Keywords: strain rate effect, steel, aluminium alloy, concrete, guard fence, vehicle collision, LS-DYNA, dynamic tensile test.

Guard fences are one of most important structures to ensure the safety of roads. Japan has the design specifications of guard fences, and it prescribes the required performance of guard fences [1]. According to the design specifications, the on-site full-scale collision experiments should be carried out to examine what happens to both guard fences and vehicles when vehicles collided with them. However, because of the huge consumption in time and cost, it is difficult in the field to measure the collision performance of guard fences for various cases. In Nagoya University, finite element (FE) models have been developed for colliding vehicles and various types of guard fences to re-enact their behaviour in on-site tests. The solutions in various simulations have been carried out using non-linear dynamic analysis software (LS-DYNA).

In the case of dynamic analysis, the stress-strain relationship of the materials of the guard fence such as steel and concrete should be clarified, and strain rate effect may influence the results of the numerical analysis [2,3,4]. In this study, dynamic tensile tests with SS400, STK400-plate and STK400-pipe, which are widely used as the materials of guard fences in Japan, are conducted to obtain the stress-strain relationship and the strain rate effect of them. Each stress-strain relationship and strain rate effect is confirmed through the tests. Dynamic tensile tests clarify that strain rate effects are clearly different in the three types of steels. These results are applied to the steel guard fence model presented in [3] to enhance it.

Three kinds of numerical collision analyses with the steel guard are performed. The first model has the strain rate effect obtained from the previous paper [4], and the second one has it obtained from the dynamic tensile test, and then the third one does not have a strain rate effect. Comparing the three results, they are obviously different to each other, and second one is similar to the on-site experimental result. From these results, the appropriate strain rate effect model must be taken into consideration when the performance of the steel guard fence is confirmed through numerical analysis.

In a previous paper [3], the dynamic tensile test with aluminium alloy was also carried out and its results were applied in the FE model of the aluminium alloy guard fence. Comparison between the numerical analysis result with the strain rate effect and the analysis result without it, shows that they are almost the same. This research indicates that the strain rate effect of the aluminium alloy is not so significant in checking the performance of the guard fence through numerical analysis. The importance of the strain rate effect is obviously different between the steel and the aluminium alloy.

In Nagoya University, the performance of concrete guard fences have been confirmed through computer simulations as well as the steel and aluminium alloy guard fence [5]. The previous paper [5] concludes that the FE models of the vehicle collision with the concrete guard fence can accurately reproduce both the vehicle and the guard fence behavior.

The purpose of the research in this paper is to check how the strain rate effect influences the analytical result. It is accepted that concrete strength is highly sensitive to the loading procedure, especially in the range of high strain rates. This research carries out the numerical collision analyses to the concrete guard fence with the strain rate effect or without it. The analysis results clarify that the strain rate effect may not be so important in the case of the numerical analyses with the concrete guard fence, because the concrete does not deform so much practically in this research. If the kinetic energy of the colliding vehicle increases or the structure of the concrete guard fence is not so strong, the guard fence is assumed to be severely damaged and the influence of the strain rate effect appears larger. This point must be confirmed in the future.

1

Japan Road Association, Design Specifications of Guard Fences, Maruzen Press, 1999 (in Japanese).

2

Itoh Y., Chunlu L. and Usami K., Non-linear Collision Analysis of Heavy Trucks onto Steel Highway Guard Fences, Proceedings of Structural Engineering and Mechanics, vol.12, No.5, pp.541-558, Kochi, 2001.

3

Itoh Y., Usami K. and Sugie M., Numerical Analyses on Impact Performance of Steel and Aluminium Alloy Bridge Guard Fences, Proceedings of Structure Under Shock and Impact 4, pp.385-394, Boston, 2000.

4

Takahashi Y., Ohno T. and Ota T., The Strain Rate Effect for Elasto-Plastic Behaviour of Loaded Armoured Concrete Beam, The Journal of Structural Engineering, Vol.37A, pp1567-1580, 1991 (in Japanese).

5

Hattori R., Itoh Y. and Kusama R., Numerical Collision Analysis of Concrete Guard Fence for Performance-Based Design, Proceedings of the Ninth East Asia-Pacific Conference on Structural Engineering and Construction (EASEC-9), Bali, 2003.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £140 +P&P)