Magnetic Field Control of the Mould Filling Process of Aluminum Investment Casting

1

• Magnetic Field Control of the Mould Filling
  Process of Aluminum Investment Casting
• S. Eckert1, V. Galindo1, G. Gerbeth1, W. Witke1,
• R. Gerke-Cantow2, H. Nicolai2, U. Steinrücken2
• 1Department Magnetohydrodynamics,
  Forschungszentrum Rossendorf
• P.O. Box 510119, D-01314 Dresden, Germany,
  http//www.fz-rossendorf.de/FWS/FWSH
• 2TITAL Ltd., P.O. Box 1363, D-59903 Bestwig,
  Germany

Sino-German Workshop on Electromagnetic
Processing of Materials Oct. 11-13, Shanghai,
China
2
Magnetic Field Control of the Mould Filling
Process of Aluminum Investment Casting

• Overview
• Motivation Aluminum investment casting as
  example of applied electro-magneto-hydrodynamics
• Measuring Techniques in liquid metals
• Numerical simulations volume of fluid
  method (VOF)
• Experiments
• Conclusions and remarks

3
Motivation
Aluminum investment casting

• casting of complex shapes
• casting unit U-bend with the filling channel
• (down sprue) and the mould being the legs
• connected by a horizontal channel
• the flow is exclusively driven by hydrostatic
• pressure

Problem
Too high fluid velocities in the early stage of
the pouring process ? entrapment of bubbles and
impurities ? worsening mechanical properties
(first) Solution
Damping of the turbulent flow with a D.C.
magnetic field
4
Strategy
Magnetic Field Control of the Mould Filling
Process of Aluminum Investment Casting

• Numerical simulation
• - finite element code FIDAP
• Labor experiments with a cold liquid metal
• - Flow observation, velocity measurements
• - Validation of numerical results
• Extrapolation to industrial scales
• - Numerical simulation
• - Magnet system design

Model experiments with GaInSn
Experiments with Aluminum melt under industrial
conditions
5
Numerical Simulation

• Finite element code FIDAP solution of the
  Navier-stokes equation for the flow and, in the
  case of applied static magnetic fields, an
  additional species equation for the electrical
  potential F
• 4 nodes (2d) or 8 node (3d) per element, bilinear
  interpolation
• Segregated algorithm for coupled equation system
• Volume of fluid (VOF) method for simulation of
  the filling process void fraction is a new
  unknown scalar
• 2 turbulence models Prandtl mixing length
  hypothesis and standard k-e

Sketch of the volume of fluid (VOF)
method Interface surface with a given constant
value of the void fraction
6
Numerical Simulation
(non-dimensional) Governing equations I

• momentum conservation time dependent
  NavierStokes equation with Lorentz force density
  term
• mass conservation

where is the Reynolds number and
is the electromagnetic
interaction parameter. is the density,
is the dynamic viscosity, is the electric
conductivity, B0 is a characteristic magnetic
field strength, L is a characteristic length and
v0 is a characteristic velocity.
7
Numerical Simulation
(non-dimensional) Governing equations II
In the case of a present static magnetic field it
is necessary to solve an additional equation for
the electric potential

• electric charge conservation taking into
  account the Ohms
• and Kirchhoffs law, following
  equation for the electric potential holds
• boundary conditions isolating walls, non-slip

8
Numerical Simulation
Material properties characteristic numbers
The system is defined through 2 independent
non-dimensional characteristic parameters Re and
N
Material properties Al GaInSn
density ? 2355 6360 Kg/m³
dynamic viscosity ? 1.45 2.164 10-3 Pa s
electric conductivity s 3.73 3.2 106 ?-1m-1
taking L0.03 m (channel height), v00.5 m/s and
B0.5 T we obtain
Characteristic parameters Al GaInSn
Reynolds number Re 24362 44085
interaction parameter N 23.8 7.6
9
Numerical Simulation
Application of a static magnetic field
down sprue
Electromagnetic force density acts as damping
force
mould model
z
y
x
pole shoes
Sketch of the casting unit
computational grid
10
3D-Numerical Simulation with the VOF Method
Application of a static magnetic field
Attenuation of the maximal velocity value at the
beginning of the filling process
t 0.15 s
B 0 T
B 0.5 T
Velocity vector plot on a cut plane in horizontal
channel. Vmax1.25 m/s corresponds to color red
11
3D-Numerical Simulation with the VOF Method
Application of a static magnetic field
t 0.25 s
B 0 T
B 0.5 T
Velocity vector plot on a cut plane in horizontal
channel. Vmax1.25 m/s corresponds to color red
12
3D-Numerical Simulation with the VOF Method
Application of a static magnetic field
t 0.5 s
B 0 T
B 0.5 T
Velocity vector plot on a cut plane in horizontal
channel. Vmax1.25 m/s corresponds to color red
13
3D-Numerical Simulation with the VOF Method
Application of a static magnetic field
t 0.75 s
B 0 T
B 0.5 T
Velocity vector plot on a cut plane in horizontal
channel. Vmax1.25 m/s corresponds to color red
14
3D-Numerical Simulation with the VOF Method
Application of a static magnetic field
t 1 s
B 0 T
B 0.5 T
Velocity vector plot on a cut plane in horizontal
channel. Vmax1.25 m/s corresponds to color red
15
3D-Numerical Simulation with the VOF Method
Application of a static magnetic field
Simulation with the help of the VOF (volume of
fluid) method Video ? left without right
with magnetic field video shows the first 2.5
seconds of the filling process
B0.5T
B0T
16
Model experiments
Perspex model
down sprue

• model fluid GaInSn,
• liquid at room temperature
• D.C. magnetic field
• up to 850 mT
• transparent walls

mould model
horizontal channel
magnet pole shoe
17
Measuring techniques

• Visual Observation
• Ultrasound Doppler Velocimetry (UDV) see
  presentation about measuring techniques
• Inductive Flowmeter (IFM)

18
Measuring Techniques
Inductive Flowmeter (IFM)

• Measurement of the perturbation of the magnetic
  field by the flow
• voltage is proportional to the flow rate
• high temporal resolution
• can be applied at high temperatures

G geometry factor
19
Experiments
Velocity measurements obtained at a pouring
experiment with a) InGaSn at room
temperature b) AlSi alloy at about 700C
20
Model experiments UDV measurements
21
Model Experiments
UDV measurements in the down sprue
B 0
B 0.85 T
22
Model Experiments
UDV velocity measurement in the model experiment
in the down sprue channel
in the horizontal channel

• dangerous peak velocity in the early stage
  removed
• velocity fluctuations become smaller

23
Experiments
Validation of the numerical simulation
Comparison of numerical and experimental results
regarding the flow rate Q as a function of the
magnetic field strength (related to the flow rate
obtained at B 0)
24
Al casting Experiments
Casting units evaluation
Visual inspection of the resulting metal surface
UV-light visualization of surface defects


B0.25 T
B0.75 T in the first 5 seconds
15.01.2014
15.01.2014
25
Al casting Experiments
Casting units evaluation - Statistics
0 T 0.25 T 0.5 T 0.75 T 0.75 T 5s
CNF 6-4 4-4 4-2 4-4 3-2
CWF 4-4 4-3 4-1 3-3 1-2
FNF 4-1 4-2 2-2 2-2 2-3
FWF 6-2 3-2 2-1 1-1 2-1
Clear tendency the d.c. magnetic field always
lead to an improved quality of the casting unit
with reduced amount of entrapped oxides
26
Conclusions I

• Velocity and flow rate measurements are needed
  for better understanding of the flow phenomena
  and filling process of the investment casting of
  Al
• Low temperature metallic melt InGaSn has been
  used for model experiments key advantage at
  this temperature a sufficient number of different
  measuring techniques are available
• Validation of numerical codes using such liquid
  metal models provides a profound basis for an
  extrapolation of the numerics to the real scale
  problem and turned out to be essential for the
  reliable simulation of the real Al casting
  process
• Main problem in the pouring process the
  occurrence of large velocity values at the
  beginning of the casting processes leads to an
  accumulated generation of vortices inside the
  pouring channel ?

27
Conclusions II

• A high rate of turbulences in the flow is
  supposed to entail the transported impurities,
  oxides or gas bubbles from the walls and the free
  surface into the bulk of the casting patterns. As
  a result the mechanical properties are
  deteriorated
• The external d.c. magnetic field damps the high
  flow velocities at the beginning of the pouring
  process. A significant reduction of the peak
  velocities, leading to a generation of vortices
  inside the pouring channel, has been shown by
  model experiments and numerics, and has been
  demonstrated in the real Al casting process
  afterwards
• As a important input for the control system, a
  contact-less flow rate sensor has been developed
  and successfully applied
• The statistics of a multitude of cast units
  showed a clear tendency of reduced oxide
  entrapment due to the magnetic field influence

28
Perspective

• Next step linear A.C. traveling field which
  brakes initially, and pumps at the end ? constant
  flow rate during the whole process

Induced electromagnetic force density compute
with a finite element Maxwell solver OPERA
Scheme of the coil system to generate the
traveling magnetic field