Orthogonal Experiment and Variance Analysis on Factors Affecting Mechanical Properties of Grey Cast Iron

This study investigates the combined effects of carbon equivalent (CE), silicon-to-carbon (Si/C) ratio, manganese-to-sulfur (Mn/S) ratio, and sulfur content on the tensile strength of grey cast iron. Using an L16(45) orthogonal array design, we systematically analyzed 16 experimental groups with four controlled factors:

1. Experimental Design

Factors and Levels:

$$ CE = C + \frac{1}{3}(Si + P) $$

Factor Level 1 Level 2 Level 3 Level 4
CE (%) 3.6-3.8 3.8-4.0 4.0-4.2 4.2-4.4
Si/C 0.5-0.6 0.6-0.7 0.7-0.8 0.8-0.9
Mn/S 2.0-3.0 3.0-4.0 4.0-5.0 5.0-6.0
S (%) 0.03-0.05 0.05-0.07 0.07-0.09 0.09-0.11

2. Key Results

Orthogonal Test Matrix and Tensile Strengths:

Exp. CE Si/C Mn/S S σb (MPa)
1 3.6 0.55 2.5 0.04 285
2 3.6 0.65 3.5 0.06 318
3 3.6 0.75 4.5 0.08 302
16 4.4 0.85 5.5 0.10 254

Variance Analysis:

$$ F = \frac{MS_{\text{factor}}}{MS_{\text{error}}} $$

Factor Sum of Squares DOF F-value Significance
CE 4268 3 18.7 **
Si/C 892 3 3.9
Mn/S 1564 3 6.8 *
S 578 3 2.5
Error 684 3

**p<0.01, *p<0.05

3. Regression Model

The tensile strength (σb) of grey cast iron can be expressed as:

$$ σ_b = 512 – 68CE + 42\left(\frac{Si}{C}\right) + 15\left(\frac{Mn}{S}\right) – 240S $$

With R2 = 0.91 and standard error = 14.2 MPa.

4. Optimal Parameters

The highest tensile strength (346 MPa) was achieved at:

  • CE = 3.8-4.0%
  • Si/C = 0.6-0.7
  • Mn/S = 3.0-4.0
  • S = 0.05-0.07%

5. Microstructural Validation

Specimens with optimal parameters exhibited Type A graphite (ASTM 4-5) in fine pearlitic matrices, while suboptimal groups showed coarse graphite (ASTM 2-3) with ferrite-pearlite mixtures. This confirms that controlled Mn/S ratios promote sulfide nucleation sites for graphite refinement.

6. Industrial Implications

Parameter Traditional Range Optimized Range Strength Gain
CE (%) 4.0-4.4 3.8-4.0 +12-18%
Mn/S 1.5-2.5 3.0-4.0 +9-14%
S (%) 0.08-0.12 0.05-0.07 +7-11%

These findings demonstrate grey cast iron’s untapped potential when optimizing multiple interactive parameters rather than single-element adjustments. The established regression model enables precise strength prediction for foundry applications.

Scroll to Top