Peter J. Zemroch

Cluster Analysis as an Experimental Design Generator, With Application to Gasoline Blend ing Experiments

This article concerns the selection of experimental design points from existing series of candidates when the design variables are too interrelated to be manipulated independently. Designs with an even spread of points are shown to estimate the parameters of an assumed linear or polynomial model reasonably efficiently while providing good tests of lack of fit. Furthestneighbor cluster analysis can be used to select the points of such a design under either the Euclidean or the Mahalanobis measure of distance. The technique is used to select the base fuels in actual series of experiments to measure the effect of blending a particular alcohol into gasolines. A new blending model parameterization is proposed, which relates the blending octane number of this alcohol to both its concentration and to the properties of the base fuel. An analagous generalized least squares model is discussed, which gives a simple expression for the expected mean squares in different error strata.

Effects of Gasoline Vapour Pressure and Ethanol Content on Evaporative Emissions from Modern European Cars

A test programme designed to investigate the influence of gasoline vapour pressure and ethanol content on evaporative emissions from modern passenger cars has been carried out by the Joint Research Centre of the European Commission jointly with CONCAWE and EUCAR. Seven gasoline passenger cars representative of current EURO 3/4 emissions technology were tested for evaporative emissions with ten different test fuels. The test fuel matrix comprised 60 and 70 kPa hydrocarbon base fuels with 5 and 10% ethanol splash blends and 5 and 10% ethanol matched volatility blends. The evaporative emission tests were carried out according to a test protocol based on the European homologation test procedure, with no additional vehicle conditioning. Although this test protocol turned out to have a considerable influence on the results, the programme has provided valuable information and several clear conclusions can be drawn. The programme confirmed that vapour pressure (DVPE) is a key fuel variable for evaporative emissions. However the effect of vapour pressure is strongly non-linear; the ethanol blends with final DVPE around 75 kPa gave considerably higher evaporative emissions than the lower volatility fuels in most of the vehicles. Differences between fuels with DVPE in the range 60-70 kPa were small. Additional tests on two vehicles performed after the main programme have raised some questions about possible effects of ethanol on carbon canister working capacity and on the role of permeation in determining evaporative emissions.

Fuel effects on regulated emissions from modern gasoline vehicles

\ The influence of gasoline quality on exhaust emissions has been evaluated using four modern European gasoline cars with advanced features designed to improve fuel economy and CO 2 emissions, including stoichiometric direct injection, lean direct injection and MPI with variable valve actuation. Fuel effects studied included sulphur content, evaluated over a range from 4 to 148 mg/kg, and other gasoline properties, including aromatics content, olefins content, volatility and final boiling point (FBP). All four cars achieved very low emissions levels, with some clear differences between the vehicle technologies. Even at these low emissions levels, all four cars showed very little short-term sensitivity to gasoline sulphur content. The measured effects of the other gasoline properties were small and often conflicting, with differing directional responses for different vehicles and emissions.

Fuel Effects on Regulated Emissions From Advanced Diesel Engines and Vehicles

The introduction of sulphur-free fuels will enable advanced engine and exha st after-treatment technologies to meet increasingly stringent exhaust emissions regulations. As these cleaner fuels and vehicles are introduced, the potential for further improvements in air quality through changes to fuel properties can be expected to diminish. Nevertheless, CONCAWE has continued to update knowledge by evaluating fuel effects on emissions from new engine/vehicle technologies as they approach the market. In this work, carried out as part of CONCAWEs contribution to the EU PARTICULATES consortium [1], two advanced light-duty diesel vehicles and three heavy-duty diesel engines covering Euro-3 to Euro-5 technologies, were tested. The fuels tested covered a range of sulphur content and compared conventional fuels with extreme fuel compositions such as Swedish Class 1 and Fischer Tropsch diesel fuels. The emissions benefits from the advanced engine/vehicle technologies operating on sulphur-free fuels are impressive and likely to bring substantial improvements in European air quality as the vehicle fleet is replaced. Particulate filters have the potential to reduce diesel particulate mass (PM) emissions by more than an order of magnitude. Capability for substantial improvements in control of NOx emissions is also evident. Fuel effects on PM and NOx emissions were also observed. In Euro-3 engines, the effects from extreme fuel changes were in the range of 10-20%. When advanced emission control technologies such as diesel particulate filters (DPFs) were used, PM emissions were so low that the impact of changing fuel properties became negligible. Extreme fuel changes continued to affect NOx emissions even with the advanced engine technologies, although these fuels also reduced maximum power. Optimisation of the exhaust after-treatment was also important, with increasing urea rate reducing NOx emissions. Further progress on NOx emissions can be expected as control of engine-out emissions improves and NOx after-treatment technology matures, with the availability of sulphur-free fuels.

Developing a Precision and Severity Monitoring System for CEC Performance Tests

The Coordinating European Council, CEC, develops performance tests for the motor, oil, petroleum, additive and allied industries. In recent years, CEC has moved away from using round robin programmes (RRPs) for monitoring the precision and severity of test methods in favour of regular referencing within a test monitoring system (TMS). In a TMS, a reference sample of known performance, determined by cross-laboratory testing, is tested at regular intervals at each laboratory. The results are plotted on control charts and determine whether the installation is and continues to be fit to evaluate products. Results from all laboratories are collated and combined to monitor the general health of the test. The TMS approach offers considerable benefits in terms of detecting test problems and improving test quality. However, the effort required in collating data for statistical analysis is much greater, and there are technical difficulties in determining precision from TMS data. This paper describes the test monitoring framework that has been developed and how it fits into the CEC test development process. Its statistical properties are evaluated and recommendations for the control-chart limits are formed. Some comparisons are made with the LTMS (Lubricant Test Monitoring System) used in North America. CEC has carried out a pilot project to assess the feasibility of running a web-based database for a bench test. The results of this were encouraging and a full trial on selected bench tests is being carried out in 2004.

The Computerized Generation of Fractional‐replicate Designs Using Galois Fields and Hadamard Matrices

Many statistical programs can now generate fractional-replicate designs, but most depend on built-in libraries of experimental plans. Relatively few have design generation algorithms that would give the user true flexibility in the numbers of tests, factors, and factor levels. Powerful search algorithms are now available for generating a wide range of blocked and fractional-replicate designs using design keys, but these are mostly regular designs with pq units and factors with p, p2, … levels, where p is prime. Many useful classes of irregular design also exist with numbers of units and factor levels which are not powers of p, and the construction rules for most of these use Galois fields and Hadamard matrices in some way. However, the underlying mathematics can be difficult. The aim of this paper is to collate algorithms to expedite the implementation of these structures in software and to present these in an accessible form to a wider community. The use of Galois fields and Hadamard matrices in generating important design classes, such as the fractional 2k Plackett–Burman and sk Addelman–Kempthorne designs, is then detailed. Combining these algorithms with design key searches gives an arsenal of methods that can generate almost all existent balanced fractional-replicate designs of sizes likely to be used in real-world experimentation. These methods have all been implemented in the KEYFINDER program, which is available from the author, free of charge. Copyright

Statistical design and analysis of a test programme to assess the volatility characteristics of ethanol/gasoline blends

Designing a measurement programme to assess the volatility characteristics of ethanol/gasoline blends posed challenges as the seven base fuel properties of interest were highly constrained and difficult to manipulate independently of one another. The target base fuel matrix was generated by augmenting a 49-fuel fraction of a 75 factorial with 11 additional fuels chosen using D-optimality. Two of the five factors were treated as pseudo-factors and used to generate the levels of two pairs of mutually constrained properties. The 60 base fuels were then blended and subsequently splash blended with 5%, 10%, 15%, 20% & 25% ethanol. The test order for the resulting 360 fuels was structured and randomized to reduce the risk of extraneous sources of variation contaminating the regression models subsequently fitted to the data. Cross-concentration models had to be fitted by generalized least squares techniques as the measured values of the principal dependent variable were structurally correlated.

CONCAWE/GFC Study on Gasoline Volatility and Ethanol Effects on Hot and Cold Weather Driveability of Modern European Vehicles

A joint test programme has been carried out by CONCAWE and GFC to evaluate the impact of gasoline volatility and ethanol on the driveability performance of modern European vehicles. Eight vehicles, three with DISI fuel systems and five with MPI, were tested for hot driveability performance. After screening tests, a subset of four vehicles was selected and tested for cold driveability. The latest test procedures developed by GFC were used for both hot (20, 30 and 40°C) and cold (+5 and -10°C: representative of moderate winter conditions) weather testing on climate controlled chassis dynamometers. A matrix of four hydrocarbon test fuels at two levels of DVPE and E70 was blended for the hot weather testing, and three fuels with varying E100 but essentially parallel distillation curves for the cold weather tests. For each hydrocarbon fuel, two other fuels containing 10% ethanol were made, one splash blend and one with matched volatility. Some tests were also carried out using 5% ethanol fuels made by blending hydrocarbon and 10% ethanol fuels. The paper describes the detailed results obtained for both hot and cold weather driveability, in terms of vehicle, fuel (volatility and ethanol content) and temperature effects.