Associate Professor S. V. GOLASHEVSKII, Assistant A. M. BUZYNNIK, Assistant D. M. LEVIN
and Assistant Yu. N. SOKOLOV.

On the question of the use of artificial liquid fuels in internal-combustion engines.

I. USE OF FUEL OIL FROM THE SEMI-COKING TAR OF SAPROPELITES FROM THE BARZAS III DEPOSIT IN DIESEL ENGINES*)

Part III.

COMPARATIVE TESTS OF SAPROPELITE FUEL OIL IN A COMPRESSORLESS PRECHAMBER ENGINE.

I. The experimental engine and its installation.

The comparative tests described below were carried out on a two-stroke compressorless Diesel engine with a Deutz PME precombustion chamber.

A vertical section of this engine is given in Fig. 26. Its principal characteristics are as follows:

Cylinder diameter \(D = 200\) mm.
Piston stroke—\(S = 300\) mm.
Rated power—\(Ne = 25\) h.p.
Rated speed—\(n = 435\) rpm.
Actual compression ratio (counting from the moment the exhaust ports close), established by filling the combustion chamber and by precise measurement of the working stroke of the piston, \(\varepsilon = 11.4\); nominal (counting over the entire piston stroke) \(\varepsilon' = 15.27\).

The volume of the precombustion chamber is 21.5% of the total volume of the compression chamber.

The engine fuel pump is driven by a conical cam mounted on the governor shaft; regulation is achieved by moving it along the axis.

The engine has two nozzles: the working one—in the cover of the pre—

Fig. 26. Vertical section of the Deutz PME engine.

) For the beginning see Proceedings of the Siberian Mechanical-Engineering Institute*, vol. I (53), issue 1, Tomsk, 1933.

measuring and starting portions, introduced directly into the main combustion chamber.

A section through the engine head, the prechamber, and the injector is shown in Fig. 27.

Fig. 27. Head, prechamber, and injector for the Deutz PME.

The engine crankcase, equipped with steel plate valves, serves as a scavenging pump.

The engine is started with the aid of air compressed to 35 atm, stored in a cylinder of 120 l capacity.

The general view of the installation of the engine described, with all the instruments used in the tests, is shown in Fig. 28.

Here:

a — engine fuel pump;

b — pressure gauge on the fuel line pump—injector;

c — measuring tank for fuel;

d — main indicator of the working cylinder;

e — indicator of the crankcase;

f — indicator for taking diagrams of the motion of the pump plunger;

g — speed reducer;

h — thermocouple for measuring the temperature of the exhaust gases;

i — device for changing the number of revolutions;

k — totalizing revolution counter;

l — arm of the band brake;

m — service fuel tank;

n — tachometer;

o — valve “A” for sampling exhaust gases when obtaining “deposit diagrams” (see below, “Characteristics of the exhaust gases”);

p — gate valve “B” with a device for the same purpose.

All instruments used in the tests are described below—in the chapter “Methods of Measurement.”

II. Starting the Engine on Sapropelite Fuel Oil: Possibilities and Conditions of Starting

The compressorless Deutz PME engine was first switched over to operation on sapropelite fuel oil on 5/I–32. The engine, previously cleaned and washed, was started on ordinary petroleum motor fuel from the upper supply tank (“m” in Fig. 28), while the measuring tank “c” was filled with sapropelite fuel oil heated to slightly above room temperature.

Fig. 28. Installation of the experimental Deutz PME engine in the internal-combustion-engine laboratory.

After 20 minutes of operation under 50% load, by switching the three-way valve on the fuel line, the engine was transferred, without preliminary unloading, to sapropelite fuel oil, with no visible changes whatsoever in the conditions of its op-

no operation was observed. The indicator diagrams that were taken likewise revealed no abnormalities. The engine operated quite smoothly, without misfires, and accepted any load up to the normal one.

After one hour of operation on sapropelite fuel oil at full load, the engine was switched over to petroleum, and then again to sapropelite fuel oil cooled to 19–20° C.

In this case as well, no abnormalities in operation were found. After 2 hours the engine was again switched over to petroleum and stopped. Inspection of the engine after stopping revealed no abnormalities.

This first experiment demonstrated the possibility of operating the engine on sapropelite fuel oil; however, it did not yet clarify the actual conditions of such operation, nor did it establish the possibility of starting the engine directly on the new fuel and the conditions required for this.

Subsequently, when conducting the main tests of the sapropelite series, repeated attempts to start the engine directly on sapropelite fuel oil did not yield positive results. Each time the engine first had to be started on ordinary petroleum fuel and only then switched over to fuel oil. This switchover was usually entirely successful, even in the case of a very short preliminary operation on petroleum without load.

During the main series of experiments on sapropelite fuel oil, at a load of 75% of normal, the engine was accidentally stopped and was then immediately restarted directly on sapropelite fuel oil.

The start was successful, but was accompanied by several very intense flashes. Thus, a well-warmed engine could also be started directly on sapropelite fuel oil.

In order to determine the possibility and the necessary conditions for starting the engine directly on sapropelite fuel oil, after completion of the main tests, special experiments were carried out.

Under conditions as nearly identical and normal as possible (starting-air pressure, duration of the starting period on compressed air, room temperature, etc.), a series of trial starts of a cold engine was made through the main injector (the so-called “starting injector” was not used, since even on ordinary fuel starting with its aid without inserting smoldering paper was unsuccessful¹), with a gradual increase in the preliminary heating of the sapropelite fuel oil used for starting. The heated fuel oil was poured into the measuring tank (“c” in Fig. 28), in which its temperature—$t_b$—was also measured. During the pumping of sapropelite fuel oil through the inj—

the temperature of it during discharge—$t_{\phi}$—was also measured, corresponding approximately to the actual temperature of the fuel supplied to the injector. At the moment of starting, indicator diagrams were taken continuously each time.

A series of trial starts, in which the sapropelite fuel oil was preheated to $35^\circ\mathrm{C}$, with the corresponding temperature after filling into the tank $t_b = 28^\circ\mathrm{C}$, and on discharge through the injector $t_{\phi} = 20^\circ\mathrm{C}$, did not give positive results: not a single flash was obtained. It was noted, however, that despite the insignificant length of the fuel line from the tank to the pump, the temperature of the sapropelite fuel oil on discharge through the injector ($t_{\phi}$) was considerably lower than its temperature in the tank ($t_b$); moreover, preliminary heating of the sapropelite fuel oil above $30^\circ\mathrm{C}$ already almost did not increase the value of $t_{\phi}$ observed during pumping. This circumstance is evidently explained by the considerable heat transfer from the sapropelite fuel oil to the body of the fuel pump, partly to the pipelines, and chiefly to the massive cylinder head of the engine as the fuel passed through the bores in the latter.

In order to raise $t_{\phi}$ above the values of about $20^\circ\mathrm{C}$ that are limiting for a cold engine, it was necessary to apply artificial heating of the engine head with a blowtorch. Under these conditions a series of trial starts was again carried out, which again were unsuccessful. Nevertheless, in some experiments it was possible to obtain 1–2 flashes, under the following conditions:

starting pressure—$p_n = 33$ atm; temperature of the sapropelite fuel oil in the tank—$t_b = 21^\circ\mathrm{C}$;     “        ”        “     after inj.—$t_{\phi} = 29^\circ\mathrm{C}$

at least 5–6 flashes were observed, recorded also by the indicator, some of which were very intense and were accompanied by a “shot” through the safety valve. The engine, however, did not start even in this case.

After each of these unsuccessful starts, the engine head was removed, and the prechamber was also taken out and inspected. The latter every time proved to be wetted with sapropelite fuel oil, but cold. On the piston crown, also, large drops of fuel were noticed almost every time.

Finally, in addition to heating the removed head, intensive heating of the prechamber was also applied.

The heated prechamber and head were quickly put in place, and the engine was immediately started. Only in this case, under the initial conditions (for the designations, see above)

$p_n = 33$ atm; $p_b = 25^\circ\mathrm{C}$ and $t_{\phi} = 40^\circ\mathrm{C}$

did the start give positive results—after several in-

intensive flashes and “shots” through the engine safety valve, the engine began to run calmly. At the moment of switching from compressed air to sapropelite fuel oil, the engine developed about 200 rpm; the atomization pressure (by the pressure gauge at the fuel pump) was approximately 70 atm.

Fig. 29. Starting diagrams of the Deutz PME engine on petroleum fuel and on sapropelite fuel oil.

The mixed indicator diagram obtained during this start is given in Fig. 29—b, on which, for comparison, the ordinary starting diagram on petroleum motor fuel is also placed.

Comparison of these diagrams indicates a noticeable increase in the delay of combustion during the first flashes of sapropelite fuel oil. In addition, along with flashes equal in intensity to those during the petroleum start, during starting on sapropelite fuel oil there are also flashes with reduced maximum pressure (down to values usual for normal engine operation) and with clearly expressed prolonged combustion at almost constant pressure.

All this indicates that the first flashes when starting a cold engine on sapropelite fuel oil can be obtained only with great difficulty, whereas starting on petroleum motor fuel usually proceeds quite freely even at reduced starting-air pressure and with fuel cooled to \(+8\)—\(9^\circ\mathrm{C}\).

Since, under operating conditions, preliminary heating of the cover and prechamber, if they have to be removed from the engine, is extremely inconvenient, experiments were undertaken to determine the possibility of starting the engine without disassembly for heating any parts.

Here starting could be accomplished only after prolonged (about 20 min.) and very intensive external heating of the engine head, the fuel pump, and the injector with a blowtorch. The initial conditions in this case were: \(p_n = 36\) atm, \(t_b = 28^\circ\mathrm{C}\), and \(t_\phi = 58^\circ\mathrm{C}\). The engine began to run immediately after slight “shots” through the safety valve.

Thus, starting a completely cold engine on sapropelite fuel oil is practically impossible. Direct starting on sapropelite fuel oil is possible only after preliminary heating of the cylinder head and prechamber removed from the engine, or after intensive external heating of the engine head, pump, and injector without any disassembly of parts. All this is extremely inconvenient under operating conditions; consequently, it is better each time to use preliminary starting on petroleum motor fuel, with the already running engine then switched over to sapropelite fuel oil.

III. Comparison of Engine Operation on Petroleum Motor Fuel and on Sapropelite Fuel Oil

1. Selection of Engine Operating Conditions

In addition to power and rotational speed, the operating regime of an engine of a given type is also determined by setting the moment at which fuel delivery begins (the angle at which the pump plunger begins to rise—$\varphi'_н$), by the pressure of its atomization (injector pressure—$p_ф$), by the general thermal condition of the engine, etc. The values $\varphi'_н$ and $p_ф$ have the greatest influence on the operating conditions of the engine and, before carrying out comparative tests, it was necessary to establish normal values for these quantities and justify their selection.

The angle at which the plunger began to rise, initially set during installation of the engine in accordance with the firm’s data—about $13^\circ$ before top dead center—proved unsuitable for our motor fuel: the engine would not take normal load, and the indicator diagram (see Fig. 30-a) showed a clear delay in combustion, with the maximum combustion pressure $p_z$ being less than the compression pressure $p_c$. By resetting the screw engagement between the main shaft and the governor shaft (the latter carries the fuel-pump cam), the angle at which the pump plunger began to rise

Fig. 30. Change in the indicator diagram depending on the beginning of fuel delivery.

was brought approximately to 35° before the upper dead center; combustion began at the dead center or even a little before it, and the maximum combustion pressure exceeded the compression pressure by 1.4–1.5 times. Under these conditions the engine acquired the ability to take a normal load with a noticeable reduction in fuel consumption. The indicator diagram obtained in this case is shown in Fig. 30—b.

The angle at which the plunger begins to rise, 35° before U.D.C., should be regarded as normal for this type of engine and fuel, since the start of combustion and the character of the indicator diagram are then normal. The somewhat exaggerated value of the maximum pressures (on the average about 55 ati) should not be a contraindication to choosing this value, since the beginning of delivery determines above all the beginning, and not the speed and character, of the combustion process, on which chiefly the magnitude \(p_2\) depends.

The value of \(p_2\), as well as the entire combustion process, is significantly influenced by the fuel-atomization pressure; its choice was made on the basis of the results of a special auxiliary series of short-duration experiments.

Before carrying out these experiments, the injector spring was calibrated and, by careful measurements, the position of the injector nut corresponding to zero spring compression was established (see Fig. 27). Then, by simple calculations, the dependence of the theoretical atomization pressure on the position of the injector nut was established, and a working diagram was drawn up (Fig. 31), which had to be used in conducting the experiments. The pressure gauge installed on the oil line from the pump to the injector made it possible to measure also the actual pressure of the atomized fuel, which differed somewhat from the theoretically calculated value.

Fig. 31. Dependence of injector pressure on spring compression.

The results of the aforementioned auxiliary tests at various \(p_{\phi}\) are presented by a number of curves in Fig. 32.

All experiments of this series were conducted at the same (approximately normal) engine load, but at each \(p_{\phi}\) the maximum overload accepted by the engine was also determined (the \(N_e\)-max. curve in Fig. 32).

The minimum value of the specific fuel consumption corresponds to values of \(p_{\phi}\) from 106 to 112 ati; however, besides this, at first glance decisive circumstance, in choosing \(p_{\phi}\) one should

one must also take into account the value of the resulting maximum combustion pressure \(p_z\), as well as the engine’s capacity for overload. The desirability of obtaining not excessively large values of \(p_z\) makes it necessary, as far as possible, to choose moderate injection pressures; however, from the character of the \(p_z\) curve it is evident that any appreciable reduction of this quantity could be achieved only by a considerable reduction of \(p_{\phi}\), which would entail a considerable increase in fuel consumption. Moreover, when \(p_{\phi}\) is reduced below 105 atm, the engine already loses the capacity to accept normal overload, as is seen from the curve of maximum powers in Fig. 32.

Taking the foregoing into account, we recognized as normal for petroleum motor fuel the injection pressure value \(p_{\phi}=109\) atm, which corresponded to a nozzle-nut height \(h_{\phi}=59.7\) mm from the igniter in the engine head (see Fig. 27).

When setting up comparative tests, it would have been possible, for each grade of fuel, to establish its own most advantageous values of \(\varphi'_{n}\) and \(p_{\phi}\) and to maintain them during the main experiments. However, the limited amount of sapropelite fuel oil available for the tests did not permit expending it on any additional tests connected with selecting the most advantageous engine operating conditions. Moreover, carrying out comparative tests under exactly the same conditions as on petroleum fuel was of definite practical interest, since it gave a characterization of engine operation on the new fuel without any changes or readjustments in the distribution organs; consequently, for operation on sapropelite fuel oil we adopted the same values of \(\varphi'_{n}\) and \(p_{\phi}\) as for petroleum motor fuel. If, when conducting tests on sapropelite fuel oil with the \(\varphi'_{n}\) and \(p_{\phi}\) selected for petroleum, appreciable deviations from normal engine operating conditions had been found, it would, of course, have been necessary to change the values of these quantities; however, as will be seen below, no need for this was established.

Fig. 32. Results of auxiliary tests for selecting the injection pressure.

Thus, for all series of comparative tests the following values were adopted:

angle of the beginning of plunger lift \(\varphi'_{n}=35^\circ\) before TDC

injection pressure \(p_{\phi}=109\) atm,

which corresponded to the position of the nozzle nut—\(h_{\phi}=59.7\) mm.

The actual values of \(p_{\phi}\) differed somewhat from \(p_{\phi}^{t}\) in the direction of increase, and were different for different fuels and even for individual experiments; this was explained by the hydraulic resistances in the oil pipeline and injector and by changes in fuel viscosity depending on the thermal state of the engine (fuel temperature).

The engine speed was maintained at the normal value in all experiments (about 435 rpm).

The engine power for each of the fuels being compared was varied in separate experiments within the range from idle running to the maximum steady overload.

Other conditions that could characterize the operation of the engine were not determined in advance, but were established automatically depending on the operating mode of the engine.

In order to eliminate the influence of the preceding warm-up on the thermal state of the engine corresponding to a given load, the latter was varied within each series of experiments in increasing order; before the start of each experiment a fully established thermal regime of the engine was attained.

To eliminate the influence on engine operation of incidental circumstances that could not be taken into account, each series of tests on a given grade of fuel was carried out on an engine that had been previously cleaned, and all experiments at individual loads were conducted directly one after another, without stopping the engine.

2. Objectives of the Tests and Selection of Characteristics

Since the principal objective of the tests carried out was to determine the suitability of sapropelite fuel oil as a motor fuel for non-compressor prechamber Diesel engines, the experiments were arranged with maximum attention given to questions of fuel delivery, its atomization, formation of the working mixture, the character and completeness of combustion, etc.

All the data obtained are of interest in comparison with the corresponding characteristics for normal (petroleum) motor fuel, and therefore two series of experiments were carried out—on petroleum motor fuel and on sapropelite fuel oil, with the results considered jointly.

The lack of special equipment did not permit an in-depth elucidation of the processes of working-mixture formation; however, under the existing conditions it proved possible to obtain and consider certain additional characteristics that make it possible, in one way or another, to judge the questions of interest.

Thus, the processes of fuel delivery and atomization were characterized by the operating conditions of the pump (plunger travel, angles of the beginning and end of delivery, taking into account the corresponding position of the regulator coupling) and by the atomization pressure (injector pressure \(P_{\phi}\)), which was determined by the tightening of the injector spring.

and by means of direct measurements with a manometer connected to the fuel line of the pump—injector. The difference between these pressures made it possible to judge the hydraulic resistances of the fuel line and injector, which, among other conditions, depend on the character of the fuel being tested (its viscosity). The temperature of the fuel delivered by the pump also characterized, to a considerable extent, the conditions of fuel delivery and, as we shall see below, had a substantial influence on the combustion process itself.

Conclusions about the character of combustion were drawn from offset indicator diagrams. In addition to the usual value of the maximum combustion pressure, the angle, in degrees of crank angle, from top dead center to the point corresponding to maximum pressure was also determined. This angle varied within considerable limits depending on the type of fuel and the degree of engine warm-up (load).

Of particular interest was the determination of the incompleteness of combustion, since it is directly connected with the conditions of formation and, in general, the quality of the working mixture. To apply here any method claiming to yield absolute data under the existing conditions (especially on a two-stroke engine) did not appear possible. Nevertheless, it was very important to obtain at least a relative characteristic of this phenomenon. Judging the completeness of combustion by the color of the exhaust gases was not sufficiently reliable because of the subjectivity of determining the very concept of “color.” Moreover, during prolonged tests, which usually began in the daytime and ended at night, observation of the color of the exhaust became completely impossible.

In order to introduce a more or less objective characteristic of the incompleteness of combustion, or, more precisely, of the presence of unburned particles in the exhaust gases, an extremely simple method was used, which nevertheless proved sufficiently characteristic.

At a certain distance from the indicator cock screwed into the exhaust pot, a sheet of chalk paper was installed. The cock was opened for a definite time, and from the intensity and character of the deposit obtained on the paper, a judgment was made about the presence of unburned particles in the exhaust gases. The use of a special scale made it possible to express this characteristic in conventional, but definite, numerical units.

Special attention was also paid to the character of deposits inside the combustion chamber; these were judged by a careful internal inspection of the engine after its operation, by photographing the most interesting parts, and by weighing the prechamber before and after the experiments.

The other characteristics obtained in conducting the experiments are the usual ones in all kinds of tests of internal-combustion engines; we shall touch on them when describing the methods of measurement and when considering the results obtained.

3. Methods of measurement, instruments, and their calibration.

The effective power was determined by means of a well-type, self-regulating brake of the Siemens system¹, which clamped a cooled cast-iron pulley mounted on the engine’s main shaft next to the flywheel. The general appearance of the brake installation is shown in Fig. 33.

Fig. 33. General view of the brake installation of the Deutz PME engine.

During the experiments, by tensioning the band (handwheel “a”) and moving the stone (handwheel “b”), such an adjustment of the brake was achieved and maintained that the arm was in constant oscillation, touching the upper or lower edge of the fork c of the fixed trunnion d. The error from the additional moment due to the reaction in trunnion d was thereby practically reduced to zero.

The dead weight of the brake was determined beforehand by weighing it on a special balanced wooden disk with a turned shaft fastened to it, resting on prisms set level. The arrangement of the brake during weighing is shown in Fig. 34.

By separate weighings, the weight reduced to points A (the place where the counterweight is suspended) on an arm from the shaft axis \(L_a = 1.003\ \text{m}\), and B—on an arm (the arm of the main load \(Q_m\)) \(L_b = 1.288\ \text{m}\), was determined. The reduced weights obtained, \(G_a = 17.15\ \text{kg}\) and \(G_b = 13.35\ \text{kg}\) (including the weight pan at point B), when checked by the equality

of the moments \(G_a \cdot L_a = G_v \cdot L_v\) gave complete agreement. At loads of 50% and higher the counterweight was not used, and the working formula for calculating the power here was:

\[ Ne = 0{,}0018\,(Q_m + 13{,}35)\,n \ . \ . \ . (12) \]

where:

\(0{,}0018\)—coefficient obtained with allowance for the length of the brake arm
\[ L_v = 1{,}288\ \text{m}; \]

\(Q_m\)—load on the brake pan, in kg;

\(13{,}35 = G_v\)—the brake’s own weight, reduced to point \(B\);

\(n\)—number of revolutions of the engine per minute.

Fig. 34. Weighing the brake when determining the reduced dead weight.

At a load of 25% the pan for weights was removed, and at point \(A\) a counterweight was fastened, suspended on a cable passed over the fixed pulley \(e\) (Fig. 23); reduced to point \(A\), its weight was \(G_o = 4{,}75\ \text{kg}\). The formula for calculating the power in this case was:

\[ Ne = 0{,}0144\,n \ . \ . \ . \ . \ . \ . \ . \ . \ . \ . (13) \]

where: \(0{,}0144\)—coefficient taking into account, in addition to the ordinary ratios and the arm \(L_a = 1{,}003\ \text{m}\), the weight of the counterweight and the brake’s own weight without the pan, reduced to point \(A\), \(G_a' = 15{,}035\ \text{kg}\).

In carrying out each of the tests, the braking load \(Q_m\) (and, consequently, the braking moment) remained constant, and the effective power of the engine could vary only as a function of the number of revolutions.

The number of engine revolutions was determined by means of a summing anchor counter coupled to the oil-pump shaft (gear ratio from the main shaft 1:3), and was checked and initially set by a stationary tachometer with chain drive from the main shaft, and later by a hand tachometer inserted into the end of the main shaft.

During the experiments, using a stopwatch with two hands, for each reading (observation) the time from the beginning of the experiment was recorded simultaneously with the counter reading. In processing the material, the mean numbers of revolutions for each interval between two readings were derived, and the mean for the experiment from these interval means was checked against the number of revolutions obtained from the initial and final counter readings for the entire experiment and the duration of the experiment. The accuracy of the calculations was to 0.1 rev/min.

Indication and characterization of the working process. Obtaining a satisfactory indicator diagram on a high-speed machine with ignition pressures up to 50 atm presented considerable difficulties. The diagrams initially obtained with Schaeffer & Budenberg and Maihak (Grösse 3) indicators (see, for example, the diagrams in Fig. 30), owing to inertial oscillations of the indicator masses, were clearly unsatisfactory. It was possible to obtain diagrams suitable for processing only after installing a high-speed indicator of the firm Dobbie Mc_Innes Ltd (Fig. 35), with a small piston modified for high pressures. The drum diameter of this indicator is 25 mm.

Fig. 35. Working-cylinder indicator, Dobbie Mc_Innes Ltd.

Nevertheless, the satisfactory external appearance of the indicator diagrams (see below, Figs. 43 and 44) does not yet give the right to regard them as sufficiently accurate, since, in addition to the influence of the masses on the tracing of the diagram, there were also oscillations and stretching of the rather long cord, as well as considerable friction in the cylinder, especially noticeable in the experiments on sapropelite fuel oil, when gumming of the working surface occurred, visible even when it was wiped before each taking of a diagram. As a result, the obtained values of the mean indicated pressure \(p_i\) must be regarded as not entirely reliable. The indicated power, as not sufficiently...

was not determined with sufficient reliability (it is enough to point out that, in approximate calculations, values of the indicated efficiency coefficient were obtained that in some cases reached 70%, with an unnaturally reduced mechanical efficiency).

Despite the insufficient reliability of the values of $p_i$, the indicator diagrams obtained (especially the mixed ones) are of considerable interest from the standpoint of characterizing the working process, and here the results obtained are already sufficiently reliable. As already mentioned above, in processing the mixed diagrams the magnitudes of the maximum pressure $p_z$ and of the angle from TDC to the occurrence of $p_z$, conventionally denoted by the letter $\varphi_{\mathrm{gor}}$, were established.

The values of the degree of pressure rise were also calculated,

$$ \lambda = \frac{p_z}{p_c}, $$

as characterizing the intensity of the combustion process.

Indicating of the crank chamber was carried out with a Schaeffer & Budenberg indicator, Crosby system, with a weak internal spring ($m = 26\ \mathrm{mm}/\mathrm{at}$).

The working springs of all indicators were checked before the experiments in the cold state on a deadweight tester, the mean scale being calculated by Mayer’s method.

Fig. 36. Indicator drive, A. Ya. Tkachenko system.

Of interest is the universal indicator drive used in the experiments (see Fig. 36), the design of which was ...

developed and implemented by the laboratory mechanic A. Ya. Tkachenko.

This drive, owing to the eccentric mechanism’s guide rod, which is movable over the sector, and to the cord being led out from the center of the shaft, makes it possible to record normal diagrams and diagrams displaced by any angle up to 90° without uncoupling the cord, which affords considerable convenience in the mass taking of various diagrams, as was necessary during the experiments. A disadvantage of this drive is the inevitable presence of a long cord, especially undesirable with a high-speed engine; however, by using a well-stretched cord, eliminating heavy couplings (hooks), and placing the longest sections of the cord in the tubes guiding it, it was possible to avoid any appreciable influence of this disadvantage.

Fuel consumption. Measurement of fuel consumption was carried out by the weight method with the aid of a system of pipelines with a siphon lowered into a tank balanced on scales. The layout of the fuel lines and the arrangement of the principal instruments associated with the fuel system are shown in Fig. 37.

Fig. 37. Diagram of fuel lines and instruments for measuring fuel consumption for the Deutz PME engine.

An additional measuring tank C makes it possible to carry out a rapid approximate determination of fuel consumption by volume, and, being filled with fuel other than that in the rest of the system, makes it possible to switch the running engine from one fuel to another.

Measurement of fuel consumption during the main tests was carried out by means of the weighing tank A. Before the beginning of the experiment, tank A was nearly balanced with weights and, when, as fuel was consumed, the pointer of the scales passed through zero, the stopwatch was started. Thereafter, stopwatch readings were taken after each 0.3 or 0.5 kg of fuel had been consumed, which made it possible to judge the uniformity of fuel consumption during the experiment.

In calculating the results, the hourly fuel consumption was determined for each of the time intervals corresponding to the consumption of 0.3–0.5 kg, and the average of these results was compared with the data,

obtained from the total fuel consumption for the entire test and the corresponding time, which made it possible to check the correctness of the calculations.

Since, with the given design of the injector, there was considerable seepage of fuel through the needle packing, the fuel consumption from the tank \(G_b\) did not represent the actual fuel consumption by the engine. The fuel that seeped through the packing in the injector was collected in a special vessel \(D\) (see Fig. 37), which was put in place from the moment the test began and was weighed before and after filling. From the weight of the fuel that had seeped during the test, the hourly drainage of fuel through the injector \(G_{\text{sl}}\) was calculated, and the difference

\[ G = G_b - G_{\text{sl}} \ldots \ldots \ldots \ldots (14) \]

gave the actual hourly fuel consumption for the engine.

Characteristics of the operation of the pump and injector. The stroke of the pump plunger (\(S_{\text{pl}}\)) and the crank angles corresponding to the beginning \((\varphi_n')\) and end \((\varphi_n'')\) of its motion were determined from diagrams of plunger motion, taken with a specially adapted indicator, the drum of which was driven by the common reducing gear, while the rod, without a spring during the recording of the diagram, was rigidly coupled to the ball turned on the handle of the fuel pump.

The motion of the plunger was shown more clearly on diagrams shifted by \(90^\circ\), typical examples of which for normal engine load are shown in Fig. 38.

Initially it was intended to record diagrams of plunger motion directly during the tests. However, it turned out that, owing to the not entirely stable position of the regulator coupling, the diagram (the fuel-pump oblique cam) came out blurred (see Fig. 38-A); recording it, owing to the high speed of the engine, was difficult, and, in addition, there was some distortion of the diagram due to inertial oscillations and lags in the connections of the entire cam—indicator mechanism. All this made it impossible to obtain, while the engine was running, satisfactory diagrams of plunger motion fully suitable for processing.

Fig. 38. Diagrams of motion of the fuel-pump plunger.

In order to obtain an accurate diagram corresponding to the given engine load (the given fuel delivery), during the tests, at each reading (8–10 times during the test), the po-

the position of the regulator coupling on a special millimeter scale fixed opposite the pointer on the tail of the regulator shaft. From these observations the average position of the regulator coupling for the run was determined and then, after the engine was stopped, the regulator was forcibly set to the position corresponding to each run, during which the diagrams of plunger motion “on a cold machine” were taken (see Fig. 38-B) and subsequently processed. Since the pump parts did not undergo appreciable heating while the engine was operating, there was no reason to fear noticeable deviations from the actual operating conditions of the engine when using the “cold” method.

The scale of ordinates of the plunger-motion diagrams was determined by measuring its stroke with an accurate depth gauge on the cold machine while simultaneously taking the diagram.

The accuracy of the results obtained reached, for the plunger stroke, 0.01 mm, and for the angles of the beginning and end of its rise, 0.5° (it would have been difficult to obtain greater accuracy here, owing to the indistinct points of tangency of the lift curve to the horizontals); these angles were determined by means of a template, analogous to the way this was done in determining the angle from the displaced diagram of the working cylinder.

Spraying pressure (nozzle pressure—$p_\phi$), in addition to the above-described calculation from the amount of spring tightening (see Fig. 31), as a result of which a value constant for both series, $p_\phi = 109$ atm, was obtained, was also measured directly by a pressure gauge installed behind the fuel pump. The pressure gauge used had previously been calibrated on an oil press, and in processing the results the corresponding correction was introduced into the obtained mean value of $p_\phi$.

Characteristics of the exhaust gases. A judgment on the mechanical incompleteness of combustion, more precisely—on the presence of unburned particles in the exhaust gases—was made, as was already indicated above, on the basis of the “deposit diagrams.” The arrangement of the devices for obtaining such “diagrams” is shown in Fig. 39.

During the principal tests the device “B,” which made it possible to trap gases and the solid and liquid particles carried by them directly in the exhaust ports, had not yet been fabricated, and the “deposit diagrams” were obtained with the aid of an ordinary indicator cock and an attach-

Fig. 39. Devices (A and B) for fixing unburned particles in the exhaust gases and installation of a thermocouple.

from valve “A.” The chalk paper was pinned to a board set at a distance of 5 cm from the outer edge of the valve, and the latter was opened for 20 sec. (the most suitable time for obtaining “clear” diagrams).

The diagrams obtained with this apparatus showed a considerable change in the character of the deposits depending on the operating conditions of the engine.

To assess the degree of incomplete combustion, a conditional scale with ten gradations (from 1 to 10) was established.

A higher number on this scale indicated greater incompleteness of combustion.

When compared with this scale, each diagram was assigned a definite number depending on the intensity of the darkened field.

The temperature of the exhaust gases was measured by two iron-constantan thermocouples installed, as shown in Fig. 39, on both sides of the exhaust channel.

The readings of these thermocouples were compared, by specially arranged experiments, with the readings of a standard thermocouple, on the basis of which the corresponding correction was introduced.

Measurement of the cooling-water consumption was carried out by direct weighing when it was drained into an iron tank set on scales. The temperature of the water leaving the head was kept as constant as possible (50–55°C), while its consumption varied in accordance with changes in the engine operating conditions.

The temperatures of the cooling water at inlet and outlet, of the fuel in the tank and before the pump, and of the ambient medium were measured with ordinary mercury thermometers, checked before the tests at 0° and 100°C.

4. Comparative tests and their results.

For the purpose of comparing the results of using sapropelite fuel oil as fuel for the above-described type of compressorless Diesel engines with analogous results for petroleum motor fuel, two series of experiments were carried out at different engine loads—on petroleum motor fuel (series H-Ne) and on sapropelite fuel oil (C-Ne). Each series included a number of 24–40 minute tests at loads of 0, 25, 50, 75, 100, and 110 percent of normal, as well as a short-term test at the maximum steady overload of the engine.

During each test, observations were made, at a signal every 4 minutes, of:

Observation	Symbol and units
injection pressure	$p_{\phi}^{m}$ atm,
position of the governor coupling	$h_{p}$ mm,
temp. of cooling water at inlet	$t_{v}^{\prime}$ °C,
“ “ “ “ outlet	$t_{v}^{\prime\prime}$ °C,
number of engine revolutions	$n$ rev/min,
temp. of exhaust gases	$t_{og}$ °C,

and indicator diagrams of the working cylinder were also taken (normal and offset), as well as of the crank chamber, and a diagram of the motion of the plunger of the fuel pump.

Before and after each test the room temperature—$t_0$, the temperature of the fuel in the tank $t_b$ and before the pump $t_m$, and the barometric pressure were observed.

The fuel consumption from the tank $G_b$ and its discharge through the injector were determined throughout the entire test.

When passing from one test to another (change in the operating regime of the engine), before observations were begun a fully established operating regime of the engine was attained; its presence was checked from the fuel-consumption data, from the character of the indicator diagrams, and from the presence of unburned particles in the exhaust gases (“deposit diagrams”).

The results obtained in carrying out the tests and after processing their materials are summarized in the table, and are also presented as a series of curves in Figs. 40, 41, 42, and 45.

Analyzing this material and comparing the operation of the engine on petroleum motor fuel and on sapropelite fuel oil, the following conclusions may be drawn.

Fuel and heat consumption. The character of the change in the total hourly fuel consumption and in the specific consumption per effective horsepower-hour (Fig. 40) is the same in both cases (petroleum motor fuel and sapropelite fuel oil) and agrees with the usual data obtained for engines of the type under consideration. The smoothness of the curves obtained and the uniform distribution of the points confirm the correctness and reliability of the results obtained.

Fig. 40. Fuel-consumption curves.

At small and medium loads the consumption of sapropelite fuel oil (by weight) is somewhat greater than that of petroleum motor fuel, which is evidently explained by the poorer conditions of atomization and combustion (see below), when at insignificant temperatures of the entire engine the fuel is supplied to the injector insufficiently heated, and the increased viscosity is manifested.

and the increase in the surface-tension coefficient of sapropelite fuel oil. At loads of about 80% the curves intersect (the consumption of petroleum motor fuel and sapropelite fuel oil by weight is the same), and thereafter—at full load and overload—the consumption of sapropelite fuel oil becomes lower.

The conclusion concerning the effect on fuel consumption of the temperature of the engine and of the fuel supplied to the injector is also confirmed by the curves of fuel temperature before the pump.

When operating on petroleum motor fuel, artificial preheating of it was not used at all; the rise of \(t_m\) before the pump with increasing engine load is explained by heating of the fuel flowing through the oil line from the nearby exhaust pipe (see the fuel-piping diagram, Fig. 37). When operating on sapropelite fuel oil, before the start of all tests the fuel in the upper tank \(A\) (Fig. 37) was heated to \(45\text{–}50^\circ\mathrm{C}\); during the tests, however, further heating of the tank was discontinued, as a result of which there is observed first a drop in \(t_m\), and then a rise (at loads of 50% and above) due to heating from the exhaust pipe. The rise of \(t_m\) of sapropelite fuel oil is especially pronounced under overloads, which, evidently, also explains the reduction in its consumption under these conditions.

If one takes into account that, in passing from the pump to the injector through the drillings in the engine head, the fuel is heated still more (see above: starting tests), and that the influence of engine warm-up at considerable loads must be still greater here, the influence of the thermal state of the engine on the conditions of atomization and combustion, and consequently also on fuel consumption, becomes more obvious.

When comparing not the weight consumption of fuel, but the consumption of heat released during its combustion, owing to the somewhat lower calorific value of sapropelite fuel oil (9520 cal/kg as against 9960 cal/kg for petroleum motor fuel), the improvement in engine economy (see Fig. 41) when it operates on sapropelite fuel oil at high loads becomes more noticeable. The effective (economic) efficiency at low loads is practically the same, both when operating on petroleum motor fuel and on sapropelite fuel oil. Beginning from 75–80%, \(\eta_e\) when operating on sapropelite fuel oil becomes higher, and at 100% load this increase reaches 4% (the adopted

Fig. 41. Heat consumption and effective efficiency.

on petroleum motor fuel as 100%, increases from 27.5% to 28.6%, with a corresponding decrease in heat consumption from 2300 to 2210 cal/e.h.p.-h. (the direct test data have been somewhat corrected in accordance with the character of the curves).

Operation of the pump and atomization conditions. The operating conditions of the fuel pump are determined primarily by the position of the governor coupling, which in both cases (petroleum motor fuel and sapropelite fuel oil) falls from the highest position at idle, as the load increases, almost to zero at maximum overload, which fully corresponds to the usual operating conditions of the engine. It should be noted, however, that at small and medium engine loads the position of the governor coupling when operating on sapropelite fuel oil is considerably lower (see Fig. 42).

Fig. 42. Characteristic of operation of the fuel pump and temperature of exhaust gases.

This circumstance is directly connected with the increased fuel consumption under the given conditions. Beginning at approximately 75% load, the position of the governor coupling is practically the same both for operation on petroleum motor fuel and on sapropelite fuel oil. The discrepancy in the values of \(h_{\text{reg}}\) at overloads should be explained either by accident, or else by the somewhat greater value of the braking torque overcome at maximum engine overload during its operation on sapropelite fuel oil (see line 7 of the summary table).

In accordance with the position of the governor coupling, the stroke of the pump plunger also changes (we note that, in this design of the fuel pump, the entire plunger stroke is useful), from minimum values at idle to maximum values at overloads. At small and medium loads \(S_{\text{pl}}\), when operating on sapropelite fuel oil, is considerably greater.

As regards the moments of the beginning and end of plunger motion, determined by the corresponding angles of the crank circle \(\varphi_{n}'\) and \(\varphi_{n}''\), the first of these (see line 27 of the summary table) is practically constant (about \(35^\circ\) before top dead center), while the second in both cases varies from maximum values (the earliest end of fuel delivery) at idle to minimum values at overloads. In accordance with the greater pump-plunger stroke when operating on sapropelite fuel oil at small and medium loads, the values of \(\varphi_{n}''\) here are somewhat smaller, which corresponds to a later end of fuel delivery.

The duration of the fuel-delivery period and the average speed of the pump plunger for the most characteristic loads of both series of experiments are given in the following table:

Table IX.

Type of fuel	Type of fuel	Petroleum motor fuel	Petroleum motor fuel	Petroleum motor fuel	Sapropelite fuel oil	Sapropelite fuel oil	Sapropelite fuel oil
Approx. engine load	Approx. engine load	0	100%	Max.	0	100%	Max.
Fuel-delivery period	In degrees of crankshaft rotation, $\varphi'_n-\varphi''_n$	20.7	31.5	35.0	24.0	30.0	33.0
Fuel-delivery period	In time, $\tau$ sec.	0.0080	0.0122	0.0131	0.0094	0.0116	0.0129
Plunger stroke $S_{\mathrm{pl}}$, m	Plunger stroke $S_{\mathrm{pl}}$, m	0.00096	0.0026	0.00279	0.00139	0.00261	0.00292
Average plunger speed $c=\dfrac{S_{\mathrm{pl}}}{\tau}$, m/sec	Average plunger speed $c=\dfrac{S_{\mathrm{pl}}}{\tau}$, m/sec	0.120	0.213	0.213	0.148	0.225	0.226

As can be seen from the table, the average plunger speed at full and maximum load is, in both cases, considerably greater than at idling. In general, when the engine operates on sapropelite fuel oil, somewhat higher speeds of plunger motion occur than in the petroleum series.

Although the injector spring was tightened identically in all experiments, the fuel pressure actually observed on the pressure gauge in the pump–injector pipeline, $p^m_{\phi}$, varied within considerable limits. The curves of variation of the quantity $p^m_{\phi}$ (Fig. 44) first of all indicate considerably higher values of this quantity when operating on sapropelite fuel oil, which is evidently explained by the greater viscosity of the latter. The character of the change in $p^m_{\phi}$ as a function of load also proved to be different.

When operating on petroleum motor fuel, an increase in load from zero to 50% is accompanied by a fall in the value of $p^m_{\phi}$ owing to the decrease in the viscosity (and consequently also in the resistance in the fuel lines) of the fuel as the engine warms up. Subsequently, the increase in the rate of fuel flow already becomes predominant, and $p^m_{\phi}$ rises again.

When operating on sapropelite fuel oil at small loads, the preliminary heating of the fuel in the upper tank still made itself felt, and immediately upon increasing the load $p^m_{\phi}$ rises because of the increase in the fuel velocity. In passing from medium load to full and maximum load, however, owing to a more intensive, than at ...

of the petroleum series of increase in fuel temperature (see Fig. 41), and also because of the sharp change in the viscosity of the sapropelite fuel oil with increasing temperature, the value of $p_{\phi}^{m}$ again decreases.

In general, it should be noted that, when the engine operates on sapropelite fuel oil, the temperature of the fuel supplied to the injector has a special effect on the atomization process; the latter is determined chiefly by the degree of heating of the engine (heat transfer from the exhaust pipe and, especially, from the cylinder head as the fuel passes through the drillings in it).

Combustion conditions, exhaust characteristics, and deposits. Figs. 43 and 44 show characteristic indicator diagrams for both series of tests. Even a cursory examination of these diagrams (especially the offset ones) shows that the combustion process proceeds, in individual cases, far from uniformly.

Fig. 43. Normal indicator diagrams of the Deutz PME engine.

Fig. 44. Offset indicator diagrams of the Deutz PME engine.

Whereas in the petroleum series, even at low loads, there is already a rapid and intense pressure jump, in the sapropelite series there is a noticeable gradual increase in the rate and intensity of combustion as the load increases.

Processing of the offset indicator diagrams gives the values of the maximum combustion pressure $p_z$ and the magnitude

the angle \(\varphi_{20p}\), whose changes, depending on the engine load, are presented in the form of curves in Fig. 45. The change in the value of \(\varphi_{20p}\) with the petroleum series is, generally speaking, insignificant. Only at idling is the lag somewhat greater than the value usual for the other experiments, \(6^\circ\) after TDC, which becomes entirely understandable if one takes into account the slight degree of heating of the engine when it is running idle.

An entirely different picture is obtained when the engine operates on sapropelite fuel oil. At idling the value of \(\varphi_{20p}\) is lowered, evidently artificially, owing to the preliminary heating of the sapropelite fuel oil in the upper tank. Subsequently \(\varphi_{20p}\) gradually falls because of the increasing heating of the engine as its load rises and because of the accompanying rise in the temperature of the fuel entering the injector. At full and maximum loads, \(\varphi_{20p}\), when operating on sapropelite fuel oil, becomes even somewhat smaller than was the case with petroleum fuel.

In complete correspondence with the magnitude of \(\varphi_{20p}\), the maximum combustion pressure \(p_2\) also changes, the fluctuations of which in the petroleum series are insignificant. Here \(p_2\) is somewhat lowered at idling (a cold engine), is almost the same at medium loads, and is again somewhat reduced at full and maximum loads (evidently because of the reduction in combustion speed with strong enrichment of the mixture). When operating on sapropelite fuel oil, \(p_2\) is comparatively low at small loads (the higher value of \(p_2\) at idling is again explained by the influence of preliminary fuel heating), then rises rapidly as the heating of the engine increases, and again decreases somewhat on transition to overloads (insufficient excess air). In general, the maximum combustion pressure at small loads is considerably lower when the engine operates on sapropelite fuel oil and, conversely, higher at full and maximum loads.

Fig. 45. Characteristic of indicator diagrams.

The considered characteristics of the combustion process from the indicator diagrams fully confirm the conclusion made above—

about the substantial influence on the process of atomization and combustion of sapropelite fuel oil exerted by the degree of heating of the engine and the associated rise in the temperature of the fuel supplied to the injector.

In processing the normal indicator diagrams, values of the mean indicated pressure were also obtained; the change in this pressure for both series of experiments is presented by the corresponding curves in Fig. 45. However, as already mentioned above, the high-speed nature of the engine and the absence of a special indicator did not permit the obtaining of fully reliable indicator diagrams; consequently, the obtained values of \(p_i\) should not be regarded as entirely trustworthy. The considerable discrepancy between the values of \(p_i\) for the petroleum and sapropelite series at high loads (and, together with this, the difference in the mechanical efficiency) cannot be considered plausible and is evidently explained by considerable friction of the indicator piston.

The incompleteness of combustion in the individual experiments of the petroleum and sapropelite series was assessed by means of “deposit diagrams” obtained during the tests through cock A (see the diagram in Fig. 39).

The evaluation of the resulting “deposit diagrams” on a conventional scale, established according to the degree of intensity of the deposits obtained by drawing gases from the exhaust ports with the aid of device B (see Fig. 39), is given as a series of figures in line 32 of the summary table. In both cases, there is an increase in the presence of unburned particles in the exhaust gases as the load is increased from idling to 50–75% of normal.

At normal loads the incompleteness of combustion is comparatively slight and reaches a maximum when the engine is overloaded.

No significant difference is observed between the degree of incompleteness of combustion when the engine operates on petroleum motor fuel and on sapropelite fuel oil. For the latter, a noticeable increase in the presence of unburned particles at medium loads is consistent with the deterioration, noted above, of the combustion conditions when the engine is only slightly heated, and with the lowering of the rate and intensity of combustion at these loads, as characterized by the indicator diagrams (increase of \(\varphi_{\text{comb}}\) and fall of \(p_z\)).

The change in the temperature of the exhaust gases for both cases is represented by the corresponding curves in Fig. 42.

No appreciable difference between the petroleum and sapropelite series was observed here.

From the standpoint of characterizing the combustion processes, the results of the internal inspection of the engine, carried out after 10 hours of operation at different loads both on petroleum motor fuel and on sapropelite fuel oil, and the nature of the deposits formed on the surfaces of the combustion chamber, are also of interest.

After operation on petroleum motor fuel, this inspection gave the following results.

The piston crown is covered with a thin coating of soot, burned out in the center, where a brownish coating characteristic of ash deposits is visible.

The cylinder head (see Fig. 46) is uniformly covered with a thin coating of soot, with noticeable brownish spots “в” on one side, where, opposite the openings in the prechamber, the soot has burned out and only a thin layer of ash remains (the light spots “a” visible in the photograph are of no significance, since they were formed as a result of oil flowing out of the injector opening during disassembly of the engine).

Fig. 46. Inner surface of the cylinder head after 10 hours of operation on petroleum fuel at various loads.

The working surface of the cylinder is quite clean and covered with a thin layer of lubricant. The piston rings did not seize.

The prechamber is almost clean on the outside, with a barely noticeable brownish coating. On the inner side of the prechamber there are noticeable ash deposits, accumulated especially strongly on its bottom in the form of a crater-like deposit, located somewhat eccentrically. The deposits in the prechamber amount to 1.2 g.

The head (the prechamber cover) on the inner side is covered with a thin layer of soot. The injector nozzle and the adjacent part of the head are covered with coarse-grained coke deposits (see Fig. 48). The injector opening itself is not coked.

After 10 hours of operation on sapropelite fuel oil, inspection of the piston crown, cylinder head, and prechamber gave exactly the same results as after operation on petroleum fuel. The working surface also proved to be quite clean and covered with a sufficient layer of lubricant. Ash deposits on the inner ...

on the surface of the prechamber after sapropelite fuel oil were, both in quantity and in character, the same as after petroleum fuel.

The inner surface of the head proved to be coked considerably more heavily. In addition to coke deposits near the injector orifice,

Fig. 47. Internal view of the bottom of the prechamber after 10 hours of operation on petroleum fuel.

Fig. 48. Inner surface of the head after 10 hours of operation on petroleum fuel.

orifices, similar to those after operation on petroleum fuel, there (see Fig. 49) significant coke deposits formed at the point where the conical surface changes into the cylindrical one. The injector orifice itself still remained clean.

Fig. 49. View of the prechamber head after 10 hours of operation on sapropelite fuel oil.

Thus, the results of internal inspections of the engine indicate that the deposits occurring after operation on sapropelite fuel oil are, in general, the same as after operation on normal petroleum fuel. The only exception is the considerably larger coke deposits in the dead angle between the head and the prechamber cup, which does not have a substantial effect on the operation of the engine.

Heat Balance of the Engine

Figure 50 presents, in the form of Sankey diagrams, the heat balances of the engine at its normal load for cases of operation on petroleum motor fuel and on sapropelite fuel oil, as well as in the form of expanded balance diagrams at various loads.

Fig. 50. Heat balances of the Deutz PME engine.

The balances at normal load were compiled not directly from the data of the corresponding tests, but in accordance with the course of the averaging curves constructed from the data of all tests in this series.

Thus, the heat consumption at normal load is, for petroleum motor fuel, 2300 cal/hp·h, and for sapropelite fuel oil, 2210 cal/hp·h.

In general, the character of the heat balances remains the same for both cases. Only in the case of operation on sapropelite fuel oil is there noticeable, at high loads, a somewhat increased heat loss with the cooling water and a reduced loss with the exhaust gases.

Operating stress of the engine and its capacity for overload. With regard to the general, external characteristics of the engine’s operation on sapropelite fuel oil, and in comparing the conditions of this operation with those on ordinary petroleum motor fuel, it should be noted that no substantial difference between the two cases was observed. Both on petroleum fuel and on sapropelite fuel oil the engine ran quite calmly, without interruptions, misfires, or excessively intense flashes, although with the somewhat “hard” running usual for this type of engine. No noticeable overloading of parts or heating of the bearings was observed in either case.

As for the maximum overload accepted by the engine, according to the test results (see the summary table), on petroleum motor fuel a load of up to 28.1 hp was steadily accepted, and on sapropelite fuel oil—up to 28.0 hp. It should be noted, however, that in the first case this power was achieved at an artificially, and considerably, increased speed as compared with the normal speed.

If attention is paid not to the power, but to the braking torque overcome by the engine (which is also important from the standpoint of the work of the gases during one cycle), it turns out that the maximum overload on sapropelite fuel oil was somewhat greater, since the braking torque in this case amounted to 46.7 kgm as against 45.2 kgm on petroleum motor fuel (an increase of 3%). In addition, the overload reached was maintained on sapropelite fuel oil somewhat more steadily than on petroleum fuel.

In both cases the maximum overload was determined by reaching the extreme position of the governor sleeve, at which any further increase in the amount of fuel delivered by the pump became impossible. If, however, without fear of overloading the engine, the governor could have been allowed to set an even greater fuel delivery (the oblique cam of the pump permitted the plunger stroke to be increased still further), still greater overloads could have been attained, both on petroleum fuel and on sapropelite fuel oil, until the deficiency of air created excessively unfavorable conditions for combustion.

Since sapropelite fuel oil, owing to its chemical composition, requires a smaller quantity of air, it should be expected that its use ought to make it possible to attain greater overloads than on petroleum fuel.

Thus, with regard to the load accepted by the engine, operation on sapropelite fuel oil gives results no worse, and even somewhat better.

Results of comparative tests of the Deutz RME diesel engine when operating on petroleum motor fuel and on sapropelite fuel oil.

No.	Designation	Symbol	Dimension	Series H—Ne: oil (instrument share of load), 0.0	Series H—Ne: 0.25	Series H—Ne: 0.50	Series H—Ne: 0.75	Series H—Ne: 1.0	Series H—Ne: max.	Series C—Ne: oil (instrument share of load), 0.0	Series C—Ne: 0.25	Series C—Ne: 0.50	Series C—Ne: 0.75	Series C—Ne: 1.0	Series C—Ne: 1.1	Series C—Ne: max.
1	Date of the experiment	—	h/min	17/50	19/00	24/I—32, 20/53	21/40	23/10	25/I, 0/10	11/30	14/10	14/55	11/II—32, 16/52	18/05	19/03	19/40
2	Time from the beginning of the experiment	—	min/sec	32′0″	32′0″	32′0″	32′0″	40′0″	24′0″	24′0″	24′0″	24′0″	24′0″	40′0″	24′0″	12′0″
3	Duration of the experiment	\( \tau \)	min/sec	32′0″	32′0″	32′0″	32′0″	40′0″	24′0″	24′0″	24′0″	24′0″	24′0″	40′0″	24′0″	12′0″
4	Barometric pressure	\(B\)	mm Hg	745	745	745	745	745	745	759	759	759	759	759	759	759
5	Room temperature	\(t_0\)	°C	20.1	20.0	22.0	20.5	17.1	16.8	21.4	18.0	18.8	19.6	19.6	21.4	21.5
6	Number of revolutions of the engine per minute	\(n\)	rev/min	433.6	438.5	432.1	431.2	431.5	446.0	427.0	426.0	436.0	439.0	432.0	430.0	429.5
7	Braking torque	\(M\)	kgm	0.0	13.0	20.6	30.9	41.1	45.2	0.0	13.0	20.6	30.9	41.1	45.2	46.7
8	Effective power	\(N_e\)	hp	0.0	6.32	12.39	18.54	24.75	28.10	0.0	6.13	12.51	18.90	24.82	27.09	28.00
9	Name of fuel	—	—	petroleum fuel	petroleum fuel	petroleum fuel	petroleum fuel	petroleum fuel	petroleum fuel	sapropelite fuel oil	sapropelite fuel oil	sapropelite fuel oil	sapropelite fuel oil	sapropelite fuel oil	sapropelite fuel oil	sapropelite fuel oil
10	Higher heating capacity	\(Q_p^v\)	kcal/kg	10640	10640	10640	10640	10640	10640	10060	10060	10060	10060	10060	10060	10060
11	Lower heating capacity	\(Q_p^n\)	kcal/kg	9960	9960	9960	9960	9960	9960	9520	9520	9520	9520	9520	9520	9520
12	Hourly fuel consumption from the tank	\(G_g\)	g/hour	1759	2345	3415	4151	5590	6945	1573	2360	3370	4648	5840	6548	6800
13	Hourly fuel drain through the injector	\(G_{cu}\)	g/hour	254	163	164	178	145	158	85	24	42	57	82	92	73

No.	Group	Quantity	Designation	Dimension	Series H—Ne, load 0.0	Series H—Ne, load 0.25	Series H—Ne, load 0.50	Series H—Ne, load 0.75	Series H—Ne, load 1.0	Series H—Ne, max.	Series C—Ne, load 0.0	Series C—Ne, load 0.25	Series C—Ne, load 0.50	Series C—Ne, load 0.75	Series C—Ne, load 1.0	Series C—Ne, load 1.1	Series C—Ne, max.
14	Fuel	Actual hourly consumption	$G$	g/hr	1505	2182	3251	4273	5845	6787	1488	2336	3328	4591	5758	6456	6727
15	Fuel	Fuel consumption per effective h.p.	$g_e$	g/h.p.-hr	$\infty$	345.2	262.5	230.5	236.3	241.5	$\infty$	381.0	266.5	243.2	232.0	238.2	240.0
16	Fuel	Temperature in expansion tank	$t_b$	°C	room temp. ~20°C						33.3	32.6	28.2	26.1	24.3	24.3	24.3
17	Fuel	Temperature of fuel before pump	$t_m$	°C	22.7	24.7	26.3	27.8	28.8	30.0	25.2	22.7	22.7	24.6	26.1	29.2	30.5
18	Indicator diagram	Pressure at end of compression	$P_c$	atm	31.0	31.0	31.0	31.0	31.5	32.0	32.0	32.0	31.0	32.0	32.0	32.0	32.5
19	Indicator diagram	Maximum combustion pressure	$P_z$	atm	46.2	50.6	50.1	49.8	46.4	44.0	40.9	38.6	38.8	47.6	49.9	46.3	48.2
20	Indicator diagram	Coefficient of pressure increase	$\sigma=\dfrac{P_z}{P_c}$	atm/atm	1.520	1.665	1.650	1.637	1.505	1.405	1.307	1.236	1.284	1.517	1.590	1.477	1.513
21	Indicator diagram	Angle of ignition lead (conditional, from TDC)	$\varphi_{\mathrm{adv}}$	degrees crank	7.7	6.15	5.9	6.0	5.8	5.6	10.3	12.7	10.7	6.9	5.4	4.8	5.2
22	Indicator diagram	Mean indicated pressure of cylinder *	$P'_u$	kg/cm²	1.43	2.65	3.31	4.34	5.66	—	1.97	2.51	3.18	4.4	4.67	4.94	5.03
23	Indicator diagram	Mean indicated pressure of the engine **	$P^n_u$	kg/cm²	0.139	0.135	0.139	0.127	0.138	—	0.082	0.083	0.126	0.132	0.125	0.104	0.113
24	Indicator diagram	Useful mean pressure **	$P_u$	kg/cm²	1.29	2.52	3.17	4.21	5.52	—	1.89	2.42	3.06	3.91	4.55	4.83	4.92
25	Pump and nozzle	Nozzle stroke	$h_p$	mm	17.0	14.5	11.5	7.0	4.2	3.0	14.1	9.7	6.1	7.7	4.4	2.3	1.25
26	Pump and nozzle	Pump plunger stroke	$S_{n.n}$	mm	0.96	1.32	1.71	2.26	2.60	2.79	1.39	1.81	2.28	2.20	2.61	2.84	2.92
27	Pump and nozzle	Angle of start of fuel delivery by pump	$\varphi'_n$	degrees crank	35.0	36.3	35.0	36.0	34.0	36.0	34.0	36.0	36.0	36.0	34.0	35.0	35.0
28	Pump and nozzle	Angle of end of delivery by pump	$\varphi''_n$	degrees crank	14.3	12.3	8.5	5.0	2.5	1.0	10.0	8.0	6.0	5.0	4.0	2.5	2.0
29	Expansion	By manometer	$P^n_\phi$	atm	121	112	108	116	122.5	129	135	127.5	142	145	139	144	134
30	Expansion	By prolonged line	$P^n_\phi$	atm	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$	$10^9=\mathrm{Const}$

No.	Group	Indicator	Designation / unit	1	2	3	4	5	6	7	8	9	10	11	12	13
31	Exhaust gases	Temperature of exhaust gases	$t_{o2}$, °C	75	81	119	147	194	221	75	87	114	145	183	208	221
32	Exhaust gases	Presence of unburned particles	On a conditional scale	1.0	3.25	5.25	5.75	3.5	9.0	1.0	2.5	6.75	2.0	3.0	8.62	10.0
33	Cooling	Hourly cooling-water consumption	$G_v$, kg/hour	182.6	238.7	275.2	360.8	366.8	475.7	197.0	235.7	308.5	341.5	613.8	428.7	373.6
34	Cooling	Water consumption per effective h.p. hour	$g_v^e$, kg/h.p.-hour	∞	37.8	22.2	19.4	14.8	16.9	∞	38.5	24.3	18.1	24.7	15.8	13.3
35	Cooling	Temperature of water at inlet	$t'_v$, °C	3.0	2.6	2.0	2.0	2.0	2.0	5.9	3.8	3.0	2.0	2.0	2.0	2.0
36	Cooling	Temperature of water at outlet	$t''_v$, °C	56.0	50.6	53.8	51.8	54.5	52.5	50.1	48.3	48.7	53.1	53.6	60.2	55.5
37	Cooling	Increase in water temperature	$\Delta t_v$, °C	53.0	48.0	51.8	49.8	52.5	50.5	44.3	44.5	45.7	51.1	51.6	58.2	53.5
38	Heat balance	Carried away with fuel*)	$Q$, kcal/hour	14980	21730	32400	42570	58200	67600	14170	22240	31700	43730	54830	61480	64060
39	Heat balance	Converted into effective work	$Q^e$, kcal/hour	0	4000	7840	11720	15650	17770	0	3880	7910	11950	15680	17130	17700
40	Heat balance	Carried away with cooling water	$Q_v$, kcal/hour	9670	11440	14270	17970	19260	24000	8730	10470	14100	17450	30650	24950	20000
41	Heat balance	Carried away with exhaust gases and residual losses	$Q_{cт}$, kcal/hour	5310	6290	10290	12880	23290	25830	5440	7890	9690	14330	8500	19400	26360
42	Heat balance	Carried away with fuel	$q$, %	100	100	100	100	100	100	100	100	100	100	100	100	100
43	Heat balance	Converted into effective work (effective efficiency), carried away with cooling water	$q_e$ ($\eta_e$), %	0.0	18.3	24.2	27.5	26.9	26.3	0.0	17.4	25.0	27.3	28.6	27.9	27.6
44	Heat balance	Carried away with cooling water	$q_v$, %	64.6	52.7	44.0	42.2	33.1	35.5	61.6	47.1	44.5	39.9	55.9	40.6	31.2
45	Heat balance	Carried away with exhaust gases and residual losses	$q_{cт}$, %	35.4	29.0	31.8	30.3	40.0	38.2	38.4	35.5	30.5	32.8	15.5	31.5	41.2

Notes: *) The values $P_i$ are not reliable, owing to insufficiently reliable operation of the indicator on the high-speed machine.

**) For the lower heating value.

Part IV.

GENERAL CONCLUSIONS AND SUMMARY

An analysis of the results of studies on the possibility of using fuel oil from the semi-coking tar of sapropelites from the Barzas III deposit in Diesel engines permits the following conclusions to be drawn:

1. The use of this grade of fuel oil as motor fuel for compressor Diesel engines and prechamber non-compressor engines of the Deutz type is entirely possible.

2. Operation of the machine under underload is associated with excess consumption, reaching at quarter load up to 14% for the compressor engine and 10% for the non-compressor engine, compared with petroleum motor fuel.

3. Under normal load, the consumption of sapropelite fuel oil for the compressor engine is 3% lower, and for the non-compressor engine 1.5% lower, while its calorific value is 4.4% lower than that of petroleum motor fuel.

4. The lower value of the theoretically required amount of air for combustion permits operation with greater excess air at the same loads and enables the engine to take a greater prolonged overload.

5. Investigation of the combustion process shows that, for sapropelite fuel oil, during operation at all loads the ignition-delay period is shorter.

This indicates that, for sapropelite fuel oil, one may expect a lower temperature for the appearance of a cold flame and a lower self-ignition temperature than for the petroleum motor fuel used.

The shorter ignition-delay period produces smoother running of the machine, which was observed in the operation of both engines.

1. Starting the compressor engine at temperatures from +15 to +20°C is carried out directly on sapropelite fuel oil.

For starting the prechamber non-compressor engine at the same temperatures, petroleum motor fuel must be used as the starting fuel.

1. When operating on sapropelite fuel oil, the compressor engine runs unevenly at idle and at quarter load. Beginning with half load and up to a 25% overload, its operation is quite stable.

The non-compressor engine operates stably and evenly at all loads.

1. Increased carbon formation, more noticeable in the compressor engine, requires attention to the need to reduce the coking tendency to the value specified by the standards.

1. The short duration of the experiments carried out does not permit an exhaustive evaluation of the operating qualities of this fuel, which may be revealed only through prolonged observations of engine operation under service conditions.

2. Along with this, it is necessary to subject the principal physicochemical properties of the new fuel to more thorough study (the appearance of a cold flame, self-ignition temperature, heat capacity, magnitude of surface tension, etc.), with the aim of establishing a solid scientific and technical basis for its rational use and for introducing the necessary design changes in existing engines in accordance with the specific properties of the new grades of fuel.

Correction. In Part II of the present work, the caption under Fig. 24 refers to Fig. 25, and the caption under Fig. 25 refers to Fig. 24.

Tomsk.

Laboratory of Internal Combustion Engines, S.M.M.I.

1. The brake was designed by Assoc. Yu. N. Sokolov. ↩↩