Impact simulation of a torque box beam column
Citation for published version (APA):
Bruijs, W. E. M. (1987). Impact simulation of a torque box beam column. (DCT rapporten; Vol. 1987.039). Technische Universiteit Eindhoven.
Document status and date: Published: 01/01/1987 Document Version:
Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:
www.tue.nl/taverne Take down policy
If you believe that this document breaches copyright please contact us at: openaccess@tue.nl
providing details and we will investigate your claim.
BEAM COLUMN
Wim Bruijs June 1987
The aim of this report is to establish the feasibility o f the PISCES-3DELK program for the crash simulation of vehicles. Therefore a simulation of the
impact of a rigid mass to a torque box beam-column has been carried out. The
results o f this simulation are in good agreement with the results of
experiments described in literature. It i s concluded that the method the
program uses t o solve the basic continuum equations is very well suited t o
Introduction
The PISCES-3DELK program Problem description The calculatjm The results Conclusions References
1
2 3 6 7 2 4 251 INTRODUCTION
This report deals with the simulation o f the crash o f a torque box beam-
column. This beam is a component o f the front car structure.
The aim o f this simulation is to test whether the PISCES-3DELK program (a
three-dimensional Euler-Lagrange coupled finite-difference/finite-element
program) can be applied to crash simulation o f vehicles.
It has been chosen to simulate the crash of a torque box beam-column,
because results of calculations (Argyris, Balmer, Doltsinis & K u n , 1986)
and experiments (Ni, 1973) are reported in literature. It is remarked that
the boundary conditions and the results o f the simulations and experiments,
are not given completely in literature.
In chapter 2 , a description of the PISCES-3DELK program is given. The
model is described in chapter 3, and in chapter 4 the calculation will be
discussed. The results of the calculation will be presented in chapter 5,
2 THE PISCES-3DELK PROGRAM
The PISCES-3DELK program is a general purpose computer program, that has
been designed to analyze continuum mechanics problems in three dimensions. The partial differential equations describing conservation of mass, momentum and energy are solved using finite element and finite difference methods. The program contains an Euler processor and a thin shell processor, that can be coupled in one single computational task.
For the crash simulation only the thin shell processor has been used. The thin shell processor employs an explicit finite element scheme.
Quadrilateral regions are modeled by a pair of triangular elements. The
shell elements are formulated in corotational coordinate systems. Each triangular element has three nodes (in the angular points of the triangle). Each node has six degrees of freedom, three rotations and three
translations. The element is of the constant strain type for the membrane strains and of the linear type for the bending strains. The displacements
perpendicular to the plane o f the element are interpolated between the nodes
using cubic polynomials. The displacements in the plane are interpolated using linear functions.
The integration points are situated at the midpoints of the sides. The
number o f Gaussian integration points over the thickness can be chosen by
the user.
The thin shell elements can interact with each other by means o f a three-
dimensional impact logic. Using the impact logic, it is possible to define a
surface by a number of nodes that have to be part of that surface. An impact
can be modeled by defining another set o f nodes, that are not allowed to
penetrate the surface.
The program version used for the calculation is version 2, level 26.
Unaccurate results lead to the addition of second order terms to the matrix that is used to calcu1afz.e the new nodal coordinate systems out of the nodal coordinate systems st the beginning of the time step. The nodal base vectors are scaled to unity length after calculation in order to avoid the
3 PROBLEM DESCRIPTION
The geometry of the torque box beam-column, which is a representative
component of the front car structure, is given in fig. i .
Impact mass: 1932 kg
Impact velocity: 4 . 7 1 m/s
a
Fly. i : The geometry of the torque box beam-column
The torque box beams are fully clamped at the lower ends. At the upper end,
the torque box beams are struck by a mass of 1932 kg with a speed of 4 . 7 4
m/s. The upper ends are allowed t o move vertically, but rotations of the
upper ends are suppressed.
The torque box beam-columns are made of aluminum 6016-T6. The strain-rate effects of the material are ignored. The density of the material is 2760
kg/m
.
The other material properties are given in fig. 2. The isotropic workhardening theory has been used.
3
Fig. 2: The material properties o f aluminum 606l-T6
Because of symmetry considerations, only a half part of one torque box beam-
column is modeled. The column is divided in 182 zones (fig. 3). Each zone i5
again divided in two triangular shell elements (except for the four
triangular zones). The mesh has been refined at the places where the largest
deformations are expected.
For the boundary conditions it is referred to fig. 4.
impact end
clamped end
Because it is not possible to define rigid bodies, the impact mass has been
modeled using shell elements too. In each node, the displacements in the x
and y direction and the rotations in the
x i
y andz
directions areprescribed to be equal to zero. This implies that the behaviour of the mass
will be close to the behaviour o f a rigid body.
The impact is modeled by defining a plane through the nodes of the impact
mass. The nodes o f the impact end (see fig. 3 ) are not allowed to penetrate
this plane.
2
t,
X:x
J -
0 : zy-symmetry displacements equal to zero: x-direction rotations equal to zero: y-direction z-direction clamped end displacements equal to zero: x-direction y-direction z-direction rotations equal to zero: x-direction y-direction z-direction impact end displacements equal to zero: x-direction y-direction rotations equal to zero: x-direction y-direction z-direction4 THE CALCULATION
The program calculates the maximum stable time increment using: d
A t =
-
Cwhere d is the height of the smallest triangular element and c is the wave
velocity :
In this equation E is the Youngs modulusr p the density and v is the Poisson
ratio.
Five integration points were used in the thickness direction.
For the crash simulation, the time increment was equal to 2.13 H S . The
program uses a safety factor o f 0.8, so the time increment used for the
calculations was 1.70 ms.
The calculation involved 6000 time increments and consumed 45.5 hours CPU
time on a VAX 11-780 computer. The impact time was 10.19 ms. The calculation
is stopped, because no interesting data were expected from a continuation o f
5 THE RESULTS
Deformations
The deformations are presented in fig. 5 . A frontal view of the deformations
i s given in fig. 6 . It is remarked that the displacements in fig. 5 are
multiplied by a factor 2, in order to get a better impression of the
deformed shape of the torque box beam-column. The displacements in fig. 6
are the real values
time = 3.4 ms time = 6 . 8 ms time = 10.2 ms
cycle 2000 time = 3.4: ms displacement upper end = 16 mm cycle 4000 time = 6.8 ns displacement upper end = 33 mm cycle 6000 time = 1 8 . 2 ms displacement upper end = 50 mm
Fig. 6 Frontal view of the stages of deformation
The deformations of the section marked in fig. 7 are given in fig. 8. The sections are viewed in the negative z-direction.
I
time = 5.1 ms
Forces
In fig. 8a, the nodal components of the reactioon force on the wall are
plotted as a function of the time (ZR, see fig. I ) . In fig. 8a and 9, the
total reaction force on the wall is given as a function of the time and the
displacements of the impact end respectively (2R, see fig. 1 ) . In fig. I O ,
the force applied by the torque box beam-column on the rigid mass, the so-
called barrier force is plotted against the time. It is remarked that less points were used €or this plot. In fig. 1 1 , the reaction force is plotted against the time using different filter frequencies as recommended by SAE J211b. In all these plots, the forces of two full torque box beam-columns are given. These results can be compared with the results of the experiments
of Ni (1973). Ni found a maximum in the barrier force of about: 89 kN at a
displacement of the impact end of about 12 mm (fig. 12). From fig. 10 it
appears that the maximum barrier force is about 90 kN at a time of about 2
ms or a displacement of the impact end of about 10 min. Because the reaction
force differs from the barrier force approximately only by a shift along the
time axis, the mnximum can also be read from the figs. 8 and 9 . This also
results in a maximum value of the force of a h u t 90 kR. At the end of the
calculations, the barrier force is about 60 kN and the displacement of the
impact end is about 50 mm. Ni (1973) found from his experiments a force of
about 60 kN at a displacement of 50 mm.
CRASH
I r- N h e-.*
h tu L 7 9 r- Cu LTi c O r- Cu d- c+
f-- Cu c3 s-- Q r- Cu ru co
I I I i I I I I I i I I I i I I O O c O m c . 4 O * a , E O .d. O o O N-
O-
c 19 O 2- c l .P ii EI
C.R ASH
i i , I Q> cn S = .* n E c - 4 O O m O Cu- - _ _ _ _ _ _ _ _ _
1
R O A D - V E H I C L E S RESEARCH I N S T I T U T E 8 7 0 6 1 1 l ! h h I m p a c t r e s p o n s e o f c u r v e d b o x b e a m F i i t e r f r e q u e n c y i n f l u e n c eI
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - - - Y 1 e Y2 -d- Y3+
Y4*
95.83 ( x = Z . 1 4 ) 86.98 (x=2.351 7 8 . 4 i ( ~ ~ 6 . 0 7 ) -0.000 ( x = O . 119) m a x . 9 8 . 1 1 ( x = 2 . 1 1 ) m i n . m e a n I e v e I c u m 3 m s c o n 3 m s i n t s c -0.000 (x=O. 1 1 9) - 0 . 0 0 0 ( * = O . 1 1 9) - 0 . 0 0 0 ( x = O . 1 1 9) H I C I : Y 1 C k N 3 ¶ 80 i 60 140 ? 20-
O O O c t) 1002
U o L LL 80 60 4 0 20 O l i I I I I I I ! 0.00 I I I L I I I I I I I I I I 1 I I I I I -90 60 I t 1 , I u 3 o O O I c C C U c I L L:
i 1 c < ti 3 0 1 . S O 3.00 4.50 6 - 0 0 7 . 5 0 T i m e ( m s e c ) 9.00 10.50 12.00Fig. 11: The reaction force as a function of the time using different filter
Displacement and velocity history
In fig. 13, the displacements of the impact end of the torque box beam-
column i.s given as a function of the time. Fig. I 4 shows the velocity
history of the impact mass.
If the surface under the curve of fig. 8b is calculated and divided by the mass of the rigid body, the result is an approximation of the decrease of the velocity of the impact mass. The resulting velocity decrease will be a too small approximation, because of the shift along the time axis of the reaction force, when compared to the barrier force. The decrease of the
velocity according to this method is 0 . 3 8 m/s (fig. 1 5 ) . The velocity
decrease resulting from the crash simulation is 0.39 m/s (fig. 1 4 )
Energy
In fig. 16 and 17, the kinetic energy of the impact mass and the total
energy are plotted as a function of the time. It is remarked that the energy
plotted is the energy of the modeled part of the construction only. Stresses
In fig. 18 the stresses are given at a time of 10.2 ms related to the
undeformed configuration. it is remarked that the stress plotted is the average mi.dplane stress of the element. This explains why the stresses are still in the elastic range.
CRASH
> c UE
X c h c*
-
W c -B c in c O c c+
m-
Q c Cu 7o
I I I i I I I I I I i I I I I OI
o o 'SZ T 'S ---+i+ 1 7O ' S -O O . - c O Q1 .4 I- O m O Cu O-
O O?SC_fS
inicrnafionai
b vCRASH
I 1 I I I I I I I I I 1 I I I 3 n O O m O N?ISCES !ntetnafionol b v
CRASH
I \ l . O
I I 1 I I I I I I I I I I I I
0OS.P- osç 'P- O09 'P- os9 't- 0 O L . t
'
O S L.!eo
OSE'P- oo+?-+ OSP 't-
_ _ _ _ _ _ _ _ _ _ _ _
ROAD-VEHICLES
RESEARCHINSTITUTE
870611I m p a c t r e s p o n s e o f c u r v e d b o x b e a m _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - - _ - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - - - Y + Yi
+
i 2-
Y3+
m a x . 105.3(t=2.04) 0.377 ( x = O , O l O ) m i n . -0.000 ( x = O . 089) -o. O00 ( x = O . 000) m e a n i e v e I c u m 3 m s c o n 3 m s i n t s c HI C/! Y 1 C k N l 1 2 0 1 O 0 80 600
40 L o LL 2 0 O - 2 0 - 4 0 - 60 - - - m / s e c . $ O i.60 1.50 A 3 . 4 0 I 0.30 0.2û 10.10.-
I
I I I I I I I I I I I I I I I
. 0 0 6 S '0015 'OOSS OOES ' 0 0 1 s '006P 'OOLP
UN U T
Xóiau3
' 0 0 1 9 a > > D O d- O NI I 1 I I i I I I I I i I I I 0 E !- . .it O d
-24-
T
:-
6 CONCLUSIONS
The aim o f this simulation was to test the feasibility of the PISCES-3DELK
program for the crash simulation o f vehicles. The following remarks are
made.
The results o f this simulation fit well with the experimental findings
reported by N i ( 1 9 7 3 ) .
The CPU time of 45.5 hours is too long for a practical application o f crash
simulation.
The CPU time necessary for the impact logic is not known. This can become an
important factor of the total CPU time of the simulation o f a complete car
crash.
In closing it ia remarked that, for this simulation a construction part of the vehicle, that can be modeled by triangular elements completely, is
chosen. For .the simulation of the impact of a complete vehicle structure,
it is desired to extend the code with other elements like beam and rigid
elements. I t can be concluded that the method the grogram uses to solve the
basic continuum equations, is very well suited to be applied t o crash
simulation. However, until the extensions mentioned earlier are made, it is
Argyris, J., Balmer, H.A., Doltsinis, J.St. & Kurz, A. ( 1 9 8 6 ) . Computer Simulation of Crash Phenomena. International Journal for Numerical Methods in Ensineerinq, 22, 497-519.
Ni, C. (1973). Impact response of curved box beam-columns with large global and local deformations. ASME/AIAA/SAE, 14th Strutures. Structural Dynamics and Materials Conf., Williamsburg, VA, U . S . A .