4.6 LEFM Fracture Toughness, Ductile-Brittle Transition & Larson-Miller Creep

4. Linear Elastic Fracture Mechanics (LEFM) & Plane Strain Toughness

LEFM quantifies the magnitude of the asymptotic elastic stress field surrounding a sharp crack tip in terms of a single scalar parameter: the Stress Intensity Factor ($K$).

     Mode I: Opening                  Mode II: Sliding               Mode III: Tearing
     (Tensile / Normal)               (In-Plane Shear)               (Out-of-Plane Shear)

          ^ σ                              --->                           (+)(-) 
          |                                --->                          Anti-plane
       +--+--+                          +------+                       +--------+
       |     |                          |      |                       |        |
       |  a  |                          |  a   |                       |   a    |
  <----+--x--+---->                <----+--x---+---->             <----+---x----+---->
       |     |                          |      |                       |        |
       +--+--+                          +------+                       +--------+
          |                                <---                           (-)(+) 
          v σ                              <---                          

Stress Intensity Formulation (Mode I)

Mode I (tensile opening) is the dominant failure mode in structural weldments:

$$K_I = Y \sigma \sqrt{\pi a}$$

where:

• $\sigma$ = Nominal tensile stress perpendicular to the crack plane.
• $a$ = Characteristic crack dimension (flaw depth for surface/edge cracks; half-length for internal embedded cracks).
• $Y$ = Geometric boundary correction factor ($Y = 1.12$ for an edge crack in a semi-infinite plate; $Y = 1.0$ for an internal center crack in an infinite sheet).

Fracture Criterion & Plane Strain Toughness ($K_{Ic}$)

Catastrophic brittle fracture occurs when the applied stress intensity reaches or exceeds the material's critical fracture toughness:

$$K_I \ge K_c$$

When plate thickness $B$ is sufficient to enforce plane strain conditions across the crack front, fracture toughness reaches an invariant minimum value known as the Plane Strain Fracture Toughness ($K_{Ic}$). $K_{Ic}$ is a fundamental material property (units: $\text{MPa}\sqrt{\text{m}}$ or $\text{ksi}\sqrt{\text{in}}$). Per ASTM E399, valid plane strain requires:

$$B, a, (W - a) \ge 2.5 \left( \frac{K_{Ic}}{\sigma_y} \right)^2$$

Crack-Tip Plastic Zone Size ($r_y$)

Because infinite stress cannot exist at an actual atomic crack tip, a localized zone of plastic yielding forms. Irwin's plastic zone radius is:

$$r_y = \frac{1}{2\pi} \left( \frac{K_I}{\sigma_y} \right)^2 \quad (\text{Plane Stress})$$ $$r_y = \frac{1}{6\pi} \left( \frac{K_I}{\sigma_y} \right)^2 \quad (\text{Plane Strain})$$

The plane strain plastic zone is one-third the size of the plane stress zone, explaining the sharp reduction in energy absorption in thick plates.

---

5. Ductile-to-Brittle Transition Temperature (DBTT) & Charpy Testing

Ferritic steels undergo a drastic transformation in fracture mechanism across a narrow operating temperature band.

    Absorbed Energy (Joules)
       ^
       |                          Upper Shelf Energy (USE): Ductile Microvoid
  USE -+ - - - - - - - - - - - - .------------------- Coalescence (Dimple Rupture)
       |                        / 
       |                       /  Transition Region (DBTT)
       |                      /   Mixed Cleavage + Shear Lips
 27 J -+ - - - - - - - - - - / - - - - - - - - - Standard Structural Code Threshold
       |                    / 
  LSE -+ - - - - - - - -.../ Lower Shelf Energy (LSE): Transgranular Cleavage
       +-------------------+--------------------------> Temperature (°C)
                         T_DBTT

Crystallographic Origin: BCC vs. FCC

• Body-Centered Cubic (BCC) Steels: Dislocation glide occurs on ${110}$, ${112}$, and ${123}$ planes. These slip systems exhibit a steep temperature-dependent Peierls-Nabarro lattice friction stress. At low temperatures, thermal energy is insufficient to assist dislocation movement; yield stress rises rapidly, exceeding the critical stress for atomic bond cleavage ($\sigma_f^*$). Cleavage occurs along low-index ${100}$ crystallographic planes, producing bright, reflective river-pattern facets with near-zero plastic deformation.
• Face-Centered Cubic (FCC) Alloys: Austenitic stainless steels (AISI 304L, 316L), aluminum, and nickel alloys possess close-packed ${111}$ planes with very low Peierls-Nabarro stress. Dislocation mobility remains high even at cryogenic temperatures (down to $4\text{ K}$), meaning FCC metals do not exhibit a DBTT and remain ductile at all temperatures.

Charpy V-Notch (CVN) Testing (ASTM E23 / AWS B4.0)

A standard $10\text{ mm} \times 10\text{ mm} \times 55\text{ mm}$ specimen with a $2\text{ mm}$ deep, $45^\circ$ V-notch (root radius $0.25\text{ mm}$) is struck by a swinging pendulum hammer ($300\text{ J}$ capacity). Key transition criteria include:

• 27 Joules ($20\text{ ft}\cdot\text{lbf}$): Standard minimum design energy required by AWS D1.1, ASME Section VIII, and API 5L at the Minimum Design Metal Temperature (MDMT).
• 50% Fracture Appearance Transition Temperature (FATT): Temperature where the fractured surface displays exactly $50%$ fibrous ductile shear lips and $50%$ crystalline cleavage.
• Lateral Expansion ($0.90\text{ mm} = 35\text{ mils}$): Direct measure of plastic deformation at the specimen base.

HAZ Local Brittle Zones (LBZs)

In multi-pass welded joints, thermal cycling produces microstructural zones highly susceptible to brittle cleavage:

• Coarse-Grained HAZ (CGHAZ): Peak temperatures $T > 1100^\circ\text{C}$ dissolve grain-pinning carbonitrides, coarsening prior austenite grains ($> 100\ \mu\text{m}$). Upon cooling, coarse upper bainite and grain-boundary ferrite form, degrading CVN toughness.
• Intercritically Reheated CGHAZ (ICCGHAZ): Subsequent weld passes reheat the CGHAZ into the dual-phase $\alpha + \gamma$ region ($750\text{--}800^\circ\text{C}$). Carbon enriches the newly formed austenite, which transforms upon cooling into brittle Martensite-Austenite (M-A) constituent islands along prior austenite grain boundaries, creating deadly fracture initiation sites.

---

6. High-Temperature Creep Deformation & The Larson-Miller Parameter

Creep is the time-dependent, permanent inelastic deformation that accumulates under sustained mechanical stress at elevated homologous temperatures, generally occurring when $T > 0.40\text{ to }0.50 T_m$ (where $T_m$ is absolute melting temperature in Kelvin). In structural steels, creep becomes active above $370^\circ\text{C}$ ($700^\circ\text{F}$); in Cr-Mo power-plant steels, it dominates above $450^\circ\text{C}$ ($840^\circ\text{F}$).

    Creep Strain (ε)
       ^
       |                                   Tertiary Creep: Microvoid Cavitation,
       |                                   Necking, Accelerating to Rupture
       |                                           /
       |                 Secondary (Steady-State):/
       |                 dε/dt = ε_ss = const    /
       |                 Balance of Hardening   /
       |                 and Dynamic Recovery  /
       |            /-------------------------
       |           /  Primary (Transient) Creep:
       |          /   Decelerating Strain Rate
       |         /
       +--------+---------------------------------------------> Time (t)
       0       t_1                         t_2                t_r (Rupture)

The Three Creep Regimes

1. Primary (Transient) Creep: Strain rate decelerates over time ($d^2\varepsilon/dt^2 < 0$) as dislocation multiplication and entanglement produce strain hardening.
2. Secondary (Steady-State) Creep: A thermodynamic balance is established between strain hardening and dynamic thermal recovery (dislocation climb past obstacles via vacancy diffusion). The creep rate reaches a constant minimum $\dot{\varepsilon}_{ss}$, modeled by Norton's Power Law: $$\dot{\varepsilon}_{ss} = A \sigma^n \exp\left( -\frac{Q_c}{R T_K} \right)$$ where $n$ is the creep stress exponent ($n = 4\text{--}7$ for dislocation climb-glide), $Q_c$ is the activation energy for creep (frequently matching self-diffusion activation energy $Q_{\text{diff}}$), $R = 8.314\text{ J/(mol}\cdot\text{K)}$, and $T_K$ is temperature in Kelvin.
3. Tertiary Creep: Dislocation-assisted grain boundary sliding leads to micro-cavity nucleation at transverse grain boundaries. Cavities coalesce into grain-boundary microcracks, effective load-bearing area drops, strain rate accelerates rapidly, and rupture occurs at time $t_r$.

Type IV Cracking in Creep-Strength Enhanced Ferritic (CSEF) Steels

In Grade 91 ($9\text{Cr-1Mo-V}$) and Grade 92 power-piping weldments, creep failures occur preferentially via Type IV Cracking in the fine-grained HAZ (FGHAZ) and intercritical HAZ (ICHAZ). Peak welding temperatures in these zones ($850\text{--}950^\circ\text{C}$) partially austenitize the steel, causing over-tempering and dissolution of fine strengthening precipitates ($\text{MX}$ carbonitrides, $\text{M}_{23}\text{C}_6$ carbides). Under service steam pressure, this narrow softened zone undergoes severe localized multiaxial strain, accelerating grain-boundary cavitation and producing premature catastrophic rupture within $30,000\text{ to }80,000\text{ operating hours}$.

The Larson-Miller Parameter ($LMP$)

Because engineering components are designed for $100,000\text{ to }300,000\text{ hours}$ ($11\text{ to }34\text{ years}$) of service, testing under exact operational conditions is impractical. The Larson-Miller Parameter relates rupture time $t_r$ and absolute temperature $T$ to establish master parametric curves:

$$LMP = T_K \left[ C + \log_{10}(t_r) \right] \times 10^{-3}$$

where:

• $T_K$ = Absolute temperature in Kelvin ($T_K = ^\circ\text{C} + 273.15$) or Rankine ($T_R = ^\circ\text{F} + 459.67$).
• $t_r$ = Rupture time in hours.
• $C$ = Dimensionless material constant (typically $C = 20$ for ferritic, martensitic, and low-alloy steels; $C = 15$ for aluminum alloys).

By evaluating the $LMP$ corresponding to a design stress $\sigma$ from a master material curve, high-temperature, short-duration accelerated tests (e.g., $1,000\text{ hours}$ at $650^\circ\text{C}$) can reliably predict long-term operational life at lower temperatures (e.g., $540^\circ\text{C}$). Solving for rupture time $t_r$:

$$\log_{10}(t_r) = \frac{1000 \cdot LMP}{T_K} - C \implies t_r = 10^{\left( \frac{1000 \cdot LMP}{T_K} - C \right)}$$

---