Fatigue crack propagation behavior and life prediction of aluminum magnesium alloy powder metallurgy components for aerospace applications

Aluminum magnesium alloy powder metallurgy components are widely used in aerospace structural components (such as unmanned aerial vehicle frames and satellite supports) due to their advantages of lightweight and comprehensive mechanical properties. However, their fatigue crack propagation behavior directly threatens service safety. Revealing the evolution law of cracks and establishing accurate life prediction models is the core issue in improving the reliability of components.

1、 Multi stage characteristics and mechanisms of fatigue crack propagation

The fatigue failure of aluminum magnesium alloy powder metallurgy components follows a three-stage mode of "initiation stable expansion unstable fracture", which is controlled by both microstructure and load environment:

Crack initiation stage (<1 mm)

Origin characteristics: Cracks often originate from surface processing defects (such as milling tool marks), internal pores (sintering residues, size>50 μ m), or interface debonding between brittle phases (such as Mg ₁₇ Al ₁₂) and the matrix.

Driving mechanism: Under cyclic loading, grain boundary slip forms a persistent slip band (PSB), accompanied by intensified local stress concentration due to oxidation. Microcracks initiate when the plastic strain amplitude exceeds the critical value (such as 0.1%).

Stable expansion phase

Low speed expansion zone (da/dN<10 ⁻⁶ mm/cycle): Cracks propagate along the transgranular path, and fine fatigue striations can be seen on the fracture surface. The expansion rate is dominated by the stress intensity factor amplitude (Δ K), which conforms to the Paris formula (da/dN=C (Δ K) ⁿ), where C and n are material constants (such as AZ31 alloy n ≈ 3.5).

High speed expansion zone (da/dN>10 ⁻⁴ mm/cycle): Cracks turn and propagate along the grain, accompanied by the generation of secondary cracks, and the fracture surface presents a sugar like morphology. At this point, Δ K approaches the fracture toughness (K ₁ C), and environmental media (such as moisture) may accelerate the expansion (stress corrosion synergistic effect).

Unstable fracture stage

When the crack size reaches the critical value (determined by K ₁ C and load conditions), the component suddenly fractures brittle, with a mixture of ductile dimples and tearing edges on the fracture surface, and the fracture energy is significantly reduced.

2、 The regulatory effect of microstructure on crack propagation

Powder particle size and densification degree

Components prepared with fine powder (<50 μ m) have small grain size (<10 μ m) and large grain boundary area, which can hinder crack propagation through the "grain boundary deflection effect" and reduce the propagation rate by 20% -30%.

High porosity (>5%) leads to an increase in stress concentration factor, and cracks are prone to initiate from pore clusters. At the same Δ K, the propagation rate is more than 50% higher than that of dense materials.

Alloy elements and second phases

The thermally stable phase formed by rare earth elements (such as Y, Nd) (such as Mg ₂₄ Y ₅) can pin the crack tip, increasing the propagation resistance; And the coarse Mg ₁₇ Al ₁₂ phase (>2 μ m) is prone to become a crack propagation channel.

Adding 0.5% Zr can refine the grain size and form dispersed ZrAl2O3 particles, increasing the fatigue crack propagation threshold (Δ Kth) from 2.5 MPa · m ¹/² to 3.8 MPa · m ¹/².

3、 Life prediction model and key technologies

Classic model based on fracture mechanics

Paris formula extrapolation method: Determine the da/dN - Δ K curve through fatigue testing, and integrate it with the initial defect size of the component (such as the maximum equivalent pore diameter measured by non-destructive testing) to calculate the number of cycles (Nf) for the crack to propagate from the initial size (a ₀) to the critical size (ac).

Probability statistical model: Introducing Weibull distribution to describe the randomness of material defects, establishing a life prediction equation with defect probability density, suitable for reliability evaluation under complex load spectra.

Emerging predictive technologies

Machine learning model: Using neural networks to train microstructure parameters (porosity, grain size, second phase volume fraction) and fatigue data, constructing a nonlinear mapping relationship, and improving prediction accuracy by 15% -20% compared to traditional formulas.

In situ monitoring technology: combined with CT scanning or digital image correlation (DIC) technology, real-time tracking of crack propagation trajectory, correction of model parameters to achieve dynamic updates of life prediction.

Surface reinforcement intervention

Laser shock peening (LSP): Introducing a 300-500 MPa compressive stress layer on the surface of the component prolongs the crack initiation life by 2-3 times while reducing the propagation rate (such as a 40% reduction in da/dN of AZ91 alloy).

Shot peening treatment: refines surface grains and closes microcracks through plastic deformation, suitable for complex curved components, and can increase fatigue limit by 15% -25%.