On May 8, 6:34 pm, Mike the Strike wrote:
FAA initial report:

IDENTIFICATION
Regis#: 747JN Make/Model: 4M Description: SCHEMP-HIRTH
NIMBUS-4M
Date: 05/04/2008 Time: 2122

Event Type: Accident Highest Injury: Minor Mid Air: N
Missing: N
Damage: Substantial

LOCATION
City: EPHRATA State: WA Country: US

DESCRIPTION
AIRCRAFT ON TAKEOFF ROLL, WING STRUCK THE PAVEMENT AND THE AIRCRAFT
CARTWHEELED, EPHRATA, WA

In somewhat of a contradiction to the FAA report, newspaper and
eyewitness accounts state that the glider was on initial climb after
takeoff and was at 40 or 50 feet when it experienced an uncontrolled
roll. Extreme deflection of the wings on impact reportedly reduced
the damage and injuries.

I think that "uncommanded" roll might be more precise than
"uncontrolled", for technical reasons.

The newspaper account said, "He noted the airport changed its
operational procedure and added obstacles on the ramp. While not the
cause, Funston said, 'It was certainly a complicating factor'."

For mental exercise, let's forget about this particular accident for a
moment and review the possiblities based on these 2 scraps of
information: uncommanded roll and new obstacles.

Let's think about the range of possibilities:
First, we can assume that it's an accident, hence something
unpredicted happened.
Therefore, what caused the upset was either unpredictable or
considered unlikely.
A: Causes of aircraft flight accidents fall into just 3 categories:
1: Something about the a/c "broke" (in the broadest sense,
including linkages not sufficiently connected)
2: Something about the pilot "broke" (again, in the broadest sense,
including natural failures of perception, which we call 'illusion.')
3: Something about the air was "broken" (again, broadly, especially
including vortices = turbulence).

There is one component of this situation that is invisible: the air.
Its motions become apparent only when dust or smoke are swept along.
Then we're aware of remarkable turbulence, which must occur when it's
windy.

Any time there is a strong wind, complex vortices are shed by
obstacles, which persist for a long distance downwind relative to the
height of the obstacle. I was years ago taught a rule of thumb that
amounts to a 1:60 ratio of height to persistence. By this rule, a 15-
foot obstacle would cause turbulence that persisists 300 yards
downwind along a plateau.

Now, anytime the wings of an aircraft experience different winds, a
roll moment will be induced. This does not require either wing to be
stalled!

For example, a friend rolled his helicopter up into a ball a few years
ago. He was crop-dusting in nearly -still air, and flew head on into a
dust devil. His airfoils obviously experienced dramatic differences
in airflow, and a dramatic uncommanded roll occurred, and in a second,
he was surrounded by wreckage, forturnately unhurt.

For another example, a man was flying a single-engine a/c to landing
in a 20-kt crosswind last November. The instructor pilot watching on
the ground said that the airplane was level at 40 ft, when it abruptly
rolled almost to vertical, struck a tip, cartwheeled, crashed inverted
an burned. All 4 occupants died.

For another example, a friend, a commercial pilot, was landing his SEL
in gusty winds last year, and as he prepared to flare, it suddenly
felt as though his left wing had suddenly lost lift. He managed to
avoid a tip-strike and accident.

Why do our wings rock when we fly through gusty winds? because the
wings experience rapid local changes in airspeed and AOA in
dimensions smaller than the wingspan. Might sometimes this be
sufficient to cause a snap roll? Why not?

I write this not to analyze Mr. Funston's accident, but because the
concepts of "uncommanded roll" and "new obstacle" inspire me to remind
fellow soaring pilots that windy conditions are always accompanied by
significant turbulence close to the ground, exacerbated by nearby
obstacles. This turbulence is invisible, and can indeed roll our
aircraft enough to cause a tip strike or loss of control.

Dan Johnson
Menomonie, WI