1017
1 INTRODUCTION
All elements, compounds and substances in nature
strive to reach the most stable form, the so-called
ground state. In this state they have the highest
stability and the lowest energy. Iron (Fe) is a stable
element when analysing its structural stability and
nuclear stability, but it is not stable on a chemical level
under normal atmospheric conditions. In nature, under
normal atmospheric conditions, iron will try to take on
a more stable form, leading to corrosion. According to
the Standard Terminology and Acronyms Relating to
Corrosion, the definition of corrosion is “the
deterioration of a material, usually a metal, that results
from a chemical or electrochemical reaction with its
environment” [1]. According to the Wartsila
Encyclopaedia of Ship Technology, corrosion is “the
process of deterioration of metals and their properties,
following a reaction with surrounding environment”
[2]. The same encyclopaedia also states that “it readily
oxidizes in moist air” [2]. To summarise, iron corrosion
is a normal, natural process in which oxygen and water
from the atmosphere help the iron to form more
chemically stable forms, such as oxides.
Objects in the maritime industry are exposed to
high relative humidity, chlorides, temperature
fluctuations, different weather conditions and
sometimes other forces of nature [3]. These harsh
conditions and heavy use can degrade components and
materials much faster than in land-based industries
[4,5]. Therefore, protective measures must be taken to
protect materials and equipment from these influences
and thus from deterioration. A statement on the most
commonly used protective measure was made more
than seventy years ago: “Paint has been used for a long
time for the protection of metals against corrosion” [6].
Originally, paint served as a barrier against the hostile
Analysis of the Surface Protection of a Maritime Object,
Case Study
M. Jurjević
1
, A. Anić
1
, B. Lalić
2
& L. Stazić
2
1
University of Dubrovnik, Dubrovnik, Croatia
2
University of Split, Split, Croatia
ABSTRACT: In the maintenance of maritime facilities, one of the most common task is painting, that is, applying
a protective hard coating to surfaces that are exposed to corrosive action. One of the main goals of painting and
corrosion protection is to provide the most economical protection of structures. This paper presents the
assessment of damage to the structure from the process of corrosive action, the control measurement of surfaces
after the removal of corrosive deposits in order to reach the conclusion that the surface is ready for paint
application. The measurements are listed in tables and are used for statistical analysis and the creation of control
charts. These measures help to decide the potential process capability and also indicate by how much the tolerance
limits exceed the actual distribution limits and whether further improvements are required. The control charts
and the QI Macros software were the basic statistical tools used. In summary, the control charts provide the
opportunity to take timely action to eliminate the cause of the defect within the process in order to minimise costs
instead of remedying the consequences of the defect.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 19
Number 3
September 2025
DOI: 10.12716/1001.19.03.37
1018
environment that surrounded the metal. Nowadays,
paints are much more than that [7,8], they are created
as electrical insulators that form a layer of electrical
resistance, they also contain soluble pigments that are
used to passivate the metal surface, and finally, paints
today serve as an additional anode for the dissolution
process [8].
The above shows the progress that has been made
recently in the protection of materials by coatings. All
this has been accompanied by corresponding
recommendations, rules and regulations [9,10], which
have also changed over time. These conditions have an
impact on the present research, which is carried out on
a submerged object subject to these strict control
requirements.
The preparation of the area to be painted and the
process of painting steel objects submerged in the sea
depend on many different factors. They are illustrated
in the fishbone diagram (Ishikawa diagram) in
Figure 1.
Figure 1. Paint preparation and application diagram
One of the aims of this research work is to
demonstrate a fast, high-quality solution that is easy to
apply under real conditions using control charts.
Control charts were first developed and applied by
Walter Shewhart [11], who used them to monitor the
current state of the process and to predict future
process phases (published in 1931 in the book
"Economic Control of Quality of Manufactured
Product") [12]. These charts are still referred to as
Shewhart charts. After him, the development of
statistical control was continued by authors like
Deming [13], Juran [14], Ishikawa [15] and others up to
the present day. This method was developed and
linked to statistical software packages that enabled the
selection, creation and analysis of control charts from
measurement data, such as SPC (Statistical Process
Control), DMAIC (Define, Measure, Analyze, Improve
and Control) [16], and QI MACROS (EXEL support)
[17]. Unlike laboratory measurements where the
conditions are controlled, statistical control in control
charts takes place due to its simplicity and precision in
representing the data during the measurement process
of the actual conditions on the measured object, which
allows control and correction within the specified
limits.
2 CONDITION ASSESSMENT OF EXISTING
SURFACE COATING SYSTEM
Before deciding on the necessary surface repairs, it is
essential to assess the condition of the surfaces. This is
usually done based on experience and with the help of
the manual for assessing the condition of hard coatings
[18]. For the assessments of the „degree of
effectiveness“ of an existing surface coating, it is
suggested that the following „rating“ be used
(explained in Table 1):
− GOOD condition with only minor spot rust.
− FAIR condition with local breakdown at the edges
of stiffeners and welded joints and/or slight rust on
20% or more of the surfaces considered, but less
than defined for POOR condition.
− POOR condition with general breakdown of coating
on 20% or more of the surfaces in question or with
hard scaling on 10% or more of the surfaces in
question.
Table 1. Definition of coating condition [19]
Rating / Condition
Good
Fair
Poor
Spot rust
Light rust
Minor
Minor
> 20%
Edges, Weld
< 20%
> 20%
Hard scale
Minor
< 10%
> 10%
General breakdown
Minor
< 20%
> 20%
Other references
ISO
RI3
RI4
RI5
European Rust Scale
RE3
RE5
RE7
Note: The lowest rating within any category shall govern the final
rating.
An example of an „Assessment Scale for
Breakdown“ of coatings is shown in the following
figures. The condition of the coating should normally
be assessed over large areas. Figure 2 shows a coating
in good condition. The condition shown corresponds
to the criteria given in Table 2, i.e. the criteria for the
evaluation of this example.
Figure 2. Coating in GOOD condition [20]
Table 2. Explanation of Coating condition GOOD [20]
Notes:
1. Minor rusting on weld seams.
2. Spot rusting.
3. Filmy deposit mush of surface.
Assessment scale
Less than 1%
Figure 3 shows a coating in fair condition. The
condition shown corresponds to the criteria given in
1019
Table 3, i.e. the criteria for the evaluation of this
example.
Figure 3. Coating in FAIR condition [20]
Table 3. Explanation of Coating condition FAIR [20]
Notes:
1. Anode working
2. White deposits 3%
3. Corrosion on edges
4. Top coat loss.
Assessment scale
5% 10%
The poor coating condition is described in the same
way as the two previous conditions. Figure 4 shows a
coating in poor condition. This condition is described
in detail in Table 4, i.e. the criteria for evaluating this
example are listed there.
Figure 4. Coating in POOR condition [20]
Table 4. Explanation of Coating condition POOR [20]
Notes:
l. Corrosion >20%
2. Hard scale >10%
3. Deformed stiffener edges
Assessment scale
20% 10%
Once the condition of the paint has been checked
against the above criteria, the surfaces to be painted are
inspected and measured.
3 MEASUREMENTS AND ANALYSIS
DESCRIPTION
The measurements were carried out on the surface of
the object in two runs, before and after painting. The
measurements are listed in tables and are used for
statistical analysis and the creation of control charts.
The statistical methods include the following steps:
Measurements, data processing and presentation,
analyses and interpretations of the values obtained.
These measures help to arrive at a decision on the
potential process capability (Cp) and also indicate by
how much the tolerance limits exceed the actual
distribution limits and whether further improvements
are required.
The potential Cp is assessed on the basis shown in
Table 5.
Table 5. Evaluation of potential process capability
USL-LSL
=
6
Cp
σ
Potential process capability
> 1.33Cp
Process has potential to be capable
1.0 < > 1.33Cp
Possible capability is questionable, and the
process should be monitored
< 1.00Cp
Very questionable potential process capability
USL-X
=
3
U
Cp
σ
Upper specification limit of potential process
capability
X-LSL
=
3
L
Cp
σ
Lower specification limit of potential process
capability
* σ - Process standard deviation, a measure of process variability
According to the data presented, Cp can be assessed
as capable, with questionable capability and with very
questionable capability. The process tolerance shows
how the measured values fulfil the standards
(tolerances) by which the process is determined.
Table 6. Evaluation of process tolerances
= min ( , )
k U L
Cp Cp Cp
Demonstrated excellence
> 1
k
Cp
Process tolerance within limits
= 0
k
Cp
Mean value is equal to one of the tolerance
limits
0 < < 1
k
Cp
Process tolerance exceeds limits
= 0; =
k
k Cp Cp
Process is perfectly centered
The Cpk value according to Table 6 indicates how
centred the process is, i.e. the position of the process
(measured values) in the tolerance field, and shows the
accuracy of the process by monitoring the lowest value.
The average of the mean values is often not centred,
which is why the capability index is used to indicate
the position of the values, i.e. the mean value and the
process deviation. The correlation between Cp and Cpk
can be expressed mathematically as follows:
(1- )=
k
Cp Cp k
(1)
The value “k“ indicates the shift of the process, i.e.
its values, from the centre when the process (the
measured values) is perfectly centred.
1020
2
; 0 1
2
−
−
=
−
USL LSL
X
kk
USL LSL
(2)
The upper specification limit (USL) and the lower
specification limit (LSL) are determined by the ISO
standard, while the lower control limit (LCL) and
upper control limit (UCL) are calculated within an
interval of ±3σ. The control limits are not related to the
tolerance limits as they are determined by the process
itself. The standard deviation is used as a measure of
the amount of dispersion in control charts. The values
obtained through data analysis are as follows:
− The upper specification limit (USL),
− The lower specification limit (LSL),
− Average, i.e. mean value,
− Standard deviation (Stdev).
The process is kept within the defined limits by
constant monitoring. For this reason, control charts are
used to represent data and are part of the statistical
control process as well as an efficient means of
obtaining information and making a quick and high-
quality decision regarding the process. If the value
rises or falls seven points in a row, a trend is created
and the process is not stable, even if all the data is
within the upper and lower limits. If the value rises or
falls by five consecutive points, resulting in a trend, the
process is in a critical state but is still stable. An
increased critical state signals the transition of the
process to an unstable state. Two or three points above
2σ should be a warning if the values are within the
control limits. This is due to a change in equipment,
measurement procedure or method. A stable process is
one in which all results, or at least a satisfactory
number of results, are within the control limits. To
ensure high quality and accurate monitoring of all
stages of the process, three statistical methods were
used: control charts (X and mR), histograms and
mathematical distribution analysis.
4 MEASUREMENTS AND ANALYSIS PRIOR TO
PAINTING
Control measurements before painting are used to
determine whether the surface is prepared to meet the
requirements. The process begins with fresh water
cleaning, degreasing and sandblasting (or hydro
blasting) to achieve a surface cleanliness level of Sa3 in
accordance with the international standard ISO 8501-
1:2007 [21]. After cleaning, the surface is protected by a
two-component zinc silicate coating, for which there
are many different manufacturers on the market today.
All equipment on the object, whether welded (beams,
clamps) or connected (pipes), must be blasted to the Sa
2.5 level. Control measurements are then carried out,
which are listed in Table 7.
Table 7. Control measurements before painting
No
ROUGHNESS
Rz (µm)
SALT CONCENTRATION
(mg/m
2
)
1
82.55
21.525
2
83.55
21.700
3
76.75
21.925
4
85
18.375
5
84.7
18.175
6
80.75
17.850
7
78.9
26.800
8
83.6
26.775
9
84.55
16.975
10
74.8
16.950
11
82.1
23.900
12
85
23.900
13
71.2
14.775
14
84.75
14.800
15
83.5
14.900
16
84.25
19.975
17
77.25
19.950
18
75.75
23.925
19
81.25
24.050
20
78.5
24.075
These measurements are used to check the
condition of the surface before applying protective
coatings. Two main areas of the control are roughness
control and analysis and salt concentration analysis.
4.1 Roughness control
Microscopic roughness is an irregular surface caused
by the treatment of a material. Since surface roughness
accelerates the corrosion process, the surface should be
protected by a coating. Before applying a protective
coating, the surface must be prepared, i.e. it must be
free of: rust, scale, dust, salts and fats. The assessment
is carried out in accordance with ISO standards 8501
[22], 8502 [23,24] and 8503 [25], while ISO 8504 [26]
contains guidelines for the preparation of steel
surfaces, i.e. for achieving a certain degree of
cleanliness. After cleaning, the surface is classified as
follows: fine, medium and rough. These gradations
represent different degrees of roughness, depending
on the required quality. The individual roughness
grades are defined in ISO 8503, while ISO 8503-1 [25]
specifies the measuring device for measuring the
surface. Figure 5 shows the surface after sandblasting
the inner surface and before applying the protective
coating.
Figure 5. Sand blasting of the inner surface [16]
When sandblasting, the abrasive must be dry and
clean and must not be contaminated, as this would
jeopardize the quality of the surface before the paint is
applied. When using abrasives, i.e. sandblasting, the
1021
size of the particles creates a roughness of at least 30
µm to a maximum of 85 µm. According to the ISO 8503
standard [25], these are acceptable roughness limits.
Using the data from Table 7 and the QI Macros
program, Figure 6 is created.
Figure 6. X chart of the roughness mean value
Figure 6 is showing an X-chart of the roughness
value. i.e. the mean value. It is consistent across all
measurements and is within the UCL and LCL (upper
and lower control limits) for all measurements. The
same sources and methods are used to create Figure 7,
which shows the mR chart of the roughness mean
value. The moving range of consecutive observations
shows that there are no significant deviations in surface
roughness. All measured data are within the limit
values and no trends can be identified.
Figure 7. mR chart of the roughness mean value
The mR chart shows only two points above the
average value, but still within the limit values. The
process is therefore stable.
The regression analysis of the roughness
measurements is shown in Figure 8, from which it can
be seen that the roughness measurements correspond
to the probability diagram of the normal distribution.
Figure 8. Regression analysis of roughness measurements
Figure 8 shows that the roughness measurements
are consistent with the probability plot for the normal
distribution. The probability plot for the normal
distribution shows a strong linear positive correlation,
represented by the line Y=0.231X-18.72. There are only
minimal deviations from the regression line, which
leads to the conclusion that the normal distribution is a
well-chosen model for the roughness measurement.
The coefficient of determination shows that 88.6% of all
deviations are interpreted by the linear regression
model, so that the correlation is very well interpreted
by the regression. This confirms that the roughness
model is representative.
The standard value Z in Figure 8 indicates the
relative position of the data and is calculated according
to Equation 3.
−
=
XX
Z
(3)
The estimation of the potential process capability Cp
and the demonstrated excellence Cpk, which were
determined using statistical analysis and the QI Macros
programme package for roughness measurements, are
shown in Table 8.
Table 8. Control measurements before painting
USL
85
Upper specification limit
LSL
30
Lower specification limit
Average
80.935
Arithmetic mean
Cp
2.27
Process is capable
CpU
0.33
Upper potential process capability
CpL
4.20
Lower potential process capability
Cpk
0.33
Process tolerance exceeds limits
Stdev
4.05
Standard deviation
The results of the roughness measurement are
shown in the histogram and in the normal distribution,
as can be seen in Figure 9.
Figure 9. Roughness measurement histogram
The histogram and normal distribution for the
roughness measurements in Figure 9 show that the
values are close to the upper limit given in Table 8.
4.2 Dust and salt concentration
Rust, salt, dust, fats and other contaminants that fall on
or come into contact with the surface affect the
cleanliness of the surface before the paint is applied.
Such surfaces are cleaned by sandblasting and
degreasing. If dust particles remain on the surface and
a coat of paint is applied, cracking and subsequent
corrosion can occur. For this reason, it is extremely
important to clean and inspect the surface as well as
possible to ensure optimum contact between the paint
and the surface. The rules for monitoring the surface
for salt and dust are laid down in ISO 8502 [23,24].
According to the ISO 8502-3 [24] standard, the dust on
the surface is monitored and graded from 1 to 5. The
maximum permissible amount of dust on a surface is
2. If this limit is exceeded, the surface must be cleaned
additionally and measurements must be taken until a
satisfactory level is reached.
1022
The actual dust and salt measurement for this
analysis is carried out on a steel surface before
applying a coat of paint using a pressure-sensitive
adhesive tape, which is shown in Figure 10.
Figure 10. Dust measurement
After removing the adhesive tape, a visual
inspection is carried out and compared with the etalon.
The standards ISO 8502-6 [27] and ISO 8502-9 [28]
describe the procedure for measuring the salt
concentration on a surface, whereby the preparation of
the measuring device is carried out in accordance with
ISO 8502-6 [27]. The surface density of the salt is
calculated according to the ISO 8502-9 [28] standard.
As already mentioned, the measurements are
carried out in accordance with ISO standards 8502-6
and 8502-9, which stipulate that the chlorine salt
content must not exceed 30 mg/m2. As shown in Figure
11, the “Bresle Method” is used for measurement and
the value of soluble salts is 27.6 mg/m2. This value is
within the limit values according to ISO standards
8502-6 and 8502-9 [27,28].
Figure 11. Salt measurement
In this case, as in the previous analysis, the data
from Table 7 were used to generate control charts X,
mR, probability paper, histogram and normal
distribution curve for the measured salt concentration
values using the QI Macros programme package.
Figure 12 shows considerable deviations in the salt
concentration measurements.
Figure 12. X chart of the salt concentration mean value
All of the measured data on the X-chart are within
the upper UCL and lower LCL limits, but with the
exception of two values that are above the UCL and
three values that are below the LCL, there is no trend
in the measured data obtained.
Since the measured values do not exceed the limits
specified in the ISO standard, it can be concluded that
the system should be closely monitored, which is
confirmed by the mR diagram shown in Figure 13.
Figure 13. mR chart of the salt concentration mean value
The figure shows that only four values are above the
upper UCL limit. These results indicate that the process
needs to be improved, i.e. the surface should be better
cleaned of salt or better protected.
The regression analysis of the salt concentration
data on the probability plot for normal distribution is
shown in Figure 14.
Figure 14. Regression analysis of salt concentration
It shows that the surface salt concentration
measurements fit best with the probability plot for
normal distribution. This normal distribution
probability plot shows a strong linear positive
correlation, represented by the line Y=0.251X-5.162.
There are only minimal deviations from the regression
line, which leads to the conclusion that the normal
distribution is a well-chosen model for roughness
measurement. The coefficient of determination shows
that 94.9% of all deviations are interpreted by the linear
regression model, so that the correlation is very well
interpreted by the regression.
1023
This confirms that the surface salt concentration
model is representative. Overall salt concentration
measurement analysis results are given in Table 9.
Table 9. Salt concentration measurement analysis results
USL
30
Upper specification limit
LSL
0
Lower specification limit
Average
20.565
Arithmetic mean
Cp
1.30
Potential capability is questionable, and the
process should be monitored
CpU
0.82
Upper potential process capability
CpL
1.78
Lower potential process capability
Cpk
0.82
Process tolerance exceeds limits
Stdev
3.86
Standard deviation
The measurements of the surface salt concentration
are shown in the histogram and in the normal
distribution and reproduced in Figure 15.
Figure 15. Salt concentration measurement histogram
It can be seen that the values are close to the upper
limit, which are listed in Table 9.
5 PAINTING
Once the surface preparation and cleaning have been
completed, the surface is ready for the paint
application, as shown in Figure 16. The paint must be
applied in accordance with the paint manufacturer's
recommendations, which comply with the ISO 2808
[29] standard, and the requirements for weather
conditions must be strictly observed.
If the paint is applied within 8 hours of surface
preparation, the following conditions must be met: The
temperature difference on the steel surface must be
within the limits, i.e. between the minimum and
maximum values, at least 3°C above the dew point and
a maximum relative humidity of 85% [29].
Figure 16. Surface before paint application
If the coating is applied more than 8 hours after
surface preparation, the temperature difference on the
steel surface must be between the minimum and
maximum values, at least 5°C above the dew point and
a maximum relative humidity of 40% (maximum 50 %)
[29]. The atmospheric conditions must be measured
during and after the application of the paint, which is
why the object must be protected. Figure 17 shows the
same surface as in Figure 16, this time after the paint
has been applied in a protected environment.
Figure 17. Surface after paint application
Before, during and after the application of the paint,
the parameters of the paint manufacturer and the ISO
2808 standard are checked by continuous condition
monitoring.
6 MEASUREMENTS AND ANALYSIS AFTER
PAINTING
The measuring devices must be tested on the test plate
before the procedure, using the etalon that comes
closest to the specified DFD (dry film density). The
quality of the measurement depends on the following
factors:
− Calibration of the measuring instrument,
− Testing the measuring instrument before each
measurement,
− Additional adjustment of the measuring
instrument, if necessary,
− Expertise of the operator.
The measurement data used for this research was
obtained by measuring the coating thickness using the
microscopic method, which comprises three
procedures: A, B, C.
− Procedure A is a general method suitable for
measuring variations in coating thickness on an
uneven surface [30].
− Procedure B is used to measure film thickness above
2µm and on solid surfaces, as the film is cut at a
specific angle [30].
− Procedure C uses a special microscope with an
additional device to monitor the surface profile of
the sample, and the method is performed on the
part that is exposed to sufficient light to obtain a
clear image in the microscope [30, 31], as shown if
Figures 18 and 19.
1024
Figure 18. Vertical coating thickness measurement
At the same time, care must be taken to ensure that
the number of samples is representative for this
method. The procedure is described in detail in ISO
standard 1463 [19].
When measuring, attention should be paid to
coatings with high elasticity, as the elasticity influences
the accuracy of the measurement results. To
successfully measure the thickness of opaque coatings,
a small area of the coating must be removed.
Figure 19. Horizontal coating thickness measurement
The difference between the surface of the coating
and the surface of the object leads to a deflection of the
light beam and thus to an absolute measurement of the
coating thickness. The measured values are listed in
Table 10.
Table 10. Control measurements before painting
No
ROUGHNESS
Rz (µm)
SALT CONCENTRATION
(mg/m
2
)
1
82.55
21.525
2
83.55
21.700
3
76.75
21.925
4
85
18.375
5
84.7
18.175
6
80.75
17.850
7
78.9
26.800
8
83.6
26.775
9
84.55
16.975
10
74.8
16.950
The data measured after painting, shown in Table 6,
were used to obtain control charts X, mR, probability
paper, histogram and normal distribution curve for the
measured coating thickness values using the QI
Macros programme package. This is done according to
the same principles as the analysis of the data obtained
before painting.
Figure 20 shows the deviations in the layer
thickness measurements. All measured data on the X-
chart lie within the upper UCL and lower LCL limits,
with the exception of one value that lies below the LCL.
Figure 20. X chart of the coating thickness mean value
The mR chart shown in Figure 21 shows five points
that form a trend indicating a critical condition, i.e.
special care must be taken when applying the coating.
Figure 21. mR chart of the coating thickness mean value
Since the measured values do not exceed the limits
specified in the ISO standard, it can be concluded that
the system should be closely monitored and the
process of applying paint to the section surface should
be improved.
The regression analysis of the coating thickness data
on the probability plot for normal distribution is shown
in Figure 22. It shows that the measurements of the
coating thickness correspond best to the probability
diagram of the normal distribution. This probability
plot for the normal distribution shows a strong linear
positive correlation, represented by the line Y=0.013X-
7.458. There are only minimal deviations from the
regression line, which leads to the conclusion that the
normal distribution is a well-chosen model for the
coating thickness measurement.
Figure 22. Regression analysis of coating thickness
The coefficient of determination shows that 91.3 %
of all deviations are interpreted by the linear regression
model, so that the correlation is interpreted very well
by the regression. This shows that the coating thickness
model is representative. Overall coating thickness
measurement analysis results are given in Table 11.
1025
Table 11. Control measurements before painting
USL
625
Upper specification limit
LSL
400
Lower specification limit
Average
546.055
Arithmetic mean
Cp
0.54
Very questionable potential process capability
CpU
0.38
Upper potential process capability
CpL
0.70
Lower potential process capability
Cpk
0.38
Process tolerance exceeds limits
Stdev
69.57
Standard deviation
The measurements of the coating thickness are
shown in the histogram and in the normal distribution
and reproduced in Figure 23.
Figure 23. Coating thickness measurement histogram
Figure 23 shows the histogram and the normal
distribution for the coating thickness measurements,
clearly showing that the values are close to the upper
limit given in Table 11.
7 CONCLUSIONS
In this work, an attempt was made to improve the
quality of the preparation and the monitoring of the
parameters that influence the quality of the painting
process. To achieve this goal, a statistical analysis was
used to allow continuous monitoring and
improvement of the process. The control charts and the
QI Macros software were the basic statistical tools
used. As part of the statistical analysis, the control
charts track whether the data complies with ISO
standards and indicate corrective actions if it is not
within the limits. This is a great help to the
manufacturer as it improves the monitoring process
and the quality of production. At the same time, the
ISO standards and the rules and regulations of the
classification societies are met. The example used in
this research shows that the control charts indicate that
the salt concentration on the surface must be closely
monitored and that the surface should be checked and
cleaned more frequently. At the same time, the
roughness and coating thickness are within the
specified limits.
In summary, the control charts provide the
opportunity to take timely action to eliminate the cause
of the defect within the process in order to minimise
costs instead of remedying the consequences of the
defect.
REFERENCES
[1] NACE International/ASTM International, Standard
Terminology and Acronyms Relating to Corrosion.
ASTM International, West Conshohocken, USA,2020
[2] J. Babicz, Wärtsilä Encyclopedia of Ship Technology, 2nd
ed., Helsinki: Wärtsilä corporation, 2015. ISBN 978-952-
93-5536-5,
[3] J.C. Kurth, P.D. Krauss and S.W.Foster, “Corrosion
Management of Maritime Infrastructure” Transportation
Research Record: Journal of the Transportation Research
Board, 2019. doi: 10.1177/0361198119855333
[4] C.G. Soares, Y. Garbatov, A. Zayed and G. Wang,
“Influence of environmental factors on corrosion of ship
structures in marine atmosphere” Corrosion Science,
51(9), 2014,. doi: 10.1016/j.corsci.2009.05.028
[5] K.A. Chandler, Marine and offshore corrosion, London:
Butterworths, 1985. doi: 10.1016/C2013-0-06267-6
[6] J.E.O. Mayne, “Corrosion Prevention by Paints”,
Transactions of the IMF, 29(1) 146-163, 1985. doi:
10.1080/00202967.1952.11869596
[7] M. Kendig and D.J. Mills “An historical perspective on the
corrosion protection by paints”, Progress in Organic
Coatings, 102, 53–59, 2017.
doi:10.1016/j.porgcoat.2016.04.044
[8] O. Ø. Knudsen and A. Forsgren. Corrosion control
through organic coatings, CRC press, 2017. doi:
10.1201/9781315153186
[9] American Bureau of Shipping. Guidance notes on the
application and inspection of marine coating systems,
ABS. USA, 2017.
[10] Performance standard for protective coatings for
dedicated seawater ballast tanks in all types of ships and
double-side skin spaces of bulk carriers, IMO Resolution
MSC.215, 2006.
[11] W.A. Shewhart, ”Quality Control Charts” The Bell
System Technical Journal, 5(4), 593-603, 1926. doi:
10.1002/j.1538-7305.1926.tb00125.x
[12] W.A. Shewhart, The Economic Control of Manufactured
Product, van Nostrand Company, New Jersey, USA. 1931
[13] W.E. Deming, ”On Some Statistical Aids Toward
Economic Production” Interfaces, 5(4), 1–15, 1975, doi:
10.1287/inte.5.4.1
[14] J.M. Juran, Juran's quality handbook, New York USA:
McGraw-Hill, ISBN 0-07-034003-X, 1999
[15] K. Ishikawa, “The quality control audit”, Quality
Progress, 20, 39-41, 1987
[16] L.M. Gaikwad, V.K. Sunnapwar, S.N. Teli and A.B.
Parab, “Application of DMAIC and SPC to Improve
Operational Performance of Manufacturing Industry: A
Case Study”, Journal of The Institution of Engineers
(India): Series C. 100; 229-238. 2019, doi: 10.1007/s40032-
017-0395-5
[17] L. Pimentel and F. Barrueto, “Statistical Process Control:
Separating Signal from Noise in Emergency Department
Operations”, The Journal of Emergency Medicine, 48(5)
628–638. 2015. doi: 10.1016/j.jemermed.2014.12.019
[18] American Bureau of Shipping, Guidance Notes on the
Inspection, Maintenance and Application of Marine
Coating Systems, ABS. USA, 2007.
[19] Metallic and oxide coatings, ISO standard 1463, 2021.
[20] American Bureau of Shipping, Guidance Notes on the
Application and Maintenance of Marine Coating
Systems, ABS. USA, 1998
[21] Rust grades and preparation grades of uncoated steel
substrates and of steel substrates after overall removal of
previous coatings, ISO standard 8501-1, 2007
[22] Preparation of steel substrates before application of
paints and related products — Visual assessment of
surface cleanliness, ISO standard 8501, 2007
[23] Laboratory determination of chloride on cleaned
surfaces, ISO standard 8502-2, 2017
[24] Assessment of dust on steel surfaces prepared for
painting — Tests for the assessment of surface cleanliness,
ISO standard 8502-3, 2017
1026
[25] Preparation of steel substrates before application of
paints and related products — Surface roughness
characteristics of blast-cleaned steel substrates, ISO
standard 8503, 2012
[26] Preparation of steel substrates before application of
paints and related products — Surface preparation
methods, ISO standard 8504, 2019
[27] Extraction of water soluble contaminants for analysis,
ISO standard 8502-6, 2020
[28] Field method for the conductometric determination of
water-soluble salts, ISO standard 8502-9, 2020
[29] Paints and varnishes — Determination of film thickness,
ISO standard 2808, 2019
[30] H. Plog, and C.E. Crosby, Guide to Methods and
Instruments for Coating Thickness Measurement.
Teddington, UK: Robert Draper, Ltd., 1971
[31] R. Kohli. “Methods for monitoring and measuring
cleanliness of surfaces,” in Developments in surface
contamination and cleaning, R. Kohli and K.L. Mittal,
Eds. William Andrew Publishing, 2012, pp. 107-178.
doi:10.1016/B978-1-4377-7883-0.00003-1
