Appunti del corso di Powertrain Testing, Calibration and Homologation

TORQUE CHAIN STRUCTURE

Regarding the control logic of an internal combustion engine, the torque chain is managed in a

way that the torque delivered is determined by the combination of two levers:

• Slow Path

• Fast Path

Torque is influenced by the amount of air because the air intake determines the amount of fuel

that can be burned, which in turn dictates the possible power output. In spark-ignited (SI)

engines, we aim for a lambda value close to one to ensure optimal combustion. By controlling

the spark advance (SA), we modify combustion efficiency. Therefore, we prefer to use air intake

control to regulate torque rather than adjusting the spark advance.

The slow path manages the amount of air entering the cylinder, which ultimately provides the

engine's power. In the fast path, we control the spark advance. While we prefer using the slow

path to maximize efficiency, there are instances where we must use the fast path.

The torque coordinator's main task is to answer the question: what torque output do we want

to deliver to the engine? Once this is determined, the system decides whether to use the fast or

slow path, based on the criteria mentioned earlier.

The driver is not the only source of external torque demand. Additional torque is needed to

ensure proper operation of various systems, such as catalyst converter heating or engine start-

up. These functions require more torque than the driver’s direct input to meet their operational

needs.

Finally, it is essential to understand how to convert the requested torque into control signals.

71

A change in the spark advance angle has an immediate effect on the torque delivered (its value

may also vary between two subsequent cycles).

In contrast, a change in the configuration of actuators controlling the quantity of air (such as

the throttle, wastegate valve, or variators) results in a slower torque change. This delay occurs

due to phenomena related to the filling and emptying dynamics of the intake manifold and

cylinders, causing the effect to manifest after several cycles.

At the same time, there is an optimal spark advance angle for each operating condition, which

delivers the maximum torque. This optimal point is known from the IMEP-MFB50 curve.

Deviating from this optimal angle reduces combustion efficiency, leading to increased fuel

consumption.

However, it is sometimes necessary to vary the spark advance angle under certain conditions

requiring instant torque adjustments. These situations include idle engine speed control,

catalyst heating, driveability filters, traction control interventions, knock control, and torque

modifications during gear shifts. AIR PATH

If no specific priorities are imposed (such as catalyst heating or fast torque correction), the

preferred method of delivering the requested torque is through the air path, as it is the most

efficient approach.

The process begins with the gas pedal input, which is interpreted as a torque request at the

wheels. This request is then translated into commands for the engine's air management system,

such as adjusting the throttle, wastegate, or other actuators controlling air intake. By

optimizing air intake dynamics, the engine achieves the desired torque with minimal losses and

maximum efficiency. 72

The gas pedal position is interpreted as a request of torque at the wheel [N/m] through a map

that depends on engine speed; at low speeds, pushing the pedal, a significant amount of torque

is wanted at the wheels while at high speeds it’s necessary to press more the pedal to have more

torque. The torque at the wheel is then a target to be achieved through the net torque requested

to the ICE, so by considering the gear ratio and the transmission efficiency. Basically now it’s

obtained a torque target at the ICE so that a certain torque at the wheel can be obtained. At

this point the mechanical efficiency in the ICE has to be considered obtaining an indicated

torque requested to the ICE and in the end also a combustion efficiency (dependent also on SA

and λ) in order to calculate the amount of air needed per cylinder per cycle, therefore to calculate

the actuation in terms of throttle and wastegate position.

How to decouple the throttle and waste gate position targets? We have cases:

1. If :

≤

,

1. (

= 100% )

2. =

ℎ ℎ,

2. If :

>

,

1. =

,

2. (

= 100% )

ℎ

And, how correct the target values in closed loop? We need a PID:

1. If :

≤

,

with (open loop only);

= 100%

73

2. If :

>

,

with (open loop only);

= 100%

ℎ

A change in the spark advance angle has an immediate effect on the torque delivered, and its

value may also vary between two subsequent cycles.

In contrast, adjusting the configuration of actuators controlling the quantity of air (such as the

throttle, wastegate valve, or variators) results in a slower change in torque. This delay occurs

due to factors related to volumetric efficiency, such as the dynamics of cylinder filling and

emptying.

Naturally, there is an optimal spark advance angle for each operating condition, which delivers

the maximum torque. This optimal point is derived from the IMEP-MFB50 curve. Deviating

from this optimal angle reduces combustion efficiency, leading to increased fuel consumption.

Despite this, reducing the spark advance is necessary in certain situations where instant torque

variation is required. These scenarios include maintaining a torque reserve (e.g., during

minimum revving speed control or catalyst heating), driveability filters (to accommodate

different driving map options), traction control interventions, knock control, and torque

adjustments (increases or decreases) during gear shifts.

74

IDLE SPEED CONTROL

It can be necessary to reduce the SA for example for knock limitations (only at high loads) or

for other reasons like engine warm up, cylinder or bank individual correction, calibratable offset

The data from a spark advance sweep test show the results in the figure. On the x-axis this

time there’s not the SA but the CA50 MFB, that has an optimal combustion phase for more or

less 8°CA50MFB, generally the optimal is for a CA50MFB between 8 ÷ 12◦ . Retarding the SA

and therefore the CA50MFB, it can be seen that the stability is reduced (retarded combustions

are less stable); indeed for the blue points that have all the same SA there is great variability

in torque production (IMEP).

The idle speed controller must be able to manage torque to maintain a stable speed condition.

Therefore, idle speed control needs the ability to generate sufficient torque to keep the engine

operating stably. When the engine approaches the instability threshold, it is essential to have

a "ready-to-use" positive torque to move away from that zone.

Idle speed control is not easy, indeed the ECU should calculate the amount of air to introduce

in the engine to achieve an indicated torque equal to the resistant torque inside the engine

because the driver is asking for 0 torque; idle means to keep constant engine speed as low as

possible while guaranteeing stable engine operation.

Having ω = const doesn’t mean that its value is constant, but that the speed is constant along

the cycle. At some point the driver can increase the torque due to friction and auxiliaries for

example by turning on the air conditioning, this brakes the equilibrium between the indicating

torque and the torque lost through auxiliaries and friction. If no action is taken, the engine

speed drops in a dead zone where the engine turns off.

75

In order to avoid this condition, it’s necessary to increase the indicating torque in a very fast

way but if the SA at idle is at its SAMBT then it’s not possible to react fast.

For this reason, at idle SAMBT is not applied but it’s chosen to be less efficient by lowering ∆SA

to guarantee a certain amount of torque reserve. The choice of how much retard the SA at idle

is obviously a compromise between the needs in terms of stability and the needs in terms of

CO2 emissions that a less efficient combustion generates.

If another load is attached to the crankshaft and the torque reserved was already used then the

engine would turn off; therefore it’s always necessary to restore the torque reserve, for example

by opening more the throttle position.

Since the response through the throttle ∆α is quite slow and the response through the spark

advance ∆SA is quite fast, it’s possible to model the control through two PID controller in which

for ∆α the integral component is the main part while for ∆SA the proportional component is the

most important. Obviously ∆SA can reduce the error but cannot set it to 0, while ∆α is slower

but can annul the error. 76

In-Cylinder Pressure Measurement

OVERVIEW

The information about the in-cylinder pressure is very important and would be good to have it

also on-board but up to now it’s not possible.

• The main reasons to measure this signal is to have a quantitative measurement of the

torque that the engine can potentially produce; the first important measurement that

can be extracted is then the indicated mean effective pressure IMEP (and indicated

torque): integral of the pressure over the volume, that is the work produced by the engine.

Obviously this is not the total work available because part of it is wasted to win friction

and drag auxiliaries.

• By measuring the indicated torque and the torque at the brake, it is possible to

calculate the mechanical efficiency.

• Starting from the pressure, it’s possible to analyse the heat release.

• Determination of combustion phasing and duration.

• Evaluation of combustion variability (pressure peak and its position, max pressure

gradient position, CoV IMEP).

• Knock intensity measurement (and control).

• Fault diagnosis (misfire, leaks, anomalous friction, . . . ).

• Pressure peak measurement

MEASUREMENT SYSTEMS

Initially, cylinder pressure measurement was performed using Watt's indicator, a mechanical

device consisting of a small piston connected to a spring. The piston’s movement, caused by

pressure changes in the chamber, was recorded by a pen on a rotating drum. However, this

system had significant limitations, including inertia, friction, inability to accur