Document Type: PM Document Date: Item Number: TRV2011/7833A

Transcription

1 WLTP-DTP LabProc_ICE subgroup chair Secretary: Stephan Redmann Federal Ministry of Transport Building and Urban Development Robert-Schuman-Platz 1 D Bonn Trafikverket Rödavägen Borlänge Street adress: Röda vägen 1 Telephone:: Text phone: trafikverket@trafikverket.se Smeds, Peter Market and Planning Direct: Mobile: Peter.Smeds@trafikverket.se Copy to: WLTP-DTP Labproc_ICE subgroup members 1(8)

2 Road Load Determination - Vehicle preparation Authors: Peter Smeds Swedish Transport Administration Iddo Riemersma Transport & Environment Background and introduction At this moment a world-harmonized light-duty test procedure (abbreviated to WLTP) is being developed. This initiative was set up to review the current test procedures that are being used in different regions (mainly Europe, USA and Japan), and to come up with an improved draft test procedure (DTP). At the same time a test cycle will be developed that is a good reflection of driving behavior in real-life. The DTP workgroup is split in 5 subgroups, that focus on specific parts of the test procedure. Within the subgroup Lab-processes for internal combustion engines (LabProc_ICE) a further subdivision into interest groups was made. This paper specifically focuses on one of these interest groups called Road Load Determination (RLD), which involves all of the technical requirements for preparation of the vehicle before the road load tests take place. The result from RLD is then used for adjustment of the chassis dynamometer during the process of homologation of the vehicle. The starting point for this interest group is to review the current road load test procedure as laid down in ISO & 2. During the expert meeting in Brussels on 5-6 October 2010 it became apparent that the current road load test procedure has a number of omissions that may result in influencing the test results. As a consequence, the road load of production vehicles may be higher in comparison to the road load of the homologation vehicle. This has a direct effect on the fuel consumption and CO 2 emissions of a given vehicle. Some first exploratory tests have shown that CO 2 figures may be 10% too optimistic, which is one of the reasons that the officially declared fuel consumption by a manufacturer does not match the customer s experience. This situation is accelerated by the global strive for energy conservation and reduced CO 2 emissions which has led to the development of new fuel saving technologies and more competition between vehicle manufactures. To avoid penalty fees, reduce vehicle taxation and improve product competiveness vehicle manufactures may be encouraged to obtain the best possible coast down figures within these current regulatory frame work conditions. Two of the important objectives of the WLTP are to create a level playing field for manufacturers and to increase the representativity of the test cycle and measurement procedure. An improved RLD test procedure that reduces the gap between the road load of a homologation vehicle and the actual road load on a production vehicle is expected to contribute in achieving these objectives. During the expert-meeting this issue was first raised by the Swedish Transport Administration and was supported by Transport & Environment. The RLD interest group decided that STA should undertake the task to investigate this, and to produce some written input for the discussion. This investigation sums up the flexibilities in the ISO test procedure as well as the tolerances that may be stretched to the most favorable end. Also the options available to address these issues properly in the DTP should be identified. This paper will present the results of the investigation and the options for the DTP, aiming at a RLD procedure which is representative for production vehicles under real-life conditions while at the same time being repeatable and reproducible. In the current outline this paper is intended only to provide input to the discussions in the DTP expert meetings; the defining of a legislative text proposal will be part of the next phase. Justification data The aim of this document is to provide information on a factual basis, so considerable effort has been put into finding data that can substantiate the gap in road load between homologation and production vehicles. However, it proved difficult to find any recent and publicly available studies that could quantify this gap. A road load verification program at EPA that dates back to 1984 revealed that the differences in coast down times measured on 24 different LD vehicles and LD trucks amounted to 7% on average [1]. The range of shortfalls was from almost 0 up to almost 15%. Interestingly, the differences found were rather unevenly distributed as 16 2(8)

3 vehicles fell within the range 0 to 7.0% while the remaining 8 vehicles showed a coast down time that was between 10 and 15% shorter than the manufacturer s declared value. A recent study performed by TÜV Nord for UBA showed the effect on CO 2 measured over the NEDC testcycle for several test parameters [2]. They showed that if the maximum allowed tolerance in road load deviation is applied (-20% at 20 km/h and -10% from 40 to 120 km/h) the CO 2 emission is reduced by 5.3% on average in a range from 2 to 11% for a total number of 5 LD vehicles. TU Graz has done emission measurements on 3 pairs of Euro 4 vehicles and their Euro 5 successors [3]. Since they were interested in CO 2 emissions under real-life conditions, all vehicles were subjected to a coastdown measurement on a flat road. The road load was calculated from the results and the vehicles were tested on the chassis dynamometer with the new settings. Over the NEDC test, the difference in CO 2 emissions between type approval value and the measurement with real-life road load was 17% on average, ranging from 9 to 24%. The difference was explained to be the result of higher driving resistance due to optimization of the tire and road surface combination, tire pressure and beneficial ambient conditions. It was even recognized that in reality the road load may even increase further by a lower tire pressure as recommended by the manufacturer, additional vehicle load, roof racks or other extra equipment that raises driving resistance. At the same time, this report acknowledges that testing a single vehicle can easily over- or underestimate the emissions when comparing these to the type approval values (e.g. due to individual performance differences or undetected malfunctions). The identified road load issues seem very specific and have therefore not been subject to any (publicly available) studies, so no scientific evidence could be found to quantify the contribution to the gap for the individual issues. The Swedish Transport Administration and Transport & Environment have not conducted any tests or measurements, though the possibilities for conducting an in-depth study with a measurement program are explored. This document is primarily based on earlier experience and discussions with engineers within the field of the subject. Identification of RLD related issues The ISO standard specifies a number of conditions and requirements that are intended to guarantee that the measured road load during the test procedure matches the road load of the final production vehicle as close as possible. However, there are a number of ways in which this standard may fall short: Before a coastdown test is performed a vehicle will have to be checked and adjusted. ISO requires these adjustments to fulfill the requirement as specified by the manufacturer (art ). This kind of terminology allows to specify settings with a wider tolerance than is technically needed. One end of the tolerance band could be used during the RLD procedure for a smooth coastdown, while the other end could then be used for a standard production vehicles. An example of such an issue is the wheel alignement. Some issues are not explicitly addressed by ISO , but still may have an influence on the road load, e.g. the preparation of the brakes to prevent parasitic losses. To allow road load tests being performed under different ambient conditions there are correction functions introduced to calculate the results back towards standard conditions. If these correction functions are not fully accurate, it may be beneficial to test at the favorable end of the allowed condition range, e.g. ambient temperature. Not all of the choices in the functional design of the homologation are yet final when the road load tests are taking place, while these choices may have a considerable influence on the road load of the production vehicle. The tire selection is a good example of this. The following sections will further address each of the identified issues. 3(8)

4 Wheel alignment Wheel alignment is an important parameter for the vehicle s ability to coast. Wheel alignment is a tradeoff between good handling and the ability to coast. During mass production in the factory it is possible to adjust wheel alignment rather tight to the specifications. However, if a wide tolerance band is specified by the manufacturer, the wheel alignment could be changed within the tolerances to a setting for better coast down results. At the same time an alignment that would increase rolling resistance also would result in a higher tire wear, so it is not assumed that there is much to be gained. Specify a fixed tolerance band for wheel alignment, either absolute or relative. Adjustment of brakes The space between a brake pad and brake disc is very small. There is a risk of contact between the pad and the disk, even when the brake pedal is not pressed. This will cause some extra friction. To avoid this, the pistons inside the brake caliper can be pushed back by a technician in the preparation shop prior to the tests. This will eliminate friction and improve the vehicle s ability to coast. In ISO , , this is referred to as parasitic drag. For reasons of test repeatability this conditioning procedure is allowed. However, if it is normal for a production vehicle to have some parasitic drag, it is not representative to neglect this resistance. The test procedure could require that as a part of the preparations of the coast down test the driver should press the brakes from a certain speed to stand still (possibly also specifying a minimum deceleration). The resulting parasitic loss (if any) will lead to a measured road load that is more representative for real life situations. However, the problem with this kind of preconditioning is that this may conflict with the repeatability of the coast down test results, since the contact between pads and discs might vary depending upon driving conditions, temperature etc. This may need some more discussion to come to a suitable solution that will do justice to the representativity of the road load test results, while the repeatability of the test can still be guaranteed. At this moment there is no data available that shows the influence of this phenomenon, but physically it is expected to be rather marginal. Before putting a lot of effort into this matter it could be sensible to first prove that the effect is substantial. Ambient conditions South of Spain, southern Italy and Arizona (USA), are some popular test tracks for coast down testing. It is presumably no coincident that these tracks are located in areas with a warm climate. One of the reasons could be that the weather conditions in those areas are more stable, but another good reason may be that the friction of bearings and other moving parts are in favor of the test results. The correction factors only take into account the effect on air density associated with a temperature and air pressure differing from standard conditions. ISO ( ) requires that the temperature during Road Load Determination should be between K (0 35 C). For testing on the chassis dynamometer the current proposal of the WLTP suggests 25 C ± 5 C. To better reflect test cell conditions the Road Load Determination temperature regulation could be adjusted to 25 C ± 5 C as well. The correction factors may be expanded in such a way that they also account for other temperature related influences on the road load. This may also be beneficial from a competition point of view for test tracks in areas with a colder climate. On the other hand, it will be difficult to derive a generally valid correction formula for all vehicles, since the influence may vary for different vehicle makes and models. Also there is probably no data available to calculate such a correction, so this would require setting up a measurement program of sufficient scale to develop some general empirical correction formulae. 4(8)

5 Tires tire wear, pressure, type and rolling resistance There are a number of issues related to the tires and the tire choice that need to be addressed. Tire wear The RLD regulations allow tires to be worn up to 50% between a new and a bald tire. (ISO , ). The resulting effect will be that the rolling resistance is lower. New tires are grinded in a machine and then broken in by mileage collection down to the lowest limit. The tires have wear indicator strips running across or parallel to the tread. The letters TWI are printed on the side of the tire. When approximately mm is left on the tread, these strips will be level with the tire surface, that the tire should be replaced. A new tire has some 8-9mm of thread dept. 50% of this means 4-4.5mm. Since vehicle users normally will replace tires before they are completely bald (typically between 2 and 3 mm), the mentioned 50% is representative for an average tire wear. Therefore it would be better to allow a maximum wear of 3 mm. In the US, the tread depth on the test vehicle must be at least half the usable tread depth (i.e. at least half the difference between new tread depth and minimum tread depth, which in the U.S. is 2/32 of an inch). To better reflect average tire wear and the associated rolling resistance, a maximum wear value of 30-40% from a new to a bald tire should be considered, e.g. 2.5 or 3 mm. Tire Pressure ISO requires the tires to be pressurized according to the manufacturer s recommended inflation pressure. If there is a difference between soak and ambient temperature, the standard provides for a pressure correction function. Since the manufacturer may indicate this recommended pressure, there is a possibility that the tires are overinflated on the test vehicle to reduce rolling resistance. There are no requirements to ensure that this recommendation corresponds to the tire pressure of the production vehicle. According to some very old formulas from the Michelin Handbook and Reimpell Fahrwerktechnik a tire that is recommended to be inflated at 1.9 bar but overpressurised to 2.2 bar will have a 6-8% lower rolling resistance [3]. A requirement should be included to specify a tire pressure that matches (within a certain tolerance) the recommended tire pressure of the production vehicle s tires. Since the recommended tire pressure is also related to the tire make and dimensions, a supplementary requirement to cover that will be needed. See also the next point. Tire Choice Vehicle models are in most cases offered with various tire specifications. The basic model of a vehicle platform might be equipped with tires having a high profile and smaller dimension (diameter) of the rim, while the more sporty version usually is equipped with low profile tires on a wider rim. Since the coast down tests are performed well in advance of the production of a new vehicle type, manufacturers claim that the final choice of tire brand, model and dimensions is not decided yet at the time of the coast down. This allows for selection of a small tire with low rolling resistance and air drag to be used for the coast down tests. There is no legislative mechanism that enforces the production vehicle to be supplied with the same tires. 5(8)

6 A requirement could be included to demand that the same tire make and model as fitted to the production vehicle will be used for the road load tests. However, it needs to be recognized that even the same vehicle model may be factory delivered with different tire types and dimensions. At the same time, tire development as well as tire costs are not stationary, so manufacturers would appreciate some flexibility. As an alternative, the manufacturer could be required to only fit tires on the production vehicle that have the same or lower rolling resistance as the tires used during the coast down tests. If necessary, he can choose to perform different tests for different tire types, as long as the rolling resistance criteria are met. It is recommended that tire dimensions and relevant properties such as rolling resistance are mentioned in the test report. Test track The design of the test track can also influence the results of the RLD. In the US legislation on road load determination it is required to only have commercially available asphalt on the test track, to prevent that special low friction asphalt is used which is not applicable for public roads. In ISO a similar but less stringent requirement is stated in paragraph 5.1.2: The road surface shall be flat, dry and hard, and its texture and composition shall be representative of current urban and highway road surfaces. This is not very restrictive, and there is no objective means to verify if this requirement is met. The different types of asphalt that are commercially applied on European roads show substantial differences in rolling resistance [3]. There are also indications that even a small inclination of the test track will influence the road load positively, though the coast down measurements are performed in opposite directions. Apparently, some second order phenomenon is interfering with the results. The same requirement as applied in the US legislation could be copied into the WLTP. Vehicle Weight During the coast down test, the vehicle is equipped with measurement equipment. To compensate for this added weight some irrelevant parts of the interior of the vehicle may be removed, e.g. the back seat. Also here, no reference is made towards the weight of the final production vehicle, it is only said that The vehicle mass shall be adjusted to meet the requirement of the intended subsequent test, including the mass of the driver and instruments. Manufactures might be tempted to offer a prototype vehicle that weighs less than the production vehicle. Similar to the tire choice issue, a requirement could be included to demand that the mass of the vehicle used for the coast down test is the same or higher than that of the production vehicle. Vehicle Body Concerning the vehicle body, different tactics are used to reduce drag, such as taping the seams between the body panels, and/or demounting roofrails and mirrors. Also the ground clearance could be lower for the road load test vehicle, which will also have a positive effect on the aerodynamic drag. This is not allowed in ISO 10521, but again there is no check with the production vehicle. Similar to the tire choice issue, a requirement could be included to demand that the aerodynamic drag of the vehicle used for the coast down test is the same or higher than that of the production vehicle. This does not necessarily have to be actually measured, but the TAA auditor witnessing the homologation process should 6(8)

7 carefully monitor differences in the bodywork and the ground clearance. It is recommended to include relevant parameters and remarks in the test report such as fitted exterior bodyparts, ground clearance, and maybe more. Transmission While the engine is decoupled from the wheels during coast down, some parts of the transmission (including the differential) are rotating with the wheels. Replacing the normally applied transmission oil by a low-viscous lowfriction alternative could potentially give a slightly lower road load, though the positive effect on road load is expected to be small. A simple statement to use a transmission oil with similar or worse lubrication characteristics as for the production vehicle will be sufficient. For that purpose some of the oil characteristics will need to be specified in the test report. Other issues There may be other issues that have a (limited) influence on the road load, e.g. applying oil to the wheel bearings to reduce friction. Alternative approach So far, a number of issues have been identified that could be solved to some extend in the DTP by adding a number of extra requirements. This could be referred to as a feed-forward approach: the requirements are aimed at the vehicle for coast-down testing, for the purpose of ensuring that the future production vehicle will match the measured road load as close as possible. Please note that there is no legislative mechanism in this feed-forward approach to actually check if the road load of the production vehicle is actually the same. At the introduction of this document it was stated that the objective is to develop a RLD procedure which is representative for production vehicles under real-life conditions while at the same time being repeatable and reproducible. An alternative method to achieve this is to use a feed-back approach: the production vehicle will also be subjected to a road load test, and results will be compared. The only thing necessary for the test procedure is to require a repetition of the emission tests when it is observed that the road load of the production vehicle is higher than that of the prototype vehicle tested earlier (surely a small threshold will be allowed). In this case, road load testing would have to be part of the CoP procedure, or even the in-use conformity checking. Whether this is a feasible option needs to be discussed, but from a scientific point of view this is more preferable than the feedforward approach. The scheme below illustrates both concepts (feed-forward is on top, feed-back at the bottom). 7(8)

8 Recommendations To guarantee the best representative results of road load tests, it is recommended to include road load tests on a production vehicle in the CoP or in-use conformity tests and demand that the road load of the production vehicle is the same or lower than measured on the earlier tested vehicle for homologation (feed-back approach). At the same time this will result in a much simpler road load test procedure, since there is no need to make all kinds of extra requirements for the preparation of the vehicle which is used for coast-down measurement. If current regulations regarding road load determination are changed, this will put an extra burden on the vehicle manufactures since the CO 2 emission targets are already fixed. Therefore the existing legislation and targets for CO 2 need to be reviewed. This review should be part of a verification program. Next step Further discussions with feedback from the automotive industry, contracting parties and the European Commission are necessary at the level of the expert meetings (or in a dedicated subgroup) to come to a solid and well-supported approach. After that, we need to agree on a legislative text which can be proposed to the DTP group as a supplement to the draft GTR. References [1] Road-load Coastdown testing of Selected 1981 thru 1984 Model Year Light-Duty Vehicle and Light-Duty Trucks, EPA, Report nr. EPA-AA-CD-84-01, April 1984 [2] Future Development of the EU Directive for Measuring the CO2 Emissions of Passenger Cars Investigation of the Influence of Different Parameters and the Improvement of Measurement Accuracy, Final report (draft), TÜV Nord / Federal Environment Agency (UBA Germany), Report nr. FKZ , 14 December 2010 [3] Measurement of CO2- and fuel consumption from cars in the NEDC and in real-world driving cycles, TU Graz / BMLFUW, M. Zallinger and S. Hausberger, report nr. I-21/09 Zall-Em 34/09/646, 17 December 2009 [4] Personal information from Dr.-Ing. Klaus-Peter Gläser - Bundesanstalt für Straßenwesen (Federal Highway Research Institute), Germany 8(8)