90 TRANSPORTATION RESEARCH RECORD 1128 Corps of Engineers Lo-Volume Road Design JOHN C. POTTER, RAYMOND S. ROLLINGS, AND WALTER R. BARKER The U.S. Army Corps of Engineers' pavement design proedures are partiularly appropriate for lo-volume road appliations beause they ere developed from traffi tests using relatively lo traffi volumes and thin pavement setions on lo-strength subgrades. The large loads and lo volumes of traffi used in their development are espeially analogous to situations enountered In mining, logging, and similar Industrial appliations, or in port failities. The flexible pavement riteria are most appropriate for thin asphalt onrete pavements in granular base ourses and subbases. Generally, rigid pavements have not been assoiated ith lo-volume roads beause of their ost. Hoever, development of roller-ompated onrete pavement onstrution has made rigid lovolume pavements feasible in many situations, and the orps design method is apable of addressing :roller-ompated onrete pavement harateristis. These design proedures, published in Army tehnial manuals, have also been omputerized. The Corps of Engineers' design riteria an no be applied quikly and effiiently, even by pavement engineers unfamiliar ith the orps riteria or programs, to design eonomial pavements for military installations and other lovolume failities. The U.S. Army Corps of Engineers design proedures for roads and streets have been developed for appliation to the lo traffi volumes generally assoiated ith military installations. A entral faet of the design riteria for both rigid and flexible pavements is a rational onsideration of the strutural requirements for pavements expeted to experiene limited traffi over their design life. These design proedures have long been desribed in the various Army tehnial manuals. No they have also been omputerized. The omputerized versions of the tehnial manuals allo faster and more error-free design. A ider range of strutural design alternatives also beomes pratial. SCOPE The design proedures for rigid and flexible pavements is the primary fous of this paper. These proedures ere developed from traffi tests using relatively lo traffi volumes and thin pavement setions on lo-strength subgrades. The large loads and lo volumes of traffi used in their development are espeially analogous to situations enountered in mining, logging, and similar industrial appliations, or in port failities. The flexible pavement riteria are most appropriate for thin asphalt onrete pavements on granular base ourses and subbases. Generally, rigid pavements have not been assoiated ith lo-volume roads beause of their ost. Hoever, development of roller-ompated onrete pavement onstrution has made rigid lo-volume pavements feasible in many situations, and the orps design method is apable of addressing roller-ompated onrete pavement harateristis. Unsurfaed or aggregate-surfaed roads subjet to heavy loads generally require ontinuous maintenane, even for lo traffi volumes. The orps riteria for these roads are also less sophistiated than for rigid and flexible pavements and have not been omputerized. ROADS AND STREETS A major element in any pavement design is the haraterization of the antiipated traffi and seletion of a design traffi loading. This task has been simplified in the orps riteria by use of a design index. The design index is a measure of the total lifetime traffi for the pavement in terms of equivalent 18-kip axle loads. Eah design index orresponds to a range of equivalent 18-kip axle loads suh as those given in Table 1. Pavements subjeted to loads from vehiles suh as forklift truks or heavy airraft, hih are signifiantly greater than the 18-kip axle load, are designed separately. The use of 18-kip axle equivalents is not suffiiently aurate for these loads. TABLE 1 FLEXIBLE PAVEMENT DESIGN INDEX AND RANGE OF EQUIVALENT 18-KIP AXLE COVERAGES Flexible Range of Equivalent Coverages Pavement (DI) Minimum Maximum 1 0 2,400 2 2,400 10,500 3 10,500 47,000 4 47,000 207,000 5 207,000 908,000 6 908,000 4,000,000 7 4,000,000 18,000,000 8 18,000,000 79,500,000 9 79,500,000 355,000,000 10 355,000,000 1,600,000,000 Norn: DI is design index. U.S. Army Corps of Engineers Waterays Experiment Station, P.O. Box 631, Viksburg, Miss. 39180-0631. The design index is determined in the same manner for both rigid and flexible pavements, exept that the ranges of
Potter et al. equivalent 18-kip axle overages for eah design index are different. (Here, a overage is defined as a suffiient number of movements or passes of the design vehile to over the entire traffi lane ith at least one stress repetition.) Traffi typial of a partiular traffi ategory and faility lass has been onverted to equivalent 18-kip axle loads so that the appropriate design index an be seleted knoing only the type of traffi (ategory) and roaday lass. Additional guidane is provided, if needed, for seleting the ategory and lass for a partiular design. The rural "road" or urban "street" lass is a funtion of the design hourly volume (DHV), for the 25-year servie life, as indiated in Table 2. The DHV is assumed to be 15 perent of the average daily traffi (ADT) for rural roads and 12 perent of the ADT for streets in built-up areas. Various geometri design onstraints, appropriate for the traffi intensity, also apply to eah lass. Parking area design is based on Class E traffi. TABLE 2 DESIGN HOURLY VOLUME AND ROAD/STREET CLASS DHV Class Road Street A 900 1,200 B 720-899 1,000-1,199 450-719 750-999 D 150-449 250-749 E 10-149 25-249 F < 10 < 25 The traffi ategory is governed by the number per day and gross eight of traked vehiles, if any. Otherise, the traffi ategory is based on the perentages of 2- and 3-axle truks in the traffi mix. The desriptions for eah ategory, and the design index orresponding to those ategories for eah lass, are given in Table 3. Thikness adjustment for traffi level has long been a part of Corps of Engineers pavement design. Both the rigid and flexible pavement design proedures are based on extensive fullsale, aelerated traffi testing to failure in a ide range of limates using a ide range of pavement strutures and load onfigurations. These proedures have been highly refined for pavements subjet to heavy airraft loads and have been transferred to the design proedures for more lightly loaded pavements. These adjustments are based on atual field performane rather than laboratory tests. Environmental effets suh as joint behavior, temperature and moisture yles, and gradients and material variability, not normally aounted for by laboratory-determined design values, are impliitly inorporated. Flexible Pavements The flexible pavement design proedure is based on the California bearing ratio (CBR) equation and uses a thikness adjustment fator alled the load-repetitions fator. Figure 1 shos the load-repetitions fator, a, as a funtion of 18-kip axle overages. The overage level orresponding to the middle of the range of equivalent 18-kip axle loads for a partiular design index (Table 1) is used to determine the appropriate a-fator. Rigid Pavements The Corps of Engineers' rigid pavement design method is based on a fatigue analysis. Tensile stress for the design load is alulated by using the Westergaard edge-loaded analytial model. When traffi loads ill not atually "traffi" the free edge of a slab suh as in parking areas or on airfields, these alulated stresses are redued 25 perent to aount for load transfer beteen adjaent slabs. Tenty five perent as seleted as an appropriate design value from model and full-sale tests on pavements by using load transfer devies that met orps riteria. Pavements suh as those on most roads and streets arry traffi loads adjaent to the outside free edge, and no redution in stress is alloed for these. The Corps of Engineers defines design fator as the modulus of rupture of onrete divided by the design tensile stress. Fullsale aelerated traffi tests ere used to develop the fatigue relation beteen design fator and overages shon in Figure 2. The failure riteria used in these tests, hih is the basis of the Corps of Engineers fatigue relationship, is the development of one or more strutural raks in one-half or more of the traffiked slabs. This analysis approah is used to selet a pavement thikness to resist the fatigue effets of the desired design repetitions of traffi load. The design fator orresponding to a given design index is hosen as ith the a-fator, but using a overage level orresponding to the midrange of the values given in Table 4. Additional adjustments are used in the design proedure for speial ases suh as subgrade pumping protetion, high strength subgrades, frost ation, and so forth. Roller-ompated onrete is of partiular interest for use on lo-volume roads. The eonomis of using it are in the onstrution method rather than in redued materials ost, and it has the potential of providing redued-ost pavements at the expense of inreased roughness and poorer appearane. To date there is no indiation that the fatigue properties of rollerompated onrete are any different from onventionally plaed portland ement onrete. Consequently, the regular Corps of Engineers' rigid pavement design proedures an be used to design roller-ompated onrete ith one restrition. The urrent methods of onstrution for roller-ompated onrete do not provide reliable load transfer aross joints beteen slabs. Therefore urrent design aids, suh as those published in tehnial manuals or the miroomputer programs, ould provide erroneous results for a ase in hih roller-ompated onrete is used for a parking area unless the user as aare of the adjustments needed. Computerization Rigid pavement design for roads and streets (inluding vehiular parking areas) is overed in Rigid Pavements for Roads, Streets, Walks, and Open Storage Areas (1). Flexible pavements are addressed in Flexible Pavements for Roads, Streets, Walks, and Open Storage Areas (2). A omplete design also 91
92 TRANSPORTATION RESEARCH RECORD 1128 TABLE 3 PAVEMENT DESIGN INDEX Pavement Design Index for Road or Street Classifiation Traffi Categorl A B D E F I (< 1% to-axle truks) 2 2 2 l l l II (S 10% 2-axle, no 3-axle truks) 3 2 2 2 2 l III (S 15% truks, S 1% 3-axle truks) 4 4 4 3 3 2 IV (S 25% truks, S 10% 3-axle truks) 5 5 5 4 4 3 IVA ( 25% truks or 10% 3-axle truks) 6 6 6 5 5 4 v (60-kilopound (kip) trak-laying vehiles): 500/day 7 7 7 7 7 200/day 6 6 6 6 6 100/day 6 6 6 6 6 6 40/day 6 6 6 5 5 5 10/day 5 5 5 5 5 5 4/day 5 5 5 5 4 4 l/dsy 5 5 5 4 4 4 VI (90-kip trak-laying vehiles): 200/day 9 9 9 9 9 100/day 8 8 8 8 8 8 40/day 7 7 7 7 7 7 10/day 6 6 6 6 6 6 4/day 6 6 6 6 6 6 l/day 5 5 5 5 5 5 I/eek 5 5 5 4 4 4 VII (120-kip trak-laying vehiles): 100/day IO 10 IO IO IO IO 40/day 9 9 9 9 9 9 IO/day 8 8 8 8 8 8 4/dsy 7 7 7 7 7 7 I/day 6 6 6 6 6 6 I/eek 5 5 5 5 5 5 Traffi limited to IOO vehiles per day. TABLE 4 RIGID PAVEMENT DESIGN INDEX AND RANGE OF EQUIVALENT 18-KIP AXLE COVERAGES Rigid Pavement Design Index 1 2 3 4 5 6 7 8 9 10 Range of Equivalent Coverages Minimum Maximum 1 45 600 13,000 130,000 800,000 3,500,000 14,000,000 40,000,000 110,000,000 45 600 13,000 130,000 800,000 3,500,000 14,000,000 40,000,000 110,000,000 300,000,000 requires use of appropriate material from the tehnial manuals Pavement Design for Seasonal Frost Conditions (3) and General Provisions and Geometri Design for Roads, Streets, Walks, and Open Storage Areas. Having a orking knoledge and understanding of the intriaies of these proedures are formidable tasks for a ne orps employee or onsultant designing a pavement for the Corps of Engineers. Errors of interpretation or appliation of the riteria or omissions of valuable options and onsiderations are easily ommitted. The srupulously preise logi of a omputer program offers the potential for reduing errors of this kind. The inreased omputational speed makes onsideration of a large number of potential pavement systems feasible. The Corps of Engineers no has a series of omputer programs for miroomputers that onstitutes a high speed, friendly, interative alternative to the tehnial manuals. Rigid pavement design for roads is overed by the RRD omputer program, and flexible pavement design for roads is overed by the FRD program. The to omputer programs ere developed speifially to operate on IBM PC's or ompatible miroomputers using FORTRAN 77. Eah of the programs has been uploaded to mainframe omputers ith only minimal effort. The omputer programs ere ritten to follo the design manual in a logial stepise proedure. Questions are asked that must be ansered so that all aspets of the design must be onsidered by the designer. In areas in hih questions are not ompletely lear, aompanying information is provided. Also, hen neessary, guidane in the seletion of design data ill be provided. The omputer programs provide for a data revie and a method of hanging any design data before exeution of the partiular design problem. The programs determine pavement thiknesses and other pertinent design information that is onsistent ith the design manuals. The produt from the program is first output to the sreen and a disk file. The operator may then diret the output from the disk file to the printer.
1.00!:! 0.80 t; u. Cl) j:: 0.80 j:: 3; a: 0.40 g 0.20 o L_..._........uu...&JL-..._..._.L----'L-_.._...LIJ'--L-_._.L...L_._.10".ooo---''--..._... 100.000 10 100 1,000 COVERAGES FIGURE 1 Load repetitions fator versus overages. gr Cl) i Cl z a: t; a: 0 t u. z Cl ;;; 2.0 1.5 1.0 1,000 10,000 100,000 FIGURE 2 Design fator versus overage. COVERAGES OR LOAD CYCLES 1,000,000 10,000,000
94 CONCLUSIONS The Corps of Engineers design riteria are appropriate for lo-. volume roads, espeially if subjeted to large heel loads. The riteria an be used to design both flexible and rigid pavements, inluding roller-ompated onrete onstrution. Although the orps riteria are relatively omplex, they have been inorporated into flexible and rigid pavement omputer programs. These programs an be used on the miroomputer at the design engineer's desk to quikly design a seletion of alternative pavement setions from hih to hoose the most eonomial setion, onfident that the Corps of Engineers design riteria are satisfied by eah of the alternatives. ACKNOWLEDGMEN'.f The authors are grateful to the U.S. Army Engineers Waterays Experiment Station and the Offie of the Chief of Engineers for administrative support and permission to publish TRANSPORTATION RESEARCH RECORD 1128 this paper. The vies of the authors do not purport to reflet the position of the U.S. Department of the Army or the U.S. Department of Defense. REFERENCES 1. Rigid Pavements for Roads, Streets, Walks, and Open Storage Areas. Tehnial Manual 5-822-6. Department of the Anny. U.S. Government Printing Offie, Washington, D.C., April 1977. 2. Flexible PavemenJs for Roads, Streets, Walks, and Open Storage Areas. Tehnial Manual 5-882-5. U.S. Government Printing Offie, Washington, D.C., Ot. 1980. 3. Pavement Design for Seasonal Frost Conditions. Tehnial Manual 5-818-2. U.S. Government Printing Offie, Washington, D.C., Jan. 1985. 4. General Provisions and Geometri Design for Roads, Streets, Walks, and Open Storage Areas. Tehnial Manual 5-822-2. U.S. Government Printing Offie, Washington, D.C., April 1977. Publiation of this paper sponsored by Committee on on Lo-Volume Roads.