1 CHAPTER 1 INTRODUCTION Among the welding typical problems and most important are the residual stress/strain and the induced dist ortions in str uctures of the components. I n order to get better understand the welding process and its effects on structures, engineers and researchers around the world, covering a large number of industries, have been trying to create algorithms and methodologies to simulate the complete welding process or just individual phases (e.g. the cooling phase). In r ecent years, due t o the high expansion of comput er softwar e, comput ations poss ibili ties, many rese archer s identifi ed the Finit e Element Analysis (FEA) as a reliable method for this purpose. Hence ANSYS is used forthe welding simulation. 1.1 Nee d of the Proje ct The need of the proj ect arise s due to the problem of ina ccuraci es that take place durin g the heating or welding process of the production and hence increase partly from the thermal distortions and partly in the form of dimensional variations due to human factors. Furthermore, with the increasing use of the automation, it is necessary to quantify and calculate the thermal distortions by means of mathematical models by using FEA, so that the required tolerances of the automation process can be achieved as efficient as possible. In order to improve planning and work scheduling by reducing the rework, to reduce the production cost significantly by reducing the measurements and rework, and to improve the quality of the weldment the significance of FEA is studied . 1.2 Scope of the Project The problem of creation of residual stress es during welding steel plates leads to dimensional inaccuracies and misalignments of structural members, which can result in comple x tasks or rewor k when toler ance limi ts are cros sed. Henc e cause incr ease in the cost of production and leads to loss of time. In fabrication sector, for example, expenses requi red for rewo rk such as straigh tening cou ld have a heavycost . Therefor e, the problems of distortion and residual stresses arealways of great concern in welding sector. In order to deal with this problem, it is necessary to interpret the extent of distortion resulting from the
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
The research work in welding simulation was initializedmany years ago.Understanding of
the theory of heat flow is essential in order to study the welding processanalytically,numerically or experimentally. Rosenthal (1946) was the first researcher who succeeded in
developing an analytical solution of heat flow during welding based on conduction heat
transfer for determining the shape of the weld pool for two and three-dimensional welds.
By using the Fourier partial differential equation (PDE) of heat conduction, he initialized
the moving coordinate system to develop solutions for the point and line heat sources and
also applied this successfully to apply to a wide range of welding problems. His analytical
solutions for the heat flow made possible for the first time the analysis of the process from
the point of view of the welding parameters namely the voltage, current, welding speed,
and weld geometry. Due to the pioneering work done by Rosenthal,significant interest in
the thermal aspects of welding was developed by many researcherssuch as Goldak (1984)
The most critical and important input data required for welding thermal analysis are the
parametersnecessary to determine the heat input to the weldment by the arc. Goldak et al.
(1984) developeda mathematical model for welding heat sources based on a Gaussian
distribution of powerdensity. They proposed a double ellipsoidal distribution in order to
capture the size and shapeof the heat source of shallow and deeper penetrations.
Some of the researchers have alsodeveloped the thermal finite element simulation to
determine the temperature distribution of ametal.Over the past forty years, finite element
techniques have been used extensively in order topredict distortion and residual stresses due
to welding operations such as the studies byFriedman (1975), Michaleris and Debiccari
(1997) and Taylor et al. (1999). The finiteelement method has already been proven to be a
successful tool for the simulation of the complexwelding process as performed by Friedman
(1975). His 2-D finite element analysis work wasthen used by Taylor et al. (1999) for the
verification of their 3-D computational modeling of weldingphenomena.
The final results of finite element analysis done by Taylor et al. (1999) were inreasonable
agreement with the result obtained by Friedman (1975). Most of the welding research work
in the past was conducted to investigate the distribution and effect ofresidual stress and
distortion of welded metal component. It is possible to estimates the welding shrinkage in a
In general, welding is defined as any process in which two or more pieces of metalare joined together by the application of heat, pressure, or a combination of both. Almostall
the processes may be grouped into two main categories: heat welding, in which the welding
process is achieved by heat; and pressure welding, where the welding process is achieved
by pressure. Heat welding is the welding which is used today. The most important welding
parameters are the welding speed and the arc energy per unit length of the weld.
[Jonsson,M., Karlsson, Lindgren L.E;1985]
During welding process, the weldment is locally heated due to the welding heat source.
Because of the non-uniform temperature distribution during the thermal cycle, incompatible
strains cause to develop thermal stresses in the components. These incompatible strains due
to dimensional changes related with solidification of the welded metal, plastic deformation
and metallurgical transformations are the major sources of residual stresses and
distortion.[M Sundar1, G Nandi, A Bandyopadhyay And S C Roy; 2005]
The parameters of the line heating process have major effect on thermal distribution and the
resulting residual deformation of the heated plate. The thermal transients are generally
dependent on various factors like torch height, gas pressure, torch speed, and nozzle size,
which then controls the residual deformation of the plate. [Biswas, P., Mandal, N.R. & Sha,
O.P. 2007]
Arc welding, which is a heat-type welding, is the most important
manufacturingoperation for the joining of structural components for a wide range of
applications, including ships, bridges, building structures, guide way for trains,
automobiles, and nuclear reactors, to give an example. It is necessary to provide a
continuous supply of either direct or alternating electric current, which is normally used todevelop an electric arc to generate enough heat to melt the metal and form a weld.
The arc welding process is a complex operation involving extremely high
temperatures, which produce high levels of residual stresses and severe distortions .
Theseextreme phenomena results in reduction of the strength of a structure, which becomes
vulnerabletocorrosion, buckling, fracture and other type of failures.
Due to its nature, the welding process is one of the complicated transient procedure which
typically occurs in a 3-D (3-dimensional) structure. A typical welding simulation is
considered to be consists of two phases one is a transient heat flow analysis phase and
another is quasistatic plastic structural analysis phase. A reasonably good welding
simulation has requred to have a fine enough mesh in order to accommodate the high
thermal gradients that are developed (in time and in space) during the transient heat flow
analysis phase. At the same time, the mesh is required to be able to solve the plastic-
structural phase of the problem. [Cristian Simion,Corneliu Manu, Saleh Baset and Julian
Millard]
During the period ofthe heating and cooling cycles of a welding process, there are many
factors which affect shrinkage of the metal and therebymaking accurate predictions of
distortion complex and difficult. The mechanical and physical properties of the metal that
affect the degree of distortion change with respect tothe application of the heat. When the
temperature of the weld gets increasedthe modulus of elasticity, the yield strength and the
thermal conductivity of the steel decrease, whereas, the coefficient of thermal expansion and
the specific heat increase.
Thechanges in temperature and stresses during welding process have been studied by
Weisman (1976). To explain the temperature changes during welding, various cross-sections are required to be analyzed as shown in Figure 3.7(a).In some distance apart of the
welding torch i.e. section along A-A, the temperature gradient, ΔT due to the welding is
nearly zero. Along section B-B, which crosses the welding arc, at this section, the
temperature change is very high and the distribution is very uneven. Along section C-C,
having some distance behind the welding arc, the temperature change becomesmore even
and less steep. Finally, along section D-D, which is very much far from the welding arc, at
that section, the temperature change due to welding has reached to nearly zero.
At present the finite element method (FEM) is the most widely used tool for solving this
kind of thermal problems When the aim of the analysis is to determine the mechanical
effects of welding(residual stress and distortions) the simple approach is to consider the
thermal and mechanical relation only, because there is a weak connectionfrom mechanics to
heat flow. Heat generated bydeformation can be neglected) and hence the most used
approach is to carry a sequentially coupled thermal and mechanical analysis, if the structure
deformation during weldingdoes not change significantly. This aspect gives also
thepossibility for the use of general purpose finite element computer codes. [Viorel
Deaconu;2007]
The results obtained by FEA emphasize the ability of this method to give quality results, in
agreement with experimental results and also offer the possibility to a better understanding
of residual stress field characteristics. Despite the limitations related to the need for information related to the welding process and for complex material data and numerical
modeling remains actually the sole method which is able to fully characterize the residual
stress field through the whole structure without any limitations related to his geometry or
shape and size. Once having residual stress distributions information, subsequent
simulations related to stress relief carried by mechanical loading or by post weld heat
treatments can be performed. [Viorel Deaconu;2007]Following are the important aspects of
FEA
4.1.1 Simulationof Butt WeldingProcess
During fabrication of welded components residual stresses are produced as a result of non-
uniform temperature distribution during the welding process and particularly the cooling
processes. The residual stresses have a significant effect on the overall performance of the
components in service, [Pornwasa Wongpanya;2009]To simplify the welding simulation, it
is bettertoperform thermaland mechanicalanalyses separately.
At first, thecomputationof thetemperaturehistory duringweldingand
ismodeledin commerciallyavailablesoftwareby using theequivalentheat
inputwhichincludesbodyheat flux.
Theamount ofheatinput,QR has been calculated by
usingstandardrelationshowninEq.1.Arcefficiency isdenotedby η,arcvoltagebyV,arccurrent byI.Typicalweldingparameterstaken in thisstudy are,arcvoltage is considered30 volts,
arccurrent to be as 200 amp and arc efficiency70%.
Heat input = efficiency x voltage xcurrent
QR = η x V x I (1)
By usingequation (1)ofheatinputthe amount of heat input is determined. Bythermal
analysisthetemperature at different pointsarenoted. This
simulatedtemperaturefieldisthenused in analysis step forcalculatingthe residual
stressesandafterthat thevaluesof residual stressesarecalculated bymeans of stressanalysis.
4.4 BoundaryConditions
If a load is applied to the model, in that case in order to stop it accelerating infinitely through the
computer's virtual ether (which is mathematically alsoknown as a zero pivot),it is necessary that at
least one constraint or boundary condition must be applied. The structural boundary conditions areusually taken in the form of zero displacements, thermal boundary conditions such as temperatures
ranges are usually specified, fluid BCs are usually specified by pressures. A boundary condition may
be specified in order to act in all directions (x, y, z), or in some cases to certain directions only. They
can be placed or applied on key points, nodes, and areas or on lines. Boundry conditions's on lines
can be represented in the form of symmetric or anti-symmetric type boundary conditions, one
allowing in plane rotations &out of plane translations, whereas the other allowing in plane
translations & out of plane rotations for a given line. The applications of correct boundary conditions
are critical related to the accurate solution of the design problem. At least one boundary condition is
required to be applied to every model, in case when even buckling & modal analyses with no loads
applied.
The welding process simulation was carried out in three different steps. As the welding
process is time dependent the transient thermal structural coupled field analysis is carried out
a) Weld temperature of 1200º C applied on weld volume
b) Room temperature 20º C applied on plate volume
c) Weld time defined as 10 secStage II – Phase Change
a) Convection applied on surfaces subjected to air
b) Phase change time defined as 100 sec
Stage III – Solidification
a) Convection applied on surfaces subjected to air
4.5 Finite Element Analysis
The Finite Element solver may be logically divided into three main parts i.e. the pre-
solver, the post-solver and the mathematical-engine. The pre-solver is able to read in the
model created by the pre-processor and also formulates the mathematical representation of
the model. All important parameters defined in the pre-processing stage are going to be used
to do this, so if something is left out, chances are the pre-solver will complain & cancel the
call to the mathematical-engine. If the model is correct then the solver proceeds to form theelement-stiffness matrix for the problem & calls the mathematical-engine which calculates
the result (temperatures, pressures, displacement etc.) The results are then returned to the
solver & the post-solver is used to calculate heat fluxes, velocities, strains and stresses etc.)
for each and every node within the component or continuum. Finally all these results are
sent to a result file, which may be read by the post-processor.
During finite element analysis of welding process, a transient heat transfer analysis is
carried out by determining the temperature distribution in all nodes. After that a static
mechanical analysis is performed. Each step of the mechanical analysis represents a time
step in the thermal analysis. Finally, at the last step, when temperatures attain their initial
values and the residual stress field is obtained as result of all intermediary analysis
During simulation of welding process, the results of the analysis are read & interpreted.
They can be represented in the form of a contour plot, a table, deformed shape of the
component or the mode shapes and natural frequencies in case if frequency analysis isinvolved. Other results are also available for thermal, fluids and electrical analysis types.
Most of the post-processors provide an animation service, which is able to produce
animation & hence brings your model to life. Contour plots are generally the most effective
way of viewing results for structural type problems. Slices maybe made through 3D models
to facilitate the viewing of internal stress patterns. All post-processors nowadays include the
calculation of stress & strains in any of the x, y or z directions, or if requred in a direction at
an angle to the coordinate axes. The principal stresses and strains can also be plotted and if
required the yield stresses and strains can also be plotted according to the main theories of
failure (Von mises, St. Venant, Tresca etc.). Other necessary information such as plastic
strain, the strain energy, and creep strain may be obtained for certain types of analyses.
Coupled field analysis consists of the following steps
1. Preprocessing:-
Create the model geometry.
Define Material Properties
Mesh generation.
2. Solution:-
Boundary conditions (loads & constraints)
Solve.
3. Post Processing:-
Review results
5.3 Modeling Geometry
The ultimate purpose of finite element analysis is to re-produce mathematically the
behavior of an actual engineering system. In other words, modeling geometry is used to
create an accurate mathematical model of a physical prototype. In the broadest sense, thismodel can comprises material properties, real constants, all the nodes, elements, boundary
conditions and other features that may be used to represent the physical system.
In terminology, the term model generation usually takes on the narrower meaning
which includes generating nodes & elements that can represent the spatial volumes &
connectivity of the actual system.
In thepresentstudy, thebutt-weldjointof twoASTM 36stainless steel plates is modeled
using a commerciallyavailable finite elementsoftware. Thetwo semi-
There are two main types of meshing as free & mapped mesh. Meshing can be made
by using AMESH, VMESH commands. But 3-D model must be meshed with only VMESH
command only. The plate is meshed by using VMESH (mapped) command. Then the solidmodel is meshed by using solid 70 element by creating the FEA model having no. of nodes
12393 and no. of elements 8250 as shown in fig.5.3 and is used for further processing then
in FEA solution the FEM model created is analysed for coupled field analysis (also termed
as coupled field analysis).After completion of simulation the results are read and tabulated
by using post-processing stage.
5.6 Boundary Conditions
The welding process simulation was carried out in three different steps. As the
welding process is time dependent the transient thermal structural coupled field analysis is
carried out by using boundary conditions as mentioned earlier [Cristian Simion, Corneliu
Manu, Saleh Baset and Julian Millard],
5.7 Solution
This is the one of the important stage in FE analysis. With the help of ANTYPEcommand the type of the solution is specified as transient. Then by using SOLVE command
program solves the analysis using numerical methods for three different stages.
5.8 Post Processing
In post processing section, the results of stress analysis were reviewed. There are two types
of post processing methods viz. /POST1 & /POST26. With /POST1 only static analysis
results can be viewed and with /POST26 mainly time dependent analysis results are seen.
/POST26 is also termed as time history post processing.
From the present study the following major conclusions are drawn -
1) In case of experimental stress analysis it will require the prototype of the structure to be
analyzed while analysis by FEA software eliminates that and so the cost & time of product
design also gets reduced. With the help of FEA software the behavior of the structure can be
studied for any number of welds. By using different mesh densities the results can be
obtained with great accuracy. Thus FEA is one of the important tool for thermal and
structural analysis which gives results very fast. Thetime and cost required to find residual
stress in the parts due to welding process using FEA software is very less as compared to the
experimental method. Calculation of Residual stress for any complicated welding process as
well any complicated structure is possible by using finite element methods without altering
the physics of the problem.
2) Themaximum strain observed in the butt welded steel plate is found to be 67.2 x10-3
at the
region of weld.The maximum temperature at 100 seconds after completion of welding
process, observed in the butt welded steel plate is found to be 450 K at the region of weld
and it goes on decreasing along the length of plate. Finally the temperature at 100 seconds at
a distance of 100 mm from the center line of the weld after completion of welding process is
found to be 331 K which is the minimum temperature along the length of the butt welded
steel plate. Distribution of temperature, strain and residual stresses with respect to distancefrom weld are showed in tabulated form as well as by using graph for better understanding
of the welding process. The maximum residual stress observed in the butt welded steel plate
is found to be 320 Mpa at a distance of 5mm from the center line of the weld.
27. Matthews, J. & Sellon, W. Arc Welding: A Basic Manual of Instruction for Learning How to
Use Arc Welding. The James F. Lincoln Arc Welding Foundation. Cleveland, OH.
28. Michaleris, P. & DeBiccari, A. 1997. Prediction of Welding Distortion. Welding Res. Suppl
76: 172-181.
29. Michaleris, P. & Sun, X. 1997. Finite Element Analysis of Thermal Tensioning TechniquesMitigating Weld Buckling Distortion. Welding Journal, 76(11): 451–457.
30. Mollicone, P., Camilleri, D., Gray, T.G.F. & Comlekci, T. 2006. Simple Thermo-Elastic–
Plastic Models for Welding Distortion Simulation. Journal of Materials Processing
Technology 176: 77–86.
31. Murakawa, H., Serizawa, H. & Shibahara, M. 2005. Prediction of Welding Hot Cracking
Using Temperature Dependent Interface Element – Modelling Of Interfaces. Math. Model.
Weld Phen. 7: 539–554.
32. P.M.Gedkar and Dr. D.V. Bhope,” Finite Element Analysis of Pipe T-Joint”, P.M.Gedkar etal. / International Journal of Engineering Science and Technology (IJEST), Vol. 4 No.01
January 2012
33. Pornwasa WONGPANYA, Welding Residual Stresses in Two Competing Single V-Butt
Joints, Journal of Metals, Materials and Minerals, Vol.19 No.1 pp.67-75, 2009.
34. Properties of ASTM 36 Mild/Low Carbon, Steel, from website.
35. Rosenthal, D. 1946. The Theory of Moving Sources of Heat and Its Application to Metal
Treatments. Trans. A.S.M.E 68: 849-866.
36. S. Sarkani, V. Tritchkov, and G. Michaelov, ―An efficient approach for computing residual
stresses in welded joints, Finite Elements in Analysis and Design, vol. 35, pp. 247-268,
2000.
37. Taylor, G., Hughes, M., Strusevich, N. & Pericleous, K. 1999. Finite Volume Methods
Applied to the Computational Modeling of Welding Phenomena. Second International
Conference on CFD in the Minerals and Process Industries, Melbourne, Australia 1-6.
39. Viorel Deaconu, Finite Element Modelling of Residual Stress - A Powerful Tool in the Aidof Structural Integrity Assessment of Welded Structures, 5th Int. Conference Structural
Integrity of Welded Structures (ISCS2007), Timisora, Romania, 20-21 Nov 2007