Appearance
A manufacturer of BTA-drilled components for aerospace applications needs to predict cutting forces and temperature distribution when drilling Inconel 625 nickel-based superalloy (Ø30 mm bore × 500 mm depth, 35 HRC) to optimize tool geometry and prevent premature tool failure. A 3D finite element model is developed in Abaqus/Explicit using the Johnson-Cook constitutive model with element elimination for chip formation. The simulation includes both cutting edges and guide pads with Coulomb friction (µ = 0.32), coupled temperature-displacement analysis, and continuous remeshing. Simulated results predict a maximum cutting edge temperature of 623 °C, feed force of 4,850 N, torque of 185 N·m, and chip morphology showing segmented chips at 0.18 mm/rev feed. The guide pad friction zone reaches 780 °C due to burnishing contact with the bore wall. Experimental validation using a Kistler dynamometer and thermocouple-instrumented BTA drill head shows prediction errors of 8.4% for torque and 4.7% for feed force — confirming the model's suitability for virtual tool geometry optimization and parameter selection.
Simulation Objectives in Deep Hole Drilling
| Objective | Why It Matters | Typical Output |
|---|---|---|
| Cutting force prediction | Tool design, machine power requirement, fixturing | Feed force (N), torque (N·m), radial force (N) |
| Temperature distribution | Tool wear, surface integrity, coolant requirement | Max temperature at cutting edge and guide pads |
| Chip morphology | Chip evacuation design, parameter selection | Chip shape, segmentation frequency, chip thickness |
| Residual stress | Fatigue life of drilled component, surface integrity | Surface and sub-surface residual stress profile |
| Straightness deviation | Bore quality prediction, process limits | Bore deviation per unit length |
| Tool wear estimation | Tool life prediction, replacement scheduling | Flank wear progression, crater wear depth |
| Guide pad contact pressure | Pad wear, bore surface finish | Contact pressure distribution, friction force |
Finite Element Modelling Approaches
| Approach | Software | Mesh Type | Element Count | Run Time | Accuracy |
|---|---|---|---|---|---|
| Lagrangian with remeshing | Abaqus/Explicit, Deform 3D | Hex (C3D8RT) or Tet (C3D4T) | 50,000–200,000 | 12–48 hours | High |
| Coupled Eulerian-Lagrangian (CEL) | Abaqus/Explicit | Eulerian workpiece, Lagrangian tool | 80,000–300,000 | 8–24 hours | Moderate–High |
| Arbitrary Lagrangian-Eulerian (ALE) | Abaqus/Explicit | Hex with adaptive meshing | 40,000–150,000 | 6–18 hours | Moderate |
| Element Elimination Technique (EET) | Abaqus/Explicit, Deform 3D | Hex or Tet with failure criterion | 50,000–200,000 | 12–48 hours | High (for chip form) |
| Analytical thermomechanical | MATLAB, custom code | N/A (analytical) | N/A | Minutes | Moderate (forces only) |
TIP
The choice between Lagrangian with remeshing and CEL depends on the research question. Lagrangian methods with element elimination produce the most realistic chip morphology and are better for studying chip formation mechanics. CEL methods are computationally more efficient (30–50% faster) and are preferred when the primary interest is forces and temperatures rather than chip shape. For industrial parameter optimization, start with an analytical thermomechanical model for quick force estimates, then use FE simulation for detailed analysis of critical parameter combinations.
Material Constitutive Models
Johnson-Cook Model
The Johnson-Cook (JC) constitutive model is the most widely used material model for machining simulations:
| Symbol | Parameter | Typical Value (42CrMo4) | Typical Value (Inconel 625) |
|---|---|---|---|
| A | Yield strength (MPa) | 520 | 690 |
| B | Strain hardening modulus (MPa) | 910 | 1,140 |
| n | Strain hardening exponent | 0.38 | 0.52 |
| C | Strain rate sensitivity coefficient | 0.014 | 0.008 |
| m | Thermal softening exponent | 1.03 | 1.36 |
| Tm | Melting temperature (°C) | 1,520 | 1,350 |
| Tr | Room temperature (°C) | 25 | 25 |
The JC flow stress equation:
σ = (A + Bεⁿ)(1 + C ln ε̇*)(1 − T*ᵐ)
where T* = (T − Tr) / (Tm − Tr) is the homologous temperature.
Johnson-Cook Damage Model
| Parameter | Description | Typical Value (Steel) |
|---|---|---|
| d₁ | Initial failure strain | 0.05 |
| d₂ | Exponential factor | 3.44 |
| d₃ | Triaxiality factor | −2.12 |
| d₄ | Strain rate factor | 0.002 |
| d₅ | Temperature factor | 0.61 |
Boundary Conditions and Contact Settings
| Setting | Typical Value | Purpose |
|---|---|---|
| Friction coefficient (cutting edge) | µ = 0.32 (Coulomb) | Chip-tool interface friction |
| Friction coefficient (guide pad) | µ = 0.15–0.25 | Lubricated guide pad contact |
| Heat partition (inelastic) | 90% converted to heat | Primary heat source in shear zone |
| Heat partition (friction) | 50% to workpiece, 50% to tool | Frictional heat distribution |
| Thermal conductivity (tool) | 85 W/m·K (carbide) | Heat dissipation through tool |
| Thermal conductivity (workpiece) | 45 W/m·K (steel) | Heat dissipation through workpiece |
| Convection coefficient (coolant) | 10,000–20,000 W/m²·K | Coolant heat transfer |
| Ambient temperature | 25°C | Initial condition |
Temperature Distribution Simulation
Temperature prediction is critical for understanding tool wear and surface integrity:
| Location | Temperature Range (Steel) | Temperature Range (Nickel Alloy) | Risk |
|---|---|---|---|
| Primary shear zone | 400–700°C | 500–850°C | Thermal softening, built-up edge |
| Cutting edge / rake face | 500–900°C | 600–1,050°C | Crater wear, edge deformation |
| Guide pad contact zone | 600–1,200°C | 700–1,400°C | Pad wear, bore surface damage |
| Bore wall (burnishing zone) | 300–600°C | 350–700°C | Residual tensile stress, white layer |
| Chip (bulk) | 200–400°C | 250–500°C | Chip colour changes (process monitoring) |
WARNING
Guide pad temperatures in BTA drilling simulations consistently exceed cutting edge temperatures by 100–300°C. This is because guide pads are in continuous sliding contact with the freshly machined bore wall under high normal pressure, with limited coolant access. The high guide pad temperature is the primary factor limiting BTA drilling speed in difficult-to-machine materials. Simulation studies on 42CrMo4 show guide pad temperatures reaching 1,190°C even when the cutting edge is at 600°C.
Chip Formation Simulation
| Chip Type | Formation Condition | Simulation Method | Indication |
|---|---|---|---|
| Continuous chip | Low feed, high speed, ductile material | Element elimination with JC damage | Good surface finish, but chip evacuation risk |
| Segmented / saw-tooth | High feed, moderate speed | Element elimination with high d₂ | Common in BTA, acceptable if well-formed |
| Broken / C-shaped | Moderate feed, chip breaker geometry | Element elimination with controlled damage | Ideal for chip evacuation in BTA |
| Serrated | High speed, low thermal conductivity | Coupled temperature-displacement | Common in titanium and nickel alloys |
Chip Formation Parameters
| Parameter | Continuous Chip | Segmented Chip | Broken Chip |
|---|---|---|---|
| Chip thickness (mm) | 0.3–0.6 | 0.4–0.8 | 0.2–0.5 |
| Shear angle (°) | 25–35 | 20–30 | 30–40 |
| Segmentation frequency (kHz) | None | 5–20 | N/A |
| Contact length (mm) | 1.0–2.0 | 0.8–1.5 | 0.5–1.0 |
| FE element count per chip | 500–2,000 | 500–2,000 | 200–800 |
Simulation Setup for BTA and Gun Drilling
BTA Drilling Simulation Steps
| Step | Description | Typical Duration (CPU) |
|---|---|---|
| 1 | Geometry creation: tool (cutting edges, guide pads, chip former), workpiece (cylindrical blank) | 1–2 hours (pre-processing) |
| 2 | Material assignment: JC parameters for workpiece, elastic-plastic for tool (or rigid body assumption) | 0.5 hour |
| 3 | Mesh generation: refined mesh in cutting zone (0.01–0.05 mm element size), coarse mesh elsewhere (0.5–2 mm) | 1–2 hours |
| 4 | Contact definition: chip-tool, guide pad-bore wall, self-contact for chip | 0.5 hour |
| 5 | Boundary conditions: workpiece fixed, tool rotation + feed assigned; coolant convection on exposed surfaces | 0.5 hour |
| 6 | Solution: explicit dynamics with mass scaling for computational efficiency | 12–48 hours |
| 7 | Post-processing: forces, temperature contours, chip morphology, residual stress extraction | 2–4 hours |
Gun Drilling Simulation Considerations
| Aspect | Gun Drilling Specifics |
|---|---|
| Single cutting edge | Model one cutting edge with V-shaped flute |
| Guide pad arrangement | Two guide pads (unlike BTA's multiple pads) |
| Coolant channel | Internal bore for high-pressure coolant delivery |
| Chip evacuation | V-groove along the drill shaft — not typically simulated |
| Depth-to-diameter ratio | Very high (up to 200:1) — typically simulate only the cutting zone |
TIP
Full-length simulation of a deep hole drilling operation (e.g., 500 mm bore depth) is computationally infeasible. Practical FE simulations model a short segment of the drilling process (5–20 mm of feed) under steady-state assumptions. The key is to ensure the simulation reaches thermal steady state, which typically occurs after 3–5 mm of tool travel for the cutting edge and 5–10 mm for guide pads. Extract forces, temperatures, and chip morphology from the steady-state region only.
Model Validation
| Method | What It Validates | Typical Error | Equipment |
|---|---|---|---|
| Dynamometer (Kistler) | Feed force, torque | 5–10% | Kistler 9271 or 9129AA |
| Thermocouple in tool | Cutting edge temperature | 10–15% | K-type embedded thermocouples |
| Infrared thermography | Tool and chip temperature field | 15–25% | FLIR or similar IR camera |
| Optical microscopy | Chip thickness, segmentation | 5–10% | Optical microscope + image analysis |
| SEM | Tool wear, chip morphology | Qualitative | Scanning electron microscope |
| Bore profilometry | Hole expansion, surface finish | 10–20% | Stylus profilometer, CMM |
| XRD | Residual stress | 15–25% | X-ray diffraction |
Published Validation Results
| Source | Material | Force Error | Torque Error | Temperature Error | Chip Morphology |
|---|---|---|---|---|---|
| Guan et al. (2023) | Inconel 625 + FeCr | 4.7% | 8.4% | N/A | Good qualitative match |
| Fandiño et al. (2021) | 42CrMo4 | N/A | N/A | ~15% (edge), ~25% (guide pad) | Good match at all feeds |
| TU Dortmund (2024) | Steel | ~10% | ~12% | ~10% | Good match for segmented chips |
| Haddag et al. (2020) | Mild steel 18MND5 | ~8% | ~10% | N/A (analytical model) | Analytical only |
Software Comparison for Drilling Simulation
| Software | Strengths | Limitations | Cost | Best For |
|---|---|---|---|---|
| Abaqus/Explicit | Coupled temp-displacement, CEL, ALE, EET, extensive material library | Steep learning curve | $$$$ | Research, detailed chip formation |
| Deform 3D | Built-in machining module, automatic remeshing, easy setup | Limited to Lagrangian, less flexible | $$$ | Industrial parameter optimization |
| Thirdwave AdvantEdge | Dedicated machining simulation, fast setup, material database included | Limited to orthogonal and 2.5D, less control over physics | $$ | Quick parameter screening |
| ANSYS Workbench | Explicit dynamics module, good for structural analysis of tool | Limited chip formation capability | $$$$ | Tool structural analysis, vibration |
| MATLAB / Simulink | Fast analytical models, no meshing | Requires detailed physics coding | $$ | Rapid force estimation |
FAQ
What is the purpose of FEM simulation in deep hole drilling?
FEM simulation predicts cutting forces, temperature distribution, chip morphology, and residual stresses without costly physical trials. It enables virtual optimization of tool geometry (rake angle, guide pad design), cutting parameters (speed, feed), and coolant strategy before production. A validated model reduces experimental DOE runs by 50–70%.
Which material model is best for drilling simulation?
The Johnson-Cook (JC) constitutive model with JC damage initiation is the most widely used and validated for machining simulations. It accounts for strain hardening, strain rate sensitivity, and thermal softening. For high-speed drilling of aerospace alloys, alternative models such as the Zerilli-Armstrong or Bammann-Chiesa-Johnson models may provide better accuracy at elevated strain rates.
What element type should be used for drilling simulation?
Eight-node hexahedral elements with reduced integration and temperature-displacement coupling (C3D8RT in Abaqus) are preferred for accuracy. Tetrahedral elements (C3D4T) are easier to mesh but produce stiffer behaviour. A minimum element size of 0.01–0.05 mm in the cutting zone is required to resolve the shear band and chip formation accurately.
How long does a typical deep hole drilling simulation take?
A 3D FE simulation of 5–10 mm of tool feed typically requires 12–48 hours of CPU time on a modern workstation (16–32 cores). Simulation time depends on element count (50,000–200,000), time increment stability (mass scaling factor), and whether chip formation is modelled. Analytical thermomechanical models produce results in minutes but provide force-only output.
What temperatures are reached at the cutting edge during BTA drilling?
Simulations and experiments show cutting edge temperatures of 400–700°C for steels and 600–1,050°C for nickel-based superalloys during BTA drilling. Guide pad temperatures are 100–300°C higher due to continuous sliding contact. These temperatures directly affect tool wear rates and surface integrity.
How is chip formation simulated in FEM?
Chip formation is simulated using element elimination (EET) or adaptive remeshing. The Johnson-Cook damage model initiates element deletion when accumulated damage reaches 1.0, causing the chip to separate from the workpiece. The element size in the shear zone must be small enough (0.01–0.05 mm) to resolve the shear band without losing too much mass.
Can FEM predict residual stress in deep hole drilling?
Yes. Coupled temperature-displacement FE simulations can predict residual stress profiles by including the cooling phase after cutting. The simulation captures thermal and mechanical loading cycles that produce the characteristic residual stress profile: tensile at the surface (200–600 MPa) transitioning to compressive in the sub-surface (−200 to −400 MPa). Validation is typically done using XRD measurement.
What is the difference between Lagrangian and CEL approaches?
Lagrangian methods mesh the workpiece material and allow it to deform and separate, producing the most accurate chip morphology. CEL (Coupled Eulerian-Lagrangian) methods model the workpiece as an Eulerian material flowing through a fixed mesh, avoiding mesh distortion issues. CEL is faster and handles large deformation well but provides less detailed chip geometry than Lagrangian methods.
What friction model is used for guide pad contact?
Coulomb friction with a coefficient of 0.15–0.25 is typically used for lubricated guide pad contact. Some advanced models separate the friction into cutting edge friction (µ = 0.32) and guide pad friction (µ = 0.15–0.20) to account for better lubrication at the guide pads. Axial and circumferential friction coefficients at the guide pads are often calibrated separately.
How is the simulation validated against experiments?
Validation compares simulated forces (from dynamometer), temperatures (from embedded thermocouples or IR camera), chip morphology (optical microscopy), and residual stress (XRD) against experimental measurements. Published validation studies report force prediction errors of 5–10% and temperature errors of 10–25% for BTA drilling simulations. A model with force errors below 10% is considered validated for industrial use.
Summary
Finite element simulation provides a powerful virtual laboratory for understanding and optimizing deep hole drilling processes. The Johnson-Cook constitutive model with element elimination is the standard approach for simulating chip formation, cutting forces, and temperature distribution in both BTA and gun drilling. Guide pad temperatures consistently exceed cutting edge temperatures by 100–300°C, making guide pad wear the primary speed-limiting factor in difficult materials. Practical simulations model 5–20 mm of feed under steady-state assumptions, with element sizes of 0.01–0.05 mm in the cutting zone. Lagrangian methods produce the most accurate chip morphology while CEL methods offer better computational efficiency for force and temperature prediction. Validated models achieve force prediction errors of 5–10% and temperature errors of 10–25% against experimental measurements. FEM simulation is now a standard tool in deep hole drilling process development, reducing physical trial requirements by 50–70% and enabling virtual optimization of tool geometry, cutting parameters, and coolant strategy before production.