2005-01-1323

Fatigue Analysis of Welding Seams and Spot Joints in

Automotive Structures
Helmut Dannbauer, Christian Gaier, Klaus Hofwimmer
Engineering Center Steyr GmbH & Co KG, St. Valentin, Austria

Copyright © 2004 SAE International

ABSTRACT prototyping”). Nevertheless, release tests are still

unavoidable before serial production starts.
For lightweight automotive structures, the stiffness and
the fatigue behaviour is greatly influenced by the A LITTLE THEORY
properties of its joints. The used joining technology, the
number and locations of the spot joints are of high WELDING SEAMS
importance for both engineers and cost accountants. An
overview of common computational procedures including Common European Standards
European and national standards is given for the
assessment of the stiffness and fatigue behaviour of thin Eurocode 3
sheet structures with spot joints and arc welds. The
influence of the quality and the size of finite shell Eurocode 3 [1] presents a general method for the fatigue
elements on the fatigue result is investigated and it is assessment of structures and structural elements made
shown, how this influence can be minimized. of steel which are subjected to repeated fluctuations of
stresses. Assessment of parent material and welds was
INTRODUCTION treated. The main provisions of Eurocode 3 rely upon a
set of fatigue resistance curves (S/N curves), equally
Since the drastic raise of fuel prices in the seventies the spaced, upon which a set of constructional details are
reduction of the vehicle weight is a basic requirement for classified. The S/N curves have a constant slope of m=3
the development of new cars, trucks and buses. (for normal stress), or m=5 (for shear stress, hollow
Nevertheless there is the general tendency in the section joints and some particular details). The slope of
6
opposite direction. The reasons for this contradiction are m=3 changes at 5 10 load cycles to m=5 and stress
8
on the one hand additional features for more comfort and amplitudes below a cut off limit at 10 load cycles will be
safety as well as increased motor torque. On the other neglected.
hand the potential for mass savings in the basic structure The assessment could be done using either nominal
is not enormous. stresses or according to the geometric stress
Vehicles have to withstand dynamical loads and their concentration concept (also called the “hot spot stress”
components have lots of arc welds, spot joints and method). For weld seams the stress components related
notches. Several technological treatments with positive to the seam direction will be used to constitute an
or negative influence to the lifetime usually are applied. equivalent stress σw and τw respectively.
How can engineers handle all important fatigue
influences during the development process? On the one σ w = σ ⊥2 + τ ⊥2 (1) τw=τ|| (2)
hand, a very good method is to perform fatigue tests with σ⊥ … Normal stress component perpendicular to weld
components or total vehicles. Many influences can be seam
considered without any mathematical model, τ⊥ … Shear stress component perpendicular to weld
nevertheless the results have absolute accuracy. seam
Unfortunately, experimental tests give only poor
information about low stressed areas, so it is very τ|| … Shear stress component parallel to weld seam
expensive and time consuming to minimize a
component’s mass based on experiments. For multi-axial loading normal stress and shear stress
Therefore, computational lifetime simulations, mainly should be assessed separately using linear damage
based on the finite element method FEM, are state of the accumulation according to Palmgren/Miner in
art to reduce or even avoid testing loops, prototyping and combination with one or two slope S/N curves.
therefore development time and costs (“virtual
The classification of load histories should be done using DS 952
rain flow or reservoir counting method.
The fatigue strength could be modified due to the In the German rail vehicle standard DS 952 [4] diagrams
influence of sheet thickness and/or by a 40% reduction are given, where permissible upper normal or shear
of the compressive part of the stress amplitude at parent stress is plotted versus stress ratio R. In each diagram
material or annealed welding seams. No further mean different curves are given for several notch categories.
stress influence will be considered in Eurocode 3. Thus again one can obtain the fatigue strength in
dependency from a notch category and the mean stress
British Standard 7608 (R value respectively).
These diagrams are given for the welds of the following
The British Standard 7608 [2] is very similar to Eurocode materials: St 37, St 52-3, AlMg3, AlMgMn, AlMg4.5Mn,
3. Again constructional details were divided into several AlMgSi1, AlZnMg1.
classes with corresponding S/N curves. In difference to The highest stress in the welding seam σmax or τmax
Eurocode 3 the S/N curves are not equally spaced and should be determined and this maximum stress must be
7
the slopes varies from m=3 over 3.5 and 4 to 8. At 10 smaller than the permissible values coming from the
load cycles the slope changes from m to m+2. No cut off diagrams.
at low stress amplitudes is allowed if only one cycle is DS 952 and DIN 15018 demonstrate similar mean stress
above the cut off limit. sensitivity at low mean stresses, but highly variable
The used stresses are the same as at Eurocode 3, at assessments at high mean stresses. The DS 952 rail
welding seams the stress components relating to the vehicle standard evaluates high mean stresses very
seam will be used. conservatively, in contrast to the DIN 15018 crane
manufacturing standard.
σ w = σ ⊥2 + τ ||2 (3)
As above linear damage accumulation according to The Radius 1mm Mean Spread Concept R1MS
Palmgren/Miner will be done, but without consideration of
shear stress separately (just within the equivalent Radaj, Koettgen, Olivier and Seeger developed an
stress). approach which allows the prediction of mean value and
The classification of load histories should be done using spread width of the dynamic strength limits of welding
reservoir counting method. seams (see [5]). During 41 test series (about 400 single
As well as at Eurocode 3 the fatigue strength could be tests) for different joint geometries, load configurations
modified too due to the sheet thickness influence and by and stress relations component S/N curves were
the reduction of compressive stress amplitudes of 40%. determined. All specimens were annealed.
No further mean stress influence will be considered in Afterwards the corresponding notch stresses at the
BS 7608. crack-endangered cross sections were recalculated
numerically for each test series. In doing so, the
DIN 15018 (equivalent stress) geometry of root and toe of the weld seam was made
with a notch radius of r=1mm. The modeled weld climb
The German crane manufacturing standard DIN 15018 angles, which have big influence to the notch stress at
[3] defines permissible fatigue stresses σD,perm(R=-1) for the toe of the welding, approximately corresponded to
St 37 and St 53-3 in dependency from notch categories the average of each individual test series.
and loading classes. Mean stress influence is considered
using the stress ratio R=σl/σu of lower stress σl to upper
stress σu. Thus one can obtain the permissible fatigue
stress σD,perm(R) for a certain notch category, loading
class and mean stress (R value respectively).
The verification of the fatigue strength will be done
proofing the following relation (4) which is similar to
v.Mises stress:
 σx   σy     τ 
  +  −  +  ≤ 1,1 (4)
2 2 2 2
σ x ⋅σ y
σ  σ  σ   
 xD, perm   yD , perm   xD , perm yD , perm   τ D, perm 
⋅ σ

σx, σy … Normal stress components in x and y

direction.
σxD,perm, σyD,perm … Permissible normal stresses
relating to σx and σy.
τ… calculated shear stress.
τD,perm … Permissible shear stress relating to τ.
Figure 1: Survival probability [%] against fatigue strength
2
[N/mm ]
 ∑
Figure 1 shows the notch stress fatigue limits calculated n σ i, j
from the experimentally determined component strength
using the above described models with r=1mm. In spite i =1 di
σj = (5)
∑
of the different joint geometries a uniform value of the n
1
notch stress fatigue limit with little scatter results. This
conclusion is valid both for a constant stress ratio R=-1 i =1 d i
and R=0. The values given in figure 1 were verified and
confirmed for other joint types using the S/N curve σj … stress component for the evaluation of the current
catalog of the DVS (see [6]). Thus it can be stated, that weld element (transverse, longitudinal and shear
for any weld joint geometry loaded with normal stress the component).
fatigue strength of root and toe can be predicted using
the described model.
σi,j … j stress components (transverse, longitudinal and
shear components with regard to the local weld
Minimization of Shell Element Size Influence
coordinate system) of the i elements at the
assessment point.
The finite element models of welded components like car
and truck frames, body in white structures, rail vehicle
bogies etc. were mostly built up with thin shell elements. di … distance of the element center of element i from the
For assessment of welding seams commonly the assessment point.
stresses of the elements adjacent to the seam will be
used. In this case the used stresses are depending n … number of elements utilized for the evaluation of
strongly on the size of the elements. Figure 2 shows the the current weld element at the assessment point.
distribution of the normal stresses perpendicular to the
seam line of an overlap joint for 4 different element sizes
(10x10mm, 5x5mm, 2.5x2.5mm and 1.25x1.25mm). On
one hand the peak stress at the joint depends on the
element size, on the other hand in a certain distance
from the joint the stress is nearly the same for all mesh
densities.

Figure 3: Automatic stress correction of FEMFAT WELD

This procedure works for both linear and quadratic shell
elements. Thus FEMFAT provides the possibility to use a
constant assessment distance from the joint
independently from the element size.
Another possibility to reduce the element size influence
is to use a grid point force dependent approach as e.g.
the equilibrium-equivalent structural stress approach
according to Dong et al. [7].
Figure 2: Stress distribution at overlap joint for different
element sizes
The fatigue software FEMFAT provides an automatic
stress correction feature for welding seams. The users SPOT JOINTS
have the possibility to define a certain distance from the
joint line. As a default the minimum sheet thickness is
used, which is the position of the weld toe between
welding seam and heat affected zone. In reality, fatigue Force Based Concepts
crack initiates at this position or at the weld root, for
which a fatigue analysis is performed too. FEMFAT Complicated automotive structures consisting of thin
interpolates the assessment stresses from the metal sheets (i.e. body in white) may contain some
surrounding elements (see fig. 3). thousands spot welds connecting the sheets. Sheets
The stresses of the elements marked with X in figure 3 usually are modeled by thin shell finite elements (e.g.
are used for evaluation of the current weld element. The NASTRAN CQUAD4), whereas spot welds are
weighted average stress components are calculated represented by one dimensional finite elements as bars,
using the following relationship: beams, rigid elements, etc. (e.g. CBAR, CBEAM, RBE,
CWELD). One dimensional finite elements can be linked
directly to the shell structure (node dependent meshing)
or by rigid body elements (node independent meshing).
The finite element analysis delivers among other results
forces and moments for one dimensional finite elements.
Radaj, Rupp et al. [8-10] proposed to use these forces
and moments for derivation of structural stresses at the
nugget boundary by an analytical model, see fig. 4.

Figure 5: Stress results in the joint area before

preprocessing for different FE-mesh size and quality

The definition and modification of the location of a

welding spot requires the respective nodes at the upper
and lower sheet-metal panel to be moved. How can be
ensured that all one-dimensional connecting elements of
a complete structural model are exactly normal to the
local panel surfaces? It is a fact that an inclination of a
one-dimensional connecting element yields to wrong
section forces.

Figure 4: Analytical spot weld model The section forces of the connecting elements only
provide information on the force flow in the spot weld
Analytical expressions for the three components of the itself between both sheets. The section forces however
plane stress state (radial stress, tangential stress and do not give any information on the loading of the welding
shear stress) as a function of all three forces and all spot in the sheet-metal panel itself (see figure 6). In case
three moments are given in [11]. It has been found that of an in-plane loading without any force flow in the spot
measured S-N curves, expressed by structural stresses, weld, such a concept would give zero damage. In reality,
fall within a narrow band for different loading directions however, the welding spot is a potential location of crack
(cross-tension load, tensile shear load) and even for initiation due to the sharp notch between the welded
different sheet thicknesses (0.8-1.2mm). Therefore it is sheets around the spot nugget.
sufficient to use one single S-N curve for the fatigue
assessment of the analytically calculated structural
stresses.

Force based concepts are rather easy to handle. No

mesh refinement is necessary around spot welds.
Therefore, a rather fast result can be obtained.
Nevertheless, it is still an open question how to fix the
diameter D of the fixed boundary, if the analytical spot
model is applied to real automotive structures. Another
open question from a practical point of view is, how Figure 6: Faulty fatigue estimation when using beam
intersections between neighboring sheets can be forces
avoided without modeling of gap elements, which is
usually too much effort. Generally, no matter which concept is chosen, the
simulation of spot-shaped joints is always a tightrope
Stress Based Concepts walk between acceptable costs and required accuracy.

Using the force based approach in conjunction with one The problems described in the previous section are
dimensional finite elements, the following problems solved by a different assessment approach, which
appear: requires a few additional steps to be performed [12]. The
basic idea is not to use the section forces of the
Size and quality of the shell elements connected to the connecting elements but the stresses in a well-defined
one-dimensional beam or rigid elements influence the ring of elements around the joining element. The
section forces as well as the local displacement behavior following steps are required:
(see figure 5). The problem to be solved is how to reach
a constant element size and quality in a complete
structural model.
Step 1: Modification of the FE-Mesh Step 3: Finite Element Analysis

For a raw FE-mesh without any connecting elements, This step generally consists of a linear or geometrically
CAD-Data may directly be put in to generate the spot nonlinear FE analysis with commercial software like
welds at the right location. Otherwise, an arbitrarily NASTRAN. As results, the linear stress distributions
meshed second generation FE-mesh, including should be generated for use in the next steps.
connecting elements, is used. A pre-processor serves to
automatically modify the local FE-mesh around the spot Step 4: Computation of Fatigue
weld. The result is a steady circular element
arrangement around the spot weld in the magnitude of The prediction of the fatigue behavior is based on the
the diameter of the spot weld (see figure 7 for an stress distributions in the element rings. The required
example). local specimen S/N curves for different loading types are
stored in a data base. They were determined by
comparing test and simulation results of specimen (H-
specimen [13], DC-specimen [14]). It was found that
differences in external loading and boundary conditions
have only little influence on the local dynamic strength
value in the assessment element ring. Similarly to the
welding seam fatigue assessment concept, the stress
tensor is transformed to the locally relevant coordinate
system. In the case of welding spots this is the radial,
tangential and shear stress. The kind of loading is
determined by the stress distribution along the ring.
Loading conditions can be pure normal tension or pure
shear tension or generally a mixture of both. The
associated S/N curve is generated by interpolation,
Figure 7: Automatically pre-processed local spot weld considering data from the DC- and H-specimen.
area of a body in white structure
Step 5: Visualization
The new mesh is automatically connected to the rest of
the structure. The constant circular FE-mesh around the For the spot welds, the computed fatigue values in terms
spot weld is the basis for a standardized assessment of damage or endurance safety factors are stored at the
quality of the element stresses. The number of the nodes of the element rings. They can be visualized
elements is increased by 24 per spot weld which can be analogously to stress results together with the fatigue
regarded as a disadvantage of the concept. results in the base material regions using conventional
post-processors. Figure 9 shows an application to a part
Step 2: Correction of Stiffness of a car- body, namely the joint B pillar.

In the course of automatic generation of the new mesh

around the spot weld, the stiffness properties of the
elements directly connected to the one-dimensional
connecting elements are adapted to correct the singular
deformation behavior (see figure 8). In the case of
welding spots, Young´s modulus is significantly
increased in comparison to regular shell elements. This
step also leads to a much improved local deformation
behavior. This makes the pre-processed model suitable
for e.g. crash investigations where deformations are
essential.

Figure 9: Damage distribution of a joint B pillar

Extensions for Self Piercing Rivets

The model for self-piercing riveted joints was developed

analogously to the model for spot welds [15]. The model
consists of two regular shell element rows around the
spot center node. The nodes of the inner row are located
on a circle with the same diameter as the rivet shaft, the
nodes of the outer row are located on a circle whose
Figure 8: Singular deformation behavior of regular DC diameter is similar to the diameter of the rivet head. The
specimen sheets are connected by beam elements with pipe cross
section. The diameter of the beam is the same as the Einner. row = 40 ⋅ Ebasic.material (6)
rivet shaft and the wall thickness of the beam is
equivalent to the wall thickness of the self-piercing rivet
(Figure 10). Eouter. row = 6 ⋅ E basic.material (7)

To consider the sheet thickness’ and the rivet diameter’s

influence the Young’s modulus of the connecting beam
will be set in dependence on the basic material, the
sheet thickness and the rivet diameter based on the
following equation:

D 
Ebeam = C ⋅ t p1 ⋅  rivet 
p2

 5.3 
(8)

where

Ebeam ........... Young’s modulus of the connecting beam in

Figure 10: Model of a self-piercing riveted joint MPa
To achieve the correct local stiffness behavior the
C................. Material constant
stiffness of the model has been adjusted by the Young’s
modulus of the surrounding shell elements and by the
Young’s modulus of the connecting beam. The values of t .................. Equivalent sheet thickness
the Young’s modulus as well as the parameters of the
formula describing the stiffness of the beam have been Drivet ............ Rivet shaft diameter
adjusted for optimal fit to test results.
p1 and p2 ... Material exponents
During the development several combinations of sheets
in a range of 0.7 – 2.7 mm have been investigated. The The material constant C has a value of 7500 MPa in
sheets are made of steel and aluminum and they are case of steel and 5000 MPa in case of aluminum. The
joined by self-piercing rivets with 3 mm and 5 mm rivet values of p1 and p2 have been adapted to test results.
shaft diameter. For the investigations I-, L- and H-
specimens have been considered (Figure 11). For predicting the fatigue life of self-piercing riveted joints
a concept similar to spot welds will be used. The fatigue
behavior of self-piercing riveted joints differs enormously
and is more complicated than that of spot welds.
Therefore, the concept has been modified for rivets.

A lot of specimens covering different combinations of

sheet thickness, sheet material and rivet diameter have
been investigated to quantify the influence of these
parameters on fatigue life. Different load types have
been investigated as cross tension load, tensile shear
load and combinations of both. It has been stated
already that the fatigue behavior of self-piercing riveted
joints depends on several parameter. The differences
are very significant in many cases. These influences are
considered by modifications of the S/N-curves. For each
load type between pure shear (0° load direction) and
pure cross tension (90° load direction) S/N-curves were
specified in 15° steps. In case of some steel types
(FeP04, DP500, Nb-Steel, PSi-Steel) the influence of the
Figure 11: I-, L- and H-specimens rivet shaft diameter has been considered by a
The stiffness of a self-piercing riveted joint depends on modification of the S/N-curves. The most important
several parameters, which cover sheet thickness, sheet influence factors which have been investigated are the
material and rivet diameter. sheet thickness and the sheet material, both for steel
and aluminum sheets. To consider the differences
between these two materials different S/N-curves were
To obtain the correct stiffness, the Young’s modulus of
the shell elements will be set in dependence on the basic defined. These S/N-curves have been included into a
database and they can be modified by the user as well.
material based on the following equations:
Another significant influence factor is the sheet EXAMPLES
thickness. If the same specimen with the same material,
rivet diameter and the same load condition but with ARC WELDS
different sheet thickness was tested, it showed that the
parameters of the S/N-curves (slope, fatigue limit, cycle A multi axial fatigue simulation of a closed rear axle
limit) were in many cases very variable. In the case of subframe was performed using the MAX module of
aluminum sheets the differences are only marginal. To FEMFAT. Seven load channels with measured load time
consider these influences modification factors have been histories were used. Non linear bearing behaviour was
introduced. These modification factors have an impact considered by separating the loading in positive and
on the parameters of the S/N-curves based on the negative direction. The position of the weldings were
following equations: defined within the FE model using node colours or
displacement coordinate system labels. The seam type
was determined using particular element material labels
σ D = σ D. ref . ⋅ t p 3 (9) according to modelling guidelines. With this FE model
linear stress analyses were performed for each unit load
case. In Figure 12 the maximum Mises stress values of
k = k ref . ⋅ t p 4 (10)
the dominant channel (loading in longitudinal direction
are shown.
N D = N D .ref . ⋅ t p 5 (11)

where

σD........... Fatigue limit of the S/N-curve of the current

spot joint with equivalent sheet thickness t

σD.ref ...... Fatigue limit of the reference S/N-curve in the

database (spot joint with 1 mm equivalent
sheet thickness)

t ............. Equivalent sheet thickness of the spot joint

k............. Slope of the S/N-curve of the current spot joint

with equivalent sheet thickness t

kref .......... Slope of the reference S/N-curve in the

database (spot joint with 1 mm equivalent
sheet thickness) Figure 12: max. Mises stress values of one load case
The stress results of the unit load cases were combined
ND .......... Cycle limit of the S/N-curve of the current spot using the measured load time histories for evaluation of
joint with equivalent sheet thickness t the fatigue life. Parent material and weldings were
assessed simultaneously in one computation run. The
ND.ref. ...... Cycle limit of the reference S/N-curve in the structural stresses at the welded joints coming from the
database (spot joint with 1 mm equivalent stress analyses were extrapolated to notch stresses by
sheet thickness) using notch factors from a database. In this database
different notch factors for many joint types were stored in
The exponents p3, p4 and p5 have been calibrated to addition to S/N curve data and Haigh diagrams. The
test results from all tested sheet materials. The spot notch factors were obtained either from submodel
database contains these exponents, so they can be analysis or experimental results.
modified by the user again. Figure 13 shows the damage values (inverse of life) of
one multi axial loading cycle. In spite of high stresses in
larger areas the damage results are focusing on the weld
joints.
The benefits of this method are:
• No care must be taken of the detailed geometry
of the welding seam in complex structures.
• Assessment of welds and parent material
simultaneously.
• Independence from the mesh density using the
automatic stress correction.
 • Assessment according to different normative Generally, it can be found, that damages simulated by
standards possible. the force based concept are mostly lower than those
• Open database (adjustable by the user) for FE ones simulated by the stress based concept. Possible
weld seam notch factors , S/N curves and Haigh reasons are:
diagrams. • The chosen diameter D=42mm of the analytical
model seems not to be suitable for body in white
structures. This value should be reduced.
• The quality of force and moment results
depends strongly on the mesh quality, which is
locally varying.
Comparisons with test results showed good correlations
with stress based simulation results. It can be assumed
that the result quality of stress based concepts is usually
better than for force based concepts. Nevertheless, the
necessity of the local finite element mesh refinement
structure around spot joints has some practical
limitations, especially the large increase of the number of
finite elements. Also the automation of the local mesh
refinement is very error-prone. Therefore force based
concepts are still superior concerning practical
applicability.

SELF-PIERCING RIVETED UNDERBODY

Figure 13: damage values for multi axial loading for one
loading cycle Finally the fatigue life prediction of a self-piercing riveted
SPOT WELDS underbody was performed by applying the newly
implemented methods. This underbody was made of
A body in white structure with spot welds has been different sheet materials (ZStE 250 steel for the basic
considered. The spot welds have been analyzed both by sheet and ZStE 340 steel for the longitudinal and lateral
the force and stress based concept as described above. supports) with different sheet thickness’ (0.70 mm for the
For a certain amount of spot welds the predicted damage basic sheet and for the lateral stiffeners and 1.14 mm for
results of both methods are almost the same, but for the longitudinal stiffeners). The sheets were connected
other spot welds there are large discrepancies in the by self-piercing rivets with 3 and 5 mm rivet shaft
predicted lifetime up to a factor 1000 or even more, see diameters. The structure was tested with a pulsating
figure 14. torsion loading of 600 Nm. Figure 15 shows the damage
distribution at corner “A” and the predicted and observed
fatigue life of the most critical rivet joint.

Figure 15: Most critical rivet joint of the underbody

(results of fatigue life prediction with FEMFAT)

It can be seen, that the very complicated stiffness and

fatigue behavior of self piercing rivets can be described
very well by the proposed stress based concept.

Figure 14: Damage results on spot welds simulated by Force based concepts are not available now and maybe
force based concept (top) and stress based concept it is also not possible to adapt force based concepts for
self piercing rivets, because such methods are too 12. Gumpinger, J.; Hahn, O.; Korte, M; Kudrnac, P;
simple to describe the behavior of self piercing rivets, Singh, S; Unger, B. “Computer Simulated Estimation
which is much more complicated than for spot welds. of the Fatigue Behaviour and Stiffness of Spot Joints
in Automotive Structures", IBEC´97 – International
CONCLUSION Body Engineering Conference, Stuttgart, Germany.
13. Singh, S; Schmid, G; Gao, S. (1991) “ Leichtbau
An overview has been given over some common durch optimierte Fuegetechnik”, lecture and paper
methods and standards for the assessment of welding on the occasion of the DVM Meeting “Bauteil 91” at
seams and spot joints including spot welds and self Berlin, Germany. Proceedings page 239-255.
piercing rivets. It has been shown, how the influence of 14. Hahn, O.; Kurzok, J.R.; Rohde, A.; Thesing, Th.
the mesh quality and the element size can be minimized. (1996): “Konzept zur numerischen Auslegung
For spot welds force based concepts and stress based punktfoermig gefuegter Duennblechstrukturen auf
concepts have been presented. It has been shown, that der Basis experimenteller Versagensanalysen”. LWF
stress based concepts usually deliver better results and
Paderborn, Germany.
can be applied for self piercing rivets too, but the effort
15. H. Dannbauer, C. Gaier, C. Halaszi, “Development of
for the local mesh refinement is very high and error-
a Model for Self-piercing Rivets to Predict Stiffness
prone. For the assessment of some thousand spot welds
in an automotive body in white structure force based and Fatigue Life of Automotive Structures”, IBEC
methods seems to be still superior because of practical 2003, Chiba, Japan, JSAE 20037136, SAE 2003-01-
reasons. Future efforts take aim to improve automatic 2857, pp. 105-110.
node-independent local mesh refinement around spot
joints.

REFERENCES
CONTACT
1. EUROCODE 3 : Design of steel structures . Part 1.1:
General rules and rules for buildings. ENV 1993-1-1. Magna Steyr, Engineering Center Steyr GmbH & Co KG
2. British Standard 7608 (1993), “Fatigue design and Steyrer Str. 32
assessment of steel structures”. A-4300 St. Valentin
3. DIN 15018 Part 1 (1984), “Cranes; principles for Austria
steel structures, stress analysis”. e-Mail: [email protected]
4. DS 952, Germany.
5. V.B. Koettgen, R. Olivier, T. Seeger (1989)
„Schwingfestigkeitsnachweis fuer Schweissverbin-
dungen auf Grund oertlicher Beanspruchungen“,
Forschungshefte, Heft 143, Forschungskuratorium
Maschinenbau, Frankfurt.
6. V.B. Koettgen, R. Olivier u. T. Seeger (1991)
„Schwingfestigkeitsnachweis für Schweissverbin-
dungen auf Grund oertlicher Beanspruchungen“,
DVS-Berichte Band 133, pp.75-83, Deutscher Verlag
für Schweisstechnik, Duesseldorf.
7. P. Dong, H. Kyuba, “Equilibrium-Equivalent
Structural Stress Approach to Fatigue Analysis of a
Tubular Joint”, International Institute of Welding
IIW/IIS Doc.XIII-1993-03; XV-1149-03.
8. D. Radaj (1990) “Design & Analysis of Fatigue
Resistant Welded Structures”, Abington Publishing,
Cambridge.
9. D. Radaj (1998) “Fatigue Assessment of Welded
Joints by Local Approches” , Abington Publishing,
Cambridge.
10. A. Rupp, K. Stoerzel, V. Grubisic (1995) „Computer
Aided Dimensioning of Spot-Welded Automotive
Structures“, SAE 950711.
11. Y. Nakahara, M. Takahashi, A. Kawamoto, M.
Fujimoto, N. Tomioka (2000) “Method of Fatigue Life
Estimation for Spot-Welded Structures”, SAE 2000-
01-0779.