No abstract

This study investigated the nucleation and growth mechanism of reaction layers and phases of hot-dipped boron steel in pure Al at 690 °C for 0–120 s. In the case of a dipping time of 30 s, reaction nuclei of width 10–15 μm and height 10 μm were formed on the steel surface in the flow direction of the liquid Al. This reaction layer was formed as a mixture of θ (Fe4Al13) phase of several nm to 2 μm, θ and η (Fe2Al5) of several nm, a columnar η region, and a β (FeAl) region of 500 nm thickness at the steel interface. At the grain boundaries of ferrite, in contact with the η phase, κ (Fe3AlC) was formed. Using the calculated Fe-Al phase diagram, it was determined that when Fe was dissolved in liquid Al from the steel above 2.5 at% (0.6 wt%), the θ phase was formed. Although most of the θ phases continuously grew toward the liquid phase, the θ phase in contact with the steel was transformed into the η phase with minimal differences in composition due to the inter-diffusion of Al and Fe. It was therefore concluded that the η phase formed at the interface became a growth nucleus and grew in a columnar form toward the steel.

Microstructural evolution and formation mechanism of reaction layer for 22MnB5 steel hot-dipped in Al–10Si (in wt %) alloy was investigated. The microstructural identification of the reaction layer was characterized via transmission electron microscopy and electron backscatter diffraction. In addition, the formation mechanisms of the phases were discussed with vertical section (isopleth) of the (Al–Si–Fe) ternary system. The solidified Al–Si coating layer consisted of three phases of Al, Si, and τ5 (Al8Fe2Si). The reaction layer on the Al–Si coating layer side is a fine τ5 phase (Al8Fe2Si) of 5 μm thickness. The layer on the steel side consisted of an η phase (Fe2Al5) of thickness of 500 nm or less. τ1 (Al2Fe3Si3, triclinic) phase of 200-nm-thickness was formed in the η phase, and κ phase (Fe3AlC) of 40–50 nm thickness was formed between η phase and steel. The τ5 phase was formed by isothermal solidification at 690 °C in the liquid Al–10 wt % Si when 3.73–29.0 wt % of Fe was dissolved from the boron steel into the Al–Si liquid bath. It was considered that the η phase was formed by the diffusion reaction of Al, Si, and Fe between τ5 and ferrite steel. κ (Fe3AlC) phase was formed by the reaction of the carbon, which is barely employed in η and τ phases, and diffused Al.

Abstract:In laser welding and hot stamping Al-Si-coated boron steel, there is a problem that the strength of the joint is lowered due to ferrite formation in the fusion zone. The purpose of this study is to develop an Al-7 wt.% Mn hot-dip coating in which Mn, an austenite stabilizing element, replaces the ferrite stabilizing element Si. The nucleation and formation mechanism of the reaction layer was studied in detail by varying the dipping time between 0 and 120 s at 773 • C. The microstructure and phase constitution of the reaction layer were investigated by various observational methods. Phase formation is discussed using a phase diagram calculated by Thermo-Calc TM . Under a 30 s hot-dipping process, no reaction occurred due to the formation of a Fe 3 O 4 layer on the steel surface. The Fe 3 O 4 layer decomposed by a reduction reaction with Al-Mn molten alloy, constituent elements of steel dissolved into a liquid, and the reaction-layer nucleus was formed toward the liquid phase. A coated layer consists of a solidified layer of Al and Al 6 Mn and a reactive layer formed beneath it. The reaction layer is formed mainly by inter-diffusion of Al and Fe in the solid state, which is arranged on the steel in the order of Al 11 Mn 4 → FeAl 3 (θ) → Fe 2 Al 5 (η) phases, and the Fe 3 AlC (κ) in several nm bands formed at the interface between the η-phase and steel.

The reinforced concrete (RC) member’s shear strength estimation has been experimentally studied in most cases due to its nonlinear behavior. Many empirical equations have been derived from the experimental data; however, even those adopted in the construction codes do not thoroughly and accurately describe their shear behavior. Theoretically explained equations, on the other hand, are aligned with the experiment; however, they are complicated to use in practice. As shear behavior research is data-driven, the machine learning technique is applicable. Herein, an artificial neural network (ANN) algorithm is trained with 776 experiment results collected from available publications. The raw data is preprocessed by principal component analysis (PCA) before the application of the ANN technique. The predictions of the trained algorithm using ANN with PCA are compared to those of formulae adopted in a few existing building codes. Finally, a parametric study is conducted, and the significance of each variable to the strength of RC members is analyzed.