Abrasive flow finishing

[jb24 at chrysler.com]

Gary Derian wrote:

>>The Reynolds number is very important in determining fluid flow but is
is the ratio of inertia loss to viscous loss in a fluid stream.  Re =
(density x velocity x length) divided by viscosity and is
dimensionless.<<

Knew this, just was trying to keep it simple.  Should have known better.

>>Adding bumps has no effect on Reynolds numbers.  Very low Re < 2,000
means
flow is always laminar.  Re > 5,000 means flow is always turbulent.  In
the
range between 2,000 and 5,000 the flow is mixed and can be pushed into
the
turbulent or laminar zone by careful design.<<

Actually, the question becomes where does Reynolds number apply?  The
boundary layer flow is what is important to Reynolds number.
Effectively the boundary layers intersect in internal flow eventually.
Other forms of drag or flow loss are calculated for external flow
differently, i.e. the so-called skin friction drag which is related to
the total surface area.  It all depends on how the coefficients of drag
are measured.  A rough surface contributes to higher drag at certain
Reynolds numbers and not at others.  Again, the bumblebee example.

>>Adding surface roughness puts more energy into the boundary layer and
may result in keeping it attached to the part.  The increase in drag
caused
by the turbulent boundary layer is more than made up by the reduction in
drag caused by the streamlines remaining attached to the part.<<

If you try to visualize why a tripped turbulent boundary layer helps
streamlines turn corners better, think about rate of pressure change on
a part.  At the leading edge, flow is pushed apart, at the trailing
edge, the ambient pressure tries to push it back together.  A tripped
turbulent boundary layer is "thinner" and therefore the streamlines are
"closer" to the part and don't have to push through as large a mushy
boundary layer to turn a corner.  A more rapid change in direction is
not long enough for the pressure to recover, and turbulent boundary
layer or not, the streamlines are not going to make the corner and then
you have big standing vortices with the streamlines making nice gradual
turns around the vortice.  Thinking of it this way helped me through my
fluid dynamics courses.

John Bucknell is jb24 at chrysler.com