This application incorporates by reference application Ser. No. 08/442,218, now U.S. Pat. No. 5,600,125, by the same inventors entitled Compensation and Status Monitoring Devices for Fiber Optic Intensity Modulated Sensors.
Four stroke direct-injection engines have been under recent intense development for both diesel and gasoline fuel applications due to significant fuel economy improvements and reduced emission levels in comparison to engines with conventional fuel injection. Direct injection diesel engines have a higher baseline thermal efficiency (about 40% peak), 20-35% better fuel efficiency, 10-20% lower CO2 emissions, near-zero evaporative emissions, and low cold-start emissions. Fuel economy improvements of as much as 35% have been recently reported, combined with a simultaneous increase in engine power and torque of 10%, for a direct gasoline injection engine. Such remarkable performance has been realized through a combination of very lean burn combustion (Air-fuel ratios as high as 40:1) and stratified charge mixing inside each engine combustion chamber.
A key component that is required for both Direct Diesel Injected (DDI) and Direct Gasoline Injected (DGI) engines is an accurate and cost-effective fuel injector. In diesel engines, the new injection must operate at extremely high pressures (as high as 30,000 psi), provide accurate and repeatable spray patterns, and be precisely timed. In addition, such an injector must operate for as many as 0.5 million miles and be of low cost.
In DGI applications, in particular, the very poor lubricating nature of gasoline and critical injector specifications make gasoline injectors difficult and expensive to manufacture.
To provide the required performance, reliability, and low-cost for both DDI and DGI injectors, disclosed below are approaches based on closed loop control of injector operating parameters, where combustion chamber and fuel pressures are used as control parameters. As detailed below a preferred way of obtaining these two pressures is to integrate two miniature fiber optic pressure sensors inside an injector. Such a “smart” injector does not need to be individually calibrated, as currently done, so its price can be significantly lower. Differences caused by manufacturing variability, aging, pressure line fluctuations, or fuel quality can be compensated for by using closed-loop control of fuel injection timing, duration and pressure. The combustion chamber pressure sensor of the smart injector provides, in addition, real-time information about cylinder pressure including peak pressure (PP), indicated mean effective pressure (IMEP), start of combustion (SOC) and location of peak pressure (LPP). When inputs from both fuel and combustion pressure sensors are used to control the injector, simultaneous benefits of reduced emissions, improved fuel economy, increased injector reliability, and reduced cost can be achieved.
Invention Overview
The fiber optic sensors utilized in the smart injector are of a novel construction aimed at high accuracy in a very small device exposed to extremely high pressures and temperatures. The sensor tip may be as small as 2.5 mm in diameter or smaller. By using a specially shaped diaphragm in the sensor and two D-shaped optical fibers, high levels of optical modulation can be realized at small diaphragm deflections. Small diaphragm deflections are required to permit high diaphragm yield strength and long fatigue life. Using a two photodiode detection technique, each sensor's signal interface/conditioner can operate accurately over a temperature range of −50 to 150° C.
In a preferred configuration two types of sensors are used in the smart injector: (1) a high pressure sensor for monitoring static fuel pressures inside the injector and (2) a sensor for detecting dynamic combustion chamber pressures.
To compensate for changing sensor response with temperature, a temperature compensation technique utilizes a combination of a thin film deposited on the diaphragm inner surface and a temperature probe mounted in the sensor housing. The thin film reflection coefficient changes with temperature thereby compensating for any intra cycle (short term) diaphragm temperature excursions above its average temperature. Any longer term errors, resulting from increased diaphragm deflection at higher average temperature and other thermal effects on the sensor head, are compensated for by adjusting the pressure sensor's gain, based on the temperature probe output.
While most direct injectors require dynamic fuel pressure information, only static pressures must be known in such approaches as fuel rails. Disclosed below is a static pressure sensor utilizing two optical fiber pairs, with one pair acting as a reference device to compensate for errors that may result from temperature effects on opto-electronic components, fiber bending, or other sources of undesirable light intensity fluctuations.
As above, to compensate for errors arising from intra-cycle diaphragm heating due to the nearby combustion gasses, the compensation technique is similar to the technique used for dynamic sensors. For the static sensor one pair of fibers is exposed to diaphragm deflection and the other pair of fibers is installed in front of a non-deflecting reflector coated with a reflection temperature dependent thin film. A separate temperature probe provides input information for additional compensation for sensor gain and offset changes arising from varying inter-cycle average diaphragm and housing temperatures.
In
The combustion pressure sensor 14 is exposed to combustion chamber gases through a short channel 22. The fiber optic lead 24 for this sensor 14 is located inside a 1 mm diameter cylindrical hole 26 extending from the sensor 14 to the strain relief 28 attached to the injector 10 near the top of the injector.
Each of the sensors is connected to its own opto-electronic module or connector as further discussed below. The connections are by the fiber optic cables 30 and 32. The opto-electronic modules contain circuity which in turn is connected to the engine Electronic Control Module (ECM). The ECM controls the injector timing, duration, and fuel line pressure in response to the output of the sensors 12 and 14 as well as other engine sensors and controls.
The preferred sensor design for the fuel and combustion pressure sensors above utilizes two or more optical fibers. These fibers are modified as illustrated in
The minimum distance between fiber cores 38 is highly desired for reduced sensor power loss and increased optical modulation, both resulting in improved sensor resolution, signal to noise ratio and output accuracy. In a typical configuration the fibers have core 38 diameters of 140 microns, cladding 36 of 170 microns and flat surfaces 34 separated from the cores by 5-10 microns. For ease of sensor assembly and repeatability, the fibers are semi-permanently bonded together during the fiber manufacturing process by a thin layer of polyamide or other temperature-capable bonding adhesive material.
In
Under optimized conditions, the curvature of the reflecting surface 48 is ellipsoidal to focus the light from the delivering fiber 42 onto the collecting fiber 44 with no external pressure applied. With the application of external pressure, the diaphragm 40 deflects and the ends of the delivering 42 and collecting 44 fibers are no longer at the foci 54 of the ellipsoidal curved reflecting surface 48 as shown in
The ellipsoidal curvature surface 48 and foci 54 locations are optimized for a given fiber size, core separation, and optical fiber numerical aperture. Although the ellipsoidal shape is optimum, more practical to manufacture parabolic or spherical surfaces for surface 48 may be substituted.
Another aspect of the present invention comprises a novel design of the sensor electronic and opto-electronic circuity for a dynamic pressure sensor, as shown in
The new design uses one LED source and two photodiode detectors in contrast to previous single detector designs. A single detector design is adequate for a maximum electronics operating temperature of approximately 80° C., however, at a maximum operating temperature of 150° C. the dark current of a PIN photodiode becomes so large as to dominate the total photodiode output.
At temperatures up to 80° C., the dark current of a typical photodiode is on the order of 1% of the total signal with 99% of the signal being proportional to the light delivered to the detector by the collecting fiber. But at 150° C. the dark current increases more than three orders of magnitude and may represent as much as 500% of the photodiode signal that remains proportional to the light delivered to the detector. At such high levels of dark current the sensor calibration is significantly compromised.
To combat the dark current effect, the second photodiode is used for the purpose of providing a differentiating dark current input to a differential amplifier. In a preferred design disclosed below, the second diode is packaged inside a common enclosure with both photodiodes formed from the same silicon wafer, so their dark currents change identically with temperature changes. One photodiode is exposed to the light delivered by the collecting fiber while the second diode is covered so its output is only due to the dark current.
In
One of the most difficult challenges in developing accurate combustion pressure sensors is the need to overcome the effect of temperature changes on the sensor's diaphragm. Under changing temperature, the diaphragm's Young's modulus and Poisson's number change resulting in increased deflection at elevated temperatures. Two types of temperature errors result: increased deflection due to increased average diaphragm temperature over many cylinder pressure cycles and the intra-cycle diaphragm temperature excursions. Disclosed below are compensation techniques that correct for both of the temperature change errors.
The short term, intra-cycle, error due to what is sometimes called thermal shock effect is compensated by virtually instantaneously reducing diaphragm reflectivity as the diaphragm temperature changes. In
The thin film 72 material is appropriately selected for a decreasing reflection co-efficient as the diaphragm deflection rate increases with temperature increases. The light intensity received by the collecting fiber 44 can be described by
Vcol(p,T)=I·C·R(T)·D(p,T),
where I is light intensity emerging from the delivering fiber 42, C is the collection efficiency dependent on fiber parameters and fiber to diaphragm distance but not dependent on diaphragm characteristics, R(T) is the reflection coefficient of the diaphragm, and D(p,T) is the function describing diaphragm deflection as a function of pressure and temperature.
Through either thin film 72 light transmission or absorption dependence on temperature, the product of the reflection co-efficient R and deflection function D can be maintained independent of temperature if:
R(T)=D−1(T)
In practice, R(T) may not be exactly an inverse of D but at least should approximate the inverse function.
At least two types of thin films 72 are suitable for temperature compensation as described above. In one type, a dielectric thin film with dichroic mirror-like characteristics is used. In this type reflection decreases with increased temperature due to the thermal expansion of the thin film material. In the alternative type, a semiconductor material such as Silicon or Gallium Arsenide is used. In this type the absorption co-efficient increases with increasing temperature. In either type of thin film 72 the film is deposited over the area illumated by the delivering fiber 42, however, to minimize internal stresses in the film, the film area should be as small as possible as should the film thickness.
The temperature compensation technique described above does not correct for increased diaphragm deflection resulting from an increased average temperature of the diaphragm-sensor assembly over many pressure cycles. To correct for this type of error a compensation technique which relies on the electronic control of sensor gain is based on the output of a temperature probe mounted inside the sensor housing as shown schematically in
A schematic block diagram of the temperature-compensating opto-electronic circuitry for the sensor of
A number of innovative designs for a high accuracy static pressure sensor suitable for fuel pressure detection in a smart injector are disclosed below. The first design utilizes a two-fiber construction similar to what is shown in
The difference between the sensor electronics is in the LED current control. While in a dynamic pressure sensor version the current is continuously adjusted based on the differential output of a minimum-maximum detector, in the static pressure sensor version the current is adjusted only during discrete periods of time when the injector is filled with fuel and the sensor is only exposed to atmospheric pressure. Suitable electronic circuitry is shown in
The second design of a static sensor uses three or four optic fibers, two detectors and one LED. One pair of optic fibers is used to detect diaphragm deflection as above and the other pair serves as a compensation device to correct for potential errors due to fiber bending, coupling efficiency changes between the LED and fibers, the temperature dependence of LED intensity or photodiode sensitivity, as well as diaphragm deflection temperature dependence.
In a temperature non-compensated version of the second design as shown in
Alternatively, as shown in
In both
By appropriately locating the pairs 86 and 88 with the tips 90 and 92 at different distances from the diaphragm 40, the pairs are positioned to straddle the peak intensity 102. The maximum slope or sensor sensitivity occurs if the fiber tip 92 is positioned inside of the peak intensity 102 at about 250 microns. With the fiber pair tip 92 positioned on the first slope 104, the distance between the diaphragm 40 and the fiber pair tip decreases as pressure increases and the sensor output is caused to decrease.
If the second fiber pair tip 90 is positioned on the second slope 106 at approximately 600 microns the sensor output is caused to increase with pressure applied to the diaphragm 40. In what is sometimes called a “push-pull” effect, the sensor output is essentially a ratio of the two outputs from the pairs of fibers 86 and 88 resulting in increased sensor sensitivity to deflection compared to a sensor operated on one slope only with a single fiber optic pair.
In the processing scheme of
Alternatively, as shown in
By contrast, a relative pressure change applied to the sensor 46 will cause one output to increase and the other output to decrease having a net result of no change in the summing 114 of the outputs. The difference 118 of the two outputs passes through offset, filtering and gain 120 to provide the corrected sensor output. With the use of a sensor 46 also having a thermocouple therein, in addition to the two fiber pairs 86 and 88, correction 122 may be made for the average temperature of the sensor 46 and applied to the offset, filtering and gain 120 of the corrected sensor output.
In
In
The opto-electronic control circuitry of
Number | Date | Country | |
---|---|---|---|
Parent | 09367141 | Oct 1999 | US |
Child | 10668056 | Sep 2003 | US |