View Q&A

In the early 1980's, Maurice Bronstad, et al, formerly of Southwest Research Institute, conducted a series of full-scale vehicle crash tests on three curved bridge railing configurations using three different test vehicles. The report references are as follows:

(1) Bronstad, M.E. and Kimball, C.E., Jr., Bridge Rail Retrofit for Curved Structures - Volume I: Executive Summary, Report No. FHWA/RD-85/040, Final Report Prepared for the Offices of Research and Development, Federal Highway Administration, Washington, D.C., September 1986.

(2) Bronstad, M.E. and Kimball, C.E., Jr., Bridge Rail Retrofit for Curved Structures - Volume 2: Technical Report, Report No. FHWA/RD-81/, Final Report Prepared for the Offices of Research and Development, Federal Highway Administration, Washington, D.C., June 1981.

The three test vehicles included an 1,800-lb mini-compact car, a 2,250-lb subcompact car, and a 20,000-lb school bus. All crash tests were performed at the target impact conditions of 40 mph and 15 degrees.

The three bridge railing configurations included:

(1) a New Jersey concrete safety shape bridge installed vertical to a super-elevated bridge deck surface,

(2) a New Jersey concrete safety shape bridge installed perpendicular to a super-elevated bridge deck surface, and

(3) a tubular thrie beam/collapsing tube retrofit bridge railing installed perpendicular to a super-elevated bridge deck surface.

The results and conclusions are as follows:

All three barrier systems contained and redirected the full range of test vehicles. In terms of stability and acceleration, the tubular thrie beam retrofit with a more vertical front face was superior but installed perpendicular to the superelevated deck.

For the safety shape testing, there was not a dramatic difference in performance for the two safety shape orientations. The preferred orientation, as determined from a 1976 concrete median barrier research program and from the later research results noted above, was to place the barrier perpendicular to the superelevation when the vehicle approach is up the superelevation. It was noted that vehicle climb was reduced by this preferred orientation in the car tests although only in the bus test was this significant. The authors stated that the school bus test was noticeably less severe in terms of vehicle redirection with the preferred perpendicular orientation. It was also observed that the small car was at the threshold of riding on top of the barrier for the barrier placed vertical to the earth and not perpendicular to the superelevated deck.

Finally, for barriers placed on the downside of the superelevation, I agree with the FHWA that those barriers are preferred to be placed vertical or perpendicular with respect to the earth and not with the superelevated deck.

FHWA Response:

This topic was discussed in the 1977 AASHTO Barrier Guide where it was concluded that the "optimum" orientation was vertical on the low side and perpendicular to the roadway surface on the high side.  However, that Guide also indicated the best "compromise" (or practical) orientation was vertical at both locations. A vertical wall (not NJ, F-, or single slope shape) would limit vehicular climb better than any shaped design.  No significant difference in crash performance between a NJ and a constant slope shape.