In the summer of 1965, the first fatality report involving an interstate median bridge pier in Kentucky caused concern among state and national officials for the safety of motorists who perchance or otherwise enter upon a collision course toward an unprotected bridge pier. A consensus of opinion seemed to indicate that some form of attenuation or deflection device was necessary.

Early innovations employed various short guardrail configurations to deflect wayward vehicles from the piers. The use of small, short sections has since evolved until present methods include surrounding the bridge pier with several hundred feet of guardrail, including ramped-end treatment (see Figure 1). The use of guardrail has not been questioned from the standpoint of safety design, yet some effort has been applied by Kentucky and other states to finding an alternative approach.

From these efforts, the use of mounds to decelerate and deflect vehicles originated. It was thought that this design concept was consistent with current safety developments as well as being an economical treatment of the problem, since most of the work involved in constructing the mound can be done during grade and drain construction using natural materials available on location. These mounds have since been constructed on certain interstate projects and on all bridge locations of the Jackson Purchase (51.4 miles) and Pennyrile (56.6 miles) Parkways in Kentucky. Having found no records of accidents involving these mounds in the interim, there has been no substantial means of evaluating the effectiveness of this innovation in preventing or reducing the severity of collisions with bridge piers. Consequently, it was decided that low speed excursions over a mound might provide a basis for evaluation.

The Division of Research made a series of driver-controlled traverses at low speeds on a typical site constructed on a 60-foot median. The purpose of this report is to summarize the results obtained from these initial, low-speed tests and, in so doing, attempt to make some determination of the reliability of this particular type of earthwork. It is anticipated that these low-speed tests will be supplemented eventually by testing at higher speeds, using some form of remote guidance system in place of the human driver. Also, it must be emphasized that conclusions drawn from this report apply only to low-speed encroachments. Additional effects which may be encountered during high-speed testing can only be hypothesized at this time.

Report Date


Report Number

No. 280

Digital Object Identifier