Flow Stress: A Unifying Parameter in Fatigue and Fracture Analysis

Introduction

Flow stress is one of those quantities that shows up everywhere in structural metals engineering, yet it rarely gets the dedicated treatment that stress intensity factor, S-N curves, or fatigue strength coefficients receive. It appears quietly in cyclic plasticity models, in fracture toughness test standards, in crack-tip field solutions, and in dynamic constitutive models — often without much explanation of what it actually represents physically or why that particular definition was chosen.

This post defines flow stress from first principles, surveys how it is modeled, and works through its specific roles in fatigue analysis (cyclic stress-strain behavior, notch plasticity) and fracture mechanics (crack-tip plastic zone size, elastic-plastic fracture parameters).

What Is Flow Stress?

Flow stress is the stress required to sustain plastic deformation at a given amount of accumulated plastic strain, strain rate, and temperature. It is, in effect, the instantaneous resistance of the material to continued plastic flow — distinct from the initial yield stress, which marks only the onset of plasticity.

For a uniaxial true stress-true strain curve, the flow stress at a given plastic strain $\varepsilon_p$ is simply the ordinate of the curve at that point:

$$\sigma_{\text{flow}} = \sigma(\varepsilon_p, \dot{\varepsilon}, T)$$

Because most engineering metals strain harden, flow stress increases with accumulated plastic strain up to the point of necking or fracture. This stands in contrast to a perfectly plastic material, for which flow stress is constant and equal to the yield stress regardless of how much plastic strain has accumulated.

A commonly used single-value approximation, particularly in fracture mechanics, is the average of the yield and ultimate tensile strength:

$$\sigma_{\text{flow}} \approx \frac{\sigma_y + \sigma_u}{2}$$

This approximation is a practical compromise — it captures a representative resistance to plastic flow over the strain range relevant to crack-tip plasticity without requiring the full hardening curve.

Modeling Flow Stress

Power-Law Hardening

The most common engineering representation of strain hardening is the Hollomon power-law relationship:

$$\sigma = K \varepsilon_p^n$$

where $K$ is the strength coefficient and $n$ is the strain hardening exponent. This form underlies the Ramberg-Osgood representation widely used in elastic-plastic stress-strain modeling:

$$\varepsilon = \frac{\sigma}{E} + \left( \frac{\sigma}{K} \right)^{1/n}$$

Rate and Temperature Dependence

When strain rate and temperature effects matter — as discussed in an earlier post on strain-rate strengthening — flow stress is more completely captured by constitutive models such as Johnson-Cook:

$$\sigma_{\text{flow}} = \left[ A + B\varepsilon_p^n \right] \left[ 1 + C \ln\dot{\varepsilon}^{\ast} \right] \left[ 1 – T^{{\ast}m} \right]$$

Here, the bracketed terms separate the strain-hardening, strain-rate-hardening, and thermal-softening contributions to flow stress. This is the same flow stress concept discussed above, simply extended to account for the rate and temperature sensitivity relevant to impact, ballistic, and high-rate forming applications.

Flow Stress in Fatigue Analysis

The Cyclic Stress-Strain Curve

Under cyclic loading, a material’s resistance to plastic flow can differ substantially from its monotonic (single-pull tensile) behavior. Repeated plastic cycling can cause cyclic hardening (flow stress increases with cycling) or cyclic softening (flow stress decreases with cycling), depending on the initial microstructural state and dislocation substructure. The stabilized cyclic stress-strain curve, typically reached after a few hundred cycles, is described by an equation directly analogous to the monotonic Ramberg-Osgood form:

$$\frac{\Delta\varepsilon}{2} = \frac{\Delta\sigma}{2E} + \left( \frac{\Delta\sigma}{2K’} \right)^{1/n’}$$

where $K’$ is the cyclic strength coefficient and $n’$ is the cyclic strain hardening exponent. These cyclic flow stress parameters, not the monotonic ones, are the correct inputs for strain-based (Coffin-Manson) fatigue life prediction, since they reflect the material’s actual resistance to plastic flow under the repeated loading the component will experience in service.

Masing Behavior and Hysteresis Loops

Many metals exhibit Masing behavior, in which the shape of the cyclic stress-strain hysteresis loop is a scaled (2x) version of the stabilized cyclic curve. This assumption allows the flow stress relationship above to be used directly to construct hysteresis loops for any loading reversal, which is essential input to rainflow-counted variable amplitude fatigue analysis and to the strain energy density fatigue methods discussed in a previous post.

Neuber’s Rule and Notch Plasticity

At a stress concentration — a fillet, hole, or notch — the local stress and strain can exceed the elastically calculated values once local yielding occurs. Neuber’s rule relates the elastically calculated (nominal) stress concentration to the actual local elastic-plastic stress and strain through the flow stress relationship:

$$\sigma \varepsilon = \frac{(K_t S)^2}{E}$$

where $\sigma$ and $\varepsilon$ are the actual local stress and strain (related to each other through the cyclic flow stress curve), $K_t$ is the elastic stress concentration factor, and $S$ is the remote nominal stress. Because $\sigma$ and $\varepsilon$ are coupled through the material’s flow stress behavior, this equation must be solved simultaneously with the cyclic stress-strain relationship — flow stress is the link that makes notch fatigue life prediction from elastically computed FE stresses possible without requiring a full elastic-plastic FE analysis.

Flow Stress in Fracture Mechanics

Crack-Tip Plastic Zone Size

In linear elastic fracture mechanics (LEFM), the first-order estimate of the plastic zone size ahead of a crack tip is derived directly from the flow stress (commonly approximated as the yield stress in this context):

$$r_p = \frac{1}{2\pi}\left( \frac{K}{\sigma_{ys}} \right)^2 \quad \text{(plane stress)}$$

$$r_p = \frac{1}{6\pi}\left( \frac{K}{\sigma_{ys}} \right)^2 \quad \text{(plane strain)}$$

The flow stress here sets the boundary between the LEFM regime (small-scale yielding, $r_p$ small relative to specimen and crack dimensions) and the elastic-plastic fracture mechanics (EPFM) regime, where plasticity is too extensive for $K$ alone to characterize the crack-tip field.

The HRR Field and J-Integral

In EPFM, the Hutchinson-Rice-Rosengren (HRR) singularity describes the crack-tip stress field for a power-law hardening material, with the flow stress curve’s $K$ and $n$ parameters appearing directly in the field equations. The J-integral, the EPFM analog of the stress intensity factor, is computed from the strain energy density ahead of the crack — connecting this discussion back to the strain energy density concepts covered in a previous post — and its relationship to crack-tip opening displacement (CTOD) is mediated by the flow stress:

$$J = m \cdot \sigma_{\text{flow}} \cdot \delta$$

where $\delta$ is CTOD and $m$ is a dimensionless constraint factor (typically 1–2, depending on hardening exponent and stress state). This relationship is the basis for converting between CTOD-based and J-based fracture toughness test results in standards such as ASTM E1820 and BS 7448.

Limit Load and Net-Section Yielding

For fully plastic or net-section yielding fracture conditions — common in ductile pipeline steels, pressure vessels, and other tough, lower-strength structural materials — failure assessment diagrams (FAD methods, as in BS 7910 and the R6 procedure) require an estimate of the limit load, which is itself computed using the flow stress (commonly the yield-ultimate average, as noted above) rather than yield strength alone, since plastic collapse involves strain levels well beyond initial yield.

Why the Yield-Ultimate Average Is Used in Fracture Mechanics

The $\sigma_{\text{flow}} = (\sigma_y + \sigma_u)/2$ approximation deserves a closer look because it appears so frequently in fracture standards (ASTM E1820 limit load equations, R6/FAD assessments). Crack-tip plasticity involves a range of plastic strains, from very small near the elastic-plastic boundary to substantial strain directly at the crack tip. Using the initial yield stress alone underestimates the material’s actual flow resistance across this strain range, while using the ultimate tensile strength alone overestimates it (since UTS corresponds to strain levels associated with the onset of necking, well beyond what is typically relevant at the crack tip in a well-behaved ductile material). The arithmetic average provides a single representative value that has been validated empirically against more detailed elastic-plastic finite element results for a wide range of structural steels.

Practical Implications for Structural Assessment

• Notch fatigue life prediction via Neuber’s rule or the related Glinka equivalent strain energy density method requires the cyclic — not monotonic — flow stress curve as input; using monotonic properties for cyclically softening or hardening materials introduces systematic error.
• Fracture toughness testing and FAD assessments require a defensible flow stress estimate; for materials with limited tensile data, the yield-ultimate average is the standard fallback, but full elastic-plastic FE analysis using the actual hardening curve is preferred when high-consequence structural margins are at stake.
• Dynamic and impact assessments require flow stress models (Johnson-Cook, Cowper-Symonds, Zerilli-Armstrong) that explicitly separate strain-hardening, rate-hardening, and thermal-softening contributions, since static flow stress curves do not capture the elevated, less ductile behavior relevant to shock and impact loading.

Summary

Flow stress — the material’s instantaneous resistance to continued plastic deformation — is a more fundamental and more broadly applicable concept than yield strength alone, and it underlies a surprising range of structural assessment tools: cyclic stress-strain behavior and Neuber’s rule in notch fatigue, the HRR field and J-CTOD relationship in elastic-plastic fracture mechanics, limit load estimation in failure assessment diagrams, and rate-dependent constitutive models for impact and crash analysis. Recognizing flow stress as the common thread connecting these methods helps clarify why certain approximations (like the yield-ultimate average) are used, and when a more detailed hardening curve is warranted instead.

References

Hollomon, J.H. (1945). Tensile deformation. Transactions of the AIME, 162, 268–290.

Ramberg, W., Osgood, W.R. (1943). Description of stress-strain curves by three parameters. NACA Technical Note 902.

Neuber, H. (1961). Theory of stress concentration for shear-strained prismatical bodies with arbitrary nonlinear stress-strain law. Journal of Applied Mechanics, 28(4), 544–550.

Hutchinson, J.W. (1968). Singular behaviour at the end of a tensile crack in a hardening material. Journal of the Mechanics and Physics of Solids, 16(1), 13–31.

Rice, J.R., Rosengren, G.F. (1968). Plane strain deformation near a crack tip in a power-law hardening material. Journal of the Mechanics and Physics of Solids, 16(1), 1–12.

ASTM E1820. Standard Test Method for Measurement of Fracture Toughness. ASTM International.

BS 7910. Guide to Methods for Assessing the Acceptability of Flaws in Metallic Structures. British Standards Institution.

Johnson, G.R., Cook, W.H. (1983). A constitutive model and data for metals subjected to large strains, high strain rates and high temperatures. Proceedings of the 7th International Symposium on Ballistics, 541–547.

Dowling, N.E. (2012). Mechanical Behavior of Materials, 4th Edition. Pearson.

Tom Irvine | VibrationData.com | Structural Dynamics, Shock, Vibration & Acoustics