The Principles of Carburetting as Determined by Exhaust Gas Analysis.
Page 19
Page 20
Page 21
If you've noticed an error in this article please click here to report it so we can fix it.
Excerpts from Mr. Dugald Clerk's Paper before the Institute of Automobile Engineers.
It is a most difficult task to design a carburetter which will automatically meet the varying demands for charge by the engine, both as to speed and volume fed per stroke. Many methods are in use, purporting so to vary the suction pressure, if I may so call it, within the pipe, as to cause the jet to respond in the manner desired. An these automatic devices act by varying the pressure due to suction, either by operating a throttle valve or by opening an auxiliary valve. By these devices petrol engine designers have succeeded in running engines with practically consecutive ignitions under very wide conditions of load and speed of rotation. A usual method of determining the degree of success of the regulating device is to watch the flame coming from the exhaust valve under the various changes of condition intended to be made ; the flame leaving the exhaust valve should not be a luminous flame, but a blue, non-luminous one. This blue, non-luminous flame is supposed to indicate fairly complete combustion within the cylinder.
On 19th and 20th March of this year, a valuable investigation was made by the Royal Automobile Club on the nature of the gases and vapours emitted by twelve motorcars under some of the conditions of ordinary running. It was shown that carbonic oxide gas was present in the exhaust from all the cars tested. In each case tested, the gases were discharged into a copper drum by means of a flexible metallic tube coupled to the exhaust pipe at a point between the engine and the silencer. and the portions of exhaust gases supplied were discharged during ten minutes, so as to ensure that the contents of the drum to be examined were average samples of the gases passing in that time from the engine. Two samples of exhaust gases were taken from each car. T had the honour to assist at these experiments, and I have studied all the analyses made from these gases. Along with others, it somewhat surprised me to find so many cars delivering more than 2 per cent, of the exhaust as carbonic oxide, because my experience of gas engine exhaust analysis had led me to expect a more perfect combustion.
As the Royal Automobile Club experiments were not sufficiently varied to enable the problem to be thoroughly understood, I proceeded to make some experiments on my own car, an 1Sh.p. SiddeIey landaulette, having a four•cylinder engine, cylinders 4 inches, stroke 4 inches. These experiments were made on three days-23rd April, 7th May, and 3rd July of this year. Eleven samples of exhaust gas were taken altogether, under various running conditions, and they were very carefully analysed for me by Mr. Horatio Ballantyne, F.I.C., F.C.S. The methods of analysis used by Mr. Ballantyne were fully discussed between us, and they are the very best available to-day.
Exhaust samples were taken while—
(a) The car was standing on the level with the engine running as slowly as possible.
(b) The car still standing, but engine running at about 600 revolutions per minute.
(c) The car running on a level at about 18 miles per hour, the throttle less than half open.
(d) The car climbing a hill, engine running about 1,000 revolutions per minute, and throttle from three-quar. ters to full open.
Table I. shows the composition of the exhaust gases on these three days :—
For the purpose of the tests, copper drums and metallic tube connections were used, exactly similar to those adopted in the Royal Automobile Club trials. The trials on 23rd April were made with the carburetter auxiliary air valve adjusted as ordinarily set by my motor man without any alteration. For the 7th: May trials, the auxiliary air valve spring was released to some extent ; and for those of 3rd July it was still further released.
Taking the analyses under (c), an examination proves that 80 per cent, of the total carbon present in the original petrol is dis charged as carbonic oxide, 5.6 per cent, is discharged as methane, so that nearly 35 per cent, of the original carbon in the petrol has been incompletely burned, or not burned at all. In this same case 27 per cent. of the total hydrogen in the fuel remains unburned. Calculating from these numbers the losses, I find that nearly 30 per cent, of the total heat possible on com• bustion is not evolved at all.
Take now the best test under (d), 3rd July. Here the engine was well loaded. Even in this case, 17.6 per cent, of the total carbon is present in carbonic oxide and methane, and 6.7 per cent, of the total hydrogen in methane and free hydrogen. This gives a loss of 10 per cent, of the total heat, due to incomplete combustion. In the best example, therefore, upon this sheet, only 90 per cent, of the total heat of the fuel can be utilised in the engine.
Undouble.dly, practically perfect results can be obtained, when petrol is properly proportioned to air. This is clearly seen from some interesting experiments described by Professor H. Hopkinson, of Cambridge, and Mr. L. G. Morse, in a paper read before the British Association this year. Table II. is reproduced from Professor Hopkinson's paper.
In these tests a four-cylinder Daimler engine was used, cylinders 3.56 in. diameter, stroke 5.11 in. The engine was run
under f till load at a constant speed between 700 and 750 revolutions per minute, and in all the tests the air taken into the engine per stroke was very nearly constant. I reproduce the curves shown in Fig. 1 from Professor Hopkinson's Paper. They show very clearly what goes on. The upper curve (A) shows the variation of power with petrol consumption. The power, it will be observed, rises rapidly when petrol consumption is changed from 0.16 to 0.19 lb. per thousand revolutions, but from that point largely increased petrol may be taken iii while the power remains practically constant. The petrol can be increased to 0.25 lb. per thousand revolutions without increasing power at all. After that increase of petrol produces diminution of power. The curve (B), however, shows variation of thermal efficiency, reckoned on indicated h.p. with petrol consumption. The maximum thermal efficiency is attained at the same point, that is' just about 0.2 petrol consumption per thousand revolutions. Any further addition of petrol rapidly reduces the thermal efficiency. The key to the position, then, so far as best results, purity of exhaust, and maximum thermal efficiency are concerned, is found in adjusting the carburetter so that the engine gives its most economical petrol consumption for a given power.
As the problem is to charge any given volume of air passing into the engine with a practically unvarying proportion of petrol vapour in an uniform manner, it seems to me that all systems of speed control must fail to obtain proportionality throughout the whole range. if the entering air could be made to drive a fan, similar say to an anemometer, with proper precautions, this fan or anemometer could be made to run at a speed of rotation very closely proportional to the volume of air supply throughout the whole range of conditions. If a spindle be driven by this anemometer, and a small chain passed from that spindle into a petrol vessel at constant level, then it would be possible to work out a carburetting contrivance which would supply the air passed through it on its way to the engine with petrol exactly in proportion to the charge volume taken in. I believe contrivances of this kind have been proposed, and it seems to me that although we have practical difficulties, if correctly carried out, they should be able to give a perfectly uniform mixture to an engine under all conditions of speed variation -and charge volume variation.
In testing any carburetter, the exhaust analysis will be found to furnish a most important aid—aid to he obtained in no other -way. Referring again to Table I., several lessons can be deduced as to the conditions of perfect combustion. The most important point apparent from these analyses is this: that complete combustion by no means follows because of excess of oxygen. That is, we may adjust a carburetter in such a way as to give a substantial excess of oxygen throughout its whole range, and yet we may not succeed in suppressing carbonic oxide or completing combustion.
It appears to me that in working out exhaust gas analyses for comparison of the merits of different cars or different adjustments, the total fuel consumed should be tested in any given experiment, and from the exhaust gas analysis there should be calculated the proportion of that fuel which is rejected without burning. This alone gives the true standard. In future comparative tests as to purity of exhaust gases, I would suggest that this method of absolute performance be adopted. Apart from the question of the carburetter, the timing of the ignition has an important effect. If combustion is to be reasonably complete in the cylinder, ignition must take place as early as possible, consistent with a proper form of diagram. So far, then, as my experiments have carried me at present, the following conditions appear to produce imperfect combustion :— 1. Too rich mixture with in-sufficiency of oxygen.
2. Too weak mixture with excess of oxygen, but too slow a rate of ignition and combustion.
3. Irregular mixture—mixture supplied too rich in composition at one part of the stroke, and too weak in another ; that is, bad mixture.
4. Engine and carburetter cold. This tends to cause imperfect combustion, due partly to low temperature, and partly to bad carburetting.
5. improper timing of ignition, and missed ignitions.
6. Igniting in the body of the cylinder, instead of in a port.
This will produce imperfect combustion at light loads.
To get the best and most rapid combustion, at light loads, it is in my view desirable that the charge should enter the cylinder by way of a port of some little length, and that the ignition of the charge should be accomplished in the port itself, and not in the cylinder. By so combining the carburetter and ignition contrivances I believe that a quite innocuous exhaust could be obtained, under light and heavy loads at all rates of speed, with a great saving in fuel consumption and wear and tear of the engine. It is well known that engines igniting in the port are more flexible. The reason of this is obvious. As the charge is reduced in volume, the proportions of exhaust products in the engine are not so reduced. They remain not quite, but nearly constant, in volume and weight. Accordingly, as the load becomes less and less, the effective charge added becomes more and inure dilute, and the mean mixture soon passes the firing point. By igniting in the port, a pocket of rich mixture is retained to the last, and this rich mixture, if fired sufficiently early, produces a practically complete combustion in the weaker mixture filling the cylinder.
There are many other interesting points to be discussed, but I reserve them until I have made further experiments, which I have now in hand.
The Discussion.
At the conclusion of the reading of the paper, Colonel Crompton drew attention to the importance of minimising the quantity, or entirely eliminating the presence, of carbon monoxide from the exhaust gases of internal-combustion engines on motor vehicles.
Dr. Hele-Shaw, who opened the discussion, also laid emphasis on the necessity of obtaining more perfect combustion in petrol engines, and said that it was the duty of the Institutien of Automobile Engineers to give due warning to all drivers of motor vehicles, and to all engine attendants who had anything whatever to do with internal-combustion engines, that, given too rich, or too poor, a mixture, the exhaust gases were, undoubtedly, highly poisonous. Dr. Hele-Shaw plotted on the blackboard some convincing comparisons of the figures given above the chart in Fig. 1. The curves, as given by Dr. HeleShaw, conveyed to the mind a much clearer impression of the state of the exhaust than did the tabulated particulars in the original -chart.
Mr. Max Laurence, who represented -the makers of the car on which the author of the paper had made his experiments, was of opinion that much better results might have been obtained with a more modern carburetter. The one used for the tests was three years old, and, on behalf of his company, he was prepared to offer the use of an engine and carburetter for further tests by Mr. Dugald Clerk, or his company would themselves carry out the experiments under Mr. Clerk's supervision, at Crayford.
Mr. Mervyn O'Gorman did not quite agree with the chairman and other speakers in proclaiming the poisonous qualities of the exhaust gases. With his usual wit he remarked that what is to one man amusing, is, in the opinion of another man, immoral ; whilst automobile engineers saw the necessity of purifying the exhaust, the alarmist's remarks, though amusing to the automobile engineers, might be taken in a much different light by the daily Press. What the motorcar owner requires is an engine which will start on the switch. To do that the carburetter must be adjusted to give a rich mixture at low speeds. He thought that much of the variation in the richness of the gases could be obviated, without complicating the carburetter, by giving to the float as little clearance as was absolutely necessary for its free movement in the float chamber. In this way the disturbance of the petrol level could be very much miniroised. He was inclined to think that the. figures of the analysis were not all due to excess of fuel. The viscosity of a fluid greatly affects the height to which the jet can be thrown with a given pressure ; the vicosity of petrol :vary considerably, and is greatly affected by atmospheric conditions. In some experiments of his own, in which he used a petrol engine as an air compressor, a sample of the compressed air was analysed and found to contain 3 per cent. of CO2, which could only have come from the lubricating oil.
Mr. Lanchester remarked that he had tested many spray carburetters, and had discarded many ; in fact, he had discarded as many as he had tested. His opinion was that the spray carburetter very often failed to atomise the fuel completely, the result being that occasional puffs of "fog" found their way into the induction pipe. If such a pipe had any " right-angled" branches, it was certain that all the cylinders would not receive a uniform mixture. The idea of Mr. Dugald Clerk's "windmill" amused him.
Mr. Tom Thornycroft also thought that " fog " was present at times, but, in order to obtain perfect diffusion of the vapour, he had introduced into the channel above the jet, layers of gauze through which the vapour was obliged to pass, and in this way more perfect diffusion of the mixture was obtained. He could not agree with Mr. Dugald Clerk on the question of a blue flame. He thought that, if a flame appeared at all, the engine was not doing its best, and he had noticed that, when such a blue flame was present, violent "popping" in the carburetter resulted. He asked Mr. Dugald Clerk if he could give any explanation of the knocking in the cylinder, and the emission of black smoke from the exhaust, when using paraffin fuel, and when pre-ignition is taking place.
Mr. D. J. Smith asked what allowance was made in the analysis for burnt lubricating oil, and whether it was possible that the varying results of the analysis might be attributable to this.
Mr. Cozens-Hardy quoted a case where he had a contract with the Napier Company to supply one of its vehicles, in which it was specified that an analysis of the exhaust gases should not reveal more than 2 per cent, of carbon monoxide. During the first trial of the car, it did not pull at all well, and the analysis showeel 4.0 per cent. of carbon monoxide, and also showed that 40 per cent, of the gases were imperfectly consumed. The vehicle was sent back for readjustment, and, without any structural alteration to the carburetter, a second trial showed 1.2 per cent, of carbon monoxide, and only 9.7 per cent, of the gases were imperfectly consumed. In this test, 3.7 per cent. of hydrogen was unconsumed, and the petrol consumption was at the rate of one gallon for each 16.3 miles, the weight of the vehicle being 49ewl. Mr. Martineau thought more perfect diffusion of the gases could be obtained by making the induction pipe of flat section with inside projections.
Mr. Lyons Sampson agreed with Mr. Tom Thornycroft about the blue flame, and was of opinion that the "popping" was caused through the mixture's being too weak, which causes the combustion to take place so slowly that, by the time the inlet valve opens for the next stroke, it is not completed, the result being a flash back down the inlet pipe, An important communication had been handed in from Mr. Brewer, but, as the hour was now getting late, the contents of this were not disclosed. Mr. Dugald Clerk rose to reply to the discussion at five minutes to ten, and stated that, although he would deal only with a few of the questions then, he wished it to be understood that he would write a fuller reply, which would be delivered at the next meeting of the institution, tie agreed with Colonel Crompton, Dr. Hele-Shaw, and Mr. Laurence, but disagreed with Mr, Mervyn O'Gorman as to the inadvisability of publicly announcing the presence of poisonous gases in the exhaust from an internal-combustion engine. Full warning was sent by the makers to the users of all suction-gas plants, and the poisonous nature of suction gas was clearly pointed out to them. He thought that the Institution of Automobile Engineers would be neglecting its duty if it did not give such similar warning to the users of petrol engines. It was quite true that many engines will pull better, and accelerate better, with a mixture, the exhaust from which, on analysis, revealed the presence of 10 per cent, of carbonic oxide, than would an engine whose thermal efficiency was high. As all knew, a broad path generally leads to destruction, and he wished to keep his audience to the narrow path as shown by the peak in the diagram (fig. I). In this way we should avoid future legislation on the sitlaject. When the Automobile Club commenced its experiments, it had km idea that so much as 10 per cent, of carbonic oxide was present in the exhaust gases in any car, and its reason for not making a public announcement at the time was that it desired to give manufacturers an opportunity of correcting the thermal inefficiency of their engines. He reminded designers that greater flexibility could be obtained by a very slight increase in the bore of the cylinders, the engine could then be run at its highest thermal efficiency, and without any emission of poisonous vapour. It was entirely a matter of adjustment of the carburetter, and his own experiments, and those of Mr. Cozens-Hardy, would indicate that much might be done by this means. He agreed with Mr. Thornycroft that the blue flame should not appear, and, in reply to that gentleman's question regarding black exhaust, he stated that this arose chiefly from a bad mixture. To Mr. D. J. Smith's query as to the effect of lubricating oil on the analysis, he would say that throughout the whole of the tests no smoke was visible from the exhaust, and he thought it might be taken that the conditions were constant throughout, and that the minimum amount of lubricating oil was being used. The proceedings were at this point adjourned to the 22nd January next.