Investigate the various parameter interactions of the FSW process
in hopes to build a model of the evolution of heat generation (heat
input) via FSW.
Continuously vary the spindle RPM at different feed rates
in Aluminum 7075-T7. All welds were performed by maintaining
a constant depth (instead of a constant z-force). Parameters
will be varied from 200-800RPM and 2-10 ipm.
In order to minimize variation in the data, four welds
were performed at each feed rate and then averaged together.
Two of the four welds were executed increasing the RPM
from low to high. The other two were executed decreasing
the RPM from high to low.
Some interesting trends were
discovered to depend heavily on RPM. These are described
In the x-direction (travel direction), forces are initially
high but decrease to a local minimum at about 200-300
RPM for all travel speeds investigated. As RPM increases
beyond this minimum, the X-force continuously rises to
a local maximum. The RPM at which this maximum occurs
are travel rate dependant. Beyond this maximum, x-forces
at all travel speeds decrease and become increasing sporadic.
Figure 1: X-force as a function of RPM
and Feed Rate
In the y-direction (perpendicular to the travel direction),
forces are initially high but decrease to zero around
200RPM. As RPM increases, the y-force returns to its original
magnitude, but in the opposite direction. This maximum
depends on the feed rate, but falls between 250-300RPM.
After reaching this maximum, the y-forces switch back
to their original direction, crossing the axis between
300 and 400 RPM. Beyond this point, the y-forces tend
to increase at low feed rates and become increasingly
sporadic at high feed rates.
Figure 2: Y-Force as a function of RPM
and Feed Rate
In the z-direction, forces are initially high at low
RPM. The force decreases nearly linearly with increasing
spindle speed. The slope of decrease is not the same for
all feed rates.
Figure 3: Z-Force as a function of RPM
and Feed Rate.
A typical plot showing tool temperatures and tool depth
throughout a weld is shown in Figure 2 above. The inset
picture shows the locations of three thermocouples, each
of which is within 0.050 in. (1.27 mm.) of the tool/workpiece
interface. Tool temperature (left) and approximate tool
depth (right) are plotted versus weld time. As can be
seen in the plot, as the tool is plunges into the workpiece
(decreasing tool depth) the tool temperatures raise over
time and then level out, reaching a final plateau near
the end of the weld. Average values for the temperatures
and z-force were taken for this last plateau and are displayed
on the plot. It can be noted that small fluctuations in
tool depth affect tool temperature. It is interesting
that the temperature at pin center is considerably higher
than the temperatures at the other two locations. This
result is consistent between welds. Current efforts are
focusing on obtaining temperature data and analyzing the
infrared images to confirm tool surface temperatures at
a number of different weld parameters.
Although much work is still to be done by way of numerical
modeling, initial results show the method to be very promising.
One numerical prediction, which correlates with the parameters
used to obtain the plot shown in Figure 2, gave the following
temperatures: Pin Center: 521 C; Root:
457 C; Shoulder: 479 C. It is
interesting here that although the temperatures are higher
than those shown in Figure 2, the trend showing the Pin
Center as the hottest, followed by the Shoulder and then
the Root as the coldest is exhibited in both the numerical
and experimental results. Further work will need to be
done to adjust the numerical heat input until the numerical
and experimental results agree.