PM samplers must sample continuously until 10 seconds after the end of the WLTP, as particles require 10 seconds to reach the filter after the last brake event. PN calculation does not completely cover all PN of the trips’ last brake event due to time delay and smoothing during the calculation period. Based on ILS data, extending the TBF export by 5 seconds would result in maximum emission factor differences of approximately 0.35 percent, estimated below 1 percent overall influence.

Figure 12.1 shows the indicative setup of the PM sampling system with sampling plane, sampling nozzle, PM sampling probe, PM cyclonic separator, PM sampling line, PM filter holder, PM sampling filter, mass or volumetric flow measurement, pump, and flow controller. Figure 12.3 shows the indicative setup of the PN sampling and measurement system with sampling plane, sampling nozzle, PM sampling probe, flow splitter, PN particle transfer tube, PN cyclonic separator, HEPA-filtered inlet for dilution air, dilution stage, Volatile Particle Remover, Particle Number Counter with internal flow controller and mass or volumetric flow measurement, and pump.

Proposal to add quality checks for applied brake torque to GTR 24, including methods for calculating Brake Torque Rising Criterion, Brake Torque Maximum Criterion, and Brake Torque Stability Criterion based on signals with frequency of 250 Hz or faster, to be recorded in Event-Based Files and reported in summary table 13.6, and to add signal filtering requirements to chapter 5.4 specifying a second order filter with cut-off frequency between 25% and 50% of the lower sampling or recording frequency to fulfill Nyquist-Shannon sampling theorem.

Proposal to establish quality criteria for applied brake torque datasets at 250 Hz sampling rate by calculating rising criterion, maximum criterion, and stability criterion for each brake event. Rising criterion measures time to reach ninety per cent of nominal brake torque and overshoot at brake event start. Maximum criterion calculates maximum brake torque weighted by nominal specific kinetic energy. Stability criterion reports average brake torque and root mean squared error weighted by nominal specific kinetic energy. Results show variation across laboratories. Further investigation necessary for different brake types, threshold definition for failed brake events, and inclusion of method and criteria reporting in event based file.

ILS 3 Findings and GTR 24 Updates addresses improvements to the Particle Measurement Programme. References to three-probe layout and smaller duct diameters are deleted; duct inner diameter is defined as 190–225 mm. A column defining numbers of decimals is added to Table 8.1. Bedding and emissions measurement sections shall use new parts. Temperature requirements at filter holder are clarified to remain above 15°C. Requirements on decimal places are added throughout Tables 13.1, 13.3, 13.4, 13.5, and 13.6. Table 13.6 now counts from 1 to 274. Unit notation is harmonized. Signal filtering requirements are added to Chapter 5.4. Various requests are rejected, including repositioning sampling nozzles and probes, PM₁₀ probe positioning, and OEM-declared threshold pressure values.

PMP Task Force 5 is working on HDV Brake Dyno Layout and HDV Brake Test Cycle. A GTR proposal is planned by the end of 2025.

Proposal to amend para. 8.2.(k) to require calculation of average drag torque over a 5 s period during constant speed phases rather than acceleration phases, amend para. 8.2.(d) to clarify brake run-out measurement procedures for drum brakes including verification measurement at final assembly, amend para. 8.2.(e) to specify assembly procedures for drum brake testing, insert para. 12.1.2.3.(c) to clarify flow measurement device calibration requirements for PM sampling, and insert para. 12.2.3.2.(c) to clarify flow measurement device calibration requirements for PN sampling, based on ECE/TRANS/WP.29/GRPE/2025/6.

This document evaluates the AIP dustIN multisampling PM10 system using an ILS-3 brake on a BMW X5 tested across the WLTC cycle with brake cooling airflow at 500, 1000, and 1500 m³/h. Testing involved six tests per flowrate comprising regular runs with repetitions. Deviations between multisampler and single sampling were less than 5 percent, covered by measurement uncertainty and partly caused by different sampling positions. No significant loss impact by the dustIN multisampler was identified. A PM2.5 campaign is currently running.