Page 89 - Automotive Engineering

CH AP TER 4 .1 Digital engine control systems

stroke. That quantity of fuel is given by the air charge Corrections of the base pulse width occur whenever
divided by the desired air/fuel ratio: anything affects the accuracy of the fuel delivery. For
A example, low battery voltage might affect the pressure in
F ¼ the fuel rail that delivers fuel to the fuel injectors. Cor-
ðA=FÞ d
rections to the base pulse width are then made using the
The quantity of air drawn into the cylinder, A, is actual battery voltage.
computed from the MAF rate and the RPM. The An alternate method of computing MAF rate is the
MAF rate will be given in kg/sec. If the engine speed is speed–density method. Although this method has es-
RPM, then the number of revolutions/second (which we sentially been replaced by direct MAF measurements,
call r) is: there will continue to be a number of cars employing
this method for years to come, so it is arguably worth-
RPM
r ¼ while to include a brief discussion in this chapter. This
60 method, which is illustrated in Fig. 4.1-4, is based on
measurements of MAP, RPM, and intake air tempera-
Then, the MAF is distributed approximately uni-
formly to half the cylinders during each revolution. If the ture T i . The air density d a is computed from MAP and
number of cylinders is N then the air charge (mass) in T i , and the volume flow rate R v of combined air and
each cylinder during one revolution is: EGR is computed from RPM and volumetric efficiency,
the latter being a function of MAP and RPM. The
MAF volume rate for air is found by subtracting the EGR
A ¼
rðN=2Þ volume flow rate from the combined air and EGR.
Finally, the MAF rate is computed as the product of the
In this case, the mass of fuel delivered to each volume flow rate for air and the intake air density. Given
cylinder is: the complexity of the speed–density method it is easy to
see why automobile manufacturers would choose the
MAF
F ¼ direct MAF measurement once a cost-effective MAF
rðN=2ÞðA=FÞ d sensor became available.
This computation is carried out by the controller The speed–density method can be implemented either
continuously so that the fuel quantity can be varied by computation in the engine control computer or via
quickly to accommodate rapid changes in engine oper- lookup tables. Fig. 4.1-5 is an illustration of the lookup
ating condition. The fuel injector pulse duration T table implementation. In this figure, three variables need
corresponding to this fuel quantity is computed using the to be determined: volumetric efficiency (n v ), intake
known fuel injector delivery rate R f : density (d a ), and EGR volume flow rate (R E ). The volu-
metric efficiency is read from ROM with an address de-
F
T ¼ termined from RPM and MAP measurements. The intake
R f air density is read from another section of ROM with an
This pulse width is known as the base pulse width. The address determined from MAP and T i measurements.
actual pulse width used is modified from this according The EGR volume flow rate is read from still another
to the mode of operation at any time, as will presently be section of ROM with an address determined from DP and
explained. EGR valve position. These variables are combined to yield
the MAF rate:
4.1.4.3 Open-loop control
RPM D
MAF ¼ d a n v R E
For a warmed-up engine, the controller will operate in an 60 2
open loop if the closed-loop mode is not available for any
where D is the engine displacement.
reason. For example, the engine may be warmed suffi-
ciently but the EGO sensor may not provide a usable
signal. In any event, it is important to have a stoichio-
metric mixture to minimize exhaust emissions as soon as 4.1.4.4 Closed-loop control
possible. The base pulse width T b is computed as
described above, except that the desired air/fuel ratio Perhaps the most important adjustment to the fuel in-
(A/F ) d is 14.7 (stoichiometry): jector pulse duration comes when the control is in the
closed-loop mode. In the open-loop mode the accuracy
MAF of the fuel delivery is dependent on the accuracy of the
T ¼ base pulse width
b
rðN=2Þð14:7ÞR f measurements of the important variables. However, any

82