Professional Documents
Culture Documents
M120427me Thesis Full Report
M120427me Thesis Full Report
M120427me Thesis Full Report
INTRODUCTION
Resistance spot welding (RSW) is a process in which contacting metal
sheets are joined by the heat obtained from resistance to electric current. Workpieces are held together under pressure exerted by electrodes. Typically the sheets
are 0.5 to 3 mm thickness range. The process uses two shaped copper
alloy electrodes to concentrate welding current into a small "spot" and to
simultaneously clamp the sheets together. Forcing a large current through the spot
will melt the metal and form the weld. The attractive feature of spot welding is
that a lot of energy can be delivered to the spot in a very short time
(approximately 10100 milliseconds), that permits the welding to occur without
excessive heating of the remainder of the sheet. The amount of heat (energy)
delivered to the spot is determined by the resistance between the electrodes and
the magnitude and duration of the current. The amount of energy is chosen to
match the sheet's material properties, its thickness, and type of electrodes.
Applying too little energy will not melt the metal or will make a poor weld.
Applying too much energy will melt too much metal, eject molten material, and
make a hole rather than a weld. Another feature of spot welding is that the energy
delivered to the spot can be controlled to produce reliable welds.
Where Q is the heat generation (J), I is the current (kA), R is the resistance of the
work piece (ohm-cm) and t is the duration of current (sec).
The secondary circuit of a resistance welding machine and the work being
welded constitute a series of resistances. The total resistance of the current path
affects the current magnitude. There are, in effect, at least seven resistances
connected in series in a weld that account for the temperature distribution and the
sum of them is expressed as R as shown in Fig 1.1. Line diagram for resistance
spot welding setup is as shown in Fig 1.2.
some elements and lacks others, while the outermost layer of the crystal is just the
opposite. This process is called micro segregation (3).
Solidification occurs when the liquid nugget reaches the liquidus
temperature of the alloy and there is a net heat loss in the liquid i.e., the heat
dissipated from the liquid is greater than that into the liquid. Under proper welding
conditions, the water cooled electrodes act as a large heat sink during welding. The
parent sheet metal also transfers heat from the sides. Partially melted grains in the
mushy zone at the nugget, heat-affected zone (HAZ) boarders serve as nuclei for
solid grains to grow. Further cooling results in columnar grains in directions
approximately normal to the fusion line. The remaining molten metal in the central
portion of the nugget solidifies last and forms equiaxed grains due to cooling when
the liquid volume is considerably small after its surroundings are solidified.
Depending on welding schedules and other conditions, various structures can be
formed in a weld nugget.
not normally recommended at this may impair the quality of the weld. Adjust the
tap switch so that proper welding current can be obtained with the percent current
set between seventy and ninety percent. The weld current should be kept as low as
possible. When determining the current to be used, the current is gradually
increased until weld spatter occurs between the metal sheets. This indicates that the
correct weld current has been reached.
storage element isn't needed. The switch causes the stored energy to be dumped
into the welding transformer. The welding transformer steps down the voltage and
steps up the current. An important feature of the transformer is it reduces the
current level that the switch must handle. The welding electrodes are part of the
transformer's secondary circuit. There is also a control box that manages the switch
and may monitor the welding electrode voltage or current. The resistance presented
to the welder is complicated. There is the resistance of secondary winding, the
cables, and the welding electrodes. There is also the contact resistance between the
welding electrodes and the work piece. There is the resistance of the work pieces,
and the contact resistance between the work pieces. At the beginning of the weld,
the contact resistances are usually high, so most of the initial energy will be
dissipated there. That heat and the clamping force will soften and smooth out the
material at the electrode-material interface and make better contact (that is, lower
the contact resistance). Consequently, more electrical energy will go into the work
piece and the junction resistance of the two work pieces. As electrical energy is
delivered to the weld and causes the temperature to rise, the electrodes and the
work piece are conducting that heat away. The goal is to apply enough energy so
that a portion of material within the spot melts without having the entire spot melt.
The perimeter of the spot will conduct away a lot of heat and keep the perimeter at
a lower temperature. The interior of the spot has less heat conducted away, so it
melts first. If the welding current is applied too long, the entire spot melts, the
material runs out or otherwise fails, and the "weld" becomes a hole.
1.6.4 Clamping
Welding times are often very short, which can cause problems with the
electrodes. They cannot move fast enough to keep the material clamped. Welding
controllers will use a double pulse to get around this problem. During the first
pulse, the electrode contact may not be able to make a good weld. The first pulse
will soften the metal. During the pause between the two pulses, the electrodes will
come closer and make better contact.
1.6.5 Fields
During spot welding, the large electric current induces a large magnetic
field, and the electric current and magnetic field interact with each other to produce
a large magnetic force field too, which drives the melted metal to move very fast at
a velocity up to 0.5 m/s. As such, the heat energy distribution in spot welding could
be dramatically changed by the fast motion of the melted metal. The fast motion in
spot welding can be observed with high speed photography.
1.6.6 Safety
It is common for a spray of molten metal droplets (sparks) to be ejected
from the area of the weld during the process. While spot welding does not generate
UV light as intensely as arc welding, eye protection is required. Welding goggles
with a 5.0 shade are recommended.
1.8 APPLICATIONS
Spot welding is typically used when welding particular types of sheet
metal, welded wire mesh or wire mesh. Thicker stock is more difficult to spot weld
because the heat flows into the surrounding metal more easily. Spot welding can be
easily identified on many sheet metal goods, such as metal buckets. Aluminium
alloys can be spotwelded, but their much higher thermal conductivity and electrical
conductivity requires higher welding currents. This requires larger, more powerful,
10
and more expensive welding transformers. Perhaps the most common application
of spot welding is in the automobile manufacturing industry, where it is used
almost universally to weld the sheet metal to form a car. Spot welders can also
be completely automated, and many of the industrial robots found on assembly
lines are spot welders (the other major use for robots being painting). Spot welding
is also used in the orthodontist's clinic, where small-scale spot welding equipment
is used when resizing metal "molar bands" used in orthodontics.
Another application is spot welding straps to nickelcadmium or nickel
metal hydride cells to make batteries. The cells are joined by spot welding thin
nickel straps to the battery terminals. Spot welding can keep the battery from
getting too hot, as might happen if conventional soldering were done. Good design
practice must always allow for adequate accessibility. Connecting surfaces should
be free of contaminants such as scale, oil, and dirt, to ensure quality welds. Metal
thickness is generally not a factor in determining good welds.
11
To study the effect of parameters weld time, weld current, electrode force
and constant squeeze time on tensile - shear strength, nugget diameter and
heat affected zone.
To study the failure analysis of the resistance spot welded joints during the
tensile-shear strength test.
To study the analysis of variance (ANOVA) for each response and develop
the mathematical model equations.
To study the optimization technique for predicting the maximum tensileshear strength, optimum nugget diameter and minimum heat affected zone.
12
CHAPTER 2
LITERATURE RIVIEW
To study the resistance spot welding process, parameters and the research
works done in this field, a detailed literature review is carried out as below.
Aslanar S et. al. [1] investigated the effect of nucleus diameter, heights of
nucleus, and nucleus size on mechanical properties i.e. tensile shear and tensile
peel strength in electrical resistance spot welding of galvanized chromided micro
alloyed steel sheets.
Aslanar S et. al. [2] investigated the Effect of welding current on
mechanical properties i.e. tensile shear and tensile peel strength of galvanized
chromided steel sheets in electrical resistance spot welding. Welding periods were
chosen as 5, 10, 12, and 15 cycles and also welding currents were increased from 4
kA up to 12 kA by rise of 1 kA. The electrode force was kept constant at 6 kN.
During these tests three types failure are occurred in sheets 1.Separation,
2.Knotting, and 3.Tearing. The results showed that the tensile- shear strength
increases with increasing weld current, the tensile- peel strength increases sharply
from 6 to 9 kA and reaches max value at 10 kA.
Luo Yi e.t al. [5] Study the mathematical regression models between
process parameters and nugget diameter, penetration ratio and tensile-shear
strength of spot welds. The method of non-linear multiple orthogonal regression
assembling design was used. Mathematical relations between experimental indexes
and process parameters can be described by the full equation of quaternionic
quadratic regression. The effects of welding process on nugget size and mechanical
properties are not varying with simple linear curve but with non-linear
characteristics. The effects of interactions must not be ignored. And the means of
contour line figures can be introduced to describe the relationship between the
interactions and the quality indexes ambiguously.
Vural, M et. al. [6] Investigated the effect of nugget diameter on the fatigue
strength of resistance spot welded joints of galvanized steel and austenitic stainless
steel (AISI 304) welded as lap joints and explained the endurance limit of the
similar steels sheet combination is higher than that different steel sheet the reason
13
for this result is heat unbalance between the steel sheets which occurs during spot
welding operations of steel sheets having different material combinations.
S.M. Hamidinejad et. al. [7] In this study, the resistance spot welding
(RSW) process of the galvanized interstitial free (IF) steel sheets and galvanized
bake hardenable (BH) steel sheets, used in the manufacturing of car bodies, has
been modeled and optimized. The quality measure of a resistance spot welding
joint is estimated from the tensileshear strength. They concluded that there is a
non-linear relationship between the welding parameters and the tensile shear
strength of the RSW joints and explained regression analysis, artificial neural
networks and genetic algorithm are used for optimization and explained failures of
the joints.
Hessamoddin Moshayedi et. al. [8] worked on numerical and experimental
study of nugget size growth in resistance spot welding of austenitic stainless steels.
In this study, two- dimensional finite element simulation of resistance spot welding
process was performed using fully coupled electricalthermal and incrementally
coupled thermalmechanical analysis on SS304L sheets to predict weld nugget
formation through temperature distributions at different welding cycles and current
intensities. Finite element modeling was used to investigate the effects of welding
time and welding current on nugget size. When the welding current exceeds a
critical value for nugget formation (6 kA for this work), it causes a rapid growth of
nugget Good agreement has been obtained between the predicted results and the
measured data, such that the maximum errors have been around 13%.
J.B. Shamsul et. al. [9] studied the influence of welding current on nugget
size and hardness distribution in the plates of Austenitic Stainless Steel (AISI 304),
which was placed as a lap joint and spot welded under varied welding current
conditions and electrode force and weld time. The results have shown that
increasing welding current increased the nugget size. The nugget size does not
influence the hardness distribution. Also increasing welding current does not
increase the hardness distribution.
Dursun Ozyurek et. al. [10] investigated the effect of the different welding
medium and welding current on the resistance spot weld quality of 304L stainless
steel and for this purpose; samples are welded in nitrogen at different currents and
then cooled in atmosphere. Samples welded in atmosphere at different currents are
14
then compared with the above samples to study the influence of different welding
medium and current.
M. Hamedi et. al. [11] worked on numerical study of nugget formation in
resistance spot welding. They concluded that as the welding process continues
(welding time increases), temperature of sheets interface rises quickly until this
area melts and nugget forms. After the formation of nugget, rate of temperature rise
is reduces. At the end of welding time, as current switches off, the weldment starts
to cool dawn. And at low electrode forces and high welding currents, the formed
fusion zone is large.
Nizamettin K. et. al. [12] has focused his study on the influence of welding
parameters on the joint strength of resistance spot-welded titanium sheets. The
results indicated that increasing current time and electrode force increased the
tensile shear strength and the joint obtained under the argon atmosphere gave better
strength. Hardness measurement results showed that welding nugget gave the
highest hardness. The argon gas used during the welding process was seen to have
no influence on the hardness values.
Emin Bayraktar et. al. [13] have contributed their research on the selection
of optimal welding conditions for getting good weld attributes and developed new
grade steels for automotive applications. The study is based on impact tensile
testing to spot welding sheets.
Bouyousfi et. al. [14] Have studied the effect of spot welding process
parameters (are intensity, welding duration and applied load) on the mechanical
properties and characteristics of the spot joints between two stainless steel sheets
(304 ASS) having the same thickness. Micro hardness and tensile test results have
shown that the weld resistance is important and highly correlated to the value of the
process parameters especially the applied load. The applied load seems to be the
control factor of the mechanical. Characteristics of weld joint compared to the
welding duration and the current intensity.
Oscar Martin et. al. [15] predicted the quality level of a resistance spot
welding (RSW) joint of 304 austenitic stainless steel (ASS). It is estimated from its
tensile shear load bearing capacity (TSLBC). The quality levels are set by
ultrasonic non-destructive testing. The weld nugget grows at the expense of heat
input that increases with increasing WT and WC. The hardness values of the weld
15
nugget do not decrease and even may be higher than that of HAZ and base metal
due to the strain hardening (revealed by the presence of slip bands in the austenite
cast microstructure of the weld nugget), caused by EF (electrode force) during
holding time.
Majid Pouranvari [16] Failure mode of AISI304 resistance spot welds is
studied under quasi static tensile-shear test. He concluded that failure location for
AISI 304 RSW in pull-out failure mode is at HAZ, adjacent to the weld nugget
Low fusion zone hardness to failure location hardness ratio increases the tendency
of spot weld failure to occur in the interfacial failure mode during the tensile-shear
test. Metallurgical characteristics of welds should be considered to predict and
analyze the spot weld failure mode more precisely.
With the objectives as summarized in (1.9), thus a study to fill the gaps seen
in literature review and for the setup newly installed in the institute, a standard to
be followed for the materials which are spot welded is to be formulated with
experimental validation.
16
CHAPTER 3
EXPERIMENTAL PROCEDURE AND SETUP DETAILS
3.1 MATERIAL
The group of alloys which today make up the family of stainless steels had
their beginning in 1913 in Sheffield, England; Harry Brearley was trying a number
of alloys as possible gun barrel steels, and noticed that samples cut from one of
these trial heats did not rust and were in fact difficult to etch. When he investigated
this curious material, it contained about 13% chromium. This lead to the
development of the stainless cutlery steels for which Sheffield became famous.
Coincidentally development work was also being carried out in France at about the
same time which culminated in the production of the first austenitic stainless steels.
The worldwide consumption of stainless steel is increasing. There is
growing demand from the building and construction industry where stainless steel
is used for its attractive appearance, corrosion resistance, low maintenance and
strength. Many other industries are adopting stainless steel for similar reasons as
well as the fact that it does not need to be treated, coated or painted when put into
service, despite the fact that it is more expensive than plain carbon steels. Evidence
to support this comes from the household appliances manufacturers, where there
are an ever increasing number of items, traditionally known as white goods being
manufactured in stainless steel.
3.1.1 Stainless steel (304)
Stainless steel types 1.4301 and 1.4307 are also known as grades 304 and
304L respectively. Type 304 is the most versatile and widely used stainless steel. It
is still sometimes referred to by its old name 18/8 which is derived from the
nominal composition of type 304 being 18% chromium and 8% nickel. Type 304 Stainless steel is an austenitic grade that can be severely deep drawn. This property
has resulted in 304 being the dominant grade used in applications like sinks and
saucepans. Type 304L is the low carbon version of 304. It is used in heavy gauge
components for improved weldability. Some products such as plate and pipe may
be available as dual certified material that meets the criteria for both 304 and
17
304L. 304H, a high carbon content variant, is also available for use at high
temperatures. Austenitic Cr-Ni stainless steel. Better corrosion resistance than Type
302. High ductility, excellent drawing, forming, and spinning properties.
Essentially non-magnetic, becomes slightly magnetic when cold worked. Low
carbon content means less carbide precipitation in the heat-affected zone during
welding and a lower susceptibility to inter-granular corrosion.
3.1.2 Properties
Resistance to corrosion
Resistance to oxidation
Easy of fabrication
Excellent formability
Beauty of appearance
Easy of cleaning
18
3.1.3 Welding
The austenitic stainless steels are considered to be the most weldable of
the high-alloy steels and can be welded by all fusion and resistance welding
processes. Two important considerations in producing weld joints in the austenitic
stainless steels are: 1) preservation of corrosion resistance, and 2) avoidance of
cracking.
A temperature gradient is produced in the material being welded which
ranges from above the melting temperature in the molten pool to ambient
temperature at some distance from the weld. The higher the carbon levels of the
materials being welded, the greater the likelihood that the welding thermal cycle
will result in the chromium carbide precipitation which is detrimental to corrosion
resistance. To provide materials at the best level of corrosion resistance, low
carbon material (Type 304L) should be used for material put in service in the
welded condition.
Table 3.1 Chemical composition of the test material (WT %)
C
Mn
P
S
0.08
2
0.045
0.03
Si
Cr
Ni
N
0.75
18-20
8-10.5
0.10
19
505 MPa
215 MPa
190 200 GPa
0.29
8 g/cc
7.2e-.005 ohm-cm
16.2 W/m-k
1400 - 1455 0C
70
20
Model
SP10P
10
460
220
6
6-position off. Load Tap
Current Control
3-0.2
Switch
Water
100-75
40
10
420
21
All S Series machines have an easy-to-read backlit liquid crystal display that can be
switched between a numerical and graphical display. All data shown on this display
is obtained in real time, with the auto ranging graphical display showing the test
curve of the specimen under test. The control unit features dedicated keys for
moving the crosshead up, down, stop, performing the test, as well as keys for load
and extension tare and crosshead return. The control panel also features an
alphanumeric keypad to allow input of test conditions, test data and the formatting
of the test report. The control unit can retain up to five separate test routines for
easy and rapid recall. While powerful as stand-alone units, these machines can be
enhanced by direct connection of a printer through which comprehensive test
reports and high resolution graphs can be quickly obtained. The S Series of testers
are also designed for users all over the world. An optional language module can be
plugged into the control panel and all data on the backlit LCD will be shown in the
selected language.
22
Load cells
10000N
405 mm
Rapid change, low profile Z type load cells
with digital encoding for automatic
recognition
and scaling available 10kN, 5kN, 2.5kN,
1kN, 500N, 250N, 100N, 50N, 10N, 5N
Measurement direct from ball screw fully
auto scaling of single measurement range
1100 mm
0.001 to 1000 mm/min
5 kN
500 mm/min
0.001 to 1000 mm/min
0.001 to 750 mm/min
1600 x 650 x 450
115 kg
3.4.3 Specifications
23
24
provided by an X-Cite 120Q metal halide lamp, a quiet and stable source of
fluorescence excitation.
25
Revolving
Hand switch
Control box
Motorized
universal
condenser
100X
26
Nugget Diameter, Heat affected zone and Tensile-shear strength in spot welding
can be summarized as shown in the ishikawa diagram. The underlined parameters
in the Fig 3.4 represent the process parameters which are taken as the most
influencing parameters on the resistance spot welding.
27
Unit
Levels
1
Weld current
kA
X1
Electrode force
kg/cm2
X2
2.5
Weld time
Cycle
X3
10
15
20
During cutting of sheets observed some burr formation on the edges of the
sheet.
First supply the coolant for electrodes and switch on the compressor for
supplying the air pressure for electrode force. The compressor capacity is
100-200 kgf.
Clean the sheets and electrodes by use of fine emery cloth for removing
dirt, oils, paint, scale and oxide etc.
Switch on the machine and set the input parameters as per the design table.
28
Thyristor contactor
Sparks
Weld signal
29
weld, such as penetration, also complement the nugget size and provide useful
information on the degree of adhesion. In fact, a weld is meant to contain all parts
of a weldment, such as the heat-affected zone (HAZ), inaddition to the nugget.
Another confusion is the use of button diameter and nugget diameter. As a nugget
and its size are usually revealed by metallographic cross sectioning, a nugget is
exposed for measuring its width as shown in fig 3.6.
30
The test specimen is obtained by lapping two sheets of metal and joining
them by a single weld. The ultimate strength of the test specimen and the nature of
fracture, whether by shear of the weld material or by tear of the parent material,
and whether a ductile or brittle fracture is obtained, should be recorded. When a
weld is under a tensile-shear loading, the weldment tends to rotate to align with the
gripped end, as in Fig 3.8.
Fig 3.8 The force of the welded joint during tensile-shear test (a) initial position
(b) nugget rotation.
In the tensile shear test, the welded joints first experience the shear stress,
which is parallel to the force direction, as shown in Fig 3.8a. In order to align with
the applied force direction, the nugget rotates first. Thus, the bending moment leads
to the tensile stress, which is perpendicular to the nugget, as shown in Fig 3.8b
31
In sample 12 knotting and pull out button failure occurred and in sample 20
and 3 tearing took place. Heat generation in between the two pieces was calculated
based on the Eqn (1.1) and setting the corresponding heat input to each run. The
heat generated for the three samples are 100%, 50%, and 75% of total heat input
respectively. The cases 7, 10 and 14 shown in fig 3.10 revealed the following:
32
CHAPTER 4
EXPRIMENTAL RESULTS AND DISCUSSIONS
4.1 INTRODUCTION
This chapter explains the analysis of variance (ANOVA) results for each
response of the process parameters in resistance spot welding and corresponding
plots. Model validation results and plots are also explained.
Randomize the order in which each set of conditions is run and collect the
data.
33
Sum of squares (SS) = sum of all the squared effects for each factor
Mean square Error (MSE) = pooled variance of samples within each level
MSA =
Eqn (4.1)
..Eqn (4.2)
34
4.2.2.3 F-Statistic
The F-statistic is a ratio of the between variability over the within
variability. If the F-statistic is large, it seems unlikely that the population means of
each level of factor A are truly equal.
The p-value is looked up in an F table and gives the likely hood of
observing an F statistic at least this extreme (at least this large) assuming that the
true population factor has equal level means. Thus, when the p-value is small (i.e.
less than 0.05 or 0.1) the effect size of that factor is statistically significant.
Fig 4.1 Typical force vs. displacement curve with their separation failure mode in
tensile-shear testing
35
Fig 4.2 Typical force vs. displacement curve with their pull out button failure
mode in tensile-shear testing
Fig 4.3 Typical force vs. displacement curve with their tearing failure mode in
tensile-shear testing
Analysis and discussion made on nugget diameter, heat affected zone and
tensile-shear strength are presented in this section. The resulting three level
factorial designs in three variables consist of the following 27 points. Experiments
36
were carried out in the run order obtained from the design-expert 7 software and
the measured response variables are as shown in the table 4.1. The response
variables are measured as mentioned in sections 3.8 and 3.9.
Table 4.1 Design matrix and measured values
S.No
X1
X2
X3
D (mm)
H (mm)
Ft (MPa)
2.5
10
2.82
0.72
174.4
2.5
10
2.38
0.41
103.8
15
3.71
1.08
348.4
20
3.09
2.20
382.8
15
3.36
1.80
356.8
2.5
20
4.33
1.12
323.6
10
2.72
0.43
124.5
10
3.88
0.94
291
2.5
20
4.95
1.41
375.2
10
15
3.24
0.59
170
11
10
2.88
0.84
206.1
12
10
4.18
0.88
285
13
15
3.19
0.73
170
14
20
3.99
0.77
252.9
15
20
4.79
1.22
334.8
16
2.5
10
5.20
0.75
288.6
17
2.2
20
3.79
0.87
244.8
18
2.5
15
3.23
0.59
166.8
19
10
2.72
0.58
121.6
20
20
3.65
1.14
383.2
21
10
3.72
0.76
223.2
22
2.5
15
5.02
1.12
356
23
20
3.94
0.86
242.7
24
2.5
15
4.17
0.97
276.9
25
20
5.00
1.29
346.4
26
15
4.42
0.80
298.5
27
15
4.19
0.90
268.5
37
SS
14.14
5.55
0.26
3.68
0.10
2.49
0.76
0.48
1.39
2.49
2.11
16.25
Dof
9
1
1
1
1
1
1
1
1
1
17
26
F
12.64
44.63
2.10
29.61
0.83
20.06
6.10
3.89
11.17
20.04
0.35
3.80
9.28
5.01
P> F
0.0001
0.0001
0.1651
0.0001
0.3762
0.0003
0.0244
0.0651
0.0039
0.0003
R squared
Adj R- squared
Pred R squared
Adj precission
Significance
Significant
Significant
Not significant
Significant
Not significant
Significant
Significant
Not significant
Significant
Significant
0.8700
0.8011
0.6920
13.212
The model F-value of 12.64 implies that the model is significant. There is
only a 0.01% chance that a model with F-Value of 12.64 could occur due to noise.
The values of Prob> F and less than 0.0500 indicate that model terms are
significant. In this case A, C, AC, A2, A2C, AB2 are significant model terms values
greater than 0.1000 indicate that the model terms are not significant. If there are
many insignificant model terms, without considering them, may improve the
model.
The Pred R-Squared of 0.6920 is in reasonable agreement with the Adj RSquared of 0.8011. Adeq-Precision measures the signal to noise ratio. A ratio
greater than 4 is desirable. In this model the ratio of 13.212 indicates an adequate
signal. So this model can be used to the design.
38
39
3. Step-wise regression
4.5.1 Backward regression method
Backward regression is considered as the robust method of model reduction
because this method starts with the full model and then the term which is not
satisfying the exit criterion is eliminated from the model. But in forward
regression technique, the terms which are satisfying the entry criteria will be added
to a base model which contains only highest significant factors. So the chance of
missing some significant interactions is more in this case compared to the
backward regression methodology step-wise and backward regression are same.
Alpha out value is a limiting value of probability that decides the exit criteria. In
this case the alpha out value is taken as 0.0500. The term P value mentioned above
is the common statistical parameter which corresponds to the F value of the term
and the degree of freedom. The software evaluates the P value of all the terms
involved in a proposed model. The details of the step-wise regression are:
Alpha out : 0.0500
P value > alpha out means insignificant terms.
Model reduction can be done by two ways, by eliminating the terms with
the p value greater than alpha out manually and also by software directly. To
reduce the model by software directly, go back and click on the model button. One
could manually deselect the insignificant terms observed from the ANOVA, but it
will be quicker with Design-Expert. Change the selection mode to step-wise. This
enables automatic model reduction by step-wise algorithm and it will give
significant terms in the model equation.
The criteria for elimination will be a probability value of 0.05 as specified in the
alpha out field. With a click on the ANOVA button, .as shown in figure 4.5, the
software gives the report on eliminated terms.
40
SS
3.95
0.33
0.010
1.16
0.20
0.10
0.072
7.407E-006
0.071
0.14
0.089
0.049
0.19
4.14
Dof
11
1
1
1
1
1
1
1
1
1
1
1
15
26
F
28.45
26.22
0.82
91.83
15.44
7.98
5.70
5.863E-004
5.63
11.12
7.04
3.89
41
P> F
0.0001
0.0001
0.3782
0.0001
0.0013
0.0128
0.0305
0.9810
0.0314
0.0045
0.0180
0.0673
Significance
Significant
Significant
Not significant
Significant
Significant
Significant
Significant
Not significant
Significant
Significant
Significant
Not significant
Std.dev
Mean
C.V %
PRESS
0.11
0.96
11.75
0.68
R squared
Adj R- squared
Pred R- squared
Adjprecission
0.9543
0.9207
0.8356
24.635
The Pred R-Squared of 0.8356 is in reasonable agreement with the Adj RSquared of 0.9207. Adeq Precision measures the signal to noise ratio. A ratio
greater than 4 is desirable. The ratio of 24.635 indicates an adequate signal.
4.6.1 Normal plot of residuals for heat affected zone
Constructing a normal probability plot of the residuals as shown in Fig 4.6
could make a typical check for normality assumption. ANOVA can be performed if
the data follows a normal distribution. Each residual is plotted against its expected
value under normality. If the residual distribution is normal, this plot will be a
straight line. In visualizing the plot, all the points are on the straight line. The
points on this plot line reasonably close to straight line, indicating that the residuals
are normally distributed.
Fig 4.6 Normal probability plot of residuals for heat affected zone
42
The Box-Cox plot for power transformation is as shown in fig 4.7. The best
lambda value suggested from the analysis of experimental data is -0.21 as shown
by the green line. The transformation log having lambda value 0 is suggested as it
falls within the lower and higher confidence interval values of -0.9 and 0.54
respectively which are represented by the red lines. Either natural log or base 10
log transformation is possible. Natural log transformation is selected for the ease of
transformation.
43
Table 4.4 shows the ANOVA results for cubic order model for heat affected
zone after log transformation. The model is developed for 95% confidence level.
The F-value of 33.93 implies that the model is significant. There is only 0.1%
chance that the F-value could occurred due to noise on the model developed. By
checking F-value and P-value it is seen that factors A, C, B2, ABC, A2B are model
significant terms. Values of pro > F less than 0.0500 indicate model terms are
significant. Values greater than 0.1000 indicate the model terms are not significant.
To improve the model, model reduction is needed by eliminating the insignificant
terms from the model.
Table 4.4 ANOVA result for heat affected zone after natural log transformation
Source
Model
A
B
C
AB
AC
BC
A2
B2
ABC
A2B
Residual
Cor total
SS
3.69
1.87
0.012
1.29
0.035
1.12E-005
0.011
0.040
0.056
0.076
0.050
0.17
3.86
Std.dev
Mean
C.V %
PRESS
0.10
-0.12
88.81
0.45
Dof
10
1
1
1
1
1
1
1
1
1
1
16
26
F
33.93
172.19
1.15
118.45
3.24
1.036E-003
0.99
3.64
5.13
7.00
4.64
P> F
0.0001
0.0001
0.2998
0.0001
0.0909
0.9747
0.3356
0.0746
0.0377
0.0176
0.0468
R squared
Adj R- squared
Pred R- squared
Adjprecission
Significance
Significant
Significant
Not significant
Significant
Not Significant
Not Significant
Not Significant
Not significant
Significant
Significant
Significant
0.9550
0.9268
0.8830
25.408
R-squared values are shown in the above table. The predicted R2 of 0.8830
is in reasonable agreement with adjusted R2 of 0.9268. Adequate precision greater
than 4 is desirable. The ratio 25.407 indicates that an adequate measure of the
signal to noise ratio.
4.6.2 Model equation after natural log transformation is
44
* EF - 0.30751 * WT +
SS
1.873E+05
51226.56
66.51
26027.87
1081.05
1181.10
562.99
237.64
422.03
1037.91
1565.09
1.889E+005
Std.dev
Mean
C.V %
PRESS
9.59
263.79
3.64
3730.8
Dof
9
1
1
1
1
1
1
1
1
1
17
26
F
226.07
556.42
0.72
282.72
11.74
12.83
6.12
2.58
4.58
11.27
P> F
0.0001
0.0001
0.4072
0.0001
0.0032
0.0023
0.0243
0.1266
0.0471
0.0037
R squared
Adj R- squared
Pred R- squared
Adjprecission
Significance
Significant
Significant
Not Significant
Significant
Significant
Significant
Significant
Not Significant
Significant
Significant
0.9917
0.9873
0.9802
47.149
Pred R-Squared of 0.9802 is in reasonable agreement with the Adj RSquared of 0.9873. Adeq Precision measures the signal to noise ratio. In this model
the ratio is 47.149, which is greater than 4, indicating adequate signal. So this
model can be used to navigate the design space.
45
46
Tensile-shear strength
Actual
Predicted
% error
Actual
Predicted
% error
Actual
Predicted
% error
2.38
2.26
5.04
0.28
0.29
-3.57
121.60
113.53
6.63
3.72
3.28
11.82
0.12
0.15
-25
103.80
105.76
-1.88
4.18
4.23
-1.19
0.53
0.48
9.43
288.60
277.33
3.90
3.19
3.38
-5.95
0.48
0.46
4.16
356.80
357.02
-0.06
4.17
4.22
-1.19
0.29
0.28
3.44
356.00
349.25
1.89
3.71
3.97
-8.20
0.72
0.54
25
244.80
242.78
-0.82
3.09
3.28
-6.14
0.20
0.16
20
323.60
323.61
-0.003
3.79
3.56
6.06
0.13
0.14
-7.69
252.90
254.39
-.058
This is as shown in table 4.6. The calculated percentage error is as shown below.
A comparative graph is also drawn which shows very less deviation for the
predicted values with that of measured values. Hence this model can be used for
predicting the performance measures of nugget diameter, heat affected zone and
tensile-shear strength in resistance spot welding process.
4.5
4
3.5
3
2.5
Actual
Predicted
1.5
1
0.5
0
0
10
Fig 4.9 Comparison plot for actual and predicted values of nugget diameter
47
0.8
0.7
0.6
0.5
Actual
0.4
Predicted
0.3
0.2
0.1
0
0
10
Fig 4.10 Comparison plot for actual and predicted values of HAZ
400
350
300
250
Actual
200
Predicted
150
100
50
0
0
10
Fig 4.11 Comparison plot for actual and predicted values of Tensile-shear strength
48
5
4.5
4
3.5
3
2.5
2
2.5
3.5
4.5
5.5
EF 2.5
EF 3
As shown in Fig 4.12 when the weld current increases, the nugget
diameter increases for low electrode forces. But with increase of electrode force
5.5
5
4.5
4
3.5
3
2.5
2
2.5
3.5
4.5
5.5
WT 15
WT 20
Fig 4.13 Effect of weld current on nugget diameter at different levels of weld time
From Fig 4.13 when the weld current increases, the nugget diameter
decreases first in 10 and 15 weld cycle times up to 4kA weld current and
afterwards increases. But for higher weld times, the reverse trend is seen.
49
400
350
300
250
200
150
100
2.5
3.5
4.5
5.5
EF 2.5
EF 3
Fig 4.14 Effect of weld current on tensile shear strength of spot weld at different
levels of electrode force
As shown in Fig 4.14 as weld current increases increase in tensile shear
strength of the joints occur. With increase of electrode force the tensile shear
400
350
300
250
200
150
100
50
2.5
3.5
4.5
5.5
WT 15
WT 20
Fig 4.15 Effect of weld current on tensile shear strength of spot weld at different
levels of weld time
Fig 4.15 is the plot of
current. Tensile shear strength increases with increase of weld time in 10 and 20
weld cycle. But, with increase in weld current, the tensile strength decreases up to
4kA and after that it starts increasing.
50
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
2.5
3.5
4.5
5.5
EF 2.5
EF 3
Fig 4.16 Effect of weld current on HAZ at different levels of electrode force
As in Fig 4.16 when weld current and electrode force increase, heat affected
2.2
1.7
1.2
0.7
0.2
2.5
3.5
4.5
5.5
WT 15
WT 20
Fig 4.17 Effect of weld current on HAZ at different levels of weld time
As shown in Fig 4.17 in the 10 cycle weld time, the HAZ first decreases
and after that it starts increasing, where as in 15 cycle, with the current increases
HAZ increases till 4kA and then reduces. So by reducing the HAZ for getting good
nugget diameter and tensile shear strength of the joint, moderate weld current and
low weld time has to be set.
51
CHAPTER 5
OPTIMIZATION RESULTS AND MICROSCOPIC
OBSERVATIONS
5.1 MULTI OBJECTIVE OPTIMIZATION
5.1.1 Optimum operating parameters
To determine the optimum operating parameters for the development of
weld zone, heat affected zone and tensile-shear strength of the joint. The
optimization feature available in Design-Expert 7 software was used. The software
searched for a combination of parameters that simultaneously satisfy the ultimate
goals and limits placed on the response and each of the parameters. The optimum
values of the parameters obtained were then assessed by composite desirability,
which carries a value from 0 to 1, to determine the degree of satisfaction of the
optimum values for the ultimate goal of response shown in table 5.1
Table 5.1 Optimization conditions
Name
Goal
Lower
Upper
limit
limit
Importance
WC
In range
EF
In range
WT
In range
10
20
Nugget diameter
In range
Tensile-shear
Maximize
103.8
383.2
Minimize
0.41
2.2
strength
HAZ
WC
EF
WT
Nugget diameter
Tensile-shear strength
HAZ
15.29
3.96
363.69
1.00635
52
As in Fig 5.2 three different layers of (i). Base metal (ii). Heat affected zone
and (iii). Fusion zone also could be observed.
53
CHAPTER 6
CONCLUSIONS
6.1 CONCLUSIONS
The main aim of the work was to find out the failures of tensile-shear strength
joints and develop model equations for each response, based on the analysis of
variance result (ANOVA) and optimize the process parameters for getting good
nugget diameter, tensile-shear strength and HAZ.
In a good joint, surface fracture occurred during the tensile shear strength
test around the weld nugget very near to the heat affected zone.
54
PUBLICATIONS
Sreenu S., Kishore N., Ramachandran N., and Allesu K., 2014, Regression
Modeling and Process Analysis of Resistance Spot Welded Joints,
Abstract accepted and paper submitted for 5thInternational Conference and
26th All India Manufacturing Technology Design and Research (AIMTDR
2014) to be held on 12-14 Dec.14 at IIT Guwahati.
Kishore N., Sreenu S., Ramachandran N., and Allesu K., 2014, Parametric
Studies and Finite Element Analysis of Welded Steel in Resistance Spot
Welding
Process,
Abstract
accepted
and
paper
submitted
for
Kishore N., Sreenu S., Ramachandran N., and Allesu K., 2014, Parametric
studies finite element analysis and regression analysis of welded steel in
resistance spot welding process, communicated to International J. of
Advanced Manufacturing Technology.
55
REFERENCES
1. Aslanlar, S., Ogur, A., Ozsarac U, Ilhan, E., Demir, Z., 2007, Effect of
welding current on mechanical properties of galvanized chromided steel
sheets in electrical resistance spot welding, J. Mater. Des., 28, pp. 27.
2. Aslanlar, S., 2006, The effect of nucleus size on mechanical properties in
electrical resistance spot welding of sheets used in automotive industry, J.
Mater. Des., 27, pp. 125131.
3. Hongyan Zhang, JacekSenkera., Resistance welding fundamentals and
applications published in CRC Press Taylor & Francis Group, Edition 2,
2006. ISBN 0-203-61639-1.
4. Larry F,Jeffus.,Welding principles and applications Indelmar Pub,
Edition 7, 2002.
5. Yi, L., Jinhe, L., Huibin, X., Chengzhi, X., Lin, L., 2009, Regression
modeling and process analysis of resistance spot welding on galvanized
steel sheet, J. Mater. Des., 30, pp. 25472555.
6. Vural, M., Akkus, A., Eryurek, B., 2006, Effect of welding nugget
diameter on the fatigue strength of the resistance spot welded joints of
different steel sheets, J. Mater. Proc. Technol., 176, pp. 127132.
7. Hamidinejad, S.M., Kolahan, F., Kokabi, A.H., 2012, The modeling and
process analysis of resistance spot welding on galvanized steel sheets used
in car body manufacturing, J. Mater. Des., 34, pp. 759767.
8. Hessamoddin, Moshayedi., Iradj Sattari-Far., 2012, Numerical and
experimental study of nugget size growth in resistance spot welding of
austenitic stainless steels, J. Mater. Proc. Technol., 212, pp.347 354.
9. Shamsul, J.B., Hisyam, M.M., Rizam, S.S., Murizam, D., Fitri, M.W.M.,
2007, Study of spot welding of austenitic stainless steel type 304, J.
Mater. Des., 911, pp. 1494-1499.
10. Ozyurek D., 2008, An effect of weld current and weld atmosphere on the
resistance spot weldability of 304 austenitic stainless steel, J. Mater. Des.,
29, pp.597-603.
56
11. Hamedi, M., Pashazadeh, H., 2008, Numerical study of nugget formation
in resistance spot welding, International Journal of Mechanics, issue 1,
Volume 2.
12. Nizamettin, K., 2007, The influence of welding parameters on the joint
strength of resistance spot welded titanium sheets, J. Mater. Des., 28, pp.
420-427.
13. Emin, B., Dominique, K., Marc, G., 2004, Applications of impact tensile
testing to spot welded sheets, J. Mater. Proc. Technol., 153-154, pp. 80
86.
14. Bouyousfi, B., Sahraoui, T., Guessasma, S., Chaouch K.T., 2003, Effect of
process parameter on physical characteristic of spot weld joints, J. Mater.
Des., 28, pp. 414419.
15. Oscar, M., Pilar, D. T., Manuel, L., Manuel, S.J., Cristina, G., Martin, F.,
Blanco, Y., 2009, Quality prediction of resistance spot welding joints of
304 austenitic stainless steel, J. Mater. Des., 30, pp. 6877.
16. Majid, P., 2011, Prediction of failure mode in AISI 304 resistance spot
welds, Association of Metallurgical Engineers of Serbia, UDC:
621.791.763, pp. 23-29.
57