1: Breakaway components produce safer roadside signs. pp 1-29, 24 FIG, 1 TAB, 17 REF (Behavior of roadside sign supports subjected to collisions was described. Support selection was predicated on texas highway department design procedures and on development of interim concepts to minimize hazards. Full-scale crash tests to observe the impact behavior of supports with new devices resulted in revised design details which are now being used. For testing, a crash vehicle was towed into a controlled collision with a sign support, and the collision was recorded on high-speed photographic film. The film was studied to observe the qualitative behavior of the sign support subjected to a collision by an automobile. Displacement-time data were obtained from the high-speed film, and accelerometers mounted on the frame of the crash vehicle were used to provide a deceleration-time trace on a recording oscillograph. An electronic computer program also was developed to simulate the collision incident. Phenomenological testing was important in the testing procedure. Improvements in camera technique and film data reduction augmented and extended the phenomenological testing. These improvements produced quantitative analytical data. The mathematical model was a product of phenomenological and quantative testing.) 2: Automobile-barrier impact studies using scale model vehicles. pp 30-41, 9 FIG, 1 TAB, 11 REF (Geometric properties of rigid concrete median barriers and trajectory parameters of a class of impacting automobiles were to be correlated with resulting trajectory and vehicle attitude parameters after vehicle-barrier impact. Correlations were to be performed by analysis of high-speed movies of an automobile-barrier impact simulation done by 1/8-scale model vehicles and barriers. Major changes in vehicle attitude were found to be caused by relatively minor barrier design changes. Objectives were not fully accomplished due to lack of data needed for vehicle and barrier scaling. Modeling techniques were advanced to the point that, when validation data become available, scale models can be used to study the efficacy of both rigid and yielding barriers.) 3: Dynamic tests of corrugated metal beam guardrail. pp 42-87 The results of eight full scale vehicle impact tests into short sections (less than 100 ft.) Of 27-in. High blocked-out corrugated metal beam guardrail are reported. Two tests were performed on free standing unanchored short sections, three tests on free standing sections, using two different end anchorage systems and three tests on simulated bridge approach guardrail flares using a cable anchor assembly on the upstream or approach end and rigid attachment to the concrete bridge rail parapet at the other end. The tests were conducted at speeds ranging from 56 to 63 mph and approach angles varying from 24 to 33 degrees utilizing 1964 to 1966 sedans weighing approximately 4500 lbs. The first two tests (131 and 132) proved that short, free standing unanchored guardrail sections up to 62.5 ft in length and ineffective under severe impact loading conditions (approximately 60 mph/25 deg). This in turn indicates that any unanchored guardrail section, regardless of length, is vulnerable when struck within 30 ft of either end. The results of test 133 and 134 indicate that short guardrail sections with sloping rail anchorage ("texas twist) are structurally adequate when struck in the center (full height section), but performance is questionable with regard to impacts into the ramped ends. As a result of test 135 through 138, an effective cable anchoring device for short free standing sections of guardrail was developed. In addition, an efficient bridge approach guardrail flare design was developed which provides relatively smooth transition from the semi-flexible blocked-out beam barrier (8- by 8-in. Posts at 6-ft 3-in. O.c.) through a semi-rigid system (10- by 10-in. Posts at 3-ft 1-1/2-in. O.c.) to a rigid reinforced concrete bridge rail. 4. New highway barriers: the practical application of theoretical design. pp 88-183, 90 FIG, 7 TAB, 7 REF (Standard barrier designs for roadsides, medians, and bridges specified by the new york state department of public works were revised after 6 years of research. Comprehensive theoretical analysis of forces generated between vehicle and barrier during impact produced four mathematical models collision with a given barrier. Models were programmed for computer solution. Three provided the force deflection curve of the barrier in the case of a pure tension rail /cable/, a combination of tension and bending /w-beam/, and pure bending /box beam/. The fourth model gave the trajectory of the vehicle using the appropriate force-deflection curve as input. Forty-eight full-scale collisions between standard- size passenger cars and various barriers were carried out. Speeds of up to 60 mph and impact angles of up to 45 deg were selected as the most severe conditions expected on a highway. Dynamic tests were conducted on various guardrail posts embedded in different soils. Contribution of the post to barrier system strength and optimum post size and embedment conditions were determined. Results of tests relating directly to barriers finally adopted were presented. Improved barriers were described, and data were given on appropriate barrier selection.) 5. Objective criteria for guardrail installation. pp 184-206, 8 REF (The establishment of a more objective basis for warrants of guardrail placement on embankments and adjacent to freeway fixed objectives is studied. A mathematical model was developed to compare the relative safety at any embankment or fixed object location with the relative safety of protective guardrail at a similar location. At any location the relative safety is dependent on two variables---accident severity and accident frequency. If two similar locations, one with guardrail and one without, have the same accident frequency, the location with the lower accident severity is safer. The relationship developed to evaluate accident severity was named the severity index and was a weighted average of the accident severity values for a given condition using the ratios of direct costs of single vehicle accidents by severity class; fatal, injury, and property-damage-only (pdo). The ratios are 25, 6, and 1 for the fatal, injury and pdo accidents, respectively. The relationship developed to evaluate accident frequency was named the probability index and is simply the number of accidents per vehicles exposed to the condition. The true comparison measure, the product of the severity index and probability index, was called the collision index. Also a field inventory of fixed objects and fixed object guardrail was made on 95% of the 1963 freeway mileage to determine the relative exposures to each fixed object and gaurdrail type. The following conclusions were arrived at based on a collision index analysis: (1) guardrail placed adjacent to bridge rail ends, abutments, and piers decreases the collision index (improves safety). (2) guardrail placed adjacent to steel signposts also decreases the collision index (improves safety). (3) guardrail placed adjacent to lightpoles increases the collision index (decreases safety). (4) steel signposts placed in the off-ramp gore areas are considerably less safe than steel signposts adjacent to the highway shoulder. And (5) the use of dimensional lumber signposts in lieu of steel signposts reduces the accident severity.)
Abstract