Ryerson test flight - Oct 29

[altonyeung]

https://www.youtube.com/watch?v=Vzpk5tPMrhM

Today we had a weird flight. The flight was first flying okay. However, at (3:53) and (4:18) the plane almost flipped upside down. From the onboard camera, it seems the plane had excessive sideslipping (caused by gust?) and causing the plane to roll.

I will post more information from the onboard telemetry data as well as the data captured by the ground anemometer.

[Frank v B]

That was a scary flight.  Yep, that was me at the sticks.  Twice the plane suddenly almost flipped on its back during the flight.
Of note in the video:
i) the right hand turn before the first flip was not initiated by me.  My plan was to fly counterclockwise circuits only.  There was never a right-turn input.  This happened in line with the clump of trees to the east of the pilot station.
ii) the second flip was similar but directly east of the pilot station and south of the clump of trees.

At first I thought it was a brown-out, then a stall, so my input was down-elevator to both speed up the plane to get out of the stall and see if the radio still worked (it did.) 

We discussed the problem at the field and now that I see the video I agree with Dr. Glenn N.'s assessment that it was a sideslip/yaw issue.  His solution was to add area to the fin.

Look carefully at the plane immediately after takeoff.  You will see it heading towards the trees on the west.  A very noticeable yaw element to the right.  There was never a right aileron or right rudder input.  Both were left inputs because the plane has a mind of its own until it is up to speed.

Alton will add fin area in front of the current fin and on top of the rudder.

Also note that when it flipped the second time I just wanted to bring the plane back in one piece.  The landing (northbound) was the opposite from the take-off (southbound) because it was the fastest way to the runway and away from the road.  The landing's first bounce was full "up" elevator.  There was no attempt to make the Goetex.  An airplane in one piece trumps and stylish landing attempt any day.

We will try again next week.

Frank

[octagon]

Good save Frank. Those guys are lucky to have a plane thanks to you. Hopefully when they go on to design full sized planes, they come up with something a little more pilot friendly!

[electroflyer]

    Frank,

   You did amazing! I truly believe that with a much larger vertical fin will tame this beast.  That was a wicked YAW on takeoff, which no doubt was because of the crosswind, but in flight a drift of the nose was definitely present and of course the Spin from which you recovered nicely. Well done!

 

   Glenn

[Frank v B]

Thanks Rob and Glenn.  Alton will add the new fin area for next Saturday.
Fingers crossed.


Frank

[Bill B]

Alton and I have been searching through data since the flight.

I've attached two sets of plots; the first shows airspeed and altitude over the flight, along with heading and track, and wind data from the ground and estimated wind data at the plane. You can see, at the time of both stalls, the heading diverged from the track quite a bit (as evident in the video). Interestingly, around 12:46:20, the plane was flying in the same direction, but did not stall. This could have been because of wind; our ground station shows that the wind picked up quite a bit just before the first stall.

The second figure shows the airspeed and heading for just the section of flight that includes the stalls, along with the roll and pitch, and control inputs (from Frank).

We're still going through the data, I just wanted to share what we have found so far.

[Frank v B]

Bill, thanks for the raw data, 

If I am right, it shows airspeed going to zero during the two events? a tail wind gust??

The last chart of control input shows the motion of a blind man moving his white cane around.  That's my flying style!

I thought about the two events (flips?) yesterday and today and came up with two more factors that could have contributed to the problem once the yaw was induced:

1) The wing has a swept back leading edge about halfway out from the fuselage.  When right yaw is induced, the port (left) outer wing half hits the oncoming wind straight on for max lift.  The left wing is the advancing wing half (travelling faster) during the yaw to the right.  The starboard (right) wing panel falls away from the oncoming wind during a right yaw and becomes the receding wing (travelling slower).  The bent leading edge will magnify this.  The left wing has more lift than the right wing, magnifying the right rolls during the event.

2) Both wing tips had small GPS boxes on the top surface.  They were square and about half the size of a match box and the same height and were 4"-6" in from the wingtip at the high point (spar?) of the airfoil.  The shape would create lots of vortices. During a right yaw event, the box on the right would blanket* the starboard wing tip out from the box and lose lift at the tip.  The left wing tip box would blanket the wind behind and to the right but leave the wing tip clean and lifting properly.  The net effect would be a roll to the right (a lever?, asymmetric lift?).  The old GPS boxes were hockey pucks and probably drew the wind around them with few vortices.  These new square ones had the aerodynamic qualities of a brick.

If you watch Mayday on TV, most airplane accidents occur when several unlikely events combine to create an untenable situation.

In our case we had 4 things go wrong:
1) a yaw to the right caused (magnified?) by too little lateral area per Dr. Glenn's observation.
2) the bent leading edge magnifying a right roll during a right yaw
3) the GPS boxes losing tip lift on the right wing magnifying a right roll
4) letting Frank anywhere near the transmitter.  ...........Over to you Andy!


Frank

* sorry, a sailing term.  In a race, when a faster boat passes to windward of another boat, his sails will blanket the slower boat's sails (no wind, no speed, lots of swearing)

[Bill B]

That is right, Frank, the airspeed did drop to zero during the two events. I didn't plot it on the chart, but the ground speed stayed around 12-14m/s during the stalls, suggesting the gust was quite strong.

In addition to your thoughts on the gps causing vortices that would blanket the wing... In a right raw situation, there would also be dirty air coming off the fuselage, firewall and motor, that would blanket the right wing as well. Also, since the plane was yawing right rather quickly (gust?), we can assume that the left wing was travelling much faster than the right wing, only making the situation worse. Possibly the right wing even had a negative airspeed for a second.

I agree that a larger vertical stab will help with this, but I'm not sure how this affects the takeoff. If the tail is too small, wouldn't we expect the plane to yaw left on takeoff? and not right, like we see in the video?

Also, I incorrectly labelled the direction of roll in the second figure, chart 3. It should be +RWD, meaning positive is right wing down.
Bill