Lecture 3 · Modelling and Simulation of Processes I – Forming (MSPO2_6I)
Ing. Martin Necpal, PhD.
Winter semester 2026/2027
Materiálovotechnologická fakulta · STU Bratislava

Piston forging – punch and workpiece in DEFORM (© SFTC)
| Time | Block | Content |
|---|---|---|
| 0 – 10 | 1 · Why it matters | Garbage in, garbage out — one part, two flow curves |
| 10 – 20 | 2 · Material window | Layout, library, Regular vs Mixture, units |
| 20 – 30 | 3 · Elastic data | E, ν, α(T), hyperelastic materials |
| 30 – 60 | 4 · Flow stress | Testing, tables, power law, Johnson–Cook, Zerilli–Armstrong, sinh-Arrhenius, choosing a model |
| 60 – 80 | 5 · Yield & hardening | von Mises, Hill 48, R-values, Barlat; isotropic, kinematic, Yoshida–Uemori |
| 80 – 90 | 6 · Creep | Creep curve, DEFORM creep models, relaxation |
| 90 – 105 | 7 · Thermal data | k, heat capacity, emissivity, density, interface HTC |
| 105 – 115 | 8 · Beyond mechanics | Grain (JMAK), phases, hardness, diffusion, EM, fracture |
| 115 – 120 | 9 · Wrap-up | Checklist per process, quiz |
Structure follows DEFORM V14 manual, Chapter 10 “Material Data”
Slab-method estimate with σ = Kεⁿ (computed for this slide)
Material window, Elastic tab — copper C10100 (DEFORM-3D)

Multiple operations — heating, forging, trimming: one material, many temperatures (© SFTC)
| Regular | Mixture | |
|---|---|---|
| Phases | one | 2 … n (austenite, ferrite, pearlite, bainite, martensite …) |
| Data | one set per tab | one set per phase + transformation data |
| Use | forming, machining | heat treatment, quenching, carburising |

Carburised and quenched helical gear — martensite fraction (© SFTC)
| Quantity | English (C10100 above) | SI in DEFORM |
|---|---|---|
| E | 16 679 ksi | 115 000 MPa |
| α | 9.44·10⁻⁶ /°F | 1.7·10⁻⁵ /°C |
| Tref | 68 °F | 20 °C |
| Stress | ksi | MPa = N/mm² |
| k | Btu/(s·in·°F) | N/(s·°C) = W/(m·K) |
| ρ·cp | Btu/(in³·°F) | N/(mm²·°C) = MJ/(m³·K) |
| ρ | lb/in³ | t/mm³ (7.7·10⁻⁹ for steel) |
Unit system is set per problem; mixing an English library entry into an SI problem → use Unit conv.
| Simulation | Plastic | Elastic | Thermal | Phase / grain |
|---|---|---|---|---|
| Isothermal forging | ● | – | – | – |
| Hot / warm forging | ● | – | ● | ○ |
| Springback, residual σ | ● | ● | ○ | – |
| Die stress (elastic die) | – | ● | ○ | – |
| Heat treatment | ● | ● | ● | ● |
| Induction heating | ○ | ○ | ● | ● + EM |
● required ○ recommended – not used
elastic + plastic + thermal + transformation strain
Elastic range: strain recovers fully on unloading

Die stress analysis with elastic dies (© SFTC)
| Material | E [GPa] | ν |
|---|---|---|
| Carbon steel | 200 – 210 | 0.29 – 0.30 |
| Stainless steel (austenitic) | 193 – 200 | 0.29 |
| Aluminium alloys | 69 – 72 | 0.33 |
| Titanium Ti-6Al-4V | 110 – 114 | 0.34 |
| Copper | 115 – 120 | 0.33 |
| WC-Co die insert | 500 – 620 | 0.22 |
Young's modulus across materials (typical handbook values)
E(T) of carbon steel — EN 1993-1-2 reduction factor kE
α(T) of several steels in DEFORM (course lecture)
Neo-Hookean vs linear elastic in tension and compression
Engineering vs true curve
Permanent strain after the yield point
Stress–strain curves vs temperature and strain rate (course lecture)
| Regime | T / Tm | Dominant effect |
|---|---|---|
| Cold | < 0.3 | Strain hardening, weak rate effect (m ≈ 0–0.05) |
| Warm | 0.3 – 0.5 | Both, recovery starts |
| Hot | > 0.5 – 0.6 | Rate + T; hardening balanced by recovery / DRX (m ≈ 0.1–0.2) |

Tensile test
ε < 0.2–0.3, ε̇ ≈ 10⁻³–10⁻¹ s⁻¹ (© ZwickRoell)
Compression (upsetting)
ε ≤ 1, friction correction

Gleeble hot compression
up to ≈ 1700 °C, ε̇ to 10² s⁻¹ (© DSI)
Split Hopkinson bar (SHPB)
ε̇ ≈ 10²–10⁴ s⁻¹
| Test | Strength | Limitation | DEFORM use |
|---|---|---|---|
| Tension | Standard, simple stress state | Necking limits strain | Sheet, cold forming start of curve |
| Compression / plane strain | Large strains, hot | Friction, barrelling → inverse correction | Bulk forming, forging |
| Torsion | Very large strains (ε > 5) | Gradient through radius | Hot rolling, DRX |
| SHPB | High rates | Small strain, adiabatic | Machining, impact (J–C) |
DEFORM flow-stress table: curves per T and ε̇ (course lecture)
Value at ε̇ = 50 s⁻¹ from the same four table points: 1012 vs 1028 MPa
Table measured to 10 s⁻¹, used at 10⁴ s⁻¹: flow stress 8.2 % too low
Process windows (solid) vs test windows (dashed) — schematic
| Material (annealed) | K [MPa] | n |
|---|---|---|
| Low-C steel DC04 | ≈ 530 | 0.22 – 0.26 |
| C45 (normalised) | ≈ 900 | 0.15 – 0.18 |
| Austenitic stainless 304 | ≈ 1300 | 0.40 – 0.50 |
| Al 5xxx | ≈ 400 | 0.25 – 0.30 |
σ̄ = c ε̄ⁿ for three hardening exponents
| Law | Equation | Behaviour at large ε | Typical use |
|---|---|---|---|
| Ludwik (1909) | \( \sigma = \sigma_0 + K\varepsilon^{n} \) | unbounded | general, with initial yield |
| Hollomon (1945) | \( \sigma = K\varepsilon^{n} \) | unbounded, σ(0) = 0 | sheet n-value, quick estimates |
| Swift (1952) | \( \sigma = K(\varepsilon_0 + \varepsilon)^{n} \) | unbounded, σ(0) = Kε₀ⁿ | sheet forming, prestrained material |
| Voce (1948) | \( \sigma = \sigma_s - (\sigma_s - \sigma_0)\,e^{-\varepsilon/\varepsilon_c} \) | saturates at σs | Al alloys, large strains |
| DEFORM power law | \( \bar\sigma = c\,\bar\varepsilon^{\,n}\,\dot{\bar\varepsilon}^{\,m} + y \) | unbounded | cold forging, heading, extrusion |
Considère construction for K = 530 MPa, n = 0.24 (DC04-type steel)
Not a straight line on log–log? The material does not follow a power law (e.g. Lüders plateau, saturation) — choose Voce or a table.
Synthetic tensile points (K = 530, n = 0.24) — regression returns n = 0.237
| Condition | m | ×10 in ε̇ gives |
|---|---|---|
| Steel, cold (20 °C) | 0.005 – 0.02 | +1 … 5 % |
| Al alloys, cold | ≈ 0 – 0.01 | ≈ 0 … 2 % |
| Warm forming (0.3–0.5 Tm) | 0.03 – 0.1 | +7 … 26 % |
| Hot forming | 0.1 – 0.2 | +26 … 58 % |
| Superplastic | ≥ 0.3 – 0.5 | ×2 … ×3 |
Strain-rate jump test (m = 0.012, cold steel): small but measurable step
All four laws fitted to the same points (ε = 0.02–0.20); scatter < 1 %
| T | c [MPa] | n | m | regime |
|---|---|---|---|---|
| 20 °C | 900 | 0.17 | 0.01 | cold |
| 200 °C | 850 | 0.15 | ≈ 0 (DSA) | avoid |
| 500 °C | 620 | 0.12 | 0.04 | warm |
| 700 °C | 330 | 0.08 | 0.10 | warm |
| 900 °C | 180 | 0.05 | 0.15 | hot → use sinh / table |
Illustrative trend for a medium-carbon steel — use measured values for real work
| Process | Why the power law |
|---|---|
| Cold forging, heading, cold extrusion | Hardening dominates, m ≈ 0; needs data to ε ≈ 2 |
| Wire and bar drawing | Accumulated strain over passes; n sets the drawing limit |
| Sheet stamping, deep drawing | Swift / Hollomon with n and r-value; n → forming limit |
| Bending, roll forming | Moderate strains; springback needs E and hardening model too |
| Warm forging | c(T), n(T), m(T) tables |
Higher n → higher forming limit in plane strain (empirical, low-carbon steels)
Limits
Summary
Check yourself
Press ↑ to return to the main lecture, or → to continue with Johnson–Cook.
| AISI 1045 | A | B | n | C | m |
|---|---|---|---|---|---|
| Jaspers & Dautzenberg | 553 MPa | 601 MPa | 0.234 | 0.0134 | 1.0 |
Johnson–Cook, AISI 1045 (Tr = 20 °C, Tm = 1460 °C)
| Symbol | Meaning | From test |
|---|---|---|
| A | yield stress at ε̇₀, Tr | quasi-static |
| B, n | strain hardening | quasi-static |
| C | strain-rate sensitivity | SHPB / torsion |
| m | thermal softening | tests at T |
| ε̇₀, Tr, Tm | reference rate (often 1 s⁻¹), reference and melting temperature | chosen / handbook |
Rate factor: linear in ln ε̇* — +12 % per 4 decades for C = 0.0134
Thermal factor: m < 1 softens early, m > 1 late; zero at Tm
| Material | A [MPa] | B [MPa] | n | C | m | Tm [K] | Source |
|---|---|---|---|---|---|---|---|
| OFHC copper | 90 | 292 | 0.31 | 0.025 | 1.09 | 1356 | J–C 1983 |
| Cartridge brass | 112 | 505 | 0.42 | 0.009 | 1.68 | 1189 | J–C 1983 |
| Nickel 200 | 163 | 648 | 0.33 | 0.006 | 1.44 | 1726 | J–C 1983 |
| 1006 steel | 350 | 275 | 0.36 | 0.022 | 1.00 | 1811 | J–C 1983 |
| 4340 steel | 792 | 510 | 0.26 | 0.014 | 1.03 | 1793 | J–C 1983 |
| 2024-T351 Al | 265 | 426 | 0.34 | 0.015 | 1.00 | 775 | J–C 1983 |
| AISI 1045 | 553 | 601 | 0.234 | 0.0134 | 1.00 | 1733 | Jaspers & Dautzenberg 2002 |
Alternative: fit all five constants together by least squares (or inverse FEM of the tests) — better overall error, less transparent
Step 3 with synthetic points (C = 0.0134): regression gives C = 0.0137
AISI 1045 at 10⁴ s⁻¹: plastic work heats the material by ≈ 258 K up to ε = 1
Known limitations
Modified forms
Failure strain falls steeply with triaxiality (4340 steel, J–C 1985)
| Operation | Why J–C | Typical ε̇, T |
|---|---|---|
| Machining (turning, milling, drilling) — DEFORM-3D Machining | Chip formation, cutting forces, tool temperature; J–C + J–C damage for segmentation | 10³–10⁶ s⁻¹, up to ≈ 1000 °C |
| Impact, ballistics, Taylor test | Original purpose; robust in explicit codes | 10³–10⁵ s⁻¹, adiabatic |
| Crash of car structures | Rate hardening of steel and Al sheet | 1–10³ s⁻¹ |
| High-speed forming (electromagnetic, explosive, hammer) | Rate + adiabatic heating | 10²–10⁴ s⁻¹ |
| Blanking, shearing, punching | Localised shear, damage model | 10–10³ s⁻¹ |
| Hot forging | Not recommended — use sinh law or tables | > 0.6 Tm |
Orthogonal cutting: extreme strain rate in the primary shear zone, high temperature at the tool–chip interface
Tips for DEFORM
Summary
Check yourself
Press ↑ to return to the main lecture, or → to continue with Zerilli–Armstrong.
Zerilli–Armstrong (1987) constants for Armco iron and OFHC copper
| Johnson–Cook | Zerilli–Armstrong | |
|---|---|---|
| Basis | Empirical fit | Dislocation mechanics |
| T–ε̇ coupling | None (product of brackets) | Built in: \(e^{-C_3T + C_4T\ln\dot\varepsilon}\) |
| Lattice | Same form for all | BCC / FCC / HCP forms |
| Grain size | Not explicit | Hall–Petch term \(k\,d^{-1/2}\) |
| Temperature | Homologous T*, needs Tm | Absolute T [K] |
| Constants | 5 (A, B, n, C, m) | 5–6 (C₀…C₅, n) |
Short-range barriers can be passed with the help of thermal vibration; long-range ones cannot (schematic)
① Thermal stress from thermal activation (activation area A)
② BCC: A independent of strain (Peierls barrier)
③ FCC: A ∝ ε̄−1/2 (forest density ρ ∝ ε̄)
④ Common athermal term
| Constant | Physical meaning | Unit |
|---|---|---|
| C₀ | athermal stress + Hall–Petch | MPa |
| C₁ (BCC) | \(M\Delta G_0/(A b)\) — thermal yield amplitude | MPa |
| C₂ (FCC) | thermal hardening amplitude | MPa |
| C₃ = β₀ | thermal softening | K⁻¹ |
| C₄ = β₁ | T–ε̇ coupling | K⁻¹ |
| C₅, n (BCC) | athermal strain hardening | MPa, – |
| k, d | Hall–Petch slope, grain size | MPa·mm½, mm |
BCC: curves shift in parallel — T acts on the yield stress
FCC: curves fan out — T acts on the hardening
Iron: thermal part vanishes with T; athermal part stays
Rate effect from 10⁻³ to 10⁴ s⁻¹: Fe ≈ +350 MPa, Cu ≈ +100 MPa (296 K)
| Metal | C₀ | C₁ / C₂ | C₃ [K⁻¹] | C₄ [K⁻¹] | C₅ | n |
|---|---|---|---|---|---|---|
| OFHC Cu (FCC) | ≈ 65* | C₂ = 890 | 0.0028 | 1.15·10⁻⁴ | – | ½ |
| Armco Fe (BCC) | k d−½ | C₁ = 1033 | 0.00698 | 4.15·10⁻⁴ | 266 | 0.289 |
| Tantalum (BCC) | 30 | C₁ = 1125 | 0.00535 | 3.27·10⁻⁴ | 310 | 0.44 |
Stresses in MPa. *Cu: σG = 46.5 MPa + k d−½ with k = 5 MPa·mm½, d = 0.073 mm. Sources: Zerilli & Armstrong 1987; Banerjee 2005 (arXiv cond-mat/0512466); Ta set as quoted in arXiv 1105.5514
DEFORM (dropdown list)
Modified ZA for hot forming (Samantaray et al. 2009)
Implemented in
Step 3–4: straight lines per strain rate (iron constants, synthetic test points)
| Operation | Why ZA | Typical ε̇, T |
|---|---|---|
| Taylor impact test (model validation) | Original validation of ZA for Cu and Fe; one test checks the whole ε–ε̇–T field | 10³–10⁴ s⁻¹, 296–1200 K |
| Ballistics, penetrators, EFP liners | BCC Fe, Ta, W: strong rate effect at high rates | 10⁴ s⁻¹, adiabatic |
| High-speed machining | Chip formation, shear bands in steels; compared with J–C | 10³–10⁶ s⁻¹, up to 1000 °C |
| High-speed forming (electromagnetic, explosive, hammer) | Rate hardening of ferritic sheet and tube | 10²–10⁴ s⁻¹ |
| Cold heading, crash, low-T service | BCC yield vs FCC hardening; ductile–brittle transition | 1–10³ s⁻¹, 200–400 K |
| Hot forging, rolling (modified ZA) | Semi-empirical fit of Gleeble data | 10⁻³–10 s⁻¹, > 0.6 Tm |
Taylor impact test: final length and mushroom diameter validate the model
Known limits
Checklist for DEFORM
Check yourself
Press ↑ to return to the main lecture, or → to continue directly with hot forming models.
Steady-state flow stress, sinh-Arrhenius (Q = 312 kJ/mol, n = 4.5, α = 0.012 MPa⁻¹)
Lines of constant Z: a lab test at low ε̇ can stand in for a fast press at higher T (C–Mn steel)
| Material | Apparent Q for hot working [kJ/mol] |
|---|---|
| Pure Al (self-diffusion ≈ 142) | ≈ 140 – 160 |
| Al alloys (6xxx, 7xxx) | ≈ 140 – 200 |
| Mg alloy AZ31 | ≈ 130 – 170 |
| C–Mn and low-alloy steels (30CrMoA: 262) | ≈ 260 – 320 |
| Austenitic stainless steels | ≈ 400 – 460 |
| Ti-6Al-4V: β / (α + β) field | ≈ 246 / 621 |
| Ni-base superalloys | ≈ 400 – 600 |
Typical literature ranges; 30CrMoA from Zhang et al. 2020; Ti-6Al-4V from “Hot deformation characteristics of Ti-6Al-4V” (CityU)
Same constants: the sinh law follows the power law at low σ and the exponential at high σ
Flow stress as a function of Z
Strain-rate sensitivity
Temperature sensitivity
m falls from 1/n (creep) towards zero (high Z)
Example 30CrMoA (Zhang et al. 2020): α = 0.011 MPa⁻¹, n = 4.40, Q = 262 kJ/mol, A = 3.67·10⁹ s⁻¹, R = 0.983
Step 3: parallel lines, one per temperature (synthetic points from the C–Mn constants)
Typical DRX-type curves: peak, then softening to steady state (illustrative, σ_p from the sinh law)
Strain-compensated Arrhenius model
Physics of the curve
Processing maps (Prasad, dynamic materials model)
DRX grain size falls with Z (illustrative constants for a C–Mn steel)
Norton–Hoff (viscoplastic power law)
Hensel–Spittel (empirical, many terms)
DEFORM flow-stress list
| Operation | Use of Z / sinh law |
|---|---|
| Hot closed-die forging | Flow stress for die fill and load; Z → DRX grain size in the part |
| Hot rolling, pass schedules | Mean flow stress per pass; Z → recrystallisation between passes |
| Aluminium extrusion | Sinh law is the classic model (Sheppard & Wright); exit temperature and speed limits |
| Isothermal / hot-die forging of Ti, Ni alloys | Low ε̇, near-creep regime; m from the sinh law |
| Process design and scale-up | Equal Z: transfer Gleeble results to the press |
| Processing maps | Choose safe T–ε̇ windows (DRX, no shear bands) |
| Creep of dies and hot tools | Same Arrhenius structure, low-stress (power-law) branch |

Hot forging sequence of a disc (© SFTC)

Ring rolling — Z changes along the roll gap (© SFTC)
Limits
Summary
Check yourself
Press ↑ to return to the main lecture, or → to continue directly.
Flow-stress options in the Plastic tab (course lecture)
| Option | Idea | When |
|---|---|---|
| Dislocation density | σ̄ = σ₀ + αMGb√ρ; ρ evolves with ε, ε̇, T | Multi-step hot forging, DRX |
| Texture / crystal plasticity | Grain orientations → anisotropic flow | Ti, Mg, sheet, earing |
| DNN predictor | Neural network: chemistry → flow curve | No test data, early design |
| User routine | Own FORTRAN subroutine compiled into the solver | Research models |
| Conversion | Equation → table; test data → fitted model | Library building |
The Material Suite (SFTC) fits flow stress from compression data, computes TTT diagrams from composition and JMAK constants.
Decision guide used in this course
Nine stress components on a volume element (course lecture)
Plane-stress yield loci (σ₃ = 0)
Yield function types in DEFORM (course lecture)
Hill quadratic (FGHLMN) dialog in DEFORM (course lecture)
Four ears from planar anisotropy
Hill 48 normal anisotropy, plane stress
Work hardening: reloading needs a higher stress (course lecture)
Isotropic (left) and kinematic (right) hardening in the deviatoric plane
Load reversal after 5 % strain: σ₀ = 300 MPa, linear hardening
Hardening rule selection (course lecture)

Multi-stage drawing of a casing — unloading after each stage (© SFTC)
Creep curves for three stress levels (schematic, computed)
Stages I, II, III (course lecture)
Creep option in the Plastic tab (course lecture)
| Model | Equation | Typical use |
|---|---|---|
| Power law (Norton) | \( \dot\varepsilon^{c} = A\,\bar\sigma^{n} \) | Steady-state creep, relaxation |
| Bailey–Norton | \( \varepsilon^{c} = A\,\bar\sigma^{n}\,t^{m} \) (m < 1) | Primary + secondary creep |
| Perzyna | \( \dot\varepsilon^{c} = \gamma\left\langle \dfrac{\bar\sigma}{\sigma_y} - 1\right\rangle^{p} \) | Viscoplasticity above yield |
| Soderberg | \( \dot\varepsilon^{c} = A\left[\exp(\bar\sigma/\sigma_0) - 1\right] \) | Low-stress creep |
| Exponential | \( \dot\varepsilon^{c} = A\,e^{B\bar\sigma}\,e^{-Q/RT} \) | High-stress creep, T-dependent |
| Tabular | \( \dot\varepsilon^{c} = f(\bar\sigma, T, \varepsilon^{c}) \) | Measured data |
Model list: DEFORM V14 manual, 10 Material Data – Creep
Stress relaxation under fixed strain, Norton n = 5 (computed)

Dead-weight creep frame (© DoITPoMS, Univ. of Cambridge)
Thermal tab: conductivity, heat capacity / specific heat, emissivity, density (course lecture)
| Property | SI unit in DEFORM | Controls |
|---|---|---|
| Conductivity k(T) | N/(s·°C) = W/(m·K) | Heat flow inside the part |
| Heat capacity ρcp(T) | N/(mm²·°C) | Heating per unit energy |
| Emissivity ε | – | Radiation to environment |
| Density ρ | t/mm³ | Mass (hammers), specific-heat option |
k(T): carbon steel per EN 1993-1-2; austenitic stainless (typical)
| Material, 20 °C | k [W/(m·K)] |
|---|---|
| Copper | ≈ 390 |
| Aluminium alloys | 120 – 200 |
| Carbon steel | 45 – 54 |
| Hot-work tool steel H13 | ≈ 25 – 29 |
| Stainless 304 | ≈ 15 |
| Ti-6Al-4V | ≈ 7 |
cp(T) of carbon steel — EN 1993-1-2 (peak at the α → γ transformation)
| Surface | Emissivity ε |
|---|---|
| Oxidised / scaled steel, hot | 0.7 – 0.9 |
| Bright machined steel | 0.2 – 0.4 |
| Polished aluminium | 0.04 – 0.1 |
| Oxidised titanium | 0.5 – 0.7 |

Open-die cogging of a hot ingot: hours of radiation cooling between passes (© SFTC)

1 · Test with thermocouples

2 · Simulated vs measured T(t)

3 · Fitted HTC (© SFTC)
Static recrystallisation, Avrami n = 1.5, t0.5 from 0.1 s to 40 s (illustrative)

Grain size in a forged Waspaloy disc (© SFTC)
Diffusional (ferrite, pearlite, bainite)
Martensite — Koistinen–Marburger
Hardness — Hardness tab

Residual hoop stress after quenching a disc (© SFTC)

Cold heading progression — hardness from strain hardening (© SFTC)
Diffusion tab
Elec./Mag. tab

Induction-hardened spur gear (© SFTC)
Upsetting: tensile hoop stress at the bulge

Cross-wedge rolling — risk of central cavity (Mannesmann effect) (© SFTC)
| Process | Flow stress | Yield / hardening | Elastic | Thermal | Extra |
|---|---|---|---|---|---|
| Cold forging, heading | Power law / table, ε to 2–3 | von Mises, isotropic | For die stress | k, ρcp (heating) | Damage C |
| Hot closed-die forging | Table σ(ε, ε̇, T) or sinh | von Mises, isotropic | – | k(T), ρcp(T), ε, HTC | Grain (JMAK) |
| Machining | Johnson–Cook / ZA | von Mises | Tool | k, ρcp | Damage, wear |
| Sheet / springback | Swift / Hollomon | Hill 48 / Barlat, kinematic, Y-U | E(ε̄p) | – | FLD |
| Heat treatment | Per phase, f(T) | Isotropic (+ TRIP) | E(T), α(T) | Per phase | TTT, latent heat, ΔV, HV |
| Induction hardening | Per phase | Isotropic | E(T), α(T) | k, ρcp | ρe(T), μr(T) |
Next lecture: preprocessing — mesh, boundary conditions, friction