Detecting Corrosion in Metal Elements of Ammunition by IR Thermography Methods Waldemar SWIDERSKI Military Institute of Armament Technology; Zielonka, Poland Phone: +48 22 7614552, Fax: +48 22 7614447; e-mail: waldemar.swiderski@wp.pl Abstract In result of long storage some metal elements of ammunition are subjected to corrosion by impact of environmental factors such as moisture and salt existing in the air. Changes of temperature can condensate water and accelerate processes of corrosion. During transport mechanical damage can happen in protecting covers of ammunition metal elements leading to formation of places susceptible to corrosion. Pyrotechnical materials undergo physicochemical changes (e.g. chemical reactions like decomposition, oxidation, phase transitions) and they can also initiate the corrosion of metal elements. As the corrosion causes the mechanical weakness of ammunition metal elements it triggers different hazards. Therefore an important matter is to detect places of corrosion by nondestructive testing methods. In order to evaluate the potential effectiveness of Infrared (IR) thermographical methods for non-destructive testing of ammunition metal elements the computer 3D simulations were carried out with deployment of specialized software ThermoCalc-6L. In this paper the results of such computer simulation are presented. Keywords: IR thermography, corrosion, computer simulation, metal elements of ammunition 1. Introduction 5th Pan American Conference for NDT 2-6 October 2011, Cancun, Mexico In the result of long storage period the ammunition is subjected to natural ageing processes and as it loses original properties in different degree and time it even may become useless to combat action. Metal elements of ammunition are subjected to corrosion. It is a progressive failure of materials caused by impact of environment. The notion of corrosion is from Latin word corrodere eat up. The corrosion concerns above all metals and alloys, though it affects non-metal materials also. There are different causes of corrosion belonging to chemical and electrochemical and biological and mechanical processes. Chemical and electrochemical corrosion is distinguished for metals. The chemical corrosion depends on reactions with oxygen (atmospheric) and chlorine and oxides sulphur and nitrogen. The main type of corrosion phenomena in case of metals and alloys is electrochemical corrosion to which belongs universal phenomenon of atmospheric corrosion caused by reaction of moist and vitiated air on metals. The electrochemical corrosion comes into being as a result of working shorted cells on point of contact between metal and electrolyte. These cells come into being in result of chemical (or physical) heterogeneity of metal, for example on point of contact of different metals but also as a result of crystalline heterogeneity in structure of metal. The most often corrosion of metals appears by surface losses of metal (stains and pits) and by reduction of strength of metals. Most often the speed of corrosion is defined by reduction of mass of a metal sample as the result of influence of corroding factor on surface and time unit. The corrosion usually begins from fine changes on attacked surface, which are most often in bent places of material, and it proceeds in depth of material too and destroys most flexible substances by corrosion. Sometimes products of corrosion create a passive layer protecting material against further degradation but in other cases it can be a next one causing corrosion factors. There are some factors important for consideration in corrosion processes: properties of metal: elements composition, electrochemical potential, contaminations, winning method, treatment method, surface quality and other;
surroundings the metal has contact with: intensity of factors which are the cause of corrosion (oxygen, water, acids and other); work conditions of metal: fatigue factors, shape of product, joining method (welding, bonding, soldering); temperature: most often corrosion processes are accelerated with growing temperature and also a harmful factor is the frequent change of metal or environment temperature; time and connected with it ageing processes of metal and protection coats and also stresses (fatigue of metal). There are two main types of corrosion depending on destruction area: surface corrosion, which is occurring on all surface of metal; localized corrosion, which is occurring in definite places of metal. Both kinds of corrosion have a lot of variations. In case of surface corrosion there is: uniform corrosion covering surface of metal on uniform layer; non-uniform corrosion occurring in certain places on surface of metal or on different depth. In case of localized corrosion it can be distinguished: point corrosion that exists in form of scattered points in different places on surface of metal; spot corrosion forming spots on surface of metal; pitting corrosion concentrated in some places of metal surface causing deep pits in result of influence of different corrosion factors on different places of surface; undersurface corrosion occurring under surface of metal or under protective layer covering the surface; crevice corrosion produced by penetration of electrolyte in cracks of construction and forming in result corrosion cells; filiform corrosion existing on surface of metal covered by protective coat (varnish) in form of characteristic not crossing threads which start from one point; intercrystalline corrosion existing on borders of crystal grains caused by separation of different phases what changes chemical composition of crystal in place of contact point between crystal grains; transcrystalline corrosion propagating deep in material by crystal grains of metal. Universally protective coats are applied: inorganic coats: metal and non-metal; organic coats: paints, varnishes, artificial materials, tar and greases. Infrared (IR) thermographic non-destructive testing (NDT), or thermal NDT, is not a new concept in evaluation of physical objects. This technique is used to monitor surface thermal radiation of objects but, in fact, it can be regarded also as one of some techniques used to see the unseen because of the fact that surface thermal radiation is conditioned by volumic temperature and in this way may be sensitive to object inner structure. Thermal NDT is implemented in two modes: passive and active [1]. Passive IR thermography is mainly used as a tool for technical diagnostics (condition monitoring, predictive maintenance) in the inspection of industrial installations which, due to their operation, are characterized by stable or low-changing temperatures which are different from the ambient temperature. This technique is particularly convenient if an anomaly ( defect ) to be detected generates heat and thus differs by its temperature from the surrounding. Active IR thermography is used in NDT of materials which might contain subsurface defects having, prior a test, the same temperature as a host material. In order to produce recognizable temperature patterns in defect sites, objects are to be stimulated by applying a certain
amount of heat energy. Thermal stimulation can be superficial or volumic depending on a physical nature of a used heat source. Superficial heating is typically performed by using optical radiation or gas (fluid). Volumic heating can be achieved by applying microwaves, ultrasonic waves etc. Subsurface irregularities represent a specific obstacle to the heat flux which is diffusing in a solid body. Thermal disturbances produce abnormal temperature signals which reach object surface and can be monitored thermographically. Active IR thermographic NDT is attractive due to its non-contact character, high productivity and possibility to inspect both metal and non-metal materials [2, 3]. From the scientific point of view, thermal NDT may be effective for checking algorithms solving the so-called inverse heat conduction problems which are of a great interest of the heat conduction theory. Such solutions characterize subsurface defects trough evaluating such their parameters as size, depth and physical nature. Because ammunition should be safe for use at temperatures between - 40 C to + 55ºC esp. in military equipment used in climatic zones attributed to Poland, the application of active IR thermpgraphy as NDT method is potentially attractive for its testing. 2. Computer simulation 2.1 Programme In order to determine the potential usefulness of thermal methods for non-destructive testing of composite pyrotechnic materials the computer simulations were carried out with the deployment of the specialized software ThermoCalc-6L (developed by V. Vavilov for needs of Military Institute of Armament Technology in Poland - MIAT). The ThermoCalc- 6L software is intended for calculating three-dimensional (3D) temperature distributions in anisotropic six-layer solid bodies which may contain up to nine subsurface defects. The corresponding mathematical heat conduction problem is modeled in Cartesian coordinates and solved by using an implicit finite-difference numerical scheme. Originally, ThermoCalc-6L was developed for simulating thermal nondestructive testing (NDT) problems where transient temperature signals over subsurface defects are of a primary interest. These signals evolve in time and diffuse in space. The unique numerical algorithm implemented in ThermoCalc- 6L, which unlike to currently available commercial software, enables the modeling of very thin defects in rather thick materials without losing the computation accuracy and allows to analyze up to nine defects within a specimen. The sample can be heated uniformly or nonuniformly with a square or cosine shape of heat pulse that gives possibilities to study defect cross-influence and lateral 3D heat diffusion [4-6]. ThermoCalc-6L is suitable to find the solution of transient heat conduction problem for a six-layer parallelepiped-shaped body that contains up to nine parallelepiped-shaped defects. The body is heated or cooled down on the front surface with an external heat pulse. The front-surface heat flux is assumed to be uniform or Gaussian-distributed in space. The heat flux center can be located at any point on a front surface. Along with the external thermal stimulation of the front surface, both the front and rear surfaces are cooled down according to the Newton s law. Thermal properties of the specimen and the defects can be specified separately in three spatial directions, thus modeling fully anisotropic material. The specimen side surfaces are adiabatic. On the boundaries between the specimen layers and between the host materials and the defects, the temperature and heat flux continuity conditions take place. The conception of the so-called capacitive defects is realized in ThermoCalc-6L. This means that, unlike resistive defects involved in some other NDT models, both defect thermal diffusivity and conductivity are taken into account. This provides the most correct description of physical phenomena occurring in the spaces occupied by defects.
2.2 Models Three models of metal samples with a paint layer (coating) in the form of rectangular prism plate were simulated in order to analyze a possibility of nondestructive testing by IR thermography methods used for detecting corrosion in metal elements of ammunition. The size of the model plates is 60x100 mm and 3 mm thick (layer of paint was simulated in three variants: 0.2 mm and 0.1 mm and 0.05 mm thick). In Model 1 (Fig. 1) three defects parallel to the surface of the sample were simulated: Defect 1 (D1) size was 20x20 mm and 0.05 mm thick and Defect 2 (D2) size was 10x10 mm and 0.05 mm thick and Defect 3 (D3) size was 5x5 mm and 0.05 mm thick. These defects simulate corrosion of metal under layer of paint. In Model 2 (Fig.1) there were simulated three defects parallel to the surface of the sample: Defect 4 (D4) size was 20x20 mm and 0.1 mm thick and Defect 5 (D5) size was 10x10 mm and 0.1 mm thick and Defect 6 (D6) size was 5x5 mm and 0.1 mm thick. These defects simulate undersurface corrosion occurring under surface of metal and were located at different depths. In Model 3 (Fig. 1) there were simulated three defects perpendicular to the surface of the sample: Defect 7 (D7) size was 40x0.2 mm and 2.5 mm thick and Defect 8 (D8) size was 40x0.1 mm and 2.5 mm thick and Defect 9 (D9) size was 40x0.05 mm and 2.5 mm thick. These defects simulate crevice corrosion. Computer simulations were performed by means of the ThermoCalc-6L software and carried out for three types of metals applied in manufacturing ammunition elements: steel and aluminium and brass. Table 1 shows thermal data of materials and the air (which simulates a defect) used for the computer simulation. Table 1: Thermal data for used materials Material Specific heat, kj/kgk Thermal conductivity, Specific density, kg/m 3 W/mK Steel 420 30 8100 Aluminium 870 120 2700 Brass 381 100 8600 Air (thin gaps) 1.005 0.07 1.2 2.3 Results A model of samples was introduced (Fig.1) to simulate heating stimulations by computer programme. The heating of all samples was modeled in two variants. In the first variant of heating a thermal pulse was applied on the front surface of the sample with the pulse heat duration τ h = 0.01 s and power density Q = 10 5 W/m 2. In the second variant the sample was heated to 50 C and the cooling was simulated to ambient temperature (20 C). A specimen was continuously cooled by the ambient air but at the beginning of the normal convection process a cooling pulse was used. In this simulation the cooling pulse duration was τ c = 1 s and power density of cooling pulse was Q = 5 10 3 W/m 2. Selected results of optimum parameters for detection of defects D1 D9 and different metals can be calculated as shown in Table 2 5 respectively. It was assumed that a defect can be reliably detected by its surface temperature footprint if the corresponding sample excess temperature T and the signal T meet the following conditions: - a sample maximum excess temperature T ( τ h ) that occurs at the end of heating is lower than the destruction temperature of the sample material T destr (+ 55ºC); - a T signal must exceed a temperature resolution of a used IR system Tres ;
- a running temperature contrast C = T ( τ ) / T ( τ ) must exceed the noise level that adheres to each material and surface condition (up to 2%). 20x20x0,1 mm 10x10x0,1 mm 5x5x0,1 mm D1 D D3 Model 1 20x20x0,1 mm 10x10x0,1 mm D4 5x5x0,1 mm D5 D6 Model 2 0,2x40x25 mm 0,1x40x25 0,05x40x25 mm D7 D8 D9 Model 3 Figure 1. Models Let us assume that T destr = + 55ºC, T 0.1 o res = C and C n = 2%. If to apply the detection criteria to the data in Table 2-4, it can be stated the following: - the sample maximum surface temperature will not exceed 50 C; - not all defects produce T > 0.1 C and C > 2%.
Table 2: Expected detection parameters in the front-surface (sample from steel, paint layer 0.2 mm thick) Heating pulse D1 D2 D3 6.7 5.8 5.6 0.01 0.47 0.48 Cooling pulse D1 D2 D3-8.4-8.49-8.46 1.63 1.01 0.89-4.00-4.06-4.05 Table 3: Expected detection parameters in the front-surface (sample from aluminum, paint layer 0.2 mm thick) Heating pulse D1 D2 D3 3.2 2.7-2.1 0.77 0.39 0.01 Cooling pulse D1 D2 D3-8.55-8.65-8.66 2.03 0.74-1.02-4.07-4.12-4.13 Table 4: Expected detection parameters in the front-surface (sample from steel, paint layer 0.05 mm thick, defects at depth of 0.75 mm) Heating pulse D4 D5 D6 2.45 1.85 0.87 0.49 0.37 0.22 Cooling pulse D4 D5 D6-2.49-2.21-1.68 1.1 0.9 0.7 0.44 0.14 0.09-0.73-0.59-0.33 Table 5: Expected detection parameters in the front-surface (sample from aluminum, paint layer 0.05 mm thick, defects at depth of 0.75 mm) Heating pulse D4 D5 D6 2.64 1.65 0.52 0.16 0.13 0.1 0.14 0.13 0.12 Table 6: Expected detection parameters in the front-surface (sample from brass, paint layer 0.05 mm thick) Cooling pulse D7 D8 D9-7.95-6.58-7.95 0.7 0.7-0.28-0.23-0.28
Fig. 2 3 show selected results obtained by computer simulation of metal samples possessing corrosion defects. Fig. 2 presents example of temperature changes on front surface of steel sample covered by paint layer 0.2 mm thick and stimulated by heat pulse duration τ h = 0.01 s and power density Q = 10 5 W/m 2. As it is visible on Fig.2 the maximum growth of surface temperature is about 15 C. Similar results were obtained for all analyzed samples. It was assumed that ambient temperature is 20 C then maximum temperature on front surface of samples heated by heating pulse is 35 C. 16 14 Temperature, C 12 10 8 6 4 2 0 0.00 0.10 0.20 0.30 0.40 0.50 0 0.70 0.80 0.90 1.00 Time, s Figure 2. Temperature changes on front surface of the sample (τ h = 0.01 s, Q = 10 5 W/m 2 ) Temperature profile of front surface of steel sample with defects (D1 D3) which is running along the sample by centre of defects and obtained for maximum difference of temperature (0.8 s after start heating) is shown in Fig. 3. 0.5 Temperature, C 0.4 0.3 0.2 0.1 0 1 10 19 28 37 46 55 64 73 82 91 100 Length, mm Figure 3. Temperature profile of front surface of the sample (steel) with subsurface corrosion (D1 D3)
3. Conclusions Results received from the computer simulation show that active IR thermography could be an effective method in detection of different types of corrosions occurring in metal elements of ammunition. However, the results also depend on limitations of testing conditions. Results of conducted simulations show that: - detecting defects of crevice corrosion type (D7 D9) is possible only by the method with cooling pulse, - defects (D1 D3) existing under paint layer were detected in all cases, - all subsurface defects (D4 D6) in steel sample were detected to 0.5 mm depth, while the largest defect 20x20 mm size (D4) was detected to 1 mm depth by the method with heating pulse, - all subsurface defects (D4 D6) in aluminum sample were detected to 0.75 mm depth but the largest defect to 1.25 mm depth by the method with heating pulse, - all subsurface defects (D4 D6) in steel sample were detected to 1.5 mm depth and the largest defect to 2.4 mm depth by the method with cooling pulse, these depths in aluminum sample were 2 mm and 2.8 mm adequately, - results received by computer simulation will be verified by planned experimental tests. Acknowledgements The research work was supported by the National Centre for Research and Development of the Republic of Poland. References 1. X. P. V. Maldague Theory and practice of infrared technology for nondestructive testing John Wiley&Sons, New York, 2001 2. W. Swiderski Applications of IR Thermography Methods for Nondestructive Evaluation of Honeycomb Type Composite Materials in Aircraft Industry Proceedings of the Fourth European Workshop Structural Health Monitoring 2008, pp.1297-1304 3. W.Świderski Lock-in Thermography to rapid evaluation of destruction area in composite materials used in military applications SPIE vol. 5132, 2003, pp. 506-517 4. W. Świderski Posibility of defect detection in pyrolitic graphite substrates by IR thermography Proceedings of 28 th International Conference of the Society for the Advancement of Materials and Process Engineering, Editor: SAMPE, Long Beach, 2007, pp.305-313 5. W. Swiderski, D. Szabra Possibility detection of defects in multi-layered composite materials used in military applications by IR thermography Proc. Vth International Workshop, Advances in Signal Processing for Non Destructive Evaluation of Materials, pp. 137-142, Québec City (Canada) 2005 6. ThermoCalc-6L, User s Manual, Tomsk, 2005