 103.0 49.6 48.2 
0.5"Sh/7"min/350 RPM 78.6 69.2 48.3 47.0 
 
Table 28: Simulated thermocouple temperature in Celsius 
 
 
Direct temperature at thermocouple: Group 2 
Factor level 
Combination 
Ad: 
Sh 
Re: 
Sh 
Ad: 
Pin 
Re: 
Pin 
0.9"/4"min/450 RPM 188.2 174.1 186.4 176.0 
0.9"/4"min/350 RPM 178.3 161.8 182.5 168.9 
0.9"/7"min/450 RPM 144.3 133.8 143.6 133.4 
0.9"/7"min/350 RPM 144.2 130.7 143.1 133.4 
0.7"/4"min/450 RPM 160.8 146.0 163.2 147.9 
0.7"/4"min/350 RPM 156.6 144.7 157.7 141.8 
0.7"/7"min/450 RPM 141.0 132.0 140.5 132.8 
0.7"/7"min/350 RPM 138.3 128.6 137.8 127.8 
0.5"/4"min/450 RPM 145.6 127.2 153.0 138.7 
0.5"/4"min/350 RPM 134.6 121.1 135.3 123.0 
0.5"/7"min/450 RPM 136.6 125.9 140.5 129.0 
0.5"/7"min/350 RPM 108.8 95.1 110.6 96.5 
 
Table 29: Direct Temperature at Group 2 thermocouples in Celsius 
 
 98
Table 30 gives the temperature calculated manually using the Payton?s equation. The values 
represent the temperature at the tool-workpiece interface. Table 31 gives the simulated tool-
workpiece interface temperature. 
Factor level combination Ad: Sh Re: Sh Ad: Pin Re: Pin 
          
0.9"Sh/4"min/450 RPM 382.49 381.17 194.21 191.44 
0.9"Sh/4"min/350 RPM 359.59 358.35 169.18 166.74 
0.9"Sh/7"min/450 RPM 308.53 306.65 166.19 161.99 
0.9"Sh/7"min/350 RPM 287.45 285.7 144.46 140.77 
0.7"Sh/4"min/450 RPM 357.93 356.33 194.21 191.44 
0.7"Sh/4"min/350 RPM 336.45 334.93 169.18 166.74 
0.7"Sh/7"min/450 RPM 307.11 304.68 166.19 161.99 
0.7"Sh/7"min/350 RPM 288.53 286.24 144.46 140.77 
0.5"Sh/4"min/450 RPM 307.6 305.61 194.21 191.44 
0.5"Sh/4"min/350 RPM 286.59 284.73 169.18 166.74 
0.5"Sh/7"min/450 RPM 280.68 277.5 166.19 161.99 
0.5"Sh/7"min/350 RPM 263.62 260.61 144.46 140.77 
 
Table 30: Temperature calculated using Payton?s equation (in Celsius). 
Factor level combination Ad: Sh Re: Sh Ad:Pin Re: Pin 
          
0.9"Sh/4"min/450 RPM 385.16 381.13 320.79 315.76 
0.9"Sh/4"min/350 RPM 355.72 352.15 301.97 298.4 
0.9"Sh/7"min/450 RPM 305.22 302.57 252.39 249.74 
0.9"Sh/7"min/350 RPM 284.35 282.55 231.33 229.53 
0.7"Sh/4"min/450 RPM 353.33 349.28 294.52 290.47 
0.7"Sh/4"min/350 RPM 333.68 330.53 279.89 276.73 
0.7"Sh/7"min/450 RPM 305.08 302.79 250.36 248.08 
0.7"Sh/7"min/350 RPM 285.44 282.03 228.02 224.6 
0.5"Sh/4"min/450 RPM 304.37 301.31 247.93 244.86 
0.5"Sh/4"min/350 RPM 282.85 280.27 229.74 227.16 
0.5"Sh/7"min/450 RPM 276.85 272.99 223.56 219.7 
0.5"Sh/7"min/350 RPM 260.41 257.32 203.87 200.77 
 
Table 31: Simulated temperature at tool-workpiece interface (in Celsius) 
 
 99
 
 
 
 
X. DISCUSSION 
 
The apparatus had seven (7) sets of thermocouples.  Paired shoulder and pin thermocouples 
captured the advancing and retreating temperatures of the shoulder and the pin. A fifth 
thermocouple in each set captured the temperature at the bottom of the pin.  The first set captured 
the temperature during the entry of the tool and the last set captured the temperature during the 
exit of the tool.  The middle sets captured the temperature during the transient action of the tool.   
Figure 49 shows the temperature distribution during the process. 
 
Figure 49: Temperature distribution during the process 
 It is to be noted that in all the experiments and for all the factor level combinations, the 
entry set was hotter than any other set and the exit set was cooler than any other set.  The highest 
temperature in the process is obtained when the tool moves in to the work piece.  This is the 
 100
result of the drilling action of the pin and the shoulder in the work piece at the same position; i.e. 
the dwell of the tool over the entry point allows the heat to build underneath the rotating tool 
during entry. 
            Thermal field mapping was done for the process.   The thermocouples used, captured the 
temperature of the advancing shoulder, retreating shoulder, advancing pin, retreating pin and 
under the pin.  The temperatures recorded were the temperature of the steel holder at fixed points 
away from the ?ideal geometrical points?.   One dimensional steady state thermodynamic 
equations were used to find the temperature at the tool tip.   These equations do not consider the 
heat dissipated in the second or the third dimensions.   But in real process, heat is lost to the 
atmosphere, the metal holder and base plate.  The calculated temperature does give an 
approximate value at the tool shoulder tip, the pin radius and the area immediately under the pin. 
             The temperatures recorded in all the experiments and the statistical test done to compare 
the difference in the peak temperatures of the advancing and the retreating sides of each pair of 
thermocouples (S1&S2, S3&S4?..) conclusively demonstrates that the advancing temperature is 
clearly higher than the retreating temperature.  For all the advancing-retreating pairs, the test 
results document that the advancing side was always hotter than the retreating with a statistical 
power of at least 90% for all the thermocouple pairs.  
             From the temperature plots that were obtained for different factor level combinations, it 
is very clear that the temperature is higher at the entry point because of the temperature dwell.  
The temperature then stabilizes as the shoulder is fully in contact with the workpiece and stays 
the same through out the process till the exit set.  The last set called the exit set sees a decrease in 
the temperature. This is because of the tool getting out of the workpiece above that set without 
the shoulder traversing through it completely.    The back half of the tool never transits over the 
 101
thermocouple set, therefore the peak temperature never occurs.   Future designs should pass at 
least two diameters of the maximum shoulder beyond the last thermocouple set. 
 
Figure 50: Bad Thermocouple in the experiment.  
             In most of the factor level combination, the plot having the average of the peak 
temperatures of the thermocouple clearly shows that the Group 4 retreating pin thermocouple 
(P11) temperature is always lower than the others (as in Figure 50).   This difference was not 
immediately apparent to the author until analysis was in progress.   Data collection happens 
quickly with this instrument, even when doing 84 weld runs. 
           For the temperatures recorded, analysis of variance was done to check the most influential 
factor on the responses for different factor level combinations.  For any factor level combination, 
the size of the shoulder had the largest impact on the entry thermocouples (S1, S2, P1, P2 and 
U1).  All other thermocouples sets are most affected by the traverse speed.  This indicates that 
more than the size of the shoulder, the traverse speed of the tool plays a vital role in the 
 102
)(*
*
][
pck
tV
HPXT
s
= recrystallization of the metal.    One interesting finding in this analysis is 
that the RPM does not have as large a statistical effect comparatively, on any of the 
thermocouple.   The results also indicate that the shoulder size and the RPM are directly 
proportional to the temperature.  As the shoulder size and the RPM goes up, the temperature of 
the work piece goes up.  But, the traverse speed is inversely proportional to the temperature of 
the work piece.  As the traverse speed goes up, the temperature goes down.   This is consistent 
with Choe?s observations [9], [22] and [23]. 
             In the literature, there is no model which reflects two important facts of the process as 
documented in these thermal data runs.  One is the asymmetry of the temperature field and the 
other is the inverse proportionality of the traverse speed.  No heat equation is modeled to adapt 
these two conditions of the process.    A new model based on Payton?s metal cutting model was 
presented and simulated here with good results.    These results might be in even better 
agreement with more powerful computers or finite element modeling software. 
            The shoulder experiment conducted had expected results.  The temperature of any 
thermocouple increased with the increasing shoulder size of the tool.   The graph plotted was 
more or less linear, steadily increasing with increasing tool size.  Earlier, from the analysis of the 
variance of the experiments, we concluded that traverse speed had more effect on the 
thermocouples than the shoulder except at the entry set of thermocouples.  Since increasing the 
shoulder size of the tool increases the temperature of the workpiece more or less linearly, the 
author expects the temperature to increase more steadily and linearly with increasing traverse 
speed. Figure 51 demonstrates an interesting observation from the data.   The peak temperature 
at a thermocouple is not obtained when the tool is exactly on top of it.   
 103
 
Figure 51: Shoulder temperature with respect to the position of the tool. 
Figure 51 shows the temperatures of S5 and S6 thermocouples from the start to the end of the 
experiment.  It is very clear that the peak temperature is obtained only after a certain period after 
the tool (trailing end of the tool) has gone past the thermocouples. To be specific, the peak 
temperature is obtained only 18 seconds after the trailing edge of the tool has gone past the 
thermocouples for the factor level combination involved. 
          The heat equation used in the model gives the temperature of the advancing shoulder, 
retreating shoulder, advancing pin and retreating pin.  The temperature of the advancing and 
retreating shoulder, more or less matches with the thermodynamically calculated values.  The 
temperature of the advancing and retreating pin is much less than the thermodynamically 
calculated temperatures.  When simulated using Multiphysics software, the shoulder 
temperatures were not exactly equal to the calculated temperatures but close to it.  The reason for 
the temperature of the pin when calculated from the heat equation being much less than the 
 104
simulated or thermodynamically calculated temperature is that the heat equation does not 
consider the effect of shoulder for pin temperature.    It is possible also that instead of a milling 
equation, an extrusion equation should be utilized in the area of flow around the pin, or perhaps a 
combination of equations as Choe attempted to do [9], [22] and [23]. 
Parameterization using tool geometry and material properties 
 The thermal model described within this paper incorporates the material properties of the 
alloys being welded and tool designs geometric properties and path.   Although the major 
experiment conducted by Kandaswaamy [37] utilized the very well known results for aluminum 
alloy 6061-T6 extensively, the researcher validated the sensitivity of the instrument to both 
material and tool changes using copper alloys and aluminum together along with various tool 
geometries, RPM?s and feed rates across the welds. 
  The model and simulation capture the ?dwell? of the tool?s motion as it passes through 
the material since the shoulder dominates the final passage over the welded junction.   It 
correctly captures the transitions during insertion and initial shoulder contact.   Most importantly, 
the asymmetrical thermal difference on the two sides of the tool, and the variation of the 
temperature from the shoulder down along the pin are accurately predicted by the simulation of 
this model presented. 
  A model such as this lets the operator chose from the classic milling conventions (feed 
per tooth, RPM, velocity) in order to ?target? a sweet spot or temperature range within the alloy 
of concern.    Although the data collected during the initial experiments involved the simple 
geometry of a ?straight? pin with no shoulder ridges or flutes, it is only a straightforward 
application of conventional milling theory to incorporate the cooling effects of flutes and scrolls 
in the design of the tool. 
 105
 
 
 
XI. CONCLUSIONS 
 
A detailed review was conducted on the previous thermal measurements and thermal models by 
various researchers.  An instrument that measures the temperature of the work piece around the 
tool was designed and built.  A full factorial, detailed observation of the thermal fields was made 
with the apparatus with the following results: 
? The highest temperatures of the friction stir welding process occur during the relatively 
long insertion into the specimen.    Once the transit starts, temperatures fall and reach as 
lower steady state transient temperature.  For aluminum 6061-T6, this occurs within 2 
diameters of the tool shoulder (on average). 
? The advancing side of the tool is always hotter than the retreating side of the tool, with a 
statistical power of at least 90% for all the experiments and factor level combinations 
done.   
? The temperature always peaks following passage of the trailing edge of the shoulder.   
This is an important consideration in applications where a run out tab is not used.    The 
weld properties at the extraction point may be very different because of the lower 
temperatures at that point. 
? The pin, well away from the shoulder, often exceeds the maximum temperature at the 
shoulder, suggesting that extrusion around the pin has a much greater heating effect than 
the rotation or radius of the shoulder. 
 106
? The rate of traverse dominated temperature responses by a large statistical margin.    
Shoulder diameter was an important, but not close, second in most cases.   RPM, to the 
surprise of the investigator, was a very distant third. 
? Temperature rose linearly as RPM and shoulder diameter increased. 
? Temperature decreased inversely as the traverse speed increased. 
? Temperature changed with material being welded, all other things constant. 
? An extensive body of thermal data is now available upon request. All thermal data is 
available from the principal advisor to modelers and simulators, along with exact 
geometry of the holder and material properties. 
? An instrument exists which can rapidly collect more data using standard commercially 
available geometries.   An upgrade to the instrument is planned when access to a 5 axis 
CNC mill is possible. 
? Payton?s model, being the only model in the literature which describes the asymmetric 
nature of the temperature field, was used to simulate the process using a multi physics 
finite element analysis tool. The results obtained were compared to the real world data for 
the shoulder with excellent results.  Software limitations precluded a good simulation of 
the pin tool area. 
? Classic metal cutting theory from the 1950s and 1960s has been successfully applied to a 
joining process.   This suggests that classic extrusion theory and software such as 
DEFORM 3D might also be very beneficially applied to Friction Stir Welding. 
 
 
 
 
 107
 
 
 
 
 
REFERENCES 
 
1. Lee, W.B., Y.M. Yeon, and S.B. Jung, Evaluation of the microstructure and mechanical 
properties of friction stir welded 6005 aluminum alloy. Materials Science and 
Technology, 2003. 19(11): p. 1513-1518. 
2. Murr, L.E., et al., Friction-stir welding: Microstructural characterization. Materials 
Research Innovations, 1998. 1(4): p. 211-223. 
3. Jata, K.V. and S.L. Semiatin, Continuous dynamic recrystallization during friction stir 
welding of high strength aluminum alloys. Scripta Materialia, 2000. 43(8): p. 743-749. 
4. Murr, L.E., G. Liu, and J.C. McClure, Dynamic recrystallization in friction-stir welding 
of aluminium alloy 1100. Journal of Materials Science Letters, 1997. 16(22): p. 1801-
1803. 
5. Murr, L.E., et al., Intercalation vortices and related microstructural features in the 
friction-stir welding of dissimilar metals. Materials Research Innovations, 1998. 2(3): p. 
150-163. 
6. James, M.N., et al., The relationship between process mechanisms and crack paths in 
friction stir welded 5083-H321 and 5383-H321 aluminium alloys. Fatigue & Fracture of 
Engineering Materials & Structures, 2005. 28(1-2): p. 245-256. 
7. Tang, W., et al., Heat input and temperature distribution in friction stir welding. Journal 
of Materials Processing & Manufacturing Science, 1999. 7(2): p. 163-172. 
 108
8. Chen, C.M. and R. Kovacevic, Finite element modeling of friction stir welding - thermal 
and thermomechanical analysis. International Journal of Machine Tools & Manufacture, 
2003. 43(13): p. 1319-1326. 
9. Chao, Y.J., X. Qi, and W. Tang, Heat transfer in friction stir welding - Experimental and 
numerical studies. Journal of Manufacturing Science and Engineering-Transactions of the 
Asme, 2003. 125(1): p. 138-145. 
10. Khandkar, M.Z.H., J.A. Khan, and A.P. Reynolds, Prediction of temperature distribution 
and thermal history during friction stir welding: input torque based model. Science and 
Technology of Welding and Joining, 2003. 8(3): p. 165-174. 
11. Song, M. and R. Kovacevic, Thermal modeling of friction stir welding in a moving 
coordinate system and its validation. International Journal of Machine Tools & 
Manufacture, 2003. 43(6): p. 605-615. 
12. McClure, J.C., et al. A thermal model of friction stir welding in Trends in Welding 
Research, Proceedings, 5th International Conference 1998. Pine Mountain, GA; USA 
13. Dong, P., et al., Coupled thermomechanical analysis of friction stir welding process 
using simplified models. Science and Technology of Welding and Joining, 2001. 6(5): p. 
281-287. 
14. Khandkar, M.Z.H. and J.A. Khan, Thermal modeling of overlap Friction Stir Welding for 
Al-alloys. Journal of Materials Processing & Manufacturing Science, 2001. 10(2): p. 91-
105. 
15. Lienert, T.J., W.L. Stellwag, and L.R. Lehman. Heat Inputs, Peak Temperatures, and 
Process Efficiencies for FSW. in 4th International Friction Stir Welding Symposium. 
2003. Park City, Utah: TWI. 
 109
16. Lambrakos, S.G., et al., Analysis of friction stir welds using thermocouple measurements. 
Science and Technology of Welding and Joining, 2003. 8(5): p. 385-390. 
17. Xu, J., et al. Experimental and Numerical Study of Thermal Process in Friction Stir 
Welding. in Joining of Advanced and Specialty Materials VI. 2003. Pittsburg, Pa: ASM 
International. 
18. Ulysse, P., Three-dimensional modeling of the friction stir-welding process. International 
Journal of Machine Tools & Manufacture, 2002. 42(14): p. 1549-1557. 
19. Chang, W.S., et al., Joint properties and thermal behaviors of friction stir welded age 
hardenable 6061Al alloy, in Thermec'2003, Pts 1-5. 2003, TRANS TECH 
PUBLICATIONS LTD: Zurich-Uetikon. p. 2953-2958. 
20. Lawrjaniec, D., et al., Numerical simulation of friction stir welding, in Thermec'2003, Pts 
1-5. 2003, TRANS TECH PUBLICATIONS LTD: Zurich-Uetikon. p. 2993-2998. 
21. Liu, S., Y.J. Chao, and C.H. Chien. Study of thermal and heat transfer phenomenon in 
friction stir welding of aluminum alloy 6061-T6 thick plates in SAE 2004 World Congress 
and Exhibition. 2004. Detroit, Michigan, USA Society of Automotive Engineers, Inc., 
Warrendale, Pennsylvania, USA. 
22. Chao, Y.J. and X.H. Qi, Thermal and thermo-mechanical modeling of friction stir 
welding of aluminum alloy 6061-T6. Journal of Materials Processing & Manufacturing 
Science, 1999. 7(2): p. 215-233. 
23. Chao, Y.J. and X. Qi. Heat Transfer and Thermo-Mechanical Analysis of Friction Stir 
Joining of AA6061-T6 Plates. in 1st International Symposium on Friction Stir Welding. 
1999. Thousand Oaks, Ca: TWI. 
 110
24. Schmidt, H., J. Hattel, and J. Wert, An analytical model for the heat generation in friction 
stir welding. Modelling and Simulation in Materials Science and Engineering, 2004. 
12(1): p. 143-157. 
25. Awang, M., et al. Thermo-Mechanical Modelling of FrictionStir Spot Welding (FSSW) 
Process: Use of an Explicit Adaptive Meshing Scheme. in SAE World Congress. 2005. 
Detroit, Michigan, USA: Society of Automotive Engineers, Inc., Warrendale, 
Pennsylvania, USA. 
26. Colegrove, P.A. and H.R. Shercliff, Experimental and numerical analysis of aluminium 
alloy 7075-T7351 friction stir welds. Science and Technology of Welding and Joining, 
2003. 8(5): p. 360-368. 
27.   Tay, A.O, ?A Numerical Study of the Temperature distribution generated during 
Orthogonal Machining?, Thesis, University of New South Wales. 
28.    Klamecki, B.E, ?Incipient Chip Formation in Metal Cutting ? A 3D Finite Element 
Analysis?, Thesis, University of Illinois. 
29.      Stevenson, M.G, Wright P.K and Chow J.C, ?Further Developments in Applying the FEM 
to the calculation of Temperature Distributions in Machining and Camparisons with 
Experiment?, Journal of Engineering for Industry , Trans. ASME, pp. 149-154. 
30.    Abdel-Hamid, ?A three dimensional finite element thermo-mechanical analysis of  
intermittent cutting process?, Dept of Mechanical Design and Production, Cairo 
University. 
31.     Shore, H, Tool and chip temperatures in Machine shop Practice, M.S Dissertation, M.I.T 
32.       Shaw, M.C, Metal cutting Principles, Oxford University Press, 594(1984). 
33.       Herbert, E.G. (1926), Proc. Instn mech. Engrs 1, 289. 
 111
 112
34.       Schwerd, F. (1937). Z.des Ver. Deutsche Ing. 77, 211. 
35.     Martelloti, M.E, An analysis of the Milling process, Part II-Down Milling, Trans ASME, 
233(1945). 
36.    Payton, L.N, Metal Cutting theory and Friction stir welding, Industrial and  systems 
Engineering, Auburn University. 
37.   Payton, L.N, Thermal field mapping of Friction stir welding process, Mechanical 
Engineering, Auburn University. 
 
 
 
 
 
 
 
 
 
 
APPENDICES 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
112 
 
APPENDIX A 
Direct Data: With 2 RPMs, 2 traverse speed and 3 shoulder diameters, there is a combination of 
12 different parameters. For each parameter, 7 experiments were done to achieve at least a 90% 
statistical power that advancing was hotter than retreating side. The peak temperatures for all the 
12 different parameters are given below. 
Nomenclature: Each Thermocouple was represented by a unique alphabet and number: 
  S ? Shoulder thermocouple. (Retreating side: S
1
, S
3?
 S
13;
  
       Advancing side: S
2
, S
4...
 S
14
) 
             P ? Pin thermocouple. (Retreating side: P
1,
 P
3?
 P
13;
 Retreating side: P
2
, P
4
? P
14
) 
           U ? Under the pin thermocouple. (U
1,
 U
2?
 U
7
) 
Factor level combination 1: 450 RPM / 7 inch per min traverse / 0.9 inch shoulder  
Entry data: 
 Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 177.5 189.01 146.8 150.1 191.3 197 123.2 124.5 203 139.2 
RUN 2 176.5 180.38 156.3 165.6 188.5 189.4 130.2 136.6 208.8 147.9 
RUN 3 191.6 198.02 146.2 153.9 207.7 216.3 136.7 140 221.2 140.8 
RUN 4 181.4 193.49 146.2 155.2 195.7 208.8 135.6 140.9 207.6 147.1 
RUN 5 183.2 186.77 144 157.5 191.7 195 131.3 145.6 210.9 153.6 
RUN 6 182.1 184.16 147.8 155 199.2 203.3 127.7 133.1 212.3 142.3 
RUN 7 189.4 193.29 150.5 157.8 198.4 205.1 132.2 134.7 217.8 136.7 
 
 
(450 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
113 
 
Traverse data: 
 
 Group 1 Group 2 Group 3 Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 127 147 130.5 133.4 141 145 139 141 
RUN 2 137.5 151.3 135.9 147.8 147 148 143 147 
RUN 3 145.8 148.7 135 140.7 141 152 132 142 
RUN 4 145 147.4 133.6 147.4 144 158 135 146 
RUN 5 145.1 145.2 135 144.2 142 156 130 144 
RUN 6 143.5 145 132.3 136.8 143 144 139 142 
RUN 7 146.4 148.1 134.1 139.8 144 147 140 140 
 
 
 Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 131.2 143.7 127 139.9 136 136 98.6 142 
RUN 2 143.2 148.4 141.4 146.3 140 144 102 150 
RUN 3 142.2 153.5 128.8 141.8 137 148 79.1 143 
RUN 4 141.2 144.5 133.1 144.8 138 141 82 141 
RUN 5 148.4 152.3 144.6 150.1 136 150 88.1 154 
RUN 6 140 141.6 132.9 140.6 139 140 100 146 
RUN 7 145.1 146 136.2 141.7 141 142 101 147 
 
?
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 136.1 132.4 126.9 132.1 
RUN 2 140.6 142.7 130.3 134.8 
RUN 3 133.1 145.1 132 132.9 
RUN 4 131.7 141.6 135.4 134.8 
RUN 5 139.8 149.7 124.5 141 
RUN 6 131.7 137.4 128 133 
RUN 7 134.9 140.1 136.3 134.3 
 
(450 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
114 
 
Exit data: 
 Exit Exit Exit 
 S13 S14 P13 P14 U7 
RUN 1 110.6 121.6 118.5 121.1 104 
RUN 2 117.1 128.9 126 133.6 108 
RUN 3 100 117.7 97.81 131.3 104 
RUN 4 117.4 119.1 86.5 132.6 105 
RUN 5 118.8 112.1 105.4 128.8 119 
RUN 6 109.1 115.2 112.9 117 106 
RUN 7 112.7 127.4 119.9 125.4 110 
 
(450 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
115 
 
Factor level combination 2: 450 RPM / 4 inch per min traverse / 0.9 inch shoulder 
Entry data 
 Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 184.7 198.9 193.7 199.2 198 224 162 165 228.8 177.7 
RUN 2 209.6 222.5 203.9 205.1 226 239 171 173 235.5 183 
RUN 3 202.4 212.4 205.1 211.1 218 229 173 176 232.8 186.9 
RUN 4 199 199.1 199.6 202.7 214 223 171 181 221.3 171.1 
RUN 5 202.3 209.3 197.1 204.3 214 219 168 171 233.2 187.3 
RUN 6 193.8 211.5 193.3 203.8 214 215 167 172 228.7 178.4 
RUN 7 197.5 207 195.8 205.7 213 214 168 173 226.7 172.1 
 
Traverse data: 
 
 Group 1 Group 2 Group 3 Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 183.5 191.7 163.3 182 184 188 179 185 
RUN 2 188.9 203.1 177.8 192.5 199 200 187 193 
RUN 3 202.9 208.7 180.6 197.8 201 202 189 198 
RUN 4 190.7 203.7 178.1 179.3 186 190 188 193 
RUN 5 190 196.1 171.8 179.7 188 189 184 188 
RUN 6 182.1 190.8 171.8 172.9 191 193 179 191 
RUN 7 199.1 201.2 175.4 183.6 188 190 188 199 
 
 Group 1 Group 2 Group 3 Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 180 188 175.4 181.2 177 180 125 193 
RUN 2 188.4 199.3 187.4 197.1 193 193 132 204 
RUN 3 192.3 204.7 189.9 200.3 193 195 132 206 
RUN 4 180.3 199.4 178.1 186 181 193 124 197 
RUN 5 187.3 192.2 175.4 183.7 181 183 129 190 
RUN 6 180.5 188.1 170.1 186.5 185 186 127 196 
RUN 7 180.4 192.9 175.5 189.8 180 193 123 196 
 
(450 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
116 
 
  
Group 
1 
Group 
2 
Group 
3 
Group 
4 
 U3 U4 U5 U6 
RUN 1 175.3 171.8 167.9 172.3 
RUN 2 183.8 176.4 170.2 178.4 
RUN 3 191.6 186.2 175.1 183.1 
RUN 4 181.1 175.6 179.9 174.5 
RUN 5 180.2 180.1 167.1 178.1 
RUN 6 179.3 185.9 161.4 165.4 
RUN 7 183 190 171.6 178 
 
(450 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
Exit data: 
 Exit Exit Exit 
 S13 S14 P13 P14 U7 
RUN 1 156 171.4 163.7 165.2 151 
RUN 2 160.8 167.5 161.1 170.3 152 
RUN 3 165.8 174.6 167.9 177 153 
RUN 4 155.4 171.2 166.9 174.4 154 
RUN 5 153.1 162.8 156.6 163.7 150 
RUN 6 139.4 157.3 140.3 164.5 134 
RUN 7 156.9 150 166.5 170.2 153 
 
(450 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
117 
 
Factor level combination 3: 350 RPM / 7 inch per min traverse / 0.9 inch shoulder  
Entry data: 
 Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 
RUN 1 189.6 191.6 130.5 140.5 177.5 201.7 128.1 140.5 215.1 
RUN 2 185.7 192 142.6 154.8 174.8 202.7 118.2 126.5 215.7 
RUN 3 197.3 199.5 141.5 152.5 48.05 207.5 125.9 130.3 222.3 
RUN 4 196.3 200.4 148.3 161.2 189 214.9 128.1 136.6 217.5 
RUN 5 186.1 193.8 138 144 40.49 198.6 117.9 120.3 216.8 
RUN 6 196.5 198.6 152 162.6 203.9 211.9 135.2 141.1 224 
RUN 7 188.5 196.7 154.6 163.2 203.8 207.3 136.4 141.2 220.8 
 
Traverse data: 
 
 Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 145.6 146.3 135.5 144.2 136.2 141.8 136.9 141.8 
RUN 2 128.5 140.8 130.1 139.1 137.8 139.7 135.8 136 
RUN 3 139.4 147.1 130.1 137 136.6 145.2 131.6 135 
RUN 4 141.9 148 130 137.3 140.6 144.8 131.5 134.5 
RUN 5 128.5 130.5 117.2 123.9 127 130.7 121.6 123.3 
RUN 6 146.6 147.4 137.3 142.9 144.9 149.3 137.9 144.8 
RUN 7 147.7 147.9 135.1 145.3 143.4 149.5 137.7 138.3 
 
 
 Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 134 145.5 133.8 148 130.9 135.9 97.73 112.9 
RUN 2 131.6 137.8 132.1 137.1 129.9 132.4 96.01 137.7 
RUN 3 137.1 138.7 134.4 136.2 131.4 139.5 98.37 135.5 
RUN 4 137.8 139.6 133.7 135.9 132.2 139.3 97.53 136 
RUN 5 124 126.8 119.4 124.2 121.4 124.1 88.41 127.5 
RUN 6 143.5 145 139 145.1 142.7 143.9 101.4 150 
RUN 7 145.4 147.6 141.5 145.1 140.4 145.1 100.7 144.2 
 
(350 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
118 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 133.2 142.7 121.4 124.9 
RUN 2 135.5 134.8 124.3 128.2 
RUN 3 126.5 142.1 132.5 125.7 
RUN 4 124.4 138.2 126 115.4 
RUN 5 111.3 126.8 118 109.6 
RUN 6 135.9 143.5 139.7 138 
RUN 7 137.6 142.9 138.3 134.3 
 
(350 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
Exit data: 
  Exit Exit Exit 
 S13 S14 P13 P14 U7 
RUN 1 94.95 103.2 96.75 98.94 86 
RUN 2 110.1 123 118.3 119.5 101.5 
RUN 3 98.48 109.4 106.8 106.8 101 
RUN 4 107.9 120 110.1 116.5 97.92 
RUN 5 95.33 103.6 98.32 104.6 87.64 
RUN 6 112.1 117.9 113.8 124.8 109 
RUN 7 108.2 110 104.8 121.2 108.3 
 
(350 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
119 
 
Factor level combination 4: 350 RPM / 4 inch per min traverse / 0.9 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 197.4 206.5 182.5 199.8 208.8 218.5 156.4 171.6 223 164.2 
RUN 2 195.2 204.1 183.3 191.8 210.9 219.3 152.1 156.2 217.3 165.6 
RUN 3 197.6 205.7 198.8 206.8 210.2 220.6 168.8 172 222.1 180.6 
RUN 4 195.3 203 184 193.6 209.8 218.8 171.9 181.9 210.9 178.2 
RUN 5 188.3 195.8 178.8 184.6 197.3 203.4 144.3 168.4 214.5 162.1 
RUN 6 193.9 204 173.1 184.8 208 211.2 151.5 157.2 223.7 169 
RUN 7 197.2 200.1 183.6 195.7 203.7 214.3 159.4 164.2 226.1 177 
 
Traverse Data: 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 185.8 192.1 163 173.4 163.2 178.1 156.9 159.9 
RUN 2 178.3 186.4 160.6 171.2 180.9 182.2 157.8 175.4 
RUN 3 199.6 199.9 172.6 191.7 173.2 193.7 162.9 171.2 
RUN 4 181.5 194.3 179.2 187.6 178 181.3 162.3 174 
RUN 5 180.3 194.4 166.1 171.7 167.1 175.6 141.3 170.5 
RUN 6 174.1 178.3 152.5 177.2 168 172.4 169.7 170.9 
RUN 7 181.8 182.5 158.6 175.7 169.7 170.9 167 172.3 
 
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 179.2 187.9 176.2 182.3 176.3 178.2 127.7 163.8 
RUN 2 174.7 181 174.1 180.8 173.7 174.5 122.7 177.5 
RUN 3 187.3 195 184.9 191.7 186.8 188.5 130.1 169 
RUN 4 171.6 186.5 191.1 197 180.6 194.3 134.9 172.7 
RUN 5 160.7 171 177 180.1 174.1 178.9 122.7 155.7 
RUN 6 171 171.1 165.2 172.4 163.2 168.1 119.7 177.5 
RUN 7 174.5 179.3 163.7 173.3 165.7 176.9 121.3 170.8 
 
(350 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
120 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 154.5 151.6 139.7 154.3 
RUN 2 166.5 161.8 152.4 166.1 
RUN 3 182.1 167.7 157.1 181.1 
RUN 4 180.8 169.8 151 168.8 
RUN 5 169.3 163 143.4 168.3 
RUN 6 160.9 157.2 159 163.9 
RUN 7 168.5 164.2 155.3 164.6 
 
Exit Data:  
  Exit Exit Exit 
 S13 S14 P13 P14 U7 
RUN 1 131.5 139.4 130.9 137.1 125.7 
RUN 2 134 148.2 142.6 142.6 145.9 
RUN 3 141 147.4 140.1 151.6 137.6 
RUN 4 144 155.2 146.6 157.3 141.1 
RUN 5 125.5 130.8 122.6 138.3 123.8 
RUN 6 131 135.6 129.2 143 130 
RUN 7 126.2 149.9 142.1 143.2 135.1 
 
(350 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
121 
 
Factor level combination 5: 450 RPM / 7 inch per min traverse / 0.7 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 168.3 176.7 138.7 159.3 185.1 201.9 126.5 130.9 198.3 134.5 
RUN 2 169 189.1 138.7 148.9 183.8 195.9 126.6 130.3 210.9 139.8 
RUN 3 172.5 184.1 147 148.5 182.9 194.6 126.3 130.3 202.2 136.5 
RUN 4 174.2 187.9 142.1 149.4 183.9 192.8 123.6 128.3 215.6 134.9 
RUN 5 176.6 182.1 147.4 154.6 189.9 196 120.9 134.5 209.8 143.4 
RUN 6 175.6 186.4 143.1 156.4 192.3 195.3 122.9 136.6 210.6 142.8 
RUN 7 174.9 177.8 148.5 154.9 190.3 202.8 120.6 132.8 204.1 139.5 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 134.2 146.7 133.2 147.1 128.2 145.5 126.2 139.7 
RUN 2 135.3 147.9 133.4 139.5 138.6 141.8 127.4 136.4 
RUN 3 138.2 141.2 130.4 137.6 134.2 139.4 129.4 133.7 
RUN 4 131.1 144.2 132.4 141.9 136 138.6 128.9 137.5 
RUN 5 135.6 146.9 135 138.6 133.9 145.1 125.7 136.5 
RUN 6 131.6 139.7 128.3 141.3 134.9 140.3 132.4 137.7 
RUN 7 129.6 144.4 131.2 141.3 137 138.6 132.9 136.6 
 
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 128.1 148.9 130.9 142.8 127.3 141.6 94.4 144.2 
RUN 2 122.8 143.3 137.5 140 125.1 135.8 99.59 143 
RUN 3 135.1 139 135.3 137.7 129.8 141.2 97.6 134.2 
RUN 4 129.8 140.7 132.1 139.5 131.6 133.5 94.63 137.9 
RUN 5 124 145.3 136.7 143.2 120.4 142.3 100.1 141.1 
RUN 6 131.1 138.5 127.2 139.8 123.5 136.1 96.79 142 
RUN 7 129.7 142.9 130.2 140.7 123 136.7 99.63 140.8 
 
(450 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
122 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 136.1 137.6 131.4 124.2 
RUN 2 134.5 140.6 134 128.3 
RUN 3 135.3 138 128.6 121.8 
RUN 4 124.6 137.1 125.5 124.1 
RUN 5 135.2 142.9 133.7 130.9 
RUN 6 129.1 140.5 131 128.3 
RUN 7 132.5 135.4 126 127.2 
 
Exit Data: 
  Exit Exit Exit 
 S13 S14 P13 P14 U7 
RUN 1 102.4 127.3 110.3 125.4 108.5 
RUN 2 109.5 120.2 119 121.1 109.2 
RUN 3 107.9 118.3 117.5 119.1 104.8 
RUN 4 115.3 121.6 120.1 124.4 113.3 
RUN 5 106.6 117.8 115.2 125.8 109.3 
RUN 6 102.7 115.2 111.1 119.4 104.6 
RUN 7 107.7 118.7 115.9 121.4 104.3 
 
(450 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
123 
 
Factor level combination 6: 450 RPM / 4 inch per min traverse / 0.7 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 161.8 166.2 154.3 164.6 169.55 181 148.3 150.9 192.4 153.3 
RUN 2 163.6 166.8 149.8 165 175.66 176.1 143.3 150.2 191.2 158.5 
RUN 3 166.4 175.1 152.8 172.8 165.22 184.5 144.3 160.9 201.5 135.6 
RUN 4 165.1 166.8 151.2 163 169.37 181.5 136.2 163.7 201.6 159.3 
RUN 5 164.2 169.1 149.6 165.7 169.15 180.3 142.8 164 197.8 160.3 
RUN 6 162.2 170.5 152.7 164.7 167.21 179 141.9 162.9 198.2 158.3 
RUN 7 164.4 178 153.1 164.9 173.55 184.5 145.1 148.4 200.1 155.9 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 150.5 163 143.1 163.3 147.22 165.7 150.5 158.6 
RUN 2 145.6 161 145.9 161.9 143.68 163.5 150.4 153.9 
RUN 3 141.8 170.8 151.4 163.9 145.63 172.4 140.5 163.3 
RUN 4 143.2 157.2 140 163.6 150.03 161.2 146.3 168.8 
RUN 5 148.7 160.1 146.1 152.1 155.27 156.3 144.5 156.6 
RUN 6 146.4 163.2 147.2 168 146.99 160 149.2 154.5 
RUN 7 153.6 164.3 148.3 153.1 157.53 161.8 152 162.4 
 
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 152.8 161 145.3 165.2 153.13 162.7 107.2 148.4 
RUN 2 143.9 155.6 145.6 162.5 152.05 157.2 106.1 147.5 
RUN 3 144.4 170.5 152.8 171.5 150.07 167.8 99.3 137.8 
RUN 4 145.4 156.4 142.1 156.9 146.43 169.8 106.4 156.8 
RUN 5 152.2 160.6 150.9 161.4 153.44 166.9 106.7 162.4 
RUN 6 149.8 161.8 146.6 161.4 153.24 156.2 108.3 159.3 
RUN 7 155.5 163.9 151.9 163.3 154.6 168.7 110.5 160 
 
(450 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
124 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 157 156.9 141 150.7 
RUN 2 154.5 154.5 141.4 148.2 
RUN 3 144.3 140.4 134.2 138 
RUN 4 145.9 144.8 140.1 149.1 
RUN 5 156.2 163.2 148.5 146.7 
RUN 6 155.4 157.6 151 152.8 
RUN 7 155 159.7 148.5 152.4 
 
Exit Data: 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 126.4 134.8 133.9 144.4 128.09 
RUN 2 126.3 136.4 129.2 145.7 121.69 
RUN 3 124.4 145.9 121.5 153.6 134.48 
RUN 4 129.4 134.9 133.3 140.6 128.07 
RUN 5 133.3 136 133.3 148 127.97 
RUN 6 128.7 129.8 125.5 139.5 129.87 
RUN 7 126.4 136.3 133.6 145.3 126.06 
 
(450 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
 
125 
 
Factor level combination 7: 350 RPM / 7 inch per min traverse / 0.7 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 179.3 183.6 139.1 148.1 182.47 192.6 121.2 134.4 202.2 134.9 
RUN 2 173.9 187.7 149 161.6 179.71 204.4 130.9 140.7 212.1 142.7 
RUN 3 172.1 190 146.3 158.6 178.01 206.7 129.5 136.5 209.3 140 
RUN 4 170.9 181.2 146.8 153 179.22 197.5 129.3 131.5 207.5 138 
RUN 5 172.9 178.5 143.9 153.1 176.41 181.6 127.2 139.5 203.1 135.4 
RUN 6 168.3 180.2 145.6 156.5 185.54 195.1 128.5 131.5 209.4 136.4 
RUN 7 173.6 181.9 147.9 154 184.25 189.9 123.4 131 206.8 142.4 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 130 134.2 124.4 139.9 132.03 135.1 120.1 130.1 
RUN 2 135.3 146.2 134.2 144.6 134.12 145 127 137.1 
RUN 3 131.1 142.1 127 137.5 133.36 141.5 129.2 129.3 
RUN 4 131 141.3 128.8 135.2 137.54 140.9 124.9 135.7 
RUN 5 129.2 140.1 128 136.1 135.84 139.5 127.6 132.5 
RUN 6 127.1 141 128.2 136.1 139.44 140.8 126.2 133.9 
RUN 7 130.3 142.1 129.6 138.4 131.71 142 122.2 132.7 
 
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 127.2 132.8 127 142.3 128.06 138.4 94.98 136.1 
RUN 2 122.9 144.4 122.3 142.5 128.28 142.4 100.8 139.2 
RUN 3 128 138.3 125 135.5 128.91 137 96.4 136.5 
RUN 4 128.9 139.1 132.1 136.4 125.07 135.9 93.28 139.6 
RUN 5 135.6 138 132.1 135.8 132.02 136.7 96.08 134.9 
RUN 6 127.9 138.6 131.8 134.3 133.83 137.3 93.49 139.2 
RUN 7 129.3 140.5 124.6 138.1 125.93 138.8 96.74 138.6 
 
(350 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
 
126 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 128.5 128.8 121.8 116.2 
RUN 2 135 143.2 137.6 132.3 
RUN 3 132 136.6 132.5 126.5 
RUN 4 125.3 133.9 128.4 125.9 
RUN 5 124.6 136.9 129.8 125.1 
RUN 6 126.9 133.8 127.9 126.7 
RUN 7 131.5 136.4 129.5 128.5 
 
Exit Data:  
  Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 107.7 119.7 107.7 113.3 100.44 
RUN 2 112.3 121.8 107.9 121.3 109.95 
RUN 3 100.4 117.8 105.3 115.8 106.96 
RUN 4 106.1 110.4 108.4 117.5 104.03 
RUN 5 105.4 118.1 94.65 118.7 106.99 
RUN 6 104.7 106.6 103.8 114.9 98.12 
RUN 7 108.7 119.7 106.8 122.6 108.26 
 
(350 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
 
127 
 
Factor level combination 8: 350 RPM / 4 inch per min traverse / 0.7 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 171.3 181.5 151 161.8 182.54 194.5 125.2 147.4 201.8 146.5 
RUN 2 176.5 188.1 154 174.4 179.61 191.9 129.7 142.3 204.5 147.5 
RUN 3 177.4 189.3 161.4 170.7 190.25 193.8 129.4 150.1 209.1 155.7 
RUN 4 183 192.9 166 186 181.79 211.5 140.5 155.7 215.4 157.5 
RUN 5 178.1 179.9 162 165.7 179.17 193.6 135.2 145.9 209.8 160.6 
RUN 6 174.9 183.2 161.1 183.1 180.68 194.9 140.9 147.2 212.5 152.5 
RUN 7 169.4 193.3 168.5 176.1 185.74 188.7 136.8 149 209.9 157 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 145.8 156.1 140.3 158.8 150.2 160.7 131.9 143.4 
RUN 2 144 163.9 134.7 154.1 149.22 164.7 149.8 152.2 
RUN 3 157.7 164 144.8 153.6 159.18 160.7 152.8 157.4 
RUN 4 141 173.1 154.7 167.7 141.04 172.6 142.7 167.2 
RUN 5 154.9 157.2 142.6 146.2 147.48 152.6 135.5 152.2 
RUN 6 164.1 164.7 145.4 158 160.62 160.7 136 166.6 
RUN 7 148.1 171.1 150.5 157.8 146.79 167 149 163 
 
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 149.1 163.3 134.6 148.6 151.89 162.4 100.6 145.9 
RUN 2 144.3 151.2 149.9 156.8 142.09 153.1 107.4 154.4 
RUN 3 153.1 160.8 130.5 157.7 152.35 154.5 111.5 158.5 
RUN 4 146.1 171.6 144.9 171 147.27 167.4 115.5 165.9 
RUN 5 148.2 151.7 143.8 148.2 146.05 157.7 106.6 162.5 
RUN 6 147.6 161.5 146 159.5 134.88 157.4 111.2 162.1 
RUN 7 150.6 168.5 143 162.5 140.62 162.3 112.8 166.6 
 
(350 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
128 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 141 149.1 128.7 145.3 
RUN 2 148.5 143.7 134.1 149.3 
RUN 3 153.1 146.2 137.7 144.3 
RUN 4 164.1 163.9 155.8 149.5 
RUN 5 162.4 142.3 140.1 131.6 
RUN 6 146.2 154.1 143.4 149.9 
RUN 7 155.9 150.6 147.9 156.7 
 
Exit Data: 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 119.4 131.8 127.9 128 116.91 
RUN 2 113.9 135.1 120.4 131.3 121.89 
RUN 3 125.3 142.7 129.3 140.2 123.31 
RUN 4 137.7 146.2 131.7 148.4 137.42 
RUN 5 126.1 140 118.9 142.9 114.54 
RUN 6 120.5 142.5 126 138.5 126.13 
RUN 7 136.3 137 123.4 145.1 131.65 
 
(350 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
 
129 
 
Factor level combination 9: 450 RPM / 7 inch per min traverse / 0.5 inch shoulder  
Entry Data 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 128.3 138.4 100.2 120.7 135.16 155.2 107.1 111.9 150.9 101.3 
RUN 2 134.1 145.5 110.6 115.9 141.33 144.2 90.54 117.3 165.9 105.5 
RUN 3 141.8 147.2 104.6 119.7 129.19 149.5 93.88 121.9 161.3 109.8 
RUN 4 125 149.4 112.8 119 130.6 152 105 111.6 164.4 120.4 
RUN 5 126.4 147 104.8 123.2 129.36 149.9 106.6 115.2 161.4 113.4 
RUN 6 126.6 140 114.4 114.7 143.46 148.1 100.2 111.9 160.2 110 
RUN 7 120.6 148 103.9 125.6 143.22 158.5 107.3 119.1 163.1 119.9 
 
Traverse Data 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 94.73 108.1 92.87 108 109.68 116.8 105.1 116.4 
RUN 2 99.21 114.6 87 95.32 100.91 121.9 104.4 119.5 
RUN 3 98.66 110.1 93.71 109.7 97.51 128.1 106.7 114.3 
RUN 4 110.5 120.8 96.79 111.9 112.83 122.5 116 120.3 
RUN 5 102.1 117.3 100.4 115.7 117.16 125.5 112.6 119.8 
RUN 6 101.1 107.2 93.28 108.3 105.24 116.8 101.7 115.8 
RUN 7 106 121.1 102 112.5 104.59 112.1 101.3 121.7 
 
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 104.2 116 91.62 112.7 105.09 121.5 81.4 110.1 
RUN 2 99.62 122.6 95.9 110.3 107.59 116.6 89.08 112.9 
RUN 3 105.9 118.4 92.75 105.4 94.67 112.3 70.82 117.6 
RUN 4 110.5 118.5 100.5 116.8 109.46 116.8 86.17 118.3 
RUN 5 112.9 125.9 100.2 113.4 105.23 121.8 84.7 126.2 
RUN 6 109 116.2 93.02 104.5 102.94 124.5 75.78 113 
RUN 7 114.7 121 101.6 111.5 102.65 126.9 87.89 126.7 
 
(450 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
130 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 98.54 103.9 104.5 95.08 
RUN 2 96.36 106.6 109.3 98.17 
RUN 3 102.8 102.6 106.4 94.67 
RUN 4 113.3 117.2 108.6 101.2 
RUN 5 107.7 117.7 110.6 93.49 
RUN 6 93.53 116.3 100.8 100.7 
RUN 7 112.2 120 121.3 110.7 
 
Exit Data 
 
  Exit  Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 90.85 94.25 73.03 99.64 86.03 
RUN 2 80.81 92.11 79.86 92.2 97.27 
RUN 3 78.9 92.07 77.31 88.56 81.47 
RUN 4 91.01 97.53 86.47 94.77 98.62 
RUN 5 91.04 98.64 80.05 90.44 95.11 
RUN 6 83.76 95.44 74.37 97.18 85.49 
RUN 7 81.91 99.63 87.97 93.86 99.41 
 
(450 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
131 
 
Factor level combination 10: 450 RPM / 4 inch per min traverse / 0.5 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 177.9 191.9 153.9 165.3 178.75 191.7 135.4 137.6 211.6 151.9 
RUN 2 166.8 185.9 141.8 164.2 170.43 181.6 123.1 143.5 197.8 128.7 
RUN 3 169.5 175.2 158.5 157.2 174.4 190.1 124.8 142.6 198.2 143.2 
RUN 4 171 189.5 152.8 178.1 168.45 187 133.6 147.2 222.1 156.4 
RUN 5 173.6 186.4 157 165.7 179.2 192 131.5 142.8 205.3 154.4 
RUN 6 166.4 177.3 151.1 162.7 174.9 186.2 127.9 131 201.9 142.4 
RUN 7 174.1 185.3 145.5 163.3 174.22 172.8 124.3 130.7 203 154.2 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 146.2 161.1 139.7 149.4 151.87 155.6 155.1 155.7 
RUN 2 124.8 160.2 111.4 139.7 131.52 163.5 144.6 163.2 
RUN 3 144.2 157.7 124.9 139.8 139.52 157.8 145.3 156.8 
RUN 4 155.2 166.2 123.4 161.4 134.9 165.4 142.2 162.4 
RUN 5 146.1 168.9 134.6 144.6 154.46 158.3 155.2 158.9 
RUN 6 148.4 160.7 130.3 142.5 146.75 158 145.4 157.8 
RUN 7 147.9 150.5 126.1 141.4 147.04 151.3 151.9 152.7 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 145.5 158.2 134.7 154.5 146.08 149.2 109.6 161 
RUN 2 140.6 156.6 138.3 154.2 122.58 156.9 92.99 134.1 
RUN 3 140.3 154.2 135.2 148.7 135.13 153.3 103.5 152.7 
RUN 4 158.5 164.6 147.8 161.8 139.24 161.1 114.4 171.5 
RUN 5 146 166.5 135.4 160.1 138.32 153.3 112.7 164.6 
RUN 6 143.6 156.7 140.7 143.6 142.66 155.2 103.3 155.7 
RUN 7 141.4 153.9 139 148.1 142.52 145.7 101.1 150.5 
 
(450 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
 
132 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 154.8 154.6 141.4 136.5 
RUN 2 116.7 119.3 113.8 115.9 
RUN 3 144.4 135 131.4 147.7 
RUN 4 159.3 155.6 150.8 143.7 
RUN 5 155.2 143.5 140.9 149.9 
RUN 6 145.1 139.6 131.5 138.9 
RUN 7 141.9 140.3 130.4 126.2 
 
Exit Data: 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 120 135.2 131.8 138.6 118.28 
RUN 2 118.3 121.3 119.6 146.8 103.83 
RUN 3 124.2 130.2 124.7 134.1 126.31 
RUN 4 113.9 126.8 126.7 148.6 104.99 
RUN 5 123.2 133.4 130.4 142.5 124.71 
RUN 6 116.4 135.9 132.1 132.3 116.39 
RUN 7 118.3 134.2 123.8 141 110.41 
 
(450 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
133 
 
Factor level combination 11: 350 RPM / 7 inch per min traverse / 0.5 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 158.2 169.7 137.8 149.9 154.34 175 128.4 130.1 189.6 142 
RUN 2 142.6 163.7 132.9 143.6 155.64 168.2 123 127.2 195.1 139.3 
RUN 3 155.1 175.6 139.9 158.4 162.85 183.9 122.7 136.2 203.1 143.3 
RUN 4 146 155.8 134.1 141.6 166.92 175.2 124.3 133.8 184.6 137 
RUN 5 159.5 171.3 144.8 148 165.82 177 128.1 138.3 194.9 141.2 
RUN 6 152.6 170.7 135.7 146.6 169.23 179.9 125.7 138.1 189.2 141 
RUN 7 152.9 164.4 133.7 144.8 161.35 171.6 125.5 138.2 197.1 139.7 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 135.8 143.5 120.1 140.3 132.42 140.1 124.6 131.8 
RUN 2 131.9 141.6 127.6 139.3 135.28 139 125.3 133.1 
RUN 3 142.2 146.9 131.8 137.5 129.47 139.6 121.9 132.9 
RUN 4 136.6 139.2 121.3 137.6 130.56 137.9 126.3 127.4 
RUN 5 132.9 142.1 128.1 130.3 134.28 138.5 127 128.3 
RUN 6 139.9 142.4 124.6 140.8 137.69 142.2 127 132.4 
RUN 7 135.5 141.6 127.8 130.4 134.45 138.8 126.6 134.9 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 130.7 141.1 128.5 139.4 131.42 136.1 92.6 137.6 
RUN 2 128.5 141.3 126.7 141.3 123.26 136.3 92.92 138.7 
RUN 3 137.3 146.3 135.4 145.1 127.02 138 96.49 141.2 
RUN 4 127.9 138.2 120.2 137.3 127.79 133.4 92.42 134.5 
RUN 5 132.5 141.7 129.8 139.5 132.79 135.3 93.35 135.8 
RUN 6 135.9 141.7 133 140.6 125.16 137.6 92.73 135 
RUN 7 132.3 142.3 129.4 140.6 122.98 137.9 93.52 140.7 
 
(350 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
134 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 130.6 136.9 127.3 129.7 
RUN 2 133.7 140.5 127.8 132 
RUN 3 137.7 137.4 129.9 130.9 
RUN 4 129.9 135.9 125.8 128.9 
RUN 5 132.9 139.8 128.7 130.4 
RUN 6 135.1 140 124.1 126.4 
RUN 7 136 142.5 130 134.1 
 
Exit Data: 
 
  Exit Exit  Exit 
  S13 S14 P13 P14 U7 
RUN 1 106.7 112.8 109.9 122.2 103.06 
RUN 2 111.8 112 111.9 124.5 111.71 
RUN 3 106 122.7 118.3 125.6 105.81 
RUN 4 104.9 115.7 113.1 122.2 106.72 
RUN 5 102.8 116.3 111.8 119.1 106.62 
RUN 6 101.6 112.2 110.3 125 101.14 
RUN 7 107.6 114.8 113.7 123.3 111.08 
 
(350 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
 
135 
 
Factor level combination 12: 350 RPM / 4 inch per min traverse / 0.5 inch shoulder  
Entry Data 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 148.6 165.5 131.7 145.5 150.16 161.8 111 132.2 173.3 131.1 
RUN 2 149.6 161.1 133.9 144.4 137.75 161.6 113.6 134 186.7 139.2 
RUN 3 165.7 168.3 148.2 150 147.92 171.7 129.7 131.3 182.3 146 
RUN 4 143.2 168.9 133.6 146.6 152.78 170.7 115.4 132.5 176.6 129.4 
RUN 5 147.9 160.4 136 146.4 161.97 164.5 115.6 135.8 186.3 135.5 
RUN 6 149.5 160.6 134.2 143.2 148.37 160.2 113.8 136.8 171.5 139 
RUN 7 137.3 159.3 135.2 138.1 153.89 185.4 113.8 126.1 179.6 132.2 
 
Traverse Data 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 117 140.3 119.9 133.1 126.43 135.8 125.3 133.1 
RUN 2 124.5 141.4 118.8 134.7 124.67 135.3 123.6 132.3 
RUN 3 142.3 146.2 129 134.5 126.34 142.1 122.2 142.9 
RUN 4 121.8 139.3 120.3 136.3 128.45 138.6 118.2 135.3 
RUN 5 125.7 141.8 121.7 135.9 124.23 134.4 119.2 136.2 
RUN 6 126.2 146.7 122 137.7 106.82 128.4 108.4 129.7 
RUN 7 121 132 116.3 129.7 125.44 138.4 124 127.5 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 117.8 137.3 120 133.5 123.19 139.3 90.67 135.8 
RUN 2 121 138.8 118.6 136.3 120.73 130.2 89.88 134.2 
RUN 3 136.2 147 133.4 140.7 132.88 139.9 92.03 151 
RUN 4 121.6 137 120.4 135.4 123.6 131.9 87.78 135.1 
RUN 5 123.7 140.1 122 137.8 120.97 131.1 88.67 138.6 
RUN 6 119.7 134.7 127.9 134 124.22 135.6 76.14 135.4 
RUN 7 120 130.2 118.9 129.2 122.8 135.8 87.94 131.6 
 
(350 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
136 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 131.7 120.5 115.7 121.1 
RUN 2 131.9 127 120.9 128.8 
RUN 3 144.6 140.6 133.7 125.1 
RUN 4 133.2 126.9 122.3 127.6 
RUN 5 135.5 130 124.1 125.3 
RUN 6 136.4 122.9 127.2 109.3 
RUN 7 122.9 128.2 119.2 127.5 
 
Exit Data 
 
  Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 103.7 119.3 102.2 118 107.76 
RUN 2 106.7 115.1 104.8 117.4 110.57 
RUN 3 118.8 123.2 121.5 135.3 110.71 
RUN 4 100.5 117.2 99.84 114.5 107.24 
RUN 5 102.9 122.4 103 119.7 107.32 
RUN 6 95.06 112.3 99.89 113.4 95.8 
RUN 7 105.9 119.1 105.6 112.6 105.96 
 
(350 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
137 
 
APPENDIX B 
Calculated Data: With 2 RPMs, 2 traverse speed and 3 shoulder diameters, there is a 
combination of 12 different parameters. For each parameter, 7 experiments were done to 
achieve at least a 90% statistical power that advancing was hotter than retreating side. 
The peak temperatures calculated for all the 12 different parameters are given below. 
Nomenclature: Each Thermocouple was represented by a unique alphabet and number: 
  S ? Shoulder thermocouple. (Retreating side: S
1
, S
3?
 S
13;
  
           Advancing side: S
2
, S
4..
 S
14
) 
             P ? Pin thermocouple. (Retreating side: P
1,
 P
3?
 P
13;
 Retreating side: P
2
, P
4
? P
14
) 
           U ? Under the pin thermocouple. (U
1,
 U
2?
 U
7
) 
Factor level combination 1: 450 RPM / 7 inch per min traverse / 0.9 inch shoulder  
Entry Data: 
 Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 361.7 383.2 331.1 334.3 379.28 395 311.2 332.5 396.3 332.5 
RUN 2 360.7 374.6 330.6 339.8 376.51 387.4 318.2 324.6 402.1 341.2 
RUN 3 365.9 382.3 330.4 338.1 375.64 394.3 324.7 328 414.5 334.1 
RUN 4 365.6 377.7 330.5 339.4 383.72 396.8 323.6 328.8 400.9 340.4 
RUN 5 367.4 371 328.2 341.7 379.72 383 319.3 333.6 404.2 346.8 
RUN 6 366.3 378.4 332 339.2 377.2 391.2 315.7 331.1 405.6 335.6 
RUN 7 373.6 377.5 334.8 342 386.36 393.1 320.2 332.7 411.1 330 
 
(450 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
138 
 
Traverse Data: 
? Group 1 Group 2 Group 3 Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 321.2 341.2 314.7 327.6 324.73 329.3 323.3 324.9 
RUN 2 321.7 335.5 320.1 332 330.97 331.9 317.5 331.2 
RUN 3 330 333 319.3 324.9 325.5 336.2 316.3 326.7 
RUN 4 329.2 331.6 317.9 331.6 328.64 342 319.6 329.9 
RUN 5 329.3 339.5 319.3 328.4 326.03 340.6 314.6 327.9 
RUN 6 327.7 339.2 316.5 331 327.66 338.7 322.8 326.5 
RUN 7 320.7 342.3 318.4 324 328.59 331.4 323.9 324.4 
 
? Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 319.2 331.7 325 327.9 323.81 334.1 286.6 330.4 
RUN 2 331.2 336.4 319.4 334.3 318.16 331.9 289.6 338.4 
RUN 3 330.2 341.5 316.8 329.8 324.83 335.8 287.1 331.4 
RUN 4 329.2 332.5 321.1 332.8 326.25 328.7 280 329 
RUN 5 326.4 340.3 322.6 338.1 323.85 337.7 286.1 342.1 
RUN 6 328 339.6 320.9 328.6 327.15 328 288 333.7 
RUN 7 323.1 334 324.2 329.7 319.03 330.1 289 335 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 329.4 325.7 320.2 325.4 
RUN 2 333.9 335.9 323.6 328.1 
RUN 3 326.4 338.4 325.3 326.2 
RUN 4 325 334.9 328.7 328.1 
RUN 5 333.1 343 317.7 334.3 
RUN 6 325 330.7 321.2 326.2 
RUN 7 328.2 333.3 329.6 327.6 
 
(450 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
 
139 
 
Exit Data: 
  Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 294.8 305.8 306.5 309.1 297.39 
RUN 2 301.4 313.1 294 321.6 301.4 
RUN 3 294.3 311.9 305.8 319.3 297.27 
RUN 4 301.7 303.4 304.5 320.6 298.21 
RUN 5 303 306.3 293.4 316.8 312.31 
RUN 6 293.3 309.4 300.9 314.9 299.55 
RUN 7 296.9 311.6 307.9 313.4 303.25 
 
(450 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
140 
 
Factor level combination 2: 450 RPM / 4 inch per min traverse / 0.9 inch shoulder 
Entry Data: 
 Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 379 383.1 377.9 383.4 395.61 412 350.3 372.7 422.1 371 
RUN 2 383.8 386.7 378.1 389.3 394.18 407.4 359.2 361 428.8 376.3 
RUN 3 386.7 396.6 369.3 395.4 395.55 417.4 361.2 363.9 426 380.2 
RUN 4 383.2 393.4 383.8 386.9 402.26 411.1 358.6 368.7 414.6 364.4 
RUN 5 376.5 393.5 381.3 388.5 391.95 407.1 355.5 369.5 426.5 380.5 
RUN 6 378 395.8 377.5 388 391.83 402.7 355.4 370.2 421.9 371.7 
RUN 7 381.8 391.2 380 389.9 391.12 401.7 355.7 360.6 419.9 365.4 
 
Traverse Data: 
 
 Group 1 Group 2 Group 3 Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 367.8 376 347.5 366.2 368.65 371.8 362.9 369.1 
RUN 2 373.1 387.4 362 376.8 362.77 384.6 371.7 377.2 
RUN 3 367.2 392.9 364.8 382 365 386.4 363 382.7 
RUN 4 354.9 387.9 362.4 373.5 370.66 374 362.3 376.7 
RUN 5 374.2 380.4 356 373.9 362.7 373.1 367.7 371.9 
RUN 6 366.3 375 356.1 367.1 365.15 377.5 363.5 375.5 
RUN 7 373.3 385.5 359.6 367.8 372.2 374.3 362.5 382.9 
? ????? ?? ?
  Group 1 Group 2 Group 3 Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 368 376 363.4 369.2 365.37 378.2 312.9 380.8 
RUN 2 366.4 387.2 365.4 385.1 360.83 381 319.8 391.7 
RUN 3 360.3 392.7 367.9 368.3 360.93 382.8 320.1 394.2 
RUN 4 368.3 387.4 366.1 374 369.44 381 312.4 385.3 
RUN 5 375.3 380.2 363.3 371.6 368.9 371 316.5 378 
RUN 6 368.5 386.1 358 374.5 373.29 373.7 314.8 383.6 
RUN 7 368.4 380.9 363.5 377.8 367.77 381.4 310.6 383.6 
 
(450 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
141 
 
  
Group 
1 
Group 
2 
Group 
3 
Group 
4 
 U3 U4 U5 U6 
RUN 1 368.6 365.1 361.2 365.6 
RUN 2 377.1 369.7 363.5 371.7 
RUN 3 384.9 379.5 368.4 376.4 
RUN 4 374.4 368.9 373.1 367.8 
RUN 5 373.5 373.4 360.3 371.4 
RUN 6 372.5 379.2 354.6 358.7 
RUN 7 376.2 383.3 364.9 371.3 
 
Exit Data: 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 340.2 355.7 341.7 353.2 344.38 
RUN 2 345.1 351.8 339.1 358.3 345.19 
RUN 3 350 358.9 345.9 365 346.11 
RUN 4 339.7 355.4 344.9 362.3 346.81 
RUN 5 337.3 347 344.5 351.7 342.84 
RUN 6 333.6 351.5 328.3 352.5 327.1 
RUN 7 341.1 354.3 344.5 358.2 346.35 
 
(450 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
 
142 
 
Factor level combination 3: 350 RPM / 7 inch per min traverse / 0.9 inch shoulder  
Entry Data: 
 Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 373.9 375.8 314.7 324.7 365.52 389.7 316.1 328.5 408.4 328.5 
RUN 2 369.9 376.3 326.8 339 362.78 390.7 306.2 334.5 409 329 
RUN 3 371.5 383.7 325.8 336.7 366.04 395.5 313.9 328.3 415.6 329.9 
RUN 4 370.5 384.6 322.6 345.5 376.97 402.9 316 324.6 410.8 330.4 
RUN 5 370.3 378.1 322.2 328.2 368.48 386.6 305.9 328.2 410.1 225 
RUN 6 370.7 382.9 326.2 346.8 381.88 399.9 323.2 329.1 417.2 343 
RUN 7 372.8 381 328.8 347.5 371.8 395.3 324.4 329.1 414.1 342.1 
 
Traverse Data: 
 
? Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 319.8 330.5 319.7 328.5 320.46 326 321.1 326.1 
RUN 2 312.7 325 314.3 333.3 312.06 323.9 320 320.3 
RUN 3 313.7 331.3 314.4 331.2 320.86 329.4 315.8 329.2 
RUN 4 316.1 332.2 314.2 321.5 314.84 329 315.7 328.7 
RUN 5 312.8 314.8 301.5 328.1 311.27 315 305.8 327.6 
RUN 6 320.8 331.6 321.5 327.1 309.17 333.5 322.1 329.1 
RUN 7 312 332.2 319.3 329.5 307.64 333.7 321.9 332.5 
? ????? ?? ?
? Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 322 333.5 321.8 336 318.9 323.9 285.7 300.9 
RUN 2 319.6 325.8 320.1 335.1 317.92 330.4 284 325.7 
RUN 3 315.1 326.7 322.4 324.2 319.43 327.5 286.4 323.5 
RUN 4 315.8 327.6 321.7 323.9 320.17 327.3 285.5 323.9 
RUN 5 312 314.8 317.4 332.2 309.38 332.1 276.4 315.5 
RUN 6 311.5 333 327 333.1 310.69 331.9 289.4 337.9 
RUN 7 313.4 335.6 319.5 333.1 328.4 333.1 288.6 332.2 
 
(350 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
143 
 
?
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 326.5 336 314.6 318.2 
RUN 2 328.8 328 317.5 321.5 
RUN 3 319.7 335.4 325.8 318.9 
RUN 4 317.6 331.5 319.3 308.7 
RUN 5 304.6 320 311.3 302.9 
RUN 6 329.1 336.8 332.9 331.3 
RUN 7 330.9 336.2 331.6 327.6 
 
Exit Data 
 
? Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 279.2 297.4 284.7 296.9 279.28 
RUN 2 284.3 307.2 286.3 307.4 294.82 
RUN 3 282.7 293.7 294.8 304.8 294.32 
RUN 4 292.1 304.2 298.1 304.5 291.2 
RUN 5 279.6 297.9 286.3 302.6 280.92 
RUN 6 286.3 302.2 301.8 312.8 302.23 
RUN 7 292.4 294.3 292.8 309.2 301.53 
 
(350 RPM / 7 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
 
144 
 
Factor level combination 4: 350 RPM / 4 inch per min traverse / 0.9 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 381.7 390.7 366.8 384.1 386.8 406.4 344.3 359.6 416.3 357.5 
RUN 2 379.4 388.3 367.5 376.1 398.92 407.3 340.1 364.1 410.6 358.9 
RUN 3 381.9 389.9 363 391 398.13 408.6 356.8 370 415.4 373.9 
RUN 4 379.6 387.3 368.2 377.8 387.83 406.8 359.9 369.9 404.2 371.4 
RUN 5 372.6 380 363 368.8 385.29 391.4 342.2 356.3 407.8 355.4 
RUN 6 378.2 388.2 357.3 369 386 399.2 349.5 365.2 417 362.3 
RUN 7 381.4 384.3 367.8 379.9 391.66 402.3 347.4 362.2 419.4 370.2 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 360.1 376.3 347.3 357.6 347.41 362.3 341.2 344.1 
RUN 2 362.5 370.6 344.8 355.5 345.16 366.5 342.1 359.6 
RUN 3 363.8 384.1 356.8 375.9 357.46 378 347.2 355.4 
RUN 4 365.8 378.5 343.4 371.8 342.22 365.5 346.6 358.2 
RUN 5 364.6 378.7 340.3 355.9 351.33 359.8 345.6 354.7 
RUN 6 358.4 362.5 346.7 361.4 352.23 356.6 343.9 355.1 
RUN 7 356.1 366.7 342.8 359.9 353.96 355.1 351.2 356.6 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 357.2 375.9 354.2 370.3 354.3 366.2 315.7 351.8 
RUN 2 362.7 369 362.1 368.8 361.72 362.5 310.7 365.5 
RUN 3 355.3 373 352.9 379.7 354.74 376.5 318.1 357 
RUN 4 359.6 374.5 359.1 385 348.57 382.3 322.9 360.7 
RUN 5 348.7 369 365 368.1 352.08 366.9 310.7 343.6 
RUN 6 359 369 353.2 360.4 351.21 356.1 307.7 365.5 
RUN 7 362.5 367.3 351.7 361.2 353.66 364.9 309.3 358.8 
 
(350 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
145 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 347.7 344.9 333 347.6 
RUN 2 359.8 355 345.7 359.4 
RUN 3 375.4 361 350.4 374.3 
RUN 4 374.1 363.1 344.2 362.1 
RUN 5 362.5 356.2 336.7 361.6 
RUN 6 354.1 350.5 352.2 357.2 
RUN 7 361.8 357.5 348.6 357.9 
 
Exit Data 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 315.7 323.6 318.9 325.1 319.01 
RUN 2 318.3 332.4 330.6 330.6 339.13 
RUN 3 315.2 331.7 308.1 339.6 330.91 
RUN 4 318.2 339.4 314.6 345.3 334.36 
RUN 5 309.7 315 310.6 326.3 317.05 
RUN 6 315.2 319.8 317.2 331 323.3 
RUN 7 310.4 334.1 320.1 331.2 328.35 
 
(350 RPM / 4 inch per min traverse / 0.9 inch shoulder) 
 
 
 
 
 
 
 
 
 
146 
 
Factor level combination 5: 450 RPM / 7 inch per min traverse / 0.7 inch shoulder  
Entry Data 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 352.5 360.9 322.9 343.5 373.09 389.9 314.5 318.9 391.6 327.7 
RUN 2 353.3 373.4 323 333.2 371.75 383.8 314.5 318.3 404.2 333.1 
RUN 3 356.7 368.3 331.3 332.8 370.88 382.6 314.3 318.3 395.5 329.7 
RUN 4 358.4 372.2 326.3 333.6 371.86 380.8 311.6 316.3 408.9 328.2 
RUN 5 360.9 366.3 331.6 338.8 377.88 383.9 308.9 322.5 403.1 336.6 
RUN 6 359.8 370.6 327.3 340.6 380.29 383.3 310.9 324.6 403.9 336 
RUN 7 359.1 362.1 332.8 339.1 378.25 390.8 308.6 320.8 397.4 332.8 
 
Traverse Data 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 318.4 330.9 317.5 331.3 312.47 329.7 310.4 323.9 
RUN 2 319.6 332.2 317.6 323.7 322.79 326.1 311.6 320.7 
RUN 3 322.4 325.4 314.6 321.9 318.43 323.6 313.7 318 
RUN 4 315.4 328.4 316.7 326.1 320.18 322.9 313.1 321.8 
RUN 5 319.9 331.2 319.2 322.8 318.14 329.4 309.9 320.7 
RUN 6 315.8 324 312.5 325.5 319.12 324.5 316.6 321.9 
RUN 7 313.8 328.6 315.4 325.5 321.19 322.9 317.1 320.8 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 316.1 336.9 318.9 330.8 315.32 329.6 282.4 332.1 
RUN 2 310.8 331.3 325.5 328 313.09 323.8 287.6 331 
RUN 3 323.1 327 323.3 325.7 317.83 329.2 285.6 322.2 
RUN 4 317.8 328.7 320.1 327.5 319.62 321.4 282.6 325.9 
RUN 5 312 333.3 324.7 331.2 308.36 330.3 288 329.1 
RUN 6 319.1 326.5 315.1 327.7 311.51 324.1 284.8 330 
RUN 7 317.6 330.9 318.2 328.6 311.01 324.7 287.6 328.8 
 
(450 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
 
147 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 329.4 330.9 324.6 317.5 
RUN 2 327.8 333.9 327.3 321.5 
RUN 3 328.6 331.3 321.8 315.1 
RUN 4 317.9 330.3 318.8 317.3 
RUN 5 328.5 336.2 327 324.2 
RUN 6 322.4 333.8 324.2 321.6 
RUN 7 325.8 328.7 319.2 320.4 
 
Exit Data 
  Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 286.6 311.5 298.3 313.4 301.73 
RUN 2 293.7 304.4 307 309.1 302.46 
RUN 3 292.2 302.5 305.5 307.1 298.06 
RUN 4 299.6 305.9 308.1 312.4 306.55 
RUN 5 290.9 302.1 303.2 313.7 302.58 
RUN 6 287 299.4 299.1 307.4 297.89 
RUN 7 291.9 302.9 303.9 309.4 297.6 
 
(450 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
148 
 
Factor level combination 6: 450 RPM / 4 inch per min traverse / 0.7 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 346.1 350.4 338.5 348.8 357.53 369 336.3 338.9 385.7 346.5 
RUN 2 347.8 351 334 349.2 363.65 364.1 331.3 338.2 384.4 351.8 
RUN 3 350.7 359.3 337.1 357 353.21 372.5 332.2 348.9 394.8 328.9 
RUN 4 349.3 351 335.5 347.3 357.36 369.5 324.2 351.6 394.9 352.6 
RUN 5 348.4 353.4 333.9 349.9 357.13 368.3 330.8 352 391.1 353.6 
RUN 6 346.4 354.7 337 348.9 355.19 367 329.9 350.9 391.5 351.6 
RUN 7 348.6 362.2 337.4 349.2 361.54 372.4 333.1 336.4 393.4 349.2 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 334.7 347.2 327.3 347.6 331.45 349.9 334.7 342.8 
RUN 2 329.8 345.2 330.1 346.2 327.91 347.8 334.6 338.1 
RUN 3 326.1 355 335.6 348.1 329.86 356.7 324.7 347.6 
RUN 4 327.4 341.4 324.2 347.8 334.26 345.4 330.5 353.1 
RUN 5 332.9 344.4 330.3 336.3 339.5 340.6 328.8 340.9 
RUN 6 330.6 347.4 331.5 352.2 331.22 344.2 333.5 338.7 
RUN 7 337.9 348.5 332.5 337.3 341.76 346 336.2 346.7 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 340.8 349 333.3 353.2 341.12 350.7 295.2 336.3 
RUN 2 331.9 343.5 333.6 350.5 340.04 345.2 294.1 335.5 
RUN 3 332.4 358.5 340.8 359.5 338.06 355.8 287.3 325.8 
RUN 4 333.4 344.3 330.1 344.9 334.42 357.8 294.4 344.8 
RUN 5 340.2 348.6 338.9 349.4 341.43 354.9 294.6 350.4 
RUN 6 337.8 349.8 334.6 349.4 341.23 344.2 296.3 347.3 
RUN 7 343.4 351.9 339.9 351.2 342.59 356.7 298.5 348 
 
(450 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
149 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 350.3 350.2 334.3 344 
RUN 2 347.8 347.8 334.6 341.4 
RUN 3 337.6 333.7 327.5 331.3 
RUN 4 339.2 338 333.4 342.4 
RUN 5 349.5 356.5 341.8 340 
RUN 6 348.7 350.9 344.3 346 
RUN 7 348.3 353 341.8 345.7 
 
Exit Data: 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 310.6 319.1 321.9 332.4 321.37 
RUN 2 310.5 320.6 317.2 333.7 314.97 
RUN 3 308.6 330.1 309.4 341.6 327.76 
RUN 4 313.6 319.1 321.3 328.6 321.35 
RUN 5 317.5 320.3 321.3 336 321.25 
RUN 6 312.9 314 313.5 327.5 323.15 
RUN 7 310.6 320.5 321.6 333.3 319.34 
 
(450 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
 
150 
 
Factor level combination 7: 350 RPM / 7 inch per min traverse / 0.7 inch shoulder  
Entry Data 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 363.5 367.8 323.3 332.3 370.46 380.5 309.2 322.4 395.5 328.2 
RUN 2 358.1 372 333.3 345.9 367.7 392.4 318.9 328.7 405.4 336 
RUN 3 356.3 374.3 330.6 342.8 366 394.7 317.5 324.5 402.5 333.3 
RUN 4 355.2 365.5 331 337.2 367.2 385.5 317.3 319.5 400.8 331.3 
RUN 5 357.1 362.7 328.1 337.3 364.39 369.6 315.2 327.5 396.4 328.7 
RUN 6 352.5 364.4 329.8 340.7 373.53 383.1 316.5 319.4 402.7 329.7 
RUN 7 357.8 366.1 332.1 338.2 372.23 377.9 311.4 319 400.1 335.7 
 
Traverse Data 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 314.3 318.4 308.6 324.1 316.26 319.4 304.3 314.3 
RUN 2 319.5 330.5 318.4 328.9 318.35 329.2 311.2 321.3 
RUN 3 315.3 326.4 311.2 321.8 317.59 325.7 313.4 313.6 
RUN 4 315.2 325.6 313.1 319.4 321.77 325.1 309.2 319.9 
RUN 5 313.4 324.3 312.3 320.4 320.07 323.8 311.8 316.8 
RUN 6 311.3 325.3 312.4 320.3 323.67 325.1 310.4 318.1 
RUN 7 314.5 326.4 313.8 322.6 315.94 326.2 306.4 317 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 315.2 320.7 315 330.3 316.05 326.4 283 324 
RUN 2 310.9 332.4 310.3 330.5 316.27 330.4 288.8 327.2 
RUN 3 316 326.3 313 323.5 316.9 325 284.4 324.5 
RUN 4 316.9 327.1 320 324.4 313.06 323.9 281.3 327.6 
RUN 5 323.6 325.9 320.1 323.8 320.01 324.7 284.1 322.9 
RUN 6 315.9 326.5 319.8 322.3 321.82 325.3 281.5 327.2 
RUN 7 317.3 328.5 312.6 326.1 313.92 326.8 284.7 326.6 
 
(350 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
 
151 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 321.8 322.1 315.1 309.5 
RUN 2 328.2 336.5 330.9 325.5 
RUN 3 325.3 329.9 325.8 319.8 
RUN 4 318.6 327.2 321.7 319.2 
RUN 5 317.9 330.2 323.1 318.4 
RUN 6 320.2 327.1 321.2 320 
RUN 7 324.8 329.7 322.8 321.8 
  
Exit Data 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 291.9 304 295.6 301.3 293.72 
RUN 2 296.5 306 295.8 309.2 303.23 
RUN 3 284.6 302 293.3 303.8 300.24 
RUN 4 290.3 294.7 296.4 305.5 297.31 
RUN 5 289.6 302.3 282.6 306.7 300.27 
RUN 6 288.9 290.8 291.8 302.9 291.4 
RUN 7 292.9 304 294.8 310.6 301.54 
 
(350 RPM / 7 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
 
152 
 
Factor level combination 8: 350 RPM / 4 inch per min traverse / 0.7 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 355.5 365.8 335.2 346 370.53 382.5 313.2 335.4 395.1 339.8 
RUN 2 360.7 372.3 338.2 358.6 367.6 379.9 317.7 330.3 397.8 340.8 
RUN 3 361.6 373.6 345.6 354.9 378.24 381.7 317.4 338.1 402.4 348.9 
RUN 4 367.2 377.2 350.2 370.2 369.78 399.5 328.5 343.7 408.6 350.8 
RUN 5 362.4 364.2 346.2 349.9 367.16 381.6 323.2 333.9 403.1 353.8 
RUN 6 359.1 367.4 345.3 367.3 368.67 382.9 328.9 335.1 405.8 345.7 
RUN 7 353.6 377.5 352.7 360.3 373.73 376.7 324.8 337 403.2 350.3 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 330.1 340.3 324.5 343 334.43 345 316.1 327.6 
RUN 2 328.2 348.2 319 338.3 333.45 348.9 334.1 336.4 
RUN 3 341.9 348.2 329 337.9 343.41 345 337 341.6 
RUN 4 325.3 357.3 338.9 351.9 325.27 356.9 326.9 351.5 
RUN 5 339.1 341.5 326.8 330.4 331.71 336.9 319.8 336.4 
RUN 6 348.4 348.9 329.6 342.2 344.85 344.9 320.2 350.8 
RUN 7 332.3 355.4 334.7 342.1 331.02 351.2 333.2 347.2 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 337.1 351.3 322.6 336.6 339.88 350.4 288.6 333.9 
RUN 2 332.3 339.2 337.9 344.8 330.08 341.1 295.4 342.4 
RUN 3 341.1 348.8 318.5 345.7 340.34 342.5 299.5 346.5 
RUN 4 334 359.6 332.9 359 335.26 355.4 303.5 353.9 
RUN 5 336.2 339.7 331.8 336.2 334.04 345.7 294.6 350.4 
RUN 6 335.6 349.5 334 347.5 322.87 345.3 299.2 350.1 
RUN 7 338.6 356.5 331 350.5 328.61 350.3 300.8 354.6 
 
(350 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
153 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 334.3 342.3 322 338.6 
RUN 2 341.8 337 327.4 342.6 
RUN 3 346.4 339.5 331 337.5 
RUN 4 357.4 357.2 349 342.8 
RUN 5 355.7 335.6 333.4 324.9 
RUN 6 339.4 347.4 336.7 343.2 
RUN 7 349.2 343.9 341.2 349.9 
 
Exit Data: 
 
  Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 303.7 316.1 315.9 316 310.19 
RUN 2 298.2 319.3 308.4 319.3 315.17 
RUN 3 309.5 326.9 317.3 328.2 316.59 
RUN 4 322 330.4 319.7 336.4 330.7 
RUN 5 310.3 324.3 306.9 330.9 307.82 
RUN 6 304.8 326.7 314 326.5 319.41 
RUN 7 320.6 321.2 311.4 333.1 324.93 
 
(350 RPM / 4 inch per min traverse / 0.7 inch shoulder) 
 
 
 
 
 
 
 
 
 
154 
 
Factor level combination 9: 450 RPM / 7 inch per min traverse / 0.5 inch shoulder  
Entry Data 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 342.5 353.9 322 334.2 342.33 363 316.4 318.1 382.9 335.3 
RUN 2 326.8 347.9 317.2 327.8 343.63 356.2 310.9 315.2 388.4 332.6 
RUN 3 339.4 359.9 324.1 342.6 350.83 371.9 310.7 324.2 396.4 336.6 
RUN 4 330.3 340 318.4 325.8 354.9 363.2 312.3 321.8 377.9 330.3 
RUN 5 343.8 355.5 329.1 332.2 353.81 365 316.1 326.2 388.2 334.5 
RUN 6 336.9 354.9 319.9 330.9 357.22 367.9 313.7 326.1 382.5 334.3 
RUN 7 337.2 348.6 318 329 349.34 359.6 313.5 326.2 390.4 333 
 
Traverse Data 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 320 327.8 304.3 324.5 316.65 324.4 308.8 316.1 
RUN 2 316.2 325.8 311.8 323.6 319.51 323.3 309.5 317.3 
RUN 3 326.5 331.2 316.1 321.7 313.7 323.9 306.1 317.1 
RUN 4 320.9 323.4 305.6 321.9 314.79 322.1 310.5 311.6 
RUN 5 317.1 326.3 312.3 314.6 318.51 322.7 311.2 312.5 
RUN 6 324.1 326.6 308.8 325 321.92 326.4 311.2 316.6 
RUN 7 319.7 325.8 312 314.6 318.68 323 310.9 319.2 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 318.7 329.1 316.5 327.4 319.41 324.1 280.6 325.5 
RUN 2 316.5 329.3 314.7 329.3 311.25 324.3 280.9 326.7 
RUN 3 325.3 334.3 323.3 333.1 315 326 284.5 329.2 
RUN 4 315.9 326.2 308.2 325.3 315.77 321.3 280.4 322.4 
RUN 5 320.5 329.7 317.8 327.5 320.78 323.3 281.3 323.8 
RUN 6 323.9 329.7 321 328.6 313.14 325.6 280.7 323 
RUN 7 320.3 330.3 317.4 328.6 310.97 325.9 281.5 328.7 
 
(450 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
155 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 323.9 330.2 320.5 323 
RUN 2 327 333.7 321.1 325.3 
RUN 3 330.9 330.7 323.2 324.2 
RUN 4 323.1 329.2 319.1 322.2 
RUN 5 326.2 333.1 322 323.7 
RUN 6 328.4 333.3 317.4 319.7 
RUN 7 329.2 335.8 323.3 327.4 
 
Exit Data: 
 
  Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 290.9 297 297.9 310.1 296.34 
RUN 2 296.1 296.3 299.9 312.5 304.99 
RUN 3 290.2 306.9 306.3 313.6 299.09 
RUN 4 289.1 300 301.1 310.2 300 
RUN 5 287.1 300.5 299.8 307.1 299.9 
RUN 6 285.8 296.5 298.3 313 294.42 
RUN 7 291.8 299 301.7 311.3 304.36 
 
(450 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
156 
 
Factor level combination 10: 450 RPM / 4 inch per min traverse / 0.5 inch shoulder  
Entry Data 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 362.1 376.2 338.2 349.5 366.73 379.7 323.4 325.6 404.9 345.2 
RUN 2 351.1 370.2 326.1 348.4 358.42 369.6 311.1 331.5 391.1 321.9 
RUN 3 353.7 359.4 342.8 341.5 362.39 378.1 312.8 330.6 391.5 336.5 
RUN 4 355.2 373.8 337 362.3 356.44 375 321.6 335.2 415.4 349.7 
RUN 5 357.9 370.7 341.3 349.9 367.19 380 319.5 330.8 398.6 347.7 
RUN 6 350.6 361.5 335.3 347 362.89 374.2 315.8 318.9 395.2 335.7 
RUN 7 358.3 369.6 329.7 347.5 362.21 360.8 312.3 318.7 396.2 347.5 
 
Traverse Data 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 330.4 345.3 324 333.7 336.1 339.8 339.3 339.9 
RUN 2 309 344.5 295.6 323.9 315.75 347.8 328.9 347.4 
RUN 3 328.5 341.9 309.1 324 323.75 342 329.6 341 
RUN 4 339.4 350.4 307.6 345.7 319.13 349.6 326.5 346.7 
RUN 5 330.3 353.1 318.8 328.9 338.69 342.5 339.5 343.2 
RUN 6 332.6 345 314.5 326.7 330.98 342.3 329.7 342.1 
RUN 7 332.2 334.7 310.3 325.6 331.27 335.6 336.1 336.9 
 
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 333.5 346.2 322.6 342.4 334.07 337.2 297.5 349 
RUN 2 328.6 344.6 326.3 342.2 310.57 344.9 281 322 
RUN 3 328.3 342.2 323.1 336.7 323.12 341.3 291.5 340.7 
RUN 4 346.5 352.6 335.8 349.8 327.23 349.1 302.4 359.5 
RUN 5 334 354.5 323.3 348.1 326.31 341.3 300.7 352.5 
RUN 6 331.6 344.7 328.7 331.6 330.65 343.2 291.3 343.7 
RUN 7 329.4 341.9 327 336.1 330.51 333.7 289.1 338.5 
 
(450 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
157 
 
  
Group 
1 
Group 
2 
Group 
3 
 Group 
4 
 U3 U4 U5 U6 
RUN 1 348.1 347.8 334.7 329.8 
RUN 2 310 312.6 307.1 309.1 
RUN 3 337.6 328.3 324.7 340.9 
RUN 4 352.6 348.9 344 337 
RUN 5 348.5 336.8 334.1 343.1 
RUN 6 338.3 332.9 324.8 332.2 
RUN 7 335.2 333.6 323.7 319.5 
 
Exit Data 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 304.2 319.5 319.8 326.6 311.56 
RUN 2 302.5 305.5 307.6 334.8 297.11 
RUN 3 308.4 314.4 312.6 322.1 319.59 
RUN 4 298.1 311.1 314.7 336.6 298.27 
RUN 5 307.5 317.7 318.4 330.5 317.99 
RUN 6 300.6 320.1 320.1 320.3 309.67 
RUN 7 302.5 318.4 311.8 329 303.69 
 
(450 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
158 
 
Factor level combination 11: 350 RPM / 7 inch per min traverse / 0.5 inch shoulder  
Entry Data 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 312.5 322.6 284.4 304.9 323.14 343.1 295.1 299.9 344.2 294.6 
RUN 2 318.3 329.7 294.8 300.2 329.32 332.2 278.5 305.3 359.2 298.8 
RUN 3 326 331.4 288.8 303.9 317.18 337.5 281.9 309.9 354.6 303.1 
RUN 4 309.3 333.6 297 303.2 318.58 340 293 299.6 357.7 313.7 
RUN 5 310.7 331.3 289.1 307.5 317.35 337.9 294.6 303.2 354.7 306.7 
RUN 6 310.8 324.2 298.7 298.9 331.45 336.1 288.2 299.9 353.4 303.3 
RUN 7 304.8 332.2 288.2 309.8 331.21 346.4 295.3 307.1 356.4 313.2 
 
Traverse Data 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 279 292.3 277.1 292.2 293.91 301 289.3 300.7 
RUN 2 283.4 298.9 271.2 279.6 285.14 306.1 288.7 303.7 
RUN 3 282.9 294.3 277.9 293.9 281.74 312.4 290.9 298.5 
RUN 4 294.7 305 281 296.1 297.06 306.7 300.2 304.6 
RUN 5 286.3 301.5 284.6 300 301.39 309.7 296.9 304.1 
RUN 6 285.3 291.4 277.5 292.5 289.47 301 286 300.1 
RUN 7 290.3 305.3 286.2 296.7 288.82 296.3 285.5 305.9 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 292.2 304 279.6 300.7 293.08 309.5 269.4 298.1 
RUN 2 287.6 310.6 283.9 298.3 295.57 304.6 277.1 300.9 
RUN 3 293.9 306.4 280.7 293.3 282.66 300.3 258.8 305.5 
RUN 4 298.5 306.5 288.5 304.8 297.45 304.8 274.2 306.3 
RUN 5 300.9 313.9 288.2 301.4 293.21 309.7 272.7 314.2 
RUN 6 297 304.2 281 292.5 290.93 312.5 263.8 301 
RUN 7 302.7 309 289.6 299.5 290.64 314.9 275.9 314.7 
 
(350 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
159 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 291.8 297.2 297.8 288.4 
RUN 2 289.6 299.9 302.6 291.5 
RUN 3 296.1 295.8 299.7 288 
RUN 4 306.6 310.4 301.9 294.5 
RUN 5 301 311 303.9 286.8 
RUN 6 286.8 309.6 294 294 
RUN 7 305.5 313.3 314.6 304 
 
Exit Data 
 
  Group 5 Group 5 
Group 
5 
  S13 S14 P13 P14 U7 
RUN 1 275.1 278.5 261 287.6 279.31 
RUN 2 265 276.3 267.9 280.2 290.55 
RUN 3 263.1 276.3 265.3 276.6 274.75 
RUN 4 275.2 281.8 274.5 282.8 291.9 
RUN 5 275.3 282.9 268 278.4 288.39 
RUN 6 268 279.7 262.4 285.2 278.77 
RUN 7 266.1 283.9 276 281.9 292.69 
 
(350 RPM / 7 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
160 
 
Factor level combination 12: 350 RPM / 4 inch per min traverse / 0.5 inch shoulder  
Entry Data: 
  Entry 1 Entry 2 Entry 1 Entry 2 
Entry 
1 
Entry 
2 
 S1 S2 S3 S4 P1 P2 P3 P4 U1 U2 
RUN 1 332.8 349.8 316 329.7 338.14 349.8 299 320.2 366.5 324.4 
RUN 2 333.8 345.4 318.1 328.6 325.74 349.6 301.6 322 379.9 332.5 
RUN 3 350 352.5 332.5 334.2 335.9 359.7 317.6 319.3 375.6 339.3 
RUN 4 327.4 353.1 317.9 330.9 340.77 358.7 303.4 320.5 369.9 322.6 
RUN 5 332.1 344.6 320.2 330.6 349.96 352.5 303.6 323.8 379.5 328.7 
RUN 6 333.7 344.8 318.4 327.4 336.36 348.1 301.7 324.8 364.8 332.3 
RUN 7 321.6 343.5 319.4 322.3 341.88 373.4 301.8 314.1 372.8 325.5 
 
Traverse Data: 
 
  Group 1 Group 2 Group 3  Group 4 
 S5 S6 S7 S8 S9 S10 S11 S12 
RUN 1 301.2 324.6 304.1 317.3 310.66 320 309.5 317.4 
RUN 2 308.7 325.6 303.1 319 308.9 319.6 307.8 316.5 
RUN 3 326.5 330.4 313.2 318.7 310.57 326.4 306.4 327.1 
RUN 4 306.1 323.6 304.6 320.6 312.68 322.8 302.4 319.5 
RUN 5 309.9 326.1 306 320.2 308.46 318.6 303.4 320.4 
RUN 6 310.5 331 306.2 322 291.05 312.6 292.6 313.9 
RUN 7 305.3 316.2 300.5 313.9 309.67 322.6 308.2 311.8 
? ????? ?? ?
  Group 1 Group 2 Group 3  Group 4 
 P5 P6 P7 P8 P9 P10 P11 P12 
RUN 1 305.8 325.3 308 321.5 311.18 327.3 278.7 323.8 
RUN 2 309 326.8 306.6 324.2 308.72 318.2 277.9 322.2 
RUN 3 324.2 335 321.4 328.7 320.87 327.9 280 339 
RUN 4 309.6 325 308.4 323.4 311.59 319.9 275.8 323.1 
RUN 5 311.7 328.1 310 325.8 308.95 319.1 276.7 326.6 
RUN 6 307.7 322.7 315.8 321.9 312.21 323.6 264.1 323.3 
RUN 7 308 318.2 306.9 317.2 310.78 323.8 275.9 319.6 
 
(350 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
161 
 
  
Group 
1 
Group 
2 
Group 
3 
 
Group 
4 
 U3 U4 U5 U6 
RUN 1 324.9 313.8 308.9 314.4 
RUN 2 325.2 320.2 314.1 322 
RUN 3 337.9 333.9 327 318.4 
RUN 4 326.5 320.2 315.6 320.9 
RUN 5 328.8 323.3 317.4 318.6 
RUN 6 329.7 316.1 320.5 302.6 
RUN 7 316.1 321.5 312.5 320.8 
 
Exit Data 
 
  Exit Exit Exit 
  S13 S14 P13 P14 U7 
RUN 1 287.9 303.5 290.2 305.9 301.04 
RUN 2 290.9 299.3 292.8 305.3 303.85 
RUN 3 303.1 307.4 309.5 323.3 303.99 
RUN 4 284.7 301.4 287.8 302.5 300.52 
RUN 5 287.1 306.6 291 307.7 300.6 
RUN 6 279.3 296.5 287.9 301.4 289.08 
RUN 7 290.2 303.3 293.6 300.6 299.24 
 
(350 RPM / 4 inch per min traverse / 0.5 inch shoulder) 
 
 
 
 
 
 
 
 
 
162 
 
APPENDIX C: 
Direct temperature plots: Average of the recorded peak temperatures of the 7 experiments for all 
factor level combinations. 
           Three different shoulders, with 2 different RPMs and 2 different feed rates were used in 
this experiment. That gives a total of 12 different factor level combinations. For each factor level 
combination, 7 experiments were done. For each experiment 7 sets of thermocouples (Entry 1, 
Entry 2, Group 1, Group 2, Group 3, Group 4 and Group 5) were used to record temperatures as 
discussed earlier in chapter 3. Recorded temperatures of all 7 experiments for 12 different 
parameters are given in Appendix A.  Averages of the peak temperatures of the 7 experiments 
for all factor level combinations are given in this appendix.  
              In the plots shown below, except for point 1(Entry 1) which acts as entry point and point 
7 (Group 5) which acts as exit point, temperatures at all other sets are more or less the same. 
Average of all advancing side and average of all retreating side temperatures in these 5 sets 
would give an appropriate value for advancing and retreating side temperatures for that given 
factor level combination. 
 
163 
 
 
 
164 
 
 
 
 
 
 
 
165 
 
 
 
 
 
 
166 
 
 
 
 
 
 
167 
 
 
 
 
 
168 
 
 
 
 
 
 
 
169 
 
APPENDIX D: 
Calculated temperature plots: Average of the calculated peak temperatures using 
thermodynamics equation of the 7 experiments for all factor level combinations.  
           Three different shoulders, with 2 different RPMs and 2 different feed rates were used in 
this experiment. That gives a total of 12 different factor level combinations. For each factor level 
combination, 7 experiments were done. For each experiment 7 sets of thermocouples (Entry 1, 
Entry 2, Group 1, Group 2, Group 3, Group 4 and Group 5) were used to record temperatures as 
discussed earlier in chapter 4. Recorded temperatures of all 7 experiments for 12 different 
parameters are given in Appendix A. One dimensional heat transfer equation was used to 
calculate the temperature at the tool tip from these recorded temperatures. The calculated 
temperatures are given in Appendix B. The calculated averages of the peak temperatures of the 7 
experiments for all factor level combinations are given in this appendix.  
              In the plots which are going to follow , except for point 1(Entry 1) which acts as entry 
point and point 7 (Group 5) which acts as exit point, temperatures at all other sets are more or 
less the same. Average of all advancing side and average of all retreating side temperatures in 
these 5 sets would give an appropriate value for advancing and retreating side temperatures at the 
tool tip for that given factor level combination. 
 
170 
 
 
 
 
171 
 
 
 
 
 
 
 
 
 
172 
 
 
 
 
 
 
173 
 
 
 
 
 
 
 
174 
 
 
 
 
 
 
 
175 
 
 
 
 
 
 
176 
 
APPENDIX E: 
Analysis of variance (ANOVA) is done for each responses involved in the experiment, to find 
the impact factors on the responses.  In statistics, analysis of variance is a collection of statistical 
models, and their associated procedures, in which the observed variance is partitioned into 
components due to different explanatory variables.  A summary of the largest impact factors are 
given in the results chapter.   ANOVA tables for every response involved in this experiment are 
given in this appendix.  STATEASE was the software used to do the analysis of variance.  The 
responses involved with the experiment are, 
Shoulder responses: S1, S2, S3? S14 
Pin responses: P1, P2, P3? P14 
Under the pin responses:  U1, U2? U7 
           The ANOVA tables of every response that are to follow are the screen shots of the result 
obtained from STATEASE software.  The most important value in the tables below is the F 
value.  F value shows the significance of a factor on the response involved.  The factor with the 
highest F-value has the strongest impact on the response.   
177 
 
 
 
178 
 
 
 
179 
 
 
 
180 
 
 
 
181 
 
 
 
182 
 
 
 
183 
 
 
 
184 
 
 
 
185 
 
 
 
186 
 
 
 
187 
 
 
 
188 
 
 
 
189 
 
 
 
 
190 
 
 
 
191 
 
 
 
192 
 
 
 
193 
 
 
 
194 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
195 
 
APPENDIX F: 
Simulation Screen Shots: 
Step 1: Model 
Simple drawing options were used to create the model with the actual dimensions of the 
instrument. The instrument is modeled in such a way that the slot for the aluminum stock already 
bears the material. 
 
Step 2: Constants 
 
196 
 
Constants should be assigned and the values of the constants are to be declared. 
Step 3: Boundary expressions 
Boundary expressions are the heat equations that apply to the surface of the material that is been 
used. This physically allows the user to select the surface where the equation is to be applied. 
 
Step 4: Subdomain settings 
 
Subdomain settings allows the user to select the material that is been used in the model. Multiple 
materials can be selected and their thermal properties are entered here. 
197 
 
Step 5: Boundary Settings 
Boundary settings allow the user to apply the boundary expression to the surface that is selected. 
In the below figure, for the 51
st
 boundary, the heat equation used is q_shoulder. The expression 
q_shoulder is already defined in boundary expression. 
 
Step 6: Meshing 
As discussed in Chapter 9 (Simulation of the process), Normal mesh is chosen for the process. 
The mesh details and the number of elements are given in Chapter 9. 
198 
 
 
Step 7: Solving 
The applied equation is solved with the model designed to get a heat distribution along the 
model. 
 
 
199 
 
Results: Advancing side shoulder temperature: Peak temperature = 385.7C  
 
Retreating side shoulder temperature: Peak temperature = 381.6 C 
 
200 
 
Advancing side pin temperature: Peak temperature = 320.6 
 
Retreating side pin temperature: Peak temperature = 315.3 C 
 
201