INVESTIGATION OF ELECTRON BEAM WELDING OF AHSS BY
PHYSICAL AND NUMERICAL SIMULATION
Raghawendra P. S. Sisodia1, Marcell Gáspár2
PhD student1, Assistant Professor2 1,2Institute of Materials Science and Technology, University of Miskolc, Miskolc
3515, Hungary
ABSTRACT
The electron beam welding (EBW) sets new standards as it facilitates very high
quality and effective welding of high strength structural steels. The technology
ensures high-quality critical welded joints in thicker structural metals. It has a high
energy density in comparison to the conventional arc welding processes (e.g.
GMAW). As a result of less overall energy input and higher velocity, the effect of
welding on the base material in the heat-affected zone (HAZ) and the distortion is
much smaller compared to conventional arc welding processes. The low heat input
result in a small HAZ and a reduced extension of critical HAZ areas which can be
favourable in high strength steels when the mechanical properties can drastically
decrease in the HAZ. In comparison with experimental studies, a numerical modelling
study can provide detailed information concerning the welding process and
parameters, and the number of costly experiments can be reduced. Finite element
modelling (FEM) of EBW enables the estimation of temperature field, time
temperature curve, weld pool geometry and welding distortion etc. The determination
of the temperature field can be very useful in terms of the further investigations since
the t8/5 cooling time can be less than 2 s during EBW. In this paper, by the application
of Sysweld software, the time-temperature curve was determined and the physical
simulation of the critical HAZ subzones were performed using a GLEEBLE 3500
physical simulator in order to analyse the properties of HAZ during extremely short
cooling time.
Keywords: Electron Beam Welding (EBW), Finite element modelling (FEM), HAZ
characteristics, Gleeble 3500 thermophysical simulator, t8/5 cooling time, 3D conical
heat source
INTRODUCTION
In keeping the view of modern world where advancement in technology is
prime requirement to drive economic growth and sustainable development objective
within the world. Several steel manufacturers have increased their interest in ultra-
high strength quenched and tempered steels [1][2]. The development of high strength
weldable steels has diversified the field of application, range of design alternatives
subjected to more severe operation conditions than previous time. Nowadays high
strength steels are gaining more popularity in welded structures, even in ship building
and machinery components. Advanced High Strength Steel (AHSS) are more
complex, particularly through their microstructures, which are usually multiphase for
MultiScience - XXXIII. microCAD International Multidisciplinary Scientific ConferenceUniversity of Miskolc, 23-24 May, 2019, ISBN 978-963-358-177-3
DOI: 10.26649/musci.2019.051
an improved combination of strength and ductility. This balance is carefully used to
meet performance requirements while maintaining excellent formability [3]. This
high strength can be achieved with minimum yield strength up to 1300 MPa, AHSS
often has other advantageous mechanical properties, such as high strain-hardening
capacity. They provide high strength to weight ratios, very good acceptable
weldability especially where overmatched welds are used, improved toughness and
sufficient deformation capacity. Financial benefits can also be realized through
reduced transportation and lifting costs (reduced weight), material savings
(smaller/lighter sections) and reduced weld volumes (thinner plates). Easier and
simple structural components and construction techniques are also possible,
especially for large structures, and establishment costs may also be reduced due to
lower dead weight [4]. High strength steels are available for structural applications,
as in bridges, buildings, offshore applications etc., all around the world. A very high
carbon equivalent value indicates poor weldability and these steels are not suitable
for structural applications, where welding is very important to assure structural safety
[5]. The future of AHSS for automotive applications is bright. The weldability of high
strength steels consists of more challenges compared to mild steels: hardening of the
heat-affected zone (HAZ) and higher cold cracking sensitivity; reduction of strength
and/or toughness of HAZ; selection of filler material (mismatch) [6]. Therefore, new
welding conceptions and special welding technology should be developed for these
steels. For the advanced high strength steels innovative welding technologies can be
needed (e.g. laser beam, electron beam, laser-GMAW hybrid welding technology)
which may have benefits for the joint properties and the production costs. Generally,
these welding processes result narrow HAZ, which is often the most problematic part
of the welded joint in AHSS. By these technologies the extension of the critical areas
(softened or hardened zones) can be minimized [7]. Regarding the economic
advantages the high productivity and the less filler material costs should be
emphasised. Due to the narrow and deep penetrated weld structure (keyhole
technique) lower amount of filler material is needed. It can be also beneficial to the
mechanical properties of the whole welded joints since in AHSS the weld can be
similarly critical as the HAZ.
Electron Beam Welding (EBW) is a highly efficient, flexible [8] and precise welding
method increasingly used within the manufacturing chain and of growing importance
in different industrial environments such as the automobile, aeronautical, aerospace
(3A), construction, and power generation sectors [9]. It is also observed from two
referred research paper that numerically simulated cooling time can be extremely
short: t8/5 = 0.49 s [10] & t8/5 = 2 s [11] for EBW.
With the development of various new high strength steel and their difficulty in
weldabilities arises the other promising welding technology that can be applied for
high and ultra- high strength steels is the electron beam welding process [12]. So far,
the studies of electron beam welding on high strength quenched and tempered steel
and thermo mechanically controlled steel with different yield strength have not been
carried out extensively, therefore their suitability should be investigated before. Since
they are very expensive technologies, therefore the application of numerical and
physical simulation is suggested before the real welding experiments and the
production.
EXPERIMENTAL METHODOLOGY
The first phase of experimental steps involves the numerical simulation that
containing steps like creation of geometry of the required model, selecting type,
order, method of element for further mesh creation, providing boundary conditions,
weld path, reference line, clamping conditions etc. Then analysis is performed which
can be either thermo-metallurgical or mechanical analysis together or separately
depending upon requirement and finally simulated data were generated with detailed
analysis. The important parameters considered for numerical simulation for S960QL
material were presented in Table 1, which seems quite satisfactory with real
experiment data with EBW for 10 mm thickness plate and velocity 10 mm/s [13].
The needed material properties of the steel S355J2 are integrated in the FE software
SYSWELD and used for the calculations because it is similarly low alloyed as
S960QL, therefore there is no a big difference in the thermo-physical properties. The
efficiency is a very important parameter and it is considered η=1 for EBW analysis.
The efficiency influenced the amount of the heat source power. [14]
Table 1
Simulation parameters
Welding
process
Heat input
(J/mm)
Velocity
(mm/s)
Simulated
time (t8/5)
(s)
Cooling
medium
EBW 600 14 2 Free air
cooling
Second phase of experimental steps is the physical simulation and the simulation
parameters were determined iteratively by running calculations with more parameters
in order to find the optimal EBW parameters [15]. Unfortunately, proper toughness
evaluation of HAZ is far more difficult than for a homogeneous material. Because of
the very local heat input, a weld offers various zones with different microstructures
and therefore different properties (metallurgical notch) [16] and the third phase is the
materials testing. A Gleeble 3500 thermomechanical simulator was used to simulate
the welding thermal cycles of the heat-affected zone of S960QL steel. Square base
specimen (10 mm x10 mm x 70 mm) were prepared and subjected to thermal cycle
[15].
The high cost of the material and the process does not allow for trial and error
approach in optimizing the process and, therefore, modelling and simulations
approaches are being used increasingly with continuous improvement brought on
better heat source modelling and more reliable predictions on the influence of process
parameters on the distribution of HAZ, residual stresses and distortions. Finite
element (FE) simulations have been the most common numerical method used to
simulate the welding process, thermal and mechanical analyses [17]. Modern finite
element codes not only allow for calculation of deformations and stresses due to the
welding process but also take into account the change of microstructure due to
different heating and cooling rates [18]. For an advanced high strength steel structure
or sample such effects shall be investigate and quantify using advanced material
modelling and FE simulation [19]. The electron beam welding (EBW) numerical
model was created using SYSWELD software. Weld plate is modelled and simulated
on 3D finite element method (FEM) using the SYSWELD software and this is
specifically used for thermal analysis. It basically involves three important steps in
welding simulation, these are modelling, analysis and post processing. The sample
geometry, materials and heat sources fitting belong to the first steps i.e. modelling in
weld simulation sequence.
Meshing
The main aim was to simulate the EBW of a 15 mm thick butt welded joint. The
meshed model of plate with butt joint used in simulation is presented in Figure 1(a).
The model contains total 49,470 nodes, forming quad triangular elements. The
dimensions of the plate used for simulation is 150 x 100 x 15 mm and the number of
plates is two for butt welded joint with a gap of 2 mm. A finer mesh has been created
in the joint area can be seen in Figure 1(a), in order to get the steep temperature
gradients with higher accuracy. The total number of elements involved in the meshing
is 52,764. After creation of the base model in the SYSWELD software, the necessary
groups needed to define the welding trajectory, starting and end points, reference line
together welding path defines the plane, respective to which heat source coordinate
frame is oriented [20]. The red line in below Figure 1(b) is referring to the reference
line. Using clamping configurations can mitigate the welding distortion of floor
structure. Here in the simulation, “Stop” under the Clamping Condition step was used
in order to block a structure in only one given direction. Thermal losses to the
surrounding environment were modelled using element groups termed “skins” which
consisted of 4-node quadrilateral elements superimposed on the free surfaces [21]. In
order to model the ambient temperatures around the model, a skin consisting of 2D
elements must be generated.
Fig. 1 Finite element model
(a) Geometry of a butt joint and (b) Reference line, skin for cooling, clamping
Heat source model
The welding operation is modelled through the Welding Wizard module, which
makes it possible to flexibly and accurately assign the necessary parameters for
processing the analytical model for volumetric heat release. Volumetric heat release
takes place as a result of the introduction of heat in accordance with several thermal
models incorporated into the program. For modelling electron beam, a conical shaped
source with 3D Gaussian distributed thermal energy density source model [22][23] is
applied to get thermal field analysis, shown in Figure 2a-b.
Fig. 2 (a) Heat source model [23] (b) Temperature field
This model is described by Eq. (1-2). At any plane perpendicular to z-axis, the heat
intensity is distributed in a Gaussian form and may be written as [23].
2 2
0 2
0
( , , ) exp( )
x yQ x y z Q
r z
+= −
(1)
the height of the conical heat source is H = Ze - Zi, the z-coordinates of the top and
bottom surfaces of the conic region, and it can be expressed as
( )0 ( ) i ee e
i e
r rr z r z z
z z
+= + −
− (2)
where Q0 is power intensity, re and ri are top and bottom radiuses, remainder (ze-zi) is
penetration, r0 is the distribution parameter and r is the radial coordinate from the
source centre.
RESULT AND DISCUSSION
Numerical Simulation
Numerical modelling of welding processes has proven to be highly efficient for
design and production engineering [24]. The numerically simulated temperature field
is presented in the cross section at the heat source in Figure 3a-b.
Fig. 3 (a) Iso section heat source [22] (b) volumetric section
Thermal analysis
After numerical simulation, one node in the HAZ is selected (Figure 4a) for
finding t8/5 cooling time and exported the curve as shown in Figure 4b, it is observed
that cooling time is 2 s.
Fig. 4
(a) Selected node in HAZ (b) Exported HAZ thermal cycle
Microscopic examination
Microscopic examinations were performed by the Zeiss Evo MA10 Scanning
Electron Microscope. The SEM images as shown in Figure 5, demonstrate the
successful physical simulation of the desired heat zone i.e. peak temperature 1350 °C
and t8/5=2.5 s, and the effect of the cooling time on the microstructure of the HAZ
subzone. It can be seen from the images that when the coarse grain zone structure
was examined, the grain size in CGHAZ is finer than in the t8/5 range of conventional
arc welding technologies in the 5…30 s cooling time range, since during EBW is less
time for grain coarsening (when the t8/5 is just around 2…2.5 s) and below also the
needle martensite’s and the block sub-substructure are well observable.
The scanning electron microscopic pictures of coarse-grained heat affected zone,
S960QL (Tmax = 1350 ºC, 2.5 s for different magnifications (M=500x, M=1000x,
M=1500x & M=2000x) are shown in Figs. 5a, b, c-d.
0
200
400
600
800
1000
0 5 10 15 20
Tem
p, °
C
t, s
t8/5=2 s
Fig. 5 Coarse grained HAZ (S960QL), Tmax = 1350 ºC, t8/5=2.5 s, (a) M=500x, (b)
M=1000x, (c) M=1500x & (d) M=2000x [15]
The hardness values thus obtained after physical simulation for CGHAZ (Figure 6a)
and the cooling time, t8/5=2 s was calculated from the equation in Figure 6b which is
440 HV10. It can be stated that CGHAZ has fulfilled the requirement of EN ISO
15614-1 standard [25] even in 2 s cooling time, which allows 450 HV10 maximum
hardness for non-heat-treated welded material group 3. Compared to the HV10
hardness of the raw material 340, the measured higher average hardness for t8/5=2 s.
With the maximum allowable hardness, the 2...2.5 s cooling time technology proved
to be the most critical, but it did not reach the 450 HV10 limit.
Fig. 6 Hardness test
(a) Thermal cycle with Rykalin-3D model (b) second degree polynomial curve [15]
0
200
400
600
800
1000
1200
1400
0 50 100 150 200 250
T, ºC
t, s
2.5 s
850 ºC
500 ºC
y = 0,0071x2 - 1,9961x + 437,48R² = 0,9947
200
250
300
350
400
450
500
0 20 40 60 80 100
HV
10
t8,5/5, s
In the case of the coarse grain zone, the effect of the cooling time t8/5 was also plotted
on the diagram Figure 6b, with the second-degree polynomial on the hardness values
[15]. So, it can be observed from above result that hardness may increase further with
shorter cooling time than 2 s. However, we did not reach the permitted maximum
value, there is a high risk in EBW that we reach this value, therefore we should
estimate higher crack sensitivity.
CONCLUSIONS
Thermo-metallurgical analysis of electron beam welding was performed on
SYSWELD software considering conical heat source model. A numerical model was
used for the accurate thermal analysis and prediction of t8/5 cooling time which is
usually shorter in case of electron beam welding compared to conventional arc
welding technologies (e.g. gas metal arc welding). and here in our results it was found
to be 2 s which is conforming to various results performed by another researchers [1]
[13]. On the basis of numerical simulation results, the time-temperature curve was
determined and various mechanical properties like hardness and microstructure of
numerically simulated thermal cycle for t8/5= 2 s cooling time in HAZ was compared
with the results of previously determined t8/5=2.5 s thermal cycle from physical
simulation using GLEEBLE 3500 HAZ tests, the following conclusions were drawn:
(a) Thermo-metallurgical simulation of advance high strength steels (S960QL) is
possible for electron beam welding using SYSWELD.
(b) To model the weld penetration for required thickness (15 mm) of material was
performed successfully and ensured the full depth penetration with the
optimized parameters by the FEM.
(c) The physical simulation results was used for comparison with numerically
simulated thermal cycle results successfully which allow studying materials
behaviour at conditions very close to real industrial processing or applications,
and by means of physical simulation a large variety of microstructures and
associated mechanical properties can be obtained and studied in a short time
and for a tiny fraction of full-scale industrial experiments.
(d) The thermal cycle exported from simulated results and found that the t8/5
cooling time is 2 s for electron beam welding is quite small and conforming to
the result.
(e) The hardness values thus obtained after physical simulation for CGHAZ, t8/5=2
s was 440 HV10. It can be said that CGHAZ has fulfilled the requirement of
EN ISO 15614-1 for required test cooling period, which allows 450 HV10
maximum hardness for non-heat-treated welded material group 3.
(f) Through microscopic examinations of the coarse grain HAZ, it was observed
that the size of the grains is finer for shorter cooling time (i.e. t8/5=2.5 s)
compared to the generally applied 5…30 s cooling time of arc welding
process.
(g) Also, the needle martensite’s and the block sub-substructure are well
observable.
REFERENCES
[1] Blacha, S., Weglowski, M.S., Dymek, S., Kopyscianski, M.: Microstructural
and mechanical characterization of electron beam welded joints of high
strength S960QL and WELDOX 1300 steel grades, Arch. Metall. Mater.
62, 2, pp. 627-634. (2017),
[2] Weglowski, M.St., Zeman, M., Grocholewski, A.: Effect of welding thermal
cycles on microstructure and mechanical properties of simulated heat
affected zone for a Weldox 1300 ultra-high strength alloy steel, Arch.
Metall.Mater. 61, pp. 127-132. (2016).
[3] https://www.autosteel.org
[4] Dobosy, Á. and Lukács, J.:Welding properties and fatigue resistance of
S690QL high strength steels, Mater. Sci. Forum, Vol. 812, pp. 29-34.
[5] Kuzsella, L., Lukács, J. and Szűcs, K.: Nil-strength temperature and hot
tensile tests on S960QL high-strength low-alloy steel, Production processes
and systems, 6 (1), pp. 67-78. (2013).
[6] Gáspár, M., Balogh, A., Lukács, J.: Toughness examination of physically
simulated S960QL HAZ by a special drilled specimen, Lecture Notes in
Mechanical Engineering F12, pp. 469-481. (2017).
[7] Maurer,W., Ernst, W., Rauch, R., Kapl, S., Pohl,A., Krüssel, T., Vallant, R.
and Enzinger, N.:Electron beam welding of a TMCP steel with 700 MPa
yield strength, Welding in The World, N°10Vol. 56, pp.85-94. (2012).
[8] Areskoug, Tekn Lic M.: A comparative study of welding super high strengh
steel with MAG and EB processes, IIW Doc.XII-2300-16/IV-1298-16/I-
1277-16/212-1450-16
[9] Schultz, H.: Electron Beam Welding, Abington Publishing, Cambridge,
United Kingdom. (1993).
[10] Enzinger, N.: Modelling and simulation of welding processes, Institute of
Materials Science and Welding, 30. Jan. (2015).
[11] https://www.scia.net/sites/default/files/thesis/reportgogou.pdf
[12] Weglowski, M.S., Błacha, S., Phillips, A.: Electron beam welding –
Techniques and Trends – Review, Vacuum 130, pp.72-92 (2016).
[13] Hesse, A.C, Pagel, T.N, Dilger, K., “Fracture toughness of electron beam
welded fine grain steels” Procedia Structural Integrity 2, pp. 3523-3530.
(2016)
[14] Slovácek1, M., Divis1, V., Junek1, L., Ochodek2, V.: Numerical Simulation
of The Welding Process - Distortion and Residual Stress Prediction, Heat
Source Model Determination, Welding in the World, Vol. 49, n° 11/12, pg.
15-29. 2005.
[15] Gáspár, M.: Physical simulation-based development of welding technology
for quenched and tempered structural high strength steels, PhD thesis,
University of Miskolc, Hungary (2016).
[16] Wiednig, C., Enzinger, N.: Toughness evaluation of EB welds, Weld World,
61: pp.463–471. (2017).
[17] Dalewski, R., Jachimowicz, J.: Numerical modelling of welded joints, Weld
Int; 25(3), pp.182–7. (2011)
[18] Siegele, D.:Welding mechanics for advanced component safety
assessment, Front. Mater. Sci., 5(2): 224–235. (2011).
[19] Willms, R.: High strength steel for steel constructions, Nordic Steel
Construction Conference, Malmo, Sweden. (2009).
[20] Monfared, A. H.: Numerical Simulation of Welding Distortion in Thin
Plates, Journal of Engineering Physics and Thermophysics, Vol. 85, No. 1,
January, pp. 187-194. (2012).
[21] Wang1 J. & Han1, J., Domblesky2, J. P., & Yang1, Z., Zhao1, Y., Zhang1, Q.:
Development of a new combined heat source model for welding based on
a polynomial curve fit of the experimental fusion line, Int. J. Adv. Manuf.
Technol. 87:1985–1997, (2016). DOI 10.1007/s00170-016-8587-3
[22] ESI Group: SYSWELD reference manual, Digital form SYSWELD
Heat_Source_models_Welding_user_meeting_2014
[23] Ferro, P., Tiziani, A., “Metallurgical and mechanical characterization
of electron beam welded DP600 steel joints, J. Mater. Sci. 47: pp.199–
207. (2012).
[24] Lacki, P., Adamus, K., “Numerical simulation of the electron beam welding
process”, Comput. Struct. 89, pp.977–985. (2011)
[25] MSZ EN ISO 15614-1: “Welding instructions and welding technology for
metals. Examination of welding technology. Part 1: Arc and gas welding
of steels and arc welding of nickel and alloys”, 2004.
ACKNOWLEDGEMENT
The described article was carried out as part of the EFOP-3.6.1-16-2016-00011
“Younger and Renewing University – Innovative Knowledge City – institutional
development of the University of Miskolc aiming at intelligent specialisation” project
implemented in the framework of the Szechenyi 2020 program. The realization of
this project is supported by the European Union, co-financed by the European Social
Fund.