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.
78 Conclusions The primary conclusions from the Phase I and II measurements of heavy-duty trucks were (1) tire/pavement noise was the predominant noise source for heavy-duty trucks; (2) engine/ powertrain was a secondary source; and (3) noise from elevated exhaust stacks occurred rarely. Further, varying factors, such as the region of the country with different truck fleets, vehicle speed operations, terrain, and operating conditions did not change these conclusions. Specific observations and conclusions from this project include the following: ⢠Tire noise was the consistent and predominant noise at/near ground level. ⢠Most trucks indicated some engine noise; some ground-level noise reflected by the pavement, and some, typically about 3ft (0.9m) above the pavement, through the front wheel well and radiator. ⢠Noise from elevated sources occurred rarely; 5 trucks out of 1,289 had levels at 12ft equal to or greater than at ground level (0.4%); 23 trucks had levels 12ft within 5 dB of ground level (1.8%); and 63 trucks had levels at 12ft within 10 dB (5.0%). Of these 63 trucks, only 56 (4.3%) trucks could be identified as having exhaust stack noise within 10 dB of the ground-level source. ⢠Vertical profiles of noise were all similar and were largely unaffected by site, vehicle operating conditions, terrain, pavement, and region of the country. ⢠For noise distribution modeling, the above variables did not require unique source heights and strength. ⢠For higher speed cruise conditions (~50 mph and above), SIP pass-bys were very close to the REMELs levels, being uniformly offset about 1 dB lower. ⢠For interrupted/grade conditions, the corresponding SIP pass-by levels were lower than the REMELs data by about 2 to 5 dB for vehicle speeds below 50 mph; above 50 mph, the regression of measured pass-by levels matched the REMELs curve. ⢠The heavy truck vertical profile could not be approximated at a 12ft (3.7m) upper source height; more accurate upper source heights ranged from 3.3ft (1m) at lower frequency to 1ft (0.3m) at higher frequency; one upper source height could not be used throughout the full frequency range. ⢠The heavy truck vertical profile could be adequately approximated with a two-point source distribution; one at ground level (primary source) and an upper source at variable height. ⢠For a simplified barrier case, the current TNM heavy truck point-source distribution under predicted barrier insertion loss by 3 to 6 dB compared to the heavy truck average profile, depending on barrier height; the TNM distribution also indicated larger noise reductions with added barrier height than the truck average profile. ⢠Based on limited data, the predominant source for medium trucks was tire noise, with most trucks also exhibiting some engine noise and possibly exhaust noise from lower height outlets. C H A P T E R 4 Conclusions and Suggested Research
Conclusions and Suggested Research 79 ⢠The variation of vertical profiles for the limited number of individual medium trucks was greater than that for heavy trucks. ⢠The beamforming array used in this research provided better resolution and dynamic range than previous NCHRP and Caltrans studies. ⢠In order to match the overall A-weighted profiles more closely, the one-third octave band profiles required some additional weighting to account for wavelength compared to source size. From this research and the previous Florida Atlantic University, Caltrans, and NCHRP work, it is clear that the actual distribution of the heavy truck source heights does not correspond to that used in TNM. Although the TNM distribution may be an expedient method to correlate model results with field verification data, there continues to be a disconnect between actual heavy truck noise generation and its representation in the model. Despite limited data, a similar disconnect may occur for medium trucks. TNM puts about 30% of the source strength at 5ft (1.5m), while the average profile indicates that the level at 5ft (1.5m) is about 7 dB lower than ground level. Further, the average profile for medium trucks could not be approximated by a point source as high as 5ft (1.5m), and, as with heavy trucks, the height of an upper source would need to vary with frequency. Suggested Research Implications for the results of NCHRP Project 25-45 suggest additional research, including reconciling the TNM heavy truck source distribution with the average profile from this research; improved (updated) definition of the lower speed heavy truck REMELs database, particularly for interrupted/grade conditions; and additional data for medium-duty trucks for source location and REMELs definition. Heavy Truck Source Distribution Reconciliation Source height distribution used in TNM does not represent the actual distribution for heavy trucks. The distribution is important because it influences both sound propagation and barrier performance. The distribution used in TNM seems to âworkâ when combined with the other aspects of the model, based on validation work carried out previously. From the simple barrier analysis of this research, it appears the TNM distribution would result in higher walls to achieve the same level of performance as that determined by the use of the average profile developed from the beamforming results. Given the cost implications of this finding, barrier performance, in regard to heavy trucks and their actual source distribution, need to be inves- tigated further. Initially, this could be done by more realistic, independent modeling of truck pass-bys on the receiver side of a barrier using the source distribution developed in this research and that used in TNM. If such modeling reconciles the differences, then measurement of the field performance of barriers in regard to heavy trucks would be recommended. These would be SIP-like measurements with simultaneous data collected on the source side and receiver side of a barrier for a select truck pass-by. More definitive would be the simultaneous measure of source regions of the truck to verify actual distribution, compared to the averages of this research. Barrier performance for a complete pass-by could be complex given that angle incidence from the source to the barrier varies as the truck approaches and recedes from the barrier, and the source itself can be directional. The highest pass-by level behind the barrier may not occur when the truck is directly opposite the barrier. Aside from the distance to the barrier and receiver, actual insertion loss would degrade as the truck approaches and recedes from being perpendicular to the barrier and the time of maximum pass-by on the source side of the barrier. These features could justify the source distribution used in TNM.
80 Mapping Heavy Vehicle Noise Source Heights for Highway Noise Analysis Lower Speed Heavy Truck Database Expansion For heavy trucks operating at speeds below about 50 mph, the existing REMELs curve for interrupted/grade operation is well above the average measured in this research. There is also considerable scatter in the data points compared to trucks above 50 mph. This discrepancy should be investigated further and possibly this region of the REMELs curves modified to reflect todayâs fleet of trucks. Based on this research, source mapping would not be necessary because the profiles at low speed were similar to the higher speeds. It is recommended that considerably more SIP measurements be conducted in this low-speed range for both interrupted/grade and cruise operating conditions. Medium Truck Source Mapping Research similar to the heavy truck noise source mapping should be conducted for medium trucks. From the limited data for medium trucks, as was found for heavy trucks, it appears the two-point source model used for TNM needs to be updated in regard to the height of the upper source and the distribution between upper- and ground-level source strengths. Further, SIP measurements should be performed in parallel to the beamforming measurements to add more data to the REMELs, because the indications are that the current REMELs curves are higher than those for todayâs medium trucks.
