|
SYNOPSIS
The float equipped Challenger II was conducting a take-off from
a small lake into a 15 mph wind. Immediately after lift-off at
full power (6300 rpm) the tachometer rpm began fluctuations of
about two hundred rpm above normal take-off rpm. The rpm increased
momentarily and would then return to slightly above normal
take-off values. The fluctuations lasted about 2 seconds followed
by a sudden and complete loss of thrust and an engine over-speed
in excess of 7000 rpm.
The aircraft was accelerating through about 42 mph in a shallow
climb attitude at a height of about 30 feet and in a high drag
(take-off flap) configuration when the thrust failed. The aircraft
immediately entered the stall regime and began to mush towards the
lake surface at a steep angle that resembled a helicopter
auto-rotation. The pilot was able to control a slight nose-up
wings level mush attitude to impact just short of the rugged
shoreline.
Total airtime was less than10 seconds. Shoreline witnesses
stated that the aircraft appeared to descend very steeply then it
disappeared in the splash. One witness thought the aircraft had
crashed and ran to contact rescue authorities. The aircraft
remained afloat and was towed back to its base.
FACTUAL INFORMATION
- The aircraft, manufactured by Quad City Ultralight Aircraft
Corp. (QCU), was powered by a Rotax 503, two-cylinder, Dual Carb,
Dual Ignition, two stroke engine rated at 52 horsepower at 6600
rpm.
- The aircraft was fitted with the 60 inch propeller and HEGAR
re-drive using a Gates PowerGrip
HTD® (High Torque Drive) 960-8M-50 belt.
-The HTD reduction drive belt was found completely devoid of
all of its *cogs.
-The reduction drive belt had been in service 2 years with a
total of 83 hours logged on it.
-QCU recommends belt replacement after 100 hours or one year in
service.
-The belt tension and condition was inspected by the
owner-pilot prior to the flight and appeared to be "tight" and in
pristine condition.
- The aircraft underwent a 50-hour inspection by an AME 10
operating hours prior to the event. The inspection included the
condition, tension and integrity of the belt.
-The aircraft was inspected after the event and no structural
damage was found.
-The PuddleJumper floats were examined after the event and no
damage was found.
-The belt tracking was found to be accurate.
-The drive sprocket did not exhibit evidence of excessive of
wear.
-Except during flight, the failed belt had not been unduly
exposed to environmental elements. The aircraft was either
hangared or an engine cover was used.
-An expert examined the failed belt (less cogs) and opined that
it did not exhibit evidence of excessive wear, delamination or
deterioration.
* Note: The PowerGrip System is described by the manufacturer
as providing positive slip-proof engagement by the belt teeth
meshing smoothly with the sprocket grooves. To better understand
this report from a Challenger user/maintainer’s perspective: The
belt teeth are termed "cogs" in accordance with common Challenger
terminology. The sprocket has protrusions which resemble teeth
that extend between the sprocket grooves.
These protrusions are termed "teeth" in this report. The
popular and incorrect belief is that these sprocket teeth drive
the belt. The positive engagement of the semi-circular belt cogs
with semi-circular grooves in the drive sprocket is what drives
the belt. It is therefore essential that the belt be tensioned to
properly seat the belt cogs firmly into the grooves. The sprocket
"teeth" would be better described as spacers between the sprocket
grooves.
RESEARCH
It was found that a substantial amount of information is
available on the care and maintenance of the
 |
| An indication of a lose belt |
Rotax 503 aircraft engine. Specific information on the Challenger
reduction drive belt inspection, tolerances, system components,
serviceability, tensioning procedures, tension values and belt
care is sparse to non-existent.
When a drive belt fails the loss of thrust is instantaneous and
virtually without warning. Drive belt failure, unlike a number of
engine failures, provides no obvious warning of an impending
failure except through detection during knowledgeable pre-flight
belt drive inspection. Information on the care and handling of
belts prior to installation was also found to be sparse to
non-existent.
Warnings not to mishandle (crimp) the belt are placed on one
particular equivalent brand of belt in 5 languages. There was no
such warning o n the failed belt. The normal fashion for a
synchronous belt to fail is by loss of cogs. Any other failure is
abnormal. The shearing of belt cogs can be caused by excessive
shock loads applied to the belt. Belt performance is generally
unaffected in ambient temperature environments of -30F to 185F.
Belt cracking can occur under extreme low temperature start-up.
If a belt is removed and reinstalled to run in the opposite
direction accelerated wear will result from a mismatch of the
established wear pattern. Synchronous belts are not date stamped
and have a shelf life of 8 years without a reduction in
performance.
One manufacturer’s policy is to not ship drive belts to
suppliers more than 5 years after belt manufacture. A mandatory or
suggested belt in-service life imposed by manufacturers was not
found. Instructions supplied with the re-drive state the belt
should last 200 hours or more but recommends replacement after 100
hours or 2 years in service.
During industrial application GT belts reportedly achieved a
service life of several thousand hours prior to routine
replacement. GT2 belts have not been on the market long enough to
establish a service life profile. Synchronous belts continue to
evolve with material technology and construction design. The first
generation (1988) Gates PowerGrip HTD® belt subsequently evolved
into an improved PowerGrip GT® version.
The GT belt has a modified cog profile resulting in complete
cog flank contact that eliminated stress
 |
| An indication of a loose belt. |
concentrations and cog deformation under load. This greatly
increased belt life and resistance to ratcheting. The Gates
PowerGrip GT2® version was introduced in February 2000. The GT2
belt features a longer and more robust life and with its
redesigned deep cog profile is capable of transmitting up to 200%
more power than previous PowerGrip GT and PowerGrip HTD belts.
When tested against competitor belts at sprocket speeds up to
9000 rpm the GT2 belt outlasted the competitor belts more than two
to one. As patented belt designs expire numerous generic versions
enter the marketplace. There are presently at least 12 different
brands of synchronous belts on the market. A recently introduced
belt features a small lateral groove along the face of the cog.
The purpose of the groove is to allow the release of trapped
air and to reduce noise generation. QCU, the manufacturer of the
Challenger Ultralight was requested to provide the research data
or to provide the foundation for the imposition of their 100-hour,
one-year in-use limitation placed on the re-drive belt.
The stated belt limitations imposed b y QCU in the Challenger
50 Hour Inspection Report was found to be arbitrary and without
foundation. The QCU imposed belt limitations remain unchanged
despite technological advances in belt design and durability.
The 880 8M 85 synchronous belt used on the Zenair 601 is about
an inch wider and 6 inches shorter than the 960 08M 50 belt used
on the Challenger. Belt life is projected to be a minimum of 500
hours of operation utilizing a Subaru 100 hp 4-stroke engine at
4500rpm. While utilizing pre-G2 belt technology and design, Zenair
belt usage to 850 hours is commonplace in applications up to 200
horsepower.
There is no specific calendar limitation imposed on the belt.
One reported premature belt failure occurred with indications that
the belt was not tensioned or tracked correctly. Information on
belt tensioning procedures is apparently not provided with the
re-drive system.
"V" belts depend on friction to transmit power. V-belt
applications such as that used on the Challenger cooling fan
sustain normal wear during service and become "loose". The common
perception that fan belts stretch with wear is only partially
correct. The belt shoulder is compressed, burnished and worn
during use causing the belt to become thinner and hence "looser"
on the pulley.
If the Challenger pulley is not compressed through shim removal
the belt will eventually slip in the pulley while under load. A
Challenger fan belt with 150 hours logged in service was examined
and found to be virtually the same length as the new replacement
however the used belt was 2 mm narrower at the shoulder.
Synchronous belts function under a different principle than
V-belts. The synchronous belt itself may wear and stretch only
minimally. It is absolutely essential that the prescribed
pre-tension value be maintained as well as the system drive
sprocket integrity and sprocket alignment (tracking).
The pitch of the belt is the distance between the center of
each cog measured on the pitch line (circular pitch). The
Challenger belts have an 8mm circular pitch. If the belt stretches
with age and use there is a corresponding belt pitch change. The
pitch change allows a mismatch to be generated between the face of
the sprocket groove and the driven side of the belt cogs. The
mismatch creates increased friction between the two surfaces thus
accelerating cog wear and sprocket wear.
Evidence of this wear manifests itself at the driven shoulder
of the belt cogs and at the shoulder of the drive sprocket groove.
In extreme cases, evidence of belt facing material may be
exhibited on the drive shoulder of the sprocket grooves.
Over-tensioning the belt can impose higher than design bearing and
shaft loads and leads to reduced belt and bearing life. When the
belt is severely under-tensioned an inherent self-tensioning
characteristic is generated by belt circumferential force against
the sprocket grooves.
Since the groove face is inclined, the belt slides up the
groove while under load which in turn increases overall belt
tension. The increased tension leads to increased stress on the
cogs, accelerated cog wear and reduced belt l ife. The sprocket
groove and cog combination will take the path of least resistance
and the cogs will ride high on the sprocket grooves and teeth or
even ratchet by them while under high torque load such as during
take-off.
In due course the drive sprocket will either damage or shave
off belt cogs in just a few revolutions of the belt. Ideal belt
tension is the lowest amount that properly seats the belt in the
sprocket grooves during maximum loading at all points around the
drive.
One method for measuring the ideal tension employs a sonic
measuring device the size of a cell phone. This device operates on
the transverse vibration of strings theory which holds that a
belt, when strummed like a guitar string, vibrates at a specific
natural frequency based on its tension, mass and span length.
During research into this event the formula to calculate belt
drive pre-tension values was entered into a computer using the
manufacturer’s engineering software. The engineering formula was
entered into the computer using the application data of a Gates
PowerGrip 960-8M-GT-50 belt driving an industrial ventilator
propeller powered by a Challenger engine using re-drive
parameters. (Gates does not endorse application of its product on
aircraft).
Using a 30 tooth, 3.01 in. DriveR sprocket and a stock 80 too
th, 8.02 in. DriveN sprocket (the Challenger uses a 78 groove
DriveN) with a 10 inch center distance the computed slack side
belt deflection was 0.150 inches (about 5/32in.) with an applied
belt deflection force of between 27.52 and 29.39 pounds for a new
belt and 23.78 to 25.65 pounds for a "used" belt.
The computed sinusoidal wave frequency, or Sonic Tension Value,
for a properly tensioned used (Challenger) GT belt is between 156
and 163 Hertz. Also determined was a handy rule-of-thumb
tensioning method which could be used by Challenger owners: The
belt, at mid-span, should deflect 1/64ths of an inch for each inch
of belt span at about 28 pounds force for a new belt and 24 pounds
for a used belt. The Challenger belt span is about 10.25 inches
which amounts to about 3/16ths of an inch cold belt deflection
measured on the slack (exhaust side) span.
Synchronous belts cannot be run at unlimited speeds and power.
The limitation of a synchronous belt drive system is the sprocket
"Rim Speed" which is almost identical to the belt speed. This
speed limitation is 6500 feet per minute (fpm) which is much
greater than speeds the Challenger 7075-T6 aluminum drive sprocket
attains at full power.
Above 6500 fpm speeds standard cast iron sprockets can develop
cracks and fracture. The actual "belt" limitation of a drive
system relates to the tension imposed on the belt by the drive
horsepower. The PowerGrip GT2 belt carries a rating of over 120hp
at a 6500fpm Rim Speed. A random Challenger take-off engine RPM of
6300 was selected and entered into the computer using a 30 groove
drive sprocket and a custom 78 groove DriveN sprocket.
The following was determined:
Rim Speed 4927 fpm with a prop Speed of 2423 rpm and a 633lb.
belt pull (Belt Shaft Load). At a take-off RPM of 6500 the Rim
Speed is 5084 fpm and a prop speed of 2500 rpm is attained. A new
GT belt was installed on the incident aircraft and pre-tensioned
to 3/16in.
The belt felt much tighter than the owner-pilot was accustomed
to during his previous pre-flight belt inspections. Because the
belt felt "too tight" for bearing li fe, the rule-of-thumb
deflection was increased an additional 1/16in. to 1/4in. The
maximum engine rpm attained after installation of the new belt was
about 200 rpm more than the average maximum attained prior to the
belt failure.
ANALYSIS
The tachometer rpm fluctuations observed by the pilot during
lift-off were caused by ratcheting and/or the initial failure of
the first belt cog or cogs. Each revolution of the belt
transported the cog-damaged sector toward and past the lower
engine sprocket drive.
 The
engine momentarily increased rpm due to lack of shaft resistance
until cogs were again in contact with the sprocket drive
grooves/teeth. The tachometer rpm indication momentarily increased
and decreased each time the cog damaged or cog-free sector of the
belt was transported past the sprocket drive. With each revolution
of the belt, the cog-damaged sector was enlarged by the
over-speeding sprocket drive that damaged or shaved off more cogs.
Finally the damaged sector was large enough to allow the engine
sprocket drive speed to exceed 7000 rpm.
All the remaining cogs were instantly shaved off when they were
transported to the high-speed sprocket drive; akin to the action
of a wood router. It was at this time that all thrust was lost.
From start to finish the process took less than 3 seconds. The
increase in maximum engine rpm attained following installation of
the new belt was evidence of the previous belt’s mismatch between
the cogs and the drive sprocket grooves due to belt pitch change.
The mismatch created additional friction that absorbed engine
torque. The float-equipped aircraft was at low altitude, in a
shallow nose-up attitude, in high drag configuration just above
the stall speed when thrust was lost. (The dead man’s curve). The
aircraft rapidly decelerated into the stall regime. Insufficient
height existed to lower the nose to effect a classic stall
recovery.
Fortunately, the Challenger’s stall characteristics allowed the
aircraft to mush with substantial elevator and flaperon authority.
As a consequence, the pilot had adequate lateral and pitch control
throughout the steep descent until splashdown.
Had the Challenger possessed classic aircraft stall
characteristics the nose would likely have dropped during the
stall and in all probability the tip or tips of the floats would
have dug in at impact causing the aircraft to flip inverted.
FINDINGS
1) The engine driven reduction belt failed under high torque at
a critical moment during take-off.
2) The lower than optimum engine rpm prior to the occurrence
was evidence of excessive belt friction.
3) The belt failure was caused by inadequate tensioning of the
belt.
4) The failed PowerGrip HTD belt was an earlier, less robust
version of current PowerGrip GT belts.
5) Synchronous belt manufacturers’ prescribed belt-tensioning
values and procedures are not readily available to aircraft
owners, pilots and maintainers.
6) The pilot was unaware that minor tachometer rpm fluctuations
to above normal values at take-off power indicated imminent belt
failure and total loss of thrust.
7) The one-year, one hundred hour service life imposed by QCU
on the Challenger synchronous belt is arbitrary and without
foundation.
8) For liability considerations the Gates PowerGrip belt
application on the Challenger (and other) aircraft is not endorsed
by Gates Rubber Company.
9) Direction of rotation should be identified on the belt and
that direction should be maintained for the life of the belt.
10) The Challenger aircraft’s docile stall characteristics
averted a potentially serious accident.
Flight Safety Suggestions for Challenger Owners, Pilots and
Maintainers.
-Discard any old drive belt or those marked "PowerGrip HTD"
regardless of the belt’s service life.
-Preflight, examine the re-drive belt tension to ensure it does
not exceed ¼" deflection. (Apply tension pressure with your thumb
till it hurts, you then have about 25 pounds pressure).
- When measuring the belt deflection pull the propeller
backward against engine compression to slacken the belt’s slack
side.
-Preflight, rotate the propeller and examine all 120 belt
"cogs" for excessive wear or damage.
-Preflight, tap or pluck the belt, if it doesn’t vibrate its
probably too loose.
-Periodically examine the belt’s neoprene backing for hairline
lateral cracking. Cracking is detected most easily where the belt
backing is stretched around the lower drive sprocket.
Alternatively, examine it around the upper sprocket.
-If any belt damage, excessive wear or cracking is found
replace the belt immediately.
-Always use smooth throttle handling to reduce the risk of
shock load failure.
-If practical, protect the belt from moisture, chemical
contaminants and direct sunlight. With proper care and maintenance
the GT belt should provide at least 5 years and 500 hours of
Challenger service. The GT2 belt should provide at least 5 years
and 700 hours of service. Four JPG photographs are attached
showing, for comparison, a new belt and the failed belt,
belt/sprocket specifications and a suggested belt tensioning
procedure.
Picture summmary
A. New and failed belt.
B. Belt pitch length is measured along the belt pitch line.
C. Preparation to measure belt deflection.
D. Prevent belt twist and apply pressure.
(Note about ¼ inch deflection. )
Editors Note.
The author spent 23 years as a professional Aviation Safety
Investigator (Canadian Aviation Safety Board) and Aviation Safety
Consultant (Accident Investigation & Research Inc.) His
professional responsibility during this period was to identify
aviation safety deficiencies. Determining "WHY" the event happened
or "WHY" a component or system failed is part of that process.
The thrust of this Challenger flight safety report was to
determine "WHY" the belt failed and to educate Challenger users
about the synchronous belt application on their aircraft.
Dissemination of this report to Challenger users will hopefully
prevent a recurrence of an event that had potentially disastrous
consequences. A draft copy of this report was distributed to the
owners of National Ultralight Inc., the exclusive Canadian
Distributor of the Challenger. They declined to comment.
|
