Content uploaded by Anand Kumar Sharma
Author content
All content in this area was uploaded by Anand Kumar Sharma on Nov 30, 2019
Content may be subject to copyright.
Hard anodisation of aluminium and its
application to sensorics
A. Rajendra*
1
, Biren J. Parmar
2
, A. K. Sharma
1
, H. Bhojraj
1
, M. M. Nayak
3
and K. Rajanna
2
Hard anodising is normally carried out with electrolytes at subzero temperatures and with a very
high current density. This may sometimes lead to burning and powdered deposits. In this work, a
method of hard anodising at a higher temperature (10uC) using a pulsed power supply is
investigated. The quality of coating obtained with this process is better than that obtained with the
conventional method. Furthermore, it minimises the refrigeration requirements for cooling the
electrolyte and eliminates the problem of burning and powdery coatings. The application of hard
anodic coating with insulation resistance 1–30 GVup to 100 V dc to a pressure sensor as an
insulating base layer is explored. The pressure sensor diaphragm is hard anodised, and this is
followed by the deposition of a thin platinum–tungsten (92Pt–8W) film as strain sensor. The
fabrication, characterisation and calibration of the pressure sensor and its applications are
discussed.
Keywords: Aluminium alloys, Anodic oxide coating, Hard anodising, Pressure sensor, Pulse power supply
Introduction
Aluminium and its alloys are the most commonly used
structural materials in aerospace industries. Hard
anodising of aluminium is carried out to improve its
surface properties (microhardness, abrasion, corrosion
resistance, etc).
Anodisation is an electrochemical process in which a
metal oxide/hydroxide coating is formed on an anodic
surface in an electrochemical cell using a suitable
electrolyte. In a conventional hard anodising process, a
coating thickness of 50–70 mm and a microhardness
y250 HV is obtained. The process is carried out with a
high current density, while the electrolyte is maintained
at subzero temperatures. As the process proceeds at
constant current density, the bath voltage increases due
to the insulating nature of the coating. After some time,
a point is reached where the dissolution of coating in the
electrolyte exceeds the coating formation. This is mainly
due to the release of a high amount of heat (contributed
by oxidation of aluminium and Joules’ heat) on the
metal/coating interface resulting in a burned, patchy or
powdery coating.
1
The pulsed power supply provides a pulsed current
superimposed on a base level current. The pulsation
allows the surface to recover from the momentary high
current heating effects and thus helps in maintaining the
lower surface temperature. The pulsed current occurs
many times a second, which allows quick and uniform
coating, even over high current density areas, without
burning. The correct combination of the pulse para-
meters can increase the thickness by approximately 30–
35% without significant increase in surface roughness.
2–4
Experimental procedure
The sheet metal coupons (50650 mm, thickness 2 mm)
and machined block material coupons (diameter 50 mm)
are prepared from aluminium alloy (AA 6061). These
coupons were processed for hard anodising as follows.
(i) Ultrasonic cleaning with trichloroethylene at
room temperature for 2–5 min.
(ii) Alkaline cleaning: sodium carbonate 20–
30 g L
–1
; trisodium orthophosphate 25–
35gL
–1
; sodium metasilicate 15–25 g L
–1
;
wetting agent 1 g L
–1
; temperature 60¡5uC;
time 3–5 min; agitation, mechanical/mild.
(iii) Water rinsing.
(iv) Acid desmutting: sulphuric acid (r51.84) 5–
15 mL L
–1
; hydrofluoric acid (40%) 10–
12 mL L
–1
; nitric acid (70%) 20–25 mL L
–1
;
temperature 25¡5uC; time 2–3 min; agitation,
mechanical/mild.
(v) Water rinsing.
(vi) Hard anodising: sulphuric acid (r51.84) 80–
120 mL L
–1
; oxalic acid 10–20 g L
–1
; current
density 444.4–888.8Am
22
; duty cycle 56%,
63%, 71% and 83%; agitation, continuous, by
compressed air; time 30–40 min.
(vii) Water rinse and drying.
1
Thermal Systems Group of ISRO Satellite Centre, Bangalore 560 017,
India
2
Instrumentation Department of Indian Institute of Science, Bangalore 560
012, India
3
Advanced Transducer Development Division, Liquid Propulsion Systems
Centre, 80 Ft. Road, HAL 2nd Stage, Bangalore 560 008, India
*Corresponding author, email rajendra@isac.ernet.in
ß2005 Institute of Materials, Minerals and Mining
Published by Maney on behalf of the Institute
Received 8 January 2004; accepted 5 April 2005
DOI 10.1179/174329405X50000 Surface Engineering 2005 VOL 21 NO 3193
Results and discussion
The experiments were conducted with the electrolyte
operating at temperatures of 10uC, 15uC and 20uC. The
effects of the following process parameters were
investigated: duty cycle, frequency, current density and
processing temperature.
Effect of duty cycle
Duty cycle (%)5On6100/(OnzOff), where On and Off
refer to pulse on and pulse off duration, respectively, in
milliseconds.
The increase in duty cycle results in higher current
density, so higher values of thickness are obtained.
However, the maximum values of microhardness are
obtained at 60–75% duty cycle. Above 80% duty cycle,
the advantage of the duty cycle decreases slowly because
there is no influence of pulse effect; that is, the benefit of
relaxation and recovery is not significant, due to a
shorter ‘Off’ period. Lower duty cycles reduce the
current density, and hence the time required to build
up the required thickness increases considerably, result-
ing in decreased microhardness. Typical relationships
among the duty cycle, microhardness and thickness are
shown in Fig. 1.
Influence of current density
Typical relationships among current density, duty cycle
and temperature are shown in Fig. 2. The current
density in the range 555.5–777.7Am
22
provides the
film with the highest microhardness. At current densities
higher than 833.25 A m
22
, burning occurs and a patchy
appearance of the coating is observed, indicating
dissolution at high current density areas.
Effect of electrolyte temperature
Experiments were conducted at 10uC, 15uC and 20uC.
The results show that the properties of a hard anodic
coating obtained at 10uC with a pulse rectifier match
those of a hard anodic coating obtained at 25uC with
the conventional process. At 10uC the rate of dissolution
is low, and a smooth surface with a dark grey to black
colour is obtained. At 15uC and 20uC, the microhard-
ness drops and the coating has a lighter appearance.
Graphs depicting microhardness v. current density and
microhardness v. duty cycle at different operating
temperatures are given in Fig. 3 (duty cycle 71%) and
Fig. 4 (current density 777.7Am
22
), respectively.
Effect of On–Off cycle frequency
Frequency (Hz) is defined as number of On–Off cycles
occurring within 1 s. Experiments were carried out at
three different frequencies (1000, 100 and 10 Hz). It was
found that higher frequencies are not beneficial for the
hard anodising process.
Figure 5 shows the variation of microhardness with
respect to change in duty cycle at 10uC. The problem of
1 Effect of duty cycle (DC) on microhardness and coat-
ing thickness at 811.03 A m
22
current density (10uC)
2 Effect of current density on microhardness and coat-
ing thickness at 63% duty cycle (DC) (10uC)
3 Effect of electrolyte temperature on microhardness at
71% duty cycle. CD, current density
4 Effect of electrolyte temperature on microhardness at
777.7Am
22
current density. DC, duty cycle
5 Effect of frequency variation on microhardness at
666.6Am
22
current density (10uC). DC, duty cycle
Rajendra et al. Hard anodisation of aluminium
194 Surface Engineering 2005 VOL 21 NO 3
burning and powdery coatings was encountered at
higher frequencies. Hence, lower frequencies in the
range 5–10 Hz were selected.
Rate of thickness build-up
In a conventional hard anodising process, the time
required to obtain a requisite coating thickness of 50–
70 mm is 60–90 min (with an operating current density
of 333.3–444.4Am
22
). With a pulse rectifier, the same
coating thickness is achieved in 30–40 min by employing
high current density (Fig. 6).
Testing and evaluation
The hard anodised test coupons obtained at optimum
operating parameters were subjected to testing and
evaluation as shown in Table 1.
Application of hard anodic coating in sensorics
Strain gauge based pressure sensors are popular choices
among pressure measurement devices. One type of
transducer employs a strain gauge bonded to a
diaphragm, whereas another method is based on the
deposition of a thin film strain gauge directly on the
diaphragm. The deflection of the diaphragm directly
induces strain on the thin film, causing a change in the
resistance values corresponding to the tensile and
compressive load. This is measured using a
Wheatstone bridge configuration and is therefore related
to the pressure being measured.
5
The conventional bonded strain gauge sensor consists
of a sensitive wire or foil and is completely attached by
an adhesive to the component whose strain is to be
measured. As the strain sensitive wire or foil conducts
electricity, it has to be electrically isolated from the
component. The component is generally made of
conducting materials such as metals. The adhesive and/
or insulating backing material provide the required level
of electrical insulation. The commonly used adhesive
and backing materials are phenolic or epoxy resins, and
this may prevent the use of such bonded gauges in
radioactive and high temperature environments. Also, as
the adhesive and backing material are in the force path,
the accuracy of measurement is limited by the thickness
and characteristics of these materials. Normally, the
force required to produce the displacement in a bonded
type foil gauge is large compared to that required in an
unbonded foil gauge. Hence, the application of hard
anodic coating to a pressure sensor as an insulator and
base layer is explored. This anodic layer has an
insulation resistance in gigaohms up to 100 V dc, and
this is an added advantage for this purpose. Figure 7
shows the variation in the insulation value of hard
anodic coating v. applied voltage (dc).
A flowchart giving details of fabrication, packaging
and testing of the pressure sensor is given in Fig. 8.
Fabrication of diaphragm
The diaphragm of the pressure sensor was fabricated
using aluminium alloy AA 6061. This material was
chosen because of its easy availability, machinability and
good fracture toughness. Figure 9 gives details of the
diaphragm.
The thickness of the diaphragm was calculated in the
standard way for the desired pressure range. The
diaphragm was polished to a mirror finish and hard
anodised to a coating thickness of 50–70 mm. The
insulation values of the hard anodic film were measured
at 30–1.5GVin the 10–100 V range.
A thin film sensing strain gauge of platinum–tungsten
alloy (92Pt–8W) was deposited on the hard anodised
diaphragm over a thin binder layer of chromium. The
platinum–tungsten alloy was selected because of its high
gauge factor (4.3). Direct current sputtering at 350 V dc
and 100 mA under a working pressure of 5610
24
torr
was used for deposition of platinum–tungsten alloy to a
thickness of 200–300 A
˚. A precision mechanical mask
was used to define the meandering path of the four
element pattern of strain gauge on the diaphragm. The
location of the four elements of the strain gauge are such
that two gauges located near the centre of the diaphragm
undergo tensile strain, whereas the other two located
near the edge will experience the compressive strain. The
pattern details are shown in Fig. 10.
Assembling, testing and calibration
The strain gauge resistance and insulation values of each
element of the strain gauge were checked, and the lead
wires were attached to the strain gauge element with
silver paste. The diaphragm was mounted on a test
fixture assembly for calibration and testing. The pressure
6 Coating thickness build-up rate at 666.6Am
22
current
density (10uC). DC, duty cycle
Table 1 Testing and evaluation of hard anodic films
Test Conditions Remarks
Visual inspection As obtained after humidity, thermal cycling and
thermovacuum test
No patch, discoloration or burning marks
observed
Adhesion test By tape peel after scribing with a sharp knife
through the deposit to base metal 11 parallel lines
approximately 2 mm apart
No detachment of coating observed
Thickness measurement ASTM-B-244 eddy current method 60–70 mm
Microhardness evaluation ASTM-E-384 .350 HV
Humidity exposure 95% relative humidity at 50uC for 48 h No degradation observed
Thermal cycling test 270uCtoz125uC, 5 min dwell time, 1500 cycles No degradation observed
Thermovacuum performance 270uCtoz125uC, 2 h dwell time, 10 cycles No degradation observed
Rajendra et al. Hard anodisation of aluminium
Surface Engineering 2005 VOL 21 NO 3195
calibration was carried out up to 5 bar in steps of 1 bar,
in both ascending and descending order. The strain
gauge output in millivolts was recorded and plotted
against the applied pressure; this is tabulated in Table 2
and shown in Fig. 11. Full scale output was recorded,
and the non-linearity and hysteresis (NLzH) were
computed using a standard program. Table 3 sum-
marises the test and calibration test results.
7 Variation in insulation values of hard anodic coating v. applied voltage
8 Process flowchart. FSO, full scale output; NLzH, non-linearity and hysteresis
9 Details of diaphragm (all dimensions in millimetres)
10 Hard anodised diaphragm with sputtered Pt–W strain
gauge elements
Rajendra et al. Hard anodisation of aluminium
196 Surface Engineering 2005 VOL 21 NO 3
Conclusions
Application of the hard anodising process to aluminium
alloys with the use of a pulsed power supply is described.
The process provides a hard anoxic film with improved
mechanical properties at relatively high temperatures
(10uC, against 25uC with a conventional dc power
supply).
The process minimises refrigeration requirements and
eliminates the problem of burning, which is frequently
encountered with a dc power supply. The process time
required to build up the required thickness is also halved
in comparison to conventional hard anodising.
The insulation values of hard anodic coating fall in
the range 1.5–30 GV, making it suitable as an insulating/
backing layer for pressure sensing element (strain gauge)
deposition.
The process sequence for fabrication of a low cost,
disposable type pressure sensor and its testing and
calibration details are outlined. The low pressure sensor
thus developed has a 0–5 bar pressure range with
NLzH better than 1.2% and sensitivity better than
0.5mVV
–1
. These properties could be improved further
by using thinner hard anodic coatings and thicker
diaphragms.
References
1. S. Wernick and R. Pinner: ‘The surface treatment & finishing of
aluminum and its alloys’, Vols 1 and 2, 4th edn, 563; 1972,
Teddington, UK, Robert Draper Ltd.
2. J. Rasmussen: Metal Finishing, 2001, 99(9), 46.
3. R. Duva: in ‘Electroplating engineering handbook’, (ed. L. J.
Durney), 684; 1984, New York, Van Nostrand Reinhold Co.
4. A. W. Brace and D. G. Sheasby: ‘The technology of aluminum
anodizing’, 2nd edn, 159; 1979, Stonehouse, Technicopy Ltd.
5. K. Rajanna, M. M. Nayak, R. Krishnamurthy and S. Mohan:
Vacuum, 1997, 48(6) 521.
Table 2 Calibration data
Pressure, bar 0 1234543210
Output, mV 0.42 1.61 2.79 3.92 5.16 6.39 5.21 3.97 2.82 1.65 0.42
Table 3 Pressure sensor test results
Parameter Value
Pressure range 0–5 bar
Excitation voltage 5 V dc
Gauge resistance 1.3¡0.2kV
Insulation resistance .50 MVat 50 V dc
Non-linearity and hysteresis Less than 1.2%
Sensitivity .0.5mVV
–1
11 Calibration curve: pressure (Pr) v. output (O/P) (mV)
Rajendra et al. Hard anodisation of aluminium
Surface Engineering 2005 VOL 21 NO 3197