Advanced Qualification of Additive Manufacturing Materials Workshop

Temperature Simulations and Measurements Manufacturing Materials Workshop

pfor Process Qualification in Powder-Bed Electron Beam Additive ManufacturingElectron Beam Additive Manufacturing

Kevin ChouKevin ChouProfessor

Mechanical Engineering DepartmentThe University of AlabamaThe University of Alabama

Assistant Director for TechnologyAdvanced Manufacturing National Program Officeg g

U.S. Department of Commerce

July 20, 2015

Disclaimer and Note• The materials presented and opinions expressed in this seminar were solely from 

the presenter himself. They do not represent the viewpoints of The University of Alabama, nor the Advanced Manufacturing National Program Office.

• The materials presented in this seminar are mainly from the following articles.– Cheng, B., S. Price, J. Lydon, K. Cooper and K. Chou, "On Process Temperature in Powder‐Bed 

El t B Additi M f t i M d l D l t d E i t l V lid ti ” J lElectron Beam Additive Manufacturing: Model Development and Experimental Validation,” Journal of Manufacturing Science and Engineering, Vol. 136, No. 6, pp. 061018 (1‐12), 2014.

– Price, S., B. Cheng, J. Lydon, K. Cooper and K. Chou, "On Process Temperature in Powder‐Bed Electron Beam Additive Manufacturing: Process Parameter Effects,” Journal of Manufacturing Science and Engineering, Vol. 136, No. 6, pp. 061019 (1‐10), 2014.Science and Engineering, Vol. 136, No. 6, pp. 061019 (1 10), 2014.

– Gong, X., J. Lydon, K. Cooper, and K. Chou, “Beam Speed Effects on Ti‐6Al‐4V Microstructures in Electron Beam Additive Manufacturing,” Journal of Materials Research, Vol. 29, No. 17, pp. 1951‐1959, 2014.

– Gong, X., J. Lydon, K. Cooper, and K. Chou, “Characterization of Ti‐6Al‐4V Powder in Electron‐Beam‐g, , y , p , ,Melting Additive Manufacturing,” International Journal of Powder Metallurgy, Vol. 51, No. 1, pp. 1‐10, 2015.

Contact information:Kevin Chou, kchou@eng.ua.edu, 205‐348‐0044

2

Quality Control in AM Material and ProcessMaterial and Process

Process ∆2 Part∆2

∆

∆f

EBAM Process Video∆1Feedstock

EBAM Process Video

GE Aviation

3∆f = Fn (∆1, ∆2, etc.)

EBAM System CharacteristicsEBAM System CharacteristicsEBAM machine

( l ld d l)(example, old model) Build chamber

High power 60 keV electron gun

Leaded glassHigh strength glass

4

No moving part Heat shieldSensor access limitation

EBAM Viewport WindowEBAM Viewport WindowMetal

ShutterLeaded GlassBuildBuild

Area

Machine Outside

Sacrificial Glass

Vacuum Glass

Heat Shield

Machine Inside

5Metalized film deposit

EBAM Process CharacteristicsSpeed Function (SF)

800

900

1000

14

16

18

300

400

500

600

700

Beam

Spe

ed (m

m/s

)

6

8

10

12

14

Beam

Cur

rent

(mA

)

B S d f ti (SF h i ht)

0

100

200

0 5 10 15 20 25 30

B

Build Height (mm)0

2

4

0 5 10 15 20 25 30

B

Build Height (mm)

Vacuum (little He)Pre-heating, sinteringW ( 650 ˚C)

Beam Speed = function (SF, height)y = 24.133x - 12.631

R² = 0.9996

y = 20.211x - 12.612R² = 0.9996

y = 17.866x - 25.3551400

1600

1800

Layer 95 (6.65 mm)

Layer 97 (6.79 mm)

Warm process (> ~650 ˚C)High scanning speedProcess parameters change

through the build

yR² = 0.9988

y = 10.587x - 0.4949R² = 0.9997

600

800

1000

1200

Bea

m S

peed

(mm

/s) Layer 153

(10.71 mm)

Layer 155 (10.85 mm)

Layer 223 (15.61 mm)

Layer 225 (15 75 mm)

6

gMetalizing filmSlow post-process cooling

0

200

400

0 10 20 30 40 50 60 70Speed Function Index

(15.75 mm)

Layer 347 (24.29 mm)

Layer 349 (24.43 mm)

EBAM Process/Material StudiesEBAM Process/Material Studies

Process PhysicsProcess Physics

Temperature

AM Part (4)Powder-Bed (1)

TemperatureSimulation (2)

Powder-Bed (1)

Ti-6Al-4V

Temperature Measurement (3)

7

(1) Feedstock Characterization

(1) Feedstock Characterization

√ Powder-Bed Particles, Porosity- Metallography, Micro-CT

√

Thermal ConductivityHot Disk Thermal Analyzer

√

- Hot-Disk Thermal Analyzer

8

Preheated PowderZ-plane

X-plane9

Porosity Study - Micro-CTy y 10 mm Ti-6Al-4V Cube (hollow) Skyscan 1172 Micro CT Skyscan 1172 Micro-CT Size/Porosity Distribution

~ 2 μm resolution~ 2 μm resolution

Loosepowders

10

Sinteredpowders

Porosity and Powder Sizey Powder-Bed Porosity, ~ 50% Particle Size Distribution Particle Size Distribution

Major: ~ 30 to 50 μm

 (b) (a)

11

Powder Thermal Conductivity TPS2500 S Thermal Analyzer (Hot Disk)

- Solid and Hollow samples

12Solid vs. Hollow Hollow sample with shell removed (1 side)

Powder Thermal Conductivityy Sintered Powder Specimens

13

(2) Temperature Simulation- Finite Element Modeling

Heat Transfer Heat Source

√√ Heat Source Material/Powder Properties√√

√ Latent Heat of Fusion√

14

Governing EquationsHeat Transfer

TcTc )()(

g q

T - Temperature

xTc

vt

TcQTk p

sp

)()()(

- Absorbed heat fluxc - Specific heat capacityρ - Densityλ - Thermal conductivity

C d f h

, ,x y zQ

Latent Heat of Fusion

d f

Assumptions:

vs - Constant speed of the moving heat source

fH T cdT L f 0

ST T f

,S

S L

T TT T T p

Heat Conduction Negligible molten flow within molten pool Radiation considered as boundary condition

1L ST T

f ,S L

L

T T TT T

ΔHf - latent heat of fusion Uniform temperature as the part initial temperature

15

Tl - liquidus temperatureTs - solidus temperaturefs - solid fraction

Heat Source Equations

Intensity distribution: a conical source: Horizontal – Gaussian distribution Vertical – Decaying with increasing of

penetration depth

η - electron beam efficiency coefficient

U - voltageU voltage

Ib - current

S - penetration depth

ΦE - beam diameterΦE beam diameter

xS, yS - horizontal position of heat source center

Hs - Gaussian heat source, Cline and Anthony

Iz - penetration function, Zäh and Lutzmannz p ,

16

Simulation Examplep

Process Simulation Animation

(3) Temperature Measurements- Near IR Thermography

Spectral Range B ild A Vi A

√√ Build Area View Access Resolutions (Spatial/Temporal)√√

Emissivity Transmission Loss

√√ Transmission Loss√

18

EBAM Temperature MeasurementsMeasurements

Near Infrared Thermal ImagerLumaSense MCS640 Spectral Range: 780 – 1080 nmSpectral Range: 780 – 1080 nm 640 by 480 FPA (Amorphous Si

based) Temperature Range: 600 to 3000 ˚CTemperature Range: 600 to 3000 C

(3 domains) Frame Rate: Max. 60 Hz Lens: ~ 500 mm Focal DistanceLens: 500 mm Focal Distance View Area: 32 mm by 24 mm Spatial Resolution: ~ 50 µm

19

Challenges:Emissivity and Transmission

Measurement Setupp

Emissivity (Single Setting, Estimated)Emissivity (Single Setting, Estimated)Transmission (Calibrated, 3 Ranges, with Glasses)

20Heat shield

NIR Video Examplesp

Build model:25. 4 mm square blockNIR Video 1

High Temperature Range

Medium Temperature RangeNIR Video 2

21

Temperature Profile Analysis (Hatch Melt)(Hatch Melt)

Single Frame (Raw Data) Average (One Video)

Transmission Loss Studyy

Controlled

Lighter side Darker side

Viewing AreaControlled Exposure

Sacrificial glass with two levels of metallization.

Thermal image on lighter side Thermal image on darker side

Controlled Exposure Experimentp pTemperature Profiles Observed Through Different

Levels of Metallization

y = 1 1602x 171 852250

2500

2200

2400

2600

y = 1.1602x - 171.85R² = 0.9822

1750

2000

de T

empe

ratu

re (C

)

1600

1800

2000

2200

mpe

ratu

re (C

)

Light SideD k Sid

1250

1500

Ligh

ter S

id1000

1200

1400

Tem Dark Side

10001000 1250 1500 1750 2000 2250 2500

Darker Side Temperature (C)

800-8 -6 -4 -2 0 2

Distance (mm)

Average Temperature Profiles Relationship Between

24

g p(6.37 mm Build Height)

pTemperatures Observed Through Different Levels of Metallization

Molten State Emissivity Estimatey

1. Identify Measured Liquidus1. Identify Measured LiquidusTemperature

2. Solve for Apparent LiquidusTemperature, (function of

d li id t tmeasured liquidus temperature, assumed emissivity, etc.)

3. Solve for True Emissivity, (function of True Liquidus( qTemperature, Apparent LiquidusTemperature, etc.)

Estimated molten pool ε: ~0.28Estimate

Tmd

Estimate Tapp (ε=1), with Tm

Estimateεtrue with true Tm, εtrue value

25

measured with Tmmeasured

mTapp

Temperature Profile Compensationp p3000

Uncompensated

2200

2600e

(°C

)Compensated for Transmission Loss

Compensated for Transmission Loss and

1400

1800

empe

ratu

re Emissivity Adjustment

600

1000

Te

Average Temperature Profile (6 37 mm Build Height)

600-7 -6 -5 -4 -3 -2 -1 0 1 2

Distance (mm)

26

Average Temperature Profile (6.37 mm Build Height)

Simulation vs. Experiment

Simulation

Set 1 Build height = 26.53 mm

Set 2 Build height = 16.87 mm

2200

2600

3000

°C)

ExperimentSimulation

Beam Speed = 728 mm/sBeam Current = 7.2 mABeam Diameter = 0.55 mm

Experiment

Set 2 Build height 16.87 mm

1400

1800

2200

Tem

pera

ture

(°

600

1000

-7 -6 -5 -4 -3 -2 -1 0 1 2

T

Distance (mm)

27

Process Parameter Effect Speed Function (SF) Index

Process Parameter Effect1000p ( )

- System Setting Beam Speed, Current 500

600

700

800

900

peed

(mm

/s)

p , Tested Range

- SF20 – SF650

100

200

300

400

Beam

Sp

00 5 10 15 20 25 30

Build Height (mm)

14

16

18

A)

4

6

8

10

12

Beam

Cur

rent

(mA

280

2

4

0 5 10 15 20 25 30

B

Build Height (mm)

NIR Images – Different SF Indices

SF 20 SF 36

SF 50 SF 65

29

Melt Pool Size Comparisons7.7 mA current, 0.65 mm diameter, 6.65 mm build height

Temperature AnalysisBeam Speed EffectBeam Speed Effect

7.7 mA current, 0.65 mm diameter, 6.65 mm build height

(4) Part Characteri ation(4) Part Characterization

Microstructures (Phases, Grain Sizes) M h i l P ti (E H YS UTS)

√√ Mechanical Properties (E, H, YS, UTS)√

32

EBAM Ti-6Al-4V Microstructure

Side surface (X-plane) Scanning surface (Z-plane)

33

Process Parameter Effects

300140

220

260

100

120

140

e, µ

m

dth,

µm

Columnar β Equiaxed β

100

140

180

40

60

80

quia

xedβ

size

lum

nar β

wid

20

60

15 25 35 45 55 650

20

40 Eq

Co

15 25 35 45 55 65Speed Function

34

Mechanical Property (Nanoindentation)Z l X plane

120

130

120

130

Z-plane X-plane

100

110

E, G

Pa

100

110

E, G

Pa

Young’smodulus

80

90

20 36 50 65Speed Function

80

90

20 36 50 65Speed Function

6.0

6.5

7.0

6.0

6.5

7.0

H d

4.0

4.5

5.0

5.5

H, G

Pa

4.0

4.5

5.0

5.5H

, GPa Hardness

35

3.0

3.5

20 36 50 65Speed Function

3.0

3.5

20 36 50 65Speed Function

Split-Hopkinson Bar Testing(P Allison)(P. Allison)

Uniaxial Tensile, Quasi-static and Dynamic Mechanical BehaviorMechanical Behavior

YS: ~975 MPa, UTS: ~1030 MPaTesting Video

3636

1 mm