2 0 1 3 E d itor: KAZMIERCZAK J.
2
THE INFLUENCE OF TYPICAL SOURCES OF TRAFFIC NOISE MODELLING METHOD ON THE ACCURACY OF CALCULATIONS OF ACOUSTICS EMISSION
2.1 INTRODUCTION
The evaluation o f acoustic influence o f road noise sources w ithin strategic acoustic m aps or other expert studies is connected w ith the necessity to m odel the noise source properly and to calculate its acoustic em ission in accordance w ith the assum ed com putational model. H owever, acoustic em ission m odelling is quite a com plex issue due to the possibility o f various road geom etry m odelling (num ber o f roadways, num ber o f lanes, etc.) as well as due to the possibility to apply different com putational methods. These factors exert a significant influence on the obtained results’ accuracy.
This article presents results o f a research experim ent the purpose o f w hich w as to analyze the influence o f different m ethods o f road noise sources m odelling as well as the influence o f the applied com putational m ethod on accuracy o f the acoustic em ission calculations.
2.2 RESEARCH PROBLEM IDENTIFICATION
In accordance w ith the Polish legislation, w hile preparing road m odels for the purposes o f strategic acoustic m aps’ creation, one uses the road netw orks layer available from the city’s geodetic resources (SIT/SIP) and represented in the form o f road axes. It shall be one axis in case o f a single carriagew ay road and tw o axes in case o f dual carriagew ay road. Therefore, the w ay o f road m odelling regardless o f the num ber o f lanes in each direction is here clearly im posed by the existing GIS model (GIS - G eographic Inform ation System). F or example, figure 2.1 presents various w ays o f a m odelling o f a road having one roadw ay and 4 lanes. In the city’s GIS model the road is described by one axis (a) o f traffic volum e accum ulated to 1 axis. However, this road m ay be m odelled in a m uch m ore precise way, e.g. as a system o f tw o directions i.e. tw o axes (b) and a system corresponding to reality w here each lane is represented by a separate line source (axis) and specified by the traffic volum e resulting from vehicles w hich actually m ove along this lane (c).
Thanks to our experience w e know that the m ethod o f road m odelling shall influence the calculations result. Yet, w e do not know how big the influence will be. The conducted research experim ent attem pts to answ er this question.
The experim ent also focused on analysis o f the results obtained w hile using the same m ethod o f road m odelling but obtained w ith the use o f various com putational m ethods. The follow ing five m ost popular com putational m ethods w ere analyzed:
• French NM PB,
• G erm an RLS-90,
• B ritish CRTN,
• Scandinavian Statens Planverk 48
• Swiss STL-86.
Fig. 2.1 Various models of one type of road:
a) one roadway as the total of all lanes, b) each direction modelled separately, c) all lanes m odelled separately
The m ethod recom m ended by the 2002/49/EC D irective for road noise calculation is the French m ethod N M PB R outes 96 (SETRA -C ERTU -LC PC-CSTB ), specified in „A rrêté du 5 mai 1995 relatif au bruit des infrastructures routières, Journal Officiel du 10 mai 1995, art. 6”
and by French standard “XPS 31-133” .
2.3 COURSE OF THE RESEARCH EXPERIMENT
The research experim ent conducted consisted in selecting hom ogeneous, in respect o f traffic volum e, various types o f roads and m aking very precise m easurem ents o f traffic volum e and equivalent level o f noise in a particular m easurem ent point. W hile selecting m easurem ent points special attention w as paid to the necessity o f ensuring low level o f the background noise during m easurem ents. Then the roads w ere subjected to acoustic m odelling in CadnaA com putational system , using various m odelling techniques and various com putational m ethods, and the obtained results w ere presented in the form o f diagram s and tables.
2 0 1 3 E d itor: KAŹMIERCZAK J.
2 .3 .1 S e le c tio n o f r o a d s e c t io n s a n d m e a s u r e m e n t p o in ts
The determ inant o f selection o f roads w here m easurem ent points w ere situated w as the road type (highw ay, national road, local road), the num ber o f roadw ays, the num ber o f lanes, traffic hom ogeneity and lack o f disturbance from other sources. Thanks to the variety o f roads it w as possible to m ake a m ore detailed analysis o f m odelling w ays o f road noise sources’
acoustic emission. The m easurem ent points w ere located near four different types o f roads:
- Point PP1 - A4 highw ay - tw o roadw ays w ith three lanes each, the section betw een W spólna junction and W irek junction,
- Point PP2 - national road no. 88 - one roadw ay w ith tw o lanes in each direction, the section betw een G liw ice and Zabrze,
- Point PP3 - district road - dual carriagew ay road w ith tw o lanes each way, B ytom ska Street, Zabrze,
- Point PP4 - district road - single carriagew ay road w ith one lane each way, K ujaw ska Street, Gliwice.
2 .3 .2 F ield m e a s u r e m e n ts o f s o u n d le v e ls
Field m easurem ents o f the levels o f sound em itted by m otor vehicles w ere m ade in m easurem ent points located near four types o f roads. Each m easurem ent point w as located 20 m etres away from the external edge o f the road at the height o f 4±0.1 m etres above the ground level.
The m easurem ents w ere m ade on w orking days from M onday to Friday, w hich was connected w ith low er traffic on w eekends. The m easurem ents w ere taken during good and stable w eather conditions, i.e. w hen it w as neither rainy nor windy.
In case o f each type o f read the m easurem ents o f sound levels as well as m easurem ents o f traffic volum e w ere taken at the same time, i.e. betw een 3pm and 4pm, that is at the time w hen the traffic w as relatively high, however, in each case the traffic w as smooth (no traffic jams). Results of m easurem ents regarding the equivalent level o f noise in each o f the m easurem ent points have been presented in table 2.1.
Table 2.1 Results of traffic noise measurements
ID Sound level, dB
LA m in LA m a x LA e q SEL
PP1 61.9 86.9 76.1 111.7
PP2 50.0 86.6 67.9 103.5
PP3 43.1 84.5 64.0 99.5
PP4 41.4 83.1 63.8 99.4
A n acoustic param eter taken into consideration in further studies is obviously the equivalent noise level (LA e q). It is used for com parison o f the results o f acoustic calculations, m ade w ith the use o f C adnaA softw are on the basis o f analysis o f particular m odels o f road noise sources, w ith the results o f acoustic m easurem ents.
2 .3 .3 T raffic v o lu m e a n d tra ffic s tr u c tu r e m e a s u r e m e n ts
The results o f traffic volum e m easurem ents constitute the basis for calculating the
source’s acoustic em ission by m eans o f com putational m ethods im plem ented in CadnaA softw are produced by D ataK ustik. GmbH. That is w hy the m ain assum ption o f the research experim ent w as a very detailed analysis o f the road traffic volum e m easurem ents and identification o f the traffic structure. Traffic volum e m easurem ents w ere m ade exactly for the tim e w hen the m easurem ent o f road noise w as made. V ehicles w ere counted for each lane separately dividing them into lightw eight vehicles (passenger cars, delivery vans) and heavyw eight vehicles (cars above 3.5 tons - trucks, buses, tractors and m otorcycles). Figure 2.2 represents the road (m easurem ent point) w ith m arked m easurem ent sections w here traffic volum e m easurem ents w ere taken, w hereas table 2.2 presents the results o f traffic volum e m easurem ents for m easurem ent point P P 3 .
Fig. 2.2 PP3 measurem ent point together with the marked m easurem ent sections
Table 2.2 The results of traffic volume m easurements - joint PP3 Section Number of lightweight
vehicles
Number of heavyweight
vehicles Total Heavyweight vehicles share [%]
A 396 10 406 2.5
B 70 1 71 1.4
C 73 0 73 0.0
D 315 9 324 2.8
2 .3 .4 T h e m e th o d o f r o a d n o is e s o u r c e s m o d e llin g
C adnaA softw are w as used for the preparation o f geom etric and com putational m odels of the exam ined sources o f road noise. It w as assum ed that each o f the road w ould first be m odelled in a sim plified m anner thus it w ould be represented by one linear source (axis) and described by the total traffic volum e on the road. Subsequent m odels w ould take m ore and m ore details into account, consequently, they w ould be m ore and m ore accurate. A n exam ple o f the m odels created for a dual carriagew ay road (B ytom ska Street in Zabrze- point PP3) has
2 0 1 3 E d itor: KAZMIERCZAK J.
been presented in figure 2.3.
M odels sim ilar to the ones presented in figure 2.3 w ere prepared for the rem aining roads analysed in the research experiment.
c) P P ;;- 3
a)
PP 3-1
d) P P ; M
b)
PP3-2
T : v
V77^7777^/7~/A
e) P P 3 - 5
Fy 7 7 ? /////ÏÏ7 7 A
Y777%9777Zffim
Fig. 2.3 Exemplary models of the noise source for a dual carriageway road (point PP3):
a) model of one roadway as the total of all lanes,
b) model of one roadway with the
“Calculate both external lanes separately"
function on, c) each of the roadways (traffic directions) modelled separately,d) as above with the
“Calculate both external lanes separately’
function on, e) each of the lanes modelled separately2 .3 .5 C o m p u ta tio n a l d e te r m in a tio n o f s e le c t e d r o a d s' a c o u s tic e m is s io n b y m e a n s o f v a r io u s m o d e llin g w a y s
In order to determ ine the sound level in m easurem ent points by m eans o f com putational m ethods, each o f the created geom etric m odels o f roads had to be supplem ented w ith detailed data connected w ith traffic param eters, road param eters, etc. Input data for the use o f com putational m ethods from C adnaA softw are w ere as follow s:
- traffic v o lu m e w ith th e p e rc e n ta g e s h a re o f h e a v y w e ig h t v eh icles
D epending on the m ethod o f a selected type o f ro ad ’s modelling, the volum e o f traffic w as specified for each lane separately or w as accum ulated for each o f the roadw ays.
M oreover, m ost o f the com putational m ethods used includes the “C a lc u la te b o th e x te r n a l la n e s s e p a r a t e l y’ option. In case w hen this option is on, the program divides the total traffic volum e o f one roadw ay equally for tw o lanes.
- a d m issib le sp e e d lim its for lig h tw e ig h t a n d h e a v y w e ig h t v eh icles
D ifferent speed lim its are adm issible for lightw eight and heavyw eight vehicles in case
o f the various types o f roads analysed. A rticle 20 o f the Traffic Code contains inform ation regarding adm issible speed o f vehicles on various road types. A dm issible speed lim its o f vehicles on the analysed roads are as follow s:
PP1 - A4 highw ay - adm issible speed: 140 km/h; for heavyw eight vehicles: 80 km /h,
PP2 - national road no. 88 - adm issible speed: 100 km/h, for heavyw eight vehicles: 80 km/h, PP3 - reduced speed lim its on the basis o f vertical signs - 70 km/h,
PP4 - built-up area from 5am until 11pm - adm issible speed lim it 50 km/h.
- ty p e of su rface of th e a n aly zed ro a d ty p e
Com putational m ethods used for the calculation o f the level o f noise em itted from various types o f roads include the option o f selecting road surface. The surface o f all the analyzed types o f roads is m ade o f smooth asphalt and their condition is good. Figure 2.4 presents an exam ple o f a dialogue box “R oads” o f the CadnaA software w hich allows to enter the data specified above.
Fig. 2.4 "Roads” dialogue box with data entered regarding:
1) traffic volume, 2) admissible speed limits, 3) type of surface
A fter one has entered all the data and m ade calculations using the “Calculate”
com m and, the field next to the receptor displayed the value o f noise level expressed in [dB ] for the selected road model. D eterm ination o f the equivalent level o f noise A in the assum ed m easurem ent points for the m odels created for each o f the types o f roads w as carried out by m eans o f five com putational m ethods:
- french m ethod - NM PB, - germ an m ethod - RLS 90, - british m ethod - CRTN,
- scandinavian m ethod - Statens Planverk 48, - swiss m ethod - STL86.
2 0 1 3 E d itor: KAZMIERCZAK J.
2.3 COMPARISON OF THE MEASUREMENTS' RESULTS AND MODEL
CALCULATIONS' RESULTS
The com parison o f acoustic m easurem ents’ results and m odel calculations’ results consisted in calculating the difference betw een the calculations’ result, by m eans o f particular com putational m ethod, and the result o f the actual m easurem ent o f the sound level. Selection o f the m ost accurate com putational m ethod and the m ost accurate m odelling w ay for a particular type o f road w as based on finding the sm allest difference betw een the calculations’
result and the result o f m easurem ent for a particular model.
Fig. 2.5 Comparison of measurements' results and model calculations' results of noise em ission from A4 highway (point PP1)
Fig. 2.6 Comparison of measurements' results and model calculations' results of noise em ission from the national road no. 88 (point PP2)
Figures 2.5-2.8 present diagram s depicting the results o f m easurem ents and calculations m ade for various m odels o f road noise sources and various m odels o f noise propagation, for all types o f the analyzed roads respectively. Red dashed line w as used to represent the noise level corresponding to the m easurem ent result. R ed column in each o f the diagram s represents the value o f sound level for the m ost accurate com putational m ethod for a particular type o f road.
C um ulative list o f the recom m ended m odels as well as road noise com putational m ethods for the four types o f roads, w hich w ere the subject o f the research experim ent, have been presented in Table 2.3 on the basis o f analysis o f the results shown in figures 2.5-2.8.
Fig. 2.7 Comparison of measurements' results and model calculations' results of noise em ission from a district dual carriageway road (point PP3)
Fig. 2.8 Comparison of measurements' results and model calculations' results of noise em ission from a district single carriageway road (point PP4)
2 0 1 3 E d itor: KAZMIERCZAK J.
Table 2.3 includes the m ost accurate geom etric m odels o f roads as w ell as com putational m ethods for each o f the analyzed types o f roads. As it may be noticed, the results o f calculations m ade in points PP1, PP3 and PP4 are slightly different from the results o f m easurem ents, w hereas in case o f point PP2 the difference is 2.7 dB.
Table 2.3 Cumulative list of the recommended geometric m odels and computational __________________ methods for various types of roads ________ ________
ID
Recommended model
Recommended computational method Measurement result[dB] s
n iso ]dB lita d[t l t 3 "slc s
la e
u - Difference [dB] Model
number Scheme
PP1 PP1-1
\lllllllllll(lllll!llllllll\
RLS-90 76,1 76,0 -0,1
PP2 PP2-1
III II I I I l iî 1II /1II III
STL-86 67,9 70,6 2,7
PP3 PP3-5
ii 1111u 11 H Y m
NMPB 64,0 64,4 0,4
PP4 PP4-3
m i n i m i litiYi m i m i m ii
CRTN 63,8 64,1 0,3
CONCLUSIONS
The conducted research experim ent as well as the analysis o f the obtained results allow ed to draw the follow ing conclusions regarding the correct m ethod o f acoustic em ission m odelling o f road noise sources:
- Com putational determ ination o f road noise em ission in CadnaA program is based on data regarding the volum e o f traffic (including the percentage share o f heavyw eight vehicles),
adm issible speed lim it and the type o f the road surface. That is why, before the m odelling com m encem ent, great em phasis should be put on credibility o f the data collected .
- In case o f modelling, in C adnaA program, o f roads o f high traffic volum e and a great num ber o f lanes (e.g. highw ays), the program raises too high the results o f noise em ission calculations in com parison to the m easurem ent result. M ethods RLS-90 and STL-86 in the model w hich is theoretically the least accurate constitute an exception w here the results o f calculations are slightly low er than the m easurem ent result.
- W hen m odelling local roads having significantly low er traffic volum e than in case o f highways, CadnaA program results o f noise em ission calculations are low er in com parison to the m easurem ent result.
- W hen considering roads o f high traffic volum e and a great num ber o f lanes (in this case a highw ay and a national road), the m ost accurate m odelling w ay turned out to be the m odelling o f one roadw ay as the total o f all the lanes and introducing for this roadw ay an accum ulated traffic o f vehicles from all the lanes. In this case the source o f noise em ission is situated in the central part o f the m odelled roadway, i.e. on its axis.
- In case o f local roads w here the traffic volum e is significantly low er than in case o f highways, the m ost accurate model included the m odelling o f each o f the roadw ays separately and introducing traffic volum e for each o f the roadw ays separately. In such a case there are a few sources o f noise em ission (depending on the num ber o f lanes) located on the axis o f each o f the lanes.
- M ethods RLS-90 and STL-86 are the recom m ended road noise com putational m ethods for roads o f high traffic volum e and a great num ber o f lanes. In case o f RLS-90 method, for the recom m ended m ethod o f point PP1 m odelling (highw ay) the difference betw een the m easurem ent result and calculations result w as only 0.1 dB. W hereas in case o f STL- 86 m ethod this difference for point PP2 w as 2.7 dB. However, it is w orth m entioning that calculations’ results obtained by m eans o f STL-86 m ethod changed slightly, w ithdraw ing from the recom m ended m odelling way. In case o f RLS-90 m ethod, the change o f the m odelling w ay from the recom m ended into a different one resulted in a increase o f the difference betw een the m easurem ent result and the calculations’ result.
- In case o f local roads o f low traffic volum e, N M PB and CRTN turned out to be the m ost accurate com putational methods. As for the N M PB m ethod, recom m ended by the EU D irective, the difference betw een the calculations’ result and the m easurem ent result for the recom m ended w ay o f point PP3 m odelling w as 0.4 dB. It is w orth noticing that that the CR TN m ethod becam e m ore and m ore accurate proportionally to the decrease o f traffic volum e on the exam ined roads and proportionally to the increase o f the m odel’s theoretical accuracy (m odelling each o f the roadw ays separately). In case o f PP 4 point (local road o f low traffic volum e) the difference betw een the m easurem ent result and the calculations’ result w as as low as 0.3 dB. O ne m ay also notice that the STL-86 com putational m ethod, w hich turned out to be the m ost accurate in case o f roads o f high traffic volum e, w as the least accurate for points PP3 and PP4 (local roads).
- The “C a lc u la te b o th e x te r n a l la n e s s e p a r a t e l y’ option available in C adnaA software exerted a small influence on the calculations’ results or increased the results in com parison to the m easurem ent result. This option divides the total traffic volum e o f one
2 0 1 3 E d itor: KAŹMIERCZAK J.
roadw ay equally for tw o lanes and it constitutes some kind o f a sim plification (it is well know n that m ore vehicles m ove on the external lanes). N onetheless, as it has already been m entioned, this sim plification did not have a significant influence on the calculations’ results or raised them too high, that is w hy it is not recom m ended to use this option for m odelling processes.
- M ethod N M PB recom m ended by the EU D irective for traffic noise calculation turned out to be the m ost accurate in case o f m odelling local roads having m oderate or low traffic volume.
- The obtained results indicate to the necessity o f conducting large-scale research in order to confirm the obtained results and conclusions.
T h is a r tic le w a s w r itte n w ith in d e v e lo p m e n t p r o j e c t no. 1 4 0 0 0 1 0 6 /2 0 0 9 : “N e tw o r k m o d e l o f c o n s u ltin g in th e p r o c e s s e s o f c r e a tio n a n d e x p lo ita tio n o f s tr a te g ic a c o u s tic m a p s ” f u n d e d b y
th e N a tio n a l C e n tr e f o r R e s e a r c h a n d D e v e lo p m e n t.
REFERENCES
1. Boczkow ski, A., K oźlik, G. W e r y fik a c ja d o k ła d n o ś c i o b lic z e ń e m is ji a k u s ty c z n e j ź r ó d e ł h a ła s u d r o g o w e g o w z a le ż n o ś c i o d p r z y ję te g o s p o s o b u m o d e lo w a n ia ź ró d ła . M ateriały XL Zimowej Szkoły Zw alczania Zagrożeń W ibroakustycznych. G liw ice-Szczyrk 27.02
02.03.2012. Str. 17-29. (2012).
2. B oczkow ski A.: Z a r z ą d z a n ie k lim a te m a k u s ty c z n y m m ia s ta z w y k o r z y s ta n ie m c y fr o w y c h m a p a k u s ty c z n y c h.Ekonom ika i O rganizacja Przedsiębiorstw a. N r 7. Lipiec 2006.
3. B oczkow ski A., K om oniew ski M.: W y k o r zy sta n ie c y fr o w y c h m a p a k u s ty c z n y c h d o r e a liz a c ji z a d a ń z z a k r e s u o c h r o n y ś r o d o w is k a p r z e d h a ła s e m. M ateriały X X X IV Zimowej Szkoły Zw alczania Zagrożeń W ibroakustycznych. G liw ice-U stroń 27.02
3.03.2006. Str. 231-237.
4. Poradnik dla pracow ników sam orządow ych uczestniczących w tw orzeniu i użytkow aniu strategicznych map akustycznych. Praca zbiorow a pod redakcją prof. Jan K aźm ierczaka.
Główny Instytut Górnictwa, K atow ice 2008.
THE INFLUENCE OF TYPICAL SOURCES OF TRAFFIC NOISE MODELLING METHOD ON THE ACCURACY OF CALCULATIONS OF ACOUSTICS EMISSION
Abstract: The implementation o f the strategic acoustic maps or evaluation o f traffic investments interactions entails the necessity o f modelling the traffic noise sources and calculating the emission o f acoustic source. The modelling o f the traffic noise seems to be the complex matter owing to different methods o f modelling the road geometry (e.g. the number o f roads and lines) and the possibility o f application different methods o f calculation. Mentioned factors significantly affect the accuracy in received results. The article below presents the course and the results o f the scientific experiment, which main aim was analyzing the impact o f different ways o f modelling the traffic noise sources and the influence o f the applied method o f calculation on the accuracy in calculations o f road acoustic emission. With particular precision there were explored the results obtained with the use o f French calculating method NMPB Routes 96 recommended by The Directive 2002/49/WE fo r traffic noise calculations.
Key words: strategic acoustics maps, traffic noise, modelling the traffic noise sources, accuracy o f the acoustic calculation, French calculating method NMPB Routes 96
WPŁYW SPOSOBU MODELOWANIA TYPOWYCH ŹRÓDEŁ HAŁASU DROGOWEGO NA DOKŁADNOŚĆ OBLICZEŃ EMISJI AKUSTYCZNEJ
Streszczenie: Realizacja strategicznych map akustycznych miast lub ocen oddziaływania inwestycji drogowych na klimat akustyczny, wiąże się koniecznością modelowania źródeł hałasu drogowego oraz obliczaniem emisji akustycznej źródła. Modelowanie hałasu drogowego stanowi dość złożony problem, z uwagi na różne sposoby modelowania geometrii dróg (tj. liczby jezdni, liczby pasów ruchu,
itp.) oraz możliwość zastosowania różnych metod obliczeniowych. Czynniki te wpływają w istotny sposób na dokładność uzyskiwanych wyników. W artykule przedstawiono przebieg i wyniki eksperymentu badawczego, którego celem było zbadanie wpływu różnego sposobu modelowania źródeł hałasu drogowego oraz wpływu zastosowanej metody obliczeniowej na dokładność obliczeń emisji akustycznej dróg. Szczególnie dokładnie przeanalizowano zbadano wyniki uzyskane przy zastosowaniu francuskiej metody obliczeniowej NMPB Routes 96 zalecanej do obliczeń hałasu drogowego przez Dyrektywę 2002/49/WE.
Słowa kluczowe: strategiczne mapy akustyczne, hałas drogowy, modelowanie źródeł hałasu drogowego, dokładność obliczeń akustycznych, francuska metoda obliczeniowa NMPB Routes 96
dr inż. Arkadiusz BOCZKOWSKI, Grzegorz KOŹLIK
Silesian University of Technology, Faculty of Organization and Management Institute o f Production Engineering
ul. Roosevelta 26, 41-800 Zabrze e-mail: Arkadiusz.Boczkowski@polsl.pl
