University of Cambridge

1
University of Cambridge Department of Materials
Science and Metallurgy
By
Miguel Angel Yescas-Gonzalez
2
CHEMICAL COMPOSITION OF CAST IRONFe C
Si Mn P S
Mgval. 3.5 2.5 0.25
0.038 0.015 0.05
Only in Ductile iron
Grey cast iron
No addition of Mg or Ce
Tensile strength 150-400 MPa
Elongation 0
Ductile cast iron
Addition of cerium or magnesium to
induce nodularisation of graphite
Tensile strength 350-800 MPa
Elongation 3-20
3
Microstructure of Ductile irons
4
A further improvement of ductile cast iron is
obtained with an isothermal heat treatment named
austempering
1. Austenitising between 850 and 950 C typically
for 60 min. 2. Quenching into a salt or oil bath
at a temperature in the range 450 - 250
C usually between 0.5 and 3 hours 3. Cooling to
a room temperature
5
Mechanical properties

• STRENGTH equal to or greater than steel
• ELONGATION maintain as cast elongation while
  double the strength of quenched and tempered
  ductile iron
• TOUGHNESS better than ductile iron and equal to
  or better than cast or forged steel
• FATIGUE STRENGTH equal to or better than forged
  steel.
• DAMPING 5 times greater than steel.
  
  

6
R. Elliott, 1988
7
Economical advantages and applications

• ADI has excellent castability, it is possible to
  obtain near-net shape castings even of high
  complex parts.
• ADI is cheaper than steel forgings
• ADI has a weight saving of 10
  
  
  
  

Gears
Automotive industry
8
Processing window

• The bainitic transformation in ductile iron can
  be described as two stage reaction

Sage I Austenite decomposition to bainitic
ferrite and carbon enriched austenite
g
g

a
r
Sage II Further austenite decomposition to
ferrite and carbide
g

a
Carbide
r
9
Closed processing window
10
Microstructure of ADI

• Bainite
• Retained austenite
• Martensite
• Carbide
• Pearlite

11
Element Cell boundary Close to graphite
Mn 1.73 0.40 Si 1.75 2.45
Mo 0.60 0.07
Element Cell boundary Close to graphite
Mn 0.81 0.57 Si 2.31 2.49
Mo 0.16 0.12
Fe-3.5C-2.5Si-0.55Mn-0.15Mo
12
o
homogenised at 1000 C for 3 days
Austempered at 350 C for 64 min
13
(No Transcript)
14
(No Transcript)
15
Variables for modelling include C, Mn, Si, Mo,
Ni, Cu, Austenitising temperature and
time Austempering temperature and time
V
a b (C) c (Mn)
g
V
a b (C) c(Mn) d (C x Mn)
g
V
sin (C) tanh (Mn)
g
16
T
Mn
C
A
INPUT
C x W
c
Mn x W
Mn
HIDDEN
sum
OUTPUT
V
g
17
Modelling with neural networks
Hyperbolic tangents
a) three different hyperbolic tangents
functions b) combination of two hyperbolic
tangents
18
Modelling with neural networks
g
Input variables
Output or target
tanh (S w x q )
h
j
i
i
ij
j
19
(No Transcript)
20
Error bars
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
Physical Model for Retained Austenite
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
Babu etal. 1993
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
40 mm
39
(No Transcript)
40
(No Transcript)