DISCUSSION.

Mr. H. W. Spangler.-It is perhaps interesting to go back to the original method of designing compound engines. They were to a very great extent designed in this way: The ratio of expansion was first determined. The square root of the ratio of expansion was taken as the ratio of the volume of the two cylinders. When it comes to triple expansion engines, perhaps the same theory would fit closely, and it would necessitate in this case that the ratio of the volumes of the cylinder should be as, or power of the volume of the low-pressure cylinder. (See Gow, Franklin Institute Journal, Sept., 1888.) If that is the case in any particular instance, the square of the volume of the intermediate cylinder should be equal to the product of the volume of the other two; and on page 583 the author refers to what is approximately the actual card. He gives for the area of the large piston 89 and the small one 40; 40 times 89 is almost the square of 60-the intermediate cylinder. So it seems to me that that is a pretty convenient way of getting at it. As to the ratio of the volumes, in these tables he has worked out from a number of examples the mean figure, and it is perhaps interesting to see how this can be given a little more easily. He says on the first page of the examples, page 494, that for 160 pounds pressure the average cylinder ratios are as 1 to 2.66 to 7.24. Now, 2.66 times 7.24 is almost exactly 19, and 19 is the ratio between the terminal pressure and the forward pressure likely to obtain in these engines. So, if you apply reasoning of that sort, it will give you as close work as he can get by the method he gives. It works very closely even in his table of results. He gives for 130 pounds 1, 2.25 and 5; 2.25 is almost exactly the square root of 5, and the same right through approximately.

With respect to the first method, which the author calls the annular ring method, that is, I think, only an amplification of the method that was proposed to Mr. Isherwood a good many years ago to explain the action of steam in a compound engine, and it seems to me to be fallacious. He says, on page 581: "Find the mean ordinate of each area; that of C will be the mean unbalanced pressure on the small piston." If that is so, B, to my mind, should be the mean unbalanced pressure of the next piston, because the work that is done in any piston is the volume traversed multiplied by the unbalanced pressure per square foot.

The

unbalanced pressure is acting on that piston. But while there is no question that the volume traversed in this particular case is the distance DE for the intermediate piston, the mean unbalanced pressure from the card is the mean unbalanced pressure of the area remaining after subtracting the small pressure, according to Mr. Whitham. I cannot see the reasoning, but it seems to me that it is not sound.

As to his method of treating curves on Figure 149, it is well known that, even if the clearance in the low-pressure cylinder is considerably less in proportion than that in the high, the lefthand side of the card in this figure that he has (149) should be at a considerable distance from the clearance; that is, he should have cut off another corner of these cards to get anything like the actual cards. The left-hand corner of the card A should be cut off to perhaps one-half the distance to FG. It makes considerable difference in the card. Of course, this is only intended as an approximation from the beginning; but it seems to me a closer approximation could be gotten if he had made a reasonable allowance for the amount of clearance.

Mr. H. H. Suplee.-Although this paper refers to triple expansion, yet I suppose the subject is really that of multiple expansion. I have some data about quadruple expansion engines which may be of interest in this connection. In the Journal of the American Society of Naval Engineers for February are given quite a number of details and data about fourteen different quadruple expansion engines, and I have reduced the cylinder areas to the common standard in order to get at their ratios. The pressures vary somewhat, which of course makes some discrepancy; but nearly all are intended to be operated at a pressure of about 180 pounds to the square inch, and, with one or two exceptions, which may be due to special conditions, the cylinder areas seemed to run almost in the proportion of 1, 2, 4, 8. Averaging the whole number, I have 1 to 2 to 3.78 to 7.70. Although I think that—with the exception of one or two that are quite different and apparently calculated for different conditions-eliminating those-it would come out almost exactly in the proportion of geometrical ratio, while the actual expansion ratio is very nearly in the same proportion as that to which Mr. Spangler has referred. This may be of interest in this connection.

Mr. F. H. Ball.-I do not see that much attention is given here to the matter of dividing the range of temperature in the different

cylinders. This formula, as I understand it, is for dividing the load, so that each cylinder shall do its share of the work. Now, by changing the point of cut-off, the amount of work of each cylinder can be varied quite considerably; and inasmuch as compound and triple expansion engines are for the purpose of preventing cylinder condensation, it seems to me important to pay some attention to the range of temperature in each cylinder. This may be covered in the table where there are different proportions given for different pressures. But I do not see anything about it in the paper; and I think there is a latitude, in proportioning these cylinders, which will permit of an equal division of the work, and at the same time, if better proportioned, for equal division of range of temperature from the highest to the lowest.

Prof. Denton.-I think that point is covered by the fact that Mr. Whitham bases his rules upon a table which is quite extensive, and taken from practice which fulfils the condition mentioned by Mr. Ball.

Mr. W. S. Doran.-It may be of interest to the members to know the size of the triple expansion engines that have just done some very wonderful work on the City of Paris. These engines have cylinders 45 by 71 by 113, 5-foot stroke; twin screws; the engines developing 20,200 horse-power. They are supposed to be very successful.

The President.—It would be interesting if we could have some additional information as to the vessels referred to the fastest of the Atlantic service at the present time.

Prof. Denton.-I would like to ask Mr. Doran what are the dimensions of the screws of that ship.

Mr. Doran.—I do not know exactly-about 20 feet diameter I think they are; 28 feet pitch; there are three blades; phosphorbronze screws.

Prof. Denton.-Plain screws?

Mr. Doran.-Three blades. The boiler pressure carried on this ship is 150 pounds.

Prof. De Volson Wood.-I have noticed in some actual triple expansion engines that the low-pressure cylinder seemed to be much larger than those usually made. There are a good many cases, I think, where there are somewhat larger proportions than those which have been intimated, and a few where they are very much larger. If such be the case, I would like to inquire of those who have knowledge in regard to triple expansion engines,

why this is the case. Why are they so made? Is it a notion of the engineer which has not been confirmed, or is there some good reason for it?

Mr. Doran.-I would like to add that the speed at which these engines run is something very unusual. They turn up about 92 revolutions a minute and a five-foot stroke.

The President.-That would be 900 feet piston.

Mr. Doran.-About 920 feet. They worked cool and quiet under these conditions.

Mr. Whitham.*-The annular ring method of proportioning the cylinders has been applied by me to many existing compound and triple expansion engines with most satisfactory results. This I did in order to check the accuracy of this method. It is only one of the many ways used, or that might be used, in proportioning the cylinders. The theory has been most ably explained by Chief Engineer Isherwood, of the Navy, and has never been proven fallacious. I have not time to here enter into a defence of this method. This way of proportioning cylinders was given to the Society because it is simple and accurate when judged from existing practice. The second method is given for the same reason. The error due to an inaccuracy in not treating the clearance in the most theoretical manner is no greater than in assuming that the steam expands by the hyperbolic law, and tends to balance it. The designer cannot lose time in exact solutions involving the use of the most complicated formulæ or methods. It has been. said that the best way to design a cylinder is to guess at the dimensions, and double them. While this is not to be recommended, no designer allows his deduced results to entirely control him when he has successful practice as a guide. The two methods are given, not to replace other methods, but as new ways in which the cylinders may be proportioned.

Prof. Denton has so fully answered Mr. Ball that I need add nothing further.

* Author's closure under the Rules.

CCCXXXIX.

IMPROVED MOTION DEVICE FOR ENGINE

INDICATORS.

BY A. W. JACOBI, NEWARK, N. J.

(Member of the Society.)

WITH the introduction in late years of large engines, and especially of the compound type, and with the increased interest Inanifested in indicator tests of engines of this class, engineers feel a need of a simple motion device-something that is easily applied, and, when in position, occupies the least possible space. It should be adjustable to a wide range of sizes and conditions, and to engines of various makers. It should be as light as is consistent with strength, and so designed as to be easily packed in boxes, thus making it convenient to be carried around. The stretch of string should be reduced to a minimum, and it should be arranged to operate any number of indicators simultaneously, at either short or long distances from the reducing motion, and take, practically, all strain off the drum spring of indicator. It should enable the operator to stop or start any one or more of the indicators independent of the rest withont leaving his position, and without stopping the reducing motion. Such a device, the writer thinks, will be appreciated by engineers and builders of steam and pumping engines, air compressors, etc.: and in presenting the following description of one recently invented by E. K. Conover, of Newark, N. J., it is believed that all the above points, as well as many additional ones, have been studied and met in this arrangement. In the accompanying cut, Fig. 131 represents a side elevation, and Fig. 132 a plan of the apparatus as applied to a tandem compound engine. Applications to other kinds of engines will suggest themselves to the engineer, so this description will cover especially this type of engine. The motion device and indicators are exaggerated in proportion to the engine in order to show them more clearly. A glance at the engraving will show that the prime feature of the device is an endless cord operated in both directions entirely by the engine, thus relieving the indicator drum spring of all strains except those due to its own propulsion.