Direct Metal Deposition: Process Control,Properties and Applications

1
Direct Metal Deposition Process
Control,Properties and Applications

• Jyoti Mazumder
• Center for Laser-Aided Intelligent Manufacturing
• University of Michigan

2
Objective

• ? Simulate the temperature and velocity field in
  DMD process.
• ? Predict the liquid pool geometry and laser
  cladding size.
• ? Dynamics relating laser input controlling
  parameters to output parameters to Fabricate
  materials with desired Properties
• - Input controlling parameters laser power,
  traverse speed, powder flow rate, laser beam
  size
• - Output parameters pool temperatures, liquid
  pool geometry and cladding size.

3
DMD Process Overview

• Blending of 5 common technologies
• Laser
• CAD
• CAM
• Feedback Sensors
• Powder Metallurgy

Laser Beam Channel
Powder Delivery Channel
Water Cooling
Shaping Gas
Changeable Tip
High power laser builds parts layer-by-layer out
of gas atomized metallic powder
4
Closed Loop Feedback Controller
Feedback Controller controls dimensional integrity
Outer Diameter of Cylinders 25.4 m
5
WHY DIRECT METAL DEPOSTION (DMD) ?

• True Metallurgical Bond
• Near Net Shape through feed back control
• Low Heat Input (smaller HAZ)
• Minimum Distortion
• Minimum Post Processing
• On the Fly Mixture
• Automated Process Control
• 6-axis DMDCAM software
• DMD Vision for small parts (turbine components
  etc.)

6
Modeling for Process Control
7
Numerical Heat Transfer and Fluid Flow Model
? What is considered? -- heat transfer, -- phase
changes, -- mass addition, -- fluid flow, --
interaction between the laser beam and the
coaxial powder flow.
? What can be simulated? -- evolution of
temperature and velocity fields, -- composition
profile, -- liquid pool geometry and cladding
size, -- evolution of liquid-gas interface.
8
Evolution of Liquid Pool
Laser beam
Solid-liquid interface
Liquid pool
Laser power 1200 W, beam diameter 1.2 mm,
scanning speed 300 mm/min, and powder flow rate
8 g/min.
9
Evolution of Liquid-Gas Interface
Liquid-gas interface
Top surface
10
Temperature and Velocity Fields for Top Surface
and Cross Section
11
Temperature, Velocity Fields and Thermal Cycles
12
Temperature along Laser Scanning Direction
Laser beam
13
Cladding Size as a Function of Time
After about 800 ms, the liquid pool reaches
steady state.
Laser power 1200 W, beam diameter 1.2 mm,
scanning speed 300 mm/min, and powder flow rate
8 g/min.
14
Summary
? Heat transfer, phase changes, mass addition,
fluid flow and interactions between the laser
beam and the coaxial powder flow were considered
in the calculation. ? The evolution of
temperature and velocity fields, liquid pool size
and cladding size during direct metal deposition
with coaxial powder injection were successfully
simulated. ? The liquid pool reaches steady
state after about 800 ms.
15
Process Control
16
Outline

• Closed loop control of molten pool temperature
• Generalized predictive controller (GPC)
  implementation
• Model identification and validation
• Temperature tracking experiments
• Deposition on a step surface with controller
• Composition sensing
• Line selection
• Composition monitoring for Fe-Cr, Fe-Ni binary
  element deposition
• Composition monitoring for Fe-Ti eutectic
  material deposition

17
Molten pool temperature control

• In real time using dSPACE 1104 controller
• State space model was identified using subspace
  method
• Generalized predictive controller was used to
  control laser power
• Dual color pyrometer was used for temperature
  measurement
• Filtered and normalized signal
  ß0.8

GPC Controller with Constraints
Laser
D/A
D/A
Data processing
18
GPC Controller with constraints-implementation
1. GPC Controller
2. A/D D/A interface to DMD process
19
Model Identification and Validation

• Diode laser power 1.1v - 900W
• H13 powder flow rate 10g/mim
• Scanning speed 650mm/min
• Standoff 20mm (beam size 2mm)

Data for model identification
20
Molten Pool Temperature Control

• Simulation
• Weight on control 100000000
• Prediction horizon 30
• Control horizon 5
• PI gain 20
• Tfilter 1 -0.8

• Experimental
• Weight on control 2105
• Prediction horizon 16
• Control horizon 5
• PI gain 1
• Tfilter 1 -0.8

21
Deposition on a Step Surface with Controller
Cladding
A
z
(a)
(b)
a
b
y
Substrate
x
(d)
(c)
Pictures of the deposition at (a) 10th layer, (b)
20th layer, (c) 30th layer and (d) 40th layer
Cladding height at different layers
22
Compositional Monitoring

• Relationship between plasma and composition

Composition1
Composition2
23
Compositional Control Identify Fe-I lines and
Cr-I lines

• Fact 90 Fe and 5 Cr in the powder
• Rule 1 no overlaps between Fe-I line and Cr-I
  line
• Rule 2 comparable strength of Cr-I line and
  Fe-I line

Cr-I 429.043nm
Cr-I 434.94nm
Fe-I 434.45nm
Fe-I 430.97nm
24
Line ratios to element composition ratios Fe-Cr,
Fe-Ni
Fe-Ni
Fe-Cr
Log-log linear relationship
25
Line ratios to element composition ratios Fe-Ti
Odd point at eutectic composition
Hypereutectic Ti56Fe44
Hypoeutectic Ti73Fe27
Eutectic Ti67.5Fe25.5
26
Summary

• Molten pool temperature control
• Molten pool temperature was successfully tracked
  to a preset temperature profile
• Controller was able to compensate the lack of
  deposition by adjusting the laser power
• Compositional monitoring
• Log-log linear relationship between the
  composition ratios and the plasma line ratios
  were obtained
• Odd point for eutectic composition was observed

27
Mechanical Properties
DMD Process Overview
28
H13 Impact Properties
DMD Process Overview
29
APPLICATIONS
DMD INDUSTRIES
AEROSPACE REMANUFACTURING
DEFENSE RESTORATION
OIL GAS SURFACE PROTECTION
MEDICAL FABRICATION
AUTOMOTIVE PRODUCT ENHANCEMENT
30
Negative CTE Structure (Completed DARPA Funded
project at UofM)
Completed Structure (2.4 in x 2.4 in x 0.5 in)
Green- Nickel, Blue- Chromium
X
31
Test Results
Strain vs. Temperature, Test 2
Strain vs. Temperature, Test 1
dL/L
dL/L
0 C
300 C
0 C
300 C
Temperature (C)
Temperature (C)
Both Results in y-direction
32
Thermal Property (P20 steel)
33
Conformal Cooling More Uniform Cooling Line
Distance to Tool Face
Part Temperature Variation due to Differences in
Water Line Distances
34
Wheel Cover Injection Mold
CAD model showing conformal water lines
Pre-DMD stage with machined water channels (P20
tool)
35
WHEEL COVER INJECTION MOLD
30 CYCLE TIME REDUCTION
Laser cut plates placed on water lines
After DMD
36
Case Study Crisper Bin Mold

• Technical challenge
• - Increase cooling rates
• - Gloss level matching

• Laser (DMD) solution
• - Conformal channels
• - 10 tools in production

• Economic impact
• - Reduced cycle time from 50 To 19 sec .

37
Door Molding Tool Part warpage
Warpage in a Conformal tool is HALF of
Conventional tool
Conventional tool
Conformal tool
38
Cooling cycle comparison between straight line
(conventional) cooling and conformal cooling
39
Mixed-material mold H13 tool steel deposited on
a copper substrate using DMD technology

• Reduce per unit injection molding cycle times by
  upwards of 25 relative to conventional
  single-material tool steel molds due to heat
  transfer improvements arising from the use of a
  copper substrate
• Mixing of materials in order to improve mold
  thermal conductivity is termed an Integrated
  Mixed-Alloy Heat Sink (iMAHS), which can be
  considered an instantiation of the broader
  category of Functionally Graded Materials.

40
Energy and Carbon Footprint Savings for
SteelCladTM DMD
41
iMAHS Technology implemented in designing a mold
cavity and a wear resistant coating
C

• Similar use-phase productivity gains to those
  observed in the case of iMAHS technology can be
  achieved in the injection molding process by
  using CCC which wrap around the contours of a
  mold cavity
• A wear-resistant coating applied to a forging
  tool via DMD for connecting rods in mass-produced
  power-train components

42
Inconel BLISK Repair

• Task
• Restoring worn out airfoils in a BLISK

• Challenge
• Minimize part distortion during re-building
• Minimize HAZ
• Maintain material integrity
• Precision part pick up

• Benefits
• Better quality
• Reduced post-machining
• Faster repair
• Cost savings

43
Spatial Control of Crystal Texture
Directional growth of grains from bottom to top
of the blade
44
Spatial Control of Crystal Texture
lt001gt crystal direction (red) preferred growth
45
Titanium parts

• Challenge
• Minimize part distortion
• Integrity of the build-up
• lt2000 ppm Oxygen

Actual part

• Benefits
• Cost savings due to material saving
• Saving in fabrication time and steps
• Provides design functionality

46
Frequently Asked Questions

• Deposition Rate 1 to 20 in3/hr
• Deposition Speed 500 to 2500 mm/min
• Beam Diameter (spot size) 1mm to 5mm diameter
• Layer Thickness 0.1 mm to 1.6 mm
• Hardness Fully Hardened as-quenched
• Powder Efficiency 40 to 75 (application
  dependent)
• Post-DMD Machining CNC/EDM/Grinding

47
THE END
Any questions?
THANK YOU for your attention