The Effect of Preparation and Cleaning Techniques

on Electrochemical Testing of Aluminium Anodes

B Linde and R Northey
Cathodic Diecasting, Australia
SUMMARY: When comparing the preparation and cleaning techniques of the various International
Standards for electrochemical testing, there are considerable differences in the methods employed. Some
standards lack sufficient detail, making them wide open to interpretation. This can result in significant
variations in the test results achieved for both anode consumption rate and anode potential.
With the intention of determining a more comprehensive and standardised approach to anode sample
preparation and cleaning, a series of A6 (AS2239) aluminium rod samples were cast and tested. Using the
test method DNV-RP-B401 Cathodic Protection Design Annex B Laboratory Testing of Galvanic
Anode Materials for Quality Control as the basis for the testing, we compared how various preparation
and cleaning scenarios affected the anode consumption rate and anode potential.
This paper reviews the test method and key variables, then analyses and compares what effect preparation
and cleaning has on anode consumption rate and anode potential.

Keywords: Aluminium, Anode, Electrochemical, Consumption, Potential

1.INTRODUCTION
The purpose of this paper is to question what effect preparation and cleaning techniques have on the electrochemical capacity
and the closed circuit potential of Aluminium anode test samples.
This testing is the first step in examining the concentration and interrelationship between key elements within Aluminium
anode alloys with the long term objective of identifying more effective combinations of elements to create a more efficient
Aluminium anode alloy. Before this process could be undertaken, it was necessary to create a comprehensive, clearly defined
procedure. Current Australian and International standards provide most of the details, but some information was missing, this
testing was conducted to aid in creating a rigid, concise internal test procedure to ensure that all of our electrochemical testing
results could be compared and reviewed with confidence that the process is clearly defined and standardised.
The paper will proceed as follows:
1. Compare and the main aspects of preparation and cleaning techniques identified in DNV-RP-B401, NACE TM0191
and AS2239
2. Review the test method
3. Review, discuss and compare the test results
4. Conclusion

2. COMPARISON OF SHORT TERM TEST METHODS

DNV-RP-B401, NACE TM0190 and AS2239 are internationally recognised standards which include short term
electrochemical testing of aluminium anodes. Laboratory based electrochemical testing is a very sensitive testing process,
where seemingly insignificant variations can affect the electrochemical capacity of the anode sample.
As table 1 illustrates, there are some similarities in the testing process of each standard, but there are also some significant
differences, including the surface treatment, where the choice of machined or as-cast samples can affect the end result. Also of
significance is the pre-test cleaning method. These two variables will be the focus of this paper, but it is worth keeping in mind
that there are a number of other variables which can also have a significant effect. Some of these variables will be discussed
towards the end of this paper.

Standard
Title

Method
Sample dimensions
(exposed to test)

Surface Treatment

Pre Test Cleaning

Post Test cleaning
Test solution

Volume of
electrolyte
Cathode

Temperature
Test Duration
Electrochemical
Capacity
Closed Circuit
Potential
(wrt Ag/AgCl Ref
Electrode)

Table 1. Comparison of Short Term Testing Methods

DNV-RP-B401 Annex B (2010) NACE TM0190 (2006)
Cathodic Protection Design
Impressed Current Laboratory
Testing of Aluminium Alloy
Anodes
Mass Loss
Mass Loss & Hydrogen Evolution
15.7cm2
16,000mm3 cubic sample cut
10mm dia (+/-1mm) x 50mm
from anode to be tested.
long (+/- 5mm)
Individual samples cut to
3,900mm2 surface area cast into
sample moulds.
Machined
Saw cut surface for samples cut
from anodes

Rinsed in tap water followed by

ethanol
Chromium trioxide, phosphoric
acid, water
Natural Seawater (adequate
purity and salinity of min 30)
or synthetic seawater ASTM
D1141
Min 10L
Steel/stainless steel Minimum
20x the exposed area of the
sample (minimum 400cm2)
20degC +/-3degC
4 days
Min 2,500 Ah/kg
Less than or equal to
-1050mV

As cast for samples cast into

moulds
Sodium Hydroxide then water

AS2239 (2003)
Galvanic (sacrificial) anodes for
cathodic protection
Mass Loss
The surface area to be used to obtain
current density is calculated from the
initial dimensions of the sample

As cast is preferred but cut and

machined are acceptable.

Nitric Acid then fresh water

Chromium trioxide (or Nitric

Acid), phosphoric acid, water
Synthetic Seawater ASTM 1141

Nitric Acid then fresh water

1.5L

Not specified

Stainless Steel beaker or Carbon

steel container

Aluminium or Stainless Steel

23degC +/-3degC
14 days
2,623 to 2,949 Ah/kg (for Al-ZnHg alloy)
-785 to -1069mV

Not specified
10 days or more
2,500 Ah/kg

Clean Sea Water (resistivity no

greater than 0.3Ohms at 20degC or
Synthetic Seawater ASTM 1141

At least -1050mV

Which method is right? There is no way to definitively answer this question. Even if a series of identical samples were cast
from a single melt (which is almost impossible) and the samples were submitted to each of the test methods listed above, the
results for each would most likely be different. There is no best or correct result.

3. TEST METHOD
3.1 DNV-RP-B401 Test Procedure
The testing procedure was based on the requirements of DNV-RP-B401. Where gaps or ambiguities were identified in the
standard, Det Norske Veritas were happy to assist and provide direction. The testing process was discussed at length with
Robin May (CMET Pty Ltd), who provided much assistance. Without his input this paper would not have been possible.
If you are interested in reviewing the test method, please refer to Annex B Laboratory Testing of Galvanic Anode Materials
for Quality Control of DNV-RP-B401 (2010) Cathodic Protection Design. The internal procedure fills some of the gaps and is
sufficiently detailed to ensure that as many details as possible are controlled.
Note that the DNV-RP-B401 standard was not followed with the post treatment cleaning. In each test the corrosion by-product
was removed with a nylon brush and running water. The DNV-RP-B401 standard specifies cleaning with Chromium trioxide,
phosphoric acid then water. The decision was made to avoid this method on OH&S grounds, as Chromium Trioxide is a
known carcinogen and explodes in the presence of organic compounds and solvents.
Secondly, the use of Chromium trioxide and phosphoric acid has been questioned, as it may remove excessive amounts of
aluminium alloy, which of course would skew the consumption rate test results.

Post cleaning is a variable which was not included in the study. It is an aspect of the cleaning process which does require
further attention. This will be investigated in the near future.
3.2 Anode Test Samples
For the purposes of this paper, a series of anode rod samples were cast in the AS2239 A6 alloy, which is derived from the
commonly used GalvalumIII alloy. A6 complies with the GalvalumIII specification, but has more stringent restrictions on
impurities. Refer to table 2 below for the A6 alloy composition.
Table 2. AS2239 A6 Alloy Composition
Alloy

Zn

In

Cd

Si

Fe

Cu

Mg

Ti

Sn

AS2239
A6

2.0
6.0

0.01
0.02

0.005
max

0.08
0.12

0.12
max

0.006
max

0.02
max

0.02
max

0.02
max

Other
Each
0.02
max

Tot.
0.05

Al
Bal

Five sets of tests were conducted using the A6 alloy and also a single test on a sample cut from a PO610 Aluminium Ingot,
which is the primary aluminium ingot used as the basis to manufacture the A6 alloy. This test was included to provide some
context for the results. It clearly illustrates the difference between an alloy designed to offer cathodic protection and the pure
ingot which does not offer cathodic protection due to the passive film forming on the surface which acts as a barrier.
Test 5 samples were cut from within an anode. It was included so that a comparison could be made between test results from
an actual anode and rod samples which are cast into small diameter rod moulds.
In order to minimise the effects of individual test sample results, 4 samples of each were tested for each of the first 5 scenarios
(refer to table 3 below). Test 6 data was based on results from a single sample.

4.TEST RESULTS
Surface
Preparation
Pre Test
Treatment

Post Test
Treatment
Echem
Capacity
(Average for 4
tests)
Potential
Readings
Polarisation
Time

Table 3. Summary of Test Results

Test 3
Test 4
Machined
Machined

Test 1
As-Cast

Test 2
As-Cast

Ethanol then
Tap water

Sodium
Hydroxide then
Nitric Acid

Ethanol then
Tap water

Sodium
Hydroxide
then Nitric
Acid

Test 5
Saw Cut

Test 6
Machined

Sample cut
from within A6
anode

Sample cut
from PO610
ingot

Ethanol then
Tap water

Ethanol then
Tap water

Nylon Brush & Running Tap Water

2,718 Ah/kg
(Average for 4
tests)

2,711 Ah/kg
(Average for 4
tests)

2,692 Ah/kg
(Average for 4
tests)

2,780 Ah/kg
(Average for 4
tests)

2,742 Ah/kg
(Single sample
tested)

2295 Ah/kg
(Single sample
tested)

-1059mV

-1062mV

-1046mV

-1054mV

-1083mV

-688mV

Almost
immediately

Almost
immediately

5+ hours

3 hours

72+ hours

Almost
immediately

Table 3 indicates that for the 5 anode alloys (Tests 1 to 5) the electrochemical test results vary by 88Ah/kg and the potential
readings at the conclusion of the 96hr test period are quite stable, varying by only 37mV. As-cast samples polarised
immediately, while machined samples took a lot longer to polarise. These results are reviewed in more detail below.

4.1 Consumption Rate

Chart 1. Consumption Rate Comparison (Ah/kg)
TEST 1 As Cast (Ethanol & Water)

2718

TEST 2 As Cast (NaOH HNO3)

2711

TEST 3 Machined (Ethanol & Water)

2692
2780

TEST 4: Machined (NaOH HNO3)

TEST 5 Machined Anode (Ethanol & Water)

2742

2800

2700

2600

2500

2400

2300

2295
2200

TEST 6 Ingot (Ethanol & Water)

To provide some context to the results, the DNV-RP-B401 standard specifies a minimum electrochemical capacity of 2,500
Ah/kg. The results in Chart 1 should be viewed with this baseline in mind. It is evident that all anode alloys (Test 1 to 5) easily
pass the DNV-RP-B401 benchmark. As expected the PO610 primary ingot (Test 6) did not. The effect the various preparation
and cleaning scenarios have in relation to each other will be discussed.
Observation 1: The chemical etch counteracts the effect of the machining. The machined sample which was treated with
Sodium Hydroxide and Nitric Acid (Test 4) was 88 units higher than the equivalent sample which was pre cleaned in ethanol
and water (Test 3).
Observation 2: A comparison between Test 3 and Test 5 indicates a difference of 50 units. Both samples were machined and
both were washed in ethanol and water. The significant difference is that the Test 3 samples were cast into a narrow diameter
rod mould, while Test 5 samples were cast into a large cross sectioned mould and the samples were cut from deep within the
anode. The significant difference in results could be due to one of two factors:
1. The difference between the microstructure of the two samples, as a result of the different mould sizes and therefore
casting modulus. This would bring into question the significance of casting rod samples to represent actual large
cross-section anode.
2. The aggressive cutting of the anode surface employed to remove the Test 5 samples from deep within the anode
caused severe damage to the outer surface of the sample which caused reduced weight loss and therefore a higher
consumption rate.
Further testing is required to identify which is the more significant factor. Follow up papers will focus on this topic.
Observation 3: Test 1 and test 2 were both as-cast with the key variable being that test 1 was cleaned in ethanol and water,
while test 2 was chemically etched. The difference between the two consumption rates was 26 Ah/kg. It appears that the
chemical etch has minimal effect on as-cast samples. This is supported later in this paper where we examine photographs of the
wear patterns at the conclusion of the 96hr test period.

4.2 Electrochemical Potential

Chart 2. Electrochemical Potential Comparison (mV)
TEST 1 As Cast (Ethanol & Water)

-1059

TEST 2 As Cast (NaOH HNO3)

-1062

TEST 3 Machined (Ethanol & Water)

-1046

TEST 4: Machined (NaOH HNO3)

-1054

TEST 5 Machined Anode (Ethanol & Water)

TEST 6 Ingot (Ethanol & Water)
-400

-500

-600

-1083

-688
-700

-800

-900

-1000

-1100

-1200

Chart 2 illustrates that regardless of whether the samples have been machined, as-cast or whether the pre test cleaning
process has been chemically etched or just washed in ethanol and water, by the end of the 96hr test period, each of the anode
alloys (Tests 1 to 5) was close to or more negative than the -1050mV mark desired for anode potential. Chart 2 also clearly
illustrates that in terms of achieving a desirable anode potential, pure aluminium ingot falls a long way short (-688mV).
When comparing the rate at which each of the anode samples polarised (refer to Table 3), machining clearly inhibits
polarization. Note that for the 3 machined scenarios (Tests 3, 4 and 5), the samples took hours to polarise, compared to the ascast samples (Tests 1 and 2), which polarized almost immediately.
This is a problem for anodes in service. Areas of anodes that have been machined, worked with a grinder or hammered often
passivate, while the remaining areas of the anode happily corrode. For this reason, many tender specifications clearly state that
anodes must not be machined, worked with a grinder or hammered.
Machining the anode surface has an effect in two ways. Firstly the folding over of the material can compress oxide layer
over oxide layer. Secondly, there is a possibility that heat generated during machining can cause precipitation of a solid
solution. The final turns in the lathe on a 10mm dia test bar become quite hot (80-120 deg C). This can encourage higher
driving potential but may also encourage a faster consumption rate. This is being investigated further.

4.3 Wear Pattern

Chart 3. Wear Pattern at Conclusion of 96hr Test

Test 1: As cast (Ethanol & Water Clean)

Test 2: As cast (NaOH HNO3 Clean)

Test 3: Machined (Ethanol & Water Clean)

Test 4: Machined (NaOH HNO3 Clean)

Test 5 Machined Anode (Ethanol & Water Clean)

Observation 1: Chart 3 indicates that the as-cast samples (Test 1 and 2) both show a similarly even corrosion pattern. There is
little difference between Test 1, which was pre-cleaned with ethanol and water and Test 2 which was chemically etched.
Notably, these samples are as-cast, but the Test 1photograph indicates that 3 of the 4 test samples have areas of passivation.
The samples were linished to remove excess aluminium just after casting. This photograph supports the previous discussion in
section 4.2 that machining damages the anode surfaces and reduces the efficiency of the anode.
Observation 2: Tests 3, 4 and 5 were all machined. The corrosion pattern is far more uneven compared to the as-cast samples
(Test 1 and 2), with large sections of passivation. As discussed previously, this is a visual representation of how machining has
a detrimental effect on the efficiency of the anode.
The A6 alloy (plus GalvalumIII and A1) does not machine as engineering grades of Aluminium would. Engineering grades
are manufactured with alloy content and modifying elements to promote a particular microstructure, one that produces even
cutting. When machining these anode alloys, it tends to build up on the tool. It then pastes itself to the gap between the
tool and the sample, which results in the swarf doing the cutting (quite brutally) rather than the tool. This causes folding over
of the oxide layers and results in the almost impenetrable finish.
During sacrificial corrosion these areas are then undermined by corroding material until the entire piece falls away (refer
to the example in Figure 1, below), hence a higher weight loss and a lower consumption rate. This is particularly evident in
Tests 3 and 5 (Chart 3), where deep craters are visible adjacent to passive areas.

Figure 1 Machined surface "undermined" by corroding material

Also, there is the problem with precipitation (solid solution) from the machining process (as discussed in section 4.2 above) it
may help to explain why some samples show grand canyon corrosion right alongside original machining marks. The
machining marks are where material has been forced over, almost welded together, the corrosion canyons where the material
has broken away.
Tests 3 and 5 samples were washed in ethanol and water, while Test 4 underwent a chemical etch. It is evident that the
chemical etch had the effect of breaking down some of the damage caused by the machining.
Observation 3: The Test 5 photograph (Chart 3), shows the samples which were cut from within an actual anode. The surface
of each sample was clearly damaged by the saw cut and large sections have passivated. It appears that well over 50% of the
surface has passivated, which is a similar percentage to Test 3, which was also machined and cleaned in Ethanol then tap
water.

5. CONCLUSION
5.1 Key Findings

The chemical etch had little effect on as-cast samples.

Regardless of whether the samples were machined, chemically etched, as-cast or cut from within an anode, if the
chemical composition is right the anode potential was achieved by the end of the 96hr test cycle.

As-cast anode samples all polarised almost immediately, while all machined anode samples took hours to polarise.

Machined anode samples had large passive areas, which consequentially reduced the efficiency of the anode. The
chemical etch appears to have repaired some of the damage caused by machining.

5.2 Discussion
Can the short term DNV-RP-B401electrochemical test in a strictly controlled laboratory environment truly represent an anode
attached to a subsea structure? A number of comments regarding the value of this testing have been made in recent times.
Some suggest that it is an expensive and time consuming exercise which does not accurately represent an in service anode.
The alternative view is that at the time of manufacture, a measure of confidence that the anodes will perform once installed is
essential. Short term electrochemical testing is one piece of the puzzle that provides this confidence.
According to the DNV-RP-B401 benchmarks of 2,500Ah/kg for the consumption rate and -1050mV for the anode potential, it
is clear that all of the anode samples tested as part of this paper were successful, whilst the ingot (not an anode alloy), was
nowhere near these benchmarks. The DNV-RP-B401 short term test should be used as a basic test for quality control purposes
only. It becomes largely irrelevant when the aim of the testing is to produce a more efficient anode. It simply does not have the
resolution required to provide conclusive results.
As the test results started to accumulate, it became clear that the parameters of our test method were insufficient for what we
were attempting to test. For our purposes the DNV-RP-B401 short term test method was far too short (4 days), and the size of
the samples and the volume of electrolyte are too small to provide clear cut results. It is clear that in order to move forward
with our testing we will need to revise some of the fundamental aspects of the test method.
This paper covered some of the variables related to sample preparation and cleaning. It is just the starting point from which
further research is required. Since writing this paper testing has commenced on a wider range of variables including mould size
(modulus), mould temperature and temperature of the alloy. All of these variables significantly influence the grain structure of
an anode and therefore the electrochemical capacity.
These variables and the contents of this paper have not even touched on the concentration of particular elements and the
interrelationship of elements such as Silicon, Iron and Indium within the alloy material. Clearly there is much research to be
done.

6.ACKNOWLEDGEMENTS
Thank you to Robin May (CMET Pty Ltd). Without his input this paper would not have been written.

7. REFERENCES
Standard
Standard
Standard

DNV-RP-B401 Annex B (2010) Cathodic Protection Design

NACE TM0190 (2006) Impressed Current Laboratory Testing of Aluminium Alloy Anodes
AS2239 (2003) Galvanic (sacrificial) anodes for cathodic protection

8. AUTHOR DETAILS
Brent Linde (B App. Sci) is the Marketing and Product Development Manager at Cathodic Diecasting,
where he has been employed since May 2008. He has been part of the Anode Manufacturing industry for
over 14 years. Over that time he has developed an intimate understanding of manufacturing and laboratory
processes.
Russell Northey, A.S.T.C. (Metals Technology). An extensive background in Casting, with Formal
Qualifications in Metallurgy, and a Founding Trade Background that goes back to 1980. Experienced in all
aspects of founding, producing Grey and Ductile Cast Irons, Wear Resistant White Irons, Low Alloy Steels,
Bronze, Aluminium, Aluminium-Bronze as well as experience in EAF Steelmaking, Secondary (Ladle
Refining) Steelmaking, and Continuous Casting of Steel Billets.
The Author has been with Cathodic Diecasting since late 2000