SAFETY TREATMENT OF ROADSIDE DRAINAGE STRUCTURES. Hayes E. Ross, Jr. Dean Sicking T. J. Hirsch

Transcription

1 SAFETY TREATMENT OF ROADSDE DRANAGE STRUCTURES by, " Hayes E. Ross, Jr. Dean Sckng T. J. Hrsch Texas Transportaton nsttute Harold D. Cooner John F. Nxon Samuel V. Fox Texas State Department of Hghways and Publc Transportaton C. P. Damon Federal Hghway Admnstraton Submtted to TRB Commttee A2A04 for revew and possble presentaton and publcaton by TRB August, 1981

2 H. E. Ross, et al 1 ABSTRACT The purpose of the research was to develop traffc-safe end treatments, for (1) cross-dranage structures and (2) parallel-dranage structures that would not apprecably restrct water flow. Cross-dranage culverts are used to convey water under the hghway. Parallel-dranage culverts are used to convey water under drveways, sde roads, ramps, or medan crossovers that abut the hghway. Prelmnary desgns were frst evaluated by computer smulaton, by use of a test pt n whch the clear open space and grate spacng could be vared, and by use of an earth berm smlar n geometry to a drveway. Promsng desgns were then subjected to full-scale prototype testng wth both subcompact and full-sze automobles. Traffc-safe culvert end treatments can be acheved as follows: Cross-dranage structures - (a) All culvert ends should be made to match the exstng sde slope wth no protruson n excess of 4 n. (10.2 cm) above grade; (b) culverts wth clear openngs 30 n. (76.2 cm) or less need no safety treatment other than as mentoned n (a); (c) culverts wth clear openngs greater than 30 n. (76.2 cm) can be made traffc-safe by grate members placed on 30 n. (76.2 cm) centers orented parallel to the flow and n the plane of the sde slope. Parallel-dranage structures - (a) The roadway sde slope (or dtch slope) shoul d be 6 to 1 or fl atter n the v c n ty of the dr veway; (b) the dr veway slope should be 6 to 1 or flatter; (c) the transton between the sde slope and the drveway slope should be rounded; and (d) safety treatment of the culvert openng should nclude an end secton cut to match the drveway slope wth cross members (grates) spaced every 24 n. (61.0 cm) perpendcular to the drecton of flow.

3 H. 1:. KOSS, et al 2 Applcaton of these fndngs wll result n mproved safety for the motorst. They wll also result n more hydraulcally effcent and more economcal safety treatments than have been used n the past. o 4, NTROD U cn ON n desgnng dranage culverts, the prmary objectve s to properly accommodate su rface runoff along the hghway r ght-of-way. However, a second mportant goal should be to provde a traffc-safe desgn that would be traversable by an out-of-control vehcle wthout rollover or abrupt change n speed. Gudelnes for desgnng traffc-safe grates have been very lmted. NCHRP publshed gudelnes for traffc-safe dranage structures n 1969 UJ. The recommendat ons dealt prma r ly wth the geometry of adjo n ng slopes. Computer smulatons have also been used to futher nvestgate the dynamc behavor of automobles traversng varous slope and dtch confguratons near drveways and medan crossovers (1,1). Crtera for the structural desgn of nlet grates was publshed n 1973 (1). However, the study dd not address the problem of grate desgn as related to safety. Recent feld revews of dranage culverts n Texas revealed that mprovements and some modfcaton of desgn detals could mprove both dranage and safety (). Many of the safety grates used n the past to cover the open ends of culverts have small openngs and the grates are easly clogged... :, wth debrs, causng water to back up and flow over the roadway, the dtch crossng, or adjacent property. n some cases safety grates do not possess enough strength to be effectve or they are used on small ppe culverts whch need no safety treatment.

4 H. E. Ross, et al 3 The objectve of ths study was to develop gudelnes for safety treatment of both cross-dranage and parallel-dranage structures that (1) can be j 1 safely traversed by an errant vehcle and (2) wll exhbt desrable hydrualc behavor. Although no hydraulc analyses were made t was assumed that hydraulc effcency ncreases as the number of grate members decrease. t was therefore a goal of the research to meet safety requrements wth as few grate members as possble. Ths paper summarzes the fndngs of two research studes, one conducted n 1979 () and the other n 1980 (2). Reference should be made to the, cted lterature for complete detals of the studes. EVALUATON CRTERA, A revew of the lterature showed that there are no natonally recognzed safety performance standards for roadsde dranage structures. Deceleraton and stablty of a vehcle durng and followng mpact are the two prmary measures of performance for safety appurtenances such as guardral s, crash cushons, etc. (). For the cross-dranage structures, performance was judged satsfactory f the vehcle smoothly traversed the culvert and the adjonng dtch slope wthout rollover for speeds from 20 mph (32.2 km/h) through 60 mph (96.5 km/h). Prevous research ( &) ndcated that a very flat dtch slope, a very.' flat drveway slope, and a very long culvert would be necessary to satsfy the above crtera for parallel-dranage structures. n vew of the economc and hydraul c mp 1 cat ons of such a des gn t was conc 1 uded that tradeoffs would be necessary to acheve an acceptable balance between the controllng elements. Performance of parallel-dranage structures was therefore judged acceptable f the vehcle smoothly traversed the adjonng slopes and culvert

5 H. E. Ross, et al 4 wthout rollover for speeds from 20 mph (32.2 km/h) through 50 mph (96.5 km/h) RESEARCH APPROACH A three-phase approach was taken n the development of safety treatments of both cross-dranage and parallel-dranage structures. n the frst two phases computer smulatons n combnaton wth a prelmnary test program were used to develop tentatve desgn concepts. n the latter phase prototypes were constructed usng the rl'sults of the prelmnary studes and tested under representatve roadsde confguratons. Cross-Dranage Structures Smulaton studes - A computer smulaton study was conducted, usng,,, ; the Hghway-Vehc1e-Object-Smu1aton-Mode1 (HVOSM) (1), to evaluate wheel drop nto varous culvert openngs on flat terran. HVOSM was also used to nvestgate the effect a ramp at the leadng edge of the culvert openng would have on vehcle behavor. Ramps havng the followng dmensons were eva 1 uated: Hor zonta 1 Ve rt ca 1 Ramp Dmenson ( n.) Dmenson (n.) A 1974 Honda Cvc was smulated n each of the computer runs snce t was assumed a mn-sze automoble would be more crtcal than a larger vehcle for the gven condtons. A speed of 20 mph (32.2 km/h) was used n each run snce t was deemed a crtcal speed. At hgher speeds t was felt the

6 H. E. Ross, et al 5 vehcle would clear the openng easer. At lower speeds, although the veh cle would tend to drop more, velocty changes would be tolerable. Prelmnary tests - n the second phase a test pt was constructed on flat terran as shown n Fgure 1 to study the behavor of a vehcle as t traversed varous openngs. The objectves of these tests were to determne prelmnary values for (1) the maxmum clear openng permssble on a nongrated culvert end and (2) the maxmum spacng permssble when grates are necessary. All runs were lve-drver tests at varous speeds and encroachment angles. Fgure 2 s a photograph of the test pt after nstallaton. A total of 31 runs were made to determne the maxmum clear openng. Test speeds ranged from 5 mph (8.0 km/h) to 35 mph (56.3 km/h);encroachment angles vared from 0 degrees to 15 degrees; and clear openngs ranged from 12 n. (30.5 cm) to 36 n. (91.4 cm). All tests were wth a 1974 Honda Cvc havng a curb weght of approxmately 1800 lb (817.2 kg). Lmtng values were determned by the severty of the rde as judged by the drver. The drver was a Texas Transportaton nsttute (TTl) techncan wth a nonprofessonal drvng hstory. Sequental photos of a 20 mph (32.2 km/h) run wth a 30 n. (76.2 cm) clear openng are shown n Fgure 3. Upon completon of the clear openng tests the pt was used to determne maxmum permssble grate spacng. A total of 22 lve-drver tests were conducted for ths purpose. Test speeds ranged from 5 mph (8.0 km/h) to 25 mph (40.2 km/h); encroachment angles vared from 0 degrees to 30 degrees; and,, grate spacng vared from 16 n. (40.6 cm) to 30 n. (76.2 cm). The grates were 3 n. (7.6 cm) schedule 40 steel ppe anchored to a steel beam wth provson to allow adjustments of the ppe to any desred spacng. Fgure 4 shows the pt setup for a 16 n. (40.6 cm) grate spacng. Each grate confguraton was evaluated wth the 1974 Honda Cvc. A 1975 Plymouth Fury r

7 .. '\U;:';:', t.. ell 6... " 10'-0" Clear Fl at Terra n--_.\'."': S- o ttl OJ - U') U 3"<P Std. Ppe Gratng; Spacng Vares Edge of Pavement.....(:.. <:;... '.. ' ". ". Fgure 1..'..;... \ \ Plan Vew of Culvert Test Pt. 18" Deep Concrete Pt _ Adjustable Cover Plate to Create Clear Openng Varable Clear Openng ' '.." Vehcle ; Encroachment Angle Vares 1,. * 1...;.,... ANGLE: S 5 PEE O:l0l: OPN'G 22 PL, j j Fgure 2. Test Pt nstallaton..,,

8 H. E. Ross, et al 7, ANGLE: 0 SPEEO:20,: OPN G;J.O Pl ' sec sec sec sec., Fgure sec. Sequental Photos of Nongrated Culvert Test, 30 n. Clear Openng, 1974 Honda Cvc, [

9 H. t. KOSS, et al 8 " ANGLE: 0 SPEED:lDf DPN'G,SPPE Fgure 4. Test Pt Wth 16 n. Grate Spacng.

10 H. E. Ross, et al 9 weghng about 4500 b (2043 kg) was also used to evaluate the larger grate spadngs. As part of the second phase of the study a lmted number of lve-drver tests were conducted to further evaluate the effects of a ramp at the leadng edge of the cul vert open ng. Based on the HVOSM results, a ramp wth a horzontal dmenson of 12 n. (30.5 cm) and a vertcal dmenson of 6 n. (15.2 cm) was selected and constructed. HVOSM ndcated ths combnaton would produce the greatest wheel hop of all combnatons consdered. The 1974 Honda Cvc and the 1975 Plymouth Fury were used n the ramp test. Each test was conducted at 20 mph (32.2 km/h). Prototype tests - Based on results obta ned from the prel m nary stud es, two cul vert structures were constructed for full-scale testng. They conssted of a 30 n. (76.2 cm) dameter corrugated steel ppe culvert and a,,,,,, 5 ft (1.5 m) wde by 3 ft (0.92 m) hgh concrete b.ox culvert wth adjonng head and wng walls. Grate members on the box culvert conssted of 3 n. (7.6 cm) schedule 40 steel ppe on 30 n. (76.2 cm) centers. Photos of both nstallatons are shown n Fgure 5. General detals of the sx tests conducted are shown n Fgure 6. Note that the culverts were subjected to tests wth both the mn-sze and fullsze automobles. n each test, wth the excepton of test 5, all four wheels of the test vehcle crossed the sloped culvert openng. n test 5 the vehcle straddled the cross member at the end of the box culvert, allowng the left sde wheels to drop approxmately 1.5 ft (0.46 m) to the dtch bottom. Sequental photos of test 3 are shown n Fgure 7. Analyss of the strength requrements of grate members ndcated that a 3 n. (7.6 cm) J.D. schedule 40 ppe was adequate for spans up to 12 ft (3.7 m). Snce grate spans on many box culverts would exceed 12 ft (3.7 m) t was

11 H. E. Ross, et al 10 a) Corrugated Steel Ppe Culvert. b) Grated Box Culvert. Fgure 5. Prototype Test nstallatons.

12 H. E. Ross, et al 11, VEHCLE DATA Make Model Year Test Weght (1 b) Test No. Velocty (mph),.... Car 1 Honda CVCC , 3, 5 20 Car 2 Plymouth Fury , 4, 6 20, 60 Metrc Conversons: 1 lb m = kg 1 mph = km/h,,. "-:.;." 'j..." Edge of Run/ay,.. ):': '. Dtch Front S19pe Approx. 5: 1 J.,t _. :.. " 112;-Ppe Culvert,.. '......,;.,'r-:'" [ 20 mph, Tests 1 and '.2. Cl '-4J, Y Box Culvert wth Gratng '. '.,.. ':: -::., vehcles towed wth cable Fgure 6 t V = 20 mph, Tests 3', 4, and 5. r-- 60 mph, Test g. Plan Vew of Ste for Prototype Tests.

13 0.000 sec sec sec sec sec Fgure 7. Sequental Photos, Test3., L, ; ;1,

14 H. Eo Ros s, et a 1 13 conclude.d that a lmted test program should be undertaken to determne ppe sze requrements for larger spans. To accomplsh ths another test pt was constructed on flat terran. The pt was of 20 ft (6.1 m) long, 10 ft (3.1 m) wde, and 1.5 ft (0.46 m) deep. A total of four full-scale vehcular tests were conducted wth a b (2043 kg) vehcle, each at 20 mph (32 km/h) and each at a head-on approach, perpendcular to the 20 ft (6.1 m) dmenson of the pt. Further detals of each test are gven n Table 1, ncludng the permanent deformatons noted after each test. Wth the excepton of test 4 the grates had a 20 ft (6.1 m) clear span. n test 4, vertcal supports consstng of 3 n. (7.6 cm) 1.0. schedule 40 ppe were placed at mdspan of each of the three grate members. The grates were attached to the walls of the pt wth a pn connecton, constructed accordng to TSDHPT standards. Parallel-Dranage Structures Smulaton studes - Desgn of a traffc-safe parallel dranage structure not only nvolves the culvert tself but adjonng slopes as well. n fact, the slopes can n many cases be a greater hazard than the culvert structure. Stud es of med an cross-over geometry po nted to the need for relatvely flat slopes to mnmze vehcle rollover (,l). To gan further nsght, HVOSM was used to examne the behavor of a vehcle traversng varous drveway condtons. Parameters nvestgated ncluded departure angle, departure speed, and the path of vehcle encroachment; the sde slopes of both the dtch and the drveway; the type of transton zone between the two slopes; depth of the dtch; and vehcle sze. These parameters are llustrated n the defnton sketch of Fgure 8. Followng s the range of each parameter evaluated: DEPARTURE ANGLE:.15 0 and head-on

15 H. E. Ross, et a1 14 Table 1. Cross Member Deflectons of Box Culvert Gratng Strength Tests TEST PPE a GRATE EST. NO MEMBERb VERTCAL DEFLECTON HORZONTAL DEFLECTON ( n. ) (n. ) (n.) { Frst Second -0 0 Thrd -15/16 3/8 { Frst -1/ Second -1/2 1/4 Thrd /8 { Frst -1 3/4 2 7/ Second -4 3/4 3 1/16 Thrd -4 1/8 1 7/16 { Frst -0 3/4 1 1/2 4 c 3 Second +0 1/2 1 7/8 Thrd +0 1/8 4 3/4 a. Schedule 40 Steel Ppe. b. Grate Members Spaced on 30 n. Centers. c. Mdspan Vertcal Supports Used On Each Grate.

16 H. E. Ross, et a1 15. SHOULDER,.;. SDE SLOPE) "--,. DTCH J /' // )-/ / / / / / DRVEWAY / / /--// (./ TRANST0;J,..../ ZONE SEE DETALS", a DRVEWAY (TYPCAL FOUR, PLACES) ' r----- ANGLED PATH () ANGLED PATH (2) ' DEPARTURE ANGLE PLAN Fgure 8. VEW Defnton Sketch

17 H. E. Ross, et a1 ( DTCH DEPTH SECTON ''A-All 16 ; 1 1 'l, 1 SECTON "8-8 11, LNES OF EQUAL ELEVATON ] ABRUPT TRANSTON DETAL Fgure 8. SMOOTH TRANSTON DETAL 2 Defnton Sketch (contnued) 1, j 1

18 H. t. KOSS, eo ctl DEPARTURE SPEED: 30 mph (48.3 km/h), 40 mph (64.4 km/h), 50 mph (80.5 km/h), and 60 mph (96.6 km/h) PATH: 15 0 angled path across transton (path 1), 15 0 angled path across dtch bottom (path 2), and head-on path nto drveway o slope (path 3) ROADWAY SLOPE: 4:1 and 6:1 DRVEWAY SLOPE: 4:1,5:1, and 6:1 TRANSTON TYPE: Abrupt and rounded DTCH DEPTH: 2 ft (0.61 m) and 3 ft (0.92 m) VEHCLE SZE: 2250 lb (1022 kg) and 4500 lb (2044 kg) A total of 68 computer runs were made to eva 1 uate the va r ous pa rame- [ ters. Prelmnary tests - Ten full-scale vehcular tests were conducted to (1) evaluate vehcle response as a functon of the drveway slope and (2) to develop a tentatve safety treatment for parallel-dranage structures. The test vehcles were 1974 and 1975 Chevrolet Vegas weghng approxmtely 2250 lb (1022 kg). n each test the vehcle was towed to the test ste along a gudance cable, released, and then allowed to traverse the test area n a free-wheel (no steer nput), no-brakng mode. A summary of the 10 tests s gven n Table 2, tests 1-1 through 7-6. Tests 1-1 through 5-1 were desgned to evaluate the relatve hazard of the drveway slope. An earth berm was constructed to smulate the drveway. The berm for tests 1-1 through 1-4 had a 3.8 to 1 slope, was approxmately 3 ft (0.92 m) hgh, and was approxmately 20 ft (6.1 m) wde at the top. Sequental photos of test 1-4 are shown n Fgure 9. After test 1-4 the berm slopes were flattened to the dmensons shown on the frst page of Fgure 10. n ths case the slope on the approach sde was

19 H. E. Ross, et al to 1. t was obvous from test 1-3 that an automoble could traverse the 6.7 to 1 slope at speeds n excess of 40 mph (64.4 km/h) wthout roll ng " over. Hence, test 5-1 was conducted at 50 mph (80.5 km/h) wth the automoble approachng from a head-on path. Although the vehcle was arborne for approxmately 75 ft (22.9 m) t remaned uprght wth no apprecable ptchng. The next ser es of tests (7-1 through 7-6) were conducted to determ ne f safety treatment of the culvert end was needed n addton to the sloped end treatment. The 6.7 to 1 drveway slope was used n each test. t was assumed that a head-on path nto the drveway culvert would be as crtcal, or more crtcal, than any other path regardng the culvert tself. Based on ths assumpton, a 24 n. (61.0 cm) dameter corrugated steel ppe culvert wth a sloped end was nstalled n the earth berm as shown on the frst page of Fguere 10. Ths culvert sze was selected snce the dameter of most drveway culverts n Texas are equal to or less than 24 n. (61.0 cm). Vehcle mpact pont for ths seres of tests was selected such that the rght sde wheels of the test vehcle traversed the center of the culvert end. Detals of the culvert confguraton for each of the culvert tests are gven n Fgure 10. Test 7-1 was conducted at 50 mph (80.5 km/h) wth an open culvert,.e., no grate members. Photos of the nstallaton are gven n Fgure 11 and sequental photos of the test are gven n Fgure 12. Large ptch and roll rates occurred after mpact wth the culvert, and the vehcle rolled over. n test 7-2 a sngle grate member was placed across the culvert as shown n detals 3 and 4 of Fgure 10. Very lttle mprovement n vehcle behavor was realzed and rollover agan occurred. Analyss of test 7-2 showed that grates spaced approxmately on 2 ft (0.61 m) centers was needed to avod excessve wheel drop and wheel snaggng.

20 H. E. Ross, et al 19 The next treatment therefore ncorporated th s featu re as shown n detal s 5 and 6 of Fgure 10. Grate members conssted of 2 lb/ft (2.98 kg/m) steel flanged channel sectons. The channel secton was chosen snce t s wdely used as a del neator post by TSDHPT and woul d therefore be readly avalable. The frst test on ths treatment, test 7-4, was conducted at 20 mph (32.2 km/h) and the results were acceptable. Test 7-5 was conducted at 50 mph (80.5 km/h) and rollover occurred due to structural falure of the grates. n test 7-6, 2-1/2 n. (6.35 cm) 1.0. standard steel ppe (schedule 40) were used as a grate member. Detals 7 through 10 of Fgure 10 show how the ppe was attached to the culvert. Although.the vehcle was arborne approxmately 65 ft (19.8 m) t remaned uprght and the test was deemed acceptable. The culvert was only slghtly damaged. Prototype tests - The fnal two tests, 9-1 and 9-2, were selected to verfy the tentatve conclusons r.eached as a result of the smulaton work and the full-scale slope and culvert testng. A full-scale prototype of a dtch-dr veway conf gu rat on was constructed as shown n F gu re 13 and the 1 j photos of Fgure 14. Test 9-1 was conducted at 40 mph (64.4 km/h) and the approach path nto the drveway was as shown n Fgure 13 such that the left sde wheels crossed the culvert. No adverse vehcle behavor occurred durng the test and the results were consdered acceptable. Test 9-2 was dentcal to test 9-1 except the speed was ncreased to 50 mph (80.5 km/h). Sequental photos of the test are shown n Fgure 15. The vehcle remaned uprght and sustaned only mnor damage. The culvert was j 1 1 only slghtly damaged and could have been used wthout repar.

21 tl' TABLE 2. SUMMARY OF FULL-SCALE TEST RESULTS ::J: ft1 TEST VEHCLE VEHCLE DRVEWAY DTCH CULVERT NO. SPEED PATH SLOPE SLOPE CONFGURATON RE SUL TS (mph) (See Fg. 1) ; 1 N/A. No Cu1 vert Satsfactory - no ro1over ;;0 0 V V. CD <+ '" ; 1 N/A No Culvert Satsfactory - no rollover ; 1 N/A No Cu1 vert Satsfactory - no rollover ; 1 /A No Cu1 vert Unsatsfactory - vehcle ptched over ;1 N/A No Cu1 vert Satsfactory - no rollover ;1 N/A (See Fg. 10) Unsatsfactory - vehcle rolled over ; 1 N/A (See Fg. 10) Unsatsfactory - vehcle rolled over ; 1 N/A (See Fg. 10) Satsfactory - no rollover : 1 N/A (See Fg. 10) Unsatsfactory - vehcle rolled over ;1 N/A (See Fg. 10) Satsfactory - no rollover ;1 6.8: 1 (See Fg. 13) Satsfactory - no rollover ; 1 6.8; 1 (See Fg. 13) Satsfactory - no rollover Metrc Conversons: 1 mph = km/h '" o

22 r H. E.Ros s, et a ntr f f, " , fgure 9. Sequental Photos, Test 1-4.

23 H. E. Ross, et al 22 24'-5" BERM 3 - RALRC)AD T1ES-- PLAN.. FOR TEST 7-1 SEE DETALS AND 2 FOR TEST 7-2 SEE DETALS 3 AND 4 = FOR TEST 7-48& 7-5 SEE DETALS 5 a 6 _::--FOR TEST 7-6 SEE DETALS 7 AND 8 '-t =jjll-=1 ll;. 11- ltllggojll A ----==J RALROAD 111' JlT - -Slll -.. TES 1111_ SECTON "A 3/4" GAL\/. CHANNEL CRMPED a SPOT WELDED 7'-0 (TYPCAL. TEST) ;;-:::S- L...-ll tkr:-;-7-:-7 770'l/ //./ ' ; / /,,//1/,1// t.../ ///.///.;/ ll /.// ll lll- 1 ==.L- 1 = 1 \l1,/ / /',.'.:::::1111=_1 \ lll = F" DETAL, "2" 7.::oTlU DcTAL 2 24" OA. GALV. CORRUGATED PPE--- Fgure 10. Berm and Culvert Detals, Tests 5-1 through /4" GAL\/. ANGLE CRMPED a SPOT WELDED (TYPCAL EA. TEST)

24 H. E. Ross, et al 7'-0" 23, ===:1\= - Jll -=D=E=TA:-::--L--:3=---:2 1/2" DA. STEEL...,'"" (SCHEO. 40) DETAL 4. 7'-0,..:D:E1:::;;-;L:;5 4 - DELNEATOR POSTS EQUALLY SPACED DETAL 6,, 1 R ; J, NOTE: SEE DETAL 9 1-8'-::::.0_.,. / FOR WELD LOCATONs Fgu'e 10. =t /:..--:.--- ;. DETAL 7., / ' ;',, /, =\ -1\\\_ 1\\ /2" DA. STEEL PPE (SCHED. 40) EQUALLY SPACED Berm and Culvert Detals, Tests 5-1 through DETAL 8 (contnued),,,

25 H. E. Ross. et al 24 =----=::-= (/) w a:: :=;.=-LjE:=========t===:::::==: PLAN DETAL 9 1/4 _ /2" DA. STEEL PPE (SCHED. 40) TYP. EA. FOR ALL 3/4" GAL\/. CHANNEL END PPES - DETAL 10 / / Fgure 10. Berm and Culvert Detals. Tests 5-1 through 7-6. (contnued) 15

26 H. E. Ross,et al 25, ",l ff ',j Fgure-ll Test nstallaton Before Test 7-1.,

27 H. E. Ross, et a Fgure 12. Sequental Photos, Test 7-1.

28 H. E. Ross, et a " DA. GAL\l CORRUGATED PPE WTH 4-2 1/2" DA. STEEL. PPES (SCHED 40).:, AVERAGE SLOPE 7.14' [ = AVERAG:: SLOPE 6.54 AVERAGE SLOPE [ PLAN.. = -' SECTON "A" Fgure 13. Test Ste Condtons, Tests 9-1 and 9-2.

29 , Ị J,j, H. E. Ross, et a1 28, Fgure 14. Test Ste, Tests 9-1 and 9-2. j j w 1 " r@

30 H. E. Ross, et al 29 r" Fgure 15. Sequental Photos, Test 9-2.

31 H. E. Ross, et al 30 F ND NGS Cross-Dranage Structures Based on the computer smulatons and the prelmnary test program, t was shown that clear openngs of at least 30 n. (76.2 em) could easly be tra ve rsed at a speed of 20 mph (32.2 km/h). A 36 n. (91.4 em) spac ng was easly traversed at 25 mph (40.2 km/h). For clear openngs n excess of 30 to 36 n. (76.2 to 91.4 em) t was shown that grates spaced on 30 n. (76.2 em) centers would provde satsfactory safety treatment. These fndngs were n fact borne out through sx full-scale prototype tests. Tests of a 30 n. (76.2 em) dameter corrugated steel ppe culvert end, cut to match a 5 to 1 sde slope, were successfully conducted. The culvert openng was readly traversed by both a full-s ze automobl e and a m n -s ze automobl e at 20 mph (32.2 km/h). Tests of a relatvely large box culvert constructed to match the exstng 5 to 1 sde slope also verfed that grates spaced on 30 n. (76.2 em) centers provde a satsfactory safety treatment. Tests of ths treatment at 20 mph (32.2 km/h) and 60 mph (96.5 km/h) by both full-sze, should be extended and anchored at the flow lne to avod any apprecable dropoff at the end of the culvert treatment. n one test, vehcle rollover, and m n -s ze automobl es were conducted. twas al so shown that the grates occurred when the left sde wheels dropped off an 18 n. (45.7 em) openng at the end of the culvert. Prelmnary tests and the prototype tests showed that 3 n. (7.6 em).d. schedule 40 steel ppe grates were of suffcent strength to support a full-sze automoble for smple-supported spans up to approxmately 12 ft (3.7 m). Addtonal full-scale tests were conducted wth a test pt to determne ppe sze requrements for larger spans. Results of these tests provded the followng gudelnes: ll

32 H. E. Ross, et al 31 Suggested Standard Schedule 40 Ppe Sze Sean Length (ft) 1.0. (n.) Up to to through f mdspan vertcal supports are used, 3.0 n. (7.62 em) 1.0. standard schedule 40 ppe can be used for spans up to 20 ft (6.1 m). Other sectons havng equvalent strengths could of course be used. Reference may also be made to an FHWA report (1) for strength requ rements of gr.ates. Results of the study to evaluate the effect of a ramp at the leadng edge > of a culvert openng were nconclusve. HVOSM results ndcated that apprecablewheel hop could be acheved by a small ramp, thus enablng the vehcle to clear larger culvert openngs. full-scale test program was made. An attempt to verfy these fndngs va a However, due n part to the test procedure, the tests dd not provde suffcent data to reach any frm conclusons. To mnmze damage to test vehcles the area behnd the ramp was not excavated and as a consequence the total wheel drop that would have occurred otherwse was unobtanable. Further evaluaton and testng of ramp treatments appear warranted. Parallel-Dranage Structures Based on the computer smulatons and the prelmnary test program t was shown that the drveway slope should be 6 to 1 or flatter to avod vehcle rollover for speeds up to 50 mph (80.5 km/h). The computer smulatons ndcated that the dtch sde slope should also be 6 to 1 or flatter. Even at these relatvely flat slopes a vehcle travelng at 50 mph (80.5 km/h) wll become arborne for approxmately 65 ft (19.8 m). The computer smulatons

33 H. E. Ros s, et a Grate members should extend to and be anchored at the flow lne. Dropoffs at the end of the culvert should be avoded. 5. Necessary grate member szes wll depend on the span of the grates, the manner n whch the grates are supported and the desgn vehcle weght. To support a full-sze automoble the followng szes or ther equvalent are adequate. Suggested Standard Schedule 40 Ppe Sze San Lenth (ft ( n.l Up to to through A 3.0 n. (7.62 cm) 1.0. standard schedule 40 ppe can be used for spans up to 20 ft (6.1 m) f a mdspan vertcal support s used. Parallel-Dranage Structures (for drveways, medan crossovers, ramps, etc.) 1. The roadway sde slope (or dtch slope) n the vcnty of the drveway slope should be 6 to 1 or flatter. 2. The drveway slope should be 6 to 1 or flatter. 3. The transton area between the roadway sde slope and the drveway slope shoul d be rounded or smoothed as opposed to an abrupt trans t on. 4. Safety treatment of the culvert openng should nclude an end secton cut to match the drveway slope wth cross members (grates) spaced approxmately every 24 n. (61.0 em) perpendcular to the drecton of flow. 5. The cross members should be desgned to support a concentrated wheel load of approxmately 10,000 lb (44,480 N) appled at mdspan. g, ', '. :

34 REFERENCES l. "Traffc-Safe and Hydraulcally Effcent Dranage Practces," Natonal < t 2. Cooperatve Hghway Research Program -- Synthess of Hghway Practce 3, Hghway Research Board, Ross, Hayes E., Jr., and Post, E. R "Crtera for the Desgn of Safe Slopng Culvert Grates," Research Report 140-3, Texas Transportaton nsttute, Texas A&M Unversty, August DeLeys, N. J., "Safety Aspects of Roadsde Cross Secton Desgn," Report No. FHWA-RD Federal Hghway Admnstraton, February Ballnger, C. A. and Gade, R. H., "Evaluaton- of the Structural Behavor of Typcal Hghway nlet Grates, wth Recommended Structural Desgn Crtera," Report No. FHWA-RD-73-90, Federal Hghway Admnstraton, December "mprovng Safety of Dranage Facl tes," Admnstratve Crcular No. 8-79, Texas State Department of Hghways and Publc Transportaton, January Ross, Hayes E., Jr., Hrsch, T. J., Jackson, Bento, Jr., and Sckng, Dean, "Safety Treatment of Roads de Cross-D ra nage Structures," Research Report 280-1, Texas Transportaton nsttute, Texas A&M Unversty, March Ross, Hayes E., Jr., Hrsch, T. J., Sckng, Dean, "Safety Treatment of Roadsde Parallel-Dranage Structures," Research Report 280-2F, Texas Transportaton nsttute, Texas A&M Unversty, June "Recommended Procedures for Vehcle Crash Testng of Hghway Appurte- " nances," Transportaton Research Crcular 191, February Segal, Davd J., "Hghway-Vehcle-Object-Smulaton Model ," Report No. FHWA-RD , Federal Hghway Admnstraton, February 1976.