Fluids Lab Report
Joe Webb
Candidate No: 16268
F1.1:
The results show that the shape with the greatest coefficient of drag
(1.70) was the triangular prism with the flat leading edge (second from top on
figure 4). This is predictable because of its flat surface giving the body excessive
form drag. The separation point will be right at the leading edge of this body and
the following turbulence will create a wide wake and make the skin friction drag
of the trailing edge insignificant. Turning the shape around 180o does not solve
the problem (top of figure 4). It greatly reduces the form drag as the mass flow of
the air is deflected by a much smaller angle, but the flat trailing edge causes a
separation point at a wide section of the body and creates another wide wake.
This turbulence lowers the pressure at the back of the body and so causes a
negative pressure that combines with the pressure from the head on air flow and
increases its CD (0.59). This shape, however, still has a lower CD than the cylinder
(bottom of figure 4). The cylinder has a preferable trailing edge to the flat one,
but it still has a relatively early separation point and has increased form drag
from a ‘flatter’ leading edge. The cylinder has a CD of 0.64. Predictably, the shape
with the lowest CD (0.33) was the aerofoil, which combines a rounded leading
edge that allows laminar flow across it, followed by a longer trailing edge that
ends in a point. This allows the body to leave the narrowest wake, as the
separation point will be right at the very end of the trailing edge. The longer
airflow attachment area, however, will increase the likelihood of significant skin-
friction drag.
These results suggest the relationship that, for a constant Reynolds
number, an earlier separation point will cause a wider wake and causes an
increased coefficient of drag
To further reduce the drag coefficient, a sharper leading edge in front of
the aerofoil may further reduce the form drag, and also a smoother surface finish
of the body may reduce skin-friction drag.
The onset airflow has a high degree of turbulence that would cause a
lower drag coefficient. However, because the airflow is kept at a constant speed,
this will cause a systematic error and won’t affect the observed trend too much.
F1.2:
In the range of Reynolds numbers used in this experiment, the coefficient
of drag remains fairly constant. This is consistent with the curves of the smooth
sphere and cylinder, because these experimental results appear within the
region of the curves that also have a constant CD. At very slow airflow, the flow
will be minimally separated at the furthest point on the trailing edge, creating
steady vortices that generate a high drag, along with the skin friction drag of the
boundary layer. As the Reynolds number (airflow velocity) begins to increase,
the downwind vortices become unstable and start to oscillate. These oscillating
vortices create a wide wake and, therefore, a large drag, but less drag than the
initial stable vortices. The constant region of the graph occurs when the wake
becomes chaotic, but the boundary layer remains laminar. At this point, the wake
is narrower than that of the oscillating vortices produced in the previous range
of Re and the CD is therefore lower. A faster airflow doesn’t increase the size of
the wake, so the form drag remains constant. At Re = 2x105, CD drops suddenly.
This is because the boundary layer transitions from laminar to turbulent flow
and the separation point is initially further along the trailing edge, leading to a
narrower wake. Increased air speed, however, counteracts this effect and the
wake gets larger, increasing CD above the previous, laminar, drag value. If the
range was increased, it is likely that a similar pattern would emerge for our
‘rough’ cylindrical body.
The high onset turbulence in this experiment causes more of a problem
than in the first experiment as the airflow is changed to give an increased drag of
a certain value. The greater the speed of the airflow, the higher the onset
turbulence, and therefore, more drag force is lost on the body.
F1.3:
The streaklines of the cylinder showed a stagnation point at the front and
a separation point past the halfway point of the cylinder. The aerofoil, however,
showed clear, attached, laminar flow with a stagnation point on the leading edge
and behind the trailing edge. The streamlines were symmetrical so the aerofoil
did not experience any lift. As the angle of attack increased, the leading
stagnation point moved below the leading edge and the trailing stagnation point
stayed behind the trailing edge relative to the onset airflow. The streaklines
became narrower above the leading edge due to an increase in velocity. This
causes a decrease in pressure and thus creates lift. Also, a separation point
moved steadily up the upper face of the aerofoil and created an increasingly
large wake, increasing the drag coefficient and, at the stall angle, began to
overcome the increased lift due to the steeper angle of attack.
The onset turbulence caused the streaklines to become wavy and
oscillatory as they got further from the source. This, however, did not obscure
the results.
F1.4:
At an angle of incidence of 0o, the pressure distribution on both sides of
the aerofoil is the same; therefore, there is no resultant lift. The distribution
shows a trend of initially increasing negative pressure, followed by a shallow
gradient of decreasing negative pressure. The centre of pressure appears to be
about a quarter of the way along the aerofoil from the leading edge, as this is
where the highest negative pressure values are. (Figure F1.4.1)
As the angle of incidence is increased, (10o to 15o), the centre of pressure
moves towards the leading edge. The pressure difference between the upper and
lower face increases so that the pressure above the aerofoil is a higher negative
pressure, and the pressure below the aerofoil becomes positive. The highest
pressure value of 1 appears to occur around the stagnation point, and remains
constant for 10o through to 20o. The pressure profile shows decreasing negative
pressure on the top face and decreasing positive pressure on the lower face aft of
the centre of pressure. (Figures F1.4.2, F1.4.3)
At 20o, the pressure difference decreases and centre of pressure seems to
move towards the trailing edge. The profile for the upper face seems to remain at
a relatively constant pressure, (~1.5), and the lower face profile behaves
similarly to the smaller incidence angles close to the leading edge, but the
pressure decreases to a more negative value towards the trailing edge. (figure
1.4.4)
These results show that lift increases with angle of attack up to an angle,
(most likely somewhere between 15o-20o), at which the lift starts to decrease
due to turbulences caused by the increased form drag of the aerofoil’s effective
profile.
F2.1:
The theoretical curve for U/Ut is given by Bt/B. This result is displayed on
figure F2.1.1 by the solid blue line. The pink points represent the experimental
results calculated from static and total pressures. These results are reasonably
consistent with the theoretical values, but the velocity decreases at a lower rate
as the passage widens. This could be due to the viscous forces in the boundary
layer extending past the centre line of the passage. If this was the case, the air
flow would slow down due to the increased area, but would also be slightly
counteracted due to being further from the walls and therefore be slowed down
less by the boundary layer.
The curve for static pressure, (figure F2.1.2), shows that a change in air
flow velocity corresponds to an opposite change in static pressure, and that the
total pressure remains constant. This is to be expected as it is predicted by the
Bernoulli equation (Po = p + 0.5ρU2 + ρgh). The ρgh term in this experiment was
omitted because the change in height was too small to make a noticeable
difference in results. However, the form of the Bernoulli equation used in this
experiment can no longer be said to satisfy the continuity equation.
Viscosity will affect the velocities within a shallow diffuser, more than an
abrupt enlargement, as the flow will remain attached to the walls and will form a
laminar boundary layer. An abrupt enlargement will most likely cause a
separation point and the following turbulence will cause some of the kinetic
energy to be dissipated and the velocity drop will be greater.
F2.2:
In figure F2.2.1, with a high Reynolds number, the curve shows a
narrower boundary layer than with a low Re, as U reaches Ufreestream at a position
around 0.7-0.8 and the gradient becomes close to infinite. Figure 2.2.2 however,
shows a shallower gradient at “Ufreestream” in the centre of the passage. Therefore,
in this case, the boundary layer extends past the centre line of the passage and
the freestream velocity is impeded by the viscous effects of the boundary layer.
True freestream velocity is therefore never reached.
The velocity profile changes because at faster airflow, the viscous forces
aren’t as dominant. This causes the boundary layer to become narrower and the
freestream velocity to be reached sooner. For the faster airflow, the profile is
very similar to the slow airflow close to the wall, but the gradient becomes
steeper as it approaches the edge of the boundary layer. This means that the
airspeed approaches a faster freestream velocity slower than if it were a slower
freestream velocity (in this case between about 0.5 and 0.95). However, this is
balanced out because after this point, the faster freestream case approaches
Ufreestream at a quicker, more constant rate (with respect to position), whereas the
slower freestream case approaches Ufreestream increasingly slowly. If the surface
were much rougher, the viscous forces would be more dominant, so the
boundary layers would get thicker.