Predictive Modeling Galvanic Corrosion risk assess… · Predictive Modeling Until the end of the 60’s no significant attention was given to corrosion in the aircraft and many other

Corrosion Risk Assessment Using Computer Modeling

Page 1111 of 6666

In the product development process designs In the product development process designs In the product development process designs In the product development process designs are often exposed to an outdoor are often exposed to an outdoor are often exposed to an outdoor are often exposed to an outdoor environment for a long time, or corrosion environment for a long time, or corrosion environment for a long time, or corrosion environment for a long time, or corrosion acceleration tests are conducted to simulate acceleration tests are conducted to simulate acceleration tests are conducted to simulate acceleration tests are conducted to simulate the actual condition the vehicle or structuthe actual condition the vehicle or structuthe actual condition the vehicle or structuthe actual condition the vehicle or structure re re re will be exposed to during its life. These will be exposed to during its life. These will be exposed to during its life. These will be exposed to during its life. These methods however require several months to methods however require several months to methods however require several months to methods however require several months to years of test time to complete. As an years of test time to complete. As an years of test time to complete. As an years of test time to complete. As an alternative approach, the BEASY Corrosion alternative approach, the BEASY Corrosion alternative approach, the BEASY Corrosion alternative approach, the BEASY Corrosion Manager computer modeling tool has the Manager computer modeling tool has the Manager computer modeling tool has the Manager computer modeling tool has the potential to significantly shorten and reduce potential to significantly shorten and reduce potential to significantly shorten and reduce potential to significantly shorten and reduce the the the the cost of testing.cost of testing.cost of testing.cost of testing.

Predictive Modeling Until the end of the 60’s no significant attention was given to corrosion in the aircraft and many other manufacturing industries. Thereafter manufacturers and engineers became increasingly aware of the impact of different types of corrosion on the lifetime of the structures; while at present, in the United States alone costs estimated in the order of 2.2 billion dollars per year are reported as coming from the aircraft industry. Nowadays, different forms of corrosion (i.e. galvanic, pitting, crevice, inter-granular, etc.) are generally recognized as key factors in limiting the operational life time of the aircraft because of uneconomical high maintenance costs. The major contributor to corrosion damage is however galvanic corrosion.

In other industries such as automotive the drive towards lighter vehicles has required the substitution of steel by lighter materials such as aluminum and magnesium. However the combination of these materials when exposed to the harsh corrosive environment typically found in vehicles can lead to significant galvanic corrosion. In the product development process designs are often exposed to an outdoor environment for a long time, or corrosion acceleration tests are conducted to simulate the actual condition the vehicle will experience on the road. Computer modeling has the potential to significantly shorten and reduce the cost of testing by

substituting for long term exposure and corrosion acceleration testing.

In the aerospace industry the currently applied “find it – fix it” maintenance concept is a common approach against corrosion where corrosion findings have to be maintained independent of their impact. However, it is widely recognized that scheduled measures are more efficient in the medium to long terms. In this aspect the benefit of modeling for prediction and optimal design becomes clear. The “find and fix” approach must be complemented by an approach based on understanding of the corrosion process and the ability to predict its behaviors. The use of corrosion prediction models can lead to the development of a cost-efficient predictive corrosion integrity and maintenance programs.

Galvanic Corrosion Galvanic corrosion (GC) occurs when different materials are in contact with a common electrolyte. GC on its own can cause severe damage in a structure due to highly accelerated corrosion rates compared to other mechanisms. Usually GC can be avoided by proper material selection and appropriate corrosion protection measures. However, combinations of dissimilar materials are often applied due to structural requirements that need to be fulfilled in the design. In this case, costly corrosion prevention systems have to be applied to avoid the access of an electrolyte. Often, GC alone may not be directly the main cause of structural failures, but its occurrence may favour proper conditions for initiation of other types of corrosion such as pitting.

The fundamental principles of GC are quite well established and generally understood, but until now the general understanding gained has not been exploited enough to narrow the gap between scientific research and engineering applications. However the progressive advance of computational modeling in the last few decades has today made it possible to model a variety of complex corroding systems, thus representing a leading edge technology not only for research in the subject, but also for direct application in engineering

Corrosion Risk Assessment Using Computer Modeling

Page 2222 of 6666

design. In addition, recent advances in numerical methods have allowed the solution of increasingly larger and more complicated structures.

One of the general objectives of modeling GC is to tailor the corrosion protection measure in order to avoid GC for hybrid structures without performing a substantial amount of laboratory testing. A crucial aspect of the BEASY computational modeling for GC is its connection with reality and its reliability as a predictive tool. For this reason the modeling tool has been extensively validated for some commonly used engineering materials and environments and further developments are ongoing.

In the following sections the BEASY galvanic corrosion modeling system is presented capable of simulating cases ranging from structures completely submerged in an electrolyte to a thin film caused by moist air or spray. The validation procedure used to support the computational modeling capabilities is described for typical GC structure environments and finally some examples/case studies are presented.

Galvanic Corrosion Model The problem formulation for the electrolyte is based on the charge conservation equation in the bulk of the electrolyte under steady state conditions.

The mathematical description of the problem is based on the 3D Poisson equation for the electrolyte potential with non-linear boundary conditions imposed by the prescribed polarisation curves on the active electrodes. In the steady state case, the governing equation for the electrolyte becomes1:

; (1)

Where is the current density given by:

, σ represents the electrolyte

conductivity, and is the electric potential in the

electrolyte at point . The integration domain of this problem is the electrolyte, and the boundaries

are defined by all the surfaces surrounding it,

including the anode, cathode and the insulating walls.

The boundary conditions applied to the active electrodes are of the generic form:

(2)

where is the current density flowing throughout the

surface in normal direction ( ), and is the

1 ( )321 ,, xxx ∂∂∂∂∂∂=∇ represents the 3D

gradient operator.

polarization potential across the interface

metal/electrolyte given by , where Vm is

the potential in the metal.

This boundary condition is actually given by the corresponding polarization curves which result from the representative polarization curves characterization of the material. The function f, usually containing

exponential factors of as prescribed by Butler-

Volmer type equations, is in general non-linear.

ModelingModelingModelingModeling ApproachApproachApproachApproach

Electrolyte environments in structures can be complex ranging from thin films on the 3D aircraft structure to areas where the structure is submerged in an electrolyte pool. The optimum choice is a modeling strategy based on both the Boundary Element Method (BEM) and the Finite Element Method (FEM). Therefore in the model developed FEM is used in general to model the thin films (Figure 1) and BEM the deeper or bulk electrolyte situations.

Figure 1 Schematic representation of galvanic corrosion in the presence of thin film electrolytes, where w << L

Validation A number of cases have been studied to validate the model and to determine the procedures to obtain the model input data. In this example case a junction between an Aluminum 2024 panel and a Carbon Fiber Reinforced Plastic (CFRP) panel is investigated.

Figure 2 Mesh discretization of the bi-metallic GC model, viewed from below the electrodes

The electrode behavior of AA2024 unclad and CFRP are described by the polarization curves which serve as an input for the active electrodes in the model. Throughout the investigation for the anode, aluminum

0=⋅∇ j

j

)(xj eV∇−= σ

)(xeV

3R∈x Ω

Ω∂=Γ

)(ˆ

VfV

j en ∆=

∂

∂−=

nσ

nj

n V∆

me VVV −=∆

V∆

9.11.

Gap = 0.2

CFRP AA2024

12.3

12.3

26.2

Top View: All dimensions in centimetres

Corrosion Risk Assessment Using Computer Modeling

Page 3333 of 6666

sheet alloy AA2024 unclad with milled surface was used. For the cathode, a ground CFRP plate where the outer resin layer was removed by grinding down into the first carbon ply (~100 µm) to achieve an almost pure carbon surface.

ConductivityConductivityConductivityConductivity Chloride concentrationChloride concentrationChloride concentrationChloride concentration

5000 µS/cm ~ 0.05 M

10000 µS/cm ~ 0.10 M

Figure 3 Polarization curves of the CFRP and AA2024 samples used for the validation test. The potential is referred to the standard calomel electrode (SCE)

The experimental work consists of two stages. The first stage involves obtaining the electrochemical properties of the AA2024 and CFRP, by means of an electrochemical cell which is a standard three electrode cell. Figure 3 shows the polarization curves obtained through experiment for unclad AA2024 in electrolytes with different Chloride concentrations and conductivities. The different conductivities are corresponding with varying of the Chloride (Cl-) content:

A representative polarization curve measurement of CFRP (cathode) is shown (exemplarily 0.05 mol/l) in Figure 3 since here the curve is not sensitive to the electrolyte composition in the investigated range.

The results are based on three case scenarios, one for each chloride concentration value shown in the table. Figure 4 shows a comparison between the experimental and numerical modeling results for the three case scenarios. The vertical axis represents total current flowing between anode and cathode while the horizontal axis represents the electrolyte thickness in µm. The curves with continuous lines and markers represent experimental results; where each marker corresponds to one case scenario in the experiment. The curves without marker symbols represent

results coming from the numerical modeling. The agreement for all the test cases in the 4980 µS/cm and 9600 µS/cm electrolytes is good.

Figure 4. Comparison between experimental and modeling results for two case scenarios of conductivity, including 4980 µS/cm and 9600 µS/cm

Further publications describing the testing and modeling procedures for thin and deep electrolytes are described in the references.

Case Study - Bolted Connection A simple bolted connection between two plates is presented to demonstrate the modeling procedure and the types of information which can be obtained from the model. The geometry of the connection can be easily generated using a CAD system or created directly in the BEASY modeling software. The other information required is the polarization data for each of the materials (which can be obtained from polarization tests) and the electrolyte resistivity.

1,00E-07

1,00E-06

1,00E-05

1,00E-04

1,00E-03

1,00E-02

1,00E-01

1,00E+00

-1 -0,9 -0,8 -0,7 -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0

Potential [V vs SCE]

log

i[A

/cm

2]

CFRP-0.05M_5080 µS/cm

AA2024-0.05M_5000 µS/cm

AA2024-0.1M_10000 µS/cm

Total Current vs Electrolyte Thickness

0

50

100

150

200

250

0 5 10 15 20 25 30 35 40 45 50

Electrolyte Thickness [µm]

Avera

ge

To

tal C

urr

en

t [µ

A]

0,05M NaCl-Experiment

0,1M NaCl-Experiment

0.05 M NaCl-Model

0.1 M NaCl-Model

Corrosion Risk Assessment Using Computer Modeling

Page 4444 of 6666

Figure 5 Predicted potentials on the surface of the metals. This can be understood as a type of galvanic stress where the larger the difference in potential the greater the driving force for corrosion

In the model the materials can be easily specified for each of the surfaces as well as the properties of any coatings or sacrificial layers applied. From a corrosion point of view it can be anticipated that the aluminum will act as an anode and corrode near the steel bolt (cathode) as its position in the galvanic table is more negative than the aluminum. However as the area of the anodic aluminum is relatively large compared with the cathodic steel the corrosion would be expected to be less than if the opposite was the case

Figure 6 The arrows are showing how the currents are flowing between the steel and aluminum through the electrolyte

In Figure 5 the model clearly demonstrates the galvanic potentials where the blue anodic aluminum potentials can be seen to contrast with the red cathodic areas of the steel. The potentials show the gradients in the electrolyte as a result of the “Galvanic Stress” caused by the metals position in the electrochemical table. Whereas in this case the behavior is largely predictable, the model clearly shows the interaction between the metal components and the extent of the interaction. This is shown more clearly in Figure 6 where the current flowing through the electrolyte can be visualized

providing a clear understanding of the root cause of the corrosion.

Figure 7 Predicted corrosion rates. In this case the red indicates areas which are corroding

Figure 7 shows the corrosion rates which provide data not only on the extent of the area expected to corrode and the actual rate of corrosion. This type of model can be easily used to assess the impact of material changes, the extent of metal areas which may be painted to reduce the corrosion to acceptable levels and to plan maintenance schedules.

Case Study - Landing Gear Aircraft landing gear are subject to harsh environments which increases the risk of galvanic corrosion. This study aimed to investigate using computer modeling, the corrosion risk and the effectiveness of mitigation measures.

Figure 8 Landing gear corrosion risk assessment model

The model was used to identify and quantify the corrosion risk which can be clearly seen in Figure 9. By simply changing the model material and coating properties alternative designs can be investigated for their corrosion risk, and mitigation measures designed and optimized.

Corrosion Risk Assessment Using Computer Modeling

Page 5555 of 6666

Figure 9 Model results showing areas of greatest corrosion risk

Case Study – Aircraft Structure Galvanic corrosion (GC) in aircraft is important as it occurs whenever dissimilar metals or certain types of composites are located close to each other. Therefore the case study was focused on the extent and rate of corrosion and how it was impacted by the geometry of the connections, the characteristic and extent of the electrolyte and the type of mitigation methods employed. A typical structure representative of a modern aircraft design was studied which consisted of a CFRP stringer and an Aluminum rib.

The modeling approach is very similar to that used for structural and stress analysis in that the geometry of the structure is divided into finite elements, each element representing the surface of the material in contact with

the corrosive electrolyte. In fact the same CAD based tools can be used to create the corrosion model as those used for the structural applications.

Figure 10 Model visualization of galvanic potential on the structure which is often referred to as a form of galvanic stress

Figure 11 Model predictions of the galvanic corrosion sites and their severity. Once a model is developed alternative materials and protection measures can be quickly evaluated

Figure 12 Overall model of the structure assembly with the aim of assessing the corrosion risk associated with coating damage

Page 6666 o

Having developed the model of the aircraft structure the user can select the type of environmental conditions the structure is expected to experience and predict the location of potential corrosion sites and assess the severity of the corrosion. The model can also be used to identify the location and extent of corrosion protection measures required against galvanic corrosion (e.g. coatings, paints etc.) and can also contribute to the reduction of materials and process development costs, as well as predicting the long term impact of coating degradation and damage.

Summary The modeling technology can be applied to a wide range of structures and components such as those found in aircraft, automobiles, ground vehicles, ships and similar structures. It is well suited to support engineers with sensitivity analyses, parametric studies, what if scenarios, optimization studies and risk assessment

The benefits include reduction in qualification time and costs, development and improvement of testing procedures, reduction of material and process development costs and decrease of maintenance costs.

Figure 13 Corrosion modeling can be applied to structures such as automobiles to assess the risk of corrosion

References 1. Modeling galvanic corrosion protection of

aircraft structures. Siva Palani, Theo Hack, Andres Peratta, Robert Adey, John Baynham, Hubertus Lohner. DOD Corrosion. Palm Springs. USA. 2011

2. Validation of a galvanic corrosion model for AA2024 and CFRP with localized coating damage. Siva Palani, Theo Hack, Andres Peratta, Robert Adey, John Baynham, Hubertus Lohner. Eurocorr. Moscow, Russia. 2010.

3. Validation of a Galvanic Corrosion Computer Model for AA2024 and CFRP with localized

coatings damage. Andres Peratta, Theo Hack, Robert Adey, Siva Palani, John Baynham, Hubertus Lohner. CMLP (2nd International Workshop on Light Weight Metal Corrosion & Modeling for Corrosion Prevention, Life Prediction and Assessment). Rome, Italy. 2010.

4. Galvanic Corrosion Modeling for Aircraft Environments. A. Peratta, T. Hack, R. Adey, J. Baynham, H. Lohner, Eurocorr (2009).

5. BEASY Corrosion Manager Software, BEASY 2012.