Category HIGHWAY ENGINEERING HANDBOOK

Locations which would benefit from the installation of supplemental advance warning devices typically exhibit safety and/or operational problems. Establishing the need for supplemental devices, therefore, requires identifying the problem locations and performing a safety and/or operational analysis. Deficient locations can be identified by a traffic safety management system, citizen complaints, employee observations, and by safety analysis during a planned resurfacing, restoration, and rehabilitation (RRR) project.

Accident-based studies are used to identify locations that can be considered haz­ardous due to a large number of accidents. These studies involve the review and analysis of systemwide accident information. To compare the accident experience of several locations, the length of time over which accidents are counted, the traffic vol­umes, and the length of roadway section involved should be the same at each location. If not, accident rates may be compared between locations, provided that a common unit of exposure (e. g., accidents per million vehicle miles for longer roadway sections, or accidents per million entering vehicles for spot locations and intersections) is used.

Potential locations can also be identified by complaints received from citizens and by observations made by employees. Often a combination of accident analysis and an investigation of complaints and observations is required for low-volume roadways. Complaints about “near misses” and observations of hazardous roadway elements can be considered indicators of site deficiencies. This type of information is treated by some agencies with the same importance as a documented accident history. Such treat­ment has the advantage of reducing the number of accidents required to identify the hazardous roadway locations.

It should be recognized that maintaining a complaint and employee observation file requires that the agency be responsive to these inputs. Complaints and observations are notifications of hazards that become a matter of public record and are available as evidence should an accident result in litigation. This alone is not a valid reason to fail to maintain a complaint and observation file. If a defect is allowed to remain for an unreasonable period of time, even if no complaints or observations were received, the courts can consider it as constructive notice and assign liability. Complaint and obser­vation files should, therefore, be maintained and a program established to respond to all complaints and to document facts and engineering decisions to minimize the possi­bility of lawsuit losses.

An opportune time to identify the need for a device is during the design phase of pro­jects primarily intended to upgrade the physical and operational characteristics of the roadway. This opportunity can be used to detect safety and operational deficiencies and to select appropriate improvements that can be incorporated into the upgrading project.

The identification of potential locations for each of the previous methods should include a field inspection to help establish the cause of the deficiency and appropriate countermea­sures. If the site inspection indicates that the deficiency cannot be readily corrected due to cost or physical constraints, then an advance warning device should be installed. If the site conditions are sufficiently unusual that an appropriate warning device is not con­tained in the federal or appropriate state MUTCD, then a supplemental device may need to be used or developed until it is feasible to take care of the underlying problem.

For example, consider a situation where a sag vertical curve was constructed to provide sufficient vertical bridge clearance on a roadway with a posted speed of 45 mi/h (70 km/h). Analysis of the areawide accidents indicated that there is a higher than expected occurrence of intersection-related and rear-end accidents at a signalized intersection immediately downstream of the bridge. A visit to the site indicated that the signal faces were not visible to approaching drivers until they were 400 ft (120 m) from the stop line. Since this distance is less than the minimum visibility distance of 460 ft (140 m) specified by Sec. 4D.15 of MUTCD, a Signal Ahead sign (W3-3) was installed [2]. The engineer determined that, although the minimum recommendations of MUTCD were being achieved, safety improvements could be achieved by providing real-time warning that a stop will be required at the intersection. Since removing the sight obstruction was not

possible, the engineer considered lowering the speed limit and/or providing additional motorist warning. Experience with lowering speed limits indicated that this countermea­sure was not an effective long-term solution. The engineer decided to install an active supplemental advance warning device with the legend “Prepare to Stop When Flashing” configured as shown in Fig. 7.1. The device was installed over the roadway, 500 ft (150 m) in advance of the stop bar, and interconnected with the traffic signal controller. The hor­izontally mounted beacons were timed to flash yellow 8 s prior to the red indication so that drivers passing the beacon at the legal speed limit would have advance warning of the required stop at the intersection. The yellow beacons continued to flash until 3 s before the end of the red indication to allow the start of queue dissipation. Motorists not encountering the flashing lights could expect not needing to come to a complete stop at the signal, while still having the signal presence reinforced by the overhead sign. The engineer plans to continue monitoring the location to determine if the active advance warning device is effective in reducing accidents.

Designing a warning device that provides a clear, unambiguous message to the motorist can be a difficult task. The difficulty is due in part to the concern of the engineer to act in a “reasonable and prudent” manner. Increasing motorist safety and minimizing lia­bility require that the device provide a readily understood and unambiguous message.

In the design of warning signs, it is important to remember that signs are designed to draw attention to themselves through contrast, color, shape, composition, reflector – ization, and illumination, with a simple message providing a clear and understandable instruction to the motorist. Sign size, symbol size, lettering size, and placement should be such to allow adequate time for proper response. Uniform and reasonable instruc­tions to the motorist will instill respect and develop willing compliance with the sign message. For these reasons, the majority of general warning signs should be designed as diamond shapes with black letters on a yellow background. Standard sign letters are prescribed in FHWA Standard Highway Signs, which should be used to develop let­tering size and style [9]. Sections 2C.01, 2C.02, 2C.03, and 2C.04 of MUTCD contain information that must be followed in the design of warning signs. In addition, Sec. 1A.11 of MUTCD lists additional publications and documents that provide requisite information for the proper design of warning signs.

Estimates by the FHWA indicate that there are an average of 15 signs per mile on the nation’s 3.8 million miles of streets and roadways [6]. The resultant 57 million traffic signs represent a huge investment in materials, labor, equipment, and maintenance costs. While this is a significant investment, improvements using standard traffic control signing are reported in the “1988 Annual Report on Highway Safety Improvement Programs” as having the highest benefit-cost ratio of any highway safety improvement [7]. Properly designed, located, and maintained standard traffic signs and other carefully conceived devices can be an effective method of increasing traffic and operational efficiency and subsequently decreasing the tort liability exposure of roadway agencies.

The concerns about tort liability judgments are valid, as the number of cases is steadily increasing. In almost every state, the shield of sovereign immunity either has been abolished by judicial decisions or has been eroded by legislative modifications to governmental immunity. In one state, for example, the legislature was instructed to enact comprehensive tort claim procedures in the near future or the doctrine of immu­nity would be abrogated by the state supreme court. In another state, the concept of sovereign immunity was declared unconstitutional [8].

A tort is a civil wrong or injury. The purpose of a tort action is to seek compensation for damages to property and individuals. The following elements must exist for a valid tort action:

• The defendant must owe a legal duty to the plaintiff.

• There must be a breach of duty; that is, the defendant must have failed to perform a

duty or performed it in an improper manner.

• The breach of duty must be a proximate cause of the accident that resulted.

• The plaintiff must have suffered damages as a result.

In highway-related tort cases, the first element is relatively easy to establish. Roadway authorities have been vested with the responsibility of providing reasonably safe travel opportunity for roadways under their jurisdiction. The failure of the roadway agency to properly perform that duty, and that this breach of duty was the proximate cause of an accident, are more difficult to establish. In most instances, establishing that a breach of the legal duty occurred becomes a major issue in tort liability cases. Plaintiffs typically will attempt to establish that the agency having roadway jurisdiction was negligent in its duty and/or a physical condition was permitted to exist that was a hazard.

Negligence is the failure to exercise such care as a reasonably prudent and careful person would use under similar circumstances. Roadway agencies can be judged neg­ligent in two ways: (1) wrongful performance (misfeasance), or (2) the omission of performance when some act should have been performed and was not (nonfeasance). Roadway agencies can, therefore, be judged negligent either by addressing a safety problem incorrectly or by ignoring it. The critical issue in highway tort liability is the care with which highway agencies perform their responsibilities. If it is judged that a reasonable standard of care was not exercised, then the responsible persons and/or organizations may be held liable for injuries and damages that resulted.

In an attempt to familiarize roadway agencies and their employees with the potential liability, and to make them aware of their duties and responsibilities to the traveling public, the NCHRP published Synthesis of Highway Practice 106: Practical Guidelines for Minimizing Tort Liability [8]. In particular, this publication advises agencies to supply a consistent highway environment for motorists. The use of standard design fea­tures and uniform traffic control devices is also emphasized.

All states are to adopt the standards of MUTCD as the basis for designing and installing traffic control devices. Some states adopt MUTCD in its entirety, while other states incorporate additional devices and practices into their manuals which address their specific roadway design and driver expectancy needs. MUTCD provides minimal require­ments, and states that do prepare their own manuals are required to conform to the nation­al standard. Additional devices not included in either the federal MUTCD or those of the states are frequently developed to provide motorist warning of roadway hazards which, ideally, should be eliminated. The reasons for not eliminating the hazard can include geo­metric constraints, planned improvements, usefulness of the condition for other purposes, burden of removing the condition, and the lack of a method to correct the situation. When the need to warn motorists involves commonly encountered hazards, such as a stop sign ahead on a rural roadway, then an appropriate warning device can be found in MUTCD. When the hazard is posed by unusual or unique conditions, however, the highway engi­neer is placed in the difficult position of identifying, or often designing, a warning device that provides a clear message to the motorist of the potential hazard. It should be empha­sized that the installation of a warning device does not remove the agency from liability, especially if it can be shown that it was reasonably possible to eliminate the hazard.

While the advantages of uniformity far outweigh the disadvantages, there are some undesirable effects when complete uniformity is maintained. One of the principal disad­vantages is that strict uniformity may result in the failure to adopt an improved device or procedure simply because it is not in common use. In addition, total uniformity would require the specification of a separate traffic control device for every conceivable road­way geometric and traffic operational condition. This would be a monumental task that undoubtedly would still not cover every situation, while simultaneously increasing the size of MUTCD with devices of limited application.

This difficulty is recognized in MUTCD, which indicates that warning signs other than those specified in the manual may be required under special conditions [2, Sec. 2C.02]. MUTCD requires exercising good engineering judgment in determining the need for other warning devices. It also mandates that the innovative devices be under­stood easily by the motorist. Ensuring that warning signs are easily understood neces­sitates that they be of standard shape and color and that the legends be unambiguous and brief. Establishing the need for distinct warning devices can be accomplished by identifying when standard devices do not properly address unusual conditions. While these conditions are unusual, they can typically be classified into the same use cate­gories that are appropriate for standard warning signs. The Traffic Control Devices Handbook [3] identifies the following uses of warning devices:

• To indicate the presence of geometric features with potential hazards

• To define major changes in roadway character

• To mark obstructions or other physical hazards in or near the roadway

• To locate areas where hazards may exist under certain conditions

• To inform motorists of regulatory controls ahead

• To advise motorists of appropriate actions

The need to provide advance warning for unusual roadway, roadside, operational, and environmental conditions has resulted in the development of a wide diversity of devices. The majority of these devices can be categorized as warning signs containing different symbols and legends. Other warning devices include flashing beacons, rumble strips, pavement surface treatments, and pavement markings. Device complexity ranges from simple passive warning signs to devices that are activated by vehicle speed, headway, or presence on one or more approaches to a potentially hazardous roadway element. Further information on supplemental warning and rumble strips can be obtained from the National Cooperative Highway Research Program (NCHRP) publi­cations Synthesis of Highway Practice 186: Supplemental Advance Warning Devices and Synthesis of Highway Practice 191: Use of Rumble Strips to Enhance Safety [4, 5].

PART 1

SIGNING

Brian L. Bowman, Ph. D., PE.

Professor of Civil Engineering Auburn University Auburn, Alabama

Part 1 of this chapter presents a comprehensive review of the design, construction, and maintenance of highway signs. Both single – and multiple-mounted sign supports are addressed, with an emphasis on highway safety. Breakaway supports with various types of slip bases, frangible bases, and post hinging systems are explained and illustrated. Commercially available devices and alternatives are identified and discussed. Guidelines on use and construction are summarized. An extensive list of references, which are noted in the text, concludes the section. Much of this material was derived from studies made by the author under a Federal Highway Administration project, NHI 38034, “Design, Construction and Maintenance of Highway Safety Features and Appurtenances.”

71 TRAFFIC SIGNING NEEDS

The capability of roadways to safely and efficiently serve vehicular traffic is depen­dent to a large extent on the adequacy of traffic control devices. The majority of motorists drive in an orderly and safe manner, provided they are given reliable regula­tory, warning, and guide information. Motorists, through training and experience, develop expectations on when and in what manner they will be provided necessary information for safely controlling their vehicles. Motorists expect that similar traffic control devices will always have the same meaning and will require the same motorist action regardless of where they are encountered. This expectation has been enhanced by the use of uniform traffic control devices which enable motorists to consistently interpret the general intent of a device by its message, shape, and color.

The advantages of traffic control device uniformity were recognized long ago. The American Association of State Highway Officials published specifications of road markers and signs for rural roadways in 1925. A manual for urban roadways was pub­lished in 1929 by the National Conference on Street and Highway Safety. The unifica­tion of the standards applicable to the different classes of roadways was addressed by a joint committee of the American Association of State Highway Officials and the National Conference on Street and Highway Safety. The joint committee developed, and printed in 1935, the first Manual on Uniform Traffic Control Devices for Streets and Highways (MUTCD) [1]. That joint committee, although subsequently reorganized and named the National Committee on Uniform Traffic Control Devices (NCUTCD), has been in continuous existence and contributes to periodic revisions of MUTCD.

The benefits of traffic control device uniformity include increasing safety by providing the road user with required information for vehicle guidance or control at the right time and place and in the proper manner. Signs should be installed only where warranted. This can include locations where special regulations apply at specific places or specific times or where hazards are not self-evident. They also provide information of highway routes, directions, destinations, and points of interest. The general standards for signs provided in Chap. 2A of the MUTCD and those sections pertaining to the particular type of sign being installed should be followed to ensure proper placement and message uniformity [2].

For proper performance, crash cushions should be placed on level terrain with a clear path between the roadway and the attenuator so the vehicle can strike at normal height, with the suspension system in a neutral state. Avoid curbs or slopes in front of the device. Install the attenuator on a smooth surface (usually concrete) so it can com­press uniformly. Conspicuous, well-delineated crash cushions are less likely to be hit than those that blend into the background. If the system is not reflective, install stan­dard object markers to improve visibility at night and during inclement weather.

Selection of the most appropriate crash cushion depends on site characteristics, perfor­mance of the systems, maintenance characteristics, and life-cycle cost. Both the geometrical conditions encountered and the space requirements for the different systems vary widely. Obstacles greater than 16 ft (5 m) wide can be shielded by systems such as arrays of sand-filled barrels, or the bullnose attenuator. Where space is limited, narrow systems are appropriate. The structural and safety characteristics of alternative systems must be carefully reviewed and matched with needs. Items to consider include impact deceleration, redirection capability, impact debris, and anchorage and backup require­ments. Table 6.10 has been prepared to compare the maintenance requirements of the different systems. Agency maintenance records should be used to establish associated costs. After potential systems have been identified for a given site, the final selection should be based on a life-cycle cost analysis. (See Chap. 10.) Costs to consider are the initial cost of the device, site preparation and installation costs, and maintenance costs, as well as the cost of accidents.

Crash cushions are impact attenuators developed to prevent errant vehicles from impacting fixed obstacles. The crash cushion should either decelerate the vehicle to a safe stop, such as in a head-on hit, or redirect it safely away from the obstacle, in the case of a side hit. Crash cushions are typically used where fixed objects cannot be removed, relocated, converted to a breakaway design, or shielded by a longitudinal barrier. Examples of application sites include exit ramp gores where a bridge rail end or bridge pier presents a hazard, and the ends of longitudinal barriers. Most crash cushions are patented systems developed and tested by the manufacturer, who can also provide design charts for selection of appropriate designs.

Most crash cushions perform their function by the principle of kinetic energy absorption or transfer of momentum. In the first case, energy is absorbed by materials or devices that crush or plastically deform, or by hydraulic devices. A rigid backup support is required for such compression crash cushions. In the second case, the momentum of the vehicle is transferred to an expendable mass, such as containers filled with sand. No rigid backup support is needed for such “inertial” barriers. Some crash cushions use a combination of these principles.

Table 6.9 provides a list of some of the most common crash cushions in use today, and their applicable test levels. A description of these systems follows.

The Advanced Dynamic Impact Extension Module (ADIEMII) is a proprietary terminal designed to shield the end of a concrete safety-shape barrier. It consists of a 30 ft-long (9.1-m) carrier beam or concrete base structure onto which 10 interlocking perlite con­crete crushable modules are mounted. Energy is dissipated by crushing the modules. Perlite is an expanded inert mineral that, when substituted for coarse aggregate in a concrete mix, results in an extremely lightweight and crushable material. Strength lev­els in the perlite concrete are closely controlled to ensure that it falls within acceptable levels compatible with acceptable vehicle deceleration limits.

The Brakemaster 350 is a proprietary design used primarily as a terminal for W-beam median barriers or as a crash cushion to shield narrow obstacles. If used to terminate a concrete median barrier, a suitable transition is required between the device and the con­crete. It may also be used to shield the end of a roadside barrier but may not be cost – effective. The manufacturer recommends use in low-frequency impact areas. This terminal consists of an anchor assembly with posts embedded in the ground, a cable/brake assembly, and W-beam panels supported by steel diaphragms that slide backward in end-on hits. When impacted end-on, the W-beam panels telescope and the cable/brake assembly absorbs most of the energy through frictional resistance. The anchor assembly also provides sufficient anchorage to redirect side-impacting vehicles. A paved instal­lation pad is not required.

The Crash Cushion Attenuating Terminal (CAT) is a proprietary, nonflared attenuator commonly used to terminate W-beam median barriers and as a crash cushion to shield narrow fixed objects. A transition design is required for the latter case. It is sometimes used to shield a W-beam roadside barrier, but in that case, a cable anchor is required at the downstream. The CAT can redirect vehicles striking its face from one side or both

sides. The CAT functions as a three-stage system, utilizing energy-absorbing beam elements, breakaway wood posts, and a cable anchorage system. The beam element is a slotted W-beam that telescopes during impact. Shearing of the steel rail between the slots dissipates energy.

The bullnose guardrail system provides a nonproprietary means for shielding an object in the median of a divided highway by constructing a thrie-beam guardrail envelope around the end. Several such designs using W-beam guardrails have been constructed by highway agencies in the past, but these did not meet the criteria of NCHRP Report 350. However, a design that has met TL-3 consists of slotted thrie – beam panels mounted on breakaway posts near the nose, followed by standard thrie-beam posts and blocks toward the back of the system. Rail tension is developed through cable anchors and struts. A set of steel retention cables is mounted on the back of the thrie-beam nose to contain vehicles in the event of rail fracture. The leading edge of the bullnose attenuator should be located a minimum distance of 62 ft (19 m) in advance of the shielded object.

The ABSORB 350 is a proprietary, nonredirective, crash cushion primarily designed to shield the ends of the Quickchange® median barrier. This is a narrow cushion that may also be used to shield ends of concrete barriers or narrow fixed objects. The system is comprised of multiple, water-filled, energy-absorbing elements; a nosepiece assembly; and a transition/attachment assembly. Three length configura­tions are available.

QuadGuard refers to a family of proprietary devices with similar design and per­formance characteristics. The design consists of several types of energy-absorbing cartridges supported by a framework of steel diaphragms and corrugated steel fender panels. A concrete pad and rigid backup are required. Crash energy is dissipated by telescoping rearward and crushing the cartridges. The devices meet TL-2 or TL-3, as indicated in Table 6.9. Many parts of the various systems are interchangeable. The standard QuadGuard is a bidirectional device used as an end treatment for a concrete barrier or narrow fixed object. The cartridges must be replaced after an impact. The QuadGuard Wide is similar but can be used to shield wider objects. The QuadGuard LMC (Low-Maintenance Cartridge) is a self-restoring, bidirectional end treatment used at locations where a moderately high frequency of impacts is anticipated. It can be used to shield rigid barriers or fixed objects and is available in two widths. Energy­absorbing components are elastomeric cylinders that are reusable after most design impacts. The QuadGuard Elite is a self-restoring, bidirectional end treatment for loca­tions with high-impact frequency. It can be used to shield rigid barriers or fixed objects and is available in two widths. The energy-absorbing components are high – density polyethylene cylinders that are reusable after most design impacts.

The Trinity Attenuating Crash Cushion (TRACC) is a proprietary system. Components include a pair of guidance tracks, an impact “sled,” intermediate steel frames, and W-beam fender panels. A concrete pad and rigid backup are required. The sled, or impact face, contains a hardened steel blade that absorbs energy by cutting metal plates on the sides of the guidance tracks as it is forced backward. The interme­diate frames that support the fender panels are free to slide backward on an end impact, but lock onto the guidance tracks on a side impact to redirect the vehicle.

The Reusable Energy-Absorbing Crash Terminal (REACT 350) is a proprietary system comprising single row of 0.9-m-diameter (3-ft), high-density, polyethylene cylinders atop steel skid rails; a restraining cable system consisting of two heavy steel wire rope assemblies along each side; a front and rear anchorage system; transition hardware; and a backup assembly. A nine-cylinder array meets TL-3 and a four-cylinder design meets TL-2. The system may be used on either a concrete or an asphalt surface if properly anchored. The polyethylene cylinders absorb energy as they slide rearward on the steel railing, and are self-restoring in many cases. The steel cables redirect vehicles in side impacts. A wider REACT that can be used to shield fixed objects up to 5 ft (1.5 m) was tested successfully to TL-3. This design consists of two parallel columns of 2-ft-diameter (0.6-m) cylinders attached to steel diaphragms mounted on an anchored monorail, which provides redirection for side impacts.

The narrow Connecticut impact attenuation system (NCIAS) is a nonproprietary, bidirectional crash cushion that consists of eight steel cylinders in a single row with two anchored wire tension cables along each side. The cylinders, which are 3 ft (0.9 m) in diameter and 4 ft (1.2 m) high, crush to absorb energy. The tension cables keep the cylinders in place and provide redirection for side impacts. The last four cylinders are reinforced with pipe stiffeners and retainers to help redirect vehicles hitting close to the rear. The NCIAS is recommended for use where shielding of narrow objects is needed and reverse-direction impacts are unlikely.

Sand-filled plastic barrels, sometimes called inertial crash cushions or inertial barriers, are used in both temporary and permanent installations to shield the ends of longitudinal barriers or other fixed objects. The sand-filled barrels dissipate energy by transferring vehicle momentum to the variable masses of sand in the barrels that are hit. Standard module masses vary from 200 to 2100 lb (90 to 960 kg). A backup struc­ture or wall is not required because the force that a vehicle exerts on the individual modules is not transmitted through the cushion. Manufacturers have developed stan­dard arrays that can be used for specific types of fixed objects as well as design charts to analyze layouts. The barrels should be set as far from the traveled way as possible to minimize hits. However, the width of the last row should always be greater than the width of the shielded object. Moisture content of the loose sand should be 3 percent or less and clean sand should be used to minimize caking. A significant variation in the density of the sand could affect performance. Frozen sand reduces safety performance but mixing rock salt (5 to 25 percent by volume, depending on climate) with the sand generally prevents wet sand from freezing. The use of sacked sand to facilitate cleanup is not acceptable.

The gravel-bed attenuator provides a means to decelerate large trucks. Basically, the truck is slowed as the wheels move through a bed of gravel. It is typically used on truck escape ramps along descending highway grades where runaway vehicles present a problem. Crash cushions previously discussed are designed to stop or redirect passenger cars and pick­up trucks. They are not applicable to large vehicles, because considerable space is required to dissipate the energy. Detailed design guidelines for the gravel-bed attenuator are provided in the AASHTO publication A Policy on Geometric Design of Highways and Streets.

The Dragnet or chain-link fence vehicle attenuator is a proprietary device consist­ing of anchor posts, energy-absorbing reels of steel tape, and a net assembly to catch the vehicle. When impacted, the chain-link fence wraps around the front of the impacting vehicle and energy is absorbed as the metal tape is pulled through a series of rollers. The system may be repaired by replacing the steel tape in the casings and resetting the chain-link fence and cable. The Dragnet may be considered for locations where impacts are expected to be head-on and the results of vehicle penetration are severe, such as for temporary road and ramp closures, or in conjunction with a longi­tudinal barrier to shield the opening between twin bridges. It is designed to stop a 4500-lb (2000-kg) passenger car impacting head-on at 60 mi/h (100 km/h). It has also been used in series to stop large vehicles where space will not accommodate a gravel bed attenuator. Such a system safely stopped at 50,000-lb (22,700-kg) tractor-trailer impacting at 90° and 50 mi/h (80 km/h). Since the Dragnet deflects significantly, it can be used effectively only at locations where a sufficient clear area exists behind it. Because of the low deceleration rates resulting, very little damage is done to impacting vehicles and serious injuries to vehicle occupants are unlikely.

The Water Twister Vehicle Arresting System (VAS) is a proprietary system consisting of a chain-link restraining net connected to two energy-absorbing base units by nylon straps. As an impacting vehicle displaces the net, the straps turn shafts connected to turbine rotors inside the base units, which contain a water/ethylene glycol solution. Rotation of the turbine blades in the fluid dissipates energy. The base units are of sub­stantial size and may require shielding.

Barrier terminals and crash cushions are developed to gradually decelerate an impact­ing vehicle to a stop or to suitably redirect it. Otherwise, untreated ends of barriers and fixed objects can cause severe accidents. A crashworthy end treatment is essential if a barrier terminates within the clear zone or other area where it is likely to be hit by an errant vehicle. Requirements for testing and performance are contained in NCHRP 350. Suitable devices must be able to perform under both head-on and side impacts, with no objects penetrating the passenger compartment or encroaching on other traffic. The vehicle should remain upright and not be redirected into adjacent traffic lanes. Occupant deceleration levels must be within target values. For longitudi­nal barriers that depend on the tensile strength of the elements, the end treatment must

6.12.1 Characteristics of End Treatments

Many types of end treatments are available. Table 6.8 provides a summary of charac­teristics and test levels for a number of them. A description follows. As indicated, many of the systems are proprietary.

The three-strand cable terminal is used at the ends of a three-cable barrier. In one version, the cable barrier is flared back at 4 ft (1.2 m) from the tangent barrier line. The three cable strands are turned down at 45° and anchored to a concrete block in the ground.

The Wyoming Box Beam End Treatment (WYBET-350) is used with the box-beam barrier. It consists of a nosepiece welded to a box beam, which is inserted into a larger tube that contains a crushable fiberglass composite tube. The device is supported by a wood post. Crushing of the composite tube dissipates the energy. It may be installed parallel to the roadway or flared out at a maximum rate of 1:10.

A barrier anchored in backslope is sometimes used in areas of a roadway cut sec­tion, or where the road is transitioning from cut to fill. A W-beam guardrail thus anchored has been successfully crash-tested to TL-3. This type of anchor can provide full shielding, eliminate the possibility of an end-on impact with the barrier terminal, and minimize the likelihood of the vehicle passing behind the rail. According to the AASHTO Roadside Design Guide, key design considerations include the following:

(1) maintaining a uniform rail height relative to the roadway grade until the barrier

crosses the ditch flow line, (2) using a flare rate within the clear zone that is appropriate for the design speed, (3) adding a rubrail for W-beam guardrail installations, and (4) using an anchor that is capable of developing the full tensile strength of the W-beam rail. Also, the foreslopes on the approach should be no greater than 1:4. If a barrier cannot be termi­nated in a backslope without violating any of these principles, a different type of end treatment may be more appropriate.

The eccentric loader terminal (ELT) evolved from efforts to improve the break­away cable terminal (BCT). The resulting device has a fabricated steel lever nose inside a section of corrugated steel pipe. A strut between the steel tube foundations for the two end posts enables these posts to act together to resist impact loads. Holes are drilled in the next four posts, one hole at ground line and one below ground, to make them break away. A blockout is added to the second post to increase curvature near the end of the rail, thus reducing rail column strength and reducing the likelihood of rail penetration. The end post is offset 4 ft (1.2 m).

The Slotted Rail Terminal (SRT-350) is a proprietary, flared, non-energy-absorbing terminal, designed to break away when impacted end-on. There are two versions, one with an offset of 4 ft (1.2 m) and another with an offset of 3 ft (0.9 m). They consist of a curved W-beam rail element in which longitudinal slots have been cut at specific loca­tions. This reduces dynamic buckling strength to an acceptable level and controls buck­ling location, so that the yaw of an impacting vehicle and the potential for secondary impacts with the bent rail are minimized. Rail tension is developed through a cable anchor system. A traversable area must be provided behind the terminal since it is designed to break away when impacted, allowing the vehicle to travel behind the guardrail.

The REGENT is a proprietary energy-absorbing end treatment. It is a flared W-beam terminal that consists of a slider head assembly, a cable anchor/strut and yoke assembly, modified W-beam rail panels, and special weakened wood posts. The post offsets cor­respond to those of the BCT, except that the REGENT uses more posts to minimize deflection and the posts are of unique design. The modified rail elements are partially crushed at two locations to induce predictable kinks in the rail in end-on hits while maintaining most of the rail’s bending strength. A traversable area must be provided behind the terminal since it is designed to break away when impacted, allowing the vehicle to travel behind the guardrail.

The Vermont low-speed, W-beam guardrail end terminal is a nonproprietary end treatment for use on roadways where impact speeds do not exceed 45 mi/h (70 km/h). It consists of a 12.5-ft (3.8-m) W-beam rail section that is shop-bent to a 16-ft (4.9-m) radius and mounted on W6 X 9 W150 X 14 steel posts with steel blocks. An anchor consisting of a steel rod and buried concrete block is attached to the rail at the third post from the end.

The Flared Energy-Absorbing Terminal (FLEAT) is a proprietary energy-absorbing end treatment that consists of an impact head installed at the end of a modified W-beam rail, a guide tube assembly, a breakaway cable anchor assembly, and a series of weakened posts. The posts may be wood or of a welded-steel breakaway design. The kinetic energy of a crash is absorbed by the head sliding along the rail element while bending it. The flattened rail exits the head on the traffic side and coils into a tight loop. Tension in the rail is developed through the cable anchor system. The terminal has been tested successfully to TL-3 and TL-2, with a total length of 37.5 ft (11.4 m) and to 25 ft (7.62 m), respectively. The TL-3 terminal can be installed with an offset from 2.5 to 4 ft (0.76 to 1.2 m), and the TL-2 terminal with an offset from 1.7 to 2.7 ft (0.51 to 0.81 m). A traversable area must be provided behind the terminal since it is designed to break away when impacted, allowing the vehicle to travel behind the guardrail.

The Beam-Eating Steel Terminal (BEST) is a proprietary energy-absorbing end treatment with an impact head mounted on the end of a wood post W-beam guardrail system. Kinetic energy is absorbed by the head, which contains three teeth that slide along the rail and cut it into four relatively flat widths. These widths are subsequently bent out of the path of the impacting vehicle. A cable provides anchorage for down­stream impacts, and a quick release attachment allows the W-beam to feed into the impact head during end-on impacts. No flare is required, but to position the impact head entirely outside the shoulder, a 1:50 flare may be desirable.

The Extruder Terminal (ET-2000) is a proprietary energy-absorbing end treatment, with an extruder head installed over the end of a standard W-beam guardrail element. Kinetic energy is absorbed by the head sliding along the rail element while flattening it and bending it away from the traffic. The extruder head includes a squeezing section and a bending section. The W-beam is fed through the squeezing section, which reshapes the rail into a flat section. Next, the bending section bends the rail around a small radius and directs it to the side, away from the vehicle. A cable provides anchorage for downstream impacts, and a quick release attachment allows the W-beam to feed into the impact head during end-on impacts. No flare is required, but to position the impact head entirely out­side the shoulder, a 1:50 flare may be desirable. Either breakaway timber posts or hinged breakaway steel posts may be used with this terminal.

The Sequential Kinking Terminal (SKT-350) is a proprietary energy-absorbing end treat­ment that consists of an impact head mounted over the end of a modified W-beam guardrail. The modification consists of punching three slots in the valley of the rail at specific locations. The impact head is forced rearward, bending the W-beam rail against the deflector plate, and absorbing the kinetic energy. A “kinker” beam in the head causes short segments of the rail to kink sequentially and bend away from the impacting vehicle. A cable anchorage system is provided to develop the tensile strength of the rail. No flare is required, but some offset is recommended to locate the edge of the impact head farther from the traveled way. Either breakaway timber posts or hinged break­away steel posts may be used with this terminal.

The QuadTrend-350 is a proprietary unidirectional end treatment for direct attachment to a vertical concrete barrier or vertical concrete bridge parapet. Additional transition guardrail sections are not needed. It employs sand-filled, energy-absorbing containers that are sacrificial and must be replaced following impact. Many of the other parts can be reused. A concrete pad is required.

The Narrow Energy-Absorbing Terminal (NEAT) is a proprietary, narrow, nonredirec­tive, energy-absorbing terminal. The NEAT is an aluminum cartridge, designed to shield the approach end of a portable concrete safety-shape barrier or a Quickchange® moveable barrier system.

A sloped concrete end treatment is sometimes used to terminate a concrete barrier, although this end tapering treatment has not met the crash-testing criteria of NCHRP

Report 350. This treatment should only be considered for locations where traffic speeds are low, 40 mi/h (60 km/h) or less, and limited space precludes the use of a tested end treatment. Other possible applications include locations where the barrier is flared out beyond the clear zone or where end-on impacts are not likely to occur. Recommended length of the taper is 20 ft (6 m) with 30 to 40 ft (9 to 12m) desirable. The height of the end of the taper should be no greater than 4 in (100 mm).

Most of the principles previously discussed for median transitions (Art. 6.9.3) apply

here as well. Transition designs should gradually stiffen the approach system to avoid

vehicle pocketing, snagging, or penetration. Some considerations of importance follow.

The concepts are appropriate for both new construction and retrofits.

• The splice between the rail of the approach barrier and the bridge rail should develop the tensile strength of the approach rail.

• Strong-post systems, or combination normal-post and strong-beam systems, can be used for transitions. These systems normally should be blocked out to avoid snagging. Also, a rub rail may be desirable with W-beam or tube-type transitions. Tapering the rigid bridge railing end behind the transition members may also be desirable. The rub rail and railing taper are specially appropriate when the approach transition is recessed into the end of a concrete railing or other rigid hazard.

• Use a gradual transition, typically 10 to 12 times the difference in lateral deflection of the two systems. Gradually stiffen by decreasing post spacing, increasing post size, and strengthening the rail (nested W-beams or thrie-beams, for example).

• Eliminate curbs, inlets, and other drainage features in front of the barrier. Keep the slope between the edge of the driving lane and the barrier to 1:10 or less.

• When possible, relocate roads that intersect near the end of the bridge and interfere with a proper transition. Crash cushions may provide an option in some cases.