1

Dr.-Ing. Tobias LooseIngenieurbüro Tobias Loose, Herdweg 13, D- 75045 Wössingenloose@tl-ing.de www.tl-ing.eu

Introduction in

Welding and Heat Treatment SimulationApplication Fields and Benefits

Workshop Welding and Heat Treatment with LS-DYNA on the

10. European LS-DYNA Conference 2015 Würzburg

Foto: ISF

Herdweg 13, D-75045 Wössingen Lkr. KarlsruheE-Post: loose@tl-ing.de Web: ww.tl-ing.eu, www.loose.at

Mobil: +49 (0) 176 6126 8671 Tel: +49 (0) 7203 329 023 Fax: +49 (0) 7203 329 025

2

Numerical Simulation forWelding and Heat Treatment since 2004

• Consulting• Training • Support• Software Development• Software Distribution

for Welding Simulation and Heat Treatment Simulation

Herdweg 13, D-75045 Wössingen Lkr. KarlsruheE-Post: loose@tl-ing.de Web: www.tl-ing.eu www.loose.at

Mobil: +49 (0) 176 6126 8671 Tel: +49 (0) 7203 329 023 Fax: +49 (0) 7203 329 025

Internet:DEeutsch: www.loose.atENglisch: www.tl-ing.euESpanol: www.loose.es

www.WeldWare.eu

www.SimWeld.eu

www.DynaWeld.eu

3

Specific Features of Welding and Heat Treatment Simulation

Material Properties– material properties depend on temperature– material properties change in thermal loading cycles

→ change of microstructure / phase transformation

History Variables– stress, strain, strain hardening– phase proportion

Material Modell– reset of material history– phase transformation– phase transformation effects

Time Schedule– History of events

in order of time

MaterialSimulation

Process Simulation

Structure Simulation

Welding & HTSimulation

4

Properties of the WeldMicrostructure - mechanical Properties

www.weldware.eu

5

Prediction of Microstructure

Prediction of microstructure in the HAZin dependence of steel grade and its chemical composition

6

Prediction of the estimated mechanical Properties in the HAZ...

• Hardness• Yield stress• Ultimate stress• Ultimate strain• Brucheinschnürung

… to avoide technological kerbs in comparison to the unwelded

base material.

8

Simulation at Mikro-NiveauSolidification, Graingrowth, Phase Transformation

10

11

Welding Process AnalysisParameter Adjustment for GMAWEquivalent Heat Source

www.simweld.eu

12

SimWeld Preprocessing

• Definition of:– wire: feed, diameter, material,– stick out– travel speed– angle of torch, stabbing, slabbing, skew– shielding gas– machine settings U, I– process type normal, pulsed U/I, pulsed I/I– pulse parameter

13

• Equivalent Heat Source• Weld Pool Geometry• Droplet• Wire Temperature• Energy, Voltage, Currency• Temperature Curve

SimWeld Results

15

Welding Structure AnalysisResidual Stress - Distortion - Temperaturein the whole Assembly

DynaWeld

www.dynaweld.eu

16

Welding Structure Analysis Using the Finite Element Method

Geometry of the Assembly - CAD

Method of the Finite Elements

FEM

Discretisation in Finte ElementsMeshing

WeldingAdjustment of equivalent Heat Source

Materials Material Properties

Process and Time ScheduleWeld Sequence, Material Assignment,

Clamps, Loads

17

The most important Items of the Welding Structure Analysis

• Geometry• Material

– dependend of Temperature– Strain Hardening

• Reset of Strain Hardening at Melting Point– Phase Transformation

• Transformation Strain• material properties depend on microstructure

and phase transformation

• Heat Source– Geometry and equivalent Heat Source Funktion– Trajectory and Travel Speed

• Mechanical Boundary Condition– Clamps– Contact

• Results– Deformation– Residual Stress, Strain, Yield Stress– global Temperature Distribution and Microstructure

18

Phase Transformation and Transformation Effects

ferritic grid krz → austinitic grid kfzVolume change during phase transformation

right: Animation of thermal strain and transformation strain during the heating an colling at different cooling rates (S355)bottom right:Transformation and Dilatation (1.4317)bottom left: Phase-Cooling-Rate -Diagram S500 comparison of given data and simulaion

19

Process Types

• Fusion Welding – Arc– Arc Welding

– Gas Metal Arc Welding

– Submerged Arc Welding

– Tungsten Inert Gas Welding

• Electro-Resistand Welding– Spot Welding

• Friction Welding– Rotational Friction Welding

– Orbital Friction Welding

– Friction Stir Welding

• Fusion Welding – Beam– Laser Welding

– Electron Beam Welding

Welding Structure Analysis with

Equivalent Heat Source

Special Analysisfor Welding Process

EM, FEM, EFG, SPH

20

Equivalent Heat Source

• The Equivalent Heat Source is a coupled function of geometry and intension of the heat generation density.

• It describes the thermal loading of a welding structure simulation.• Its aproach is to generate the same heat input as the real proccess.• It is an engineering aproach and not real physics.• It covers fluid dynamic effects like convection in the weld pool.• Any funktion is allowed.• SimWeld can predict the heat source parameter for GMAW.

Foto: ISF

21

Validation

22

• Plate with the dimension270 x 200 x 30 mm3 with V/U-shaped groove

• austenitic stainless steel (316LNSPH, kf = 275 MPa)

• 2 Layer, Filler materials is the same as the base material: 316L

• TIG Welding with U = 9 V, I = 155 A, v = 0,67 mm/s

ValidationIIW Round Robin Versuch

Comparison Measurement and Simulation

Loose, T. ; Sakkiettibutra, J. ; Wohlfahrt, H. : New 3D-Calculations of residual stresses consistent with measured results of the

IIW Round Robin Programme. In: Cherjak, H. (Ed.) ; Enzinger, N. (Ed.) :

Mathematical Modelling of Weld Phenomena Bd. 9, Verlag der Technischen Universität Graz, 2010

23

Welding direction

ValidationIIW Round Robin Versuch

SYSWELD

LS-DYNA

Lateral residual stress Longitudinal residual stress

24

Temperature – Equivalent StressLateral Residual Stress - Longitudinal Residual Stress

25

Validation Nitschke-Pagel Test

Distortion w:Experiment: 0,34 mmSysweld: 0,32 mmLS-DYNA: 0,34 mm

Loose, T.: Einfluß des transienten Schweißvorganges auf Verzug, Eigenspannungen und Stabiltiätsverhalten axial gedrückter Kreiszylinderschalenaus Stahl, Diss, Karlsruhe, 2008

26

Welding of a T-Joint

• Double sided T-Joint a = 4 mm• Plate S355 thickness 8 mm• 3 Tacks double sided • Travel speed 80 cm/min• Current: 390 A• Voltage: 30 V

• Start Time Tack 1: 0 s• Start Time Tack 2: 20 s• Start Time Weld 1: 1000 s• Start Time Weld 2: 1023 s• Weld 1 and Weld 2 have

the same travel direction Foto: Volvo

27

Process Simulation with SimWeld

Input-Parameter SimWeld

28

SimWeld Results

• a = 4,4 mm• I = 390 A• V = 29,2 V

29

Temperature

Tack 1 Tack 2

Weld 1 Weld 2

30

z-Distortion at Evaluation Pathtransformed to flat left side

31

Multi LayeredWelds

32

Weld of a Pipe with 40 mm Wall Thicknessmade of Alloy 625

60 Layer - GMAW 93 Layer - TIG

33

Weld of a Pipe with 40 mm Wall Thicknessmade of Alloy 625 - 60 Layer GMAW

Temperature Layer 44 Equivalent Plastic Strain

34

2D plain strainPlate: 300 x 80 mmStiffner: 150 x 24 mmFillet Weld: a = 13 mmMaterial: 1.4301

Tack Weld a = 1,4 mm With failure criterion strain:KFAIL = 0,25 m/m

Initial gap between stiffner and plate: 0,1 mm

The plate is fixed with symmetry boundary conditions on the left and right side.

Angular Distortion of multilayered Welds2D-Modell LS-DYNA - Simulationsmodell

35

Example Angular Distortion of multilayered Welds2D-Model LS-DYNA - Temperaturefield

36

Example Angular Distortion of multilayered Welds2D-Modell LS-DYNA – plastische Dehnungen

37

Heat Treatment Simulation

38

Types of Heat TreatmentProcess steps

• Heating– Austinitsation– grain size– heat generation by induction

• Carburization– change of material– change of composition

• External Pressure– phase transformation

influenced by stress

• Quenching– phase transformation

influenced by thermal cycle

• Tempering / Anealing– smoothing of Martensit / Bainit– stress release

Case Hardening

Quenching

Press Hardening

Induction Hardening

AnenalingPost Weld Heat Treatment

Tempering

39

Quenching of a Gear made of S355Temperature – 20 times scaled displacement

40

Quenching of a Gear made of S355Martensit - 20 times scaled displacement

41

Quenching of a Gear made of S355Temperature – Surface and mid-cross-section

42

Quenching of a Gear made of S355Temperature Curve

EdgeMiddle

43

Martensit (right)

Hardness HV (bottom left)

Yield (bottom right)

Quenching of a Gear made of S355Results of Heat Treatment Simulation

44

Process ChainFormingWeldingHeat Treatment

45

Simulation of the Process Chain

MaterialSimulation

Process Simulation

Structure Simulation

Welding & HTSimulation

Forming

CrashClamping and Predeformation

Cutting andGrinding

Assembly

46

Deep-Drawing of a Cup from a Laser Welded SheetTask and Model

Welding:• Two sheets (S355) with 1 mm wall thickness are laser weldedForming:• The welded and distorted sheet is

clamped• a globular die is pressed slow in the

sheet.

Model:• Shell-elements are used for the sheet, solid elements are used for the clamps and the die• Same material model (*MAT_244) is used in all steps• History variables, phase proportions and deformations are kept from one step to an other• Welding: implicit analysis, Forming / Crash: explicit analysis

47

Stresses and Strains in Midsurface of Shell

After welding and coolingtop left: Longitudinal stresstop right: Effectiv stress (v. Mises)bottom right: plastic strain

48

Microstructure

After welding and coolingtop left: Ferrit proportiontop right: Bainit proportionbottom right: Martensit proportion

49

Deep drawing – effectiv stress

50

Stresses and Strains in Midsurface of Shell

top left: effectiv stress bevor unclamping 200 .. 1100 N/mm²

bottom left: effectiv stess after unclamping 0 .. 200 N/mm²

bottom right: plastic strain after unclamping 0 .. 0.65 m/m

51

Microstructure during Deep-Drawing

top left: Ferrit proportiontop right: Bainit proportionbottom right: Martensit proportion

52

Thinning of the SheetInfluence of Material Property Change from Welding

53

Manufacturing of a BoxTask and Model

Forming:• The roof geometry is made by forming a 3 mm thick sheet (1.4301)Assembly:• Add the sidewallWelding:• Weld the sidewall to the roofClamp and predeformation:• press the sidewall on measureAssembly:• Add the bottom plateWelding:• Weld the bottom plate to the sidewallUnclampingModel:• Solid-element model• Material model (*MAT_270) is used in all steps• History variables and deformations are kept from one step to an other• Implicit analysis in all steps

54

Benefit of a Continous Simulation of Process Chain

• Precalcualtion of the final state of the assembly:– geometry

– residual stresses

– microstructure

• Complete simulation of the entire manufacturing process• Take into account the two way impact of single manufacturing tasks

• Enables the design of the manufacturing process• Enables the desing of compensation methods for requested conditions

55

Thanks for your Attention!