"Multi-Technique Problem Solving in an Industrial
Environment"
Shell, Thornton Laboratories, Chester-
Wednesday 6th of January 1999
"Diffusion Bonding of Al/Li Alloys"
Chris Pratchett1, Bob Wild1, Simon
Church2 & Steve Harris2
1: Interface Analysis Centre, 121 St Michael’s Hill, Bristol
2: Soweby Research Centre, British Aerospace, Filton, Bristol
| |
 |
| |
Figure 10: Diffusion Bonding Rig at
Bristol University |
| |
|
Diffusion bonding aluminium alloys is difficult due to the formation of
a stable Al2O3 oxide on the surface of the material. The oxide film blocks the movement of
aluminium across the bond interface during the bonding process, resulting in the formation
of weak joints of in most cases no bond at all. The oxide film has to be modified in order
to allow the bonding process to take place and this can be done using either mechanical or
chemical pretreatments prior to bonding. Using a specially developed bonding press, with
pretreatments developed at BAe and the IAC as series of bonding trials were performed at
Bristol University.
At set of standard bonding conditions were developed for the process,
these were as follows:
 |
Heat treat at 530°C for 30 mins, water quench, anneal at
185°C for 5 hours.
|
|
Surface Pretreatments: Mechanically polish with 1200 SiC
grit or Chemically etch with treatment A or B
|
|
Press Parameters: Temperature (500°C to 600°C), Time (10
mins to 120 mins), Pressure (0.5 MPa to 3 Mpa)
|
| |
|
|
|
Figure 11:
Effect of Bonding Pressure on Joint Shear Strength |
|
| |
|
|
|
Figure 12:
The Effect of the Surface Pretreatments on Joint Strength |
|
| |
|
Using 1" diameter samples, an optimum set of bonding parameters
was developed and the most effective bonding treatment determined. The results obtained
from these experiments are shown below in Figures 11 and 12.
TEM analysis of the bonded panels showed that even using the optimum
set of bonding conditions, there were still micro-voids alone the bondline. These voids
changed their shape and composition when the panels were heat treated after bonding. In
order to examine these features in more detail, specimens were fractured in-situ and
analyzed using AES.
The samples fractured in-situ showed clear differences to specimens
fractured in air then analysed. There are clear differences in the Li and Al signals in
the 25-75 eV region of the Auger spectrum.
 |
|
 |
Figure 12: TEM showing
Micro-Voids at the DB Interface |
|
Figure 13: AES Spectra from Al/Li samples
fractured (a) in-situ and then (b) exposed to air |
| |
|
|
In-situ fractured samples show spectra
indicating a Li (metal ) rich layer at the locus of failure, which is then converted to a
Al/Li oxide based film when the samples are exposed to air. The thin Li rich layer (~ 2nm)
must form at the interface during the bonding process, but is normally not detected, since
the surface composition is altered when the samples are fractured. Similar processes may
occur in the voids during the heat treatment process, with the conversion of a lithium
rich metallic layer to a mixed Al/Li oxide film.
|
