CHAPTER 17 EMERGENCY ESCAPE RAMPS

Transcription

1 CHAPTER 17 EMERGENCY ESCAPE RAMPS 17.0 INTRODUCTION Where long, descending grades exist or where topographic and location controls require such grades on new alignment, the design and construction of an emergency escape ramp at an appropriate location is desirable to provide a location for out-of-control vehicles, particularly trucks, to slow and stop away from the main traffic stream. Out-of-control vehicles are generally the result of a driver losing braking ability either through overheating of the brakes due to mechanical failure or failure to downshift at the appropriate time. The loss of stopping capability of a heavy vehicle on a downgrade resulting in an "out-of-control" situation is a relatively infrequent event. The results of that event, however, in many cases are spectacular and very costly in both lives lost and property damage. Even the best road design cannot fully compensate for the "out-of-control" problem in mountainous terrain, and vehicle performance standards can provide for heavy vehicle control on long and/or steep downgrades only when the use of gear shifting and braking are properly combined to reduce speeds. Static signing and stopping areas (turnouts or pull off areas) located before severe downgrades, to be used for voluntary or mandatory brake inspections and for cooling of brakes to restore their stopping capabilities, are the methods most commonly used to provide proper information to the drivers and provide an opportunity for checking the operation of the equipment prior to descent. In addition, information about the grade ahead and the location of escape ramps can be provided by diagrammetric signing. Refer to Publication 212, Official Traffic Control Devices. The Department has constructed and maintains emergency escape ramps throughout the Commonwealth. Reports and evaluations indicate that these escape ramps have reduced property damage and more importantly have saved lives. The design criteria presented in this chapter have been developed by the Department through research by The Pennsylvania State University. The goal of this research project was to understand the physical characteristics of the stopping mechanism and to provide a means for adequately designing and maintaining gravel arrester beds. Fullscale testing was performed at operational gravel arrester beds within the state as well as at two research gravel arrester beds located at The Pennsylvania Transportation Institute's (PTI) Research Facility. Additional information and details can be obtained from The Pennsylvania Transportation Institute, A Field and Laboratory Study to Establish Truck Escape Ramp Design Methodology, Report No. FHWA-PA , October 1988 and the AASHTO Green Book DYNAMICS OF A VEHICLE The effectiveness of gravel arrester beds in stopping runaway vehicles results from the interaction between vehicle motion and gravel movement. The motion can be predicted if the forces acting on the vehicle can be predicted because Newton's law gives the deceleration if the forces and masses of the vehicle are known. Resistance forces that act on every vehicle and affect the vehicle's speed include engine, braking and tractive resistance forces. Engine and braking resistance forces can be ignored in the design of emergency escape ramps because the ramp should be designed for the worst case; that is, the vehicle is out of gear and the brake system has failed. Tractive resistance forces contain four subclasses: inertial, aerodynamic, rolling and gradient. Inertial and negative gradient resistance forces act to maintain motion of the vehicle while rolling, positive gradient and air resistance forces act to retard its motion. The 2004 AASHTO Green Book, Chapter 3, Exhibit 3-65 illustrates the action of the various resistance forces on a vehicle in motion. 17-1

2 Inertial resistance force can be described as a force that resists movement in a vehicle at rest or keeps a vehicle in motion, unless the vehicle is acted upon by some external force. Inertial resistance force must be overcome to either increase or decrease the speed of a vehicle. Rolling and positive gradient resistance forces are available to overcome the inertial resistance force. Rolling resistance force is a general term used to describe the resistance to motion at the area of contact between a vehicle's tires and the roadway surface and is only applicable when a vehicle is in motion. It is influenced by the type and displacement characteristics of the surfacing material of the roadway. Gradient resistance force is due to the effect of gravity and is expressed as the force needed to move the vehicle through a vertical distance. For gradient resistance force to provide a beneficial force on an escape ramp, the vehicle must be moving upgrade against gravity. In the case where the vehicle is descending a grade, gradient resistance force is negative, thereby reducing the forces available to slow and stop the vehicle. It is influenced by the total mass (weight) of the vehicle and the magnitude of the grade. The remaining component of tractive resistance force is aerodynamic resistance force. Air causes a significant resistance at speeds above 80 km/h (50 mph) and is negligible under 30 km/h (20 mph). The effect of aerodynamic resistance has been neglected in determining the length of the arrester bed in this chapter NEED AND LOCATION Each grade has its own unique characteristics. Highway alignment, gradient, length and descent speed contribute to the potential for out-of-control vehicles. For existing highways, operational problems on a downgrade will often be reported by law enforcement officials, truck drivers or the general public. A field review of a specific grade may reveal damaged guide rail, gouged pavement surfaces or spilled oil indicating locations where drivers of heavy vehicles had difficulty negotiating a downgrade. For existing facilities, an escape ramp should be provided as soon as a need is established. Crash experience (or, for new facilities, use crash experience on similar facilities) and truck operations on the grade combined with engineering judgment are used frequently to determine the need for a truck escape ramp. Often the impact of potential runaway trucks on adjacent activities or population centers will provide sufficient reason to construct an escape ramp. Likewise, for Interstate highways on extended lengths of maximum or near maximum descending grades, emergency escape ramps should be added where an evaluation indicates they are required. Unnecessary escape ramps should be avoided. For example, a second escape ramp should not be needed just beyond the curve that created the need for the initial ramp. While there are no universal guidelines available for new and existing facilities, a variety of factors should be considered in selecting the specific site for an escape ramp. Each location presents a different array of design needs; factors that should be considered include topography, length and percent of grade, potential speed, economics, environmental impact and crash experience. Ramps should be located to intercept the greatest number of runaway vehicles, such as at the bottom of the grade and at intermediate points along the grade where an out-of-control vehicle could cause a catastrophic crash. A technique for new and existing facilities available for use in analyzing operations on a grade, in addition to crash analysis, is the Grade Severity Rating System. The system uses a predetermined brake temperature limit (260 C (500 F)) to establish a safe descent speed for the grade. It also can be used to determine expected brake temperatures at 0.8 km (0.5 mi) intervals along the downgrade. The location where brake temperatures exceed the limit indicates the point that brake failures can occur, leading to potential runaways. Escape ramps generally may be built at any practical location where the main road alignment is tangent. They should be built in advance of horizontal curves that cannot be negotiated safely by an out-of-control vehicle and in advance of populated areas. Escape ramps should exit to the right of the roadway. On divided multilane highways, where a left exit may appear to be the only practical location, difficulties may be expected by the refusal of vehicles in the left lane to yield to out-of-control vehicles attempting to change lanes. Although crashes involving runaway trucks usually can occur at various sites along a grade, locations having multiple crashes should be analyzed in detail. Analysis of crash data pertinent to a prospective site should include evaluation of the section of highway immediately uphill including the amount of curvature traversed and distance to and radius of the adjacent curve. 17-2

3 An integral part of the evaluation should be the determination of the maximum speed that an out-of-control vehicle could attain at the proposed site. This highest obtainable speed can then be used as the minimum design speed for the ramp. The 130 to 140 km/h (80 to 90 mph) entry speed, recommended for design, is intended to represent an extreme condition and therefore should not be used as the basis for selecting locations of escape ramps. Although the variables involved make it impractical to establish a maximum truck speed warrant for location of escape ramps, it is evident that anticipated speeds should be below the range used for design. The principal factor in determining the need for an emergency escape ramp should be the safety of the other traffic on the roadway, the driver of the outof-control vehicle and the residents along and at the bottom of the grade. An escape ramp, or ramps if the conditions indicate the need for more than one, should be located wherever grades are of a steepness and length that present a substantial risk of runaway trucks and topographic conditions will permit construction TYPES OF EMERGENCY ESCAPE RAMPS Emergency escape ramps have been classified in a variety of ways. Three broad categories used to classify ramps are gravity, sandpile and arrester bed. Within these broad categories, four basic emergency escape ramp designs predominate. These designs are the sandpile and three types of arrester beds, classified by grade of the arrester bed: descending grade, horizontal grade and ascending grade. Typical escape ramps are shown in the 2004 AASHTO Green Book, Chapter 3, Exhibits 3-67 and The gravity ramp has a paved or densely compacted aggregate surface, relying primarily on gravitational forces to slow and stop the runaway. Rolling resistance forces contribute little to assist in stopping the vehicle. Gravity ramps are usually long and steep and are constrained by topographic controls and costs. While a gravity ramp stops forward motion, the paved surface cannot prevent the vehicle from rolling back down the ramp grade and jackknifing without a positive capture mechanism. Therefore, the gravity ramp is the least desirable of the escape ramp types. Sandpiles, composed of loose, dry sand dumped at the ramp site, are usually no more than 120 m (400 ft) in length. The influence of gravity is dependent on the slope of the surface. The increase in rolling resistance is supplied by loose sand. Deceleration characteristics of sandpiles are usually severe and the sand can be affected by weather. Because of the deceleration characteristics, the sandpile is less desirable than the arrester bed. However, at locations where inadequate space exists for another type of ramp, the sandpile may be appropriate because of its compact dimensions. Descending-grade arrester-bed escape ramps are constructed parallel and adjacent to the through lanes of the highway. These ramps use loose aggregate in an arrester bed to increase rolling resistance to slow the vehicle. The gradient resistance acts in the direction of vehicle movement. As a result, the descending-grade ramps can be rather lengthy because the gravitational effect is not acting to help reduce the speed of the vehicle. The ramp should have a clear, obvious return path to the highway so drivers who doubt the effectiveness of the ramp will feel they will be able to return to the highway at a reduced speed. Where the topography can accommodate, a horizontal-grade arrester-bed escape ramp is another option. Constructed on an essentially flat gradient, the horizontal-grade ramp relies on the increased rolling resistance from the loose aggregate in an arrester bed to slow and stop the out-of-control vehicle, since the effect of gravity is minimal. This type of ramp is longer than the ascending-grade arrester bed. The most commonly used escape ramp is the ascending-grade arrester bed. Ramp installations of this type use gradient resistance to its advantage, supplementing the effects of the aggregate in the arrester bed, and in general, reducing the length of ramp needed to stop the vehicle. The loose material in the arresting bed increases the rolling resistance, as in the other types of ramps, while the gradient resistance acts downgrade, opposite to the vehicle movement. The loose bedding material also serves to hold the vehicle in place on the ramp grade after it has come to a safe stop. Each of the ramp types is applicable to a particular situation where an emergency escape ramp is desirable and should be compatible with established location and topographic controls at possible sites. The procedures used for analysis of truck escape ramps are essentially the same for each of the categories or types identified. The rolling resistance factor for the surfacing material used in determining the length needed to slow and stop the runaway safely is the difference in the procedures. 17-3

4 17.4 ELEMENTS OF DESIGN The principal factor in determining the need for an emergency escape ramp should be the safety of the other traffic on the roadway, the driver of the out-of-control vehicle and the residents along and at the bottom of the grade. To safely stop an out-of-control vehicle, the length of an escape ramp should be sufficient to dissipate the kinetic energy of the moving vehicle. A "last chance" device at the end of the ramp, such as a mound or a row of barrels, should be considered when the consequences of leaving the end of ramp are serious. There are numerous elements which affect the performance of emergency escape ramps. The depth, length and slope as well as the gradation, density and type of material are important factors in the performance of an arresting bed. Resistance forces limit the maximum speed of an out-of-control vehicle. Speeds in excess of 130 to 140 km/h (80 to 90 mph) will rarely, if ever, be attained. For the escape ramp to be effective, it must stop the largest vehicle expected to use the ramp; generally a truck, such as a WB-15 (WB-50) or a WB-18 (WB-60). Access to the ramp should be made obvious by exit signing, with sufficient sight distance in advance, to allow the driver of an out-of-control vehicle time to react, so as to minimize the possibility of missing the ramp. Advance signing is needed to inform drivers of the existence of an escape ramp and to prepare drivers well in advance of the decision point so that they will have enough time to decide whether or not to use the escape ramp. Regulatory signs near the entrance should be used to discourage other motorists from entering, stopping, or parking at or on the ramp. The path of the ramp should be delineated to define ramp edges and provide nighttime direction. Illumination of the approach and ramp is desirable. Design recommendations for emergency escape ramps are divided into the following subsections: (1) Basic Bed Length, (2) Barrels, (3) Mounds, (4) Length Design with Combination of Bed, Mounds and Barrels, (5) Bed Design and (6) Incidental Items. A. Basic Bed Length. The basic bed length (L) required without mounds or barrels is given by a third-order equation: where: L = Basic bed length (m (ft)). V = Entry speed (km/h (mph)). A,B,C,D = Constants given in Table L = A + BV + CV 2 + DV 3 (Eq. 17-1) The above equation is used to calculate the basic bed length for entry speeds up to 140 km/h (90 mph). However, the values of the constants used with the equation are different. In Table 17.1, a set of values is given for entry speeds from 50 to 100 km/h (30 to 60 mph) and another set of values for entry speeds from 101 to 140 km/h (61 to 90 mph). To calculate L, the entry speed and bed grade are chosen, and then the constants can be determined from Table METRIC EXAMPLE: To find the length required for an entry speed of 100 km/h at 0% grade and 10% grade, the constants are first found from Table % grade 10% grade A 0 = A 10 = B 0 = B 10 = C 0 = C 10 = D 0 = D 10 =

5 By substituting these values into Equation 17-1, lengths of 103 m for a bed with 0% grade and 79 m for a bed with 10% grade are found. As shown here, if conditions dictate a shorter bed length, the grade of the bed is required to be steeper. 1. Graphs for Designing the Bed Length. Figure 17.1 is a plot of Equation 17-1 using the values from Table 17.1 and thus can be used as an alternative to Equation 17-1 for designing bed length. In the example given above, the 100 km/h line would be followed to the 0% grade curve where the length of 103 m is indicated. For a 10% slope, the 100 km/h line would be followed to the 10% grade curve, where the length of 79 m, is shown. The basic bed length is also given in Table 17.2 which results from using Equation 17-1 and the constants from Table The basic bed length is given for entry speeds of 2 km/h increments from 50 to 140 km/h. This table is to be used as a quick reference during preliminary design and for review of final plans for emergency escape ramps. ENGLISH EXAMPLE: To find the length required for an entry speed of 60 mph at 0% grade and 10% grade, the constants are first found from Table % grade 10% grade A 0 = A 10 = B 0 = B 10 = C 0 = C 10 = D 0 = D 10 = By substituting these values into Equation 17-1, lengths of 310 ft for a bed with 0% grade and 240 ft for a bed with 10% grade are found. As shown here, if conditions dictate a shorter bed length, the grade of the bed is required to be steeper. 1. Graphs for Designing the Bed Length. Figure 17.1 is a plot of Equation 17-1 using the values from Table 17.1 and thus can be used as an alternative to Equation 17-1 for designing bed length. In the example given above, the 60 mph line would be followed to the 0% grade curve where the length of 310 ft is indicated. For a 10% slope, the 60 mph line would be followed to the 10% grade curve, where the length of 260 ft, is shown. The basic bed length is also given in Table 17.2 which results from using Equation 17-1 and the constants from Table The basic bed length is given for entry speeds of 2 mph increments from 30 to 90 mph. This table is to be used as a quick reference during preliminary design and for review of final plans for emergency escape ramps. 17-5

6 17-6

7 17-7

8 CONSTANT TABLE 17.1 (METRIC) VALUES FOR CONSTANTS USED TO CALCULATE BASIC BED LENGTH IN EQUATION TO 100 km/h PERCENT GRADE OF BED A B C D TO 140 km/h CONSTANT PERCENT GRADE OF BED A B C D

9 TABLE 17.1 (ENGLISH) VALUES FOR CONSTANTS USED TO CALCULATE BASIC BED LENGTH IN EQUATION TO 60 mph CONSTANT PERCENT GRADE OF BED A B C D TO 90 mph CONSTANT PERCENT GRADE OF BED A B C D

10 ENTRY SPEED (km/h) TABLE 17.2 (METRIC) BASIC BED LENGTH FOR ENTRY SPEED FOR SOME BED GRADES BASIC BED LENGTH (m) PERCENT GRADE OF BED

11 ENTRY SPEED (km/h) TABLE 17.2 (METRIC)(CONTINUED) BASIC BED LENGTH FOR ENTRY SPEED FOR SOME BED GRADES BASIC BED LENGTH (m) PERCENT GRADE OF BED

12 ENTRY SPEED (mph) TABLE 17.2 (ENGLISH) BASIC BED LENGTH FOR ENTRY SPEED FOR SOME BED GRADES BASIC BED LENGTH (ft) PERCENT GRADE OF BED

13 ENTRY SPEED (mph) TABLE 17.2 (ENGLISH)(CONTINUED) BASIC BED LENGTH FOR ENTRY SPEED FOR SOME BED GRADES BASIC BED LENGTH (ft) PERCENT GRADE OF BED

14 B. Barrels. Impact attenuator barrels can be used to help decelerate trucks where insufficient space is available for proper bed length. The following three equations give the change in speed, average deceleration load and time to travel the 0.9 m (3 ft) through a row of barrels: V f = DV e METRIC: g = (D 2-1)V e 2 (Eq. 17-2) t = / (D + l)v e ENGLISH: g = (D 2-1)V e 2 t = / (D + l)v e where: V f = Exit speed after barrel row (km/h (mph)). V e = Entry speed into barrel row (km/h (mph)). D = Factor given in Table g = Deceleration, g. t = Time to travel length of barrels (s). METRIC EXAMPLE: Three or more barrels across by four rows deep, would permit an additional reduction of 50 km/h when combined with the last 4.4 m of the bed for a kg vehicle and would cause a 4.7-g load. Otherwise, 21 m of bed without the barrel barrier would be required to achieve the same results at 0% grade. Because of the larger deceleration loads at speeds above 70 km/h, a design as given in Figure 17.2 is suggested to reduce the maximum to below 9-g. This design is good for vehicles traveling as fast as 100 km/h. If speeds are higher, the designer should follow the procedures given in the next section, on barrel curves, to design the appropriate configuration. ENGLISH EXAMPLE: Three or more barrels across by four rows deep, would permit an additional reduction of 30 mph when combined with the last 14 ft of the bed for a 80,000 lb vehicle and would cause a 4.7-g load. Otherwise, 65 ft of bed without the barrel barrier would be required to achieve the same results at 0% grade. Because of the larger deceleration loads at speeds above 45 mph, a design as given in Figure 17.2 is suggested to reduce the maximum to below 9-g. This design is good for vehicles traveling as fast as 60 mph. If speeds are higher, the designer should follow the procedures given in the next section, on barrel curves, to design the appropriate configuration. 1. Graphs for Designing the Barrel Array. The equations given in the previous section are presented graphically in Figure This figure presents, in the same manner as that used in the Energite System manual for the inertial barrier system, the design for a kg (80,000 lb) vehicle using 0.64 m 3 (22.6 ft 3 ) barrels filled with river gravel (AASHTO No. 57). Sand should not be used in the barrels since sand would contaminate the gravel bed. The maximum deceleration forces acceptable for a vehicle's occupants are 12-g. The figure gives deceleration force levels up to 12-g. However, a deceleration force of 9-g is desirable for design. To use Figure 17.3, the same procedure is followed as was used to determine the bed length. A vehicle impacting a three by three barrel array at 50 km/h (30 mph) can be considered as an example. Using Figure 17.4, the initial impact speed (50 km/h (30 mph)) can be located on the baseline (point 1). A straight vertical line can be drawn from that point until it intersects the horizontal line for three-barrel rows. The deceleration (4.7-g) is read here. From that point, another line can be drawn parallel to the nearest exit speed line down to the baseline (point 2), which gives an exit speed of 38 km/h (23 mph) from the first row. The procedure is repeated for the second row, giving a deceleration of 2.6-g and an exit speed of 28 km/h (17 mph) at point 3. Repeating this procedure for the third row gives 1.5-g and 21 km/h (12 mph) at exit

15 TABLE 17.3 D FACTORS FOR BARRELS VEHICLE MASS ROW (Vehicle Weight) 820 kg (1800 lb) 2040 kg (4500 lb) 6580 kg (14,500 lb) kg (33,000 lb) kg (41,000 lb) kg (80,000 lb) ONE BARREL TWO BARRELS THREE BARRELS

16 For this case, the speed is decreased from 50 km/h to 21 km/h (30 to 12 mph) over 3.0 m (10 ft). Also, from Figure 17.1, 3.0 m (10 ft) of a 0% grade bed will remove 21 km/h (12 mph) from the vehicle speed; i.e., the combination of three rows of barrels and 3.0 m (10 ft) of bed will remove a total of 50 km/h (30 mph). In another example (not marked on the figure) of a single row of two barrels with an impact speed of 60 km/h (40 mph), a 5.3-g deceleration and an exit speed of 49 km/h (33 mph) will be found. C. Mounds. Mounds, which are depicted by Figure 17.5, are treated in a manner similar to that used for the barrels, and the same speed change equation is used; however, a different deceleration equation is needed because, while the average deceleration is calculated in the same manner, the peak is different. Using Equation 17-2, the equations for mounds were developed and are given below: METRIC: 1. Full Mound: g ave = (D 2-1)V e 2 g peak = (D 2-1)V e 2 V f = DV e (Eq. 17-3) t = / (D + 1)V e 2. Half Mound: g ave = (D 2-1)V e 2 g peak = (D 2-1)V e 2 t = / (D + 1)V e ENGLISH: 1. Full Mound: g ave = (D 2 2-1)V e g peak = (D 2 2-1)V e t = / (D + 1)V e 2. Half Mound: g ave = (D 2-1)V e 2 g peak = (D 2-1)V e 2 t = / (D + 1) V e where: V f = Exit speed from mound (km/h (mph)) V e = Entry speed into mound (km/h (mph)) D = Factor given in Table g = Deceleration, g. t = Time to travel mound(s) (s)

17 17-17

18 17-18

19 17-19

20 17-20

21 17-21

22 MOUND TYPE 820 kg (1800 lb) TABLE 17.4 D FACTORS FOR MOUNDS VEHICLE MASS (VEHICLE WEIGHT) 2040 kg 6580 kg kg (4500 lb) (14,500 lb) (33,000 lb) kg (41,000 lb) kg (80,000 lb) FULL HALF Tests have shown that mounds should not be placed nearer than 30 m (100 ft) into the bed. When a truck is still riding on top of the bed, it will ride up over the mound, giving high vertical acceleration. In the case of the full-sized mound, the truck can become airborne; however, once the truck sinks into the bed, it generally plows through the mounds. Mound usage should be avoided, if possible. If they are used, however, they should be placed in the bed such that they will be hit at slow speeds, 40 km/h (25 mph) and less. Although barrels are more expensive than mounds, barrels are recommended. 3. Graphs for Designing Mounds. In the same manner used for the barrel designs, given previously, Figure 17.6 is used in the design of mounds. Figure 17.5 shows the cross sections for a full and a half mound. An example is shown in Figure 17.7 for a kg (80,000 lb) vehicle traveling 70 km/h (45 mph) into a full mound. It has a peak deceleration of 2.8-g and exits at 61 km/h (38 mph). Although a deceleration force of 2.8-g is acceptable for the driver, the larger forces are on only the first (front) axle, and for a kg (80,000 lb) vehicle, this deceleration gives a force of approximately 1020 kn (240,000 lb). D. Design with Combination Bed, Mounds and Barrels. A bed design in combination with mounds or barrels or both requires the use of Figures 17.3, 17.6 and Figure 17.8 is a replot of Figure 17.1 with the x and y axes interchanged. This figure can be used to illustrate the design; an example is shown for the 0% grade. METRIC EXAMPLE: Suppose a bed is proposed to stop a kg vehicle traveling at 100 km/h on a 0% grade. Consider a mound placed at 30 m and three rows of barrels at the end. For this example, Figures 17.9, and are used. First, from Figure 17.10, if the speed is reduced to 50 km/h when the barrels are reached, the exit speed due to the barrels alone is 21 km/h. Since the barrels are 2.7 m long, the 0% curve of Figure shows that the remaining 21 km/h will be eliminated by the bed while traveling the 2.7 m. Thus, enough length of the bed plus the mound will be needed to get the speed down to 50 km/h. Using Figure 17.11, a horizontal line should be drawn from the 100 km/h point until the 0% grade line has been intersected. The 0% grade line is then followed for 30 m from 103 m to 73 m, at which point the full mound has been reached. A horizontal line is then drawn to find that the speed is 86 km/h. Using Figure 17.9, an entrance speed of 86 km/h gives an exit speed of 75 km/h. When Figure is re-entered at 75 km/h, the 0% curve is followed to 50 km/h, which is the speed reduction previously found for the three rows of barrels. Thus, the first 30 m of the bed (see Figure 17.11) plus 32 m (from 53 m down to 21 m) plus the 2 m at the end adds up to require a bed of 64 m. ENGLISH EXAMPLE: Suppose a bed is proposed to stop a 80,000 lb vehicle traveling at 60 mph on a 0% grade. Consider a mound placed at 100 ft and three rows of barrels at the end. For this example, Figures 17.9, 17.10, and are used. First, from Figure 17.10, if the speed is reduced to 30 mph when the barrels are reached, the exit speed due to the barrels alone is 12 mph. Since the barrels are 9 ft long, the 0% curve of Figure shows that the remaining 12 mph will be eliminated by the bed while traveling the 9 ft. Thus, enough length of the bed plus the mound will be needed to get the speed down to 30 mph

23 17-23

24 17-24

25 17-25

26 17-26

27 Using Figure 17.11, a horizontal line should be drawn from the 60 mph point until the 0% grade line has been intersected. The 0% grade line is then followed for 100 ft from 311 ft to 211 ft, at which point the full mound has been reached. A horizontal line is then drawn to find that the speed is 51 mph. Using Figure 17.9, an entrance speed of 51 mph gives an exit speed of 45 mph. When Figure is re-entered at 45 mph, the 0% curve is followed to 30 mph, which is the speed reduction previously found for the three rows of barrels. Thus, the first 100 ft of the bed (see Figure 17.11) plus 105 ft (from 158 ft down to 65 ft) plus the 5 ft at the end adds up to require a bed of 198 ft. When space permits, the bed length should be designed as required for the entry speed without regard to reduction in distance due to barrels or mounds. The barrels in this case would be provided as an added safety benefit at the end of the ramp. E. Bed Design. The method for designing the length of an arrester bed was addressed in the previous section, along with the need to consider the grade of the bed and the use of barrels and of gravel mounds in the overall bed design. Other characteristics of the bed which must be considered for an effective design are the depth and width of the bed and drainage features. Also essential to the overall arrester bed operation is the installation of a concrete anchor block 15 to 30 m (50 to 100 ft) in front of the bed to act as a dead man for the tow vehicle used for truck extraction. 1. Bed Depth. A minimum of 1070 mm (42 in) is the recommended depth for beds of river gravel. Testing has shown that a bed with 915 mm (36 in) of river gravel gives the same results as a bed as deep as 2440 mm (8 ft). The minimum recommended depth includes 150 mm (6 in) to allow for compaction when the gravel contains many fines, especially if the bed is located where the potential for heavy use is great. Frequent use results in the significant increase in fines content, which decreases the effectiveness of the bed. Smooth, rounded, uncrushed gravel of approximately a single size is the most effective arrester bed material. The best size is approximately 13 mm (0.5 in) in diameter. The river gravel graded to AASHTO No. 57 was found to be the best of those materials tested if it had been washed so that fines were removed. A greater percentage of larger or smaller stones decreased the effectiveness and increased planning. Crushed stone should not be used because it provides a drag factor of about half that of rounded gravel and compacts more quickly. To be effective, crushed aggregate would require longer beds and fluffing almost weekly. Rounded river gravel produces higher decelerations than the more angular crushed aggregate because the truck sinks into the river gravel more, transferring more energy to the stones over a shorter distance. 2. Bed Width. In general, a width of 6.6 to 7.5 m (22 to 25 ft) is recommended on the basis of an entry design which has a horizontal straightaway of at least one truck length, directing the vehicle in a straight line through the center of the bed. If the top of the bed is at ground level, a minimum width of 6.6 m (22 ft) would be sufficient. Conversely, if either side of the bed is a drop-off of any height, the bed should be wider. The vehicle's direction of momentum as it enters the bed is the direction it will travel through the bed; this direction should be the bed center line. 3. Bed Drainage. Proper drainage so that water does not stand in the bed is important. A 305 mm (12 in) base layer of large (at least 75 mm (3 in) in diameter) crushed limestone aggregate (AASHTO No. 1) will effectively drain the arrester bed. The stones should be confined to the layer by covering them with geotextile material (Type 3) to separate the larger stones from the river gravel. The cross slope of the base should be toward one side, with either subgrade drains or a crown for removing any water from the arrester layer. The river gravel covering this sloping base layer should have no cross slope at the top surface; i.e., the bed should be filled such that the top surface has no cross slope. F. Incidental Items. 1. Service Road. A service road located adjacent to the arrester bed is recommended so the tow and maintenance vehicles can use it without becoming trapped in the bedding material. The typical width of this road is 3.0 m (10 ft)

28 17-28

29 17-29

30 17-30

31 17-31

32 17-32

33 17-33

34 17-34

35 17-35

36 2. Concrete Anchor Block. Every escape ramp must include a concrete anchor block to which a tow vehicle can attach when it pulls an arrested vehicle from the gravel bed. The anchor provides a necessary dead man which can withstand the retraction loads required to pull an arrested vehicle from the ramp bed. Retraction loads range from % of the gross vehicle mass (weight) and are generally about 50%. Thus, for example, a load of more than kg (40,000 lb) may be required to extract a loaded tractor trailer. During some tests, towing service personnel who did not believe the anchor would be necessary found that the tow truck, rather than the captured vehicle, was moved during extraction attempts. In one such case, the paved approach was damaged. Thus, it is recommended that every escape ramp include an anchor block such as that shown in Figure 17.12, and should be located at the center and flush with the road surface 15 to 30 m (50 to 100 ft) from the start of the bed. Additional blocks may be needed, depending on the length of the escape ramp. Towing services should be required to use the anchor block. 3. Gravel Fluffer. To preserve its original density, the gravel bed should be fluffed by a device capable of breaking up the compacted areas. A typical mechanical device is shown on Figure This device, designed by PTI, consists of a sled with prongs that extend down into the gravel bed. As the sled is pulled through the bed, its prongs break up the compacted areas. One important element of a fluffer is adequate mass (weight) to break up the compacted area at the required depth. 4. Extraction. Extraction should be performed with a wrecker and a winch. The front of the wrecker must be chained to a dead man anchor block, and the winch must be used with a block and tackle that has at least a two to one mechanical advantage. After the vehicle being extracted from the bed begins to move, it can be raised onto 50 mm 150 mm (2 in 6 in) boards to greatly reduce the drawbar pull required to remove the vehicle from the bed and, correspondingly, the pull loads on the vehicle and its axles. At least two boards per wheel set should be used so that, as a wheel rolls off one board, it will roll onto the next board. As the wheel then rolls clear of the first board, that board can be placed in front of the second board, which can then be moved in front of the first board, and so on, in a leapfrog fashion. When the captured vehicle has wheel flaps, an important note of caution must be heeded. Flaps must be removed or tied such that they will not wrap around the wheel between the stones and the tire. Only a turning wheel will ride up onto the aggregate rather than dig into them, and a barrier between the stones and tire will prevent the wheel from turning. Consequently, when the wheel flaps provide such a barrier, the wheel generally digs deeper into the stones, thus creating a need for greater force to pull the vehicle. G. Maintenance. All of the beds were found to compact with time depending on the durability of the stone. During the design phase, consideration should be given to access by maintenance vehicles to fluff and maintain escape ramps in good operating condition. For maintenance details and requirements, refer to the Maintenance Manual

37 2 Lids x 914 x 19 Steel Plates Opening 51 Diameter SECTION B-B B A 76 B A 38 PVC Pipe Outlet Thru Embankment No. 43 Bars at Bend on Bars 305 SECTION A-A 1829 Note: All Dimensions are in Millimeters (mm) Except as Noted. FIGURE (METRIC) Design of a Concrete Anchor Block Dead Man 17-37

38 2 Lids - 2' x 3' x 0.75" Steel Plates Opening 2" Diameter SECTION B-B 4' 11" 10.5" 1'-10" 10" 4" 8" 2" 2" 0.75" 3.25" 1.50" B A 3" B 8" 2' 8" 2' 8" 2" 2" 2" 2" 11" 11" A 1.5" PVC Pipe Outlet Thru Embankment 10" 4' 2 No. 12 Bars at 10' 1' Bend on Bars 1' SECTION A-A 6' FIGURE (ENGLISH) Design of a Concrete Anchor Block Dead Man 17-38

39 17-39

40 BLANK PAGE 17-40