This report included an analysis of 110 accidents involving breakaway-cable-terminal (BCT) end treatments and 36 accidents involving median-breakaway-cable-terminal (MBCT) end treatments as used in kentucky. The primary data base consisted of kentucky accident records for the years 1980-87, with a few accidents that were identified before 1980. An attempt was made to document each accident with a police report, photographs, and a maintenance repair form. BCT end treatments evaluated included those with the terminal section installed as follows: (1) straight with no offset, (2) flared 6 feet at the end by using a 4.5-Degree simple curve over 125 feet, and (3) flared 4 feet with a parabolic curve over the last 37.5 Feet. Proper performance was based on a determination of whether the posts broke away as designed and/or the vehicle was redirected after impacting the guardrail. Results indicate that proper performance ranged from 60% for end sections with no offset to 69% for a "simple curve" offset, and 79% for a parabolic flare offset. Only 10 impacts were documented for small cars, and the BCT performed improperly in four of those accidents. Evaluation of the BCT end treatment indicates that it may be used where geometrics permit. Where those geometrics are not present, the turned-down end treatment proposed in a previous report should be used. The MBCT end treatment performed properly in 63% of the accidents. Problems related to stiffness of the end treatment were most apparent when impact angles were shallow. A recommendation was made to contour grade gore areas where possible and to install a crash cushion where the need for a barrier could not be eliminated. For MBCT installations at median piers and median width of 20 feet or less, crash cushions were also recommended. A turned-down end-treatment design was proposed for consideration at median piers where the median width was greater than 20 feet.
Abstract