Agilent

Small Engine Dynamometer Testing

Performing Dynamometer Testing on Combustion Engines

Application Note

Dr. Gitano Horizon

University Science Malaysia
About the Author
Dr. Horizon worked at Seagate R&D centers in the US and factories around
the world including Mexico, Singapore, Malaysia, Thailand, and China
(1992 – 2001). Since 2001 he has been working as a consultant focusing on
controls and mechatronics systems. He was a lecturer at Colorado State
University (2002 – 2004) and has been a professor at the University Science
Malaysia since 2006. Dr. Horizon also founded a consulting company, called the
Focus Applied Technologies Sdn. Bhd. in 2008 as a university technology spinoff
company.

Areas of specialization: mechanics, dynamics, mechatronics, internal

combustion engines, microcontroller applications

For full biodata, visit http://www.skyshorz.com/personal/bio.php

2
Measuring Engine Performance with Agilent USB DAQ Modules

Fuel

Scale Serial port

USB port
Throttle
position sensor
USB DAQ

Test Load cell

Dynamometer
engine

Engine Speed wheel

temperature and pickup
sensor

Figure 1. Schematic overview of test system

Testing of engine performance is often important in the development of engine

and fuel technologies. Many parameters affect an engine’s performance: the
basic engine design, compression ratio, valve timing, ignition timing, fuel,
lubricant, and temperature.

The most critical factor governing an engine’s power, fuel consumption and
emissions is the amount of fuel consumed relative to the amount of air ingested.
This is expressed as the air/fuel ratio; that is, the mass of air used divided by
the mass of fuel burned. Larger engines typically have computer-controlled fuel
injection governing the air/fuel ratio while smaller engines still use carburetors
Figure 2. Motorcycle engine mounted
for controlling the air/fuel ratio. Proper tuning of an engine’s carburetor requires
for testing
careful measurements of torque, speed, temperature, and fuel consumption
as a function of throttle position. Agilent USB data acquisition (DAQ) modules
provide an inexpensive, flexible solution for engine testing.

To test engine performance in the laboratory, the engine is coupled to a

dynamometer, which may be either an engine dynamometer (connected directly
to the engine output shaft) or a chassis dynamometer (connected to the drive
wheels). The dynamometer, which provides a load to the engine, can be easily
controlled to allow testing under a wide range of speeds and torque.

Figure 3. Top view of dynamometer

(top) and engine (lower half)

3
 Overview of Key Measurements
The most critical performance measurements are output torque and speed. The
engine’s brake power (output power) is determined from the engine speed and
torque using the following equation:

Pb = T * ω

Pb is brake power in watts

T is shaft torque in Nm
ω is shaft rotational speed in rad/sec

Speed is measured with a pickup that monitors the spinning dynamometer.

Typically, a toothed steel wheel is placed on the shaft and an inductive pickup
is placed close enough to the wheel such that it generates pulses as the teeth
pass by. This time-based pulse signal is analyzed in order to measure the period
between pulses and converted to frequency. Shaft speed is directly proportional
to the measured frequency. If the engine output is taken directly from the
crankshaft, then the dynamometer shaft speed equals the engine speed. If the
output shaft is geared (with respect to the engine speed), then the dynamometer
Figure 4. Speed wheel (top) and speed must be multiplied by the gearing ratio to determine the engine speed.
pickup (below)
Torque is measured by attaching a load cell (force transducer) to the
dynamometer housing. The load cell strains slightly with applied load, and the
strain is measured using strain gages connected to a preamplifier that outputs
a voltage proportional to the load. In accordance with Newton’s first law, the
measured torque on the dynamometer housing equals to the torque on the
dynamometer shaft.

Another useful parameter to measure is the throttle position. Many fuel-injected

vehicles include a throttle position sensor (TPS) that relays the throttle position
to the electronic control unit (ECU) of the fuel-injection system. The TPS is
basically a potentiometer configured as a voltage divider that rotates with the
throttle. For carburetted engines, a TPS can easily be added to the throttle
Figure 5. TPS pot (right) and throttle linkage, allowing direct measurement of throttle position.
linkage (left)
Generally, a controller provides servo control of the load in order to maintain
constant speed or torque. A typical test might first set the dynamometer
controller to a specific speed and then measure torque variation as a function of
throttle setting. Alternatively, the throttle may remain fixed (for example, wide
open) and the speed control set point varied. The resulting speed and torque
values are measured, and power is calculated i.e., speed times torque.

Figure 6. Controller in the foreground,

engine and dynamometer behind

4
 During testing, important parameters such as fuel consumption, exhaust
emissions, and air flow may also be measured. These provide quantitative
measures of how efficiently an engine is running. Several temperatures, such as
exhaust gas temperature, lubricant temperature, and cylinder head temperature,
may also be worth measuring. The temperature measurements are useful in
determining if the engine is being operated safely or is overheated. Often, engine
parameters such as air/fuel ratio vary from test to test and the resulting effects
on fuel consumption, overall efficiency, or power are measured.

This type of testing requires fairly low data-rate measurements because all
variables are averaged over several engine cycles. Data acquired in this manner
are referred to as “cycle-average data.” Relatively slowly changing values, such
as engine speed, fuel flow, engine temperature, and manifold pressure, are
measured directly. Torque fluctuates over the cycle of an engine, being highest
during the power stroke and lowest during the compression stroke. Fluctuating
signals may be filtered electronically to remove this higher-frequency variation,
and various digital techniques (such as exponential averaging) may also be
applied once the data has been acquired.

Dynamometer

Speed wheel Engine

Drive shaft

Speed pickup

Load cell

Figure 7. Side view of dynamometer and engine

5
The Measurement System
Given the volume of data and the number of real-time calculations required, a
computer-controlled data acquisition (DAQ) system is generally used to
calculate, display, and archive test data from tests. Agilent DAQ modules
devices such as the U2355A multifunction DAQ module and U2802A 31-chan-
nel thermocouple input module are good examples. When used together, these
devices are especially helpful in engine testing that requires measurements of
speed, torque, throttle position, and temperature of cylinder and exhaust.

The speed, torque, and throttle position signals are measured using the analog
input ports of the U2355A. Shaft speed is calculated from the period of the
sinusoidal speed using the formula:

Speed = 60 / ( n * T )

Speed is shaft speed in RPM

T is the period of the sine wave
n is the number of teeth on the speed wheel

Depending on the engine characteristics, this speed value may fluctuate based
on the timing of each sample acquisition. For example, a single-cylinder engine
will slow down significantly during compression and speed up during the
expansion or power stroke. To reduce this fluctuation, it is common to perform
an exponential average of the calculated speed value. A simple way to
implement exponential averaging is to declare the averaged value equal to a
weighting factor times the current value, then add one minus the weighing
factor times the previous averaged value. In a scripted language this might be
represented by the following statement.

AverageSpeed = Speed * 0.1 + PreviousAverage * 0.9

PreviousAverage = AverageSpeed

The torque value — minus an offset or zero level — can be multiplied by a

calibration factor, as follows:

Torq = (TorqVoltage – VoltageOffset) * NmperVolt

Torq is the value of the shaft torque in Nm

TorqVoltage is the voltage read from the load cell
VoltageOffset is the “no load” torque reading
NmperVolt is the calibration factor in Nm per volt

Using Agilent’s VEE programming language, we can perform both a

linear calibration of the TPS and an exponential average (Figure 8). In this case,
the current TPS voltage is multiplied by 60 percent and added to the previous
value (from the shift register), which is multiplied by 40 percent. This effectively
yields an exponential average of the TPS voltage, thereby reducing the influence
of electronic noise and high-frequency variations in the signal. The results are
then sent to the TPC calibration formula, where they are linearly interpolated
between the high and low TPS calibration levels. The result is given as a
percentage of full throttle: zero percent when the throttle is closed and 100
percent when it is wide open.

6
Figure 8. Exponential averaging and linear calibration of TPS value

The torque value can also be exponentially averaged, as with the speed value.
Power may then be calculated from the measured speed and torque values:

Power = Torq * AverageSpeed * 6.28 / 60

Power is the shaft power in watts

6.28 / 60 converts the speed from RPM to radians/sec

Fuel consumption is generally calculated based on changes in the weight of the

fuel tank over a given period of time (for example, one minute). Precision scales
can be used to report the weight of a fuel tank through a serial port. The scale
weight can be polled for the initial fuel tank weight and then polled again later
after the running engine has consumed part of the fuel. The fuel consumption
is then computed as follows:

FuelConsumption = ( Weight1 – Weight2 ) / FuelPeriod

FuelConsumption is the fuel consumption rate in gm/sec

Weight1 is the initial reported fuel scale weight in grams
Weight2 is the reported fuel scale weight after time FuelPeriod
FuelPeriod is the time between the two weight readings in seconds

Once the data has been properly calculated, the results are presented in
numerical and graphical forms on a computer display. The data is also saved to
a data file for later analysis.

Several temperatures are generally monitored during testing. For example,

lubricant temperature is monitored to ensure that the engine has achieved
normal operating temperature. If the lubricant is too cold, engine friction will be
unusually high, adversely affecting power and fuel-efficiency measurements.
Another example is cylinder-head temperature, which is especially crucial
during high-power operation. Running with a high air/fuel ratio (for example,
too little fuel for complete combustion of the oxygen present in the air) is
referred to as lean operation. While lean operation is desirable at low loads, as
it is cleaner and more efficient, it can rapidly overheat and destroy an engine at
high loads. Therefore, it is important to monitor the cylinder-head temperature
at high power conditions to ensure that the temperature is low enough to avoid
engine destruction.

7
Previously, either a dedicated thermocouple reader was required, or a special
amplifier was needed. With the Agilent U2802A, however, we can now read
thermocouples directly, greatly simplifying the system.

Sample data and screen shots

Figure 9. Computer screen with Agilent VEE front panel during engine testing

Figure 10. Post-processed data (from run shown above) with power calculated

Figure 11. Agilent VEE 8.5 program used during engine testing

8
The Testing Sequence
A typical test will consist of running the engine at a number of speeds and loads
under varied engine conditions. For example, when investigating the effect of
carburetor tuning, it is important to test at both high and low speeds and high
and low throttle settings. Table 1 shows an example set of test points.

Table 1. Example test conditions

Speed Torque Power
(rpm) (Nm) (W)
1500 1 157
1500 8 1257
2500 4 1047
2500 8 2094
4000 5 2094
4000 12 5027
6000 10 6283

At each test point, the system software sets the dynamometer speed and then
opens the throttle until the desired torque is achieved. The engine is then
operated at the desired condition long enough to consume a significant amount
of fuel. This generally takes one or two minutes, which is enough time to obtain
results from exhaust-gas emissions analysis. Once the test point is finished, the
system proceeds to the next test point. Once all of the test points have been run,
carburetor tuning may be adjusted and a new test performed using the entire set
of test points.

Trends in fuel consumption, emissions, or engine temperature can then be seen

as a function of the carburetor settings, making it possible to determine the
optimum settings. Sample results from an example study are shown in Figure 12.

Stock Carburetor Tuning

Tuned Carburetor

Figure 12. Brake-specific fuel consumption vs. engine power for stock and tuned
carburetor

9
Offline Analysis
Often, it may be useful or necessary to calculate other factors at a later time.
Examples include overall thermal efficiency of the engine, the brake specific fuel
consumption, and the brake mean effective pressure.

The overall fuel conversion efficiency, ηt, (sometimes called brake thermal
efficiency), is the power output divided by the chemical energy of the fuel
consumed. Chemical energy of the fuel can be calculated by multiplying the
mass consumed per unit time (for example, FuelConsumption mentioned
previously) times the heating value (the chemical energy, typically approximately
42 kJ/gram for hydrocarbon fuels) of the fuel Qlhv:

ηt = Pb / ( FuelConsumption * Qlhv)

A convenient way to compare engines of different sizes is through brake-specific

fuel consumption (BSFC), which is the fuel consumption rate divided by the
power output:

BSFC = FuelConsumption / Pb

Finally, a common measure of an engine’s power density is the brake mean

effective pressure (BMEP), which is calculated from the brake power divided by
the engine’s displacement volume times the engine speed. One additional factor
is the number of revolutions per power stroke. This factor, denoted n, is 2 for
four-stroke engines and 1 for two-stroke engines:

BMEP = Pb * n / ( Vd * Speed / 60 )

BMEP is the brake mean effective pressure in Pascals

n is the number of revolutions per power stroke
Vd is the displacement of the engine in m3
Speed is the engine speed in RPM

Conclusion
Given the world’s limited petroleum resources and the growing concern over
environmental issues, it is important and beneficial to tune engines to ensure
the proper balance of power, fuel consumption, and emissions. Agilent USB
DAQ modules such as the U2802A and the U2355A allow rapid development of
flexible systems capable of measuring the various parameters required for proper
engine tuning. These systems also have the added advantage of simultaneously
receiving data from other sources such as serial devices. Finally, the U2802A has
the ability to directly read thermocouples, simplifying system design and helping
ensure safe operation of the engine under test. This provides an excellent
solution for seamless integration into existing engine dynamometer test beds.

10
 Agilent Email Updates Remove all doubt www.agilent.com
For more information on Agilent Technologies’
www.agilent.com/find/emailupdates Our repair and calibration services
products, applications or services, please contact
Get the latest information on the will get your equipment back to you, your local Agilent office. The complete list is
products and applications you select. performing like new, when prom- available at:
ised. You will get full value out of
www.agilent.com/find/contactus
your Agilent equipment through-
out its lifetime. Your equipment
Americas
Agilent Direct will be serviced by Agilent-trained
Canada (877) 894-4414
technicians using the latest factory
www.agilent.com/find/agilentdirect Latin America 305 269 7500
calibration procedures, automated
Quickly choose and use your test United States (800) 829-4444
repair diagnostics and genuine parts.
equipment solutions with confidence. You will always have the utmost
Asia Pacific
confidence in your measurements.
Australia 1 800 629 485
China 800 810 0189
Agilent Agilent offers a wide range of ad-
Open Hong Kong 800 938 693
ditional expert test and measure-
India 1 800 112 929
www.agilent.com/find/open ment services for your equipment,
Japan 0120 (421) 345
Agilent Open simplifies the process including initial start-up assistance,
Korea 080 769 0800
of connecting and programming onsite education and training, as
Malaysia 1 800 888 848
test systems to help engineers well as design, system integration,
Singapore 1 800 375 8100
design, validate and manufacture and project management.
Taiwan 0800 047 866
electronic products. Agilent offers Thailand 1 800 226 008
open connectivity for a broad range For more information on repair and
of system-ready instruments, open calibration services, go to:
Europe & Middle East
industry software, PC-standard I/O Austria 0820 87 44 11
and global support, which are Belgium 32 (0) 2 404 93 40
combined to more easily integrate test Denmark 45 70 13 15 15
system development. www.agilent.com/find/removealldoubt
Finland 358 (0) 10 855 2100
France 0825 010 700*
*0.125 € fixed network rates
Germany 01805 24 6333**
**0.14€/minute
Ireland 1890 924 204
Israel 972-3-9288-504/544
Italy 39 02 92 60 8484
Netherlands 31 (0) 20 547 2111
Spain 34 (91) 631 3300
Sweden 0200-88 22 55
Switzerland (French) 41 (21) 8113811(Opt 2)
Switzerland (German) 0800 80 53 53 (Opt 1)
United Kingdom 44 (0) 118 9276201
Other European Countries:
www.agilent.com/find/contactus
Revised: October 24, 2007

Product specifications and descriptions

in this document subject to change
without notice.

© Agilent Technologies, Inc. 2008

Printed in USA, June 6, 2008
5989-8070EN