Powertrain & Calibration 101
Powertrain & Calibration 101
Powertrain & Calibration 101
Calibration 101
John Bucknell
DaimlerChrysler
Powertrain Systems Engineering
December 4, 2006
Powertrain & Calibration Topics
Background
Powertrain terms
Thermodynamics
Mechanical
Design
Combustion
Architecture
Cylinder Filling &
Emptying
Aerodynamics
Calibration
Spark & Fuel
Transients &
Drivability
What is a Powertrain?
Engine that converts thermal energy to
mechanical work
Particularly, the architecture comprising all the
subsystems required to convert this energy to
work
Sometimes extends to drivetrain, which
connects powertrain to end-user of power
Characteristics of Internal
Combustion Heat Engines
High energy density of fuel leads to high power
to weight ratio, especially when combusting with
atmospheric oxygen
External combustion has losses due to multiple
inefficiencies (primarily heat loss from
condensing of working fluid), internal
combustion has less inefficiencies
Heat engines use working fluids which is the
simplest of all energy conversion methods
Reciprocating Internal
Combustion Heat Engines
Characteristics
Slider-crank mechanism has high mechanical
efficiency (piston skirt rubbing is source of 50-
60% of all firing friction)
Piston-cylinder mechanism has high single-
stage compression ratio capability – leads to
high thermal efficiency capability
Fair to poor air pump, limiting power potential
without additional mechanisms
Reciprocating Engine Terms
Vc = Clearance Volume
Vd = Displacement or Swept Volume
Vt = Total Volume
TC or TDC =
Top or Top Dead Center Position
BC or BDC =
Bottom or Bottom Dead Center
Position
Compression Ratio (CR)
Vd Vc
CR
Vc
Further explanation of aspects of Compression Ratio
Reciprocating
Engines
Most layouts created
during second World
War as aircraft
manufacturers
struggled to make the
least-compromised
installation
Thermodynamics
Otto Cycle
Diesel Cycle
Throttled Cycle
Supercharged Cycle
Work = P dV
Power = Work/Unit Time
Work Cycles / Second
Power
Re volutions / Cycle
TDC BDC
TDC BDC
150% additional
1200 -90
1150 -105
distance for cold-start 1100 -120
can be enhanced by IMEP 4-2-1 1044.1 1122.8 1188.5 1226.6 1269.2 1290.5 1337.9 1390.1 1445.7 1427 1445.8 1435.4 1411.7 1337.9
1102.5 1162.2 1225.5 1252.3 1248 1262.4 1320.9 1403.6 1403.5 1406.3 1398 1367.2 1294.6
3-8% across 60% of
IMEP 4-1 Cast
PMEP 4-2-1 -5.3 -9.7 -14.2 -19.7 -23.0 -29.9 -38.4 -52.3 -64.0 -78.5 -90.8 -107.9 -122.8 -136.2
-12.5 -16.8 -20.8 -26.1 -32.0 -40.3 -54.0 -68.6 -81.0 -89.0 -99.8 -111.5 -119.5
the operating range.
PMEP 4-1 Cast
Engine Speed (rpm)
Momentum Effects
Pressure loss influences dictate that duct
diameter be as large as possible for minimum
friction
Increasing charge momentum enhances cylinder
filling by extending induction process past
unsteady direct energy transfer of induction
stroke (ie piston motion)
Decreasing duct diameter increases available
kinetic energy for a given mass flux
Therefore duct diameter is a trade-off between
velocity and wall friction of passing charge
Pressure Wave Effects
Induction process and exhaust blowdown
both cause pressure pulsations
Abrupt changes of increased cross-section
in the path of a pressure wave will reflect
a wave of opposite magnitude back down
the path of the wave
Closed-ended ducts reflect pressure waves
directly, therefore a wave will echo with
same amplitude
Pressure Wave Effects con’t
Friction decreases energy of pressure
waves, therefore the 1st order reflection is
the strongest – but up to 5th order have
been utilized to good effect in high speed
engines (thus active runners in F1 in Y2K)
Plenums also resonate and through
superposition increase the amplitude of
pressure waves in runners – small impact
relative to runner geometry
Effects of Intake Runner Geometry
Tuning in Production I4 Engine
470
450
Air Mass per Cylinder (mg)
430
410
390
370
350
1200 1600 2000 2400 2800 3200 3600 4000 4400 4800 5200 5600 6000 6400
Trapped Mass 372 381 373 421 428 402 397 430 454 453 458 460 431 401
Engine Speed (rpm)
Aerodynamics
Losses due to poor aerodynamics can be
equal in magnitude to the gains from
pressure wave tuning
Often the dominant factory in poorly
performing OE components
If properly designed, flow of a single-entry
intake manifold can approach 98% of an
ideal entrance on a cylinder head port
(steady state on a flow bench)
Aerodynamics con’t
Flow Separation
Literally same phenomenon as stall in wing
elements – pressure in free stream insufficient
to ‘push’ flow along wall of short side radius
Recirculation pushes flow away from wall,
thereby reducing effective cross-section: so-
called “vena contracta”
Simple guidelines can prevent flow separation
in ducts – studies performed by NACA in the
1930s empirically established the best duct
configurations
Aerodynamics con’t
Wall Friction
Surface finish of ducts need to be as smooth
as possible to prevent ‘tripping’ of flow on a
macro level
Junctions & Bends
Everything from your fluid dynamics textbook
applies
Radiused inlets and free-standing pipe outlets
Minimize number of bends
151
150
Corrected Power (c Bhp)
149
148
147
146
145
0 2 4 6 8 10 12 14 16
Back Pressure (in-Hg)
560
550
540
530
520
510
Torque (ft-lbs)
500
490
480
470
460
450
440
430
420
1600 2000 2400 2800 3200 3600 4000 4400 4800 5200 5600
Engine Speed (rpm)
1200
1100
250 110
120
100
1000
275
90
900 80
70 300
60
800
240
50
700 40
BMEP SI [kPa]
30
600
250
500
20
400
275
300
300
10 350
200
400
450
500
100 600
700
0
1200 1600 2000 2400 2800 3200 3600 4000 4400 4800 5200 5600 6000 6400
d Speed [rpm]
Summary
Component’s Relative Impact on
Performance
1. Cylinder Head Ports & Valve Area
2. Valve Events
3. Intake Manifold Runner Geometry
4. Compression Ratio
5. Exhaust Header Geometry
6. Exhaust Restriction
7. Air Cleaner Restriction
Powertrain Closing Remarks
Powertrain is compromise
Four-stroke engines are volumetric flow rate devices
– the only route to more power is increased engine
speed, more valve area or increased charge density
More speed, charge density or valve area are
expensive or difficult to develop – therefore
minimizing losses is the most efficient path within
existing engine architectures
Highest average power during a vehicle acceleration
is fastest – peak power values don’t win races
Break
Calibration
What is it?
Optimizing the control system (once hardware is
finalized) for drivability, durability & emissions
It’s just spark and fuel – how hard could it be?
Knowledge of Thermodynamics, Combustion and
Control Theory all play in
Fortunately race engines have no emissions
constraints and use race fuel (usually eliminates any
knock) – therefore are relatively easy to calibrate
Calibration Terms
Stoichiometry – Chemically correct ratio of fuel to air
for combustion
F/A – Fuel/Air Ratio
Mass ratio of mixture, a determination of richness or
leanness. Stoichiometry = 0.0688-0.0696 FA
Lambda – Excess Air Ratio
Stoichiometry = 1.0 Lambda
Rich F/A – F/A greater than Stoichiometry
Rich < 1.0 Lambda
Lean F/A – F/A less than Stoichiometry
Lean > 1.0 Lambda
Calibration Terms continued
Brake Power – Power measured by the absorber
(brake) at the crankshaft
BSFC - Brake Specific Fuel Consumption
Fuel Mass Flow Rate / Brake Power
grams/kW-h or lbs/hp-h
LBT Fuelling – Lean Best Torque
Leanest Fuel/Air to Achieve Best Torque
LBT = 0.0780-0.0800 FA or 0.85-0.9 Lambda
Thermal Enrichment – Fuel added for cooling
due to exhaust component temperature limit
Injector Pulse Width - Time Injector is Open
Calibration Terms continued
Spark Advance – Timing in crank degrees prior to TDC
for start of combustion event (ignition)
100%
98%
96%
94%
Torque Delta Factor From LBT
92%
90%
88%
Spark Held Constant During Fuel Air Sw eep
86%
84%
82%
80%
78%
76%
0.0660 0.0690 0.0720 0.0750 0.0780 0.0810 0.0840 0.0870 0.0900 0.0930 0.0960 0.0990 0.1020 0.1050 0.1080 0.1110
F/A FN
1856 RPM, 70 kPa MAP 3296 RPM, 98 kPa MAP 3296 RPM, 56 kPa MAP 3296 RPM, 84 kPa MAP
4544 RPM, 70 kPa MAP 3296 RPM, 98 kPa MAP 2688 RPM, 70 kPa MAP
Spark Advance vs Torque
102%
100%
98%
96%
Torque Delta from MBT
94%
92%
90%
88%
86%
84%
-22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
Delta Spark Advance From MBT
Control System Types
Alpha-N
Engine Speed & Throttle Angle
Speed-Density
Engine Speed and MAP/ACT
MAF
Engine Speed and MAF
Alpha-N
Fuel and spark maps are based on
throttle angle – which is very non-linear
and requires complete mapping of
engine
Good throttle response once dialed in
Density compensation (altitude and
temperature) is usually absent – needs to
be recalibrated every time car goes out
Speed-Density
Fuel and spark maps are based on MAP –
density of charge is a strong function of
pressure, corrected by air temp and coolant
temp therefore air flow is simple to calculate
Less time-intensive than Alpha-N, once calibrated
is good – most common type of control
Needs less mapping – can do WOT line and mid-
map then curve-fit air flow (spark needs a little
more in-depth for optimal control)
MAF
Fuel and spark maps are based on MAF –
airflow measured directly
MAF sensor isn’t the most robust device
Pressure pulses confuse signal, each application has to
be mapped with secondary damped MAF sensor (usually
a 55 gallon drum inline)
Least noisy signal is usually at air cleaner – so separate
transport delay controls need to be calibrated for
transients and leaks need to be absolutely eliminated
Boosted applications usually add a MAP as well
Control System Components
Fuel System
Injectors, Fuel pump & Regulator
Basic Sensors
Manifold Absolute Pressure (MAP) or Mass Air
Flow (MAF)
Crank Position (Rpm & TDC)
Cam Position (Sync)
Air Charge Temp (ACT)
Engine Coolant Temp (ECT)
Knock Sensor
Lamda Sensor
Fuel System
Injectors
Volumetric flow rate solenoids, linear relationship
between pulsewidth and flow for given pressure delta Pulsewidth + Battery Offset
Battery offset is time necessary to open and close Pulsewidth
solenoid – time is fixed for any voltage
Duty cycle is injector on time – it’ll go static above 95%
Pintle Height
Bernoulli relationship for different pressure deltas –
allowing differing flow rates for a given injector
High impedance injectors have lower dynamic range
and lower amperage and thus less heat in controller
Fuel Pump & Regulator
Pressure needs to be sufficiently high to prevent vapour
lock (>4bar) and low enough that engine can idle
V P
In-tank regulation adds least heat but has line-loss as
flow rate increases, ie fuel pressure changes with flow
2
2
Manifold-referenced regulation can help injectors
achieve higher flow rates at elevated boost or lower V P
1
flows at low vacuum – making calibration more 1
complicated
Bernoulli Effect of Fuel Pressure
Sensors
Manifold Absolute Pressure (MAP)
A variable-resistance diaphragm with perfect vacuum on one side
and manifold pressure on other
Mass Air Flow (MAF)
A heating element followed by a temperature-sensitive element.
Heated element is maintained at a constant temperature and
based upon the measured downstream temperature the mass flow
rate can be determined
Crank Position
High resolution for spark advance, less-so for crank speed and
with once-per-rev can indicate TDC
Cam Position
Low resolution for syncronization for sequential fuel injection and
individual cylinder spark
Air Charge Temp and Engine Coolant Temp
Thermistors used for air density correction and startup enrichment
Sensors, cont
Knock Sensor
A piezoelectric load cell that measures structural vibration.
Knock is a pressure wave that travels at local sonic velocity and
‘rings’ at a frequency that is a function of bore diameter
(typically between 14-18kHz). When the structure of the
engine (typically the block) is hit with this pressure wave it rings
as well, but at a frequency that is a function of the structure (ie
materials and geometry). A FFT analysis of different mounting
positions (nodes not anti-nodes) is necessary to determine the
‘center frequency’ to listen for knock (which is measured via in-
cylinder pressure measurements) without picking up other
structure-borne noise.
Sensors, cont
Lamda Sensor (EGO)
Compares ambient air to
exhaust oxygen content
(partial pressure of oxygen).
Sensor output is essentially
binary (only indicates rich or
lean of stoichiometry).
Wide-band Lamda Sensor EGO Schematic
(UEGO)
Compares partial pressure of
oxygen (lean) and partial
pressure of HmCn, H2 & CO
(rich) with ambient. Gives
output from ~0.6 to 2 Lamda.
UEGO Schematic
Calibration Goals
Combustion & Thermodynamics
Work, Power & Mean Effective Pressures
Knock, Pre-Ignition
Burn Rate
Transients
Wall film
Thermal Enrichment
Drivability
Knock
Causes of Knock
Knock = f(Time,Temperature,Pressure,Octane)
Time – Higher engine speeds or faster burn rates reduce knock
tendency. Burn rate can come from multiple spark sources,
more compact combustion chambers or increased turbulence
Temperature – Reduced combustion temperatures reduce knock
through reduced charge temperatures (cooler incoming charge
or reduced residual burned gases), increased evaporative
cooling from richer F/A mixtures and increased combustion
chamber cooling
Pressure – Lower cylinder pressures reduce knock tendency
through lower compression ratio or MAP pressure
Octane – Different fuel types have higher or lower autoignition
tendencies. Octane value is directly related to knocking
tendency
Knock continued
Effects of Knock
Disrupts stagnant gases that form boundary layer at
edge of combustion chamber, increasing heat transfer
to components and raising mean combustion
chamber temp that can lead to pre-ignition
Scours oil film off cylinder wall, leading to dry friction
and increased wear of piston rings
Shockwave can induce vibratory loads into piston pin,
piston pin bore and top land - reducing oil film
thickness and accelerating wear
Shockwave can be strong enough to stress
components to failure
In-cylinder Pressure Measurement
Piezoelectric pressure
transducers develop
charge with changes
in pressure
Installed in
combustion chamber
wall or spark plug to
measure full-cycle
pressures
Typical pressure probe installation
Targets Valve