NCHRP 633 Impact of Shoulder Width and Median Width on Safety N. Sta ma tia dis Unive rsity o f K e ntuc ky K T C Pa rtne ring Co nfe re nc e Aug ust 2010
Background (1/2) Ob je c tive s o f ro a dwa y de sig ns Sa fe ty E ffic ie nc y Huma n a nd na tura l e nviro nme nt fit Ne e d to e va lua te a lte rna tive s T ra de o ff g e o me tric e le me nts Guide line s vs. sta nda rds
Background (2/2) Pro je c t issue s Uniq ue ne ss Co nte xt Sa fe ty implic a tio ns
Research Objectives Unde rsta nd re la tio nships a nd q ua ntify tra de o ffs fo r de sig n e le me nts De ve lo p info rma tio n re so urc e s a nd de c isio n to o ls fo r de sig ne rs
Study Approach L ite ra ture re vie w Da ta a c q uisitio n a nd a na lysis Mo de l de ve lo pme nt a nd e va lua tio n Guide line de ve lo pme nt F ina l re po rt
Research Focus Multi-la ne rura l ro a ds Da ta o f inte re st L a ne width Sho ulde r width a nd type Me dia n width a nd type Cle a r zo ne
Literature Review Sa fe ty implic a tio ns fro m de sig n e le me nt tra de o ffs No t muc h o n multi-la ne rura l ro a ds Hig hwa y Sa fe ty Ma nua l AMF va lue s 2 la ne rura l ro a ds
Data Da ta fo r MN, CA, K Y 1991-2002 pe rio d Da ta o f inte re st L a ne width Sho ulde r width a nd type Me dia n width a nd type Cle a r zo ne (K Y o nly)
Data Distribution Va ria b le CA MN K Y L e ng th (mi) 835.84 975.16 576.08 Se g me nts 2,726 4,385 930 Numb e r o f c ra she s 30,413 16,244 30,788 Numb e r o f injury c ra she s 7,676 2,173 10,428 Se g me nts with no c ra she s 68% 80% 63%
Data Issues Da ta issue s Princ . Arte ria l 4 la ne s 12-ft la ne s 8-ft sho ulde rs Guide line s fo r 4-la ne rura l ro a ds with 12-fo o t la ne s
Methodology (1/5) All crashes Divided Undivided Single Multi All Same for Injury only crashes
Methodology (2/5) Ne g a tive b ino mia l e b 0+b 1 lnADT + b 2X2+b 3X3+…+b nXn E [N] = L whe re E [N] numb e r o f c ra she s pe r ye a r L se g me nt le ng th ADT a ve ra g e da ily tra ffic Xi e xpla na to ry va ria b le s
Methodology (3/5) Va ria b le s c o nside re d F unc tio na l c la ss Rig ht sho ulde r pa ve d L e ft turn la ne pre se nc e Me dia n b a rrie r pre se nc e Sho ulde r width Me dia n width
Methodology (4/5) Ac c ide nt Mo dific a tio n F a c to rs (AMF ) Use c o e ffic ie nts = e b i AMF = e 0.407 =1.50 b i = 0.407 the n AMF
Methodology (5/5) Guide line s Re vie w NCHRP 633 mo de ls Appra ise c urre nt kno wle dg e Co nsult HSM mo de ls Use e xpe rt pa ne l re vie w Re c o mme nd AMF
Shoulder Width Width impa c ts c ra she s Po sitive e ffe c t Wide r sho ulde rs re duc e c ra she s Wide r sho ulde rs e nc o ura g e hig he r spe e ds I nte ra c tio n with la ne width a nd numb e r o f la ne s
Shoulder Width AMF Ave ra g e sho ulde r width (ft) Ca te g o ry 0 3 4 5 6 7 8 Undivide d 1.22 1.00 0.94 0.87 0.82 0.76 0.71 Divide d 1.17 1.00 0.95 0.9 0.85 0.81 0.77 No te s: AMF fo r a ll c ra she s a nd se ve ritie s Divide d: L e ft a nd rig ht sho ulde r widths Undivide d: Rig ht sho ulde rs widths
Medians Me dia n e ffe c t Cro ss me dia n c ra she s Me dia n re la te d c ra she s T o ta l e ffe c t unkno wn Me dia n b a rrie r pre se nc e
Median Width AMF Me dia n width (ft) Ca te g o ry 10 20 30 40 50 60 70 80 Multi-ve hic le 1.00 0.91 0.83 0.75 0.68 0.62 0.57 0.51 No te s: AMF fo r a ll se ve ritie s No e ffe c t o n sing le ve hic le c ra she s
Median Barrier I mpa c t o n c ra she s unkno wn I nc re a se due to pre se nc e De c re a se o n se ve rity Me dia n b a rrie r type Me dia n b a rrie r pla c e me nt Da ta a nd mo de ls inc o nc lusive
Application AMF e stima te c ho ic e impa c t Sing le e le me nt N = [AMF i / AMF j ] – 1 Multiple e le me nts AMF i x AMF j
Application Example Wide n sho ulde r fro m 4 to 8 ft o n a fo ur la ne undivide d ro a d AMF 4 = 0.71; AMF 8 = 0.94 N = (0.71/ 0.94)-1 = -0.24 24% c ra sh re duc tio n pe r ye a r pe r mile
NCHRP 633 vs. HSM (1/2) Sho ulde r width Simila r tre nds Divide d: Sa me ma g nitude Undivide d: L a rg e r diffe re nc e s No AMF fo r sho ulde rs o ve r 8 fe e t HSM sho ulde r re la te d c ra she s o nly
NCHRP 633 vs. HSM (2/2) Me dia n width Simila r tre nds HSM sma lle r re duc tio ns HSM me dia n re la te d o nly c ra she s a nd b a rrie r pre se nt
Conclusions AMF c a n b e use d fo r a ll c ra she s All fo r 4-la ne rura l ro a ds with 12- ft la ne s Suppo rtive o f HSM Additio na l wo rk o n me dia n b a rrie r T ype a nd pla c e me nt Cra sh type s a nd se ve rity
Recommend
More recommend