Analysis of Displacement of A Concrete Barrier On
Analysis of Displacement of A Concrete Barrier On
Analysis of Displacement of A Concrete Barrier On
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/263550661
CITATIONS READS
2 200
1 author:
Leon Prochowski
Military University of Technology
22 PUBLICATIONS 11 CITATIONS
SEE PROFILE
All content following this page was uploaded by Leon Prochowski on 05 April 2015.
ANALYSIS OF DISPLACEMENT
OF A CONCRETE BARRIER ON IMPACT OF A VEHICLE.
THEORETICAL MODEL AND EXPERIMENTAL VALIDATION
Leon Prochowski
1. Introduction
Road infrastructure (in much the same way as vehicles) incorporates active and passive safety
devices. The latter group contains various safety barriers. These are important for safety as more
than 15% vehicles driving off the road lane collide with such structures [7].
We distinguish permanent and temporary (removable) barriers. The latter are mostly concrete
blocks used for temporary traffic re-routing or securing road work sites. Concrete barriers consist
of segments connected with couplings featuring high tensile strength. Such combination acts as an
articulated joint with limited mobility. By retaining their integrity on impact from a vehicle, the
components perform their protective role. There are various types of concrete barriers with
different profiles (e.g., New Jersey, STEP, or profile F) [9].
Although many research institutes perform computer-assisted simulation of car crashes
(including collisions with rigid structures), most of them focus on safety features of vehicles and
structural aspects of barriers. There are few studies devoted to the practical use of traffic barriers,
including barrier behaviour on impact.
Our purpose is to build a practical model for researching certain aspects of such behaviour. We
considered various barrier base surfaces and displacements depending on impact energy. We also
performed validation calculations based on results of many experiments.
The collision of a car with a traffic barrier creates great threats resulting from:
- difficult to predict vehicle trajectory after rebounding from or sliding along the barrier,
- barrier component displacement as a result of vehicle impact.
L. Prochowski
Regarding the former point, barriers are designed and installed so that they return the vehicle
back into traffic heading in its original direction (parallel or nearly parallel to the barrier).
Typically, such trajectory poses least threat to other vehicles. The other aspect of the problem,
lateral displacement of the barrier, is worth considering as another threat, particularly where the
barrier defines the outer lane limit or the median.
Fig. 1. The following quantities describe such displacement: a - maximum lateral displacement, b - length of the
active (movable) section of the barrier, E - angle of vehicle approach to the barrier with velocity v
Letters A and B in Fig. 1 represent different surface strips with a clear demarcation line.
Lateral barrier displacement on impact of a vehicle, also shown in Fig. 2, is typically observed
for a few sections of the barrier (section b in Fig. 1). The maximum displacement is also referred
to as barrier deflection (letter a) or working width.
Based on results of traffic monitoring and concrete barrier placement studies, we built a model
serving a purpose of describing the movements of the vehicle and barrier segments on collision.
An important feature of the model is that it describes the development of situation from the initial
contact until disengagement.
The PC Crash application we used for this purpose enables linking structure segments to
represent a concrete barrier placed directly on ground. The model addresses the differences
between the surfaces in areas A and B were considered in this model and outlined in Fig. 1.
Our mathematical model of the trajectory of the vehicle and the displacement of the barrier
components was based on studies [12, 16-18]. We used the Kudlich-Slibar model to describe the
collision. [13, 17]. We did not analyze the vehicle deformations described in this paper.
Figure 3 represents a model of collision with a barrier consisting of rigid segments and shows
the following:
- outlines of segments 9, 10 and 11 and segment couplings s10 and s11,
- frozen vehicle body position,
- vehicles centre of inertia trajectory.
400
Analysis of Displacement of a Concrete Barrier on Impact of a Vehicle. Theoretical Model and Experimental Validation
Fig. 3. Momentary vehicle position during collision with a barrier segment (see the description in the body of this paper)
We considered forces and moments acting on barrier segment couplings as external factors
affecting each segment individually. The following equation describes the moment of resistance of
a coupling calculated for each axis of the local coordinate system,
M (M ) (M M M ) S M D for M ! M M , (1)
M (M ) 0 for M d M M , (2)
where:
M - angle of barrier segment rotation relative to the adjacent segment [o],
M M - threshold angle, the exceeding of which results in an increase of the moment of resistance,
S - moment of resistance gain factor in the coupling [Nm/o],
M - angular velocity of rotation of adjacent barrier segments,
D - coupling damping coefficient.
Once the segment rotation angle reaches M Mo so that M (M0 ) M max , the coupling is broken
and from now on M (M ) 0 . Obviously, M0 ! M M . The maximum moment of resistance of the
coupling is limited by the couplings tensile force limit. The tensile force limit for the coupling
used in the model, the exceeding of which ruptured the barrier, was 350 kN (i.e., close to the value
declared by manufacturers of reinforced concrete barriers [15]).
The barrier displacement resistance is produced by friction between barrier components and the
underlying surface and the moment of resistance acting in segment couplings. We considered the
following in our selection of test conditions:
- conclusions from traffic collision studies [13],
- Polish and European standard PN EN 1317,
- possibility of calculation result validation.
401
L. Prochowski
displacement of the barrier, so we disregarded such scenarios in our work. However, a heavier
vehicle and a 10 degree or larger angle produces a clear barrier displacement, which can pose a
threat to the traffic.
The coefficient of friction between the barrier and the underlying surface (strip B in Fig. 1)
addresses the results of study [5]. We measured the position of the barrier segment from the idle
state up to the speed of approx. 2.5 m/s and obtained 0.2-0.9 friction coefficient values. The
characteristic values observed for the segment moving at 1m/s velocity are as follows:
- ground surface: 0.7-0.75,
- asphalt covered with dry and wet sand: 0.5-0.6.
We used vehicle crash descriptions contained in the PC Crash applications library of vehicles
in our calculations.
In our model, the barrier consisting of a kinematics chain of solids can be displaced by impact.
Each segment can move, rotate or turn over.
Fig. 4. The position of the barrier hit by the vehicle at a 16 degrees angle, vehicles centre of inertia trajectory and
vehicle contours at 0.5 s intervals. The visible part of the barrier consists of segments 6-12
Figure 4 shows some calculation results for the final barrier position and vehicle positions
calculated at 0.5 s interval. The maximum barrier displacement is measured from the barrier axis
in pre-collision state to the farthest coupling position in the end of the simulation. This distance is
measured orthogonally to the original position of the barrier (Fig. 4).
402
Analysis of Displacement of a Concrete Barrier on Impact of a Vehicle. Theoretical Model and Experimental Validation
In the last application, the model is used for evaluating convergence in mean square according
to [11].
2
1,8
1,6 Friction 0,2
0,4
Displacement, m
1,4
0,6
1,2
Friction 0,8
1
Polynom.(Friction 0,2)
0,8
Polynom.(0,4)
0,6 Polynom.(0,6)
0,4 Polynom.(Friction 0,8)
0,2
0
0 20 40 60 80 100 120 140
Energy, kJ
Fig. 5. Maximum Barrier Displacement vs. Impact Energy
Figure 5 compares calculation results for 4 friction f coefficient values: P B 0.2; 0.4; 0.6 i 0.8 .
We considered a few impact angles ( E ) ranging from 8 to 24 degrees. Also, we addressed
80-110 km/h impact speeds of cars from commercial segments B and C. The comparison
demonstrated that a second-order polynomial depending on the coefficient of barrier friction
against the underlying surface can describe the relationship between the displacement and energy.
See Tab. 1 for the functions we used to approximate the results shown in Fig. 5.
Tab. 1. List of functions used to approximate the relationship between the maximum lateral barrier displacement (y
expressed in meters) and the energy of impact (x in kJ)
Convergence factor (R2) values close to 1 confirm that the combination of functions provides
an efficient displacement/energy model for B and C type vehicles.
We should always try to validate the model by referring it to empirical results. For this purpose
we used results of vehicle-barrier crash tests. [6, 10]. Although the literature is abundant,
descriptions detailed enough for the purposes of our work are rare.
Table 2 provides the results of two experiments compared with results calculated on the basis
of the course of the approximating functions listed in Table 1 above, containing quite detailed
descriptions. So compared, the values of the maximum barrier displacement in the experiment to
calculation result approximation layout are similar one to another.
403
L. Prochowski
Tab. 2. Comparison for validation of calculation results based on experimental research results
Impact Barrier; Max. displacement
Vehicle Impact Impact Max.
speed segment calculated using
weights, angle, energy, displacement References
V, length the approximation
type deg. kJ empirical, m
km/h and type from Tab. 5
Chevrolet
Cheyenne 3.81m.
25 100 137 0.81 m 0.78-0.79 m [6]
1995 r Type F
2000kg,
Suzuki
2.5 m .
Metro, 15 78 14 0.21 m 0.14-0.15 m [10]
RS
910kg
Figure 6 shows the results of experiments for many vehicles weighing from 800 to 1200 kg and
a wide range of impact speeds (see paper [10]), depending on impact energy. The figure shows
also the course of the experiment result approximating function, i.e. lateral barrier displacement
dependence on vehicle impact energy.
0,4
0,35
Displacement, m
0,3
0,25
Empirical
0,2
Involution fun.(empirical)
0,15
0,1
0,05
0
0 10 20 30 40 50
Impact energy, kJ
Then, Fig. 7 and Tab. 2 compare the course of the approximating function shown in Fig. 6 with
approximation of our calculations.
0,8
0,7
Displacement, m
Tab. 2. List of approximating functions shown in Fig. 7 and convergence factor values
404
Analysis of Displacement of a Concrete Barrier on Impact of a Vehicle. Theoretical Model and Experimental Validation
Figure 7 shows the courses of the two functions from Fig. 5 we used to approximate the
calculation results for barrier displacement on the underlying surface with friction factor
P B 0.6 i 0.8 (typical for concrete or asphaltic concrete either clean or contaminated with sand).
Also, the figure shows the course of the function we used to approximate the empirical results
from Fig. 6 (for concrete surfaces). We extrapolated the function for an impact energy range much
wider than that shown in Fig. 6. The extrapolated course of approximation also corresponds to our
calculation results. It confirms that the theoretical results are consistent with the empirical ones.
7. Conclusion
Our calculations and results demonstrate that our model describing the process of a vehicle
collision with a removable concrete barrier is feasible. The results of comparing our theoretical
results with the empirical ones confirm, in principle, that the two are consistent, although, the
latter originated from multiple experiments.
We demonstrated that lateral barrier displacement depends on vehicle impact energy. This
relationship is particularly useful for traffic safety studies since it defines characteristic traffic
parameters (vehicle weights and speed). Basing on our calculations, we determined that a second-
order polynomial can be used for describing the lateral displacement of a barrier placed on various
types of surfaces for a wide range of impact energies (actually, from 0 to 120 kJ). The courses of
the approximating function shown in Fig. 5 represent an efficient model for determining the
maximum lateral displacement of a concrete barrier on impact of a B or C segment vehicle.
We give our credits to Mr. Piotr Golonka, our engineer, for performing a part of computations
we used in this paper.
Acknowledgements
Research study financed from funds for science in years 20072010 as a development project.
References
[1] Bartusiak, W., Ulaniuk, M., Safety barriers Delta Bloc, Magazine Autostrady, No. 5, 2005.
[2] Faller, R. K., Rohde, J. R., Rosson, B. T., Smith, R. P., Addink, K. H., Development of a TL-
3 F- Shape Temporary Concrete Median Barier, Report TRP-03-64-96, University of
NebraskaLincoln, USA 1996, http://engineering.unl.edu/specialty- units/MWRSF /MwRSF-
Downloads/TempBarrier/TRP-03-64-96.pdf .
[3] Grzebieta, R. H., Zou, R., Corben, B., Judd, R., Kulgren, A., Tingval, C., Powell, C.,
Roadside Crash Barrier Testing, Proceedings ICRASH, Melbourne 2002.
[4] Grzebieta, R. H., Zou, R., Jiang, T., Carey, A., Roadside Hazard and Barrier Crashworthiness
Issues Confronting Vehicle and Barrier Manufactures and Government Regulators,
19 Conference ESV, Washington 2005.
[5] Jackowski, J., Determination of the factor of friction between the concrete barrier and the
underlying surface, Sprawozdanie z bada, WAT, Warszawa 2009.
[6] MacDonald, D. J., Kirk, A. R., Precast Concrete Barrier Crash Testing, FHWA Final Report
SPR330, USA 2001.
[7] Minton, R. Cuerden, R., Behaviour of SUV and MPV- type Vehicles in Collisions with
Roadside Safety Barriers, 20 Conference ESV, Lyon 2007.
[8] Mikoajkw, L., Safety traffic barriers, WKi, Warszawa 1983.
[9] Mikoajkw, L., Traffic safety devices in Poland, Infrastruktura, Vol. 1-2, 2006.
[10] Navin, F., Kymchuk, R., Romilly, D., Thomson, R., Reconstruction of Accidents Involving
Highway Barriers, SAE 930656.
405
L. Prochowski
406