Diagnosis of Engine Power and Compression Balance

Frequently when a multicylinder internal combustion engine fails to deliver its rated power, the problem arises from the weakness or malfunctioning of one or two cylinders out of a total of, for example, six cylinders. In the case of a diesel engine, malfunctioning of a cylinder may be due, for example, to underfueling caused by failure of the fuel injector, or to worn piston rings and valves which cause loss of compression. Evidence of a malfunction may be roughness of engine operation or poor cylinder compression.

It is of course desirable to be able to identify a weak cylinder, and there are highly skilled mechanics who are able to do so using time consuming techniques. This solution is not, however, satisfactory to an engine manufacturer where many engines must be tested, or to service facilities where skilled mechanics often are not available. Devices such as thermocouples, vibration detectors and cylinder pressure sensors may be used in a laboratory but they generally are not desirable or available elsewhere.

U.S. Pat. No. 4,064,747 describes a method of identifying or isolating a weak or defective cylinder by measuring the instantaneous speed of the engine crankshaft as the engine goes through a complete operating cycle. When the instantaneous engine speed is plotted against time, a cyclically varying curve is obtained, and the maximum speed attained during the expansion stroke of a defective cylinder will normally not be as high as the speed attained for a normal cylinder. This method is not always accurate however, because it does not take into account variations in other engine parameters, such as the masses of the operating parts, inaccurate speed measurements, variations in cylinder compression work, etc.

It is a general object of this invention to provide an improved system for testing the power and compression balance among the cylinders, which eliminates the foregoing disadvantages.

In accordance with the present invention, means is provided for sensing the instantaneous engine speed and, during an acceleration run of the engine, measuring and storing the instantaneous engine speeds during at least one full cycle of the engine. At substantially the same engine speed during a deceleration run of the engine, the instantaneous speeds are again measured over at least one operating cycle. A function representing the change in kinetic energy during each firing interval is computed during acceleration and also during each firing interval during deceleration. For each firing interval, the two functions are subtracted to obtain the work from each cylinder.

The foregoing and other objects and advantages of the present invention will be better understood from the following detailed description taken in conjunction with the accompanying figures of the drawings, wherein:

FIG. 1 is a view of an engine including a diagnostic system in accordance with the present invention;

FIG. 2 is a block diagram showing the diagnostic system in greater detail;

FIG. 3 is a block diagram of a speed measuring part of the system;

FIGS. 4 to 6 are curves illustrating the operation of the engine and the system; and

FIGS. 7-A to 7-S show a Metacode setting out the operation of a processor of the system.

With reference to FIG. 1, an engine 10 is illustrated which may be a standard internal combustion engine such as the NH series, six cylinder, in-line reciprocating piston, diesel engine manufactured by Cummins Engine Company, Inc. Such an engine includes a head 11, a block 12, an oil pan 13 and a rocker housing 14 fastened to the upper side of the head 11. The pistons (not shown) of the engine reciprocate within cylinders (also not shown) and are connected to rotate a crankshaft 66. A flywheel on the crankshaft has a ring gear 62 attached to it, teeth 63 on the gear 62 being selectively engaged by a starter motor (not shown) for starting the engine.

A plurality of fuel injectors 16 inject metered quantities of fuel into the cylinders after inlet air within the cylinders has been compressed sufficiently to cause compression ignition of the resultant combustable mixture. The injectors 16 may be a unit type embodying the features of the injectors shown in U.S. Pat. No. 3,351,288. A common fuel supply rail 17 connects the injectors 16 with a fuel supply system including a fuel pump 18 of the character shown in the U.S. Pat. No. 3,139,875. The fuel pump 18 draws fuel 19 from a reservoir or fuel tank 21 and forms a regulated fuel source for the fuel supplied to the rail 17. A throttle is incorporated in the fuel pump 18 and permits the operator of the engine to regulate the fuel pressure delivered to the injectors. Also connected to each of the injectors 16 is a fuel return rail 24 which carries fuel from the injectors 16 to the tank 21.

The engine 10 further includes a turbocharger unit 31 which may have a conventional design. The unit 31 includes a turbine that receives engine exhaust from an exhaust manifold 32 and it further includes a compressor that is connected by a duct 33 to an air intake manifold of the engine.

The engine 10 further includes a lubricant system for circulating a lubricant such as oil through the various operating parts of the engine. The lubricant system includes a pump 41 that draws the lubricant from a reservoir in the crankcase and pan 13 and pumps the lubricant under pressure to a lubricant rifle passage 42 in the block. The pressure in the rifle 42 is regulated by a pressure regulator valve 43 connected in a bypass line 44 that is connected across the pump 41.

A number of mechanical couplings, illustrated by dashed lines in FIG. 1 and indicated by the reference numerals 67 and 69, connect the crankshaft 66 with the fuel pump 18 and the lubricant pump 41, respectively.

A diagnostic system in accordance with the present invention is provided, and includes a cycle event marker (CEM) sensor 51 which is preferably mounted in the rocker housing 14 and responds to the movement of an operating part of the engine. For example, the CEM sensor 51 may be a magnetic coil proximity type sensor that is mounted adjacent the rocker arm that actuates the injector 16 of the number one cylinder. This rocker arm pivots during injection which occurs toward the end of the compression stroke of the piston of the number one cylinder and this movement causes the sensor 51 to generate a CEM signal toward the end of the compression stroke of the piston of the number one cylinder. The CEM signal is utilized in testing engine parameters as will be subsequently described.

The diagnostic system still further includes an engine speed sensor 61 that is mounted adjacent to the outer periphery of the flywheel ring gear 62 of the engine 10. FIG. 3 illustrates an example of the sensor 61 and the circuits connected to it. The sensor 61 has two spaced elements 91 and 92 which in the present specific example, are variable reluctance magnetic sensors. The teeth 63, moving clockwise, generate signals first in the element 91 and then in the element 92. An oscillator 93 is connected to a counter 94 which is controlled by the tooth pulses from the elements. A pulse from the element 91 operates through circuits 96 and 97 to enable or start the counter 94 and a pulse from the element 92 operates through circuits 98 and 97 to disable or stop the counter. The count associated with each tooth is read by the processor 29. Each count is directly proportional to the time interval (.DELTA.t) for a tooth to move from one element 91 to the other element 92, and inversely proportional to the instantaneous speed of the ring gear. A factor for converting counts read to engine RPM may be provided as an input to the processor 29 based on physical measurements, such as the spacing X between the elements 91 and 92 and the radius R of the elements 91 and 92 or may be computed within the processor based on signals from the cycle event marker sensor 51. The CEM sensor 51 is connected through circuits 95, similar to the circuits 96-98, to a CEM counter-register 100. The signal from the oscillator 93 is connected through the divider 99 to the register 100, and the register 100 output is connected to data lines of the processor 29.

The diagnostic system further includes a number of other engine sensors including a fuel pressure sensor 27 connected in the rail 17, a lubricant pressure sensor 46 connected in the rifle passage 42, and an intake manifold air pressure sensor 34 connected in the intake manifold. The sensors 51 and 61 are connected to a counter-timer module 22 and the sensors 27, 34 and 46 are connected to an A/D convertor 23, the components 22 and 23 being connected to control and data lines of the processor 29. The processor 29 provides outputs of a readout device 70 which may provide, for example, visual indications and permanent records.

FIG. 2 illustrates the diagnostic system in greater detail. The processor 29 includes a processing unit 71 and a memory unit 72. An operator interface 73 is connected to the unit 71 and forms means whereby the operator may insert information and instructions and includes the readout 70. The processor utilizes the signal from the CEM sensor 51 which is shown mounted in a position to sense the movement of a rocker arm 74 for an injector plunger 76. A cam 77 moves the plunger 76 in an injection stroke toward the end of the compression stroke.

The components 22, 23, 29 and 73 may comprise, for example, standard products of Texas Instruments Company.

FIG. 4 illustrates the variation in the engine torque output, at the crankshaft and the ring gear, for a six-cylinder four-stroke engine. The torque varies or fluctuates as shown about the mean absorbed torque with the engine operating at a steady speed. The crankshaft makes two complete revolutions for each engine cycle, and the firing interval of each cylinder is 120.degree. long. Torque peaks 101 to 106 appear during the combustion strokes of the cylinders, and the relatively low peak 104 illustrates the characteristic of an underfueled cylinder. Top dead center (TDC) at the start of the combustion stroke of the number one cylinder is indicated by the number 108. If a cylinder firing interval is divided into three equal 40.degree. segments, for a normal cylinder, about 52% of the total work is produced in the first segment and 87% of the total is produced by the end of the second segment. For a cylinder that is weak due to improper burning, for example, about 40% of the total work is produced during the first segment and 80% is produced by the end of the second segment.

FIG. 5 shows the torque output vs crank angle of a single cylinder over one firing interval, and illustrates the difference in torque between engine acceleration and deceleration. The acceleration curve 111 represents the condition when power is being generated as during a full open throttle and free engine acceleration, and shows the high peak torque caused by burning of fuel. The peak on the deceleration curve 112 is produced by the expansion of air in the cylinder without fuel combustion, as during deceleration when no power is being generated. The curve 112, particularly the shoulder 113 shows the effect of the compression in the next cylinder in the firing order and the inertia torques. Of course, if the compression and torque influences shown in curve 112 vary from cylinder to cylinder, the curve 111 would also vary and be an unreliable indication of power balance. Prior art systems based only on acceleration rate are not able to remove those factors from consideration.

When the deceleration curve 112 is subtracted from the acceleration curve 111, the torque or work output due to the combustion of fuel is derived for a single cylinder, and the torque for a normal cylinder over one cycle is represented by the curve 114 in FIG. 6. By subtracting the acceleration and deceleration curves to produce the curve 114, in accordance with this invention, a number of factors are eliminated, such as apparent speed variations due to ring gear faults, variations in the inertia of the rotating engine parts, variations in the compression work of the cylinders, and engine friction. The integral of the curve 114 thus represents the work produced by a single cylinder over one firing interval. The curve 116 is a similar curve but shows a deficient cylinder, and shows the lower work output or torque of a weak cylinder.

To determine the work produced by each cylinder in accordance with this invention, and thereby to determine the power balance of the cylinders, the speed signals from the sensor 61 and the CEM signals from the sensor 51 are transmitted to the processor 22 which operates in accordance with the Metacode shown in FIGS. 7-A to 7-S.

The Metacode is an abstracted flow chart of the steps to be executed by the processor; a complete program will be obvious to those skilled in the art from the Metacode and the present description. While the Metacode is a sufficient basis for the preparation of a program to carry out the invention, the following discussion is provided to aid in understanding the system and the Metacode.

Broadly, the present invention comprises measuring the instantaneous engine speed utilizing the sensor 61, the speed measurements being in terms of .DELTA.t or the time interval for a tooth 63 to move from one element 91 to the other element 92. The angular distance, measured in ring gear teeth 63, from a CEM signal to the next subsequent TDC, and the angular distances between adjacent TDCs, are determined to identify the TDC locations. During an acceleration run from a selected speed, the time interval data are obtained for at least a full cycle of engine operation. A function representing the change in kinetic energy from each TDC location to the next subsequent TDC is determined by squaring the angular speeds at both TDC points and finding their difference. Similar time interval measurements at the same engine speed are made during deceleration, and the functions representing the energy over the firing intervals are determined. Then, for each firing interval, the change in kinetic energy on deceleration is subtracted from the change in kinetic energy on acceleration. This change in kinetic energy is related to the work done by the engine by the relationship ##EQU1## The work done during a firing interval is predominantly influenced by the combustion of fuel in the cylinder whose power stroke occurs during the firing interval being examined. The works for the various cylinders may then be compared or ranked to obtain the power balance.

The kinetic energy at any instant is a function of the angular speed of the crankshaft 66 and the ring gear 62. In a test of a free engine, that is where no torque (T) is externally absorbed by a dynamometer or other load, the torque fluctuation is represented by a variation in engine acceleration as related by the function T=I.alpha. where I is the inertia and .alpha. is the angular acceleration. In the method disclosed herein of measuring the engine speed, the sensor 61 and the related circuitry determine the time interval .DELTA.t required for a tooth to move from one sensor element to the other. The time interval is an inverse function of ring gear angular velocity .omega. as follows:

The kinetic energy is

or where I is the engine inertia and K.sub.1 is a constant.

Thus, the instantaneous engine speed is related to the instantaneous torque output and to the kinetic energy of the engine.

While the Metacode includes a number of definitions and comments, the following additional commentaries may be useful. CEMRK is produced by the sensor 51 which, in this example, produces a pulse just ahead of top-dead-center of the number one cylinder. The dual-pole sensor is the speed sensor 61. The variable DLTWSQ represents .DELTA..omega..sup.2 which is a function of the change in KE over an angular interval. TIMINT is a time interval .DELTA.t reading from the dual-pole sensor. PZX stands for positive zero crossing. When the kinetic energy over an engine cycle is plotted vs. time, the point at which torque stops being absorbed (negative KE) and starts to be produced (positive KE) is marked by a zero crossing from negative to positive (PZX). This crossing substantially coincides with the top dead center (TDC) when compression ends and expansion begins. Thus the TDC location of each firing interval may be located by plotting the change in KE and identifying the positive zero crossings. The crossing data are stored in (ARRAY) PZX. The OFFSET means the distance, measured in ring gear teeth, from CEMRK to the next PZX (or TDC). As to PROBCAL, the RPM equals PROBCAL divided by .DELTA.t and is a function of X/2.pi.R (see FIG. 3 for X and R).

FIG. 7-C shows the power balance routine. The operator is prompted to insert various information describing the engine under test. Certain data may be stored on a conventional media such as cassette tape or floppy disc. These data may be accessed according to an engine model number on code supplied by the operator. Thus the operator need not supply the detailed information needed to perform the test. NUMCYL means the number of engine cylinders and FIRORD means the firing order. The test to be performed is the power balance and one or more runs or samples may be made. As to the threshold values, in an acceleration run of the engine from, for example, 600 RPM to 2400 RPM, the time interval data for the power balance test may be collected at approximately 1000 RPM. An engine of the type previously described accelerates at the rate of about 200 RPM per engine cycle; consequently the threshold instantaneous speed to initiate data collection may be approximately 800 RPM. During the subsequent deceleration run, the threshold instantaneous speed may be set at about 1050 RPM in order to accumulate data at about 1000 RPM, because an engine decelerates much more slowly than it accelerates. The system then calls the SETUP subroutine (shown in FIG. 7-D) and then calls the EXPANSION subroutine (shown in FIG. 7-I). SUMINT means twice the number of teeth on the ring gear, that is, the number of teeth passing the sensor 61 in one full engine cycle.

In the SETUP subroutine (FIG. 7-D), LOSPEED indicates a threshold value such as 600 RPM. TIMINT readings are taken over at least an engine cycle and stored in a buffer. The gear tooth corresponding to the occurrence of the CEM mark is determined. The positive zero crossings, which correspond to the top-dead-centers, are determined and the corresponding gear teeth are identified. This information and the firing order enables the TDC locations to be identified in terms of ring gear teeth. These data are utilized in the power balance test as previously noted. The SETUP subroutine calls up a number of other subroutines which accomplish these steps and are shown in the drawings.

The ACQUIRE subroutine, (FIG. 7-F) acquires the .DELTA.t values and the CEM values over at least one engine cycle, from the counters that receive the oscillator 86 cycles during the .DELTA.t and the CEM time intervals. The CALIBRATE subroutine (FIG. 7-G) determines a factor used in determining RPM from the .DELTA.t measurements. The SMOOTH subroutine (FIG. 7-H) smooths and edits the data to eliminate wild points, in accordance with standard techniques. The EXPANSION subroutine computes the work done during a firing interval, and in EXPWORK, the work during acceleration is subtracted from the work during acceleration for each interval. In the IDENTIFY subroutine (FIGS. 7-I and 7-K) the cylinders and their TDC locations are identified relative to the CEMRK. The commentaries for the subroutines of FIGS. 7-L to 7-Q describe the functions.

The operation of the system may again be briefly summarized as follows: The instantaneous speed data are accumulated in the form of time intervals .DELTA.t, the data being received from the sensor 61 and the counter timer 94. Other data are initially loaded into the system such as the number of points or .DELTA.t measurements to be taken and the engine threshold speeds at which the measurements are to be taken. The values are stored in a buffer TIMINT which also receives the CEM signal. The buffer stores the .DELTA.t values, and it provides an index identifying the location of the .DELTA.t value that occurs at the same time as the CEM signal. The SETUP routine accumulates and processes the data in preparation for the acceleration and deceleration runs. The data in the TIMINT buffer are edited and smoothed using standard techniques. The instantaneous kinetic energy is computed from the inverse of .DELTA.t.sup.2 values. The processor utilizes the CEM signal and the zero crossing data, and provides an identification of the cylinder top dead center next following the CEM signal. The offset is computed in ring gear teeth from a CEM signal to the next positive crossing (PZX or TCD). From the engine firing order and the total number of teeth on the ring gear the processor computes the number of ring gear teeth between TDCs, which is the interval between power strokes (INTBPS). Thus, the number of gear teeth from the CEM to the TDC of each cylinder is computed. The processor also computes the instantaneous engine RPM using the PROBCAL conversion factor.

After the data are accumulated during the acceleration run, the time interval measurements at the top dead centers of the cylinders are utilized to compute the change in kinetic energy from each TDC to the next TDC. Similarly, after the subsequent deceleration run, the changes in kinetic energy between TDCs are computed during deceleration. The kinetic energy changes are correlated with the associated cylinders or firing intervals, utilizing the OFFSET and the INTPBS data. Using the measurement for each cylinder or firing interval, the kinetic energy change during deceleration is subtracted from the kinetic energy change during acceleration, to produce the work of each cylinder. To compare the cylinders, the work values may be averaged and the average values of the cylinders may be ranked.

Instead of accumulating data in only one acceleration run and only one deceleration run, there may be a number of acceleration runs and the data averaged, and a number of deceleration runs and the data averaged.

The system may also provide for determination of the compression balance of the cylinders as shown by the routines of FIGS. 7-R and 7-S. The processor receives the time interval and the CEM signals, and the OFFSET and the INTBPS factors. A low threshold speed is set in it, which is below normal low idle speed. When the speed is sufficiently low and the fuel is cut off to produce deceleration at low speed, the variation in inertia forces between cylinders is small as compared to the gas forces, and consequently the kinetic energy change over a segment of a firing interval, such as 40.degree., preceeding TDC represents the compression work done on the gas. The processor receives the data, divides each firing interval into a number of segments, and computes the change in kinetic energy in each segment. There may be, for example, three equal 40.degree. segments in each interval. The data for the segments just preceeding the top dead centers are compared and ranked in order to determine the relative work of compression for the cylinders.

It will be apparent from the foregoing that a novel and useful system has been provided for diagnosing the health of the cylinders of an engine. The system determines the power balances and is able to identify a weak cylinder, by sensing the kinetic energy of the engine. By this method, a number of factors that could lead to error are eliminated by the subtraction of the energy during deceleration from the energy during acceleration. The work from each cylinder is thus determined. It is preferred that the work be determined by first measuring the changes in kinetic energy over the firing intervals both on acceleration and deceleration and then taking the difference, because the method may be carried out in a straight forward manner as described. However, the invention is also broad enough to encompass other systems for measuring the work, such as by taking a number of readings at a number of angular positions over each interval both during acceleration and deceleration, correlating the readings at each angle, taking the difference at each angle, and computing the work from the differences. The system also provides for determining the compression balance by measuring the changes in kinetic energy over the part of each interval that just preceeds top-dead-center during the compression stroke.

When diagnosing power balance or compression balance, the values for the cylinders may be produced for evaluation by the system operator, or the processor may automatically rank the cylinders, or the values may be compared with acceptable reference values. Of course other calculations or tests may be performed based on the kinetic energy calculations.