Johnson-Cook Constitutive Plasticity Model Calculator

Comprehensive Johnson-Cook Dynamic Plasticity Constitutive Model Guide

Predicting the non-linear mechanical response of metallic structures subjected to violent, high-energy dynamic loading—such as projectile impact, hypervelocity penetration, automotive crash, aerospace bird strikes, and metal machining chip formation—requires constitutive formulations that capture the complex interplay between strain hardening, strain rate sensitivity, and thermal softening. The Johnson-Cook model, published in 1983 by Gordon R. Johnson and William H. Cook at Honeywell and the US Air Force Armament Laboratory, is universally adopted in modern computational mechanics.

1. Mathematical Formulation and Multiplicative Decomposition

The Johnson-Cook constitutive law represents the von Mises equivalent flow stress $sigma$ as an uncoupled, multiplicative product of three distinct physical phenomena:

where the terms are physically defined as follows:

A. Strain Hardening Term ($left[ A + B varepsilon_p^n right]$)

Derived from classical Ludwik-Hollomon empirical power-law hardening:

• $A$ [$text{MPa}$]: The initial quasi-static yield strength at the reference temperature $T_{room}$ and reference strain rate $dot{varepsilon}_0$.
• $B$ [$text{MPa}$]: The strain hardening coefficient representing the slope of plastic work-hardening.
• $n$ [dimensionless]: The strain hardening exponent (typically $0.15 le n le 0.45$). If $n = 0$, the material exhibits elastic-perfectly plastic behavior.
• $varepsilon_p$: The equivalent effective plastic strain.

B. Strain Rate Sensitivity Term ($left[ 1 + C ln(dot{varepsilon}^*) right]$)

Accounts for dislocation velocity drag and thermal activation barrier hopping at elevated deformation rates:

• $dot{varepsilon}^* = frac{dot{varepsilon}}{dot{varepsilon}_0}$: Dimensionless normalized strain rate.
• $dot{varepsilon}_0$: Calibration reference quasi-static strain rate (standardized at $1.0text{ s}^{-1}$ or $10^{-3}text{ s}^{-1}$).
• $C$: Dimensionless strain rate sensitivity coefficient (typically $0.005 le C le 0.05$).

For quasi-static loading where $dot{varepsilon} le dot{varepsilon}_0$, the logarithmic term is bounded to zero ($dot{varepsilon}^* ge 1 implies ln(dot{varepsilon}^*) ge 0$) to prevent unphysical strength reductions below the static yield strength.

C. Thermal Softening Term ($left[ 1 - (T^*)^m right]$)

Accounts for thermal activation aiding dislocation climb and glide, weakening crystal lattice resistance as temperatures approach melting:

• $T^* = frac{T - T_{room}}{T_{melt} - T_{room}}$: Dimensionless **homologous temperature** ($0 le T^* le 1$).
• $T_{room}$: Reference ambient room temperature (standardized at $293.15text{ K} = 20^circtext{C}$).
• $T_{melt}$: Liquidus melting temperature of the alloy in Kelvin.
• $m$: Thermal softening exponent. When $m = 1$, softening is strictly linear; when $m < 1$, rapid softening occurs at initial temperature increases.

2. Adiabatic Heating & The Taylor-Quinney Coefficient

During dynamic deformation at strain rates exceeding $dot{varepsilon} ge 10^2text{ s}^{-1}$, the plastic deformation timescale ($t sim 10^{-4}text{ s}$) is vastly shorter than the thermal diffusion timescale across the specimen. The process becomes **adiabatic**: virtually none of the dissipated plastic work escapes via conduction.

The mechanical plastic work per unit volume converted into thermal energy is governed by the **Taylor-Quinney coefficient ($beta$)**:

where:

• $beta approx 0.90$ (typically 90% of plastic work is dissipated as heat, with 10% stored in crystal dislocation lattice defects).
• $rho$ is material density ($text{kg/m}^3$).
• $c_p$ is specific heat capacity ($text{J}/(text{kg}cdottext{K})$).

This self-induced heating raises local temperatures by hundreds of degrees, triggering severe thermal softening that precipitates catastrophic **adiabatic shear bands (ASBs)** during ballistic impact and high-speed metal cutting.

3. Standard Material Parameter Reference Library

4. Step-by-Step Worked High-Strain-Rate Calculation

Evaluate the dynamic flow stress of Titanium Ti-6Al-4V under a high-speed ballistic impact event:

• Parameters: $A = 792text{ MPa}$, $B = 510text{ MPa}$, $n = 0.26$, $C = 0.014$, $m = 1.03$, $T_{melt} = 1793text{ K}$
• Current state: Equivalent plastic strain $varepsilon_p = 0.25$
• Dynamic strain rate $dot{varepsilon} = 2,500text{ s}^{-1}$ ($dot{varepsilon}_0 = 1.0text{ s}^{-1}$)
• Current temperature $T = 350text{ K}$ (moderate prior adiabatic heating, $T_{room} = 293text{ K}$)

Step 1: Compute Strain Hardening Component

Step 2: Compute Strain Rate Multiplier

Step 3: Compute Thermal Softening Term

Step 4: Calculate Total Flow Stress

Notice that despite moderate thermal softening ($0.9652$), strain hardening ($+45%$) and high-rate sensitivity ($+11%$) elevate the material's dynamic flow stress to $1,229text{ MPa}$, a $55%$ increase over static yield.

5. Extension: Johnson-Cook Dynamic Fracture Damage Model

In hydrocodes, the plasticity model is paired with the Johnson-Cook fracture model to simulate spallation, armor perforation, and petaling. The equivalent plastic strain at failure $varepsilon_f$ is:

where $eta = sigma_m / sigma_{eq}$ is the **stress triaxiality**. Cumulative damage $D = sum frac{Delta varepsilon_p}{varepsilon_f}$ tracks microvoid coalescence; when $D = 1.0$, element erosion is executed.

6. Experimental Calibration via Split-Hopkinson Pressure Bar (SHPB)

Extracting the five Johnson-Cook parameters requires an interconnected sequence of specialized mechanical testing machines:

• Quasi-Static Tensile/Compression Testing ($10^{-4}text{ to }10^{-1}text{ s}^{-1}$): Conducted on screw-driven universal testing machines at room temperature. The initial yield offset ($varepsilon_p = 0$) defines parameter $A$. Fitting the true stress-plastic strain curve up to necking isolates hardening modulus $B$ and exponent $n$ via linear regression on $ln(sigma - A) = ln B + n ln varepsilon_p$.
• Split-Hopkinson Pressure Bar (Kolsky Bar, $10^2text{ to }10^4text{ s}^{-1}$): A gas gun propels a striker bar against an incident elastic bar, launching a 1D compressive stress wave through a miniature cylindrical specimen sandwiched against a transmitter bar. Strain gauge records on the elastic bars determine dynamic flow stresses across several high strain rates. Plotting $sigma / (A + Bvarepsilon_p^n)$ against $ln(dot{varepsilon}^*)$ yields the slope $C$.
• High-Temperature Dynamic SHPB ($300text{ K to }0.8 T_{melt}$): High-frequency induction heating or radiant infrared furnace setups rapidly heat the specimen immediately prior to pneumatic bar contact. Testing at multiple elevated temperatures isolates the softening exponent $m$ from $ln[1 - sigma / (sigma_{ref}(1 + Clndot{varepsilon}^*))] = m ln T^*$.

7. Modified Johnson-Cook Variants (Rule-Jones & Calamaz Models)

While the standard 1983 model offers exceptional computational efficiency, experimental mechanics has revealed limitations in specific extreme regimes:

• Strain Rate Saturation: Dislocation drag transitions to relativistic phonon drag above $dot{varepsilon} > 10^5text{ s}^{-1}$. The modified **Rule-Jones model** replaces logarithmic scaling with an exponential power-law: $f_{rate} = [1 + dot{varepsilon}^*]^C$ or adds a high-rate quadratic enhancement term $C_2 [ln(dot{varepsilon}^*)]^2$.
• Coupled Strain Hardening and Thermal Softening: In machining titanium alloys, cyclic adiabatic shear bands induce localized chip serration. The **Calamaz model (2008)** introduces a hyperbolic tangent strain-softening function $D(varepsilon_p, T) = 1 - frac{p tanh(q varepsilon_p)}{1 + (T^*)^r}$ that captures dynamic recrystallization (DRX) at elevated cutting speeds.

8. Benchmark Taylor Anvil Impact Test Validation

The definitive empirical validation of Johnson-Cook parameter sets is the **Taylor cylinder impact test (1948)**. A flat-ended cylindrical projectile of length $L_0$ and diameter $D_0$ is fired at high velocity ($V_0 approx 150 dots 350text{ m/s}$) against a rigid, unyielding armor plate anvil. The resulting deformed mushroomed length $L_f$, foot diameter $D_f$, and plastic wave front boundary provide an integrated test covering strains from 0 to 1.5, strain rates up to $10^5text{ s}^{-1}$, and adiabatic heating exceeding several hundred degrees. An accurate Johnson-Cook parameter set reproduces the final profile within 2% to 3% in hydrocode simulations.