Journal of Transportation Technologies, 2011, 1, 1-6
doi:10.4236/jtts.2011.11001 Published Online January 2011 (http://www.SciRP.org/journal/jtts)
Copyright © 2011 SciRes. JTTs
Dynamic Detection on Airflow for Vehicle Intake System
Based-on Hot-Film Anemometry Sensors
Rong-Hua Ma, Tsung-Sheng Sheu
Department of Mechanical Engineering, ROC Military Academy, Kaohsiung, Taiwan, China
E-mail: rh.ma@msa.hinet.net
Received November 12, 2010; revised December 17, 2010; accepted December 28, 2010
Abstract
The goal of this study is to develop an airflow meter sensor for vehicle intake system detection in internal
combustion engines. The study uses micro-electromechanical process technology to develop a hot-film flow
meter with an alumina substrate and platinum film heater; the hotline method is used to create a micro air-
flow anemometry meter sensor relying on variations in resistance of the platinum film corresponding to dif-
ferent wind velocity at the set temperatures. The micro-sensor is less bulky and simpler structure than ordi-
nary meters, and its small size enables it to provide good sensitivity and measurement precision. The alumina
plate used in this study is produced by polishing an alumina substrate, a platinum film is then deposited on
the plate to complete the micro-heater used in the sensor. Resistance on the sensor side varies as gas flows
through the sensor, and the instrument determines airflow velocity on the basis of the changes in resistance
caused by gas flow differences. Airflow velocity form 10 m/s to 60 m/s are used to test. Resistance displays a
regular slope, indicating the relationship between airflow velocities varies remain predictable throughout the
sensing range. Therefore, the sensor can achieve its airflow measurement purpose completely.
Keywords: Vehicle Intake System, Detection, Anemometry, MEMS
1. Introduction
Due to the growth of vehicle density worldwide, plus
surging demand for motor vehicles in emerging markets,
there will inevitably be stricter vehicle performance re-
quirements in the years to come. In these circumstances,
the effective monitoring of engine air intake volume and
the resulting energy conversion efficiency will depend
crucially on the measurement of air flow rate.
Electronic fuel injection control systems are ubiqui-
tous in motor vehicles, and detective sensors play a vital
role in these systems. Although fuel injection has been
used on and off for a number of years, the current thrust
in emission control and fuel conversation has sparked
renewed interest and a number of cars are now being
offered with fuel injection systems. Fuel injection can
administer a much more closely controlled mixture.
While this mixture is not always ideal, it more closely
approximates what is required. This permits better gas
mileage, smother operation, more power and lowered
exhaust emission levels, etc. Taking air flow meters as a
typical example, flow meters are used to determine the
amount of air taken into the cylinder, enabling an auto-
motive electronic fuel injection system to control fuel
injection time on the basis of airflow and engine rpm
signals. Apart from being much less bulky than ordinary
conventional meters, the small size of micro-sensors
made using micro-electromechanical technology enables
them to achieve greater precision and sensitivity. Fur-
thermore, micro-electromechanical technology and pro-
cesses can overcome past mechanical processing obsta-
cles. Due to micro-electromechanical process technology
is highly compatible with semiconductor processes, ele-
ments can be integrated with semiconductor ICs to form
single-chip systems achieving the system miniaturiza-
tion.
In today’s fast-changing technological society, people
enjoy many more conveniences and comforts than in the
agricultural society of the past. Automation deserves
much of the credit for this progress, and the automation
that has brought modern convenience depends heavily on
sensors. Sensors serve to measure target data, such as
temperature, pressure, sound waves, concentrations of
specific gases, wind speed, flow velocity, and magnetic
R.-H. MA ET AL.
Copyright © 2011 SciRes. JTTs
2
fields, etc. This data must be processed by the conversion
of input signals to output signals, amplification, and in-
tegration before it can be interpreted by the system so as
to respond correctly and appropriately to external physi-
cal phenomena. As a consequence, the ability of sensors
to quickly obtain correct data plays a key role in automa-
tion, control, and actuation. This is how the need for air
stream and airflow direction sensors arose.
Airflow rate and direction sensing are important in
many fields, including the monitoring on vehicle engine
performance. In past literature on the subject, flow me-
ters were usually classified as thermal and non-thermal
sensors on the basis of their sensing method. Among
non-thermal gas flow sensors, designs integrating a pie-
zoresistive structure with a cantilever arm are attracting
favorable attention due to their high sensitivity, high sta-
bility, and good linearity.
1.1. Micro-Electromechanical Systems
The physicist Richard P. Feyman first proposed shrink-
ing science and engineering to a microscopic scale at a
physics conference in 1959, and suggested that this
would be a technological milestone for the future. The
invention of the integrated circuit spurred the rapid ex-
pansion and development of the electronic element in-
dustry. More than forty years later, the electronics indus-
try is developing process technologies capable of achie-
ving even tinier dimensions [1-3]. Apart from enabling
many technical improvements, this miniaturization trend
has also made electronic elements increasingly inexpen-
sive. Micro-electromechanical system (MEMS) technol-
ogy is a derivation of semiconductor process technology,
and therefore draws on the advantages of semiconductor
manufacturing technology [4,5]. MEMS technologies are
used in an extremely wide range of interdisciplinary ap-
plications spanning the areas of physics, chemistry, elec-
tronics, machinery, optics, and materials science, etc.
MEMS also represents an improvement on conventional
mechanical processing technologies, and offers the ad-
vantages of small dimensions, high accuracy, large-scale
batch production, and low costs. Furthermore, MEMS
technology allows the integration of micro-sensors with
other circuits or sensors in order to achieve more power-
ful functions and greater stability and reliability [6,7].
Adamec et al. [8] fabricated a multi-axis hotwire ane-
mometer with four thermo resistor to evaluate flow di-
rection with a power consumption of 25mW. Makinwa
et al. [9] presented a circular-type thermal flow sensor
consists of a heater to detect flow direction and velocity.
Recently, non-thermal gas flow meters have been devel-
oped. These devices have the advantages of lower power
consumption and an improved potential for integration
with other sensors. In spite the previous studies investi-
gated many thermal and non-thermal types of flow sen-
sors, those based on hot-films have not been well dis-
cussed.
2. Design
The flow sensors of the study are fabricated on alumina
and utilize platinum resistors as heating and sensing de-
vices. The sensors tested with different micro-heater and
sensing mechanism orientations, and investigated the
sensing characteristics of single-chip and double-chip
sensors of different sizes. The micro-gas sensors investi-
gated in this study employed an alumina substrate on
which platinum was deposited to produce a micro-heater
and sensing mechanism (Figure 2). Because micro-heaters
can continuously produce a controllable, constant tem-
perature environment, providing a stabile power system
to the micro-heater can maintain the sensor's operating
temperature at the desired level. Wind tunnel experi-
ments were used in conjunction with a connected sensor
to observe the amount of temperature varies.
2.1. Hotline Airflow Anemometry Meter Design
Principles
Hot-line type anemometry meters include several com-
monly-seen structures: heater, heat sensor, and tempera-
ture compensation resistor. Hotline meters employ three
chief sensing principles: constant current, constant tem-
perature, and constant power.
1) When there is a constant current, temperature dif-
ferences caused by the fluid generate different signals.
2) When there is a constant heater temperature, dif-
ferences in output power can be used to determine the
flow rate.
3) When there is constant power, information con-
cerning the fluid can be determined from the effect of the
fluid on resistance and current.
The hotline wind speed meter proposed in this study
uses a platinum heating resistor that generates heat when
a current is flowing. Forced convection will remove the
heat, changing the resistance of the heat sensor. In a
normal environment, the problem of heat dissipation
must be overcome. In Figure 3, the sum of natural con-
vection QN, heat conduction QC, thermal radiation QR,
and forced convection QF is the total heat output of the
heater (see Equation (1)). A temperature compensation
mechanism can be used to deal with the problems of
natural convection and thermal radiation and maintain a
balance with the environmental temperature.
PQNQFQCQR
 (1)
R.-H. MA ET AL.
Copyright © 2011 SciRes. JTTs
3
Because the hotline anemometry meters are prone to
detect errors when the environmental temperature varies,
a temperature compensation mechanism must be used.
Since temperature and resistance will have a linear rela-
tionship within a certain range, appropriate calculations
can be used to correct the data in line with the environ-
mental temperature. Because the temperature compensa-
tion resistance of a meter is the product of the same
processes, it must be very close to the resistance of the
meter itself. Equation (2) shows the general relationship
between resistance and temperature:

TC
RT Tc RR
  (2)
RT: resistance at temperature T
RC: resistance at temperature TC
α: resistance-temperature coefficient
2.2. Design of Hotline Airflow Anemometry
Meter Dimensions
Engine loading is transmitted to the electronic control
unit by means of intake manifold sensors signals. The
anemometry meter sensor is located in electronic con-
trol unit which a protected area away from excessive
heat and is connected to the system by means of a wir-
ing harness plug. This hotline airflow anemometry me-
ter has a rectangular shape (see Figures 1 and 2). The
gray portion consists of the platinum electrodes; all di-
mensions are in mm. The double chip has dimensions of
2 mm × 6 mm, and the single chip has dimensions of 6 mm
× 6 mm.
3. Fabrication and Experiments
Figure 4 shows the chief process steps in the production
of the airflow anemometry meter proposed in this study.
The alumina substrate was first sent to a wafer fab for
polishing (to reduce surface roughness) and cutting to the
desired size (50 mm × 50 mm). Before electron beam
evaporation, microlithography was employed to deline-
ate the platinum portions. A spin coater was used to ap-
ply the photoresist on the substrate. HMDS was first ap-
plied for 30 sec. at 3,500 rpm as an adhesive layer, and
the substrate was dried for 1 min. at 110˚C before appli-
cation of photoresist using the same parameters. The
photoresist soft bake required a constant temperature of
110˚C for 3 min. Exposure was performed after the com-
pletion of coating. After confirming no defects, a devel-
oper consisting of AZ-400K developer mixed with water
in a 1:3 ratio was used to perform development. A diffu-
sion pump and booster pump were used to evacuate the
Figure 1. Double chip mask pattern (2 mm × 6 mm).
Figure 2. Physical drawing of single chip (6 mm × 6 mm).
Figure 3. Principle of hotline wind speed meter operation.
vacuum chamber to a background pressure of 2 × 10-6
Torr. Before deposition of the platinum, a layer of chro-
mium with a thickness of 0.02 μm was deposited on the
alumina substrate as an adhesive layer, and electron beam
(E-beam) evaporation was used to deposit a platinum
layer 0.1 μm in thickness. The lift-off method was used
to produce parallel electrodes and micro-heaters in the
R.-H. MA ET AL.
Copyright © 2011 SciRes. JTTs
4
(a)
(b)
Figure 4. During lift-off process: (a) Before lift-off; (b) Af-
ter lift-off.
shape of the pattern (Figure 4). The double chip’s mi-
cro-heaters had resistance values of 20 , 30 , and 40
, and the single chip’s micro-heaters had resistance
values of 40-80-40 , 80-80-80 , and 40-120-40 . It
can be seen from Figure 5 that the fabrication process
involved the deposition of platinum micro-heaters on an
alumina substrate.
4. Results and Discussion
This study performed wind speed testing using single
and double chips separately. The double chips were
mounted on a flat surface, and were oriented with either
front sensing and rear heating or front heating and rear
sensing (Figure 6). Testing was performed using chips
with different resistances in different environmental
temperatures to compare the chips and perform optimi-
zation (Figures 7-10). As shown, the resistance signals
of the increases approximately linearly with increasing
airflow velocity, thus confirming the stability of the sen-
Figure 5. Overview of fabrication steps for airflow meter
sensor.
Figure 6. Schematic diagram of sensing mechanism.
Figure 7. Front 30 heating, rear 40 sensing.
sor for airflow rate measurement applications. It can be
clearly seen from Figures 7 and 8 that front sensing and
rear heating yields the best performance. Figures 9 and
R.-H. MA ET AL.
Copyright © 2011 SciRes. JTTs
5
Figure 8. Front 30 sensing, rear 40 heating.
Figure 9. Front 20 sensing, rear 30 heating.
10 show the results of pairings of 20 and 30 and of
20 and 40 , which yielded results similar to those of
front sensing and rear heating. The single chips were
then subjected to wind tunnel testing. Because it was
known from testing the double chip that front sensing
and rear heating yielded the best performance, the single
chips were tested only in the front sensing and rear heat-
ing orientation. The results show the faster heat transfer
and the faster response to the airflow rate. Curves were
plotted of the relationship between the airflow anemom-
etry and measured resistance (Figures 11-13). It can be
found the response of the single-chip type of the flow
sensors are faster due to their stronger heat conduction
effect. The results reveal that the sensitivity increases as
Figure 10. Front 20 sensing, rear 40 heating.
Figure 11. Front 40 sensing, rear 80 heating.
the resistance of the sensing element increases.
5. Conclusions
This study successfully demonstrated MEMS-based tech-
logy to produce an airflow anemometry sensor, and also
varied parameters including resistance, size, direction,
interval, and angle, and therefore deforms the piezore-
sistors patterned on their upper surfaces to observe the
effect on sensing characteristics, and determine the best
parameter values and greatest sensitivity. This work will
aid the design of future car and motorcycle airflow ane-
mometry meters. The double chip sensor was less sensi-
tive than the single chip sensor, which indicates that dis-
R.-H. MA ET AL.
Copyright © 2011 SciRes. JTTs
6
Figure 12. Front 40 sensing, rear 120 heating.
Figure 13. Front 120 sensing, rear 40 heating.
tance will influence sensitivity. The distance between the
electrodes and sensing electrodes was found to be the
main influencing factor, and the higher the environ-
mental temperature, the better the performance.
6. References
[1] S. Kim, T. Nam and S. Park, “Measurement of Flow Di-
rection and Velocity Using a Micromachined Flow,”
Sensors and Actuators A, Vol. 114, No. 2-3, 2004, pp.
312-318. doi:10.1016/j.sna.2003.12.019
[2] L. Qiu, S. Hein, E. Obermeier and A. Schubert, “Micro
Gas-Flow Sensor with Integrated Heat Sink and Flow
Guide,” Sensors and Actuators A, Vol. 54, No. 1, 1996,
pp. 547-551. doi:10.1016/S0924-4247(97)80012-X
[3] F. Mailly, A. Giani, R. Bonnot, P. Temple-Boyer, F.
Pascal-Delannoy, A. Foucaran abd A. Boyer, “Ane-
mometer with Hot Platinum Thin Film,” Sensors and Ac-
tuators A, Vol. 94, No. 1-2, 2001, pp. 32-38. doi:10.1016/
S0924-4247(01)00668-9
[4] T. Neda, K. Nakmura and T. Takumi, “A Polysilicon
Flow Sensor for Gas Flow Meters,” Sensors and Actua-
tors A, Vol. 54, No. 1, 1996, pp. 626-631. doi:10.1016/
S0924-4247(97)80027-1
[5] K. A. Makinwa and J. H Huijsing, “A Wind-Sensor In-
terface Using Thermal Sigma Delta Modulation Tech-
niques,” Sensors and Actuator A, Vol. 92, No. 1-3, 2001,
pp. 280-285. doi:10.1016/S0924-4247(01)00584-2
[6] P. Bruschi, A. Diligenti, D. Navarrin and M. Pioto, “A
Double Heater Integrated Gas Flow Sensor with Thermal
Feedback,” Sensors and Actuators A, Vol. 123-124, pp.
2005, pp. 210-215.
[7] C. Tropea, A. L. Yarin and J. F. Foss, “Springer Hand-
book of Experimental Fluid Mechanics,” Springer, Berlin,
2007. doi:10.1007/978-3-540-30299-5
[8] R. Adamec, D. V. Thiel and P. Tanner, “MEMS Wind
Direction Detection: From Design to Operation,” Pro-
ceedings of 2nd IEEE Sensors Conference, Toronto, 22-
24 October 2003, pp. 340-343.
[9] K. A. A. Makinwa and J. H. Huijsing, “Constant Power
Operation of a Two-Dimensional Flow Sensor Using a
Thermal Sigma–Delta Modulation Technique,” Proceed-
ings of 18th IEEE Instrumentation and Measurement
Technology Conference, Budapest, 21-23 May 2001, pp.
1577-1580.