Appendices to NCHRP Research Report 842 (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

C1 APPENDIX C STATISTICAL ISOLATED PASS-BY DISCUSSION TABLE OF CONTENTS COMPARISON OF TEST SITES ............................................................................................C2 Northern California Test Sites ......................................................................................C2 North Carolina Uphill and Downhill Sites ..................................................................C3 North Carolina Level Sites .......................................................................................... C11 North Carolina Slower Speed Sites ............................................................................ C18 COMPARISON OF MEASUREMENT SITE AVERAGES ............................................... C21 COMPARISON OF TESTING METHODS ......................................................................... C27

COMPA Northern Under th vehicle s a given v Northern NB1 me Lakeville equal at line follo NB1 hav produces speed of than the 505 SB1 trend line at 505 SB Figure C sites RISON OF California e SIP proce peed, fitted ehicle type California a rged due to and 505 N 90.3 dBA a ws the SIP ing a steep consistently 60 mph, the ground PCC and 2 sites s of both pa 1 produces 1: Maximum TEST SIT Test Sites dure, the m with a logar . The data f re shown in the small B1 sites sho nd 90.5 dB Reference C er slope. T higher lev tire/pavem at 505 NB1 shown in Fi vements are lower pass- pass-by l ES aximum pa ithmic regre or the heav Figures C1 number of wn in Figur A for Lakev urve, with hroughout els than the ent noise at , which is c gure C2, bot similar to t by levels co evels versus C2 ss-by level ssion, and c y truck pas and C2, wi pass-by eve e C1, the le ille and 50 the Lakevil their respe projection o the Lakevil onsistent w h pavement he SIP Refe nsistent with speed at th s for a spec ompared to s-by levels th the result nts measure vel at the av 5 NB1, resp le results ha ctive speed f the 505 N le site with ith the comp s are OGAC rence Curve the OBSI l e Lakeville ific site ar the SIP Ref along the fo s for the two d on the f erage spee ectively. N ving a flatt ranges, th B1 levels. older DGA arison of F , but 505 S ; however, t evels 4 dB l and 505 N e plotted ag erence Curv ur pavemen test days a irst day. Fo ds are essen either regre er slope and e Lakeville At the OBS C is 3 dB h igure C1. Fo B1 is newer he newer O ower. B1 measure ainst e for ts in t 505 r the tially ssion 505 site I test igher r the . The GAC ment

Figure North C To consi same pav Curve, a differenc however, octave ba C4 and C exhaust n as in the overall le level dat typically those of noise in frequenc C2: Maxim arolina Uph der the effe ement were s shown in e is indicate the speeds nd levels. S 5, respectiv oise. These higher band vel also ind a from Run do not show Runs 7 and the 500 or 6 ies for truck um pass-by ill and Dow ct of uphill plotted alon Figure C3 d between u are also sli ome of the ely. In the u spectra ten s, as shown icated eleva s 7 and 26 the elevat 26 of Figu 30 Hz band tires and pa levels versu nhill Sites versus dow g with the r . For Sites p and down ghtly highe spectra for phill data se ded to show by Run 61 ted low and of Figure C ed low and re C4. The s. These fr rticularly fo C3 s speed at th nhill operati egressions t NC1 and hill. The do r. The effec the uphill a t, two pass-b higher leve in Figure C high freque 4. In the do high frequen downhill sp equencies c r traction tir e 505 SB1 a on, the pass hrough the NC2 with wnhill leve t of grade i nd downhil ys were ide ls in the low 4. Others, ncy levels, wnhill case cy levels, a ectra also an be assoc es often use nd SB2 mea -by levels o points and th the noisier ls are actual s also appa l cases are s ntified as h er frequenc such as Run relative to t (in Figure nd the spec typically dis iated with t d on truck d surement si n nominall e SIP Refer pavement, ly a little hi rent in one- hown in Fi aving percep y bands, as 72, with h he more av C5), the sp tra are more play some ire tread pa rive axles.1 tes y the ence little gher; third gures tible well igher erage ectra like tonal ssage

Figure C tine PCC Figure C uphill 3: Maximum 4: One-thir pass-by no d octave ba ise levels v nd spectra f C4 ersus speed or pass-bys for Sites N at PCC Sit C1 and NC2 e NC1 – tra with trans nsverse tine verse and

Figure C downhill The com pavemen 3.5 dB q NC5 rev though th identical, two sites this diffe spectra f respectiv frequenc powertra and 36 w and down passage results (in 5: One-thir parison of S t is about 5 uieter, in ter eal that dow e downhill this implie , due to inc rence in uph or Sites NC ely. For do ies than tho in noise con ere all note hill directio and its first Figures C8 d octave ba IP data for dB quieter t ms of SIP r nhill levels speeds are g s that the op reased powe ill versus d 4 and NC wnhill, the se of Figu tribution for d to possib ns display t harmonic i and C9), th nd spectra f Sites NC4 han Sites NC esults. Unli are clearly reater. Sinc erating cond rtrain-relate ownhill ope 5, as shown spectra of re C8 and the downhi ly have som onality in bo n the 800 t is pavemen C5 or pass-bys and NC5 i 1 and NC2 ke Sites NC lower than e the OBSI l ition is con d noise goi ration is app in Figure Figure C7 t also less ll measurem e amount o th of the fre o 1000 Hz t tends to em at PCC Sit s shown in , in regard t 1 and NC2 uphill by evels for Si tributing to ng uphill. A arent in com s C7 and C end to hav in the high ents. For th f perceptibl quencies as bands. As n phasize tre e NC2 – tra Figure C6. o OBSI leve , the data fo 1 to 3 dB, o tes NC4 and the differen s with Site paring som 8 for dow e lower lev frequency e uphill resu e exhaust n sociated wit oted in reg ad-related n nsverse tine In this case l, and abou r Sites NC4 n average, NC5 are al ce between s NC1 and N e of the pa nhill and u els in the l , indicating lts, Runs 30 oise. Both u h tire noise ard to the O oise. and , the t 1 to and even most these C2, ss-by phill, ower less , 32, phill tread BSI

Figure C AC Figure C downhill 6: Maximum 7: One-thir pass-by no d octave b ise levels v and spectra C6 ersus speed for pass-b for Sites NC ys at Site N 4 and NC5 C4 – new with newer er S9.5 AC S9.5 and

Figure C For Sites the down are actua higher th spectra f similar to and 69. F exhaust s have a un beamform spectra ( frequenc for eight note is th be localiz 8: One-third NC6 and N hill levels b lly about 1. an uphill in or Site NC7 Runs 8 an or Run 13 ystem. For R iquely high ing measur shown in F ies, consiste of the pass- e unusually ed to the fro octave ban C7 on I-74 y 1 to 2 dB, 5 dB higher the absenc do not ind d 72 of Figu , the truck w un 69, noth peak in the ement, and igure C11 nt with incr bys at this s high level i nt of the ca d spectra fo shown in F on average than uphill e of conseq icate elevat re C10. Ho as using it ing was not 250 Hz ban the source o ) are sprea eased powe ite, specific n the 800 H b; however, C7 r pass-bys a igure C9, th . In this case , suggesting uences due ed levels in wever, there s compressi ed at the tim d. Unfortun f this noise d out with rtrain noise ally for Run z band of R the cause o t Site NC5 – e uphill SIP , the OBSI that the do to the oper the lower a were two r on brake an e of measu ately, this b is unknown elevated l . Subjective s 65 and 75 un 21. From f the noise i newer S9.5 levels are a levels for do wnhill SIP ating condit nd higher f ather uniqu d appeared rement; how and is out o . In the uph evels in th ly, exhaust of Figure C the contou s unknown. AC and up gain higher wnhill Site levels shou ion. Most o requency b e cases: Run to have a n ever, the sp f the range o ill direction e low and noise was n 11. Of parti rs, this nois hill than NC7 ld be f the ands, s 13 oisy ectra f the , the high oted cular e can

Figure Figure C9: Maxim C10: One- um pass-by third octave noise levels band spectr C8 versus spee a for pass-b d for Sites N ys at Site N C6 and NC C7 – DGAC 7 with DGA and downh C ill

Figu Similar t PCC Site to 2 dB, noise wa Site NC1 NC10 on This even Spectra f are quite (in Figur trailer tir for this t more typ frequency re C11: One o most of t NC10 are l as shown in s subjective 1 had high ly had one t (Run 40) or the down similar to th e C13) was es was the a ruck for bo ical of the , which has -third octav he other up ower than u Figure C1 ly thought to er levels in case of exh is indicated hill case are ose of Run also quite hi pparent sou th trailer ax pass-by sp been identi e band spec hill/downhil phill levels 2. For the u occur. As the lower f aust noise, w in Figure C shown in F 13 of downh gh with a le rce of noise les. The oth ectra meas fied with po C9 tra for pass l pairs, the for Site NC phill directi for the other requency ba hich occur 12 as the lou igure C13. ill Site NC vel of 98.4 d , which wa er two runs ured at thi wertrain-rel -bys at Site N downhill p 11. In this c on, 10 case pairs of up nds than th red under c dest data po The spectru 7 (see Figur BA at 70 m s confirmed in Figure s site and ated noise. C6 – DGA ass-by level ase, the diff s were note hill/downhi e downhill ompression int of 100.7 m character e C10). Run ph. Tire fla by the beam C13 (Runs indicate les C and uphil s for the gr erence is ab d where ex ll, the spectr Site NC10 brake opera dBA at 69 istics for Ru 66 at Site N pping noise forming re 30 and 29) s low and l ound out 1 haust a for . Site tion. mph. n 40 C10 from sults were high

Figure C PCC Figure C downhill 12: Maximu 13: One-th m pass-by n ird octave oise levels band spectr C10 versus speed a for pass- for Sites N bys at Site C10 and N NC10 – g C11 with gr round PCC ound and

A final u differenc identical site by a levels. Th differenc apparent direction Figure C pavemen North C The SIP northern Site NC3 were not section. S shown in approxim NC9 sho Site NC9 in the Sit the SIP respectiv lower fre phill/downh e in slope o for the two bout 1.5 dB e spectra fo es in the uph exhaust no and five in 14: Maxim t arolina Lev results for part of Nort was selecte as likely “i ite NC9 wa Figure C ately 0.5 to uld be even was 5 dB q e NC9 even spectra. Re ely. Similar quency con ill case is sh f the sites sites. For t , which m r these site ill and dow ise was note the downhil um pass-by el Test Site Sites NC3 h Carolina a d just beyo nto the thro s close to le 17. The av 1.5 dB even lower in com uieter. Exha ts. The influ sults for Si to the uphi tent and hig own in Fig between up hese sites, h ay have red s follow the nhill spectra d less for b l direction. noise levels s and NC9 ar round Elkin nd the crest ttle” as an vel; howeve erage SIP with the sl parison to ust noise w ence of the te NC3 an ll/downhill p her frequen C11 ure C14 for hill and dow owever, the uced the ef trends of th were not a oth directio versus spe e shown in , there were of hill, as s uphill site, r, it was in levels for ight uphill g Site NC3 if as noted onc slight uphil d Site NC9 airs, the sp cy content t Sites NC12 nhill is sli OBSI leve fect of uph e other uph s apparent a ns with on ed for Site Figure C1 no complet hown in Fig as they wer a long, sligh Site NC9 eometry. In both sites w e in the Site l grade at Si are shown ectra from S han the spe and NC13 ght, and lev ls were high ill operation ill/downhill s in the prev ly three tru s NC12 and 5. Due to t ely flat sites ure C16. In e about to b tly uphill s are lower principle, t ere level, as NC3 pass-b te NC9 was in Figure ite NC9 ten ctra for Site . In this case els are virt er for the u on the pa pairs excep ious cases. cks in the u NC13 with he terrain i of any dist this case, tr egin a dow ection of I-7 than Site he levels for the paveme ys and six t also appare s C18 and d to have m NC3. How , the ually phill ss-by t that Also, phill AC n the ance. ucks nhill 7, as NC3, Site nt at imes nt in C19, ore, ever,

the runs by spectr C15, ther effects o loading a throttle ( pavemen Figure C tine PCC shown in Fi a shapes at e is a large f throttle po nd/or thrott as much) a t effects from 15: Maximu and S9.5 A gure C19 w Site NC9 w r scatter in sition in go le settings t the time o site-to-site m pass-by n C pavement, ere all noted ere more lik the levels fo ing up the prior to rea f maximum may be mu oise levels respectively C12 to have exh e those of S r Site NC9 slight grade ching the a pass-by le ddled by sm versus speed aust or som ite NC3. Fo than for Si of Site NC rray locatio vel. If this all differen for Sites N e exhaust n r SIP levels te NC3. Th 9. For som n, they may is the case, ces in opera C3 and NC9 oise. Other shown in F is may be d e trucks, d not be int then evalu ting conditi with trans pass- igure ue to ue to o the ating ons. verse

Figure C grade Figure 16: Photogr C17: Photog aph of I-77 raph of I-7 measureme 7 measurem C13 nt Site NC3 ent Site NC9 – transvers – S9.5 AC e tine PCC pavement a at the cres t slight upgr t of a ade

Figure C pavemen Figure C slight uph 18: One-th t and hill cre 19: One-thi ill ird octave b st rd octave b and spectr and spectra C14 a for pass-b for pass-by ys at Site N s at Site NC C3 – tran 9 – S9.5 A sverse tine C pavemen PCC t and

C15 As with the previous two sites, Sites NC15 and NC16 have some ambiguity in their operating conditions. The SIP results for these sites are shown in Figure C20. Site NC15 is apparently flat; however, it is also a local low spot. Prior to the array position, there is a very slight downhill approach. Past the array at a little distance off, the highway begins a gradual uphill section. There are only two pass-bys for which exhaust noise was noted. The average SIP levels are almost the same as those of the slight uphill/downhill pair of Sites NC12 and NC13 (see Figure C14). However, the variation in the data points from the regression line at Site NC15 is greater than the Sites NC12 and NC13. This increased variation may be due to variation in operating conditions for individual trucks that are cruising after the downhill section and those that are beginning to accelerate in anticipation of the approaching uphill section. The spectra for this site are also varied, as shown in Figure C21. Two of these spectra were noted for exhaust noise: Runs 26 and 13. Both of these tend to have higher levels in the lower frequency bands than the typical pass-bys, such as Run 35. Runs 13, 26, and 70 all have pronounced peaks at very low frequency, 80 or 100 Hz. These are likely related to the engine firing frequency of about 87 Hz for a V8 engine at 1,300 revolutions per minute (RPM). The level of these peaks could be indicative of a lack of attenuation provided by the exhaust system. Site NC16 is level opposite the array; however, there is an uphill portion prior to the measurement location, as can be seen from the photograph of Figure C22. With the lower posted speed and the uphill portion prior to the measurement site, the speeds were lower than Site NC15. Exhaust noise was noted for 14 of the trucks, the highest for any site. This site also appeared to have more local truck traffic than on the interstate, which seems to be correlated with the occurrence of noisier exhaust systems. A majority of the higher SIP levels in Figure C20 can be associated with observations of exhaust noise. The one-third octave band spectra for four of the pass-bys with noted exhaust noise are shown in Figure C23, along with one more typical pass-by without exhaust noise. Run 14 is of particular interest, as the low frequency noise is unique to all of the measurements from North Carolina. For this truck, the level at 63 Hz is remarkably high and actually determines the overall A-weighted level. As noted regarding Site NC15, this is probably due to engine combustion noise that is not attenuated by the exhaust system. This behavior is also present for the trucks of Runs 37, 60, and 68, although the very low frequencies are much lower. Compared to the more typical Run 69 spectrum, these four trucks with noted exhaust noise also have elevated levels below 800 Hz and above 1,000 Hz. The preponderance of exhaust noise at Site NC16 and the large scatter in the SIP levels may also be due to variation in operating conditions. Subjectively, some of these trucks were slow coming up the grade prior to the array, and these trucks may well have been into the throttle to get back up to speed after the hill.

Figure C pavemen Figure 20: Maxim t C21: One-th um pass-by ird octave b noise levels and spectra C16 versus spe for pass-by ed for Site s at Site NC s NC15 and 15 –AC pav NC16 with ement and l AC evel

Figure C uphill Figure C prior uph 22: Photog 23: One-thi ill raph of me rd octave ba asurement S nd spectra f C17 ite NC16 – or pass-bys SR 211 le at Site NC vel AC pave 16 – level A ment with p C pavement rior with

North C The SIP ramp site Site NC8 result, th out with vehicles 45 mph, in the Sit (REMEL (accelera accounts shown in the Site N the REM differenc levels be higher. In transvers Pavemen accelerat add to th levels of REMELs Figure arolina Slow results from (Site NC8) , the vehicl e levels are levels in th were either and levels m e NC14 resu s)2 curves ting) condit only for tir Figure C24 C8 results ELs cruise a e between a tween Site N principle, e tine PCC t.5 Since thi ing trucks is is suppositi the SIP R cruise (see C24: Maxi er Speed S the two slo are clearly es were obv well above e 89 to 95 d under cruise ore closely lts also. As for heavy tr ions are sh e noise. At , with the S correspond nd interrupt cceleration C8 and RE the data from , which is t s does not o higher than on as the le eference C Figure C25 mum pass-by ites w speed sit higher than iously accel the SIP Ref BA range, conditions follow the point of ref ucks on av own in Figu first glance ite NC14 le ing to the in ed flow curv and cruise i MELs inte Site NC8 ypically co ccur, it cou it is today vels under urve, which ), the Site N noise level C18 es are show the SR 211 erating mor erence Curv adding to th or mild acc SIP Referen erence, the R erage pavem re C25, alo , the REM vels corresp terrupted fl es is about s about 4 dB rrupted flow should be h nsidered to ld be inferre , more than the mostly does not C14 SIP lev s versus spe n in Figure location (S e in order t e. There are e scatter o eleration up ce Curve. T eference En ent under ng with th ELs results onding to t ow. At 30 m 5 dB, while at 30 mph is 3 dB, w igher than th be higher i d that the 1 20 years lat cruise cond include po els are abou ed for low s C24. The l ite NC14). o merge on also some f these data to the post here is gene ergy Mean cruise and e SIP Refer seem to ex he REMEL ph, the dif for Sites NC . The differ ith the REM e REMELs n level than 993-1995 n er. The resu itions of 35 wertrain no t 2 dB lowe peed Sites N evels for th For the on- to I-77, and trucks that . For Site N ed speed lim rally less sc Emission L interrupted ence Curve plain the re s cruise data ference bet 8 and NC1 ence in abs ELs data b data since TNM Av oise produce lts of Site N mph matc ise. Relativ r at 35 mph. C8 and NC e on- ramp as a stand C14, it of atter evels flow that sults and ween 4 the olute eing it has erage d by C14 h the e to 14

Figure C interrupt The spec overall le from 32 (shown i C27. On level, abo line, but to have e NC14, o (Run 73) from Run bands, as where ex start at t measurem of the av 25: REME ed flow, and tra for Site vel than di to 33 mph. n Figure C2 e of these tr ut 8 dB abo none were n xhaust noise nly one truc is shown i 73 are con well as bel haust noise he traffic l ents. Comp erage line, w Ls database tire noise s NC8 genera d the highw In this rang 4). The resu ucks (Run ve the aver oted as havi ; however, k was iden n Figure C2 sistently hi ow 80 Hz. T was noted. F ight about ared to the ith two actu results fo ource region lly had high ay sites, as e, all of th lts in the 3 48) was not age line. Th ng exhaust the overall l tified as hav 8, in comp gher from hese trends or some of 750ft (229m average lin ally 0.5 and C19 r heavy tru only (SIP R er levels in shown in F e trucks we 4 to 35 mph ed for exha e others in F noise. Only evel was abo ing exhaus arison to ot 1,000 to 4,0 are consist the pass-by ) from th e, these pas 1.5 dB low cks and av eference C the lower f igure C26 f re close to range varie ust, and thi igure C27 one other tr ut 7 dB bel t noise, and her trucks a 00 Hz and ent with the s, the trucks e array. Th s-bys were t er than aver erage pave urve) requency ba or trucks in the average d more, as s truck prod had levels a uck in the d ow the aver the spectru t similar sp also in the results from accelerated ese were n ypically wi age. ment for cr nds compar the speed r regression shown in F uced the hi bove the av ata set was n age line. Fo m for that eeds. The l 200 and 25 the other from a stan oted during thin 2 dB o uise, ed to ange level igure ghest erage oted r Site truck evels 0 Hz sites, ding the r less

Figure C range Figure C range 26: One-thi 27: One-thi rd octave ba rd octave ba nd spectra f nd spectra f C20 or pass-bys or pass-bys at Site NC8 at Site NC8 – on-ramp – on-ramp in 32 to 33 in 34 to 35 mph mph

Figure C range COMPA Beyond t Carolina purpose, averages curves. T speed-ind does not REMELs contribut actual me measured condition Californi measured The regre truck cur with TNM the level reference (downhil offset by 28: One-thi RISON OF he SIP ana sites to the the regress . The averag he “tire no ependent te actually cor pass-by da e to the ove asured tire under acce s going up a (only) we in North C ssion lines ves. The OB Average s from the R values (Sit l) are below about 1 d rd octave ba MEASUR lysis, it is in REMELs re ion lines ge es are then c ise” curve i rm used to respond to ta at higher rall. The sp noise. REM leration from a hill. The re measure arolina was for Sites NC SI levels f Pavement. F EMELs dat e NC7 dow the REME B from Sit nd spectra f EMENT SI structive to sults for hea nerated in t ompared to s actually th account for measured tir speeds in w eed depend ELs “interru a traffic-c grades rang d for this 3%. 6 and NC7 or these sec or these da abase for in nhill). In F Ls cruise/S e NC10, si C21 or pass-bys TE AVERA compare th vy trucks s he SIP ana the REMEL at used in powertrain e noise only hich other ence of the pted” flow ontrol devic ed from 3% sub data se are plotted tions are wi ta, the pass- terrupted fl igure C30, IP referenc milar to th at Site NC GES e average hown previo lysis for ea s cruise, in the SIP ana noise is rem , but rather sources bes “tire noise is actually a e and truck to 7%, an t. From Ta in Figure C thin 1 dB o by measure ow (Site NC the SIP reg e curves. T e offset be 14 – SR 211 pass-by leve usly in Fig ch site are terrupted flo lysis in wh oved. The “ to the loga ides tires ar ” curve is s composite s under stea d the four ble 1, the 29 with the r less of the ments follow 6 uphill) an ression line he uphill S tween REM in 44 to 46 ls for the N ure C25. Fo used for the w, and tire ich the con tire noise” c rithmic fit o e thought n imilar to th of data for tr dy state “cr sites in Nor maximum g REMELs h level assoc very close d the cruis s for Site N ite NC11 li ELs cruise mph orth r this site noise stant, urve f the ot to at of ucks awl” thern rade eavy iated ly to e/SIP C10 ne is and

interrupte REMELs flow curv for all thr by 1.5 to reference and uphil when po curves al Site NC9 they equa a given s may be o drivers w those goi may not may also expected compress systems i from the Figure C results d curves. T curve and e. The resu ee of these 2 dB. The curves, by l are similar wertrain no most conve . At low spe l the downh ite grade, e perating the ell below th ng up the s be in the thr be a facto for tire no ion brakes n good wor roadside. 29: Compar he rate of in transitions f lts for Sites sites are sim downhill re as much as 2 to the REM ise may be rge. An inte eds, the Sit ill levels. T .g. downhil ir vehicles e speed lim teeper inclin ottle, like th r for Site N ise alone. may increas king order, ison of Site crease of th rom being b NC4, NC5, ilar and are sults of Site to 2.5 dB a ELs counte becoming resting beh e NC9 resul hese results l, operating differently a it of 70 mph e of Site N ose going d C4. The i As the fast e, resulting the use of c s NC6 and C22 e Site NC11 elow the cru and NC9 ar below that r NC4 fall w t the lower rparts, in tha more signif avior is show ts are equal suggest that conditions s a function (50 to 60 m C5. Those c ownhill in S ncreasing in er trucks g in higher po ompression NC7 SIP re regression, ise curve to e shown in F eported for ell below t speeds. The t the differe icant. At h n in Figur to the uphil even thoug for the vehi of speed. F ph) are we loser to the ite NC4. Th level with ain speed wertrain no braking wo gression cur however, i being abov igure C31. the TNM A he REMEL trends betw nce increas igher speed e C31 for th l levels, and h it might b cles would or Site NC ll into the th speed limi e use of com speed is going down ise. For tru uld not be e ves to REM s greater tha e the interru The OBSI l verage Pave s cruise and een the dow es at lower s s, the regre e gradual u at higher s e thought th be fixed, dr 9, it may be rottle, simi t (65 to 70 pression b greater than hill, the us cks with ex asily identif ELs heavy n the pted evels ment SIP nhill peed ssion phill peed, at for ivers that lar to mph) rakes that e of haust iable truck

Figure C truck res Figure C truck res 30: Compa ults 31: Compar ults rison of Sit ison of Site es NC10 an s NC4, NC5 C23 d NC11 SI , and NC9 S P regressio IP regressi n curves to on curves to REMELs h REMELs h eavy eavy

C24 More evidence of the convolution of operating conditions and truck speeds can be concluded for the results of Site NC1, NC2, and NC3, shown in Figure C32. All three sites are transverse tine PCC, with a range in OBSI levels of less than 1 dB. For the downhill Site NC2, the levels are actually higher than uphill Site NC1. Similar to Figure C31, these data suggest that for trucks operating at or above the speed limit and going downhill, there is a significant occurrence of compression brake usage. Further, the gradient of level versus speed for Site NC2 is almost the same as that of Site NC4 (shown in Figure C31). For Site NC1, the gradient is more, as would be expected, and parallels that of the REMELs interrupted flow curve. The behavior for Site NC3 is almost inexplicable and indicates virtually no increase in level with speed. This site is just beyond a crest in the hill and is located approximately where a truck would begin a downhill section. For trucks in the speed range of 56 to 60 mph, the levels are similar to those of uphill Site NC1, and operators are likely into the throttle. For trucks already near or at the speed limit and facing a downhill stretch, they are likely off the throttle and nearly coasting in anticipation of gaining speed going downhill. The results shown in Figure C33 for Sites NC12, NC13, and NC15 also suggest a mixture of operating conditions influencing the results. These sites are located in close proximity to each other and have the similar pavements. Site NC12 is towards the end of a gradual downhill section that transitions into a gradual uphill section (Site NC13). Site NC15 is a perfectly flat section of pavement that follows a gradual downgrade and precedes a gradual upgrade. The results for these sites are less than 1 dB apart throughout the speed range. The gradients for the sites do not demonstrate typical tire noise behavior, as the increase with speed is well below that known to occur. For the trucks at Site NC13 and NC15, and even for those at Site NC12, the sight of the uphill section could be resulting in operators of the slower trucks getting into the throttle in order to maintain speed on the upgrade. For the operators of the faster trucks, given the slightly sloping terrain, they may be just maintaining a cruise condition throughout the sections. The net result is a low noise level/speed gradient, as indicated.

Figure C truck res Figure C heavy tru 32: Compar ults 33: Compa ck results ison of Site rison of Sit s NC1, NC2 es NC12, N C25 , and NC3 S C13, and N IP regressi C15 SIP re on curves to gression cu REMELs h rves to REM eavy ELs

The resu pass-bys regard to at 35 mph the interr gradient cruise an probably almost 4 the REM as betwe is high co accelerat opposite than that speed sit However slow spe suggestin 55 mph. Figure C heavy tru lts for the s from the o the OBSI r and is grea upted flow behavior is d moderate higher than dB lower th ELs interrup en the Site N mpared to ion is simil of the REM expected fo es. For Site N , there is so eds fall ab g that the tr 34: Compa ck results lower speed n-ramp (Sit esults, the tr ter than wh REMELs c dissimilar t acceleration TNM Ave an the REM ted flow cu C14 curve the current d ar. The curv ELs curves r tire noise C16, the le me indicati ove the RE ucks below rison of Sit Sites NC8 e NC8) are ansverse tin at would be urve, the le o the REM . The pavem rage Pavem ELs cruise rve is about and REMEL ata, there is ature of bo . For Site N , as was not vels follow on of opera MELs cruis the 55 mph es NC8, NC C26 , NC14, an all for truc e PCC pav expected fo vels for Site ELs curve. ent is about ent. The p curves. The 4 dB at 30 s cruise. Th a possibilit th the Site C14, the slo ed in regard the REMEL tional diffe e curve, an speed limit 14, and N d NC16 are ks under a ement on th r TNM Ave NC8 are 1 For Site NC 1 dB quiete ass-by leve offset betw mph, which is implies th y that the d NC8 and S pe of the le to Figures s curve for rences as a d those at are more int C16 SIP reg shown in cceleration. e ramp prod rage Pavem .5 to 5 dB 14, the tru r than Site N ls for Site N een the Sit is about the at even if t ifference be ite NC14 re vel versus s C31 and C cruise withi function of faster spee o the throttl ression cu Figure C34 As discuss uced high l ent. Compar lower. Also cks are a m C8; so, it i C14 are 1 e NC8 curve same differ he REMELs tween cruis sults is als peed data is 32 for the h n less than speed; leve ds are belo e than those rves to REM . The ed in evels ed to , the ix of s still .5 to and ence data e and o the less igher 1 dB. ls at w it, over ELs

C27 COMPARISON OF TESTING METHODS The results from all three types of measurements, including OBSI, SIP, and beamforming are presented in Table C1. The OBSI values represent the overall A- weighted level measured at 60 mph for sites with higher speed limits and 35 mph for Sites NC8 and NC14. The SIP data are presented several ways: the SIP Index, SIP level at the average speed of the data for each site, and SIP level at either 65, 55, or 35 mph, depending upon the site. The source mapping results presented are the average profile level for each site at heights of 0 and 5ft (0 and 1.5m), taken from the average profiles shown in Figure 49 of the main report. For each metric summarized in Table C1, the highest levels were found at Sites NC1 and NC2, which consisted of transverse tine PCC. For the OBSI measurements, the uphill Site NC1 had higher levels, but the downhill Site NC2 resulted in higher SIP levels at average speed, as well as higher profile levels. The lowest levels for each metric were measured at Site NC14, which was a DGAC pavement. However, the vehicle speed at this site was slower than the other sites, which affects the results. Excluding the three sites with slower speeds, Site 505 SB1 resulted in the lowest OBSI levels, the lowest SIP levels at average speed, and the lowest profile levels. As noted above at 505 SB1, the average profile reduced by 4 dBA at 5ft (1.5m), which was the smallest reduction. The maximum reduction at 5ft (1.5m) of 6.8 dB was measured at NC5. As shown in Figure 50, this indicates that the fall-off rate of noise levels is slower at 505 SB1. Therefore, the presence of engine noise and other mid-height sources at or below 9ft (2.7m) is more apparent than at the other sites. Considering the OBSI level at 505 SB1 was the lowest for all faster speed sites, it is possible that the higher tire noise at the other sites was masking other noise sources, such as engine noise, and at 505 SB1 where the tire noise was reduced, the other noise sources became more apparent. With the relatively higher OBSI levels of 505 SB2 compared to 505 SB1, it may be expected that SB 5052 would have a profile more similar to the 505 NB1 and Lakeville. However, there may be some porosity in the OGAC of 505 SB2 that absorbs some of the noise produced by low height sources compared to the pavements of Lakeville and 505 NB1.

C28 Table C1: Summary of measurement results by site Test Site OBSI, dBA SIP Level at Avg. Speed, dBA SIP Avg. Speed, mph SIP Reference Level, dBA SIPI, dB Profile Level at 0ft, dBA Profile Level at 5ft, dBA Lakeville 106.6 90.3 51.4 87.8 2.5 95.9 91.0 505 SB1 100.4 86.0 57.3 89.6 -3.6 93.1 89.1 505 SB2 104.4 88.8 56.8 89.2 -0.4 95.8 91.5 505 NB1 103.6 90.5 58.4 89.6 0.9 96.9 91.8 Site NC1 106.5 93.2 65.9 91.6 1.6 101.1 95.1 Site NC2 106.0 94.0 66.8 91.8 2.2 101.9 96.9 Site NC3 105.7 91.9 63.4 91.0 0.9 99.6 93.9 Site NC4 101.5 90.9 67.7 92.0 -1.1 97.9 91.4 Site NC5 101.3 91.7 62.6 90.8 0.9 98.6 91.8 Site NC6 102.5 91.4 58.4 89.7 1.7 99.3 94.1 Site NC7 104.0 91.3 64.0 91.1 0.2 99.0 93.7 Site NC8 99.0a 84.4 29.0 78.7 5.7 91.8 87.1 Site NC9 100.7 90.7 61.6 90.5 0.2 98.3 91.6 Site NC10 103.7 91.8 67.8 92.0 -0.2 99.2 92.8 Site NC11 104.5 91.9 63.4 91.0 0.9 99.2 93.6 Site NC12 102.6 90.2 63.8 91.1 -0.9 97.9 92.6 Site NC13 104.1 89.7 61.3 90.4 -0.7 97.0 91.8 Site NC14 97.9a 82.5 38.0 83.0 -0.5 89.1 84.4 Site NC15 103.3 89.9 60.7 90.2 -0.3 97.1 90.6 Site NC16 103.9 88.9 53.2 88.2 0.7 93.7 88.9 a Vehicle speed during OBSI testing was 35 mph, instead of 60 mph, due to slower posted speed limits at these sites.                                                              1 P. Donavan, “Generation of Noise by Truck and Car Tires on Various Types of Asphalt Concrete Pavements,” Proceedings of Inter-Noise 2006, Honolulu, Hawaii, December 2006.  2 Fleming, G., Rapoza, A., and Lee, C., U.S. Department of Transportation, Report No. DOT-VNTSC-FHWA-96-2, Development of National Reference Energy Mean Emission Levels for the FHWA Traffic Noise Model (FHEA TNM), Version 1.0, 1996.