Novel zinc-based alloys used to improve the corrosion protection of metallic substrates

Abstract The protection of metallic structural components against corrosion is fundamental to preserve their mechanical properties in aggressive environments. Zn-based coating represents one of the most used techniques to make protective coatings for metallic substrates. In the present paper, two types of novel zinc-based coating are proposed, by employing either a tin addition or an aluminium-tin-copper addition to the traditional zinc bath. The behaviour of steel-coated specimens under bending is experimentally and numerically investigated by considering different bath dipping times. A quite satisfactory agreement between experimental and numerical results is observed, especially under plastic behaviour regime.


3
The inner zone of coating is characterized by high contents of iron. The content of zinc is predominant in outer zone, being the surface composition closes to the chemical composition of the bath: as a matter of fact, its formation is due to the solidification of the wet layer after extraction from the bath.
Between inner and outer zones, there is the stability of intermetallic phases according to the equilibrium diagram [5].
The  phase has an iron content equal to about 5-6 wt%.
The external phase, named  phase, is characterized by a low value of iron content. Usually all phases with an iron content greater than 12wt% are named  phase.
In order to optimise both mechanical and chemical properties of coating, alloys of metallic elements may be added to the galvanizing bath.
Influence of copper, cadmium and tin additions on both morphology and thickness of galvanized coatings was studied by Katiforis  to the practice of engineering failure analysis.

Specimens
Two series of rectangular 80x25x3mm specimens are obtained from two hot-rolled ipersandelin plates (named Support 1 and Support 2 in the following). The mechanical properties of Supports 1 and 2 are reported in Table 1 and Table 2, respectively. - phase, characterized by high content of iron (equal to about 7-12% in weight), located in the inner zone of the coating; - phase, characterized by low content of iron (lower than 5-6% in weight), located in the outer zone of the coating.

Experimental tests
On each Series of specimens, three experimental bending tests were performed for the five values of dipping time examined, by employing an electromechanical non-standard device, shown in

Results
The experimental results related to the above bending tests, in terms of bending moment against half bending angle of specimen, are shown in Figure 5 for each considered dipping time.

9
As can be noted, the increasing dipping time leads to higher bending moment values, and this effect is more pronounced for Series 2 specimens due to their higher bath reactivity.    By assuming plain strain condition, a 2D model with 4-nodes plates finite elements is employed. Due to the symmetry, only one half of the specimen is modeled (Fig.8).

Finite element model
The specimen has a length of 50mm, being such a value equal to the calibrated length of the bent specimen (Fig.4).

Figure 8.
Three layers are piled up for meshing the thicknesses of coating corresponding to the experimentally observed  ,  and  phases for Series 1 specimens, whereas  , lamellar and  phases for Series 2 specimens, whose thicknesses are plotted in Figure 1. A layer of a thickness equal to 3mm is used for meshing the substrate. The bonds between the phases and between the  phase and the support are perfect.
In order to numerically simulate the experimental boundary condition, the following conditions are modeled: -for nodes along x2-axis (Fig.8), displacements along x1-axis and rotation around x3-axis are taken equal to zero;

11
-for nodes along the opposite extreme, rotation around x3-axis is applied (Fig.8), by increasing the half bending angle from 0 to 35 degrees, degrees with an increment equal to 1 degree at a time.
For Series 1 specimens, the following stress-strain relationships are assumed: (i) for Support 1, a constitutive law typical for ductile steels, with the mechanical properties listed in Table 1; (ii) for both  and  phase, constitutive laws typical of elasticplastic materials, with the mechanical properties listed in Table   1 (properties independent of the dipping time); (iii) for  phase, a constitutive law that changes by varying the dipping time, according to Figure 9. The assumption (iii) is justified by the fact that, for dipping time equal to 180, 360, and 900 s, the  phase experimentally shows a large region characterized by a nonoriented morphology, that strongly modifies the mechanical behaviour of such a phase, while columnar crystals are predominant for lower dipping times.
For Series 2 specimens, the following stress-strain relationships are assumed: (i) for Support 2, a constitutive law typical for cast irons, with the mechanical properties listed in Table 2; (ii) for both  and  phase, constitutive laws typical of elasticplastic materials, with the mechanical properties listed in Table   2 (properties independent of the dipping time); (iii) for lamellar phase, a constitutive law that changes by varying the dipping time, according to Figure 10.
The assumption (iii) is justified by the fact that, for dipping time equal to 180, 360, and 900 s, the lamellar phase changes its morphology. On the other hand, the lamella is characterized by a different chemical composition than the matrix, and its formation is probably due to presence of an eutectic liquid during the coating formation in the beginning stages (up to 60s). Exposing the lamella and its matrix at high temperature for more time (180, 360 and 900s) implies the activation of diffusion phenomena which lead to a better homogenization of the chemical composition, changing the lamellar phase morphology and its mechanical properties.

Figure 10.
A numerical nonlinear static analysis is performed.
In Figures 11 and 12

Figure 12.
A satisfactory agreement between the above results can be observed for Series 1 specimens in Figure 11, especially in plastic region (that is, for half-bending angle higher than 7 degrees). The phase thickness increasing with dipping time leads to appreciably higher bending moment bearing.
However, a satisfactory agreement between numerical and experimental results can be also noted for Series 2 specimens ( Fig. 12), except for a dipping time equal to 15s. Figure 12 shows a stronger influence of phase thickness on mechanical behaviour in the case of Series 2 specimens with respect to the Series 1 specimens but, in such a case, a decrease in bending strength is observed for high values of bending angle.

CONCLUSIONS
In the present work, the mechanical behaviour of ipersandelin steel plate specimens galvanized with two different types of improved zinc-based coatings has been analysed. In particular,   Table 1. Mechanical properties of Support 1 and intermetallic phases  and  related to Series 1 specimens.              ZnAl5Sn1Cu0.5%