Gentlemen,
After much exchange of ideas and techniques with several of you,
I finally cut into my cabin door this past weekend to
replace a cracked window. I discovered this crack—only visible by looking
at an angle into the sandwich between the inner and outer skins in the upper
back corner—during pre-flight. The crack did not extend into the
transparent part of the window. This was one of the early “fast build”
doors from Lancair; purchased in 2004-2005. I will post a detailed report when
I’m done, but thought you’d be interested in what I found to
date…I was surprised.
As this fault was observed directly opposite a door latch, I
assumed it resulted from the holes drilled through the Plexi to accommodate the
latch. As the door/window was still intact (not blown out), I was careful to
grind away carbon from the outside to free the window without further damage so
we could confirm this. As I removed the carbon, I exposed a 4” crack
starting about an inch below the latch and extending toward the top of the door
and around the corner radius. The crack was about ¼” “under”
the carbon skin, and had not progressed to the edge of the Plexi at either end.
It did not originate at the bolt holes, but was generally parallel to the edge
of the Plexi and about ½” farther from the edge than the bolt holes. The
crack appeared to be caused by forceful flexing of the Plexi parallel to its
edge.
I am not a structural engineer, but believe
this was caused by one or more of three possibilities
(Clark Still gets credit for suggesting the second one).
First hypothesis: the latch exerts a twisting moment on
the door frame when closed and the door is pressurized. Possible, but I checked
the latch alignment and function carefully during initial installation. All
of the latches are aligned the same, and work very easily…no binding and
no stress upon opening or closing. They also engage fully, and only very
slightly “pull” on the door frame when going over-center. No other
latches showed any damage or stress in the area, and one of the forward latches
is also bolted with holes drilled through the Plexi. The process of
pressurizing is a gradual increase in force on the door/window, allowing them
to conform—within their stress limits. I therefore believe it
unlikely this crack was caused by the door latch or associated mounting holes
in the Plexi.
Second hypothesis. The window fracture was
caused by taxiing the airplane with the door open, allowing it to flex against
the air strut during upswings. This is a possibility. Clark Still observed
during our travel to Alaska last summer that I often taxi with the door fully
open (I think he’s jealous because he needs to secure the door to start
his IV-P Turbine). I have done this from the first flights, but have always
closed, or held the door partially closed, during any taxi over rough surfaces…and
this has been very infrequent. I don’t think this was the cause, but the
failure mode I observed in the window was likely caused by door frame flexing
that this situation would cause, so it’s a possibility.
Third hypothesis. Starting with the
FB doors, Lancair re-oriented the lift strut location so that its attach point on
the door frame is just above the ”offending” latch
location, and it’s mounted to the door stiffener. On at least two known
occasions, other people have opened the canopy door in the wind and allowed the
door to extend forcibly against the lift strut internal stop. This
would cause a sudden twisting of the door frame in the location of the crack.
Although both violent door openings that I am aware of occurred
some time ago, I believe this to be the cause of the failure--subject to
further review/ideas/comments from you.
A couple of related observations. The door was well constructed
in the areas I opened. I found no voids or weak areas. Holes in the Plexi for
through-bolts had been drilled to approximately ½”, filled with Hysol and
redrilled. All four of them were intact, and these “spacers” quite
forcefully held the inner and outer skins together. The window had been bedded
against the outer skin with Hysol, and bedded with micro around the edges. The
window was well attached to the door skins, although it could be separated with
a thin putty knife. The bond appears to be mechanical—the Hysol adheres
by conforming very closely to the rough surface without making a chemical or
adhesive attachment. More to follow as I have data…
Bill Harrelson and I took a lot of pictures, and I’ll take
more as we complete this job. I’ll write up the details for posting on
the LML and LOBO web sites for those interested. I’d like to extend a
special thanks to Laurie Fitzgerald (Australia), Brent Regan and Brad Simmons
for their exceptional advice, pictures, information—and encouragement. As
with most things done on this airplane, I’m combining their excellent
recommendations to hopefully make a long-lasting and reliable fix.
See you at Oshkosh!
Bob