10/5/2012

Casting Forming & Welding

Casting, Forming & Welding
(ME31007) 

JJinu Paul
u au
Dept. of Mechanical Engineering

Course details: Welding

Topic Hours
1. Introduction to welding science & technology 2-3
2 Welding Processes 4
3 Welding Energy sources & characteristics 1-2
4 Welding fluxes and coatings 1
5 Physics of Welding Arc 1
6 Heat flow in welding 2
7 Design of weld joints 1
8 Defects, Testing and inspection of weld joints 1-2

9 Metallurgical characteristics of welded joints, Weldability

and welding of various metals and alloys 2

Total 19
2

1
 10/5/2012

Welding Lecture ‐ 9
04 Oct 2012, Thursday, 8.30‐9.30 am
y

Heat flow in welds

3
3

The welding thermal cycle

• Thermal excursion Weld temp. ranges from the
ambient temp. of the work environment to above the
liquidus temp. (possibly to boiling point and above for
some very high-energy-density
hi h d it processes))
• The severity of this excursion  in terms of the
– temp. reached
– time taken to reach them
– the time remain at them
completely determines the effects on structure (both
microstructural for material changes and
macrostructural for distortion)
• To quantify the thermal cycle mathematically, we
need temp. distribution in time and space co-
4
ordinates

2
 10/5/2012

Thermal cycle characterization via

thermocouples

Temp.
p

5
Time

Thermal cycle- quasi-steady state

• Thermocouples  at various points along weld path
• Approach of the heat source  rapid rise in
temperature to a peak  a very short hold at that
peak  then a rapid drop in temperature once the
source has passed by
• A short time after the heat from the source begins
being deposited,  the peak temperature & rest of
the thermal cycle, reaches a quasi-steady state
• Quasi-steady state  balance achieved between the
rate
t off energy input
i t and
d the
th rate
t off energy loss
l or
dissipation
• Quasi-steady state  temperature isotherms
surrounding a moving heat source remain steady and
seem to move with the heat source (away from edges)
6

3
 10/5/2012

Thermal cycle- quasi-steady state

HAZ

S Fusion
zone

Temperature isotherms surrounding a moving heat source

remain steady and seem to move with the heat source 7

Time-Temperature curves
Tp Time= ti

Temperature

Time 8

4
 10/5/2012

Time-Temperature curves
• The peak temp. decrease with increasing distance from
the source, and more or less abruptly
• The maximum temperatures
p reached ((TmA TmB, Tmc)
decrease with distance from the weld line and occur at
times (tmA, tmB, tmc) that increase. This allows the peak
temperature, Tp to be plotted as a function of time
• Peak temp. separates the heating portion of the welding
thermal cycle from the cooling portion,
• At a time when points closest to a weld start cooling, the
points farther away
p y are still undergoing
g g heating. g This
phenomenon explains
– certain aspects of phase transformations that go on in the
heat-affected zone,
– differential rates of thermal expansion/contraction that lead to
thermally induced stresses and, possibly, distortion
9

Spatial isotherms
Tp (Peak temperature)

Heating zone C li zone

Cooling

Cooling zone

Heating zone

10

5
 10/5/2012

The generalized heat flow equation

• Temp. distribution Controls microstructure,
residual stresses and distortions, and
chemical reactions (e.g., oxidation)
• The influencing parameters
– the solidification rate of the weld metal,
– the distribution of peak temperature in the HAZ
– the coolingg rates in the fusion and HAZ
– the distribution of heat between the fusion zone
and the heat-affected zone
• Requires mathematical formulation to
quantify the influence of these parameters 11

The generalized heat flow equation

Heat supplied + Heat generated = Heat
consumed (for temp rise, melting) + Heat
transferred via conduction + Heat loss via
convection & radiation

12

6
 10/5/2012

The generalized heat flow equation

x = coordinate in the direction of welding (mm)
y = coordinate transverse to the welding direction (mm)
z = coordinate normal to weldment surface (mm)
T = temperature of the weldment, (K)
k(T) = thermal conductivity of the material (J/mm s-1K-1) as a
function of temperature
(T) = density of the material (g/mm3) as a function of temp.
C(T) = specific heat of the material (J/g-1 K-1),
) as a function of
temperature
Vx, Vy, and Vz = components of velocity
Q = rate of any internal heat generation, (W/mm3)

13

The generalized heat flow equation

• This general equation needs to be solved for one, two, or
three dimensions depends on
– Weld geometry,
– Whether the weld penetrates fully or partially
– Parallel sided or tapered, and
– Relative plate thickness
• 1-D solution  thin plate or sheet with a stationary source or
for welding under steady state (at constant speed and in
uniform cross sections remote from edges) in very thin
weldments
• 2-D l ti  thin
2 D solution thi weldments
ld t or iin thi
thicker
k weldments
ld t
where the weld is full penetration and parallel-sided (as in
EBW) to assess both longitudinal and transverse heat flow
• 3-D solution  thick weldment in which the weld is partial
penetration or non-parallel-sided (as is the case for most
single or multipass welds made with an arc source) 14

7
 10/5/2012

Weld geometry and dimensionality

of heat flow
(a)2-D heat flow for
full-penetration
welds
ld iin thi
thin plates
l t
or sheets;
(b)2-D heat flow for
full-penetration
welds with parallel
sides (as in EBW
and some LBW)
(c)3-D heat flow for
partial penetration
welds in thick plate
(d)3D, condition for
near-full penetration
welds 15

Rosenthal’s Simplified Approach

• Rosenthal’s first critical assumption  Energy input

from the heat source was uniform and moved with a
constant velocity v along the x-axis of a fixed
rectangular coordinate system
• The net heat input to the weld under theseconditions is
given by
Hnet = EI/v (J/m)
where  is the the transfer efficiency of the process E
and I are the welding voltage (in V) and current (in A),
respectively, and v is the velocity of welding or travel
speed (in m/s).
16

8
 10/5/2012

Rosenthal’s Simplified Approach

Assumption 2  Heat source is a point source,
with all of the energy being deposited into the
weld at a single point
– This assumption avoided complexities with density distribution of
the energy from different sources and restricted heat flow
analysis to the heat-affected zone, beyond the fusion zone or
weld pool boundary.
Assumption 3  The thermal properties (thermal
conductivity, k,
conductivity k and product of the specific heat and
density, Cp) of the material being welded are constants
Assumption 4  Modify the coordinate system from a
fixed system to a moving system

17

Rosenthal’s Simplified Approach

• The moving coordinate system  replace x with
 (Xi),
(Xi) where  is the distance of the point heat
source from some fixed position along the x
axis, depending on the velocity of welding, v
=x-vt
where t is the time

18

9
 10/5/2012

Rosenthal’s Simplified Approach

• This equation can be further simplified, in
accordance with Rosenthal
Rosenthal, if a quasi
quasi-stationary
stationary
temperature distribution exists.
• Temperature distribution around a point heat
source moving at constant velocity will settle
down to a steady form, such that dT/dt = 0, for
q/v = a constant
constant. The result is

19

Rosenthal’s solution
• Rosenthal solved the simplified form of the heat flow
equation above for both thin and thick plates in which
the heat flow is basically 2-D
2 D and 3-D
3 D, respectively
respectively.
• For thin plates,

q = EI/v
 EI/v = heat input from the welding source (in J/m)
k = thermal conductivity (in J/m s-1 K-1)
 = thermal diffusivity = k/pC, (in m/s)
R = (2 + y2 + z2)1/2, the distance from the heat source to a particular
fixed point (in m)
K0 = a Bessel function of the first kind, zero order 20

10
 10/5/2012

Rosenthal’s solution
• For the thick plate,

where d = depth of the weld (which for symmetrical

welds
ld iis h
halflf off the
h weld
ld width,
id h since
i w = 2d)

21

Rosenthal’s solution
• Above equations can each be written in a simpler form,
giving the time-temperature
time temperature distribution around a weld
when the position from the weld centerline is defined by a
radial distance, r, where r2 = z2 + y2
• For the thin plate, the time-temperature distribution is

22

11
 10/5/2012

Dimensionless Weld Depth Vs

Dimensionless Operating
Parameter
• B
Based d on R
Rosenthal’s
th l’ solution
l ti off th
the simplified
i lifi d three-
th
dimensional heat flow equation, Christiansen et al.
(1965) derived theoretical relationships between a weld
bead’s cross-sectional geometry and the welding
process operating conditions using dimensionless
parameters.
• The
Th theoretical
th ti l relationship
l ti hi b between
t th
the di
dimensionless
i l
weld width, D, and dimensionless operating parameter,
n, is shown, where

23

Dimensionless weld depth (D) Vs

process operating parameter n
Christiansen (1965)

• Width of the weld bead can

be determined (w = 2d)
• Width of heat-affected zone
can be determined
24

12
 10/5/2012

Dimensionless Weld Depth Vs

Operating Parameter n

d = depth of penetration of the weld,

U = welding speed (m/s),
s = thermal diffusivity (k/C) of the base material (as a solid),
Q = rate of heat input to the workpiece (J/s),
Tm = melting point of the base material (the workpiece), and
To = temperature of the workpiece at the start of welding.
For a symmetrical weld bead,
bead the width of the weld bead w = 2d 2d,
 Cross-sectional area of the weld bead can be determined

Can be applied to the heat-affected zone by simply substituting

TH for Tm where TH is the temperature of some relevant phase
transformation that could take place 25

Assignment -2
1) Find w and d for symmetrical weld bead as shown in
figure.
2) Find
Fi d th
the width
idth off HAZ ((phase
h ttransition
iti ttemp = 730 C)

Material steel with Tm= 1510 C

E=20 V
I= 200 A
Welding speed (v or U) =5 mm/s w
T0= 25 C
Arc efficiency =0.9 d
K=40 W/mK
C = 0.0044 J/mm3. C
t=5 mm
26

13
 10/5/2012

Welding Lecture ‐ 10
05 Oct 2012, Friday, 11.30‐12.30 am
y

Heat flow in welds

27
27

Effect of welding parameters on heat

distribution

• The shape of the melt , size & heat distribution,

is a function of
1. Material (thermal conductivity, heat capacity,
density)
2. Welding speed, and
3 Welding
3. W ldi power/energy
/ d
density
i
4. Weldment plate thickness

28

14
 10/5/2012

Effect of thermal conductivity (and

material property) on heat distribution

• Increasing thermal conductivity

– tends to cause deposited heat to spread
– Smaller welds for a given heat input and
melting temperature
• For a given heat input, the lower the
melting point, the larger the weld

29

Material Thermal 
diffusivity 
=k/C
(mm2/s)
Aluminium 84
Carbon steel 12
Austenitic  4
steel

Effect of
th
thermall
conductivity
(and material
property) on
heat distribution
30

15
 10/5/2012

Effect of welding speed on

Shape of Fusion/HAZ
Increasing velocity

Velocit 0 Low Medium High Very high

y
Plan Circle elliptical Elongated Tear Detached
view ellipse drop tear drop
3D
3-D H i
Hemi- P l t
Prolate Elongated
El t d 3D 3D ttear
view spherical spheroidal prolate tear drop
spheroidal drop

Effect of welding speed

32

16
 10/5/2012

Tear drop formation at very high

velocity
Continuous tear drops Detached tear drops at very high velocity

Weld direction
Weld direction

Effect of welding speed

• For a stationary (spot) weld, the shape, is round
(plan view)
view), and approximately hemispherical in 3 3-D
D
• Once the source is moved with constant velocity,
the weld pool and surrounding HAZ become
elongated to an elliptical shape (plan view), and
prolate spheroidal in 3-D
• With increased velocity,y, these zones become more
and more elliptical
• At some velocity (for each specific material), a tear
drop shape forms, with a tail at the trailing end of
the pool.
34

17
 10/5/2012

Effect of welding speed

• Increasing velocity  elongates the teardrop

more and more,  narrows the fusion and heat-
affected zone  overall melted volume constant
• Very high welding speeds  the tail of the
teardrop weld pool detaches  isolate regions
of molten metal  lead to shrinkage-induced
cracks along the centerline of the weld

35

Efect of the thickness of a

weldment

•Thick weldment  Small weld pool and heat-

affected zone 36

18
 10/5/2012

Effect of energy density

• Increased energy density  increases the
g  increases the amount
efficiencyy of melting,
of melting (especially in the depth direction)
 decreases the heat-affected zone.
• Shape of weld pool & HAZ will be distorted
by any asymmetry around the joint.
• Asymmetry might be the result of the relative
thermal mass (e.g.,
(e g thickness) of the joint
elements as well as their relative thermal
properties (Tm, k & C)

37

Simplified Equations for Approximating

Welding Conditions
1) Peak Temperatures  Predicting metallurgical
transformations (melting, austenitization, recrystallization of
cold-worked material, etc.) at a point in the solid material
near a weld requires some knowledge of the maximum
temperature reached at that specific location.
 For a single-pass, full-penetration butt weld in a sheet or a
plate, the distribution of peak temperatures (Tp) in the base
material adjacent to the weld is given by

38

19
 10/5/2012

Peak Temperature

To = initial temperature of the weldment (K)

e = base of natural logarithms = 2.718
 = density of the base material (g/mm3)
C = specific heat of the base material (J/g K- I)
h = thickness of the base material (mm)
y = 0 at the fusion zone boundary and where Tp = Tm
Tm = melting (or liquidus) temperature of the material
being welded (K)
Hnet = EI/v (J/m)
39

Width of the Heat-Affected Zone

• Peak temperature equation can be used to
calculate the width of the HAZ.
• Define Tp  Tre or Tau
• The width of the HAZ is determined by the value of
y that yields a Tp equal to the pertinent
transformation temperature (recrystallization
temperature, austenitizing temperature, etc.).
• Equation cannot be used to estimate the width of
the fusion zone, since it becomes unsolvable when
Tp =TTm
• (Remember the assumption in Rosenthal’s solution
of the generalized equation of heat flow,  Heat
was deposited at a point, and there was no melted
region, but just a HAZ)
40

20
 10/5/2012

Assignment 3- Part A
A single full penetration weld pass is made on steel using the
following parameters.
Tm= 1510 C,C E=20 V V, I= 200 A
A, Welding speed (v or U) =5
mm/s, T0= 25 C, Arc efficiency =0.9, C = 0.0044 J/mm3. C,
t=5 mm, Hnet= 720 J/mm
a) Calculate the peak temperatures at distances of 1.5 and 3.0
mm from the weld fusion boundary
b) Calculate the width of HAZ if the recrystallization temperature
is 730 C
c) Find the influence on the width of HAZ if a preheated sample
is used (Assume preheat temp =200 C)
d) Find the influence on the width of HAZ if the net energy is
increased by 10%
41

Solidification rate
• The rate at which weld metal solidifies can
have a profound effect on its microstructure,
microstructure
properties, and response to post weld heat
treatment.
• The solidification time, St, in seconds, is
given by

42

21
 10/5/2012

Solidification rate
• L is the latent heat of fusion (J/mm3),
• k is the thermal conductivityy of the base material
-1
(J/m s K ), -1

• , C, Tm and To are as before

• St is the time (s) elapsed from the beginning to
the end of solidification.
• The solidification rate, which is derived from the
solidification
lidifi ti titime, h
helps
l d determine
t i th
the nature
t off
the growth mode (with temperature gradient) and
the size of the grains
(will be discussed in detail on a separate lecture)
43

Cooling Rates
• The rate of cooling influences
– the coarseness or fineness of the resulting structure
– the homogeneity
homogeneity, the distribution and form of the phases and
constituents in the microstructure, of both the FZ and the HAZ for
diffusion controlled transformations
• Cooling rate determines which transformation (phase) will
occur and, thus, which phases or constituents will result.
• If cooling rates are too high in certain steels  hard,
untempered martensite  embrittling the weld  adds
susceptibility
p y to embrittlement byy hydrogen
y g .
• By calculating the cooling rate, it is possible to decide
whether undesirable microstructures are likely to result.
• Preheat  To reduce the cooling rate. Cooling rate is
primarily calculated to determine the need for preheat.
44

22
 10/5/2012

Cooling Rates-Thick plates

• For a single pass butt weld (equal thickness),

For thick plates,

R = cooling rate at the weld center line (K/s)

k = thermal conductivity of the material (J/mm s-1 K-1 )
T0 = initial plate temperature (K)
Tc = temperature at which the cooling rate is calculated (K)
45

Cooling Rates-Thin plates

For thin plates,

h = thickness of the base material (mm)

p = density of the base material (g/mm3)
C = specific
f heat off the base material (J/g
( / K-11)
pC = volumetric specific heat (J/mm3 K-1)

Increasing the initial temperature, T0 or applying preheat,

decreases the cooling rate
46

23
 10/5/2012

Relative plate thickness factor 

 ≤ 0.75 Thin plate equation is valid

  0.75 Thick plate equation is valid

47

Assignment 3- Part B

Find the best weldingg speed

p to be used for welding g
6 mm steel plates with an ambient temperature of
30 C with the welding transformer set at 25 V and
current passing is 300 A. The arc efficiency is 0.9
and possible travel speeds ranges from 5-10 mm/s.
The limiting cooling rate for satisfactory
performance
f i 6 C/s
is C/ att a temperature
t t off 550 C.
C

48

24
 10/5/2012

Assignment 3- Part B-Solution

49

Assignment 3- Part B-Solution

50

25