BARON B55 TAKEOFF INCIDENT
REPORT TO BEECH AIRCRAFT OWNERS GROUP:
My Beech B55 Baron
flies often from a private runway in Virginia, and
we had just installed a new JPI-760 Twin engine monitor
to track temperatures of exhaust gases, cylinder heads,
oil, ambient air, and to monitor fuel flows. Several
recent discussions on an internet world-wide (Proper takeoff mixture and
fuel flow settings... The use of the engine
monitor... Engine failures... come to mind)
and twice-annually recurrent training made a big difference
in the outcome of an engine failure on takeoff recently.
Early in my departure from Virginia to Nova Scotia
via Nantucket, I noticed the takeoff fuel flow on
the right engine factory fuel flow guage (it's pressure
sensing, but marked as flows) were VERY low... 12GPH...
when it should have been 21GPH, and the brand new
EDM 760 Twin monitor was peaking at max EGT value
for all six cylinders, while the good left engine
was mid-range and normal (but I have less than two
hours of looking at it, as it was just installed).
Nevertheless it was pretty hard to miss the two obvious
clues: Split fuel flow needles on the primary
flow guage and a bright orange right engine
on the JPI, to say nothing of the real obvious fact
that the plane was not accelerating properly (I first
rechecked the gear and flaps to see if I had left
them hanging out, but no) and one engine was popping
(afterfire or backfire type sounds) and shaking.
Damn, I think first, a fuel pump failure or fuel leak.
(Two flight hours before this, we had run a leak test
on the new fuel lines; while doing that, we noticed
that the right boost pump would not bring up the flow,
and we found and fixed a small leak). But there were
no flames or smoke obvious, so I turned on LOBOOST
and then HIBOOST with only a small improvement up
to 15GPH. Engine was still sputtering, and I was making
every effort to prevent detonation (yet still climb),
so I reduced power on both sides to keep flying as
coordinated as possible ... ultimately down to about
2500RPM and 18-20MP and things settled out, but still
popping, engine very rough. We were able to climb
very slowly. I called CHO tower (15 miles from my
runway) and said that I was IFR to Nantucket, but
I needed to land immediately. Weather was clear with
good visibility, mimimal wind.
By then (30 seconds to one minute after the event)
I was beginning reject my fuel leak hypothesis and
starting to suspect a piece of trash in the fuel line
that had broken loose from a two-hour before installation
of the new JPI flowmeters. In that installation, the
main fuel lines to the engine are cut, flowmeters
added, and the system rechecked.
All seemed normal in the several hours I flew it post-installation.
(But that was not true! My neighbor who knows nothing
about aircraft but can fix his tractor said - after
the fact - that the plane did not sound right on my
approach the day before! From now on, he will call
if hears anything strange! ... and maybe noise-cancelling
headsets aren't always a good idea.)
I live on a grass strip that's really smooth, but
since it has been so terribly dry here the runway
grass is dried up, and much of its cushion has disappeared,
and the strip feels rougher than it normally does.
My first guess was that the vibration shook loose
a piece rubber from the inside of the fuel hose feeding
the flow divider and stuck in exactly the wrong place
Throughout all EGTs were flat across, but the right
engine was very hot, yet all CHTs were flat, but I
was pretty busy and did not focus on the exact numbers.
The worst part is that the problem occurred - or I
first noticed! - after liftoff and just before - maybe
5 Kts before - blueline (Lordy, I HATE IT WHEN THAT
HAPPPENS !) with full fuel and two up, small baggage
load, at 82F ... but since the failure was partial
I elected to keep going, which was not that hard to
do, but violating every preplanned "below blueline,
land it" scenario I had thought out in years
and years of thinking about it.
We did wake up anyone who was napping as it took a
number of miles to get to pattern altitude as we worked
our way out of the low ridges.. To have a preplanned
"worst case" route, incidentally, was taught
me in in the 1960s in the Cassiar Mountains of British
Columbia and the Yukon Territory by a great bush pilot.
He said always know - ahead of time - which valley
you can fly out of, or turn in. It will be too late
to figure it out when the engine burps.. He's right.
My normal takeoff procedure is as follows: leaving
my 3000 ft farm grass strip in 82F temps amd nearly
calm breezes, I did a normal runup at 1700 RPM or
so and all was well. Coming up to full takeoff power,
I always pause at 2000RPM for a thorough (engines,
temps, pressures) recheck before releasing brakes,
say out loud "Two good ones" and then up
to full power and another glance at the three important
guages (RPM, MP, Fuel Flows) while looking forward
and rolling. I leave the fingers of my throttles hand
forward of the handles below blueline (which slightly
blocks the pilot's view of the flowguages), preparing
to pull back and land without further thinking if
something breaks before blueline. When I get to blueline
in ground effect I get the gear except that when on
short runways I sometimes get it sooner, looking for
speed. At blueline, I switch my hand to behind the
throttles, pushing lightly forward and always say
But I was initially worried about having a fuel leak
and could not figure out if it was a leak why the
engine was rough and the EGTs were so hot, and why
the boost pumps would make no permanent difference
in flows (even with a leak, I would have expected
some permanent change in pressure) but I asked the
Charlottesville tower anyway "if the fire truck
wasn't busy would he please meet me at the center
taxiway ramp and coverup my right engine and us if
the engine caught fire after stopping?". They
said "Sure". We were all very relaxed...
That was silly of me (and inconsiderate, because I was so casual that the crash crew showed up but NOT WEARING their protective fire gear!!). I don't have any idea why I
did not say the "E" word but they got the
picture (and they know my aircraft). I also was not
as excited as always thought I'd be, just working
the problem, and trying to give them back their runway
as soon as possible. An interesting and polite idea,
but dumb, even though - by then - I was pretty sure
that we were not going to have a bad day. I probably
should have stopped it on the runway. But it only
took ten seconds more to clear the runway and taxi
up to the waiting fire truck, who I had asked to cover
up my right engine if it lit off.
I signed off the radio at the flare and shut off the
master at touchdown, and I had my passenger pop open
the door just after the touchdown, and had briefed
him to get out and walk to the rear as soon as we
stopped. He did great and was cool as could be, but
I forgot to tell him that he had to undo his seatbelt
first! Being a nice guy, as soon as he got it loose
he started to hang the shoulder harness where he found
it, and I finally had the sense to tell him "to
hell with stowing the belts, just get out" Should
have done that sooner, and been clearer with him,
So... nothing lit off and the firemen went back home,
and we found a friend and mechanic who was working
on his airplane; his quick analysis reconfirmed our
earlier suspicion of a blockage. After restarting
the plane I realized that it ran pretty well up to
2000RPM or so, but started to burp, just above that..
So we thought it was kind of neat that I could send
George Braly at GAMI the stored data and he can tell
me how well I can actually remember what happened!
That will be very interesting....
AND SO.... WE SENT OUT THE DATA, and included two
E-mails follow, first mine to:
Hello... George Braly [at GAMI], John Deakin [AvWEB],
and John Eckalbar [who writes great books on flying
Beech singles and twins]:
Attached is a JPI data file from the partial engine
failure at takeoff on my Baron 55 (TCM IO-470s @ 260HP
each) on Saturday.
I'd like to know how much I hurt the engine? (I'm
guessing not much as the temps got to 1600F, but only
briefly) and for George, I'm just curious if you can
tell how much power the right one was putting out?
Maybe you don't have enough data....
This JPI is brand new to me and this takeoff was only
my second or third with the new instrument. As I read
the data, I'm seeing 108 seconds of a Baron fighting
to fly followed by a out of whack but controllable
approach and landing. How do you see it?
I'm pretty sure now - in hindsight - that I missed
seeing the low flows on the initial takeoff roll.
Initially, I had thought the power loss occurred just
below blueline, but looking now at the recorded flows,
I think the 12 GPH flow on the right was constant
more or less throughout the flight. Am I right?
I think I know why, too. If I had my optional factory
twin needle EGT I'd have seen the overtemps immediately.
As I have said before, I find digital data hard to
read quickly, and so I saw the split flow needles
first, (I think!) and ultimately I caught the BRIGHT
ORANGE!! bars on the right engine, but my sorry excuse
is that it was a new presentation. Not a very good
reason, I'm afraid. Thanks for looking at this.
THE RESULT OF REPAIR WORK... on Monday following:
We had a brand new but completely blocked fuel line
because the new - and incorrectly installed - fuel
line nipples downstream of the flowmeters had made
a little "heart valve" out of the interior
wall. We could not even blow through it, and the other
engine was not a whole lot better.. Replaced both
and all is well. Baron flies great, idles and runs
smooth, too. Temps flat across, flows normal.
FOLLOWED BY RESPONSES FROM BRALY, DEAKIN, AND ECKALBAR:
(for a large high resolution version (287K) of this
FROM JOHN DEAKIN:
(My thanks to him for the excellent graph)
I've graphed out your data, and put a few notes on
it. You're home free. You actually got a nice LOP
climb setting, with 3 jugs staying under 350, and
the other three dropping a bit more. Your right engine
was putting out a nice 61%, or 182 HP. It would have
been rough, as three jugs (3, 4, 6) were quite LOP,
and the others were barely LOP. As you took the power
up for takeoff, both rose to 16.7, then the left one
kept on going (normally), while the right dropped
back to a very steady 12.2.
On takeoff, your crosscheck should be, "MP, RPM,
But you know that, now! <grin>
FROM GEORGE BRALY:
It looks to me like when the right engine went lean,
it went lean of peak. Didn't hurt a darn thing. You
could have run it that way (except for the roughness)
for hours and it would have been fine.
Also, I plotted it out. You only had 3 of six jugs
that went lean. The other three stayed rich. THAT
is a bit puzzling. [...as it turned out, some loose
rubber hose lining went downstream and clogged three
injectors] Looking at it, it is absolutely certain
that you were simply lean of peak. Notice the two
vertical reference points where the fuel flow increased
slightly [that's where the boost pumps were turned
on] - - and you had a corresponding rise in EGTs...
the only way that can happen is if the cylinders were
lean of peak. I think three of the cylinders were
so lean that they were no longer firing... or at least
doing so intermittently, and that is why it felt so
Absolutely certain that there was no harm done to
the engine when your engine experienced the large
loss in fuel flow on takeoff. Having good data sure
FROM JOHN ECKALBAR:
Congratulations on the successful outcome. Pilots
often wonder how they will react when the cookie starts
to crumble. It's always a real personal discovery...and
totally unpredictable, in my experience at least.
Sometimes everything slows down and gets real clear,
and sometimes you can't remember your wife's name.
I'm real glad it all turned out well for you.
I just saw George's note, and for what it's worth
I agree totally. The data looks like a big experiment
in extreme LOP operations...low power, no harm. EGTs
barely flirted with 1600F and CHTs were all low. One
of the beauties of the JPI is that when your heart
rate gets back into double digits you can
coolly analyze what was going on.
Your takeoff strategy is exactly what I've advocated
for years. I like the hand position routine especially.
Seems like that gets you poised to do what you have
to do. Looking at the data though, it looks to me
like proper fuel flow was never present on the right
engine. Looks to me like your takeoff chant might
have gone like this:
"Throttles up. RPMs good.
MPs good. Fuel flow...oops. Abort." That's the
virtue of having the hand ahead of the throttles.
Anyhow, I'm real glad we can all learn from this.
SO, WHAT DID I LEARN??
Here's the trap I got myself into: 600 hours in a
PBaron teaches (so does FlightSafety) to pause at
2000 RPM and spool up the blowers, then full power.
My later BPPP recurrent classes all teach "No
rolling takeoffs" also with a last minute power
check - just a pause on the way up - of all engine
instruments at 2000RPM. I always follow that with
a fullpower recheck of the three primary power gauges
only (the three above the levers -- RPM, MP, and Flows)
both at initial roll and a glance again just before
redline. The hand-on-throttle technique is widely
taught, however this incident also taught me that
my fist significantly blocks the view of the flowmeters,
particularly when they are at full takeoff flows and
pointing to the bottom of the gauge). The redline
glance caught the failure on the split flow needles
-- or maybe it was just a bit later -- but I cannot
tell from the data if I'm exactly correct about that.
No question that I missed seeing the failure for several
So, from now on, after going to full power I remove
my hand from the throttles, and point to each
power guage - this takes two seconds - then replace
my hand where I had it. Removing my hand forces me
to uncover and look at the hidden guages.
All this normal procedure (right or wrong) with a
new installation that had a really serious defect
in the only fuel hose to each engine, and a pilot
who has spent a lot of money on a new device who got
distracted by it for a few seconds on the initial
roll when he should have been looking at the primary
and familiar flow gauges.
But what a great device is the JPI EDM 760 Twin unit!
It traps data every six seconds. That function had
nothing to do with why I bought it but it saved a
major engine overhaul, because we know for sure that
the engine was not hurt at all.
A FINAL COMMENT FROM JOHN DEAKIN:
There are large numbers of pilots who prefer to sit
in position, and run up to some fairly high power
setting, then release the brakes and go. I can't really
fault that, but I don't do it, or teach it, unless
runway length is really critical. For the very reasons
that "trapped" you.
You've already done a high-power runup (usually),
so sitting on the end of the runway is just making
noise, burning fuel (trivial, of course), and adding
to the time that the engine is running at power without
cooling flow. Those few seconds can also drive the
tower batty, because they count on people moving out,
so the next one on final doesn't have to go around.
Your usual high-power runup does NOT check anything
between about 2,000 RPM and full takeoff power, so
that's another reason you're probably just wasting
effort on the runway at 2,000. You ain't a'checkin'
anything you ain't checked already. Get it up to full
power smoothly, but with plenty of time left before
"decision speed" to do a quick check of
fuel flow, RPM, and MP.
AND MORE STUFF I AM REMINDED OF
Finally, I was once again reminded that the world
of aviation is just full of knowledable folks like
Braly, Deakin, Eckalbar, Livermore, Shannon, Osborne,
Sobota, Truffer, Bedell, Hegg, and many others, like
Scott in JPI's technical support department, who all
have better things to do with their time than help
someone they have never met or even spoken to, but
and... thanks for the extraordinary Advanced
Pilots' Seminar in Ada, Oklahoma.
and .... like all the instructors who have pulled
engines to idle or cutoff on me over the years: Takes
guts to do that down low, and when you did it right
after rotation and before blueline the student may
even have learned something. I hope I can remember
next time, too..
And I sure hope there ain'tagonna be a next time,
Thanks also to a very chagrined mechanic and his inspector
(twin-rated himself, and understands what I had to
deal with), both of whom shall remain nameless; I'll
be back to their shop, because I know they will be
more careful next time... and, besides, I like - and
still trust - them.
Live and learn, with emphasis on the former! I hope
this helps someone else.. Fred