Purpose: To evaluate the suitable range of the Double-Angle Method (DAM) for B₁ mapping in low flip angle VFA-T₁ quantification using a five-level phantom-human brain model system.Methods: We scanned five types of samples on a 3 T scanner using the same DAM sequence (TR/TE 9000/15.9 ms, flip angles 60°/120°): pure water phantom, NaCl water phantom, NaCl+Gd water phantom, a healthy human brain, and a brain with hemorrhage. Each sample had 9 slices scanned. Using slice 6 as the reference, we calculated the mean (μ), standard deviation (σ), coefficient of variation (CV), relative range (RelRange), and 90th percentile width (P90) after background removal. We performed statistical analysis both between slices and between groups.Results:Between slices (Pure water phantom): The average CV across 9 slices was 5.65%, and the ICC was 0.261. This shows that differences between slices need to be accounted for (e.g., using covariates).Between groups (Slice 6): The range of group means was 3.2%. The CV increased stepwise: 4.7% (pure water) → 9.2% (NaCl) → 10.5% (NaCl+Gd) → 14.8% (healthy brain) → 11.3% (brain with hemorrhage). Human brain CVs were still below the 20% clinical tolerance limit. Adding Gd caused less than 0.5% extra CV.Conclusion: DAM shows good stability across different materials under 60°/120° conditions. It can be used directly in scenarios where CV ≤ 10%. For CV > 10%, parallel transmission or second-order shimming is needed.