Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 25
In addition to the AADTi, TrafLoad uses classification counts collected at Level 1A sites to derive or infer monthly traffic distribution factors to be used at all Level 1, 2, and 3A sites. Sim- ilarly, TrafLoad uses classification counts collected at Level 1A and 2A sites to derive hourly distribution factors (HDFs) for these sites and to infer HDFs for many other sites.1 The monthly and hourly distribution factors provide the Pavement Design Guide software with the ability to estimate how the pavement load varies by time of day and time of year. This information, in turn, is used by the software in analyzing the effects of diurnal and seasonally varying envi- ronmental factors that affect the pavement's susceptibility to damage. Since all estimates of AADTi for Level 1, 2, and 3A sites are developed by direction or lane, TrafLoad sets an accompanying set of directional distribution factors (DDFs) to 1.0, indicating to the Pavement Design Guide software that the AADTi represents traffic in one direction only. (If not provided with these DDFs, the Pavement Design Guide software would assume that the estimates represent two-way traffic, and an appropriate default DDF would be used.) Additional information about the traffic data that TrafLoad produces for Level 1, 2, and 3A classification sites is presented in Sections 3.33.5. Level 3B Classification Sites The Pavement Design Guide software requires, and TrafLoad produces, just two pieces of information for Level 3B sites: · Total (two-way) annual average daily truck traffic and · The "Truck Traffic Classification" group to which the site is assigned. This information is discussed further in the second part of Section 3.5. 3.3 Level 1 Classification Sites Level 1A sites are sites at which continuously operating AVCs have been used to collect a min- imum of 1 week of classification counts for 12 consecutive months. The first two subsections below discuss these data and the several uses that are made of them. Level 1B sites are classi- fication sites that are on the same road as a Level 1A site and reasonably near that site; these sites are discussed in the third subsection below. Continuous Classification Counts The principal goal of the continuous classification-count program is the creation of factors needed to estimate annual average daily truck volumes from short-duration classification 1 For sites for which HDFs are not provided to the Pavement Design Guide software, the software uses its own set of HDF default values. 2-25
OCR for page 26
counts. The same information is also used to estimate seasonal fluctuations in truck volumes so that these changes can be accounted for in the design process. To accomplish this goal, it is necessary to measure day-of-week and seasonal variation in truck traffic and to develop factors that can be applied to short-duration counts. As illustrated by Figures 3.1 and 3.2, truck volumes vary significantly by time of day (TOD) and day of week (DOW), and different patterns exist for local (predominantly business-day) trucks and for long-distance trucks. A sufficient number of continuous-count locations are needed to measure each of the differ- ent truck volume patterns found in a state or region. This means that continuous counters should be placed on different functional classes of roads and in different geographic locations within each state. It is especially important to be able to measure the differences in truck vol- ume patterns between roads that carry primarily local truck traffic and those that serve through traffic. A good rule of thumb is that the continuous classification-count program should be roughly the same size as the traditional continuous volume count program. (The latter program, con- ducted with automatic traffic recorders, is frequently called the ATR program.) In fact, the design of the continuous classification-count program is very similar to the design of the ATR Figure 3.1 Typical TOD Patterns Percent of Daily Traffic 9 Through Trucks Business-Day Trucks 8 Rural Cars Urban Cars 7 6 5 4 . 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour of Day 2-26
OCR for page 27
Figure 3.2 Typical DOW Patterns Daily Traffic Ratio 1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 Recreational Car 0.5 Traditional Car Business-Day Truck 0.4 Through Truck 0.3 Sunday Monday Tuesday Wednesday Thursday Friday Saturday Day of Week program. While the recommended continuous-count program requires a significant number of count locations, it is important to note that continuous classifiers also serve as ATRs. Thus, it is possible to use the classification counters in place of ATRs at the same time as they are used to supply continuous classification data. Such a step significantly reduces the number of contin- uous counters an agency needs and reduces unnecessary duplication of the data-collection effort. Permanent, continuous counts also provide an excellent source of information on truck vol- ume trends. In particular, all highway agencies should monitor the total volume of heavy trucks. The trend in truck volume should be examined both for each individual roadway on which a permanent data-collection device is located and for each of the geographic areas in the state. (Truck traffic tends to vary with both the economic activity taking place in a geo- graphic region and the amount of through traffic passing through the area.) Also of interest are changes in the mix of trucks. Changes in truck size and weight laws can have significant effects on the total number (and percentage of) large trucks of specific designs. FHWA's Traffic Monitoring Guide provides wide latitude in the selection of locations where permanent classifiers are placed. For the purposes of both general monitoring and pavement design, permanent classifiers should be placed on a variety of roads throughout the state. Thus, some classifiers should be on Interstate highways and other major routes that carry 2-27
OCR for page 28
heavy through-traffic volumes. Others should be placed so that trucking patterns specific to within-state movements of freight can be monitored. Lastly, where possible, urban locations should also be monitored, so that urban truck volumes can be measured. Uses of Data from Continuous Classification-Count Sites The highest quality estimates of AADTi are those that are developed using 12 months of data from a continuous classification-count site. Such a site is referred to as a Level 1A classifica- tion site. For such a site, estimates of AADTi, monthly traffic distribution factors by VC, and hourly distribution factors are developed entirely from the classification counts obtained for the site. In addition, classification counts obtained at these sites are used for developing several types of traffic ratio.2 Two of these types of traffic ratio are monthly and DOW traffic ratios that are used for seasonal and DOW factoring of short-duration classification counts obtained at Level 2 classification sites. The use of separate sets of monthly and DOW traffic ratios for this purpose makes it possible to adopt independent definitions of the seasonal and DOW factor groups that are used for this purpose. The development and use of the seasonal and DOW fac- tor groups are discussed in the first and third subsections below, and a related concept, VC groups, is discussed in the second subsection. Monthly traffic ratios that are used for adjusting any short-duration classification count should (if pos- sible) be current year traffic ratios; that is, they should be developed from data that are collected over a 12-month period that includes the months during which the short-duration counts are collected. The resulting estimates of AADTi generally will be better estimates of AADTi for that 12-month period than estimates developed using traffic ratios developed using data from earlier 12-month periods. In particular, current-year traffic ratios will provide a better adjust- ment for any unusual conditions affecting truck volumes at the time that a short-duration count is collected. (Such conditions include unusual weather, a poor harvest, or the beginning of a sharp recession or of an economic recovery.) A third type of traffic ratio developed from classification counts obtained at Level 1A classifi- cation sites is TOD traffic ratios. These traffic ratios are combined with partial-day classification counts obtained at manual-count sites to estimate 24-hour traffic volumes, by VC, at these sites. The development of TOD factor groups is discussed in the fourth subsection below. A fourth type of traffic ratio developed from classification counts obtained at Level 1A classi- fication sites is applied to data from Level 1B classification sites and is described in a subse- quent section discussing such sites. 2 TrafLoad users should think of the "traffic ratios" used by TrafLoad as "factors." The technical dis- tinction between "traffic ratios" and "factors" is presented in Section 4.1 of Part 1, along with an expla- nation of the advantage of using traffic ratios. 2-28
OCR for page 29
DOW Factor Groups As observed earlier (see Figure 3.2), "through trucks" and "business-day trucks" have very different DOW volume patterns. The volume of through trucks varies only slightly from day to day, while the volume of business-day trucks drops substantially on Saturdays and Sun- days. For purposes of the present discussion, we shall replace the term "business-day" with "business-week" to emphasize that our interest is in the drop-off in activity during the week- end (and not the drop-off at night). The volume pattern shown in Figure 3.2 for business-week trucks provides a good example of the importance of DOW factoring. For a site where nearly all trucks are business-week trucks, estimates of AADTi derived from weekday classification counts without DOW factor- ing will tend to be overestimates. (The plot in Figure 3.2 indicates that, on average, the over- estimates will be in the 15- to 25-percent range.) DOW factoring is designed to correct for the overestimates that are reflected in weekday classification counts. In order to use DOW factoring effectively, it is necessary to distinguish VCs and sites where business-week trucking predominates from VCs and sites where through trucking predomi- nates. Since nearly all single-unit trucks are used primarily for local service, business-week trucking is likely to be dominant for FHWA Classes 57 at nearly all sites. In the case of buses and combination trucks (Classes 4 and 813), the situation is more com- plex. At sites on the Interstate system that are more than 200 miles from a major urban area, most vehicles in these classes are likely to be through vehicles. On the other hand, in major urban areas, vehicles in these classes are more likely to be providing service of a more local nature, especially on roads that carry little or no through traffic. In these areas, the volume of vehicles in these classes is likely to be appreciably lower on weekends than on weekdays. A schematic summary of the above observations is presented in Table 3.1. The table indicates that Class 57 vehicles are likely to exhibit a business-week volume pattern at nearly all sites at which significant numbers of these vehicles operate, while other truck and bus classes are likely to exhibit a business-week pattern at some sites and a through pattern at other sites. For this Table 3.1 Commonly Observed DOW Volume Patterns, by VC VCs DOW Volume Pattern 5-7 4 and 8 - 13 Business-Week Pattern X X Through Pattern - X Key: X Pattern is likely to exist at many sites. Pattern occurs only under unusual circumstances. 2-29
OCR for page 30
reason, when developing DOW factor groups, attention should be focused on the bus and combination-truck classes, particularly on the most important of these classes (usually Class 9). The general approach to developing DOW factor groups is to start by distinguishing two or three such groups: · One consisting of sites at which buses and combination trucks exhibit a business-week pattern; · One consisting of sites at which they exhibit a through pattern; and · Perhaps, one consisting of sites at which they exhibit an intermediate pattern. For the purpose of developing DOW traffic ratios for each of these factor groups, there should be a minimum of three continuous classification-count sites in each group, with a larger num- ber (five to eight) used wherever possible. For many states, the two or three DOW factor groups described above will suffice. However, states with large numbers of continuous classification-count sites may wish to consider estab- lishing additional DOW factor groups. One possibility that might be considered is increasing the number of groups corresponding to intermediate "through"/"business-week" patterns. Another possibility for increasing the number of factor groups involves identifying and dis- tinguishing different DOW patterns for through trucks. In particular, the dip in through-truck volume that, in Figure 3.2, is shown as occurring on Monday is actually affected by distance from the trucks' origins and destinations. As a result, these DOW patterns may vary by road orientation (North-South versus East-West) or by direction of travel. These differences are likely to be most significant in the mountain and western plains states, where many trucks are traveling to and from relatively distant origins and destinations. VC Groups In concept, it would be desirable to develop separate sets of DOW traffic ratios for each VC. However, attempting to do so may produce zero values for some DOW traffic ratios for uncommon classes, resulting in division by zero when traffic counts are subsequently divided by these ratios. To avoid this problem, a set of user-defined vehicle-class (VC) groups are established. In TrafLoad, VC groups used in the factoring of classification counts are called "Type 1 VC groups" to distinguish them from the WIM VC groups discussed in the preceding chapter. The Type 1 VC groups are also used in the development and application of seasonal and TOD traf- fic ratios. When defining Type 1 VC groups, a general rule is that VCs that have appreciably different DOW or seasonal volume patterns should be assigned to separate VC groups. If TOD factor- ing is to be used, this rule also applies to TOD volume patterns. As observed previously, FHWA VCs 57 tend to exhibit business-week and business-day volume patterns that are 2-30
OCR for page 31
stronger than those exhibited by the other VCs, suggesting that these three VCs generally should be assigned to a separate Type 1 VC group from the other VCs. When first setting up Type 1 VC groups, two VC groups may be found to be sufficient. How- ever, TrafLoad allows users to define a larger number of VC groups, and some users may wish to take advantage of this capability to divide the VC groups further. The one limitation in this process is that VCs that are rarely used should be assigned to VC groups that include one or more VCs that are frequently used. If this is not done, there is a small possibility that the TrafLoad factoring procedure will be forced to terminate abnormally in order to avoid divid- ing by zero.3 Seasonal Factor Groups As discussed above, the DOW factor groups should be designed to group sites with similar DOW patterns of truck volume. Similarly, the seasonal factor groups should be designed to group sites with similar seasonal (or month-of-year) patterns of truck volume. Toward this end, the research team makes several observations about seasonal variations in truck volumes: · As in the case of automobiles, seasonal variations in truck volumes tend to be weaker in urban areas than in rural areas. · In many areas, the highest truck volumes occur during the MayOctober period, and the lowest volumes occur in January. · Local influences (commodities carried, harvest season, etc.) can produce substantial site- to-site variation in the timing and intensity of the seasonal peak in truck volumes on rural non-Interstate roads. · The greater diversity of trucks using the Interstate system mutes the effects of local influ- ences on seasonal variations in truck volumes, producing more consistent seasonal patterns. The above observations suggest that, for many states, the development of seasonal factor groups might begin with the creation of an urban group and a rural Interstate group. Some consideration might also be given to creating a third group consisting of sites whose seasonal variations are in between those of the first two groups. This group might include urban IS sites with relatively high volumes of through trucks. 3 As an example, consider a DOW factor group that contains no Level 1A site at which any vehicle in a specific VC group was observed on a Sunday. In this case, the Sunday traffic ratio for this DOW fac- tor group and this VC group will be zero. Assume that a 7-day classification count is obtained for a Level 2 site that has been assigned to this DOW factor group and that one or more vehicles in that VC group are observed on Sunday at this site. Then TrafLoad will be unable to factor the Sunday count for this VC group at this site. Similar problems can also be constructed for seasonal and TOD factoring, and they are even more likely to occur if combined monthly/DOW traffic ratios are used for factoring counts obtained at Level 1B sites. 2-31
OCR for page 32
The remaining issue is how to develop seasonal factors to be applied to classification counts obtained at rural Level 2 sites that are not on the Interstate system. A simple alternative is to create a single rural non-Interstate factor group for this purpose.4 Issues relating to the devel- opment of seasonal factor groups for classification counting are discussed further in the Traf- fic Monitoring Guide (pp. 4-22 through 4-32). As in the case of DOW factors, TrafLoad develops separate sets of seasonal factors for each seasonal factor group and each Type 1 VC group. TOD Factor Groups TrafLoad uses TOD factoring to convert partial-day classification counts (collected at Level 2B sites) to estimates of 24-hour traffic volumes by VC. As in the case of seasonal and DOW traf- fic ratios, TrafLoad uses classification counts from Level 1A sites to develop several sets of TOD traffic ratios. In particular, for each of the user-defined Type 1 VC groups, TrafLoad develops a separate set of 24 TOD traffic ratios (or "hourly fractions") for each user-defined TOD factor group. Since partial-day classification counts are almost always collected on a weekday, only weekday classification counts are used in developing TOD traffic ratios. At a minimum, the TOD factor groups should be designed to distinguish sites at which business- day trucking predominates from sites at which through trucking predominates. Additional TOD factor groups may also be created to represent intermediate situations and/or more extreme cases of business-day or through-trucking patterns. It is likely that the TOD factor groups frequently will be identical to the DOW factor groups (discussed earlier), but the soft- ware allows the user to identify differences where appropriate. (For example, sites on a road that is used primarily to access a truck terminal or warehouse that operate 24 hours per day, 5 days per week, might be treated as "through-trucking" sites for the purpose of TOD factor- ing but not for the purpose of DOW factoring.) FHWA VCs 57 almost always exhibit a business-day volume pattern (just as they almost always exhibit a business-week volume pattern). For this reason, when developing TOD fac- tor groups, attention should be focused on the bus and combination-truck classes, just as in the case of DOW factor groups. TOD factoring is performed only if there are sites at which partial-day classification counts are collected. If no such sites exist, it is not necessary to define TOD factor groups. Level 1B Sites Consider a classification site that is not a Level 1A site but that is on the same road as a Level 1A site. If it is believed that the two sites are sufficiently close that most trucks that pass 4 A more ambitious alternative would be to examine the seasonal patterns of all Level 1A rural non- Interstate sites, and, on the basis of this review, create two or more separate rural non-Interstate fac- tor groups. If this alternative is adopted, it will then be necessary to determine how to assign rural non-Interstate Level 2 sites to factor groups. However, if this assignment is performed well, the result- ing estimates of AADTi for these sites are likely to be better than those that would result from using a single rural non-Interstate factor group. 2-32
OCR for page 33
one of the sites pass both sites, then the site in question qualifies as a Level 1B site that is "asso- ciated" with the Level 1A site, and this information should be provided to the TrafLoad soft- ware. TrafLoad is capable of producing high-quality AADTi estimates for Level 1B sites, with the quality of these estimates depending on the similarity of the truck traffic at the two sites. TrafLoad has two procedures for producing AADTi estimates for Level 1B sites. The choice as to which procedure to use for a particular site is made automatically by TrafLoad. Brief, somewhat technical descriptions of the two procedures and how the choice is made are pre- sented below. The monthly traffic distribution factors and hourly distribution factors for any Level 1B site are assumed to be the same as those for the associated Level 1A site. Direct Scaling The simpler of the two procedures for estimating AADTi at a Level 1B site is "direct scaling." TrafLoad uses direct scaling whenever a) Classification counts at a Level 1A site associated with a Level 1B site have been obtained for the same hours and dates as the classification counts that were obtained at the Level 1B site and b) Both sites have no more than one lane in each direction. Under these circumstances, the ratios of the counts at the two sites are used to scale the AADTi at the Level 1A site to produce estimates of the AADTi at the Level 1B site.5 Separate scale fac- tors are used for each Type 1 VC group. Factored Counts The second procedure for estimating AADTi at a Level 1B site is a factoring procedure. For this purpose, for each Level 1A site, a set of combined monthly/DOW traffic ratios is developed. For each Level 1A site, each direction, and each Type 1 VC group, 84 such ratios are devel- oped, corresponding to all combinations of the 12 months and 7 days of the week.6 Each of these ratios is developed by TrafLoad by obtaining monthly average day-of-week traffic (MADW) for a given direction and VC group and dividing by AADT for that direction and VC 5 If the user has requested that AADTi be estimated by direction (rather than by lane), then direct scal- ing would be appropriate even if Condition (b) does not hold. However, the current version of TrafLoad does not perform direct scaling in this case. 6 The use of 84 combined monthly/DOW traffic ratios allows the factoring procedure to reflect the com- bination of monthly and DOW variations in volume better than can be done with separate monthly and DOW traffic ratios (12 monthly ratios and 7 DOW ratios). However, combined traffic ratios can- not be used in conjunction with monthly and DOW factor groups that are developed independently of each other. Hence, TrafLoad uses combined traffic ratios for factoring counts from Level 1B sites and separate monthly and DOW traffic ratios, developed using data from groups of Level 1A sites, for factoring counts from Level 2 sites. 2-33