Testing of Emission DLL s for. CNOSSOS-EU Road, Rail and Industry Noise Sources

Transcription

1 Testing of Emission DLL s for CNOSSOS-EU Road, Rail and Industry Noise Sources General Stapelfeldt Ingenieurgesellschaft mbh, Dortmund, (SIG) undertook independent testing of emission DLL s developed and provided by DGMR for use in CNOSSOS based noise exposure analysis. Development of the DLL s is based on Chapter III of the JRC Reference Report Stylianos Kephalopoulos, Marco Paviotti, Fabienne Anfosso Lédée (2012), Common Noise Assessment Methods in Europe (CNOSSOS-EU), EUR EN [1] Program design guides were submitted by DGMR in a series of corresponding.pdf files. The referenced versions were: N SJS.pdf for roads, N SJS.pdf for rail, and N SJS.pdf for industry To support using the DLLs in an instantly available manner DGMR provided a separate tool named CNOSSOS_DLL_CONSOLE.EXE, which addresses the 3 source DLLs. Testing Methodology The DLL s under investigation were: CNOSSOS_ROADNOISE_DLL.dll CNOSSOS_RAILNOISE_DLL.dll CNOSSOS_INDUSTRIAL_NOISE_DLL.dll The 3 DLL s can be addressed via CNOSSOS_DLL_CONSOLE.EXE. This software tool takes XML files as input data and writes XML as output files. For the purpose of testing a large variety of parameter settings for the emission calculation, existing software is used to write the XML input file, call CNOSSOS_DLL_CONSOLE.EXE and read the XML output file for each test case. Return values are compared to result values which have been created in an alternative manner. For road and railway noise, the 2 sets of results were collected in XLS files which are delivered as part of this report. Doc. Rev

2 Testing emission DLL s CNOSSOS Source Modules The DLL s take basic input data from extra XML files, which were delivered together with the DLL s: CNOSSOS_Road_Params.xml CNOSSOS_Road_Surfaces.xml CNOSSOS_Rail_ehicles.xml CNOSSOS_Rail_Track.xml CNOSSOS_Industry_Catalogue.xml These files contain data subject to potential changes by national authorities and any results documented in this report might not be reproducible after such changes. Testing the Road Noise DLL For a range of test cases the CNOSSOS_DLL_CONSOLE.exe was used to generate road emission data for various combinations of input parameters. Chosen octave bands of those results were compared to results calculated in an alternative manner, i.e. manually, by a small software program, or by AWK scripts (AWK is a UNIX command based script language by Alfred. Aho, Peter J. Weinberger und Brian W. Kernighan). The focus of comparison was on the change in noise emission levels which occur due to modifications of individual input parameters. Potentially critical input parameters were selected with respect to the underlying formulas. Details of the process are described further in the following section Details of Road Noise Testing. To be consistent with the testing strategy used in testing the railway noise DLL it was decided to independently program the source model and create an XLS output table with results produced by the source DLL as well as by the independent code. For this alternative testing approach, the number of checks was limited to several thousand sets of input data. This was done by defining lists of potential input choices for each source parameter. By combining these values the various test cases are generated. Some input parameters are linked in this process to avoid unreasonable combinations and also to limit the number of total number of combinations. The road noise testing procedure is described in more detail in the Appendix at the end of this report. Testing the Railway Noise DLL For a range of test cases the CNOSSOS_DLL_CONSOLE.exe was used to calculate railway emission data for various combinations of input parameters similar to the approach described before. 2 Doc. Rev

3 The full octave spectrum of those results was compared to results calculated in an alternative manner, i.e. by using an alternative software program which does not refer to the DLL provided by DGMR. Test cases were generated in an equivalent manner to those described for the road source DLL above. Further details of the railway noise testing procedure is described in the Appendix at the end of this report. Testing the Industrial Noise DLL This test did not include any comparison of separately calculated values as no numerical processing is used in the emission DLL provided by DGMR. The DLL was linked to an alternative software program to test its function by calling sets of emission data related to specific keywords, describing the source. Testing Results The findings of the testing and debugging of the DLL s led to corrections in the documentation as well as the DLL code. This resulted in revised DLL s being released by the developer. Information on the testing results is set out below. Road Noise DLL In Chapter III of the JRC Reference Report then relevant octave bands of road traffic noise source emission was defined in the range of (Hz) while the DLL s use the correct range of (Hz). The documentation has been corrected. Clarify frequency range for temperature correction Parameters α and β of the source surface description needed to be addressed by index reflecting a potentially national data set. There the notation was changed from e.g. 01 to NL01. Railway Noise DLL For traction noise the values for Accelerating and Decelerating were swapped In case of idling vehicles, only calculate traction noise Missing initialization of all results to zero (prevents random results for unused results) An inconsistency in the XML data base was found for ehicle ID 5, 6, 7, 15, 24 and 25 where undefined traction noise indices were used Doc. Rev

4 Testing emission DLL s CNOSSOS Source Modules The summation of Lw from Lr and LH was changed from energetic to numeric according to the DGMR documentation ( N002-05) On 4 th of July 2014, the summation of Lw from Lr and LH was re-changed to numeric and the corresponding documentation was changed as well ( N002-07). Also a misinterpretation of the input dimension for number of switched was corrected for the DLL as well as for the testing software. Industrial Noise DLL The DLL did not respond to individual source descriptors but did deliver emission values for the 1 st source index only The time correction was subtracted instead of being added to the source level Summary The testing of the source emission DLL s did lead to corrections in the documentation as well as in the software code and underlying XML files. Final testing results for road- and railway source noise showed identical results for total sound power level for all test cases within a tolerance of 0.01 db. CNOSSOS_INDUSTRIAL_NOISE_DLL.dll did return expected source emission levels in accordance with the data base used by the DLL. All DLL s showed stabile response when being linked to external software. Dortmund, the 7 th of July, Dipl.-Ing. H. Stapelfeldt 4 Doc. Rev

5 Appendix Details of Road Noise Testing SIG used the basic xml input file implemented as part of the CNOSSOS Road Source Module, namely CNOSSOS_Road_Input.xml, for verification. This xml file reflects the parameters contained within the CNOSSOS-EU Road database developed under the project The content of CNOSSOS_Road_Input.xml was used to check result output of CNOSSOS_DLL_CONSOLE.EXE, internally calling CNOSSOS_ROADNOISE_DLL.DLL, against results created by manually applying the formulae in [1], chapter 3. This evaluation was undertaken using a variety of approaches including manually, by AWK scripts and by short FORTRAN programs. In the initial phase of testing the input data file values CNOSSOS_Road_Input.xml were varied depending on the formulae and tables set out below. Input parameter values were preferably selected to match with skip values in the formulae. Rolling Noise The formulae for rolling noise are given in chapter III.2.3 of [1]. The formulae mentioned in the sub chapters following are meant with respect to chapter III.2.3 of [1]. For all parameters not explicitly mentioned we assume the settings of XML file CNOSSOS_Road_Input.xml to be valid for the evaluation verification. The following tables describe the changes with respect to XML file CNOSSOS_Road_Input.xml Rolling noise tyre correction This term is only applied on vehicles of category m=1 of table III.1 (light motor vehicles). It is evaluated using formulae (III-7), (III-8) and (III-9). The verification has been done for following parameters (in notation of the XML file): 40, 50, 60, 70, 80, 90, 100 km/h FStud 0.5 Q 1000 Tstudded 4 Frequencies 63, 125, 250, 500, 1000, 8000 Hz Doc. Rev

6 Testing emission DLL s CNOSSOS Source Modules and 70 km/h FStud 0.2 Q 1000 Tstudded 2 Frequencies 63, 125, 250, 500, 1000, 8000 Hz Temperature correction This term applies only for categories 1, 2 and 3. It corresponds to formula (III-10) with K=0.08 db/ C for category 1 and K=0.04 db/ C for category 2 and 3. It has been verified for temperature 15 C and categories 1, 2 and 3. Rolling noise acceleration correction The noise acceleration correction term depends on a distance parameter referring to the next junction and the type of junction which is given in (III-17). It has been verified for Speedariations Distance 20, 50, 70,100, 120 m Speedariations Type 1, 2 Rolling noise correction This term refers to formula (III-19).It depends on the category, frequency, velocity and the road surface. The evaluation was done for all categories and frequencies and for special velocities and road surfaces given in the following tables. Surface Ref 1, 2, 3 v 70 km/h Surface Ref 1 v 40, 90 km/h 6 Doc. Rev

7 Total rolling noise correction This is the sum of the terms Rolling noise tyre correction, Temperature correction, Rolling noise acceleration correction and Rolling noise correction mentioned above and corresponds to formula (III-6). It has been verified for XML input file CNOSSOS_Road_Input.xml. Besides we checked the sums for the following parameter changes with respect to CNOSSOS_Road_Input.xml: Speedariations Distance 20, 70 Speedariations Type 1, 2 40, 50, 60, 80, 90 Surface Ref 2, 3 Total rolling noise factor This is the sum of the first two terms of the right hand side of (III-5). It has been evaluated and verified for the settings of XML input file CNOSSOS_Road_Input.xml and for parameter v changed from 70 to 40 and 80 km/h. Total rolling noise This is the complete formula (III-5) and has been verified for XML file CNOSSOS_Road_Input.xml and the changes mentioned in item Total rolling noise correction. Propulsion Noise We have the same conventions with respect to the notation as for Rolling noise above. Propulsion noise acceleration correction This term refers to (III-18) and has been verified for the same cases as mentioned under Rolling noise acceleration correction. Doc. Rev

8 Testing emission DLL s CNOSSOS Source Modules Propulsion noise gradient correction This term depends on the velocity, category and the slope and is given in formula (III-13), (III- 14), (III-15) and (III-16). We checked it in the following cases for m=1, 2, 3 and all frequencies: Slope 10 % 70, 80 km/h Slope -2% 70 km/h Slope -8 % 10, 20, 70, 80 km/h Slope 12 % 70 km/h Propulsion noise road correction This term refers to (III-20) and has been verified for all frequencies, categories and for Surface Refs 1, 2, 3, 4a, 4b, 5. Total propulsion noise correction This term is the sum of Propulsion noise acceleration correction, Propulsion noise gradient correction and Propulsion noise road correction and refers to formula (III-12). We checked this term for the cases mentioned under Total rolling noise correction. Total rolling noise factor This term is the sum of the first two terms of the right hand side of formula (III-11). The verification has been done for cases mentioned in Total rolling noise factor mentioned above. 8 Doc. Rev

9 Total propulsion noise This term is the sum of the terms "Total propulsion noise correction" and "Total propulsion noise factor" and refers to (III-11). It has been verified for the cases mentioned in Total rolling noise correction above. Total Lwwqline_im factor This term is the second term in the sum on the right hand side of (III-1), it depends on the traffic flow Q and the velocity of the category under investigation. We verified it for the settings of XML file CNOSSOS_Road_Input.xml and for special values of Q and changed with respect to this input file. For selected categories these changes were: 40, 50, 60, 80, 90, 100 km/h Q 500, 1500, 2000 Q , 80 km/h Total result This term is the left hand side of (III-1) and has been verified for XML file CNOSSOS_Road_Input.xml. We varied these settings and changed the following parameters for selected categories: Speedariations Distance 20 Speedariations Type 1, 2 C 40, 80 km/h Cumulated total result For a given frequency this is the energetic sum of the values Total result of all categories (1, 2, 3, 4a, 4b, 5) and has been verified for XML file CNOSSOS_Road_Input.xml. Doc. Rev

10 Testing emission DLL s CNOSSOS Source Modules Result comparison per single vehicle category in XLS file In a final extra test a large number of emission data resulting from variations of input parameters per single vehicle category was checked against emission values calculated by separately programmed alternative software. The full list of results is given in file CNOSSOS_ROAD_EMISSION_TEST.XLSX which is reported alongside this note in the project Final Report. An extract of the reporting spreadsheet is set out in Figure 2. Header content and parameter variations are explained in the tables underneath the figure. To reduce the number of combinations some parameters are linked, i.e. a fixed combination of input values is used. Figure 1 Extract of road source module test results 10 Doc. Rev

11 ariation of vehicle data XLS header name Descriptor ariations of content DLL /Test method DLL bases emission calculation Test emission results when using alternative software CAT ehicle category 1 2, 3, 4a, 4b Q Q vehicle per h 500, 1000, 5000 ehicle speed (km/h) 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130 FS FStud studded tyres (%/100) 0, 0.5 ariation of road data XLS header name Descriptor ariations of content SRF Surface Ref NL01, NL02,,NL13 TEM Temperature ( C) -5, 2.5, 15, 30 MST Tstudded usage (month) 0, 1, 1.5 GRD SLOPE Gradient (%) -15, -5, 0, 5, 15 DJU Speedariations Distance Distance crossing (m) 200, 50, 10, 50, 200 TJU Speedariations Type. Type of crossing 1, 2 Cases of linked attributes 1. CAT Q FS 2. TEM MST 3. GRD DJU TJU Doc. Rev

12 Testing emission DLL s CNOSSOS Source Modules Details of railway noise testing Testing of the railway source module utilized the underlying XML database which is used by the external software tool as well as by CNOSSOS_ROADNOISE_DLL.DLL. The xml database reflects the design of the railway database delivered by the project. A total of 39,744 different test case scenarios were generated by systematically combining a range of potential attribute settings as described in the tables below. For each case the octave band emission data was calculated with help of the DLL and a second time by using the external software. Octave values were summed up to a total Leq values which were compared to each other. The following parameters for rail vehicle and rail track data were set to the values given in the table below to create test case. However some of the attributes were linked to each other while defining the combinations. The full list of results is given in file CNOSSOS_RAIL_EMISSION_TEST.XLSX which is reported alongside this note in the project Final Report. An extract of the reporting spreadsheet is set out in Figure 2. Header content and parameter variations are explained in the tables underneath the figure. Figure 2. Extract of railway source module test results 12 Doc. Rev

13 ariation of the vehicle data XLS header name Descriptor ariations of content SRC Source Height A, B RI Reference ID 3 to 29 ehicle speed 30, 120, 260 RC Running condition Accelerating, Decelerating, Constant Idling IDL Idling time 0, 0, 0, 1 Q Number of vehicles 1, 10 ariation of track data including geometry relation to the receiver position XLS header name Descriptor ariations of content TCR Curve Radius 250, 500, 750, 1000, 1500, 2000, 2500, 3000, 5000 TTTI Track Transfer ID 1, 2, 3, 4, 5, 6, 7, 8, 9 TSTI Structure Transfer ID 1, 2, 3, 1, 2, 3, 1, 2, 3 TRRI Rail Roughness ID 1, 2, 3, 4, 1, 2, 3, 4, 1 TINI Impact Noise ID 1, 2, 3, 1, 2, 3, 1, 2, 3 TBCI Bridge Constant ID 1, 2, 3, 4, 1, 2, 3, 4, 1 THA Horizontal Angle -45, -45, 90, 135 TA ertical Angle -45, 45, 0, 45 Cases of linked Attributes 1. RC IDL 2. THA TA 3. TCR TTTI TSTI TRRI TINI TBCI Doc. Rev