RESEARCH. o ~, 'DEPARTMENTAL SURFACE IMPROVEMENTS TEXAS HIGHWAY. for SKID RESISTANCE GUIDE ~.. ' Number: Center For, \\:!... ;...,,:!.!:\l\!

Transcription

1 \.. ', 'DEPARTMENTAL \\:!.... ;...,,:!.!:\l\!::: RESEARCH Number: 45-2 SKID RESISTANCE GUIDE fr SURFACE IMPROVEMENTS n TEXAS HIGHWAY Center Fr, Highway Researcb Library TEXAS HIGHWAY DEPARTMENT

2 CENTER FOR TRANSPORTATION RESEARCH LIBRARY L9519 SKID RESISTANCE GUIDELINES FOR SURFACE IMPROVEMENTS ON TEXAS HIGHWAYS BY B. F. McCullugh Supervising Design Research Engineer K. D. Hankins Assciate Design Engineer Research Reprt 45-2 Determining and Evaluating Skid Characteristics f Texas Pavements Research Prject Cnducted By Highway Design Divisin, Research Sectin The Texas Highway Department In Cperatin with the U. S. Departmerit f Cmmerce, Bureau f Public Rads August 1966

3 ACKNOWLEDGMENTS Acknwledgment is extended t the Plice Department and t the City Traffic Engineering Sectin in Austin, Texas, fr their assistance in prviding access t the city accident files, especially t Messrs. W. H. Klapprth and W. T. Nuckls f the Traffic Engineering Sectin and t Sergeants Cutler and Wilsn f the Plice Department. The assistance f the San Antni Plice Department is gratefully acknwledged and a special thanks t Mr. C. F. Braunig, District Traffic Engineer f the San Antni District, wh cllected and cmpiled the accident data frm that city. The Skid Resistance Advisry Cmmittee fr this prject has been especially helpful and acknwledgment is given t Mr. R. O. Lyttn f the San Antni District and Mr. J. H. Aiken f the Wac District. This prject was cnducted in cperatin with the Bureau f Public Rads and the suggestins f Bureau persnnel is acknwledged.

4 TABLE OF CONTENTS LIST OF FIGURES ABSTRACT I. II. III. INTRODUCTION Backgrund..... Objective.. METHOD OF ANALYSIS Selectin f Test Sectins Selectin f Accident Type PRESENTATION OF RESULTS Direct Cmparisn Cumulative Cmparisn Summary..... ii.iii IV. DISCUSSION OF RESULTS 22 Accident Minimum 24 Cmpsite Minimum Implicatins f Cmpsite Minimum.. 25 V. CONCLUSIONS AND RECOMMENDATIONS 29 BIBLIOGRAPHY 3 i

5 LIST OF FIGURES Figure N. Page N. 1 Number f Rural sectins Cmpared with 7 Cefficient f Frictin at 2 M.P.H Number f Rural Sectins Cmpared with Cefficient f Frictin at 5 M.P.H Study f Average Accumulative Percent Accidents and Cefficient f Frictin fr Austin and San Antni n IH 35., 1 4 Cmparisn f Ttal Accidents and Cefficient f Frictin at 2 M.. P.H Cmparisn. f Ttal Accidents and Cefficient f Frictin at 5 M.P.H Variatin f Accident Rate f Ttal Accidents with Cefficient f Frictin at 2 M.P.H Variatin f Accident Rate f Ttal Accidents with Cefficient f Frictin at 5 M.P.H Cmparisn f Fatal and Injury Accidents and Cefficient f Frictin at 2 M.P.H.. Cmparisn f Fatal and Injury Accidents and Cefficient f Frictin at 5 M.P.H Variatin f Accident Rate f Fatal and Injury Accidents with Cefficient f Frictin at 2 M.P.H Variatin f Accident Rate f Fatal and Injury Accidents with Cefficient f Frictin at 5 M.P.H Study f Average Accumulative Percent Accidents and Cefficient f Frictin at 2 M.P.H study f Average Accumulative Percent Accidents and Cefficient f Frictin at 5 M.P.H. Cmparisn f Stpping Distance Vehicle vs Trailer Results Accumulative Percent f Rural Sectins Cmpared Cefficient f Frictin at 2 M.P.H. Accumulative Percent f Rural Sectins Cmpared Cefficient f Frictin at 5 M.P. H. with 26 with 27 ii

6 ABSTRACT This reprt pertains t the selectin f a minimum skid resistance fr use as anther guideline fr surface imprvements by the Texas Highway Department. This prblem was apprached frm an accident standpint as well as frm a design standpint, since experience n several sectins f radway indicated a sharp reductin f accidents after surface imprvements. Skid resistance and accident data were cllected n 517 pre-selected rural sectins that represented a sample f Texas Highways. The skid resistance values were btained thrugh the use f a twed trailer emplying the. lcked wheel principal n artifically wetted pavements. An analysis f this data shwed that the pssibility f a radway sectin having a high accident rate increased as the cefficient f frictin decreased. On the basis f this study, cmpsite skid resistances f.4 and.3 fr testing velcities f 2 and 5 miles per hur, respectively, were selected as guidelines fr cnsidering surface imprvements. In additin, skid resistance values f.31 and.24 at 2 and 5 miles per hur, respectively, were recmmended as minimum values. iii

7 1 I. Intrductin Backgrund When discussing skid resistance f radway surfaces with interested persnnel, the first tpic inevitably relates surface frictin and accidents. After btaining equipment t measure skid resistance such as the tw wheeled test trailer recently develped by the Texas Highway Departmentl, the questin immediately frms - hw can accidents be reduced with this device? It is realized that the causes fr.accidents are much mre cmplex than the influence f the deceleratin induced t the vehicle by the pavement surface after the brake applicatin. Mr. K. A. Stnex, f the General Mtrs Crpratin, defines skidding as the mtin f a vehicle under cnditins f partial r cmplete lss f cntrl caused by the sliding f ne r mre wheels f a vehicle. 2 Hw the vehicle reacts when skidding is a functin f many variables f the vehicle itself, such as brake distributins, lad distributins, and braking cnditins. It has been fund that if the frnt wheels are lcked and the rear wheels rll freely the curse f the car is straight. If the rear wheels are lcked and the frnt wheels rll freely the car switches ends by sliding sideways t a psitin apprximately 18 t the psitin befre brake applicatin, much as an inverted pendulum swings abut its pivt. 2 Many variatins,can be expected if variatins are experienced in the number and extent t which wheels lck. Since cefficient f frictin is related t skidding, and studies have fund that wet pavement cefficients are much lwer than dry pavement cefficients, it seems indicative that skidding is related t weather cnditins. It is a cmmn pinin that the number f accidents as well as the injury rate increases n wet radways. Investigatrs bth in this cuntry and abrad have made detailed studies f accidents, skidding, and weather cnditins. Mr. C. G. Giles and Barbara E. Sabey, f the Rad Research Labratry, united Kingdm, reprts

8 2 that eight percent f the ttal number f persnal injury accidents in dry cnditins invlved skidding; whereas, 27 percent f the ttal number f accidents n wet rads invlved skidding. 3 virginia investigatrs reprt.66 percent f the ttal number f accidents n rural rads in dry cnditins were skidding accidents and percent f the ttal number f accidents n wet rads were skidding accidents. 4 Virginia als reprts that skidding f sme nature ccurred in 35 percent f 37,57 accidents during 1956 in Virginia. Dr. Brun Werner reprts that frm 1953 t 1956 in pen cuntry ne accident ut f every 4 r 5 invlves slippery rad cnditins as a cause in Germany.5 The Rad Research Labratry f the united Kingdm reprts that urban areas, even thugh they have the majrity f accidents, are by n means the chief areas, since ver 1/3 f all skids n wet rads ccur n rural rads. 3 The enumerated studies indicate that skidding accidents ccur chiefly n a wet surface, and cefficients are lwer under wet cnditins. Therefre, it must fllw that many accidents are in sme way related t skid resistance. A few years ag a slick sectin f pavement in San Antni, Texas, was "deslicked" by using a methd f sawing small clsely spaced lngitudinal grves in the cncrete surface. 6 This peratin increased the cefficient f frictin frm.32 befre sawing t.42 after sawing as measured by the "stpping distance" methd at 3 mph with wet pavement cnditins. Immediately after grving wrk, the accident rate n the sectin decreased sharply, but within a year the grves began t plish and the accident rate started t increase. Recently sectins f the sawed cncrete were verlaid with asphaltic cncrete cntaining tw different types f aggregate. Cefficient f Frictin values were btained n the verlaid sectins and als n the sawed sectin. Using the sawed sectin as a "befre" cefficient and the cefficients btained n the verlaid sectins as "after" values the fllwing is given:

9 3 TABLE 1 Cefficient *Accidents Percent Reductin sectin Befre After Befre After** In Accidents (1963) (1964) Sawed Cncret * Accidents are given per 1 millin vehicle miles based n ne year f bservatins. ** "After Accidents" have been extraplated t a yearly basis frm seven mnths f infrmatin. -. Even thugh there is a reductin in accidents n the sawed cncrete frm 1963 t 1964, the reductin f accidents n the verlaid sectins is cnsiderably greter. These studies lead t the questin, hw can skid resistance measurements be used t decrease accidents? The mst bvius answer is t establish a minimum cefficient f frictin value. Technically this minimum cefficient shuld vary as t lcatin such as hills, intersectins, r curves and prbably between rural and urban area. The AASHO Guide 7 refers t stpping sight distance and superelevatin, bth f which use minimum cefficient values described as "stpping distance" f a vehicle. Other agencies have set minimums; amng thse are England 3, Michigan 8, and virginia 9 Objective The bjective f this reprt is t establish a guide fr a minimum cefficient value t be used n Texas Highways.

10 4 This minimum cefficient value will be selected using minimum design requirements, accident infrmatin and ecnmics. Althugh the value shuld vary with lcatin, this reprt will nly cnsider general guidelines fr a statewide value. Future studies will encmpass the mre specific applicatins.

11 5 II. Methd f Analysis During the initial backgrund study and after a review f the infrmatin available, it was fund that cnsiderable research wuld be necessary fr a cmplete study f the influence f surface frictin n accidents. In line with the bjective f this study, it was felt that sufficient infrmatin culd be accumulated t establish a guide fr a needed minimum cefficient f frictin value. This wuld be dne quickly with the thught f a cmplete study shrtly after. Selectin f Test Sectins The advisry cmmittee fr this prject recmmended skid resistance tests be perfrmed n 517 sectins which were previusly selected by Texas Transprtatin Institute in cnnectin with pavement structure research study.1 This decisin was based n tw reasns: 1. It was felt that existing data culd be used fr bth prjects, and therefre, n duplicate data gathering wuld be required. 2. The sectins selected wuld prvide a representative sample f Texas Highways. Since the 517 test sectins are predminantly rural, fifteen additinal sectins were als selected in urban areas t prvide a crss-check f any cnclusins r bservatins. The urban sectins were als used t develp study guidelines fr the rural sectins. Urban. The urban study was perfrmed n Interstate Highway 35 in tw cities--austin and San Antni. A preselected number f skids were perfrmed n selected radway sectins in bth cities. The radway sectins were selected n the basis f cnstructin prject bundaries. In general l when the sectins are linked end t end they cmpse the entire distances between city limits. Infrmatin was cllected within each sectin bundary n each f the several accident types. The accident infrmatin was

12 6 btained directly frm the plice department files fr each f the tw cities. In bth cities, the skid resistance values were btained n the utside lanes and these were used fr cmparisn purpses. Rural. The rural tests were made n the 517 preselected sectins. These sectins are als being used as a part f an verall skid study f Texas materials which will be reprted at a later date. Tests in rural sectins cnsisted f respective averages f five skids made fr each f the fllwing cnditins at 2 miles per hur in the inside wheel path, 5 miles per hur in the inside wheel path, and at 2 miles per hur with tbe left wheel r test wheel between the wheel paths. Rural accident infrmatin was btained frm an annual reprt by the Texas Highway Department which cmpiles all accidents reprted by the Texas Department f Public safety.ll Figures 1 and 2 prtray the inside wheel path cefficient f frictin frequency distributins fr the 517 test sectins fr testing velcities f 2 miles per hur and 5 miles per hur, respectively. Nte that the statewide average based n this sample is.56 at 2 miles per hur and.391 at 5 miles per hur. Selectin f Accident Type It is difficult t define accidents caused by skidding, fr in almst every accident the brakes will be applied. In an emergency situatin, brake applicatin may result in several different circumstances: (1) aviding the accident, (2) speed reductins sufficient t result in nly minr damage, r (3) cmplete lss f cntrl thereby leading t an accident. All skidding accidents may nt be caused by brake applicatins in that it has been reprted that 35% f skidding accidents ccurred befre brake applicatin. 4 There is then indecisin as t which accident type t select fr crrelating with the cefficient f frictin. Shuld the study be based n ttal accidents, assuming that the percent f skidding accidents ccurring within the ttal number stays cnstant; r rain

13 1i ' r r r ' ' 81 7' 'I Ttal (All Pavement types) f/) C + Q) en... a:: -... Q) E z 6 I /I '\ " Surface Treatment 4 I /,..; \ 2 M.P.H. In Wheel Path 2 I II 7' 1- _.. _. _. - \ 1\ ar...-- I' I - I I I-=: Cefficient f Frictin at 2 M.P.H. NUMBER OF RURAL SECTIONS COMPARED WITH COEFFICIENT OF FRICTION AT 2 M.P.H. Figure I J

14 1Ir------r ,, I If \ Ttal (All Pavement Types) en c.2 -u IU (J)... :J a: -... IU. E :J Z 61 Treatment 4 I A I '<. 5 M.P.H. In Wheel Path 2 I II F / I 7' ',\: 'c \ I r1:..-- J'::-=...,. ::=:-::--;: Cefficient f Frictin at 5 M.P. H. NUMBER OF RURAL SECTIONS COMPARED WITH COEFFICIENT OF FRICTION AT 5 MPH. Figure 2.8

15 9 accidents, assuming that skidding ccurs mre frequently n the wet pavement; r skidding accidents which l in the present case, are hand-selected taking cnsiderable time and expense? The accidents selected in the urban areas were f three types; skid accidents, rain accidents and ttal accidents. In an effrt t purify the skid accidents selected, nly thse accidents in which skidding was invlved were selected. These accidents wuld be f the type in which the driver applied the brakes fr speed reductin, lst cntrl f the vehicle because f skidding and the vehicle hit anther bject. The rain accidents actually ccurred in a rain r mist. The accidents are als a functin f the vehicular density, especially thse accidents n freeways with shrt gap r headways as cmpared with thse accidents which ccur in sparsely travelled sectins. The sectin length als wuld influence the number f accidents. Therefre, in an effrt t standardize the data, the accidents per length per average daily traffic were selected n a ne-year basis. This standardizatin is called Accidents Per 1,, Vehicle Miles. ll Figure 3 shws the results f the urban study which was used as a pilt study fr the selectin f accident type. The data indicates that the three accident types are clsely assciated. Skid accidents trailed ttal accidents in average cumulative percent and the rain accidents curve was the mst variable f the three. This infrmatin led t the decisin f selecting ttal accidents as a study tl fr the rural investigatin. One limiting feature is the pssibility that all ttal accidents are nt reprted depending upn the severity f the accident and the driver invlved. Hwever, accidents in which an injury r fatality ccurs wuld in all prbability be reprted. Therefre, n the basis f this hypthesis, fatal and injury accidents were als included as a study tl.

16 1 Ir r---=---r A=----r r tl CD '... Q) > «- fit -c: Q) - u «8 6 -c: Q) u... Q) a.. CD > - ::J E ::J U U «4 2 Rin Cefficient f Frictin t 5 M. P. H. STUDY OF AVERAGE ACCUMULATIVE PERCENT ACCIDENTS a COEFFICIENT OF FRICTION FOR AUSTIN a SAN ANTONIO ON IH 35 Figure 3

17 11 III. Presentatin f Results Althugh numerus reprts have assciated accidents and skid resistance, n studies have been reprted in this cuntry that attempt t arive at a minimum skid resistance value thrugh the use f accident data. In this study, the accident data fr the test sectins described previusly was used t investigate the effect f skid resistance n accidents. In rder t develp. future study guidelines, tw different methds f analysis were used. One methd was t directly cmpare accident qata n a sectin f radway with its cefficient f frictin; whereas, the secnd methd used a cumulative frequency distributin curve. Direct Cmparisn In this methd, the accident rate per ne hundred millin vehicle miles fr each given test sectin was pltted in terms f cefficient f frictin fr the test sectin. This type f analysis was run using bth ttal accidents fr a sectin and the fatal injury accidents fr the sectin. Ttal Accidents. Figures 4 and 5 shw the ttal accidents experienced n a sectin f radway in terms f the cefficient f frictin fr that radway at 2 miles per hur and at 5 miles per hur, respectively. Althugh there is a wide scatter f pints, the data des indicate that accidents are in general terms, inversely prprtinal t the cefficient f frictin, r in ther wrds, there is an indicatin that the accident rate increases as the cefficient decreases. Even thugh the number f sectins experiencing accidents d nd increase with a decrease in skid resistance, it appears the number f accidents n anyne sectin des increase as the skid resistance decreases. Since the statistical term "standard deviatin" is a measure f scatter abut the mean, this methd can be used t measure the accident rate "variatin" frm the mean.

18 (f) z 4J U U <t -l <t.!! u.s:::: > c.2 I " " C :> '" '" c a. III '" " u T 2 OF OF x I x I x I x x N 12 X x Z III U u. U. f'.: :t a: :E III U Q.. z :::c - N,... C :::.:,. :..... '.. :. X,.. -. en.... :: z.. <.P <[ &a: C'4 Z '" a:: Q... <t.. (.) 1 III i r IQ : i.'" ".., ". - u... C " I :,.. \, u... U W. : a.:.:..... U a::.. <t C Z <.!) X :) I-. :: :' :..,, Z /....",. W.. (.) -'... u... C U u r<1 w cae X (.) u. ", N COEFFICIENT F FR ICT I ON AT 2 M. P H. TOTAL ACCIDENTS AND COEFFICIENT... u. FRICTION AT 2M. P. H. u. FIGURE 4,. u... -;;; I 1.!!IOOO '" 9 8 <t r<) N C1> ex> t- <.D It) <t r<) N 7 (SClII Cli J!4C1/\ UOIII! pcljpunh ClUO Jad suapijj'1 ) SlN31:):)'1 l'111 5 C 4 <t 3 COMPARISON Z en cae C A. U

19 x I' - r Cf) en., 11., 1 U., > 9 z c 8 w u 7 u <t "., 6 "...J c <t I "., 5 c.,. 4 - en 3 c., " U 2 u <t 1 " I '.. x.*. *. '. j. x.'1' :... '.. :.. J"':.:. \... A. ".. 6 :'.. _ :... '..>. "...., ,.., ".. ' A.... I. _.'. ',. I' ,.'. ' '.-. ' " '.... "., "1":.:..: '..'..,'... j' " ' :':"." I '. '. '. e_.... I x ' '. x COEFFICIENT OF FRICTION AT 5 M P. H. COMPARISON OF TOTAL ACCIDENTS AND OF FRICTION AT 5 M. P. H. FIGURE 5 COEFFICIENT v.l

20 14 The cefficients were therefre gruped int classes with the class interval equal t.15 as shwn by the vertical dashed lines in Figures 4 and 5. The mean f the accidents in anyne class has been pltted as a large circle and a pint three standard deviatin away frm the mean has been pltted with a large "X". Figures 6 and 7 indicate a plt f the standard deviatin and cefficient f frictin. Using the larger variatins as a guide, larger accident rates can be expected in the lwer cefficient ranges and this rate decreases t a relatively cnstant value between cefficients f.3 t.4 at 5 miles per hur and frm O. 45 t.55 at 2 miles per hur. Fatal and Injuries. Figures 8 and 9 prtray the number f fatal and injury accidents experienced in a sectin f radway in lieu f the ttal accidents used in the previus graphs. Figures 8 and 9 are fr 2 miles per hur and 5 miles per hur, respectively. Figures 1 and 11 indicate the variatin experienced in this accident type. The trends and bservatins nted with ttal accidents are verified by these graphs. At 2 miles per hur, a rapid inciease in fatal and injury accidents is experienced when the cefficient f frictin decreases belw a value f.45 t.55. At 5 miles per hur, the increase in fatal and injury accidents is experienced as the cefficient f frictin decreased belw.3. Cumulative Cmparisn Anther methd f cmparisn is t use the data and cnstruct a cumulative frequency distributin curve. One cmplicating factr in making this type f analysis is that the data frms a nrmal distributin pattern rather than a factrial type arrangement, as wuld be the case fr a planned experiment (see Figures 1 and 2). Therefre, a cefficient value near the mean skid resistance f all the test sectins will shw a much greater ttal number f accidents due t the greater number f test sectins in this range. In rder t ffset this phenmenn, an average accident rate fr each cefficient range was calculated n the basis f the number f sectins in the particular

21 z (J) _ z C Q) IJJ= ::! > - Q)... - «;.c Q) Q)... a.> 2. C Q) ' r :::««UI C ' -Q) - e «.- C) e z c= e.- (J) LL C) 1 e « COEFFICIENT OF FRICTION AT 2 M.P.H. VARIATION OF ACCrENT RAT'E OF TOTAL ACCIDENTS WITH COEFFICIENT OF FRICTION AT 2 M. P. H. FIGURE e OL- -L L- -L L- L L----L COEFFICIENT OF FRICTION AT 5 M.P.H. VARIATION OF ACCIDENT RATE OF WITH COEFFICIENT OF FRICTION TOTAL ACCIDENTS AT 5 M. P. H. FIGURE 7

22 III IV en z IV U W.s= 1 IV 9 U > U 4: c: 8 I I I I 1 >- a:: 7 ::> "U J CII 6 z "U c: ::;, :I: IV Z c: «...J CII Q. «5 4 III 3 4: -c:: LL x IV x "U x U 2 u «1 lij - I. x '.. '...'......'... I.. - j ". 1 '. '. ' '... I.....:,.. t,x' '.' J..' ' 'v. :. -'. It.... fl.., I!\. -w......,.. \;J ', '. I. :.",,':,....' I. 1-,-.:..: J...,".;; :... ". I..:, r' "-,.' -'..:',I.'...,.'.,},. x. -.[ I COEFFICIENT OF FRICTION AT 2 M. P. H. COMPARISON OF FATAL AND INJURY ACCIDENTS AND COEFFICIENT OF FRICTION AT 2M. P. H. FIGURE 8 (J)

23 17 x x : I() I «Q Z C tit... Z III Q U U C. % 1ft... C I- v.., N Q m <Xl x I'- ID I() v x. x.. :. '.'. ".. :.. : :,--,'...'.,,..... : -A...,... :. -'!.; rtl... :-:.'" '::': ':..'.., N : I : z I U a:: I.L. I.L. I Z W U I.L. I.L. W U Q Z C... C..... C z... U Z.. III u.. z tit.. III.. U C a u I.L. SlN3al'V,U:lnrNI an'v l'vl'v.:l

24 <:) <:) <:) 5 <:) <:) COE FFIC I E NT OF FRICTION AT 2 M. P. H. VARIATION OF ACCIDENT RATE OF FATAL AND INJURY ACCIDENTS WITH COEFFICIENT OF FRICTION AT 2 M. P. H. FIGURE 1...J «(I)"1:I 2"1:1- LLWC: UJ C :::I ::t:.- U 2 U CP «c: CP OC:; <:) 1 I:J <:) <:) 5 <:) COEFFICIENT OF FRICTION AT 5 M.P.H. VARIATION OF ACCIDENT RATE OF FATAL ACCIDENTS WITH COEFFICIENT OF FRICTION AND INJURY AT 5 M. P. H. FIGURE "

25 19 range. The frequency curve was then cmpiled n the basis f these values. As a result, a cefficient range at the extremes f the distributin has the same influence as ne near the mean. Figures 12 and 13 shw these cumulative percentages fr 2 and 5 miles per hur, respectively. Bth ttal accidents and fatal and injury accidents are shwn n these graphs. At 2 miles per hur, the slpe f the line starts decreasing between.45 and.5 fr the ttal accidents, but in the case f fatal and injury accidents, there is n sharp break in the curve until a value between.5 and.6 is reached. At 5 miles per hur, the slpe begins t decrease at a cefficient value between.3 and.35 fr bth the ttal accidents and the fatal and injury accidents. The pint f change is much mre clear cut at 5 miles per hur than at 2 miles per hur. Summary It is interesting t nte that the tw methds f analysis give critical values apprximately equal althugh the values vary fr speed as wuld be expected. Fr 2 miles per hur, the data indicates that an increase f accidents is experienced when the cefficient f frictin decreases belw a value that is in a range f.45 t.55. At 5 miles per hur, this value appears t be between.2 and.3. Frm purely an accident standpint, the previusly enumerated values may be cnsidered as minimum levels fr radways in the State.

26 1i , r _r------li_=== -CP ' CP > «Fatal a Injury 81 L _ _i A r c CP 6 1 I/C «>, -c CP CP a.. 41 H CP > - ::J E :::::J «21 / / a Injury,.e Cefficient f Frictin at 2 M. P. H. STUDY OF AVERAGE ACCUMULATIVE PERCENT ACCIDENTS a COEFFICIENT OF FRICTION AT 2 M.P.H. Figure 12 I\)

27 1 -Q) Ot Q) > «8 ēn +c Q) ' «+c Q) Q) CL 6 4 Q) > +- :::J E :::J «2 I Y r-fatal 8 Injury O...e:' Cefficient f Frictin at 5 M. P. H. STUDY OF AVERAGE ACCUMULATIVE PERCENT ACCIDENTS a COEFFICIENT OF FRICTION AT 5 M.P.H. Figure 13.8 N

28 22 IV. Discussin f Results At this pint, it is again emphasized that the accident data btained in this reprt was in n way crrelated t driver characteristics vehicular characteristics, r gemetric characteristics. The authrs fully recgnize that pavement skid resistance is nly ne f many factrs cntributing t accidents, but the results f this study pint ut its imprtance. The size f sample used in this study crrelated with experience n specific lcatins certainly lend credence t the apprach used. In this chapter, a cmpsite minimum cefficient f frictin t be used as anther guide fr skidprfing will be derived frm bth design and accident minimum cefficients f frictin. A design minimum cefficient f frictin is stated in AASHO's plicy n gemetric design which is the basis f design in mst states. This minimum cefficient is based n an assumed speed, an adequate perceptin, reactin time, and the assumptin f gd brakes r gd vehicular characteristics. The minimum was established n the basis f data derived by the stpping distance vehicle methd. The values btained by this methd are prbably mre clsely assciated with accidents, in that it parallels the "panic stp" situatin. Since very little stpping distance data was available in this state, and it is necessary t crrelate the tw methds, the wrk perfrmed at Tappahannck, virginia, was used as a basis. 12 Figure 14 is a crrelatin f the tw methds in which the three sedans were selected frm the stpping distance infrmatin, and the New Yrk, Prtland Cement Assciatin, and Bureau f Public Rads trailers were selected frm the trailer infrmatin. These three trailers were chsen, because their design is similar t the design f the Texas trailer. Using the abve relatinship in cnnectin with the curve shwing the skid resistance in terms f velcity

29 Q)....- Q) E! I-.6 I c. I.L.. Q) 1. I Taken Frm Cmparisn Study at.4 - -t I I I I I J-- Tappahannck, Virginia. Cmparisn f grand means at 2 M.P.H. PCA- BPR and New Yrk COMPARISON OF STOPPING DISTANCE VEHICLE VS. TRAILER COEFFICIENT OF FRICTION Figure 14 [\) UI Average with 3 Vehicles. -g I I and 4 M.P.H. perfrmed n 5 sites. I ;: c < I I I I I ' Cefficient f Frictin - Stpping Distance Vehicle

30 24 that was used as the criteria fr stpping distance in the AASHO Guide, a minimum design cefficient as measured with the trailer methd may be btained. The minimum design cefficients n wet pavement using the stpping distance methd are.4 and.3 at 2 and 5 miles per hur, respectively. Using these values with Figure 14, the minimum design cefficient as measured by the trailer methd is fund t be.31 and.25 fr 2 and 4 miles per hur, respectively. On the basis f straight line extraplatjn a minimum design cefficient value f.24 at 5 miles per hur wuld be btained using the stpping distance cefficient f.3 at 5 miles per hur. Therefre, the minimum cefficient f frictin frm the design standpint wuld be.31 and.24 at 2 miles per hur and 5 miles per hur, respectively, as measured by the trailer methd. Accident Minimum Frm strictly an accident standpint, the data discussed in the previus chapter indicate accident rates increase when the skid resistance decreases belw the values in a range f.45 t.55 at testing speeds f 2 miles per hur and.2 t.3 at 5 miles per hur. Althugh it is nt knwn t what extent all the accidents used in the analysis are related t skid resistance, the inter-relatin between the tw factrs is evident, and it shuld be taken int cnsideratin. Cmpsite Minimum Frm a lgical standpint, the design cefficient fr stpping distance as established by AASHO is an abslute minimum value that can be tlerated n the highway' system. Cnsideratin f the accident data brings frth the questin, is this the minimum cefficient value t be maintained fr safety? It is evident frm the data that the cmpsite minimum shuld be greater than the design minimum frm a standpint f driver safety. Althugh this variatin in design accident minimums maybe hypthesized t be a result f vehicle characteristics r driver characteristics, the fact is that the variatin exists and a cmpsite minimum must be determined.

31 25 The selectin f a cmpsite minimum will nt eliminate all accidents, but certainly the severity f the accident can be reduced. A reductin in impact velcity due t an imprved skid resistance has a large effect n the energy f impact, since the energy is decreased by the square f the velcity. The number f vehicles is increasing n Texas highways and recently the speed limits were increased. Many radways experience small headways and gaps and increasing the cefficient value cannt be expected t prevent all accidents, but it can be expected t reduce the severity and number f accidents. Cmbining the design minimum f.31 at 2 miles per hur with an accident minimum f.45 t.55 at 2 miles per hur, it is recmmended that the Texas Highway Department use as a guide a minimum cefficient value f.4 at 2 miles per hur. Even thugh the cefficient at 2 miles per hur is related t cefficient at 5 miles per hur, it is als recmmended that the Highway Department use as a guide a minimum cefficient value f.3 fr 5 miles per hur. These cmpsite minimums shuld be used as anther guide fr scheduling surface imprvements n a sectin f highway. Implicatins f Cmpsite Minimum After selectin f a cmpsite minimum, the immediate questin arises, what percent f the highway system is belw this minimum? Figures 15 and 16 indicate a cumulative frequency distributin f the cefficient f frictin fr the 517 pre-selected rural sectins at testing velcities f 2 miles per hur and 5 miles per hur, respectively. If the selectin f these test sectins represents a sample frm all Texas highways, then it may als be assumed that this selectin represents all highways in the state. Entering Figure 15 with the cmpsite minimum value f.4, it may be pstulated that 27% f Texas highways d nt meet this minimum value at the present time. Upn first cnsideratin, this percentage seems largej hwever, if Figure 15 is entered with the abslute design minimum f.31, it is fund that eight percent f Texas highways d nt cmply with the design minimum at the present time. Investigating these

32 en c -(.) Q) Cf) / / v- :;, a:: - 1: G) u G) n. 4 2 L--\. V / V Median =.47 Mean = Cefficient f Frictin at 2 M.P.H. ACCUMULATIVE PERCENT OF RURAL SECTIONS COMPARED WITH COEFFICIENT OF FRICTION AT 2 M.P.H. Figure J5 N ')

33 c '" -CJ Q) en c "'- a:: - -c Q) (,) Q) Q / I 6 / I 4 2 / V V Median =.36 Mean =.391 LJ v Cefficient f Frictin at 5 M.P. H. ACCUMULATIVE PERCENT OF RURAL SECTIONS COMPARED WITH COEFFICIENT OF FRICTION AT 5 M.P. H. Figure 16 I I I I\l...,

34 28 percentages fr a testing velcity f 5 miles per hur, Figure 16 is entered with the cmpsite minimum cefficient f.3. This analysis shws that 32% f Texas highways d nt meet this value, and again using the abslute design minimum f.24, the data indicates 14% f Texas highways are deficient. These cmpsite minimums s.huld be used as anther guide fr surface imprvement. Ecnmical cnsideratins alng with ther verall prject needs wuld preclude the use f these values as abslute minimums at the present time, but the cmpsite minimums shuld be used as psitive guides fr surface imprvements where feasible. The design minimums shuld be cnsidered as abslute minimums.

35 29 v. Cnclusins and Recmmendatins On the basis f this study the fllwing cnclusins and recmmendatins are warranted: 1. The number f accidents experienced n a highway is related t the magnitude f the surface's skid resistance. The smaller the skid resistnace, the greater the chance f a high accident rate. 2. The cmpsite minimum cefficients f.4 and.3 at testing velcities f 2 and 5 miles per hur, respectively, shuld be used as anther guide fr prgramming pavement surface imprvements. When the skid resistance decreases belw this value, surface upgrading shuld be cnsidered. 3. The design cefficients f.31 and.24 at 2 miles per hur and 5 miles per hur shuld be cnsidered as abslute minimum values. When the radway skid resistance decreases belw these values, an immediate surface imprvement prgram shuld be undertaken. 4. It is recmmended that the Maintenance Engineer in each District institute an inventry prgram t determine the level f skid resistance n each prject in his area. This inventry shuld be kept current in rder t 9rvide anther methd f evaluating the prgram needs fr seal cats, verlays, etc.

36 3 BIBLIOGRAPHY 1. McCullugh, B.F. and Hankins, K.D., IITexas Highway Department Skid Test Trailer Develpment ll, Research Reprt 45-1, April Stnex, K. A., IIElements f Skidding U, prceedings, First Internatinal Skid Preventin Cnference, Part I, August Giles, C. G. and Sabey, Barbara E., IISkidding as a Factr in Accidents n the Rads f Great Britain ll Prceedings, First Internatinal Skid Preventin Cnference, Part I, August Mills, Jr., J. P., IIVirginia Accident Infrmatin Relating t Skidding ll, Prceedings, First Internatinal Skid Preventin Cnference, part I, August Wehner, Brun, IIAccidents Invlving Slippery Rad Cnditins in Germanyll, prceedings, First Internatinal Skid Preventin Cnference, Part I, August Hilgers, H.F. and McCullugh, B.F., IIS1ich When Wet ll, Texas Highways, January IIA Plicy n Gemetric Design f Rural Highways II - AASHO, Finney, E. A. and Brwn, M. G q JlRelative Skid Resistance f Pavement Surfaces Based n Michigans Experience", Reprt Number 295, August Dillard, J. H. and Alwd, R. L., "prviding Skid Resistant Rads in Virginia ll, July "preliminary Listing f Data frm Phases I and II fr Applicatin f AASHO Rad Test Results t Texas Cnditins" I December (Unpublished) 11. "Highway Traffic Accident Tabulatin and Rates by Cntrl and Sectin", Texas Highway Department, Divisin f Maintenance Operatins, Traffic Engineering sectin, 1962.

37 12. Dillard, J. H., and Mahne l D. C. I IICmparisn f several Methds f Measuring Rad Surface Slipperiness Tappahannck, Virginia, May