Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.

com, ISSN 1743-355X

A combined design methodology for impressed

current cathodic protection systems
V.G. DeGiorgi," E.D. Thomas II,* K.E. Lucas,* A. Kee"
"Mechanics ofMaterials Branch, Code 6382, Naval Research

*Environmental Effects Branch, Code 6310, Naval Research

Za6orafor% %%j/zmgfoM, DC2037J, [V&4
'Geo-Centers Incorporation, Fort Washington, MD 20744, USA

Abstract

The corrosion prevention capability of shipboard impressed

current cathodic protection systems are effected by the
geometry of the ship hull, conductivity of the surrounding
water, material polarization response and any material
interactions. Changes in system configuration and service
environment from those used in design analysis require the
capability to analyze system performance under varied
circumstances. While individually limited, experimental
scale modeling and computational boundary element
modeling techniques can be used in concert to design a
robust system and to provide a means for quick analysis of
system and environmental changes. A methodology for
system design based on these two analysis techniques is
presented.

1 Introduction

Impressed current cathodic protection (ICCP) systems take

advantage of natural electrochemcial reactions of materials
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X

336 Boundary Element Technology

to minimize corrosion damage. In an ICCP system, an

external source of electrons is provided to the
metal/electrolyte combination, for example the hull of a
ship in the ocean. In order to achieve protection from
corrosion the source of electrons must be sufficient to raise
the potential of the structure to a level at which negligible
corrosion occurs. Major factors which must be considered
in the design of such systems are the geometry, the
conductivity of the electrolyte, the polarization behavior of
the metal, the location and polarization of any other
exposed metal and the magnitude of the power source
supplying the electrons. The use of shipboard ICCP systems
is well established. In the past decade considerable effort
has been spent in the development of design methodologies
which have a scientific rather than designer expertise basis.
Two such methodologies are physical scale modeling and
computational modeling using boundary element
techniques. Physical scale modeling has been used to
design ICCP systems which have been installed on U S Navy
ships. However, it is felt that computational modeling
capability is still required. Failure of individual
components in an system or unexpected damage require
the capability to quickly evaluate the system. Frequently
re-evaluation is required in a short period of time and the
time required for experimental evaluation is not available.

2 Scope of Work

The purpose of this paper is to review the status,

advantages and limitations of physical scale modeling and
computational modeling as applied to shipboard ICCP
systems. The verification work performed for physical
scale modeling and computational modeling is briefly
highlighted. A unique combined computational and
experimental design methodology is presented. This
method which combines the advantages of the individual
methods should result in a more robust system design than
could be created independently by either method.
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X

Boundary Element Technology 337

3 Physical Scale Modeling

Physical scale modeling is a experimental technique in

which the linear dimension of the structure and the
conductivity of the electrolyte are scaled by the same
factor. Ditchfield et al [1] examined the theoretical
considerations of this mechanical scaling factor. Based on
modeling of electrochemical cells the physical scale model
and full size structure must maintain:

(1) identical current density values at points

(2) identical potential differences at points
(3) identical polarization potentials at the
anode and cathode
(4) identical potential drop across the
electrolyte.

Based on theoretical considerations, a correct scaled model

results in:

(1) measured potential values identical to

those found in the full size structures
(2) values under flow conditions can be
obtained from models by applying flow
velocities for the full size structure (flow
velocities are not scaled).
(3) measured electrical fields in the model are
higher than those found in the full size
structure but true current densities can be
calculated from model results.

Item (1), measured potentials equal to potentials on the full

size structure, is one of the most important features with
regards to applications of scale modeling techniques.
As an experimental process, physical scale modeling
was first introduced in 1985 for shipboard ICCP systems [2].
The scaling factor used is dependent on the model size and
magnitude of electrolyte conductivity which can be handled
in the experimental facility. Verification of the procedure
has included comparison with ship data for a British
destroyer [2] and U S Navy CG class system [3,4]. Scaling
factors of 1/60, 1/96 and 1/100 have been used for
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X

338 Boundary Element Technology

shipboard systems. Work at the Naval Research Laboratory

concentrates on 1/96 scaled models and electrolyte.
Physical scale modeling techniques were used to
design a two power zone system for the USS Princeton, CG-
59 [3]. The ICCP system is a modification of the previously
installed system on the CG class. Prior to sea trials, the USS
Princeton was modified for data gathering. Additional
instrumentation and an incomplete coating system were
installed. The incomplete coating system resulted in a hull
condition between minimum and maximum damage
definitions. The USS Princeton was later evaluated
computationally. Physical scale and computational modeling
were performed independently. As can be seen in Figure 1,
good agreement was achieved between full size ship data,
physical scale modeling data and computational results.
In summary, important advantages of physical scale
modeling include:

(1) ability to duplicate complex geometries

(2) ability to incorporate actual materials and
material combinations to capture any galvanic
interactions
(3) lack of requirement of polarization response
data for the materials and material combinations
in use

2.0^
1.8- • Ship Data - CG59
31.6- # PSM-3% Distal
A BEM-3%

&1-2H
01.0-
0.8*'
0.6-j—
0.0 100.0300.0 400.0
200.0 500.0 600.0
Frame
Figure la - Static flow and minimum damage conditions.
Measurements taken at a depth of 10 feet.
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X
Boundary Element Technology 339

2.0-,
• Ship Data - CG59
1.8-
• PSM-15%Distal
§1.6-1 A BEM-15%
D)

&1-2H
.§ 1.0-
0.8-
0.6.
0.0 300.0
100.0 400.0 500.0
200.0 600.0
Frame
Figure Ib - Static flow and maximum damage conditions.
Measurements taken at a depth of 10 feet.

4 Computational Modeling

The use of boundary element methods to model corrosion

systems which are governed by LaPlace's equation is well
established [5]. The requirements for applicability of
LaPlace's equation are an electroneutral electrolyte, an
absence of polarization concentration gradients and an
absence of electron sources or sinks. Shipboard ICCP
systems can be mathematically defined to meet these
requirements. Preliminary analysis by DeGiorgi et al [6]
and Zamani [7] validated the use of boundary element
methods for the design and analysis of shipboard ICCP
systems. Zamani analyzed a prototypic Canadian ship
modeling the propeller and rudder as equivalent cathodic
areas on the ship hull. DeGiorgi et al analyzed U S Navy
ships modeling the propeller and rudder as stylized but
distinct appendages to the hull. Zamani and Chuang [8]
used boundary element methods to determine the optimum
control system for a shipboard ICCP system. Trevelyan and
Hack [9] have done preliminary work on the incorporation
of stray current into a boundary element code.
General information learned about computational
modeling based on series of verification analyses
performed at the Naval Research Laboratory are
summarized here. The commercial boundary element code
BEASY-CP [10] was used for all work however the
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X

340 Boundary Element Technology

methodology presented should not be limited to this code.

Minimum and maximum paint damage conditions were
examined for static and dynamic flow conditions. Damage
patterns were supplied to the analyst and were developed
based on dry-dock information from shipyards.
Computational results which were compared in detail with
physical scale modeling results included potential profiles,
total current and current distribution to selected
components. Good agreement was obtained for the CG class
hull analysis. Representative comparisons for the CG class
between computational and physical scale model data are
shown in Figure 1.
As part of the analysis of the CG class hull, a mesh
refinement study was completed. It was determined that a
relative high level of refinement was required for accurate
results. It was found that geometric modeling of the bilge
keel was required for the generation of an accurate
potential map. A strong correlation between the accuracy
of the computational results and the accuracy of the
polarization response used in the analysis was observed.
The ability to obtain accurate and appropriate data is
complicated by the sensitivity of polarization response to
environmental factors and testing procedures [11,12].
In a design analysis, less than accurate polarization
response may be used to develop potential plots which will
identify 'good' and 'bad' areas of protection. While the
absolute magnitude of the results may not be correct, the
trends shown are correct. The designer can use these
results to determine the number and placement of anodes
required to provide protection.
An analysis of U S Navy aircraft carrier CVN hull class
was performed to determine the effects of scale on
modeling parameters determined by the CG hull class
analyses. Scale is addressed directly by the increased size
of the CVN, approximately 4 times the surface area as the
CG. Scale is addressed by the inclusion of relatively small
features such as the bilge keel on the CVN model.
Representative comparisons between physical scale and
computational results are shown in Figure 2.
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X
Boundary Element Technology 341

2.5-1

2.0-

• •

0.5.
0.0 50.0 150.0 200.0
100.0 250.0 300.0
Frame
Figure 2a: Dynamic flow and minimum damage conditions.
Measurements taken at a depth of 10 feet.

5.5-,

4.5- # PSM - Run2xx

A BEM Results
0)3.5-

J2.5-

.*
V
0.5
0.0 50.0 150.0 200.0
100.0 250.0 300.0
Frame
Figure 2b: Dynamic flow and maximum damage conditions.
Measurements taken at a depth of 10 feet.

Good agreement was observed in most instances.

Differences between computational and experimental
results were determined to be the results of model
simplification or material characterization. Potential
contours along regions of small geometric detail indicated
that more refinement could improve results. The
polarization response does not incorporate material
interaction. The formation of a protective film, which has
been observed in testing, is not addressed in the
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X

342 Boundary Element Technology

polarization response. Effects of initial current level on

subsequent polarization response, an observed phenomenon
was not incorporated into the response used. The
polarization response at high current levels, such as
observed near anodes in the computer model, was not
incorporated into the response used. These issues are not
generally addressed in any laboratory testing program for
polarization response and may be difficult to quantify.
In summary, the advantages of computational
modeling include:

(1) ease of model manipulation for addition or

movement of anodes
(2) ability to examine potential maps of the
entire structure to determine under and over
protection regions based on a generic
polarization response
(3) the ability to quickly evaluate changes in
anode, reference cell or damage configurations.

5 Combined Design Methodology

The proposed combined computer-physical scale modeling

procedure is shown in Figure 3. A two phase methodology
is proposed in which computational modeling is used for
preliminary system design and evaluation while physical
scale modeling is used for final verification. Computational
modeling is used to evaluate multiple candidate ICCP
system designs. Changes in anode number and placement,
reference cell number and placement and power zone
definitions can be readily evaluated. The final design as
determined by computational analysis is then evaluated by
physical scale modeling. This eliminates uncertainty due to
polarization response definitions used in the computational
analysis. The uncertainty in polarization response is one of
most severe limitations of accurate computational
evaluations. In addition, a final check of the proposed ICCP
system design by physical scale modeling will identify any
inaccuracies which may have been inadvertently
introduced in the computer model by model simplification
or material characterization. Final anode placement,
reference cell placement and power supply magnitudes will
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X

Boundary Element Technology 343

be validated by examination of physical scale modeling

results. The combined procedure results in significant cost
and time savings with an increase in design confidence.

Phase I Phase II
Computational Physical Scale
Modeling Modeling

Create Boundary Use Best

Element Model Design from
Basic Geometry Phase I

Model Perform
Shakedown,
Experimental
Addition of
ICCP System Evaluation
h h
rep(sat as repe<at as
neeessary necessary
1 i
Analyze Modify
Proposed System as
System
Required

Figure 3: Combined experimental and computational design

methodology

6 Summary

Physical scale modeling can be an expensive and extremely

time consuming method to evaluate multiple ICCP system
designs. Computational modeling on the other hand is well
suited for the quick evaluation of multiple ICCP system
designs once the hull geometry has been defined.
Conversely the accuracy of computational modeling is
critically dependent on accurate characterizations of
polarization response while physical scale modeling does
not depend on the mathematical representation of
polarization response. Therefore while a comparison of
 Transactions on Modelling and Simulation vol 13, © 1996 WIT Press, www.witpress.com, ISSN 1743-355X

344 Boundary Element Technology

system attributes may be quickly obtained using

computational modeling, the absolute accuracy of results
may be limited. The accuracy of physical scale modeling
results is not affected by the ability to correctly define the
polarization response and in this respect the method is
more suited to determining absolute current requirements
than the computational modeling method. Combining the
two methods results in a more robust methodology with
fewer intrinsic limitations than either methodology when
considered separately.

References

1. Ditchfield, R. W., McGrath, J. N., and Tighe-Ford, D. J., /. of

Applied Electrochemistry, Vol. 25, 54-60, 1995.
2. McGrath, J. N., Tighe-Ford, D. J., and Hodgkiss, L.,
Corrosion Prevention and Control, April, 33-36, 1985.
3. Thomas, E. D. and Parks, A. R., Corrosion 59, Paper 274,
National Assc. of Corrosion Engineers, Houston, TX, 1989.
4. Parks, A. R., Thomas, E. D., and Lucas, K. E., Corrosion 90,
Paper 370, National Assc. of Corrosion Engineers, Houston,
TX, 1990.
5. Munn, R. S., Computer Modeling for Corrosion. ASTM STP-
1154, American Society for Testing and Materials,
Philadelphia, 215-228, 1992.
6. DeGiorgi, V. G., Kee, A., and Thomas, E. D. II, Boundary
Elements IV. Computational Mechanics Publications,
Southampton, England, 679-694, 1993.
7. Zamani, N. G., Applied Math, and Computation, Vol. 26, No.
2, 119-134, 1988.
8. Zamani, N. G. and Chuang, J. M., Optimal Control App. and
Mff/Wj, Vol. 8, 339-350, 1987.
9. Trevelyan, J. and Hack, H. P., Boundary Element
Technology IX. Computational Mechanics Publications,
Southampton, England, 347-356, 1994.
10. "BEASY-CP Users Manual," Computational Mechanics
International, Billerica, MA, 1990.
11. Jones, D. A., Principles and Prevention of Corrosion.
Maximilan Publishing Co., New York, 1992.
12. Hack, H. P. and Scully, J. R., Corrosion, Vol. 43, No. 3, 79-
90, 1986.