Low Pressure

Indicating
Optimization of gas
exchange by a suitable
combination of pressure
indicating, analysis and
simulation

Dipl.-Ing. C. Burkhardt
EnginOS GmbH,
D-73760 Ostfildern

Dipl.-Ing. M. Gnielka,
Dr. sc. techn. C. Gossweiler,
Dr. techn. D. Karst,
Dipl.-Ing. M. Schnepf,
Dr.-Ing. J. von Berg,
Dipl.-Ing. P. Wolfer
Kistler Instrumente AG

Special Print
920-246e-01.04
 -1-
Institute for Internal Combustion Engines and Thermodynamics, Graz University of Technology
Institut für Verbrennungskraftmaschinen und Thermodynamik, Technische Universität Graz
9th Symposium "The Working Process of the Internal Combustion Engine" September 2003
9. Tagung "Der Arbeitsprozess des Verbrennungsmotors" September 2003

Optimization of gas exchange by a suitable combination

of pressure indicating , analysis and simulation
Ladungswechseloptimierung durch die geeignete Kombination
von Indiziermesstechnik, Analyse und Simulation

Dipl.-Ing. C. Burkhardt
EnginOS GmbH, D-73760 Ostfildern

Dipl.-Ing. M. Gnielka, Dr. sc. techn. C. Gossweiler, Dr. techn. D. Karst,

Dipl.-Ing. M. Schnepf, Dr.-Ing. J. von Berg, Dipl.-Ing. P. Wolfer

Kistler Instrumente AG, CH-8408 Winterthur

Abstract
The design of an efficient gas exchange for internal combustion engines requires close coordination
between pressure measurement, heat release rate and gas exchange analysis, and gas exchange
simulation. Moreover, to meet the needs of low pressure measurement, particular attention must be
paid to the type and positioning of the pressure transducers used.

Through a suitable combination of pressure indicating, analysis and simulation the time needed to
develop an optimized gas exchange can be significantly reduced. To this end, a complete pressure
analysis - i.e. a full analysis of the combustion and the gas exchange processes - is required for
calculation of the residual gas masses and the rate of heat release.

The simulation inputs deriving from the analysis of pressure measurements, combustion and gas
exchange serve not only to adjust the geometry model but also to calibrate and check the heat release
model. This is an important prerequisite to permit the definition of a suitable camshaft geometry for
example, through the variation of simulation parameters.

Solutions to the overall problem of low pressure measurement and the combination of analysis and
zero-dimension gas exchange simulation will be demonstrated by an example of a four-stroke single-
cylinder gasoline engine. The study assesses the suitability of various types of low pressure
transducers and the effect of their position on low pressure measurements. Thanks to the development
of miniaturized transducers for absolute and dynamic measurement, low pressure measurements can
now be taken close to the valve. Investigations have shown that these transducers are capable of
performing highly precise measurements without the need for heat-protection by a switching adapter,
an advantage which considerably simplifies their mounting and operation.
 -2-

1 INTRODUCTION
One method of analyzing the working process of developed engines is by means of energy balances
according to the First Law of Thermodynamics, and this analysis is used for a basic assessment of the
engine design. In this regard the initial conditions for the engine cycle calculation are of fundamental
importance, that is to say, the mass as well as condition and composition of the working gas at the
intake. These are normally determined by a gas exchange calculation. Alternatively, quantification by
loss analysis of theoretically avoidable individual losses in the operating cycle may offer the
possibility for improving different parts of the cycle. In addition to analysis of the losses in the high
pressure part of the operating cycle (e.g. losses due to incomplete combustion, heat transfer or
friction), also the losses from the actual gas exchange are analyzed (PISCHINGER, R., KLELL,
SAMS 2002).
The gas exchange calculation is normally carried out using the flow equation and the intake and
exhaust pressure curves, which are known either from a simulation of the intake and exhaust system or
from a low pressure measurement. Low pressure measurements require extremely high demands on
the low-pressure sensors. For example, as there are gas temperatures of more than 700 °C in the
exhaust system, the sensors are screwed into a cooling adapter at a considerable distance from the
exhaust valve. Far better is to screw the sensor into a cooled switching adapter in which the indicating
bore opens only during the short measuring period.
A new generation of water-cooled low-pressure sensors is now available, which dispenses with the
need for expensive thermal protection by means of a switching adapter, thus facilitating installation
and operation. Even measurements close to the valve have now become possible through the
development of miniaturized pressure transducers for static measurement.
This study assesses the suitability of low-pressure sensors for precision measurements and the
influence of the pressure transducer position on the measuring results from a single-cylinder spark-
ignition engine up to a maximum speed of 5 000 rpm. The test bed was made available by Prof. Dr.
Michael Bargende of the Stuttgart Research Institute for Automotive Engineering and Motor Vehicles
(FKFS). The gas exchange calculations were made using the Tiger software produced by EnginOS
GmbH, Ostfildern.

2 MOTIVATION

2.1 Gas exchange optimization

To reduce the toxic constituents NOx and HC of the exhaust gas, engine manufacturers rely among
other things on optimization of the gas exchange. An effective measure, particularly with partial load,
is internal exhaust-gas recirculation controlled by the gas exchange, whereby the increasing residual
exhaust gas content in the cylinder reduces the combustion temperatures and thus the NOx emissions.
In addition, the higher average HC concentration of the residual exhaust gas compared with the
exhaust gas reduces the HC emissions (PISCHINGER, F. 1988).

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 -3-

The exhaust gas content of the cylinder charge influences not only the emissions, but also the thermo-
dynamic efficiency of the complete engine in constant-volume combustion. The reason for this is the
higher specific thermal capacity of the residual exhaust gas during compression, which reduces the
isentropic exponents of the charge and thus the compression work. In the expansion phase, with an
unchanged air-fuel ratio, the reduced fuel mass results in lower process temperatures and a
corresponding rise in the isentropic exponents. The two effects are cumulative and lead to an increase
in efficiency, except where there is a shortage of air in the case of high residual exhaust gas contents.
For example, the variable valve control in spark-ignition engines is an efficient way of controlling the
internal exhaust-gas recirculation. With high exhaust-gas recirculation rates, an increase in the internal
efficiency of up to about 10 % is possible (PISCHINGER R., KLELL 2001).
Optimization of the gas exchange, particularly with variable valve timing gear, is possible by using a
suitable combination of pressure indicating, analysis and simulation. Simulation requires accurate
determination of the residual exhaust gas content at the moment when the intake valve closes as an
important initial condition for the working process calculation. This can be achieved by the combined
gas exchange and combustion process calculation, called in the following text complete analysis. The
measurement must provide accurate indications in respect of the low pressure, particularly of the
pressure level.

2.2 Methods for calculating the residual gas content

Where the charge-cycle port present a complex geometry or where gas-dynamic effects in the ports
constitute a major influence (e.g. at high speeds), the three-dimensional computational fluid dynamics
(CFD) calculation of gas exchange and scavenging becomes increasingly important. However, this
very expensive and accurate method of calculation cannot be used as an optimizing tool on the test bed
because, given the many measuring points normally used for calculating the residual exhaust gas
content, the expense of modeling and calculation would be unacceptably high.
Very often, however, a one-dimensional gas exchange calculation will satisfy the requirements for
values which are as accurate as possible for volumetric efficiency, performance and general efficiency,
even when gas-dynamic oscillations in the pipelines become stronger with increasing speed.
When determining the residual exhaust gas content, data quality depends primarily on the simulation
model used. Simple approaches treat the residual exhaust gas as a mass which returns unmixed to the
combustion chamber. A significant improvement is achieved under the assumption of an ideal mixture
(perfect mixing) of exhaust gas and fresh gas (PISCHINGER, R., KLELL, SAMS 2002).
The three-dimensional and one-dimensional flow simulations are used to calculate the low pressure
curves in the intake and exhaust systems.

2.3 Gas exchange calculation based on the low pressure processes measured
The EnginOS Tiger software uses the low pressure curves measured in the intake and exhaust systems
to calculate the gas exchange. The sequentially actuated mass flows, the residual exhaust gas content
and the pressure in the relevant cylinder are calculated on the basis of the flow equation.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 -4-
In a combined heat release rate and gas exchange analysis, referred to below as complete analysis,
high pressure and gas exchange of the operating cycle are calculated in a single step. Only by this
means it is possible to determine the exact gas composition at the moment when the intake valve
closes and thus to provide the initial conditions for a subsequent calculation of the heat release rate.
The complete analysis allows a comprehensive interpretation of both combustion (internal energy
balance) and gas exchange.
The following input quantities are required for calculating the gas exchange:
Valve lift as a function of the crank angle of the valves involved in the gas exchange. These data
can either be derived from the theoretical design calculations or entered in the form of the valve lift
measured on the test bed.
Valve timing
It is important to indicate the valve timing, since a significant mass already flows through the
valves (particularly in the exhaust system) shortly after the valves open, as a result of the high
pressure gradients.
Flow coefficient
These values are obtained within the scope of a stationary flow test on a special test bed with
various valve strokes. The flow coefficient describes the ratio of the actual mass flow to the
theoretical mass flow (isentropic flow).
Initial cylinder pressure value measured at the moment the exhaust valve opens
Low pressure measured as a function of the crank angle in the intake and exhaust system
The calculation of the gas exchange is based on the outflow equation after Saint Venant and Wantzel
for an isentropic change of state of a compressible medium (FROMMELT, BARGENDE 1997),
(BURKHARDT 2001).
During the valve overlap, in which the gas composition and thus the caloric data continuously change,
there are frequent backflows of burnt gases into the cylinder or the intake system. This is taken into
account in the software by application of the model "Complete mixing of fresh gas and exhaust gas"
(Perfect Mixing).
The iteration for calculating the fresh-air mass has a decisive effect on the gas exchange results.
Iteration takes place here on the basis of adaptation of the aspirated-air temperature at a specified air-
fuel ratio, and not on adaptation of the fresh-air mass at a specified intake temperature. The iteration
loop is ended when the fresh-air mass calculated deviates by less than 1 % from the fresh-air mass of
the previous iteration step, and the mass in the cylinder at the moment the exhaust valve opens differs
by less than 1 % from the cylinder mass at the moment the intake valve closes. The results calculated
include residual exhaust gas mass and intake temperature.
An FVV research project just completed (KÖHLER, BARGENDE 2003) includes evidence that this
approach using the gas exchange analysis up to speeds of 6 000 rpm produces very close conformity
with a one-dimensional gas exchange calculation (GT-Power simulation software).

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 -5-

2.4 Requirements for the use of gas exchange analysis

As deviations due to erroneous valve lifting curves, flow coefficients and valve timing influence the
quality of the result of the gas exchange analysis, they must be carefully avoided. The necessary flow
coefficients are usually determined on a static air-injection test bed, and therefore do not necessarily
correspond to actual engine operation. Deviations may therefore occur, particularly in the range of
smaller valve lifts.
The theoretical valve lifting curves, particularly over a broad speed band may also vary as a result of valve
timing gear elasticity. Frequent consequences of these variations are a shift in the valve timing, a change in
the valve spacing and / or valve lift. This influence can be corrected with the software used and the actual
valve lifting curves then determined (BURKHARDT 2001).
A very important and sensitive
plausibility check is to compare the
cylinder pressure calculated in a gas
exchange analysis with the cylinder
pressure measured. Incorrect input data
can thus be easily identified, provided
that several error influences are not
superimposed. As shown by a
comparison of the cylinder pressure
curves from the test series (Fig. 1), the
calculated curve coincides closely with
the measured curve, which is an
indication of correct input data, parti-
cularly of the preset valve lifting curves.
The calculation results presented for the
performed test series (Section 4) meet
Fig. 1: Comparison of cylinder pressures (measured - calculated)
this criterion.

The question also arises as to which requirements are placed on low pressure indication for the
application on gas exchange analysis. In this regard, the accuracy of the absolute pressure level
measured must be examined, as well as any phase shift and change in amplitude.
The installation must allow for the following, partly conflicting requirements:

• Protection of the sensor against high temperatures in order to measure the pressure level as
accurately as possible.

• Installation of the sensor as close to the valve as possible in order to avoid phase and amplitude
errors.
These questions are investigated in the following sections.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 -6-

3 REQUIREMENTS FOR LOW PRESSURE INDICATION

3.1 Specification of suitable pressure transducers

An accurate gas exchange analysis presupposes high precision pressure measurement in both the
intake and exhaust systems. Kistler oil-filled piezoresistive sensors (Types 4045 and 4075) set the
benchmark in this regard. They guarantee high precision during engine operation as a result of their
installation in cooling or cooled switching adapters.
The disadvantage of these systems is their size, which makes measurement close to a valve more
difficult. This can have a negative influence on the measuring result, particularly in highly dynamic
processes, such as in Formula One applications. For this reason, customers are showing increasing
interest in smaller pressure sensors which can be mounted directly in the intake and exhaust systems
without an adapter. There is a clear trend towards miniaturized, uncooled, low-pressure sensors, with
the possibility of installation in the cylinder head.
A further important requirement concerns the stability of the pressure indication. Here, the shift in the
indicated pressure level during the measurement (caused, for example, by thermal loading of the
pressure transducer) must be as low as possible. Investigations into the possibilities for accurate
measurement of gas exchange pressure curves show that calculation of the gas mass is still sufficiently
accurate, when the deviation in pressure level is not more than 10 mbar in the intake and exhaust
systems (WIMMER, BERAN, FIGER 2000). When a 2 bar transducer is fitted into naturally aspirated
engines, this represents a measuring accuracy of 0,5 % FSO (Full Scale Output), or 0,2 % FSO in the
case of a 5 bar transducer used in supercharged engines.
In order to minimize the thermal zero offset between normal conditions and engine operation
condition and to increase the measuring accuracy of the pressure level, the sensor is normally
calibrated before measurement in the warm but not running engine, in which it is set to ambient
pressure. For even greater accuracy, particularly in the exhaust system with its high temperatures,
calibration must be carried out in the running engine using a combined cooling and switching adapter,
in which the adapter is briefly switched over automatically to atmospheric pressure before and after
the measurement. Any drift can then be mathematically corrected in the indicating system
(FROMMELT, BARGENDE 1997).
Particularly in the case of absolute pressure sensors, the thermal sensitivity shift can result in major
errors of pressure level and amplitude during measurement. This problem can be overcome by
transducers which are measuring relative pressure. The thermal sensitivity change can be very much
reduced through the use of a cooling adapter, which maintains an almost constant operating
temperature. Other sensor characteristics, such as linearity error and sensor hysteresis, have a minor
influence on the results, but the total error should be less than 1 % FSO.
Further requirements of the temperature-compensated and calibrated measuring chain are a good
frequency response i.e. uncorrupted signal with regard to amplitude and phase, together with low
acceleration sensitivity, even in highly dynamic engine operation. Thus, in addition to a sensor with an
appropriately high natural frequency, a suitable amplifier is also required with a constant frequency
response of up to at least 10 kHz.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 -7-

Table 1 contains the minimum Technical data Value Unit

standard specifications to be met by Stability of the zero point <±0.5 %FSO
the pressure transducers used in a Thermal zero offset <±0.5 %FSO
gas exchange analysis. The data for
Thermal sensitivity change <±1 %
the thermal zero offset and thermal
sensitivity change refer to the Linearity error <±0.3 %FSO
operating temperature range of 20 °C Natural frequency >20 kHz
to 120 °C. The use of a cooling Acceleration sensitivity <0.3 mbar/g
adapter permits the temperature
range to be restricted and this leads Table 1: Standard specifications required for pressure transducers to
to a considerable improvement in the be used in a gas exchange analysis
specifications.

3.2 State of the art and new developments at Kistler

The state of the art is represented by Kistler oil-filled, piezoresistive, absolute-pressure sensors. The
pressure is transmitted by an oil medium between a steel diaphragm und a piezoresistive silicon chip.
Sensor Type 4045 has an M14 connecting thread and is available in 2 bar, 5 bar and 10 bar pressure
ranges. The M12 Type 4075 includes a 10 bar version. Because of the significantly lower thermal
loading up to about 120 °C and the high accuracy requirement, sensor Type 4045 is often fitted in the
intake system without an additional adapter. The 2 bar version is used in naturally aspirated engines
and the 5 bar version in supercharged engines.
For applications in the exhaust system with its high thermal loading, it is advisable to use either the
sensor Type 4045 integrated in a cooling adapter or the 10 bar version of sensor Type 4075 in a
switching adapter with integrated water cooling (Fig. 2).

Fig. 2: Existing oil-filled transducers (top) and newly developed oil-free pressure sensors (bottom)

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 -8-

Fig. 3 shows the design of the novel pressure sensors compared with the standard oil-filled
transducers. The newly developed sensors dispense with the fluid transmission medium. Pressure acts
on the piezoresistive chip either directly or via a diaphragm.

State of the Art New Developments

Oil-filled transducer,e.g. Type 4045 Type 4005 for intake Type 4047 for exhaust
(max. temp. 120 °C) (max. temp. 120 °C / 220 °C) (max. temp. 180 °C / 300 °C)

Pressure Pressure Pressure

Oil Filling
Screen

Reference Pressure

High accuracy High accuracy High temperature, cooled

Harsh environment conditions Miniaturization Harsh environment conditions

Fig. 3: Design of the standard oil-filled transducers compared with the newly developed oil-free pressure sensors

For the newly developed pressure sensors, there is a distinction between their use in the intake and
exhaust systems. An M5 absolute pressure sensor Type 4005 without adapter is used in the intake
system with its relatively moderate thermal loading. The sensor is designed for the 2 bar, 5 bar and
10 bar pressure ranges. With this type of sensor, the pressure acts directly on the silicon chip (DCE =
Direct Chip Exposure), which requires various protective mechanisms, particularly for applications in
engines with exhaust-gas recirculation. A version is also offered with a special chip in SOI technology
(Silicon On Insulator) for high temperature applications.
In the exhaust system, effective sensor cooling is essential because of the enormous thermal loading.
To avoid the use of a cooling adapter, the newly developed Type 4047 M14 absolute-pressure sensor
is provided with integrated water cooling. In addition, the piezoresistive chip is realized in high
temperature resistant SOI technolgy (VON BERG, GNIELKA, SONDEREGGER 2003). This sensor
is also designed for the 2 bar, 5 bar and 10 bar pressure ranges. A set-back, thermal-protection
installation usually results in a significant increase in accuracy in determining the pressure level.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 -9-

4 CARRYING OUT THE TEST

4.1 Test bed and measuring point plan

The influence of the pressure transducer position on
the low pressures measured and thus on the BMW F650 GS - Otto
calculation results of the gas exchange analysis was Max. power at 6 500 rpm 37 kW
investigated on a single-cylinder spark-ignition engine Cylinder 1
(Table 2). Stroke 83 mm
Bore 100 mm
The new generation of Type 4005 A DCE uncooled, Compression ratio 11.5 : 1
piezoresistive miniature pressure transducers was used Valves 4
in the intake system, and the Type 4047 A5 water-
Table 2: Test engine data
cooled, piezoresistive sensors for the exhaust system.
As the measuring point plan (Fig. 4) shows, there was
space for only one pressure transducer of the new generation directly at the cylinder head in both the
intake and exhaust systems. The influence of the sensor position on the results of the gas exchange
analysis can be assessed by simultaneous pressure measurement at three different locations in the
intake pipe and in the exhaust pipe.

Fig. 4: Measuring point plan with details of the pressure transducer types used

Reliable cooling and thermal protection of the piezoresistive pressure transducers is necessary,
particularly in the exhaust system. As already mentioned, the Kistler Type 7531 combined switching
and cooling adapters are suitable for the Type 4075 A10 high-precision oil-filled pressure transducers.
These adapters allow fully automatic control in test bed operation and limit the measuring time in
order to protect the sensors.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 - 10 -

As shown in Fig. 5 (right), the combined switching and cooling adapters in the exhaust take up a considerable
amount of space, thereby preventing an installation close to the valve. The exhaust pipe prevents a view of the
exhaust-side position in the cylinder head of the Type 4047 A5 pressure transducer. On the intake side (as
shown in Fig. 5 left) besides to the three sensors, the uncooled, piezoresistive Type 4005 A DCE miniature
pressure transducer can be seen in the cylinder head (near position), slightly below the intake pipe flange.
intake (cylinder head and near) exhaust (near and distant)

Fig. 5: Adaptation of the intake and exhaust system

4.2 Influence of a phase shift on the gas exchange analysis

The following calculations are based on the low pressure measurements with the oil-filled sensors, as
these offer the highest accuracy consistent with state-of-the art technology.
The measuring results presented here using the high-precision, oil-filled pressure transducers illustrate
the influence of the measuring point location on the phase shift and amplitude change of the pressure
signal. Fig. 6 right shows examples of the low-pressure curves measured in the intake pipe and in the
exhaust pipe for a speed of 5 000 rpm with full load.
When the exhaust valve opens (EO), a compression wave occurs due to the difference between the
high pressure in the cylinder and the low pressure in the exhaust pipe. This discharge compresses and
accelerates the exhaust gas from the quiescent state.
When the intake valve opens (IO), the pressure gradient (Fig. 6 left) causes a back flow of the cylinder
charge into the intake pipe until during the intake stroke the cylinder charge flows back into the
cylinder as a result of the reversing pressure gradient.

Fig. 6: Influence of the sensor position on the indicated cylinder pressure at full load, 5 000 rpm

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 - 11 -

The different sensor positions result in differences in phase and amplitude; to a lesser extent, this may
also show up in the calculated mass flow curves (Fig. 7). Here, the calculation is based on the pressure
curves in the same positions in the intake and exhaust systems, e.g. intake "distant" and exhaust
"distant".
A phase shift of and/or amplitude change of the exhaust pressure has very little effect on the calculated
exhaust mass flows (Fig. 7 below left), as the pressure difference between exhaust and cylinder is very
large at the moment when the exhaust valve (EO) opens.

Fig. 7: Influence of a phase shift on the calculated mass flows and cylinder mass

By contrast, a phase shift of the intake pressure has a stronger influence on the calculated intake mass
flows, particularly in the initial phase of the inflow shortly after the intake opens (IO). This is due to
the slight pressure difference between intake system and cylinder (Fig. 7 top left).
The influence of a phase shift of the intake pressure can also be seen in the calculated curve of the
cylinder mass (Fig. 7 right) during the gas exchange after the top dead center (TDC).
Fig. 8 shows the influence of the phase shift on the calculated residual exhaust gas content for the full
load curve. No relationship between the calculated residual exhaust gas content and the speed can be
detected but there is a tendency for the calculated residual exhaust gas content to increase with
increasing phase shift, i.e. towards "distant" – "distant" position.

Fig. 8: Influence of a phase shift on the calculated residual exhaust gas content

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 - 12 -

4.3 Influence of a shift in pressure level on the gas exchange analysis

As the following analysis will show, a shift in pressure level affects the result of gas exchange analysis
significantly. Such a shift may be caused, for example, by a thermal zero offset or a thermal sensitivity
change of the low pressure sensors. In order
to investigate this influence on the results of
the residual exhaust gas content, the intake
and exhaust system pressures were raised in
10 mbar steps. The sensor position
corresponded to the cylinder head position
shown in the measuring point plan (Fig. 4).
As Fig. 9 shows, a shift in pressure level in
the intake pipe and exhaust pipe influences
the residual exhaust gas content calculation.
The gas exchange analysis reacts particu-
larly sensitively to a shift in the exhaust
pressure level. Fig. 9: Influence of a shift in pressure level in the intake and
exhaust pipes on the calculated residual exhaust gas content

A shift in pressure level also affects the calculated mass flows in the intake and exhaust systems. The
top part of Fig. 10 (left) shows the changes in the curve of the mass flows when the pressure level in
the exhaust system is raised by 50 mbar and 100 mbar respectively. This action leads to an increased
phase shift in the intake mass flow and to a significantly changed exhaust mass flow in the valve
overlapping phase. The bottom part of Fig. 10 left shows that a shift in pressure level in the intake
system influences the curve of the mass flows less strongly.

Fig. 10: Influence of a shift in pressure level on the calculated mass flows and cylinder masses

Similar conclusions may be drawn concerning the influence on the calculated cylinder mass, which
reacts significantly more sensitively to a shift in the pressure level in the exhaust system than to a
corresponding shift in the intake system (Fig. 10, right).

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 - 13 -

For this reason, care must be taken to ensure accurate measurement of the pressure level, with a
recommended measuring accuracy of ±10 mbar, particularly with regard to the low pressure indication
in the exhaust. For example, when a piezoresistive pressure transducer with a measuring range of 5 bar
is used, it is essential to ensure that the sum of the errors due to instability of the zero point, thermal
zero offset and thermal sensitivity change does not exceed 0,2 % FSO.
This measuring accuracy can best be achieved on the exhaust side with a combined cooling and
switching adapter, in which a level correction can be carried out for each operating point by the
automatic atmospheric measurement. However, because of its size, the switching adapter cannot be
mounted in an ideal position close to the valve. But, as shown, this has little influence on the quality of
the result and there is no need for a delay correction.

4.4 Suitability of the newly developed pressure transducers for gas exchange analysis
The requirements demanded of piezoresistive sensors for low pressure indication will depend on the
particular application. For example, good dynamic properties of the pressure transducer are necessary
for vibration analyses. Allowance can more easily be made for a shift in pressure level due to thermal
influences (thermal zero offset, thermal sensitivity change).
Requirements are significantly higher for application in a gas exchange analysis. The pressure level
must be known to an accuracy of ±10 mbar, in order to accurately determine the residual exhaust gas
content at the moment the intake valve closes. These specifications are best met by oil-filled pressure
transducers. However, their disadvantage is that for indications in the exhaust system they must be
protected by rather costly and complicated cooling and switching adapters. This makes installation
close to the valve hardly feasible. With fast-running Formula One engines, the additional mass of
adapter and sensor can lead to vibration cracks in the exhaust system. Normally, therefore, their
operating time must be restricted.
The new generation pressure transducers offer clear advantages in application. The uncooled miniature
pressure transducers M5 Type 4005A2/5/10 can be installed close to the valve at least in the intake
system. They have an excellent dynamic behavior and are extremely accurate. These sensors have
already been installed in high-speed Formula One engines and have passed this suitability test with
flying colors. Future development is being concentrated on a miniature pressure transducer suitable for
high-temperature applications, in which measurements close to the valve will also be possible in the
exhaust system.
Where installation close to the valve is not possible (for example in the cylinder head), the water-
cooled M14 sensor Type 4047 is available for both the intake and exhaust systems. With its excellent
dynamic behavior, it is also suitable for a demanding gas exchange analysis, with engine speeds up to
6000 rpm. Its particular advantage lies in its integrated water-cooling. As this type of sensor does not
need bulky adapters, it is ideal for use in confined spaces. Future development is being concentrated
on a pressure transducer suitable for highly dynamic operation.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 - 14 -

Whichever type of sensor is used in the gas exchange analysis, it should be positioned close to the
valve on the intake side, and placed in a thermally protected position on the exhaust side. Appropriate
calibration of the sensors prior to use is also important for accurate measurement of the pressure level.
A shift in the pressure level due to a thermal zero offset or a thermal sensitivity change must not
exceed ±10 mbar. This minimum requirement for optimizing the gas exchange in developed engines is
satisfied, in particular, by the newly developed uncooled M5 DCE transducer Type 4005.

5 CONCLUSIONS
Optimization of the gas exchange in engines on the test bed is becoming increasingly important, but
requires a suitable combination of pressure indicating, analysis and simulation. In order to determine
accurately the residual exhaust gas content at the moment the intake valve closes – an essential initial
requirement for the working process calculation – stringent requirements are placed on the quality of
the low pressure indication used. In particular, the change in pressure level during the measurement
must be kept as small as possible. A measuring accuracy of ±10 mbar is necessary for the gas
exchange analysis. Piezoresistive sensors with special cooling and switching adapters are therefore
recommended, particularly in the exhaust system with its high thermal loading.
Investigations on a single cylinder engine revealed an extraordinarily small influence of a phase shift
of the low pressure curves on the calculated residual exhaust gas content. A phase shift has very little
influence on the calculated mass flow in the exhaust system, but to a greater extent in the intake
system during the initial phase of intake, due to the low difference in pressure with respect to the
cylinder. Thus, to obtain reliable and accurate results from a gas exchange analysis, the transducer
must be placed close to the valve on the intake side and in a thermally protected position on the
exhaust side. Initial investigations on Formula One engines with a mounting situation comparable to
the "near" position produced excellent calculation results for the mass flow and the residual exhaust
gas, based on the Tiger software.
Customers are showing increasing interest in pressure transducers which can be mounted directly in
the intake or exhaust systems without an adapter. The new generation of uncooled miniature pressure
transducers for the intake side is being further developed for use on the exhaust side. This involves
SOI technology for manufacturing chips capable of use up to at least 300 °C. At the same time, work
is being undertaken on sensor-integrated cooling. The reduction in mass resulting from the removal of
the adapter will serve to avoid vibration cracks in the exhaust system.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 - 15 -

One of the motivations for the application of the low pressure indication still is calibration of the
simulation with the measuring results, within the framework of an engine development. Normally, this
type of application concentrates on assessment of the stationary operating points of an engine, as with
the vibration and gas exchange analysis, where the trend is towards optimization of the gas exchange
on engines on the test bed.
The pressure of stricter regulations in exhaust gas legislation means that greater importance will be
attached in the future to the assessment of engines under transient conditions. In 2008, the stationary
13-step test will be replaced by a transient test, heralding new challenges for sensor developers.

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 - 16 -

6 LITERATURE
Rudolf Pischinger, Manfred Klell, Theodor Sams
Thermodynamics of Internal Combustion Engines
Second revised edition. Der Fahrzeugantrieb. Springer Vienna, New York, 2002
Franz Pischinger
Combustion Engines Lecture Paper Vol II
Chair for Applied Thermodynamics. RWTH Aachen, 1988
Univ.-Prof. Dr. Rudolf Pischinger, Dr. Manfred Klell
Potential of New Engine Concepts from the Thermodynamic Aspect
Institute for Internal Combustion Engines and Thermodynamics, Graz University of Technology
8th Symposium, The Working Process of the Internal Combustion Engine
Köhler U., Bargende M.
A Generally Valid Exhaust Residual Gas Model
5th International Stuttgart Symposium Automotive Engeering and Combustion Engines, Stuttgart 2003
A. Wimmer, R. Beran, G. Figer, J. Glaser, and P. Prenninger
Possibilities for the Accurate Measurement of Gas Exchange Processes
4th Int. Symp. for Combustion Diagnostics 2000, Baden-Baden, pp. 63-71
J. von Berg, M. Gnielka, C. Sonderegger, G. Kaden, C. Cavalloni, B. Mukhopadhyay, E. Obermeier
Piezoresistive Low-Pressure Sensor for Harsh Environment Applications
11th Int. Sensor 2003, Nuremberg, pp. 219-224
Christine Burkhardt
Determining Effective Valve Lifting Curves by Measuring High and Low Pressure Processes
8th Symposium, The Working Process of the Combustion Engine, Graz 2001
Frommelt A.; Bargende M.
Possibilities and Limitations of Low Pressure Indication for Gas Exchange Analysis on Motor Car
Engines
Haus der Technik, Essen, 1997
Basshuysen, R. and Schäfer, F.
Combustion Engine Manual
ATZ/MTZ Reference Book, 2002

9th Symposium "THE WORKING PROCESS OF THE INTERNAL COMBUSTION ENGINE", Graz, 2003
 Kistler worldwide
Europe
Germany Austria Denmark/Finland/Norway/Sweden
Kistler Instrumente GmbH Kistler GmbH Kistler Nordic AB
Daimlerstrasse 6 Lemböckgasse 49f Aminogatan 34
DE-73760 Ostfildern AT-1230 Wien SE-431 53 Mölndal
Tel. (+49) 7 11 34 07 0 Tel. (+43) 1 867 48 67 0 Tel. (+46) 31 871 566
Fax (+49) 7 11 34 07 159 Fax (+43) 1 867 48 67 17 Fax (+46) 31 871 597
[email protected] [email protected] [email protected]

France Italy Netherlands

Kistler SA Kistler Italia s.r.l. Kistler B.V. Nederland
ZA de Courtabœuf 1 Via Paolo Uccello, 4 Newtonlaan 115
15, avenue du Hoggar IT-20148 Milano NL-3584 BH, Utrecht
FR-91953 Les Ulis cédex Tel. (+39) 02 481 27 51 Tel. (+31) 30 210 63 67
Tel. (+33) 1 69 18 81 81 Fax (+39) 02 481 28 21 Fax (+31) 30 210 66 66
Fax (+33) 1 69 18 81 89 [email protected] [email protected]
[email protected]

Switzerland/Liechtenstein United Kingdom

Kistler Instrumente AG Kistler Instruments Ltd.
Verkauf Schweiz Alresford House, Mill Lane
PO Box, Eulachstr. 22 Alton, Hampshire GU34 2QJ
CH-8408 Winterthur Tel. (+44) 1420 54 44 77
Tel. (+41) 52 224 12 32 Fax (+44) 1420 54 44 74
Fax (+41) 52 224 14 21 [email protected]
[email protected]

Asia America
Japan Korea, Republic of USA / Canada
Kistler Japan Co., Ltd. Kistler Korea Co., Ltd. Kistler Instrument Corp.
MT Building 3rd Floor, Bow Building 75 John Glenn Drive
7-5, Shibadaimon 2-chome 1580-1, Seocho-3 dong, Amherst, NY 14228-2171
Minato-ku, Tokyo 105-0012 Seocho-ku, Tel. (+1) 716 691 5100
Tel. (+81) 3 35 78 02 71 Seoul, Korea 137-875 Fax (+1) 716 691 5226
Fax (+81) 3 35 78 02 78 Tel. (+82) 2 597 6013 [email protected]
[email protected] Fax (+82) 2 525 6015
[email protected]
China, People’s Republic of
Kistler China Ltd. India Other countries
Unit D, 24 / F Seabright Plaza Kistler Instruments (Pte) Ltd.
9–23 Shell Street India Liaison Office Kistler Instrumente AG
North Point, Hong Kong 2B Century Plaza Export Sales
Tel. (+852) 25 91 59 30 560/562 Anna Salai PO Box, Eulachstr. 22
Fax (+852) 25 91 18 85 Teynampet, Chennai 600 018/India CH-8408 Winterthur
[email protected] Tel. (+91) 44 5213 2089 Tel. (+41) 52 224 11 11
Fax (+91) 44 5213 2331 Fax (+41) 52 224 15 49
Representative Office Beijing [email protected] [email protected]
Mat1000 ©2004, Kistler Instrumente AG

Tel. (+86) 10 8225 2163

Fax (+86) 10 8225 2124 Singapore
[email protected] Kistler Instruments (Pte) Ltd.
50 Bukit Batok Street 23
#04-06 Midview Building
Singapore 659578
Tel. (+65) 6316 7331
Headquarters Fax (+65) 6316 7332
www.kistler.com
[email protected]
Switzerland
Kistler Instrumente AG
920-246e-01.04

Eulachstrasse 22, CH-8408 Winterthur

Tel. (+41) 52 224 11 11
Fax (+41) 52 224 14 14
[email protected]