Evaluation and Comparison of Roadside Crash Injury Metrics (2023)

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.

172 12 Compare FSM and Alternate Metrics in Roadside Crash Tests Introduction and Objective This chapter summarizes the available MASH crash tests, MASH test sample selection, and analysis of the available electronic MASH crash test data for the sample tests. The candidate injury metric values computed for this sample of MASH crash tests were ultimately used to evaluate the implications of various proposed changes to the existing MASH occupant risk procedures. Methods The overall approach was to determine the available MASH tests, select a sample of these tests considering relevant factors, and analyze the associated electronic crash data for the sample tests. Additional details are provided in the sections below. 12.2.1 Determine Available MASH Full-Scale Crash Tests and Associated Characteristics There are three primary crash testing facilities in the U.S. that conduct full-scale crash tests of roadside safety hardware: MwRSF, TTI, and FHWA FOIL. The research team contacted these facilities regarding the availability of existing MASH roadside hardware crash test data, and all three test facilities agreed to provide electronic data for a selected number of MASH tests. The goal was to determine the total number of MASH tests that have been conducted by each facility and obtain summary data for these tests to facilitate the selection of an appropriate subset to assess the implications of any potential changes to the existing MASH criteria. The MwRSF maintains an online, publicly available database of all crash test reports (UNL MwRSF 2020). This database was used to develop a list of MASH tests of non-proprietary roadside hardware conducted by MwRSF. For TTI and FOIL, there is no analogous publicly available database of conducted crash tests. The research team contacted staff members of these two facilities directly, and each facility was able to provide an Excel spreadsheet containing summary data for available crash tests. Note that the information available in the spreadsheets varied, both across testing facilities and (in some cases) across tests from the same facility. Both spreadsheets contained data for MASH and pre-MASH crash tests. Using the information available in each spreadsheet, the research team identified and isolated the relevant MASH tests. In addition to the total number of available MASH tests, the research team collected data describing each test. Data collected included test vehicle type, the type of roadside safety hardware device tested, test outcome (pass/fail), and the reported MASH occupant risk values. For the MwRSF tests, the available PDF test reports were accessed to collect this information manually. The TTI spreadsheet contained data elements corresponding to test vehicle type, test outcome, roadside hardware device tested, and the reported occupant risk values. The FOIL spreadsheet contained data on test vehicle type and roadside hardware device tested, but no data on test outcome or reported occupant risk values. Data on the test outcome and reported occupant risk values were obtained with the requested electronic test data. The reported occupant risk values are summarized by roadside hardware device type and test vehicle type.

173 12.2.2 Pilot MASH Crash Test Data and Computational Validation Procedure Electronic sensor data for a small number of MASH crash tests were first obtained from each of the three primary test facilities. This pilot data were used develop analysis methods to assess the implications of any potential changes to the existing MASH criteria, including developing/adapting code to batch process the available data, as data formats differed by test facility. The pilot data were also used to validate our computation procedure for the existing occupant risk metrics for roadside crash tests. We developed our own computation procedures since we were computing metrics beyond those computed by TRAP (i.e., OLC and VPI). For each pilot test, occupant risk values were computed using our analysis methods and, using percent difference, compared to the reported occupant risk values in the corresponding test reports. As the TTI tests and a majority of the FOIL tests used TRAP to compute the reported occupant risk values, TRAP was not used to recompute all the pilot test occupant risk values. TRAP, however, was used selectively in cases where discrepancies in computed and reported values were present. 12.2.3 MASH Crash Test Sample Selection Process The research team selected a 25% sample of the available MASH crash test cases for use in this analysis. Sample cases were selected considering the aspects summarized in Table 12-1. Table 12-1. Summary of MASH test selection considerations. MASH Test Selection Aspect Details Data Availability Sensor data for the crash test must be available electronically for the case to be included. Data Quality Available sensor data will be checked for signal quality issues prior to use in the analysis. Any signal errors present will be remedied (if possible) prior to use or the case will be excluded from the analysis. Roadside Hardware Device Type Crash tests of a variety of roadside hardware devices will be included. Device categories include metal beam barrier, concrete barrier, cable barrier, end terminals, transitions, temporary barrier, bridge rails, and work zone devices. To the extent possible based on data availability, an effort will be made to include a range of different test devices within each of the categories (e.g., a w-beam, box beam, and thrie beam test for the metal beam barrier category). Test Outcome Both passing and failing MASH tests will be considered, including those that did not meet the current MASH occupant risk criteria. The reason for the failed test will be evaluated on a case-by-case basis to determine whether the test should be included or not (e.g., a test failing due to vehicle rollover is not likely to be useful for computation of the injury risk metrics). Occupant Risk Values The sample of tests should have a range of occupant risk values and ideally one that approximates the range of occupant risk values for the entire population of MASH crash tests. Test Facility While we do not expect large variations across test facilities, tests from all three test facilities will be included. The research team presumed that electronic data were available for all identified MASH tests from each of the crash test facilities. Any tests with missing electronic data were documented, and

174 replacement tests selected if there was a significant number of missing data cases. Data quality was verified as part of the analysis process, and any excluded tests documented. We, however, expected a small number of tests, if any, to be excluded for data quality issues. Cases were primarily selected to fulfill the roadside hardware device type, test outcome, occupant risk value, and test facility aspects described in Table 12-1. 12.2.4 MASH Crash Test Sample: Computations and Preliminary Analysis The collected electronic sensor data for the selected MASH tests were used to compute the six metrics listed below: • OIV • RA • MDV • ASI • VPI • OLC Unlike the NHTSA full-scale vehicle crash tests, crash tests with roadside hardware are routinely conducted at oblique angles. OIV, RA, and ASI include at least the lateral and longitudinal vehicle accelerations within the computation procedure, so the computation of these metrics was not altered. As MDV, VPI, and OLC do not inherently include both the lateral and longitudinal vehicle accelerations, these metrics were computed separately for the lateral and longitudinal directions. The computation of all the metrics was completed using MATLAB software (Version 2020a). Sensor data provided by TTI and MwRSF were already filtered using an SAE J-211 four-pole Butterworth filter with channel frequency class (CFC) 180; these data were used directly for the computation of the metrics. The sensor data obtained from the FOIL tests were filtered to CFC 180 prior to using MATLAB to compute the metrics. The existing roadside metrics (OIV, RA, and ASI) were computed according to the MASH procedures. For the other metrics (MDV, VPI, OLC), the longitudinal values were computed using the validated procedures previously described (Section 5.2.1). For the lateral direction, the computation of MDV, VPI, and OLC was identical to that used for the longitudinal direction but used the lateral vehicle acceleration data in place of the longitudinal vehicle acceleration data. Since there was a relatively small number of failing MASH tests, all the available MASH crash test data were converted to the appropriate format for analysis, and all six metrics were computed for each test. The ranges of computed metric values for the available tests were summarized via scatter plots. Where appropriate, existing MASH preferred and maximum limits are indicated, such as for OIV and RA. Future work will include verifying the pass/fail status of each test (a number of the TTI tests did not indicate test outcome) and deciding whether a failing test should be included or excluded based on the reason for the failure.

175 Results 12.3.1 Available MASH Full-Scale Crash Tests Table 12-2 summarizes the available MASH crash test data for MwRSF, FOIL, and TTI by roadside hardware device type tested. Table 12-2. Summary of non-proprietary MASH tests conducted by device type and test facility. Roadside Hardware Device Type Test Facility MwRSF FHWA FOIL TTI Metal Beam Barrier 39 2 31 Concrete Barrier 5 0 7 Cable Barrier 20 0 1 End Terminal 10 8 22 Transition 14 0 18 Temporary Barrier 21 0 15 Bridge Rail 3 0 40 Work Zone Device 5 0 18 Other 8 0 81 Total 125 10 233 Note that the values in Table 12-2 exclude any crash tests with proprietary roadside hardware devices. Also, the large number of “other” roadside hardware tests present in the available TTI data are predominately sign/luminaire supports (41 tests), mailboxes (24 tests), and gates (seven tests). As expected, the vast majority of MASH tests have been conducted by MwRSF and TTI. A subset of the tests in Table 12-2 will be selected and used for analysis to assess the implications of any potential changes to the existing MASH criteria. 12.3.2 Available MASH Full-Scale Crash Test Characteristics Table 12-3 summarizes the occupant risk values for the available MASH tests, combining the tests from TTI, FOIL, and MwRSF. For each occupant risk metric, the mean and ranges are reported based on the n available MASH tests that reported that metric. In addition, the associated threshold value is reported for each metric for comparison purposes. While all of these U.S. MASH tests use the OIV and RA values to evaluate occupant risk, most MASH tests also report the analogous European metrics (THIV, PHD, and ASI), so those metrics have been included in Table 12-3 as well. Table 12-3: Summary of occupant risk parameter values for available MASH tests by testing facility. Test Facility Parameter n (TTI / MwRSF / FOIL) Mean Minimum / Maximum Threshold Value TTI, MwRSF, and FOIL Longitudinal OIV [m/s] 226 (103 / 113 / 10) 5.16 0.2 / 21.9 12 Lateral OIV [m/s] 222 (99 / 113 / 10) 4.60 0 / 15.1 Longitudinal RA [G] 226 (104 / 112 / 10) 8.33 0.1 / 39.7 20 Lateral RA [G] 7.23 0.1 / 20.41 THIV [m/s]* 218 (104 / 106 / 8) 6.92 0.28 / 23.75 9.17 PHD [G]* 9.88 0.1 / 42.9 20 ASI* 187 (112 / 67 / 8) 0.91 0.04 / 2.96 1.0/1.4/1.9 *Note: These metrics are reported but not used in evaluating MASH tests

176 Note that not all of the available TTI, FOIL, and MwRSF tests had occupant risk values reported. Since the MwRSF test occupant risk values were obtained from publicly available crash test reports, the vast majority of these tests have occupant risk values reported. Similarly, the FOIL tests have nearly all the occupant risk values reported, as FOIL provided test summary reports for all available MASH tests. Table 12-3 excludes any MASH tests with heavy vehicles where occupant risk values were reported since the occupant risk values are only required for the passenger vehicle MASH tests. The number of heavy vehicle tests present in the data, however, was small (i.e., 20 of the TTI tests and two of the MwRSF tests). Both the mean OIV and RA values in Table 12-3 are well below the corresponding maximum threshold values. Examining the occupant risk ranges, there are tests with values in excess of the corresponding threshold values. This is especially true for the THIV, PHD, and ASI metrics. Note that Table 12-3 includes data for both passing and failing MASH tests. Table 12-4 summarizes the available occupant risk values by test vehicle type. Again, both passing and failing MASH tests are included. Although the small passenger car has traditionally been considered the critical case in terms of occupant risk, the small passenger car (1100C) occupant risk average values in Table 12-4 are not always higher than the corresponding large pickup truck (2270P) values. Table 12-4: Summary of occupant risk parameter values for available MASH tests by test vehicle. Test Vehicle Parameter n (TTI / MwRSF / FOIL) Mean Minimum / Maximum Threshold Value Small Passenger Car (1100C) Longitudinal OIV [m/s] 78 (34 / 40 / 4) 6.13 0.3 / 21.9 12 Lateral OIV [m/s] 75 (31 / 40 / 4) 3.90 0.0 / 12.40 Longitudinal RA [G] 77 (34 / 39 / 4) 8.11 0.1 / 25.55 20 Lateral RA [G] 6.02 0.2 / 14.70 THIV [m/s]* 76 (35 / 37 / 4) 7.33 0.67 / 17.40 9.17 PHD [G]* 9.15 0.3 / 23.79 20 ASI* 70 (37 / 29 / 4) 0.87 0.04 / 2.94 1.0/1.4/1.9 Large Pickup Truck (2270P) Longitudinal OIV [m/s] 148 (69 / 73 / 6) 4.65 0.2 / 19.0 12 Lateral OIV [m/s] 147 (68 / 73 / 6) 4.95 0 / 15.1 Longitudinal RA [G] 149 (70 / 73 / 6) 7.81 0.1 / 39.7 20 Lateral RA [G] 7.86 0.1 / 20.41 THIV [m/s]* 142 (69 / 69 / 4) 6.70 0.28 / 23.75 9.17 PHD [G]* 10.27 0.1 / 42.9 20 ASI* 117 (75 / 38 / 4) 0.93 0.05 / 2.96 1.0/1.4/1.9 *Note: These metrics are reported but not used in evaluating MASH tests 12.3.3 MASH Test Selection Using the available data from each of the three crash test facilities and the selection aspects detailed in Table 12-1, the research team has selected a minimum sample of 90 MASH tests for analysis. The vast majority of the tests in the minimum sample are indicated as a “passing” test (77 of 90) with the remaining having missing values. The distribution of test facility and roadside hardware device for the selected tests are shown in Table 12-5.

177 Table 12-5. Summary of selected non-proprietary MASH tests by device type and test facility. Roadside Hardware Device Type Test Facility MwRSF FHWA FOIL TTI Metal Beam Barrier 10 1 7 Concrete Barrier 4 0 1 Cable Barrier 5 0 1 End Terminal 3 7 9 Transition 4 0 5 Temporary Barrier 6 0 1 Bridge Rail 2 0 10 Work Zone Device 2 0 2 Other 2 0 8 Total 38 8 44 The selected passing (or no test outcome indicated) tests represent approximately 25% of the available MASH tests, and there is a wide distribution of tested device type. For the tests with reported occupant risk values, the ratio of the reported value to the associated threshold value was computed and used to aid in the test selection. As a general rule, the research team first selected all tests where one or more occupant risk values were 80% of the associated threshold or more prior to selecting other cases for the sample. In addition to the minimum sample of 90 MASH tests, there is a total of 37 tests that are indicated as “failing”; these tests will be evaluated on an individual basis prior to including (or excluding) from the sample. At a minimum, we planned to include all MASH tests that fail only due to one or more of the occupant risk criteria, but there may be other failed tests that could be included in the sample. Table 12-6 summarizes the distribution of the reported MASH occupant risk metrics for the minimum MASH test sample (e.g., only “passing” / no outcome indicated tests) as well as a sample that includes all of the failing tests. In general, the mean occupant risk values are slightly greater than those for all available MASH tests, and the minimum sample data including all the failing MASH tests cover essentially the full range of occupant risk values present in the available data. Table 12-6: Summary of occupant risk parameter values for minimum MASH test sample. Test Type Parameter Mean Minimum / Maximum Passing or Missing indication (n = 90) Longitudinal OIV [m/s] 5.32 0.73 / 11.3 Lateral OIV [m/s] 5.46 0.09 / 12.4 Longitudinal RA [G] 9.17 0.7 / 20.34 Lateral RA [G] 8.29 0.43 / 20.41 Including Failing Tests (n = 127) Longitudinal OIV [m/s] 5.57 0.73 / 19 Lateral OIV [m/s] 5.20 0 / 15.1 Longitudinal RA [G] 9.94 0.7 / 39.7 Lateral RA [G] 8.61 0.43 / 20.41 12.3.4 Validation of Occupant Risk Values for Pilot MASH Tests The pilot MASH tests data included electronic sensor data for 20 roadside crash tests as summarized in Table 12-7. Since there was such a large number of TTI and MwRSF tests, only five from each test lab were initially selected. As FOIL only had a small number of MASH tests available, we obtained electronic data for all 10 of the available MASH tests.

178 As expected, the provided electronic data varied in format with the MwRSF data generally provided in Excel format and the TTI and FOIL data typically in ASCII / comma-delimited text files with headers. All relevant data were stripped of any associated header information and converted to a CSV format prior to processing using MATLAB. Sensor data provided by TTI and MwRSF were already filtered using an SAE J211 CFC 180 four-pole Butterworth filter. The FOIL sensor data were filtered in the same manner prior to computing the occupant risk metrics. The primary occupant risk values (i.e., OIV, RA, and ASI) were computed for each test and compared to the corresponding values reported by each test facility. For the TTI, MwRSF, and the first two FOIL tests, the comparison values were obtained from provided or publicly available MASH crash test report summary sheets. For the remainder of the FOIL tests, FOIL provided a summary document for each test that included a copy of the TRAP program output. Table 12-7. Summary of pilot MASH crash test data obtained. Test Facility Test Number / Designation MASH Test Designation Device Type TTI 476460-1-4 3-11 Rigid Barrier/Bridge Rail 478730-2 3-10 Cable Barrier 469467-7-1 3-62 Luminaire 490026-3-1 3-33 End Terminal 490022-6 3-20 Transition MwRSF 2214NJ-2 4-12 Rigid Barrier 4CMB-4 3-10 Cable Barrier MGSCO-2 3-11 Metal Beam Barrier NYBBT-3 3-34 End Terminal WITD-1 3-11 Temporary Barrier FOIL 16010 3-11 Metal Beam Barrier 16015 3-11 Metal Beam Barrier 17004 3-32 Metal Beam Barrier/Terminal 17007 3-32 Metal Beam Barrier/Terminal 17010 3-33 Metal Beam Barrier/Terminal 17012 3-36 Metal Beam Barrier/Terminal 18003 3-31 Metal Beam Barrier/Terminal 19008 3-31 Metal Beam Barrier/Terminal 19010 3-32 Metal Beam Barrier/Terminal 19012 3-33 Metal Beam Barrier/Terminal Table 12-8 summarizes the longitudinal and lateral OIV magnitudes for the pilot MASH tests. All of the computed OIV values were within 6% of the reported values, and the vast majority were within 2% of the reported value. Note that some of the tests had OIV values reported in feet per second; these were converted to meters per second prior to reporting them in Table 12-8. In addition, the precision of the computed values was often greater than the precision of the reported OIV values and, in many cases, rounding the computed value to the same precision of the reported OIV value would result in no error present. For the FOIL tests 19008, 19010, and 19012, the TRAP output provided was incorrectly based on positive longitudinal acceleration (i.e., the vehicle was accelerating instead of decelerating). The TRAP program was rerun with the corrected longitudinal acceleration values to produce the output shown in Table 12-8. Table 12-9 summarizes the longitudinal and lateral RA magnitudes for the pilot MASH tests. Similar to the OIV, all of the values were within 9% and a vast majority were within 1% of the

179 reported value. The three FOIL tests with incorrect longitudinal acceleration (noted previously) show the revised RA values based on TRAP output. In addition, a fourth FOIL test (17012) reported a lateral RA of 5.6 G while our procedures computed 4.79 G with the available sensor data. The TRAP program was run on the sensor input data we used to generate the lateral RA of 4.79 G and TRAP produced 4.8 G (shown in Table 12-9). One possibility is that we have incorrect y-direction acceleration data for this test. Nevertheless, our procedures produced nearly the same RA produced by TRAP using the same input data. Table 12-8. Summary of computed and reported OIV values for MASH pilot tests. Test Facility Test Number / Designation Longitudinal OIV [m/s] Lateral OIV [m/s] Computed Reported % Difference Computed Reported % Difference TTI 476460-1-4 4.32 4.30 0.47 9.21 9.20 0.11 478730-2 6.68 6.71 0.45 0.42 0.4 5.0 469467-7-1 0.77 0.79 2.53 0.02 0 N/A 490026-3-1 7.2 7.19 0.14 2.12 2.1 0.95 490022-6 6.42 6.4 0.31 8.39 8.41 0.24 MwRSF 2214NJ-2 1.88 1.99 5.53 4.13 4.15 0.48 4CMB-4 6.58 6.58 0.00 3.17 3.17 0.00 MGSCO-2 6.25 6.25 0.00 6.29 6.29 0.00 NYBBT-3 5.21 5.23 0.38 4.44 4.44 0.00 WITD-1 4 4.01 0.25 6.21 6.21 0.00 FOIL 16010 4.77 4.8 0.63 3.76 3.8 1.05 16015 4.70 4.7 0.00 4.22 4.2 0.48 17004 11.4 11.4 0.00 0.66 0.7 5.71 17007 11.8 11.8 0.00 0.04 0 N/A 17010 7.50 7.5 0.00 1.28 1.3 1.54 17012 6.10 6.1 0.00 5.12 5.1 0.39 18003 10.00 10 0.00 0.47 0.5 6.00 19008 9.05 9.1* 0.55 2.33 2.3* 1.30 19010 11.83 11.8* 0.25 1.03 1.0* 3.00 19012 7.77 7.8* 0.38 2.07 2.1* 1.43 *Note: These reported values differ from those reported by the test facility. Table 12-9. Summary of computed and reported RA values for MASH pilot tests. Test Facility Test Number / Designation Longitudinal RA [G] Lateral RA [G] Computed Reported % Difference Computed Reported % Difference TTI 476460-1-4 5.62 5.6 0.36 9.64 9.6 0.42 478730-2 6.13 6.1 0.49 3.99 4 0.25 469467-7-1 0.84 0.8 5.00 0.65 0.6 8.33 490026-3-1 6.33 6.3 0.48 4.96 5 0.80 490022-6 6.13 6.1 0.49 6.32 6.3 0.32 MwRSF 2214NJ-2 22.36 22.39 0.13 8.83 8.84 0.11 4CMB-4 6.405 6.4 0.08 6.307 6.31 0.05 MGSCO-2 12.69 12.73 0.31 10.36 10.31 0.48 NYBBT-3 9.02 9.02 0.00 6.53 6.53 0.00 WITD-1 6.71 6.66 0.75 20.48 20.41 0.34 FOIL 16010 6.30 6.1 3.28 6.05 6 0.83 16015 5.91 5.9 0.17 6.72 6.7 0.30 17004 13.61 13.6 0.07 5.50 5.5 0.00 17007 16.83 16.8 0.18 5.82 5.8 0.34 17010 8.84 8.8 0.45 4.79 4.8* 0.21 17012 9.14 9.1 0.44 6.22 6.2 0.32 18003 17.56 17.5 0.34 4.25 4.2 1.19

180 19008 11.99 11.9* 0.76 6.56 6.5* 0.92 19010 15.05 14.9* 1.01 6.33 6.3* 0.48 19012 6.20 6.2* 0.00 5.12 5.1* 0.39 *Note: These reported values differ from those reported by the test facility. Table 12-10 summarizes the ASI values for the pilot MASH tests. All of the computed ASI values were within 8% of the reported values, and most were within 2% of the reported value. Two of the MwRSF tests and two of the FOIL tests did not report the ASI values (denoted with a +); for these tests, the TRAP program was used to compute the ASI value shown in Table 12-10. Table 12-10. Summary of computed and reported ASI values for MASH pilot tests. Test Facility Test Number / Designation ASI Computed Reported % Difference TTI 476460-1-4 1.85 1.85 0.00 478730-2 0.86 0.86 0.00 469467-7-1 0.13 0.14 7.14 490026-3-1 0.78 0.83 6.02 490022-6 1.93 1.92 0.52 MwRSF 2214NJ-2 0.60 0.61+ 1.64 4CMB-4 0.733 0.73 0.41 MGSCO-2 1.11 1.11 0.00 NYBBT-3 0.85 0.83+ 2.41 WITD-1 1.32 1.32 0.00 FOIL 16010 0.56 0.53+ 5.66 16015 0.60 0.60+ 0.00 17004 1.27 1.27 0.00 17007 1.30 1.3 0.00 17010 0.584 0.58 0.69 17012 0.845 0.85 0.59 18003 1.022 1.02 0.20 19008 0.86 0.82 4.88 19010 1.206 1.21 0.33 19012 0.67 0.68 1.47 + Note: No corresponding values reported by test facility. These values computed using the TRAP program. 12.3.5 Computation of Metrics for Sample MASH Tests The validated MATLAB code was used to compute the occupant risk metrics for all the available MASH sample tests. Note that valid data were only available for 124 MASH tests. The data provided for three of the tests were either missing or repeated data from a different test designation and have been omitted. A graphical summary of the computed values is shown in Figure 12-1 through Figure 12-5. All the plots combine the lateral and longitudinal values for the same metric, and the corresponding MASH limits are shown for OIV and RA. Note that the ASI metric is not included in these plots as this metric provides a single value representative of the entire crash and not lateral and longitudinal directions independently. For simplicity, the magnitudes of the OIV, MDV, and OLC are shown; these values are typically negative. Table 12-11 provides a numerical summary of all the computed values. Once the potential MASH modification options were developed, these computed values were used to determine the impact of any proposed modifications to existing MASH occupant risk procedures on the resulting test outcome, that is, pass or fail (see Chapter 13).

181 Figure 12-1. Longitudinal and lateral occupant impact values in sample MASH tests (n = 124). Figure 12-2. Longitudinal and lateral RA values in sample MASH tests (n = 124). 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 La te ra l O IV [m /s ] Longitudinal OIV [m/s] Sample MASH Tests MASH Upper Limit MASH Preferred Limit 0 5 10 15 20 25 0 5 10 15 20 25 30 35 40 45 La te ra l R A [m /s ] Longitudinal RA [m/s] Sample MASH Tests MASH Upper Limit MASH Preferred Limit

182 Figure 12-3. Longitudinal and lateral OLC values in sample MASH tests (n = 124). Figure 12-4. Longitudinal and lateral MDV values in sample MASH tests (n = 124). 0 5 10 15 20 25 30 0 5 10 15 20 25 La te ra l O LC [G ] Longitudinal OLC [G] 0 5 10 15 20 25 30 0 5 10 15 20 25 30 35 La te ra l D V [m /s ] Longitudinal DV [m/s]

183 Figure 12-5. Longitudinal and lateral VPI values in sample MASH tests (n = 124). Table 12-11. Summary of computed occupant metric values for sample MASH tests. Metric Min / Max Average Median 5th Percentile 25th Percentile 75th Percentile 95th Percentile Longitudinal OIV 0.18/12.58 5.72 5.42 1.88 4.40 6.62 11.05 Lateral OIV 0.001/12.44 4.71 4.61 0.18 2.75 6.57 9.36 Longitudinal RA 1.0 / 39.62 8.94 7.72 2.10 5.77 10.51 18.60 Lateral RA 0.21 /20.48 7.48 6.83 1.14 5.37 9.43 15.23 Longitudinal OLC 0 / 19.13 6.23 5.28 0.54 3.89 8.04 14.42 Lateral OLC 0 / 25.49 5.76 4.20 0.03 1.83 8.47 14.92 Longitudinal DV 0.77 / 30.21 13.36 12.23 3.32 8.82 16.81 27.57 Lateral DV 0.12 / 25.13 11.56 12.22 1.08 8.42 15.06 20.18 Longitudinal VPI 0 / 6344.8 700.1 384.0 139.1 228.2 811.1 1901.0 Lateral VPI 0 / 3977.6 484.0 344.8 72.0 164.5 601.2 1367.4 ASI 0.02 / 2.94 1.07 0.93 0.22 0.62 1.39 2.05 One relevant outcome of these computations was the finding that both MDV and VPI values are sensitive to the analysis window selected. Many MASH tests record data for multiple seconds, and many up to 5 or 10 seconds surrounding the crash event. These long durations mean that the sensors can include data when the test facilities are applying the brakes to stop the vehicle after the crash event. When using the full extent of electronic sensor data available for all tests, 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 1000 2000 3000 4000 5000 6000 7000 La te ra l V PI [m /s 2 ] Longitudinal VPI [m/s2]

184 approximately half of the tests had very large MDV and VPI values. The research team has examined these cases and appropriately truncated the data to include only the data relevant to the crash event. These revised values are reflected in the data shown in Figure 12-1 through Figure 12-5 as well as in Table 12-11. While in many cases it is relatively clear when the crash event ends, adopting a metric that is sensitive to the analysis window selected could introduce additional subjectivity into the analysis procedures. In contrast, the other metrics (OIV, RA, ASI, OLC) were insensitive to the analysis window selected. The values of these metrics changed on average less than 0.01% when computed using the full extent of electronic sensor data or the truncated electronic sensor data. Conclusions A sample set of 124 MASH full-scale crash tests was selected from the three primary roadside safety crash test facilities. The crash tests were selected to represent a variety of tested roadside hardware devices as well as a range of possible occupant risk values, skewed toward the higher occupant risk values. A set of 20 pilot MASH crash tests was used to validate the occupant risk computations as this effort required computation of alternate metrics not included in the standard occupant risk software (i.e., TRAP). There was good agreement between the computed values and the test facility reported values for the pilot tests. The same computation procedures were then used to compute all the candidate injury metric values for all 124 tests. These values will be used to evaluate any implications of any potential changes to the MASH occupant risk criteria. Based on the computations for the candidate injury metrics, the MDV and VPI metrics were found to be sensitive to the analysis window selected. As a result, using the MDV or VPI metric in the occupant risk procedures would introduce an additional amount of uncertainty in the computed values. In contrast, the other metrics (OIV, RA, ASI, OLC) were insensitive to the analysis window selected.