Notes on FEM modelling of residual stresses in Ti-6Al-4V micro-turning (scale effect included)

Notes on FEM modelling of residual stresses in Ti-6Al-4V during micro-turning (scale effect included)

  • Source context

    • Study focuses on three-dimensional finite element modelling of micro-turning Ti-6Al-4V with coupled thermo-mechanical transient analysis.

    • Aims to predict surface and sub-surface residual stresses (axial, radial, circumferential) during micro-turning, incorporating scale effects via strain-gradient considerations.

    • Strain-gradient modification applied to Johnson–Cook (JC) flow model to capture scale effects; JC damage model used for material separation.

    • Nano-indentation used for validation of residual stress predictions.

  • Key findings (highlights)

    • At low feed, radial and axial residual stresses shift from compressive to tensile at a depth of about 12 μm due to rubbing/ploughing and heat generation.

    • Compressive residual stress increases with feed rate (due to higher strain rate and chip load); circumferential (hoop) residual stress is markedly higher (~10^3 MPa) than radial or axial stresses.

    • At low feed (3 μm/rev), maximum axial residual stress occurs at about 4 μm from the surface due to dislocation density accumulation from scale effects; at 12 μm depth, axial stress changes to tensile.

    • Nano-indentation measurements show good agreement for maximum compressive circumferential residual stress; deviations attributed to pile-up/sink-in effects during indentation.

    • Overall, a 3D FEM approach with scale-aware constitutive modelling captures depth-varying residual stresses and chip morphology under micro-turning conditions.

  • Methods: modelling approach and domain

    • Process: micro-turning of Ti-6Al-4V with a WC tool treated as rigid; workpiece modeled as elastic–visco-plastic.

    • Geometry: machined domain width = 50 μm (depth of cut), length of cut = 50 μm; tool rake angle = 7°, clearance angle = 7°, nose radius = 3 μm.

    • Mesh and elements: C3D8RT (eight-node thermally coupled brick elements); minimum mesh size = 2 μm; total ~130,000 elements; reduced integration to ease computational cost; hourglass control factor = 0.5.

    • Time integration: explicit time stepping; equation of motion at nodes given by Ma + Fint(n) = Fext(n) with explicit discretization.

    • Time-step stability: Δt = Lc / x0, where Lc is the minimum element length and x0 is the wave speed of the deformed material.

    • Boundary conditions: surface-to-node contact between tool and workpiece; rigid tool reference point; friction coefficient μ = 0.24.

    • Thermal aspects: convection with h = 40 kW/m^2K on the top surface and chip outer surface; conduction within the shear zone and workpiece; diffusion neglected.

    • Cutting conditions studied: speeds = 60 m/min; feeds = 3, 10, 20 μm/rev; depth of cut = 50 μm.

  • Material models and constitutive laws

    • Workpiece material: Ti-6Al-4V (Ti-6Al-4V alloy, ASTM Grade 5); WC tool included for heat transfer boundaries; temperature-dependent properties provided in the study.

    • Tool properties: WC assumed to be rigid in the model; focus is on workpiece thermo-mechanical response.

    • Johnson–Cook (JC) flow model for flow stress σr: ilde{r} = (A + B ilde{oldsymbol{ ho}}^{n})\, iggl(1 + C \,\ln\frac{\dot{\varepsilon}}{\dot{\varepsilon}0}\biggr)\Bigl(1 - T^\Bigr)^m, T^ = \frac{T - T{\mathrm{room}}}{T{\mathrm{melt}} - T_{\mathrm{room}}}

    • Parameters (typical JC form): A, B, n, C, m; Troom = 25°C; Tmelt ≈ 1600°C (values provided in the paper’s table).

    • Scale effect: strain-gradient plasticity is introduced to modify JC flow to capture micro-scale size effects.

    • Gradient-modified flow stress, with gradient length and geometry-dependent terms (Equations (4)–(7) in the paper):
      r = r{JC} \,\sqrt{1 + 18 a^2 G^2 b \, / \, (r{JC} L)},
      L = h\sin\alpha,
      \tan\alpha = \frac{r\cos\alpha}{1 - \sin\alpha},
      r = h t.

    • Here r_JC denotes the conventional JC flow stress, L is the length of the primary shear zone, h is a dimension (e.g., depth of cut), α is a geometry-related angle, a, G, b, t are additional material/geometry constants as defined in the study. The exact forms are as given in the article (Equations (4)–(7)).

    • JC damage model for material separation (damage initiation and evolution):

    • Damage initiation criterion uses an equivalent fracture strain ef expressed through parameters D1–D5 and stress triaxiality g, along with strain rate and temperature terms.

    • The general form used in the study (Equation (8)) is reported as:
      ef = \left[ D1 + D2 \, e^{D3} \right] \, (1 + D4 \ln\frac{\dot{\varepsilon}}{\dot{\varepsilon}0}) \, \left(1 + D5 \frac{T - T{room}}{T{melt} - T{room}} \right),
      D1, D2, D3, D4, D5 \text{ are JC damage parameters (Table 3)}.

    • Friction model at tool–workpiece interface:

    • Coulomb friction with μ = 0.24; stick-slip behavior is captured via the relationship:
      s{crit} = K = 0.577 \; ry,
      s = K, \ \text{Slipping zone} , \ s = l p s, \ \text{Sticking zone} .

    • Where ry is material yield strength and ps is contact pressure (as defined in the paper’s Figure/section).

    • Material properties (Table references in the paper): temperature-dependent elastic moduli and thermal properties for Ti-6Al-4V and WC (details provided in the paper’s Table 1).

  • Boundary conditions and heat transfer details

    • Surface-to-node contact between tool and workpiece with a frictional interface (μ = 0.24).

    • Convective boundary condition on exposed surfaces: h = 40 kW/m^2K on the top surface and outer surface of the chip.

    • Heat transfer to/from shear zone to workpiece/tool via conduction; diffusion is neglected in this analysis.

  • Numerical solution and FE settings

    • Explicit time integration to handle highly nonlinear thermo-mechanical coupling during cutting.

    • Governing equations discretization follows the weak form of mechanical and thermal equilibrium; for each node the discrete form is:
      M a + F{int}(n) = F{ext}(n) \quad (t+1)
      where M is mass matrix, a is nodal acceleration, Fint is internal force, Fext is external force.

    • Time-step stability criterion used as:
      \Delta t = \frac{Lc}{x0},
      where Lc is the minimum element length and x0 is the wave speed of the material.

  • Experimental validation setup (nano-indentation)

    • Nano-indentation (Berkovich tip) used to measure maximum compressive residual stress in the machined Ti-6Al-4V.

    • Indentation protocol follows Suresh and Giannakopoulos approach for residual-stress-assisted indentation (Reference [27]).

    • A Punch/Punch-like analysis is used to relate indentation load and depth (Kick’s law) and to determine residual stress via a sphere/offset contact model with an indenter angle a.

    • Idealized relationships (simplified):

    • Kick’s law for sharp indentation: P = C h^2,

    • Equivalent biaxial residual stress contribution and adjusted indentation depth differences in stressed vs. unstressed zones are used to extract r_{R X,0}.

    • Equations (11)–(24) in the paper relate indentation load P, penetration h, geometry factors (indenter angle a, contact area A), and residual stress r_{R X,0} along X-direction (circumferential hoop stress) to estimate the compressive residual stress.

  • Experimental validation: micro-turning experiment details

    • Equipment: Mikrotools DT110 micro-machining center (Singapore).

    • Workpiece: Ti-6Al-4V hollow cylindrical rods (outer diameter ~7 mm, wall thickness ~3 mm, length ~200 mm).

    • Cutting tool: Tungsten carbide (WC).

    • Surface preparation: Polished to low roughness (grades P1200 to P3000) and electropolished to remove native oxide.

    • Indentation/indentation results: indentation tests performed in stressed and unstressed zones; maximum circumferential residual stress from nano-indentation around 1463 MPa experimentally, while the FEM predicted ~1000 MPa (deviation ~31% attributed to pile-up/sink-in effects and model assumptions).

    • Comparison: FEM results for the maximum circumferential residual stress at feed 20 μm/rev, Vc = 60 m/min show good trend agreement with experimental data from Hascelik et al./other micro-turning studies, noting differences due to strain-softening effects not captured by JC, etc.

  • Results: residual stress, temperature, and chip morphology

    • Contour plots for a representative case (feed = 20 μm/rev):

    • Heat flux (due to plastic deformation) distributed in the primary shear zone; maximum Von Mises stress along the shear plane in the deformation zone.

    • Peak cutting temperature around 599°C at a certain distance from the cutting edge (chip–tool interface region).

    • Plastic strain distribution indicating localized plastic deformation in the primary shear zone.

    • Effect of feed rate on residual stresses and surface state:

    • Lower feed (3 μm/rev): heat transfer per unit area is higher at low chip load, leading to higher temperature near the interface; axial residual stress more prone to become tensile at greater depths due to rubbing/ploughing.

    • Higher feed increases chip load and friction, increasing cutting forces and circumferential residual stress; hoop residual stress remains the dominant component (≈10^3 MPa range) compared with radial and axial stress.

    • Depth-resolved residual stress patterns:

    • Circumferential residual stress is most compressive near the surface and remains substantial in subsurface layers.

    • Maximum axial residual stress at low feed (3 μm/rev) occurs at about 4 μm depth; at deeper depths (≈12 μm) the axial stress tends to center around tensile state due to size effects and rubbing/ploughing interactions.

    • Depth profiles: residual stress vs depth show a transition from compressive to tensile in axial and radial directions at around 12 μm depth under low-feed conditions; the circumferential stress remains largely compressive and high.

  • Validation and key conclusions

    • The 3D FEM framework with scale-aware JC and damage models successfully predicts surface/subsurface residual stresses and their depth profiles under micro-turning of Ti-6Al-4V.

    • The scale effect (via strain-gradient modification) is essential to capture the observed size-dependent strengthening and the depth distribution of residual stress.

    • The nano-indentation-based residual-stress validation shows reasonable agreement (within ~31% for the maximum compressive circumferential residual stress), supporting the modelling approach.

    • Conclusions drawn by the authors:

    • Circumferential residual stress is significantly higher than radial/axial due to dominant circumferential force components during micro-turning.

    • Increasing feed strengthens compressive circumferential residual stress due to higher chip load and deformation.

    • Axial and radial residual stresses can become tensile at depth (~12 μm) due to rubbing/ploughing and strain-gradient effects.

    • The nano-indentation-based validation provides support for the predicted magnitude and depth location of residual stresses, with acknowledged limitations from pile-up/sink-in effects and model assumptions.

  • Notable references and context

    • The study builds on broader micro-machining/residual-stress literature: 2D orthogonal models, edge-radius effects, and experimental validations have shown that tool geometry, lubrication, cutting parameters, and scale effects strongly influence residual stress distributions.

    • Prior works cited include analytical strain-gradient models, JC-based mdoelling, and finite-element validations of surface residual stresses in Ti alloys (Ti-6Al-4V) and other materials.

    • The work emphasizes the importance of capturing scale effects at micro-scale for accurate residual-stress predictions, beyond conventional 2D plane-strain or purely macroscopic models.

  • Practical implications and future scope

    • The approach provides a framework to predict surface integrity in micro-machining of Ti-6Al-4V, relevant for precision micro-components (e.g., micro screws, micro gears, micro pins).

    • The model can help in optimizing process parameters (feed, speed) to minimize detrimental residual stresses and improve fatigue life.

    • Future improvements suggested include incorporating strain-softening behaviour, enhancing friction/contact models, and refining gradient-length formulations for better accuracy.

  • Tables and figures (as referenced in the study)

    • Table 1: Material properties of Ti-6Al-4V (workpiece) and WC (tool) with temperature-variant values (density, elastic modulus, Poisson ratio, thermal conductivity, specific heat, inelastic heat fraction).

    • Table 2: Johnson–Cook material parameters for Ti-6Al-4V (A, B, n, C, m, Tmelt, Troom).

    • Table 3: Johnson–Cook damage parameters for Ti-6Al-4V (D1–D5).

    • Table 4: Cutting conditions with feed variation (speed fixed at 60 m/min; depth of cut 50 μm).

    • Table 5: Cutting conditions with speed variation (feed fixed, depth of cut 50 μm).

    • Table 6: Indentation parameters from nano-indentation measurements (Young’s modulus, hardness, stiffness, maximum load, depth, contact area, residual stress).

    • Figures 1–12: Process flow, mesh/topology, mesh undercut, contact conditions, heat transfer boundaries, contour plots for heat flux, stress and temperature, cutting force trends, residual-stress depth profiles, nano-indentation schematic, and validation plots.

  • Summary takeaway for exam preparation

    • A 3D FEM model with coupled thermo-mechanical analysis can predict micro-turning residual stresses in Ti-6Al-4V when scale effects are included via strain-gradient modifications to the JC flow model and JC damage criteria are used for failure.

    • The dominant circumferential stress component arises from cutting kinematics and tool-workpiece interactions; the depth-dependent transition from compressive to tensile residual stress in axial/radial directions is tied to size effects and rubbing/ploughing phenomena.

    • Validation via nano-indentation is feasible and can quantify maximum compressive residual stress, though pile-up/sink-in effects can cause deviations from FEM predictions.

  • References (contextual)

    • The paper cites key works on micro-scale effects (strain-gradient plasticity, edge radius effects, and micro-turning residual stress modelling) and links simulation outcomes to experimental nano-indentation measurements and micro-turning experiments.

  • Important equations to remember (sampled from the paper)

    • Johnson–Cook flow stress (standard form):
      \sigmar = (A + B \varepsilon^n) \left(1 + C \ln\frac{\dot{\varepsilon}}{\dot{\varepsilon}0}\right) (1 - T^)^m, T^ = \frac{T - T{room}}{T{melt} - T_{room}}.

    • Gradient-enhanced flow stress (partial representation):
      r = r{JC} \sqrt{1 + \frac{18 a^2 G^2 b}{r{JC} L}},
      L = h \sin\alpha, \quad \tan\alpha = \frac{r \cos\alpha}{1 - \sin\alpha}, \quad r = h t.

    • JC damage initiation example form (illustrative):
      ef = \bigl[D1 + D2 e^{D3}\bigr] \left[1 + D4 \ln\frac{\dot{\varepsilon}}{\dot{\varepsilon}0}\right] \left[1 + D5 \frac{T - T{room}}{T{melt} - T{room}}\right], \text{with } D1, D2, D3, D4, D5 \text{ from Table 3}.

    • Contact mechanics and friction (stylized):
      \mu = 0.24, \quad s{crit} = K = 0.577 ry, \quad s = K \; (\text{stick}), \; s = S_l \; (\text{slip}).

    • Time-step stability (explicit):
      \Delta t = \frac{Lc}{x0}.

    • Indentation-based residual-stress relation (conceptual):
      r{R X,0} = \text{(function of P, h, a, A, }\alpha, h1, h2, P{avg}, D) }.

    • Final relation for residual stress estimation from indentation depth changes (Equation (24) in the paper):
      r{R X,0} = P{avg} \sin\alpha \Bigl[ 1 - \left( \frac{h1}{h2} \right)^2 \Bigr]^{-1}.

  • Practical exam-oriented takeaways

    • When analyzing micro-scale machining of Ti-6Al-4V, include scale effects to capture true residual-stress distributions; JC alone may underpredict or misrepresent local strengthening and damage initiation.

    • Expect a strong circumferential residual stress component near the surface, with possible transition to tensile states for axial and radial stresses at several micrometers depth due to size effects and rubbing.

    • Use nano-indentation as a complementary validation tool for surface/sub-surface residual stresses, but account for pile-up/sink-in effects in interpreting the indentation data.