Geometric Design of Driveways (2010)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

93 CHAPTER 4 Data Collection and Analysis The issues related to design of the vertical alignment of driveways that were selected for study fall into the following three categories.  Driveway grades and measured vehicle ground clearance  Driveway grades and signs of inadequate ground clearance  Driveway grades and speeds of entering vehicles The following sections in this chapter discuss the procedures and findings from the Task 6A research activities. 4.1 DRIVEWAY GRADES AND MEASURED VEHICLE GROUND CLEARANCE There are two modes in which the underside of a vehicle can drag or hangup. One mode occurs when the road profile creates a sharp vertical crest, which causes the underside of the vehicle between the front and rear axles to drag on the pavement surface. The other mode occurs when the road profile creates a sharp vertical sag, which causes the underside of the vehicle either to the front of the front axle or to the rear of the rear axle to hang up. Exhibit 4-1 displays both of these conditions. To determine the change in vertical profile at which the underside of the vehicle will drag, one makes x- and y-coordinate measurements of the critical points on the underside of a vehicle that will define a profile or silhouette of the vehicle's underside. Then one conducts a geometric analysis to determine the least change in profile grade that will cause the underside of the vehicle to come in contact with the driveway surface. Exhibit 4-2 displays the geometry of this analysis. EXHIBIT 4-1 Two modes of vehicle underside dragging WB=wheelbase OHF= front overhang OHR= rear overhang CREST: Underside will drag if the axle-to-axle ground clearance is inadequate. WB/2 driveway WB/2 roadway curb SAG: Underside will drag if the axle-to-bumper ground clearance is inadequate. WB curbroadway OHFOHR driveway

94 EXHIBIT 4-2 Vehicle ground clearance geometry Selecting and Locating Vehicles The project oversight panel directed the contractor to define the ground clearance dimensions of at least three vehicles, and a fourth if the budget allowed. The project oversight panel specified that the vehicles to be defined include a small automobile and a Class A motor home (“diesel pusher”), and the contractor suggested a pickup truck pulling a trailer and a beverage delivery truck. To locate vehicles to measure, the contractor contacted nearby automobile dealers, beverage distributing companies, and recreational vehicle dealers. The ground clearance of one automobile was measured on a dealer's lot, and another was measured on a dealer’s showroom floor. The beverage delivery truck was measured inside the distributor’s warehouse. The motor home was measured on a dealer's lot. Dimensions for the pickup truck and trailer were obtained from manufacturers’ literature. Measuring Vehicle Ground Clearances To measure the underside in hard-to-reach areas, a technician fabricated a specially designed measuring jig. This jig, shown in Exhibit 4-3, consisted of a black rigid flat base, a silver vertical rod at each end of the base, and an orange rigid parallel bar with bushings on each and that allowed the bar to slide up and down on the two vertical rods. To measure the vertical clearance at any given spot, two people slide the rigid parallel bar up to contact the underside of the vehicle, then make a measurement from the ground up to the top of the rigid bar. Determine the change of grade G2-G1 at which underside hangup will occur. driveway G 1 G2 WB B X1 X2X0 XR OHF OHR

95 EXHIBIT 4-3 Measuring vehicle ground clearance Vehicle Ground Clearance Measurement Findings Exhibit 4-4 shows the resulting x- and y-coordinates of the points that define the underside profile of the four measured vehicles. From these measurements, the profile or grade change at which the vehicle would drag in both crest and sag conditions was computed.

96 EXHIBIT 4-4 Measured coordinates of vehicle undersides (0 .2 9, 0 .9 6) (7 .1 8, 1 .5 5) (9 .3 6, 0 .9 6) (2 6. 31 , 1 .0 4) (2 8. 26 , 1 .6 0) (3 0. 36 , 1 .0 8) (3 9. 56 , 1 .7 5) Class A diesel motor home (diesel pusher), unoccupied Alfa See Ya’! Gold® (x , y c oo rd in at es ) 0’ 10’ 20’ 30’ 40’ 50’ Y X front: x = 0’ rear: x = 40.66’ (ladder extra) ladder 0’ 10’ 20’ 30’ 40’ 50’ Y X (x , y c oo rd in at es ) (0 .7 1, 1 .5 5) (2 .9 1, 1 .5 5) (5 .5 1, 0 .8 7) (1 9. 26 , 1 .1 1) (4 2. 26 , 1 .6 4) (4 0. 17 , 0 .9 9) Tractor with 10-bay beverage trailer, about 5/8 loaded International tractor, Centennial Body trailer (0 , 1 .2 7) (1 5. 26 , 1 .5 8) (4 6. 36 , 1 .6 4) (5 3. 46 , 1 .3 6) (9 .3 6, 0 .9 2) (4 8. 52 , 0 .9 4) front guard: x = 0’ rear (A =0 .0 , 0 .7 1) (D =5 .1 6, 0 .5 0) (F =8 .2 9, 0 .4 6) (J =1 3. 46 , 0 .6 1) (K =1 5. 04 , 0 .8 5) rear = 15.79’ Cheverolet Camaro 1998 (E =6 .9 6, 0 .3 7) (B =1 .4 6, 0 .3 7) (G =1 0. 96 , 0 .4 8) A B D E F G J K 0’ 5’ 10’ 15’ Y X front: x = 0’ (x , y c oo rd in at es ) (H =1 2. 29 , 1 .0 4) (C =3 .8 8, 1 .0 4) C H (A =0 .0 , 0 .4 2) (D =4 .2 4, 0 .4 5) (G =1 4. 34 , 0 .7 2) Cheverolet Corvette Z06 2008 (B =1 .1 8, 0 .2 6) (E =1 0. 53 , 0 .5 0) rear = 14.57’ 0’ 5’ 10’ 15’ Y X (x , y c oo rd in at es ) front: x = 0’ A B D E FC G (F =1 1. 82 ) (C =3 .0 2)

97 4.2 DRIVEWAY GRADES AND SIGNS OF INADEQUATE GROUND CLEARANCE Visible scrape marks on the surface that result from the dragging of vehicle undersides can be clear indicators that the profile geometry of an existing driveway is too abrupt. The project oversight panel directed the contractor to measure the profiles of driveways with scrape marks that the contractor encountered during the course of conducting the research. A few of these driveways were measured by one person with a 24-inch digital level, while most of them were measured by two-person crew with land surveying equipment. Often, two profiles were measured. For instance, for driveway with visible scrape marks on the entry side, the entry-side edge and the driveway centerline were profiled. One of the driveways with visible scrape marks that the contractor measured is shown in Exhibit 4-5. EXHIBIT 4-5 Example of a driveway with visible vehicle underside scrape marks The crest and or sag breakover angles near scrape marks on each driveway were calculated. For those driveways with a crest breakovers close to a sag breakovers, the investigator was not able to determine with certainty if the scrape marks were the result of the crest or the sag profile. The 31 driveways that were measured are listed in the Exhibit 4-6. The individual data forms for each measured driveway are presented in Appendix F.

98 EXHIBIT 4-6 Driveways with visible scrape marks that were measured Street Block Site Name Notes Breakover Grade Crest Sag AUSTIN 1 Balcones N 5206 Highland Park Baptist Church south exit drive na 16.8% 2 First HEB shopping center west drive na 9.4% 3 Hancock W 3339 Russells’ Bakery continuous drive na 21.3% 4 William Cannon W 1021 Genie Car Wash west drive na 17.0% 5 William Cannon W 2501 Stonegate One, medical offices middle drive, above the sidewalk na 13.5% FAYETTEVILLE-SPRINGDALE 6 Cliff Aqua, multifamily na 8.6% 7 Cliff Lapis, multifamily 11.3% 13.4% 8 Cliff E Peridot, multifamily na 15.1% 9 Crossover N 1831 Automatic Car Wash south driveway 16.5% na 10 Dickson W 800 SE Building, classrooms na 11.1% 11 Gregg S 41 Myers' Apartments na 10.2% 12 Hyland Park 2730 single-family residence na > 20% 13 Lafayette Valero, gas station middle drive 10.6% 18.5% 14 Mission 1813 Tim's Pizza west drive na 11.2% 15 North St North Street Condos 10.9% na 16 Rock Cliff 583 single-family residence na > 20% 17 Rock Cliff 599 single-family residence na > 20% 18 St Charles Colliers’ Drug na 14.4% 19 Sapphire Aqua, multifamily 13.9% na 20 Sapphire Goldrush, multifamily 14.2% 12.6% 21 Sixth O'Reilly's east drive 16.5% na 22 Sunbridge W 6 Arthritis Center 13.1% na 23 Sunbridge E 18 McClelland's Fly Shop 11.4% 17.2% 24 Sunbridge E 114 Sunbridge Center 12.0% na 25 Sunbridge E 158 VA Dental ? 9.7% 13.9% 26 Sunbridge E 180 VA Outpatient 11.4% 14.2% 27 Sunset 2255 Fuji Restaurant west drive 14.0% na 28 Sycamore Royal Cleaners west edge na 20.0% TULSA 29 71st E 6550 Hausam Realty, Arvest Bank ? 9.0% 16.4% 30 Archer E 6616 Super 8 Motel na 16.9% 31 Mingo Union Plaza shopping center west drive ?10.5% 16.8% Minimum Observed Breakover Grade 10.6% 8.6%

99 The vehicle geometry that causes the undersides of vehicles to drag on the pavement surface is a combination of ground clearance height and either the wheelbase or the overhang length. Obviously, there are many possible combinations of height and length that could cause the underside to scrape the pavement surface. The driveways at which these measurements were made are traversed mainly by private automobiles and similar sized vehicles. When determining the grade on either side of a breakover point, the contractor often computed the average grade of the driveway surface within four to eight feet of the scrapes or the breakover point. 4.3 DRIVEWAY GRADES AND SPEEDS OF ENTERING VEHICLES Most of the research activity was directed toward measuring and comparing the speeds and elapsed times of vehicles turning right and turning left into driveways having different vertical alignment or profiles. The project oversight panel had directed the contractor to examine this aspect of traffic operations because of the perspectives of various interest groups. Some advocates for bicyclists, pedestrians, and pedestrians with disabilities are concerned that vehicles enter driveways at speeds they consider excessive and create a hazard. On the other hand, those focusing on motorists’ are concerned that the more time it takes for a vehicle to enter a driveway, the more exposed that vehicle is to being struck by other through vehicles. So there are the following two underlying questions. 1. To what extent does the vertical alignment affect the speed and the elapsed time of vehicles turning right or turning left into a driveway? 2. What effect do these differences have on the exposure of all users (bicyclists, motorists, pedestrians, pedestrians with disabilities)? Criteria for Suitable Sites The researchers determined that the sites selected for the study of speed and elapsed time as vehicles entered driveways of different grades should possess attributes that are representative of a broader population of driveways. To the extent possible, the various driveways selected should have somewhat similar attributes, in order to reduce the variability among the attributes of the sites at which the data would be collected. Even though it was theoretically desirable to find sites having the same widths, entry shapes, and shape dimensions, the researchers recognized that it was highly unlikely that this could be perfectly achieved. It was decided that one factor that could increase the similarity among the sites, in terms of

100 characteristics such as volume and speed of traffic on the through street, would be to select some driveways along the same street. The researchers developed an initial set of criteria for identifying potentially suitable driveways for data collection. The criteria evolved during the course of the search, with some of the evolution affected by what traits were more frequently encountered. The following criteria helped identify a pool that is typical of those driveways serving small- to medium-sized commercial and professional office developments that became quite common in the latter part of the 1900s along non-fringe suburban multilane arterial roadways. The term “non-fringe suburban” was selected to indicate land that was not at the edge of the developed urban area, where conditions approach those of an open, rural highway, yet not in or near the downtown urban core, where speeds are typically lower and congestion is greater. General Traits 1. The site has space to accommodate people and equipment collecting the data, with a clear line of sight to the driveway entry 2. The driveway has sufficient volume to make the time spent in data collection productive 3. The driveway is not built to appear like a street (note: this tends to exclude driveways to large commercial developments, such as large shopping centers) 4. Through-street posted speed limit is 40 or 45 mph Plan View Design 5. Driveway is either 2 or 3 lanes wide 6. The driveway does not have pavement markings that would conflict with the standard marking the contractor installs at each site 7. Driveway throat length (connection depth) is not less than 23 ft, measured from face of curb 8. Driveway entry transition shape is curved (i.e, not tapered/triangular) with a radius of 13 to 19.5 ft 9. Driveway intersects street at or close to a 90O angle 10. Both the driveway and the through-street are fairly straight in the immediate vicinity of where they connect 11. Driveway connects to a multilane street 12. The width of the through-street outer lane from curb face to lane line is between 10.5 and 13.5 ft (e.g., no shoulder, bike lane, or auxiliary right-turn lane) 13. The through-street has a separate left-turn lane or a two-way left-turn lane (TWLTL) Vertical Alignment

101 14. No vertical lip at the roadway-driveway interface 15. The driveway does not slope markedly downward from the through-street into the site 16. The street grade is relatively flat, not steep Operations - Driveway Interaction with Other Traffic 17. Driveway is not signalized 18. Driveway traffic operations are not often affected by a nearby traffic signal, such as the backup queue from a nearby signalized intersection 19. Enough separation so driveway traffic is not often affected by any other driveway or street Searching for Suitable Data Collection Sites Searches were conducted for driveways suitable for data collection in the following locales. Arkansas: Bentonville, Fayetteville, Rogers, Russellville, Siloam Springs, Springdale Missouri: Springfield New Jersey: Montville, Parsippany, Wayne New York: Roslyn Heights, Yonkers Oklahoma: Broken Arrow, Jenks, Sapulpa, Tulsa Texas: Austin The process of searching for suitable data collection sites and making detailed inspections and measurements lead to the following observations about driveways. Some driveway plan design elements, as constructed and in-place, are irregular. Specifically, highly irregular and variable entry radii were encountered. A common manifestation of this was a curved entry shape in the form of a spiral, not a curve with a constant radius. This caused some potential sites to be excluded from further consideration. Driveway grades are seldom constant across the width of the driveway. This is inherent in the geometric nature of one plane surface (the driveway surface) intersecting another plane surface (the edge of a roadway) on a grade. Unless the cross slope of the driveway exactly follows the grade of the street, laws of geometry cause the driveway grade to vary across the width of the driveway. In some areas, it may be common practice to construct the outer one to two feet of the outside lane (i.e., gutter area) with a greater cross slope than that of the rest of the lane. Since this construction practice makes it difficult to quantify the street cross slope and the actual grade change perceived by the driver at the street-edge interface with the driveway-end, otherwise desirable sites were excluded from further consideration due to the increased gutter cross slope.

102 Even though a designer may specify a measurement to a hundredth of an inch, roadway construction is seldom that precise. This is not to imply that designers should be less precise; rather it is to state that an expectation of construction to that precision is unrealistic. And even if a roadway were constructed with a high precision, settling or other material deformation would eventually bring about a change of dimensions. Specific to this study, the researchers observed that the rutting and shoving of asphalt concrete surfaces created slight variations in the cross slope over the width of a lane. Selecting Suitable Data Collection Sites Recognizing that the only way to obtain a perfect set of data collection sites would be to fund and construct the driveways specifically for this project, the researchers exercised judgment to evaluate potential driveway sites. After conducting visual inventories along many miles of roadway in a number of cities, a candidate short list of driveways with relatively similar characteristics evolved. All of the selected driveways serve small to medium-sized commercial or office tracts abutting non- fringe suburban arterial roadways with speeds of 40 or 45 mph. (At one driveway site, either the posted speed limit was incorrectly noted during an initial search, or the speed limit was changed to 50 mph.) All of these driveways connect to multilane (4 or 6 through lanes) arterials with either a raised median or a two-way left turn lane (TWLTL). After considering the various attributes associated with the driveway sites on the candidate list, certain sites were selected for actual field data collection and analysis. The researchers measured driveway attributes such as width, entry radius, and profile grades at each site. The sites selected in Austin, Texas were all along the same arterial roadway. The sites selected in Tulsa and in the suburb of Broken Arrow, Oklahoma were all in the southeast part of the metropolitan area, where Tulsa and Broken Arrow abut. One of the sites was in Fayetteville, Arkansas. The selected driveways were grouped into one of three categories shown in Exhibit 4-7.  The steeper driveways have grades up from the gutter line of 12.5% to 15.5%, with changes of grade between roadway cross slope and the driveway grade (i.e., breakover) between 13.5% and 19.0%.  The moderate-grade driveways have grades up from the gutter line between 6.0% and 9.0%, with breakovers between 5.0% and 10.5%.  The flatter driveways have grades up from the gutter line between 1.5% and 5.0%, with breakovers between 3.5% to 6.5%.

103 EXHIBIT 4-7 Driveway grade groups Exhibit 4-8 lists the sites selected for study. Exhibit 4-9 shows example site photographs. Photographs of all sites are in Appendix G. No lip or other abrupt vertical profile element roadway Cross slope Breakover ∆ Flatter 1.5%-5% ∆ = 3.5%-6.5% Steeper 12.5%-15.5% ∆ = 13.5%-19% Moderate 6%-9% ∆ = 5.0%-10.5%

104 EXHIBIT 4-8 Driveways selected for speed and elapsed time studies Site Description Street Alignment Speed Limit (mph) Outer Lane Width (ft) See Note Street Cross Slope Grade Change Near Gutter Line Dway. Grades Throat Length (ft) Throat Entry Traffic Pattern Rt. Turn Entry Radius (ft) Dates of Studies STEEPER turn Sep 18 conflict Jan 7 Jul 29 turn Sep 15 conflict Jan 5 thru Feb 9 free Mar 15 thru May 13 free MODERATE mixed Mar 14 free 64 turn 16 Aug 12 free thru Sep 16 conflict turn Sep 17 conflict Jul 30 FLATTER turn Nov 16 free May 14 thru Feb 27 free Jul 1 turn Jan 6 conflict turn Mar 16 free 1.6%-2.1% 3.7% 7.1% 6.5' / 2.2% 4.5' / 0% Red Robin Straight, G  -0.4% 45 11.5 5.1% -2.0% 5.0% 58 16 17 66 18 41 19.5 13 29 Okla. Central Credit Union Straight, G  0.4% 45 McAlisters, Meineke Straight, G  -1.0% 6.4% 12' / 5.5% -3.8%/ -1%45 13 3.0% 9' - 0.8% 52 15.5 12.6% 29 8.7% 10' / 12.8% 6.0% 10' / 1.1% J D China 5.2% 19.5 4.7% 20' - 2.0% 6' 43 19 Hollywood Video - Southcross Plaza Wendys 13 -2.1% 6.5% 4.4% 11' / 1.3% 40 13.8% 10.1% small shopping center- HEB grocery Arvest Bank 10.0%Straight, G  -2.4% 40 11.5 -1.8% 10.5% Stonegate One - Austin Pain Assoc. Genie Car Wash 40 11.0 Straight, G  2.6% Straight, G  -0.9% 40 12.0 11.2 Straight, G  0.6% 40 12.6 45 40 Straight, G  -1.2% -1.2% R=2292 ft, G ≈ -0.5% 40 11.5 2.0% 11.0 -4.0% Shell gas; self storage 17.0% -2.6%/ -7% 1' Straight, G  0.0% 40 11.5 -0.5% Straight, G  -0.6% 23 12.8% 2' / 15.6%-4.2% -3.1% 18.6% 15.5% 6.5' / 0.3% 6' / 13.8% 48 4815.8%Straight, G 1.6% 40 13.5 13.2% 4' / 3.4% 6' / 5.3% 8' Union Plaza - Mardells 13.5 19 19

105 EXHIBIT 4-9 Examples of speed data collection sites Descriptions of Steeper Sites The Stonegate One professional offices in Austin consist of a series of upscale looking buildings in a strip mall arrangement. The driveway at which data were collected serves medical offices. Stonegate One is on West William Cannon Drive, which has four lanes and a raised median. This roadway is abutted by mostly small- and medium-sized commercial and office tracts. The vacant tract across the street was undergoing site grading and construction when data were collected. Genie Car Wash in Austin offers both self-service and attendant car washes on a stand-alone tract. It is on West William Cannon Drive, a six-street roadway with a raised median. Due to the raised median, only right turn movements into the site are possible. The roadway is abutted by mostly small- and medium-sized commercial and office tracts. In the immediate area, a multifamily area and the back side of some single family lots abut the street. One-story professional offices are across the street. Union Plaza shopping center is a medium-size center occupying the northeast corner of South Mingo and East 71st in Tulsa. It is anchored by a large hobby-and-crafts store and a large bookstore. The driveway at which data were collected is on Mingo, a four-lane roadway with a TWLTL. In the immediate vicinity, South Mingo is abutted by a variety of commercial land uses, and a high school and a

106 large church building. (Note that at this site, data were collected on Saturday.) Across the street, there are small stores on outparcels, with a large discount store behind them. The Arvest Bank branch office is on the northeast corner of East 61st and 89th East in Tulsa. The tract is connected to one adjacent site, a small one with commercial tenants. The driveway at which data were collected is on East 61st, a four-lane roadway with a TWLTL. This roadway is abutted by mostly small commercial and professional sites. The playground for a school is across the street. Descriptions of Moderate Sites The Oklahoma Central Credit Union branch in Broken Arrow occupies a stand-along site on South Aspen, a four-lane roadway with a TWLTL. The roadway is abutted by mostly small- and medium-sized commercial and office tracts. The tract to the south (behind the field of view in the photograph) is vacant. Across the street, there is a one-story thrift store. McAlister’s Deli in Broken Arrow shares a driveway with a Meineke Car Care Center to the south. It is on Aspen, a four-lane roadway with a TWLTL. The roadway is abutted by mostly small- and medium-sized commercial and office tracts. A Walmart is behind the site, and a car wash is across the street. The small shopping center on the northeast corner of West William Cannon Drive and South First in Austin is anchored by a HEB grocery store. The driveway at which data were collected is on Cannon, which has six lanes and a raised median. Cannon is abutted by mostly small- and medium-sized commercial and office tracts. Hollywood Video in Austin is one of the many tenants in Southcross Plaza, an approximately ¼- mile long strip center along West William Cannon Drive, which has six lanes and a raised median. The roadway is abutted by mostly small- and medium-sized commercial and office tracts. Some of the land across the street in undeveloped, and some is occupied by a shopping center with a grocery store. Descriptions of Flatter Sites Wendy’s Restaurant in Tulsa is connected to other commercial tracts on the south side of East 71st Street, which has six lanes and a raised median. The roadway is lined on both sides by a variety of commercial uses. J. D. China Restaurant is on a stand-along tract on West 6th Street, a four-lane roadway with a TWLTL, in Fayetteville. The roadway is in an area lined on both sides by mainly small commercial

107 tracts. The tract immediately across the street is occupied by a hardwood mill, with a solid wood fence along the right-of-way line. The Shell gas station and the self-storage units share a driveway on the south side of East William Cannon Drive in Austin, and the driveway is also connected to a strip shopping center to the east. Cannon has six lanes and a raised median. The roadway is abutted by mostly small- and medium-sized commercial and office tracts. In this section, apartment complexes are across the street. The Red Robin Restaurant is on the south side of Kenosha in Broken Arrow (an extension of E. 71st in Tulsa), a four-lane roadway with a TWLTL. The tract is connected internally to a tract to the east. The roadway is abutted by a variety of commercial tracts on both sides. The tract immediately to the west (to the right in the photograph) is undeveloped. Verifying the Vertical Alignment In order to define the profiles of each studied driveway, the contractor had taken elevation readings with surveying equipment at the observed break points (i.e., points at which changes in the profile were observable) along the profiles of each driveway. The project oversight panel expressed concern that the contractor may have not taken elevation readings at intervals spaced closely enough to precisely define the profiles of the driveways. As a check, the project oversight panel asked the contractor to resurvey three driveways with readings at more closely spaced intervals. The contractor actually resurveyed seven driveways at more closely spaced intervals. Exhibit 4-10 shows one of the profiles generated from the initial or previous survey and from the checking re-survey. To illustrate how the information from the initial or previous survey can be compared with the later re-survey, the grades at the Arvest driveway were originally reported, based on surveying readings taken at points with observable changes of grade, as having a street cross slope of 1.2% and a driveway grade of 12.6%, creating a breakover angle of 13.8%. From the more detailed re- survey, shooting elevation readings at one foot intervals near the roadway edge, the street cross slope was found to be 1.15%, the driveway grade was 12.52%, and the resulting breakover grade was 13.67%.

108 EXHIBIT 4-10 Profile of Arvest driveway Arvest driveway, E. 61st St., Tulsa, OK C ur b Elev. of dway 14.2' from Rt edge 6. 03 6. 08 6. 09 6. 10 6. 12 6. 13 6. 15 6. 15 6. 05 5. 94 5. 82 5. 71 5. 59 5. 49 5. 36 5. 25 5. 14 5. 01 4. 78 4. 53 4. 26 4. 11 Distance from curb face -1 0 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 12 14 16 17 .3 Elevation of dway Rt edge 6. 24 6. 28 6. 30 6. 31 6. 32 6. 34 6. 35 6. 35 6. 23 6. 09 5. 96 5. 83 5. 70 5. 57 5. 44 5. 31 5. 17 5. 04 4. 79 4. 51 4. 20 4. 07 Average elevation 6. 14 6. 18 6. 20 6. 21 6. 22 6. 24 6. 25 6. 25 6. 14 6. 02 5. 89 5. 77 5. 65 5. 53 5. 40 5. 28 5. 16 5. 03 4. 79 4. 52 4. 23 4. 09 Values from previous measurements, based on level readings only at observed breakpoints. Breakover grade change = Elev. calculated from breakpoints 6. 14 6. 15 6. 16 6. 17 6. 18 6. 19 6. 20 6. 22 6. 23 6. 24 6. 25 6. 12 6. 00 5. 87 5. 75 5. 62 5. 50 5. 37 5. 25 5. 12 5. 00 4. 75 4. 50 4. 25 4. 09 Difference between calculated and actual 0. 00 0. 00 0. 00 0. 00 -0 .0 1 -0 .0 1 -0 .0 2 -0 .0 2 -0 .0 2 -0 .0 2 -0 .0 2 -0 .0 3 -0 .0 3 -0 .0 3 -0 .0 3 -0 .0 3 -0 .0 4 -0 .0 2 0. 02 12.6% 12.52% Calculated grade 13.67% -1.2% 13.8% -1.15% Arvest Driveway Elevations 4.00 4.50 5.00 5.50 6.00 6.50 -15 -10 -5 0 5 10 15 20 25 Distance from Curb Face E le va tio n Closely spaced Only breakpoints

109 After resurveying five of the driveways at which data had been collected, the contractor had yet to find a driveway where the breakover sag at the gutter line had been ameliorated by rounding. At two of the sites, the contractor observed rounding where a driveway ascending from the roadway gutter suddenly flattened as it met a sidewalk. At the Stonegate driveway, this rounding was determined to be 0.5 inch. At the Hollywood driveway, although rounding was visible, it was so slight that it was not detected with the surveying equipment. At the sixth site to be surveyed, McAllister's, the survey readings identified a flat gutter pan that had the effect of flattening or rounding the sag profile by about 0.5 inch at a point a few inches in front of the curb line. A seventh site at which data had been collected, Union Plaza shopping center, was surveyed to quantify the grades that were causing the undersides of vehicles to drag on the driveway surface. From this survey, it was noted that the cross slope at the gutter pan was actually steeper downward than the cross slope of the roadway, which created a dip of slightly less than 0.5 inch, the opposite of rounding. In general, these exhibits indicate that the profiles made from the readings of the initial survey were close to the profiles made from the follow-up checking survey. Of the seven sites at which data were collected, the checking re-surveys found that one of them had slight rounding of the sag, one of them had a dip at the sag, and the other five had no noticeable adjustment of the profile at the sag point where the street cross slope abuts the driveway grade. At one site, the survey also identified pavement rutting in the outer lane of the through roadway. Data Collection Procedures Prior to the field data collection, project oversight panel members had suggested that the contractor consider using contact closure switches to record speed of vehicles entering the driveway. After evaluating alternative methods, the contractor proposed patterns of contact closure switch pairs to record the speed and elapsed time between successive stations as vehicles turned right or turned left into the driveway. To collect data, the contractor made precise measurements to set the location of pairs of contact closure switches, then taped the switches to the roadway and driveway surfaces. A pair of switches constituted a sensor. Wiring attached to the switch ends was also taped to the surface, and the wiring for each turning movement was connected to a data logger, which in turn was connected to a laptop computer loaded with a program specially designed to receive and store the readings generated in the data logger. The data logger allowed input signals from the switches to be processed and by means of knowing elapsed time over a set distance, calculated vehicle speed. A person operating the computer would key the devices to record data when a turning vehicle approached the sensors. A camcorder was aimed to

110 include Sensors 2 and 3 in the field of view. Exhibit 4-11 displays these two patterns, one for right- turning and one for left-turning vehicles. EXHIBIT 4-11 Sensor layout diagrams To record the data from right-turning vehicles, three sensors (i.e., pairs of contact closures switches) were deployed. These pairs were named Sensors #2, #3, and #4. Initially, the left turn data were collected with four pairs of contact closure switches, numbered #1 through #4. Due to ongoing problems with the switches, the decision was made to eliminate pair #1. This was done to eliminate the long electrical wiring required to reach these switches, the associated demand for power, and the longer signal transmission distance. It was hoped that this would improve the reliability of the remaining three left-turn switch pairs. Note that the pairs of switches actually recorded the speed of the vector perpendicular to the orientation of the switches, which may in some cases be slightly less than the actual forward speed of the vehicle. The switches at Sensor #3 recorded the vehicle speed vector toward pedestrians on the sidewalk. Data Collection Problems and Adjustments At some sites, data were collected on multiple dates. The main reason for repeat visits to the site was a technical failure, either with the wiring leading to the contact closure switches or the software. Repeat visits also had to be made because of damp weather and because of vehicles damaging the contact closure switches. The manufacture of the contact closure switches stated that they were intended to be used by vehicles going straight, and the tire movement of turning vehicles could cause problems. 25’ RT. TURN INTO DRIVEWAY 3 4 2’ 15’ 25’ 2 25’ 2’ ROADWAY LT. TURN INTO DRIVEWAY 3 4 11’ 15’ 25’ 2 1 5’

111 For the first three studies conducted, Sensor #4 was positioned 25 ft back from the roadway curb line. From observations during this data collection, it was concluded that at this distance, some drivers were beginning to react to maneuvers or traffic conflicts in the driveway throat ahead. This was affecting speeds differently at different sites. Therefore, this distance was adjusted to 15 ft back from the roadway curb edge. At the Stonegate driveway, there were numerous marks from the scraping of vehicle undersides at the locations for Sensors 3 and 4. During the first data collection trip, some vehicles scraped the sensors. Based on this experience, in the second and third studies at the site, Sensor 3 was shifted two feet closer to the curb, so the lead switch aligned with the curb face. During the second study, Sensor 4 was shifted two feet farther into the driveway throat, so the leading switch was 17 ft from the curb face. During the third study, Sensors 2 and 4 were both shifted up (i.e., in advance) two feet, to preserve the standard spacing between sensors. At the Union Plaza site, the slightly wider outer lane on Mingo Road caused left turn Sensor 2 to be struck by so many through vehicles that the sensor was damaged during the February study. During the repeat left turn study in March, left turn Sensor 2 was shifted two feet, so the lead switch was 15 ft from the curb face instead of the normal 13 ft. Because the locations of some sensors were moved, adjustments were made during the analysis. These are discussed later in this report. Achieving a More Common Entry Throat Width To help confine those vehicles turning into the driveway to a common width at the various sites, the contractor created a driveway centerline by installing a 15 foot long strip of 4 inch wide yellow pavement marking tape. To partially compensate for variations in the radii among the different sites and for the construction of slightly irregular radii, the contractor placed the yellow pavement marking tape at the greater of either an offset distance of 13 ft from the straight edge of the driveway, or after measuring back from the face-of-curb (FC) edge a distance of 13.2 ft, an offset distance of 14.2 ft from the entry radius. These 13.2 and 14.2 ft distances were chosen to replicate the throat width available 70° into a right turn having a 20 foot radius into a 13 ft wide entry lane (see Exhibit 4-12). The intent of the pavement marking tape was reinforced by the practice of positioning a blocking vehicle in the driveway exit lane (see Exhibit 4-13). This vehicle essentially parked in the exit lane until such time as another vehicle trying to leave the site pulled up behind the blocking vehicle. When this occurred, the blocking vehicle drove away and then quickly returned to the blocking position. This practice was followed at all sites except Union Plaza, where the volume of exiting traffic was sufficient to

112 perform the blocking task. A small piece of white pavement marking tape was placed to help the driver of the blocking vehicle stop close to the same spot each time. EXHIBIT 4-12 Width available 70° through a 90° right turn 2’ ROADWAY LEFT TURN INTO DRIVEWAY 3 4 11’ 2 Position vehicle to discourage corner cutting. Position sensor one lane f rom the curb. EXHIBIT 4-13 Position of blocking vehicle Offset to CL: max. of either 13’ OR * 14.2 ft @ 13.2 ft back from FC 18.79 20.0 6. 8420O 1.21’ 13.16’ ROADWAY 10’ 15’ D R IV EW AY *13.2 ft back from FC *14.2 ft *13.0 ft 6. 84 6. 84 D R IV EW AY

113 Exhibit 4-14 shows two people installing a pair of sensors at a data collection site. Exhibit 4-15 shows a site with data collection in progress. Note that the computer operators were partially screened from the view of drivers with a three-sided, 30 inch high barrier on a frame weighted to remain steady in the breeze. EXHIBIT 4-14 Installing contact closure switches EXHIBIT 4-15 Data collection in progress

114 Data Reduction and Analysis After collecting the field data, files were downloaded from the laptop computers, and the strings of data were formatted into columns in a spreadsheet. Those reducing the data meticulously examined spreadsheet entries and viewed video of the vehicles turning into the driveways. The person reviewing the video tapes noted when right turning vehicles swung wide in the through lane and crossed the white lane line. The reviewer noted when either left or right turning vehicles crossed the yellow driveway centerline that had been installed by the researchers. Also, reviewers noted when there was interference with entering vehicle, such as a pedestrian walking in front of it. Such cases were flagged for exclusion, so the analysis would consider only unimpeded vehicles that made turns from and into the provided lane width. Considerable effort was directed to screening the data to remove erroneous readings. Some examples follow. Sometimes, the person operating the laptop computer collecting right turn data might incorrectly assume that an approaching vehicle was about to turn right, and set the system to record data. These through vehicles would trigger a reading on the Sensor 2 pair, but not on the following Sensors 3 and 4. In this event, Sensors 3 and 4 would have unrealistic speed readings. Screening identified these events. Another not uncommon event was that one of the switchers in a sensor pair was stuck closed, having not rebounded from a previous tire strike. In such instances, an unrealistic speed would be generated. Again, screening identified such instances. Another source of bad data was the result of entering vehicles turning wide. For instance, a vehicle turning left into a driveway might almost miss the sensors, but barely clip the lead switch with the inside front tire. Continuing in an arc, the front tire would miss the trailing switch. But the inward-tracking rear tire would cross both switches, triggering the hit on the trailing switch. This would generate unusually low readings, in the neighborhood of 2 mph. A checking routine was coded in the spreadsheet to identify suspect readings. The measured elapsed time between two sensors was compared to the elapsed time calculated from the average of the speeds at two successive sensors. When this difference was relatively high, speeds and videos were checked to determine whether the readings were reasonable. After many reviews of each file, statistical analyses were performed to compare data. For the study at the HEB driveway, the spacing between Sensors 3 and 4 was 25 ft, not the 15 ft used later on, so the recorded elapsed time was multiplied by 3/5. The values of the elapsed times at the January and July Stonegate and the March Union Plaza sites were proportionally adjusted to reflect the necessary

115 repositioning of certain sensor pairs. Due to the problems previously mentioned, data from the following sites and dates were not used: Stonegate September, Genie September, and Hollywood September. The data in each of the three grade groups (Flatter, Moderate, Steeper) were initially evaluated. Right turn data and left turn data were evaluated separately. Then, comparisons were made among the three grade groups. Initial Results, Observations, and Considerations During the course of the field data collection and the subsequent analysis, factors that could possibly affect driver behavior and the resulting speeds and elapsed travel times at specific sites were identified. At the driveway sites, the data collectors observed that even with a 19 ft radius and a 13 ft wide entry throat, it appeared from drivers’ facial expressions and driving behaviors that many turning right into the driveway felt constrained. The entering drivers’ seemed concerned with the proximity of their left front bumper with the left side of the blocking vehicle in the exit lane. Some entering drivers seem to slightly halt at this point during their turn maneuver. The analogous phenomenon for left turning vehicles entering the driveway was much less pronounced. Also, the observed directional patterns (predominately through or predominately turning) of vehicles immediately after entering the driveway, and the presence or absence of traffic conflicts in the driveway throat, may somewhat affect the speeds at Sensor 4. The following Exhibit 4-16, separately for each of the three grade groups, describes the directional traffic patterns and conflicts at each site. Exhibit 4-17 lists the three grade groups, with the number of readings and the average speed and elapsed time values. Values are reported separately for right- and for left-turning vehicles entering the driveways.

116 EXHIBIT 4-16 Driveway throat traffic patterns at study sites Most traffic entering the driveway turns. Drivers encounter traffic conflicts, with some sight distance limits. StonegateUnion Plaza Traffic entering the driveway must choose from multiple options. Drivers encounter random traffic conflict patterns. Genie Car Wash ? Most traffic entering the driveway proceeds straight. Occasional traffic conflicts are mainly from the right. Most traffic entering the driveway proceeds straight to a window or turns left to park. Few traffic conflicts. O. C. Credit Union Traffic entering the driveway must turn; most go to the right. Infrequent traffic conflicts. Traffic patterns are diverse. Many are to/from the ATM to the right. Traffic conflicts are common. Hollywood VideoMcAlister’s HEB Most traffic proceeds counterclockwise around the building. Occasional conflicts are mainly from the left. Most traffic proceeds straight into the parking lot. Few traffic conflicts. Most traffic entering the driveway turns. Traffic conflicts are common. ShellWendys Traffic entering the driveway turns left. There are no traffic conflict patterns. Red Robin J D China MODERATE FLATTER STEEPER LESS CONFLICT MORE CONFLICT Most traffic entering the driveway proceeds straight. Cross traffic is almost non-existent. Most traffic entering the driveway proceeds to drive-up tellers. Minimal traffic conflicts from the right. Arvest Bank

117 EXHIBIT 4-17 Speed and elapsed travel time from individual sites Rt 2 Rt 3 Rt 4 Elap 2-3 Elap 3-4 Lt 2 Lt 3 Lt 4 Elap 2-3 Elap 3-4 turn conflict 58 58 57 59 55 66 64 71 62 70 15.4 4.6 5.8 1.43 1.71 10.0 7.3 6.1 1.29 1.73 turn 78 85 78 79 78 conflict 14.7 5.3 5.4 1.51 1.67 thru 83 83 83 80 81 95 112 123 88 112 free 13.9 5.1 6.4 1.79 1.53 9.4 8.9 9.2 1.04 1.08 thru 48 51 49 42 47 free 9.5 9.7 8.2 1.09 1.09 number = 219 226 218 218 214 209 227 243 192 229 mixed 36 40 40 40 40 77 74 69 61 74 free 13.4 5.6 7.5 1.62 1.42 8.8 9.8 10.1 1.07 1.04 64 turn 16 86 88 84 87 88 60 67 72 72 73 free 14.8 5.7 7.4 1.65 1.37 9.5 11.4 11.4 1.12 0.86 thru 47 47 40 0 40 conflict 14.1 7.2 7.6 -- 1.20 turn 167 164 163 166 164 162 151 179 135 170 conflict 15.1 5.4 6.8 1.51 1.45 10.7 9.8 8.6 1.01 1.09 number = 336 339 327 293 332 299 292 320 268 317 turn 61 62 64 61 61 121 114 117 115 115 free 13.9 5.7 6.7 1.75 1.49 11.1 10.9 10.1 0.94 0.89 thru 19 24 24 18 23 42 42 40 39 42 free 12.7 5.0 7.0 1.40 1.48 8.4 8.9 9.1 1.14 1.07 turn 77 94 80 73 81 0 65 61 0 56 conflict 13.8 5.5 6.8 1.47 1.39 -- 10.7 10.1 -- 0.89 turn 84 85 83 81 82 18 20 20 17 18 free 14.8 5.5 7.9 1.41 1.34 9.2 10.5 11.1 0.99 0.90 number = 241 265 251 233 247 181 241 238 171 231 Total number of all = 796 830 796 744 793 689 760 801 631 777 13.5 19 19 4815.8%13.5 13.2% 4' / 3.4% 6' / 5.3% 8' 2312.8% 2' / 15.6%-4.2% -3.1% 18.6% 15.5% 6.5' / 0.3% 6' / 13.8% 48 12.0 11.2 12.6 Shell gas; self storage 11.5 11.5 Union Plaza - Mardells Arvest Bank 10.0% Stonegate One - Austin Pain Assoc. Genie Car Wash 11.0 11.0 17.0% -2.6%/ -7% 1' -1.2% 13.0 -2.1% 6.5% 4.4% 11' / 1.3% 40 15.5 4.7% 20' - 2.0% 6' 43 19 3.0% 9' - 0.8% 52 J D China 5.2% 10.1% 11.5 -1.8% 10.5% Hollywood Video - Southcross Plaza Wendys -0.5% 2.0% McAlisters, Meineke 6.4% 12' / 5.5% -3.8%/ -1%13.0 13.8% 29 6.0% 10' / 1.1% 5.1% 12.6% 29 18 41 19.5 13 19.5 for each site, the sample size on top, mean on bottom 1.6%-2.1% 3.7% 7.1% 6.5' / 2.2% 4.5' / 0% -2.0% Rt. Turn Entry Radius (ft) 58 16 17 Throat Entry Traffic Pattern Dway. Grades Grade Change Near Gutter Line small shop. center- HEB grocery Site Description Outer Lane Width (ft) See Note Street Cross Slope Throat Length (ft) 668.7% 10' / 12.8% NOTE: Outer lane width is measured from the lane line to the entry-radius tangent FLATTER MODERATE STEEPER Red Robin 11.5 -4.0% 5.0% Okla. Central Credit Union From observations in the field during data collection and from an analysis of the data, the following remarks are offered about factors that may have affected drivers’ speeds as they entered the driveways. Steeper Sites The higher level of throat traffic conflict may have contributed to lower speeds at Sensor 4 at Stonegate and at Genie. The lower level of conflict may have contributed to higher Sensor 4 speeds at

118 Union Plaza. Also at Union Plaza, the dip in the cross section at the gutter line may have caused drivers to proceed at lower speeds at Sensor 3 than they would have if the cross slope had remained constant. Moderate Sites At McAlister's, the flattened gutter pan may have caused left-turning drivers to proceed at higher speeds at Sensor 3 than they would have if the cross slope had remained constant. The data also suggest that the higher level of traffic conflict in the Hollywood throat caused speeds at both right- and left-turn Sensor 4 to be lower. Flatter Sites The pavement rutting in the outside lane at J. D. China may have caused left-turning drivers to proceed at lower speeds at Sensors 2 and 3 than they otherwise would have. The only Flatter site which seems to have been affected by the absence or presence of throat congestion was Red Robin, which had little congestion and faster speeds at Sensor 4. Shell was the only site at which it was thought that occasional traffic congestion in the outer through lane may have affected driveway flow. From time to time, vehicles that turned right into the next driveway downstream from the subject driveway were observed to begin decelerating in advance of the subject driveway.