Multiple Vehicle Collision on Interstate 95

8

Multiple Vehicle Collision on Interstate 95

Fairfield, Connecticut January 17, 2003

〔图〕插图

Highway Accident Report

NTSB/HAR-05/03

PB2005-916203

Notation 7734

iii

Highway Accident Report

Contents

Acronyms and Abbreviations $\ldots$ v

Executive Summary $\ldots$ vi

Factual Information $\ldots$ 1

Accident Narrative $\ldots$ 1

Preaccident Events $\ldots$ 2

Injuries $\ldots$ 4

Medical and Pathological Information $\ldots$ 6

Survival Aspects $\ldots$ 6

Restraint Use and Survivable Space $\ldots$ 6

Emergency Response $\ldots$ 8

Driver Information $\ldots$ 10

1996 Freightliner Driver $\ldots$ 10

Chevrolet Tahoe Driver $\ldots$ 11

Vehicle and Wreckage $\ldots$ 14

1996 Freightliner Tractor and Flatbed Semitrailer $\ldots$ 14

Chevrolet Tahoe $\ldots$ 16

Meteorological Information $\ldots$ 17

Highway Information $\ldots$ 18

Forensic Evidence $\ldots$ 19

Accident History $\ldots$ 20

Roadway Condition $\ldots$ 20

Work Zone Construction $\ldots$ 24

Lighting $\ldots$ 27

Variable Message Signs $\ldots$ 28

Incident Management $\ldots$ 29

Motor Carrier Information $\ldots$ 32

Tests and Research $\ldots$ 33

Conspicuity Testing $\ldots$ 33

Speed Testing $\ldots$ 34

Deceleration Testing $\ldots$ 34

Flare Testing $\ldots$ 35

Previous Safety Board Actions $\ldots$ 35

Barrier and Heavy Truck Compatibility $\ldots$ 35

Seat Belt Usage $\ldots$ 36

Analysis $\ldots$ 38

Exclusions $\ldots$ 38

Toxicology $\ldots$ 38

Vehicle Factors $\ldots$ 38

Motor Carrier Factors $\ldots$ 38

Emergency Response $\ldots$ 39

Accident Discussion $\ldots$ 39

Driver Issues $\ldots$ 40

References

1

Highway Accident Report

Factual Information

Accident Narrative

Two consecutive accidents occurred within about 11 minutes in the early morning hours of January 17, 2003, on Interstate 95 (I-95) near Fairfield, Connecticut. (See figure 1.) Witnesses reported that at the time of the accidents, light snow was falling, the roads were wet and icy, and snow covered the roadway shoulders.

〔图〕插图

Figure 1. Map of accident location.

About 4:50 a.m.,$^{1}$ a 1996 Freightliner tractor flatbed semitrailer, loaded with five portable compressor units, was involved in a nonfatal multivehicle accident. The truck was traveling in a work zone on I-95 north, near milepost (MP) 26.6, at a driver-estimated speed of 50 mph, when it slid out of control approximately 1,150 feet south of the exit 24 southbound off-ramp. The vehicle entered the median, overturned and overrode the portable concrete median barrier (see figure 2a), and collided with a southbound 1997 Dodge Avenger sedan. The sedan was subsequently struck from behind by a southbound 2001 Freightliner tractor/refrigerated trailer combination unit, which then struck the 1996 Freightliner tractor. The three vehicles came to rest blocking the southbound lanes of the highway. The drivers of the two combination units and the Dodge sedan received minor or

Factual Information

3

Highway Accident Report

〔图〕Southbound= postcrash direction of travel一一一一一一一一一Tractor and semitrailer一一一一口口一一一一口一一一一一Northbound一一一一一一一一一一一一一一一\n

2a. 1996 Freightliner tractor and semitrailer impact and overrun median barrier.

〔图〕一TractorSemitrailer一一一一一一一一一一一一一一一二一一一一一一一一\n

2b. Tractor and semitrailer separate, causing semitrailer to encroach into northbound traffic lanes.

〔图〕一TractorSemitrailer一一一一一一一0B一一一Chevrolet\n

2c. Chevrolet collides with semitrailer followed by secondary impacts with median barrier.

Figure 2. Representation of multivehicle accident sequence.

Factual Information

4

Highway Accident Report

〔图〕2018.12.15\n

Figure 3. Chevrolet Tahoe at final rest in northbound roadway.

Injuries

Table 1 summarizes the injuries resulting from the two consecutive multivehicle accidents. It is based on the International Civil Aviation Organization injury criteria, which the Safety Board uses in accident reports for all transportation modes.

Table 1. Injuries.$^{2}$

Factual Information

S

Highway Accident Report

〔图〕Dodge area of impactwith barrier walNORTH1996 FreightlinerDodge, and2001 FreightlineSouthbound 1-95area of impactSemitra

Figure 4. Highway forensic evidence from 4:50 a.m. and 5:01 a.m. accidents.

Factual Information

7

Highway Accident Report

〔图〕Chevrolet frontMale 2DriverFront seatsseriousmale 21injuriesfataRear seatsMale 20Male 19Male 21Male 19fatalfatafataserio

Figure 5. Chevrolet Tahoe seating chart.

Factual Information

9

Highway Accident Report

Table 2. Emergency response timeline for 4:50 a.m. and 5:01 a.m. accidents.

The Bridgeport Operations Center fielded the emergency calls and dispatched first responders to the 4:50 a.m. accident scene on I-95 south. The 5:01 a.m. accident involving the Chevrolet occurred while first responders were enroute to the southbound accident. According to the CSP, both accident scenes were treated as one large event. The responding agencies were able to communicate with each other and allocate resources. The CSP added that response time was negatively affected by three factors:

· Two of the closest patrol officers were detained at the Bridgeport Hospital with the driver from the 3:56 a.m. accident, who happened to have an outstanding arrest warrant.

· Road and weather conditions were poor.

● The traffic queue that formed after the southbound accident caused additional highway congestion.

Factual Information

10

Highway Accident Report

Driver Information

1996 Freightliner Driver

The 33-year-old 1996 Freightliner driver was an owner/operator who was driving his personal truck tractor and a flatbed semitrailer owned by Arrow Trucking Company of Tulsa, Oklahoma. He held a valid California Class A commercial driver's license (CDL) with hazardous materials and doubles/triples endorsements, issued July 6, 1999, and expiring May 13, 2004. The driver also held a valid medical certificate, issued May 30, 2001, and expiring May 31, 2003. He had attended a professional truck driving school from May 24 through June 24, 1999, and then worked as an employee driver for Arrow Trucking for 1 year. He subsequently purchased the 1996 Freightliner and worked for Arrow Trucking as an owner/operator. Table 3 lists his traffic violations prior to the Fairfield accidents. The 1996 Freightliner driver indicated to Safety Board investigators that he had prior experience driving commercial vehicles in winter weather conditions and snow.

Table 3. Prior moving violations for 1996 Freightliner driver.

The 1996 Freightliner driver's 72-hour history was reconstructed using a combination of his daily logs and other records that were found in the tractor cab following the accident. (See table 4.)

Factual Information

11

Highway Accident Report

Table 4. Timeline of 1996 Freightliner driver's activities leading up to accident.

Chevrolet Tahoe Driver

At the time of the accident, the 20-year-old Chevrolet driver held a valid California noncommercial Class C driver's license, which was issued on June 19, 1998, and set to expire on April 13, 2006. The driver's California record indicated one violation on July 3, 2002, for "unsafe speed for road conditions." The driver did not own the Chevrolet, which belonged to another fraternity member.

Factual Information

12

Highway Accident Report

Safety Board investigators reconstructed the activities of the Chevrolet driver during rush week and prior to the accident using information gathered during interviews with the driver's dormitory roommate; the chapter president of Delta Kappa Epsilon; and the driver's friend, also a Yale student. (See table 5.) The driver's sleep/wake pattern was also derived from this information. (See figure 6.)

Table 5. Timeline of Chevrolet driver's activities leading up to accident.

Factual Information

13

Highway Accident Report

〔图〕插图

Figure 6. Chevrolet driver sleep pattern prior to accident.

According to the president of the fraternity chapter to which the Chevrolet driver belonged, the Fairfield accidents occurred during “rush week,” a period when pledges are considered for membership. The chapter president stated that the Chevrolet driver volunteered to serve as one of three designated nondrinking drivers for the Thursday night rush activity in New York City. The fraternity members arrived at an Irish pub in New York City between 11:30 p.m. and midnight. They were later joined by several pledges who had taken the train to the city. The chapter president stated that he had observed the Chevrolet driver drink one beer while at the pub.

The fraternity members and pledges departed New York City for New Haven about 3:45 a.m. The pledges who had arrived by train returned by car because the first train for New Haven did not leave until 5:00 a.m. The fraternity chapter president, who was also a designated nondrinking driver and who was the last to leave for New Haven, stated that he experienced slippery road conditions along I-95. He briefly skidded while traveling at 65 mph, then reduced his speed to 50 mph and experienced no further problems. He said that he also observed a tractor-trailer skidding on the roadway. He learned that the Chevrolet had been involved in a collision when he approached the accident site.

The CSP and the Safety Board interviewed four of the surviving five Chevrolet passengers. Two of the passengers, who had been seated in the cargo area, stated that they were asleep when the accident occurred. Another passenger, also seated in the cargo area, had sustained serious head injuries and had no recollection of events prior to the accident. The right rear-seat passenger stated that he had slept most of the way back from New York, except when a friend shared a sandwich with him after a fast food stop along I-95.

Factual Information

14

Highway Accident Report

Vehicle and Wreckage

1996 Freightliner Tractor and Flatbed Semitrailer

The 1996 Freightliner tractor, model D120064ST, had a wheel base of 235 inches and an unladened weight of 17,000 pounds. The odometer indicated that it had been driven 655,420.3 miles. The tractor was powered by a 12.7-liter, 360-horsepower Detroit Diesel Corporation 60-Series engine, with a 10-speed Rockwell Automation, Inc., manual transmission. The 2000 Fontaine Trailer Company flatbed semitrailer was 48 feet long and had an unladened weight of 14,000 pounds. It carried five portable generators, each weighing 2,265 pounds (total weight, 11,325 pounds), for a total trailer weight of 25,325 pounds.

The tractor and semitrailer combination unit was equipped with an air braking system that consisted of 15-inch drum brakes on the steer axle and 16.5-inch drum brakes on the second, third, fourth, and fifth axles. The tractor did not have an antilock braking system (ABS); the semitrailer was equipped with ABS. The tractor had an engine retarder,$^{4}$ also called a “Jake” brake, which could be set to operate on two, four, or six cylinders. The retarder switch was in the “off” position when the Safety Board inspected the tractor. Manufacturers generally advise drivers not to use a retarder under wet or slippery roadway conditions. In fact, a Safety Board investigation of a motorcoach accident that occurred in Canon City, Colorado, in December 1999, revealed that an enabled retarder most likely triggered the loss of control and eventual crash of the motorcoach on a snow-covered and mountainous roadway.$^{5}$

The truck tractor had scrape marks along the entire passenger side as a result of overturning during the 4:50 a.m. accident sequence. The steer axle and engine compartment received extensive contact damage. The engine was crushed downward. In addition, the fifth wheel $^{6}$ plate was torn off the tractor's frame. On the semitrailer, the area where the fifth wheel was connected was damaged, the landing gear was damaged,$^{7}$ the rear impact bar was separated,$^{8}$ and the surface of both the underside and flatbed area was damaged. The bars that had attached the rear impact bar to the semitrailer were bent rightward.

The tractor was equipped with a Detroit Diesel Electronic Control III Module (ECM).$^{9}$ The ECM's optional memory feature, which records operating information about the engine and the vehicle, was not activated. There was extensive damage to the brakes on the fourth axle and to the right chamber of the second axle. The brakes that could be

Factual Information

15

Highway Accident Report

inspected and measured were found to be properly adjusted. No preexisting defects were found in the lines, valves, couplings, or brake chambers. The tires on the tractor and semitrailer were found to be in good condition, with little sign of uneven wear; tread depths were within regulations.$^{10}$

The fifth wheel slide $^{11}$ was in the “off” position, which meant that the flatbed semitrailer had been locked to the tractor prior to the accident. Cruise control was in the “off” position. The windshield wiper selection knob was in the “low speed” setting. The factory fog light switch was in the “off” position, and a second after-market add-on fog light switch was in the “on” position.

Retroreflective tape on the semitrailer was 3 inches wide and ran down both sides in 18-inch segments spaced 17 inches apart. (See figure 7.) It also ran fully along the rear of the semitrailer and rear impact bar. Each segment of the tape consisted of 7 inches of white material and 11 inches of red material. The left side of the semitrailer—the area impacted by the Chevrolet—also included two circular light covers. The smaller of the two covers was amber, about 2 inches in diameter, and cracked and damaged. The larger light cover was red, about 3 inches in diameter, and undamaged.

〔图〕、--\n

Figure 7. Retroreflective tape on rear of flatbed semitrailer.

Factual Information

16

Highway Accident Report

Chevrolet Tahoe

The rear impact bar of the semitrailer was embedded in the Chevrolet's radiator and engine compartment, approximately at the center of the vehicle. The engine hood was detached from the vehicle and deformed. The top portion of the engine compartment was severely damaged, as were various components mounted above the engine—the air conditioning unit, brake booster, ECM, throttle, air filter, and air intake. The windshield was broken and detached from the vehicle. The engine compartment firewall was pushed rearward and down toward the passenger compartment. The left side A-pillar was deformed and pushed rearward. The front portion of the roof was peeled back. The driver's side door was detached. The left rear door was pushed rearward into the left rear quarter panel and was in contact with the left rear tire. (See figure 8.)

〔图〕Green painttransferContacdamageFront axleEngine blockRoof supportGreen paintbracetransferLeft rearpassengerdoorRear axle

Figure 8. Diagram of Chevrolet Tahoe wreckage.

Safety Board investigators examined the Chevrolet's wheels, power train, steering mechanism, brakes, and sensors. No preexisting defects were found to be associated with these systems. Light bulbs extracted from the Chevrolet were taken to the Safety Board's

Factual Information

17

Highway Accident Report

Materials Laboratory for testing. Inspection of the headlight bulbs revealed that the high beams were on at the time of the accident, as evidenced by hot stretching on the bulb filaments.12 The Chevrolet was equipped with an after-market brush guard over the front exterior lighting area. (See figure 9.) The upper and lower metal bars on the guard were approximately 1 1/2 inches in diameter, separated by five smaller metal bars of approximately 3/8 inch diameter. One of the smaller diameter bars and a small portion of the thick upper bar were directly in front of the headlight lenses.

〔图〕插图

Figure 9. Chevrolet Tahoe with brush guard.

The Chevrolet was equipped with a sensing diagnostic module (SDM) located under the driver's seat. An SDM interprets a crash pulse (the time over which a crash occurs) and determines whether to deploy the airbags. The module can store one nondeployment event,$^{13}$ which is overwritten by subsequent nondeployment events. An SDM can also store one deployment event, as well as related information that occurs within 5 seconds of that event. In this case, one nondeployment event was stored because the airbags did not deploy. SDM data obtained by the CSP indicated that the Chevrolet reached a maximum change in velocity of 29 mph (5 g's) about 272.5 milliseconds after the algorithm was enabled, at which time power was lost and no additional data were recorded. The SDM can record up to 300 milliseconds of data once the algorithm is enabled.

Meteorological Information

The National Weather Service (NWS) reporting facility at Igor Sikorsky Memorial Airport, Bridgeport, Connecticut, is located approximately 7 miles east of the accident site. Between 11:17 p.m. on January 16 and 3:16 a.m. on January 17, 2003, the airport

Factual Information

19

Highway Accident Report

〔图〕插图

Figure 10. Overturned portable concrete barrier segments viewed from northbound lanes of I-95.

Forensic Evidence

Physical evidence of the initial accident involving the 1996 Freightliner consisted of furrow marks on the northbound side of the median, contact damage to the concrete median barrier, and pavement scrape marks in the southbound lanes. (See figure 4.) The furrow marks began in the median adjacent to the left lane, veered off toward the concrete median barrier at approximately a 23° angle, and then continued into the southbound traffic lanes. The pavement scrape marks traversed the three southbound lanes and extended a distance of approximately 63 feet. Additional contact damage to the top of the concrete median barrier consisted of green paint transfer that began about 80 feet beyond the furrow marks and continued for a distance of 101 feet.

The Chevrolet left pavement scrape marks on the left lane at the point of impact with the 1996 Freightliner flatbed semitrailer; additional forensic evidence included intermittent contact damage to the median barrier beyond the impact with the trailer. (See figure 4.) The pavement scrape marks were 10 feet long and were located approximately 150 feet north of the furrow marks. The contact damage to the median barrier—consisting of tire marks that extended about 500 feet—was located 215 feet north of the furrow marks.

Factual Information

20

Highway Accident Report

Accident History

In the 28 miles of I-95 from the New York State line to the eastern boundary of Fairfield, Connecticut, which includes the accident location, ConnDOT accident records indicated 40 traffic fatalities from 1998 to 2002—a 5-year average of 0.63 fatalities per 100 million vehicle miles traveled (VMT). The rest of the Connecticut urban interstate system averaged 0.596 fatalities per 100 million VMT; the national urban interstate system averaged 0.640 fatalities per 100 million VMT.

From 1998 to 2002, a total of 5,743 fatal and nonfatal accidents were recorded within the 5 miles of I-95 north that encompass the accident site. Table 6 categorizes these accidents.

Table 6. Accidents occurring within 5 miles of I-95 north accident site, 1998–2002.$^{15}$

Roadway Condition

The 1996 Freightliner driver stated to Safety Board investigators that the accident occurred because he had lost control of his vehicle immediately after crossing an $overpass^{16}$ and encountering black ice.$^{17}$ Several witnesses interviewed by the Safety Board reported experiencing sideslip or skidding while driving on I-95 north in the vicinity of the accident site. Witnesses also stated that there was no evidence of plowing or road treatment in the northbound or southbound lanes along the area of the Fairfield accidents.

According to the CSP, 18 other traffic accidents occurred on I-95 between midnight and 4:50 a.m. on January 17 “within several miles” of the Fairfield collisions, including one at 3:56 a.m. According to the police report for the 3:56 a.m. accident, a Ford

Factual Information

28

Highway Accident Report

Following the accidents, ConnDOT added a special provision to its standard specifications that requires a contractor to immediately notify the agency of a lighting outage or pole knockdown. The contractor is responsible for repairing the lighting system before the normal nighttime turn-on of the lights. If this cannot be achieved, the lighting must be operational prior to the next normal nighttime turn-on, within a maximum of 24 hours.

Variable Message Signs

ConnDOT had placed 17 variable message signs (VMSs) along the I-95 corridor. (See appendix D for sign placement.) Eleven signs were located on I-95 north, and six were located on I-95 south. Northbound, the nearest VMSs were located approximately 3.1 miles (VMS #55) and 4.67 miles (VMS #54) in advance of the 1996 Freightliner and Chevrolet accident sites, respectively. At the time of the two accidents, these two signs were not in use and did not display any messages.

The 1996 Freightliner accident occurred about 4:50 a.m.—at which time the Bridgeport Operations Center was notified that the left and center southbound lanes of I-95 were blocked. The center was not informed that the semitrailer also obstructed the left northbound lane. The operations center activated 23 VMSs on I-95 south; I-91; Routes 1, 8, 15, and 80; and State Road 796. Table 7 shows the messages that were displayed at approximately 5:01 a.m. on VMS #75, #76, and #77, the southbound signs closest to the accident site.

Table 7. Variable message signs displayed on I-95 south at 5:01 a.m., January 17, 2003. The signs alternated between frames 1 and 2.

The Bridgeport Operations Center received a 911 call at 5:01 a.m. regarding a northbound accident blocking the left and center lanes of I-95, at exit 24. It took steps to activate VMS #55 and #54 at 5:06 a.m. By 5:21 a.m., the two signs displayed messages warning of the accident. (See table 8.) According to ConnDOT, the delay in display of the northbound sign messages was a result of the VMS software system still completing the earlier command to activate the 23 southbound signs. The VMS system took approximately 31 minutes—from 4:50 to 5:21 a.m.—to activate all 23 southbound signs, at which time the northbound signs were activated.

Table 8. Variable message signs displayed on I-95 north at 5:21 a.m., January 17, 2003. The signs alternated between frames 1 and 2.

Factual Information

30

Highway Accident Report

〔图〕Second flare pattern25two flares set45mid-point inleft lane55Initial flare patternfive flares setdiagonally acrossleft l

Figure 11. Diagram of flare placement.

Analysis

46

Highway Accident Report

LOS programs would have called for the application of straight salt rather than a mixture of abrasives and salt.

The snowfall on the night of January 16–17 was light, which probably influenced ConnDOT's decision to initiate minimum (25 percent) treatment coverage. However, over time, even light snowfall could have overlaid the 7:2 mixture and reduced the capacity of the sand to provide traction. Also, though traffic was not heavy during the early morning hours, the traveling speed was likely sufficient to rapidly disperse the minimal amount of 7:2 mixture applied. CSP transcripts and witness statements indicated that—despite roadway treatment—the pavement remained slippery and hazardous, and accidents continued to occur. The Safety Board concludes that the sand and salt mixture that ConnDOT used to treat the light-to-moderate snow and icy roadway conditions prevailing in the early morning hours of January 17, 2003, was not as effective as straight salt because such a mixture tends to be more rapidly dispersed by high-speed traffic. The Safety Board believes that ConnDOT should develop an LOS program in accordance with FHWA and AASHTO guidance and incorporate the program elements, including a discussion of the limited usefulness of abrasives for roadway treatment, into the Snow and Ice Guidelines.

The Safety Board will also inform AASHTO of the circumstances of this accident and will advise the association to encourage its members to implement LOS programs that use the most effective snow and ice treatment materials for given weather, roadway, and traffic conditions in accordance with the FHWA Manual of Practice for an Effective Anti-icing Program and the AASHTO Guide for Snow and Ice Control.

Median Barrier

When the driver of the 1996 Freightliner tractor flatbed semitrailer lost control of his vehicle near MP 26.6, the truck entered into a counterclockwise skid toward the unpaved and depressed highway median. It collided with and overrode a 32-inch-high portable concrete barrier system. The collision overturned and laterally deflected five barrier segments toward the south and visibly cracked one segment. The Safety Board examined issues associated with the type of barrier used at the accident site and its placement on an unpaved surface.

In 2002, heavy trucks accounted for between 26 and 28 percent of the 125,300 vehicles that traveled daily on the accident segment of I-95. However, the 32-inch-high standard portable concrete median barrier in use at the accident site was designed only to redirect cars and private-use trucks under 4,400 pounds. Safety Board investigators calculated that the 1996 Freightliner, including the compressors and the flatbed semitrailer, weighed over 42,000 pounds and had a center of gravity approximately 4 inches higher than the top of the median barrier. In addition:

Analysis

49

Highway Accident Report

barrier prior to the Fairfield accidents. Figure 13 illustrates how a two-stage process would have avoided placing the barrier on the unpaved median.

〔图〕32-inch-hightemporary barrierTwo inside northboundTwo inside southboundtraffic lanesEarthen mediantraffic lanesPavedPave

Figure 12. Cross-section of positioning of portable concrete barrier at time of accidents.

〔图〕插图 〔图〕插图

Figure 13. Cross-section of I-95 illustrating exemplar stages 1 and 2 of resurfacing project. (Not to scale. View limited to two inside lanes for clarity.)

As a result of the Fairfield accidents, ConnDOT has adopted a procedure for reviewing and approving revisions to construction projects. This new procedure requires that the Offices of Construction and Engineering explicitly approve any proposed revision to a construction contract. All construction revision submissions must include reasons for the change, proposed procedures, affected locations, cost estimates, and a completion date. Requiring that highway engineers, knowledgeable in State and national guidelines, review proposed revisions to construction projects should mitigate practices that contradict, or are not covered by, established guidelines for highway construction.

56

Highway Accident Report

Recommendations

As a result of its investigation, the National Transportation Safety Board makes the following safety recommendations:

New Recommendations

To the Federal Highway Administration:

Conduct crash testing of standard and high-performance portable concrete median barriers on unpaved surfaces to determine whether they meet the test level criteria of National Cooperative Highway Research Program Report 350, Recommended Procedures for the Safety Performance Evaluation of Highway Features. (H-05-27)

To the Connecticut Department of Transportation:

Develop a level-of-service program in accordance with Federal Highway Administration and American Association of State Highway and Transportation Officials guidance and incorporate the program elements, including a discussion of the limited usefulness of abrasives for roadway treatment, into the Snow and Ice Guidelines. (H-05-28)

Revise your State Traffic Commission Regulations to include the control of traffic through traffic incident management areas, as described in the Federal Highway Administration Manual on Uniform Traffic Control Devices. (H-05-29)

Work with the Connecticut Department of Public Safety to provide coordinated training on traffic incident management for both your own staff and that of the Connecticut State Police. (H-05-30)

To the American Association of State Highway and Transportation Officials:

Establish warrants in the Roadside Design Guide regarding the selection and use of high-performance barriers, including 42- and 50-inch-high concrete barriers, that are capable of redirecting heavy trucks. (H-05-31)

Appendix B

63

Highway Accident Report

Snow Control Goals, Regular Level of Service$^{9}$

Snow Control Goals, Modified Level of Service

NYSDOT recommends specific guidelines for the application of salt for light-to-moderate accumulating snow and rapidly accumulating dense snow:$^{10}$

Light to moderately accumulating normal snow-The initial application shall be 225 pounds of salt per mile, per lane. Follow-up applications shall be as necessary, and at 115 pounds of salt per mile, per lane. If the pavement temperature is over 30 degrees Fahrenheit and/or traffic volume is high, the initial application rate may be reduced.

Rapidly accumulating dense snow, freezing rain, pack, glaze and sleet (high ice content)–The initial application shall be 270 pounds of salt per mile, per lane. Follow-up applications shall be as necessary, and the rate of application will vary with the observed condition of the snow or ice on the pavement surface. If surface is still glazed or packed, the follow-up application shall be 270 pounds of salt per mile, per lane (the same as the initial application). If the surface is mealy or slushy, the follow-up application shall be 115 pounds of salt per mile per lane.

The Highway Maintenance Guidelines offer the following guidance on the use of abrasives (sand)/chemical mixtures:$^{11}$

The use of “sweetened” mixtures like 50-50 (1 part abrasives and 1 part salt) is wasteful and inefficient. If spread at the normal application rate for abrasives, this mixture will place 40% more salt on the road than a normal application of pure salt. The effectiveness of that salt is reduced by the presence of the abrasives.

Abrasives, overall, are more costly (based on initial application, reapplication, cleanup and other factors) and less safe to use than salt (if abrasives were used

Appendix C

66

Highway Accident Report

Table C-1. Barriers crash tested and approved at TL-5 and TL-6.

The crash tests performed for the barriers identified in table C-1 were done on a flat paved surface when accelerating the tractor-trailer to the desired speed to provide for unrestricted trajectory of the vehicle following impact. The barriers were anchored to the surface during the tests.

The performance level of each test vehicle varied with each crash test. The 50,050-pound tractor-trailer rolled over when crash tested into the 42-inch-high vertical concrete parapet. The 80,080-pound truck rolled over when crash tested into the Texas type HT (modified T5) barrier. Under TL-5 and TL-6, the crash test criteria in NCHRP Report 350 allow the vehicle to roll over.

Portable Concrete Barriers

Table C-2 identifies the types of 42-inch-high portable concrete barriers that have been crash tested and approved under NCHRP Report 350 at TL-3 and TL-4. None of the portable concrete barriers (at 32 or 42 inches high) were crash tested with tractor-trailers. The tests were performed with the barriers resting on a flat paved surface. The barriers were connected to each other either by a channel or a pin-and-loop design; however, they were not anchored to the paved surface. The barriers were connected and arranged in a series of 20. The crash test criteria under NCHRP Report 350 allow the test vehicle to penetrate the barrier, but it may not roll over—nor can the barrier cause excessive forces on the vehicle occupants.

Appendix C

67

Highway Accident Report

Table C-2. Barriers crash tested and approved at TL-3 and TL-4.

The connection between the 42-inch-tall F-shape portable concrete barriers consisted of two sets of two perforated C-shape steel channels with the open sides alternately positioned such that one leg of each channel fit between the legs of the mating channel on the adjacent barrier segment. The barriers were not connected by a pin-and-loop design.

68

Highway Accident Report

Appendix D

Variable Message Sign Placement

I-95 North

I-95 South