NSI CAP Prop Failure Analysis Followup
Gary Palmer
As promised, here is some follow-up analysis on my August 17th loss of a
blade on my NSI CAP model 160 cockpit adjustable propellor. For those
who missed the original posting, this prop was installed on a Lancair
235 with a Lycoming O325-C2C engine. The failure happened on the third
flight with a total of 1.5 hours hobbs meter time, and the aircraft was
substantially damaged (a write-off). Fortunately I received only minor
injuries. The failure was a transverse crack of the retaining collar
that captures the Warp Drive blades, starting immediately adjacent to
the flange that transfers centrifugal loads to the hub.
At the time NSI claimed my failure was the first and only one, but based
on subsequent research, I have second hand confirmation of other
failures, at least one of which involves current litigation. The prop
was sent to the Canadian Transportation Safety Board labs for
independent analysis, and at the time I had proposed three possible
causes; being either materials flaw, machining flaw, or design flaw. NSI
had claimed that improper assembly was the cause, and in so doing
admitted to a couple of other failures that they had previously denied.
In the meantime I have had some private email dialog with Paul Messinger
who is an EAA Technical Counsellor, and retired aeronautical engineer.
Neither Paul nor several of his engineer friends could agree with NSI's
contention of improper assembly leading to the failure.
I have some preliminary information from the TSB labs based on
dis-assembly of the hub, and before any metallurgical analysis that I
feel is essential to communicate to the homebuilt movement at large. A
second blade collar assembly in the hub had an identical crack that was
a hairs breadth away from failure. This pretty well rules out a
materials flaw, and there were no obvious machining marks, although
there is a stress relief groove machined in the collar, which is where
the crack started.
I relayed this information to Paul Messinger, along with a nagging
concern I had expressed earlier to both Warp Drive and NSI with an in
flight resonance of the Warp Drive blades which I had always experienced
with my earlier ground adjustable Warp Drive prop during low power
descent at approx 2200 rpm. (BTW, this vibration is commonly observed on
Warp Drive blades, but I don't know if it only occurs with high speed
aircraft). I had recommended that the blades in this application be
stiffened to keep this vibration out of the operational rpm range. The
opinion expressed by both NSI and Warp drive was that I could
operationally pass through this region quickly with the variable pitch
prop and should hence experience no problem. Paul's response to this was
immediate, and forceful (see verbatim copy later in this post).
In essence, Paul reminded me that resonance in any mechanical system is
no trivial matter, and based on his experience could easily lead to
extreme loads and hence a fatigue failure ("The amplification of forces
under resonance can quickly exceed
> any practical materials strength. In the dynamics lab I have seen stress
> amplifications of over 1000 times and failures in the blink of an eye.").
Also of significance is his observation that the prohibited operational
rpm range on a Mooney is due to resonance of an amplitude undetectable
by the pilot, hence the concern over the forces involved when a pilot
can in fact detect a resonance (vibration). I am reminded of the NASA
videos at Oshkosh showing the instantaneous explosion of a wing
encountering flutter (i.e. resonance) in wind tunnel tests.
Reflection on Paul's comments brought other information to the surface
as I recalled articles in the Lancair Newsletter of the exhaustive
vibration testing done by MT and Hartzell on their proven designs to
ensure there were no critical resonance frequencies anywhere in the
operational envelope of their specific propellors on this specific
airframe. Clearly experience has taught the professionals that this
degree of testing is essential in an area as critical as a variable
pitch prop. In retrospect, I really wish I had been more cautious, and
insisted on hard documented proof of proper testing from NSI.
As I considered this issue further, I realized that in my exploration of
the flight performance envelope of the prop, I had encountered a much
stronger resonance on the ground as I increased pitch, and I immediately
backed off, and then checked the blades for any externally visible signs
of damage. Of course there were none, but this may have been the point
at which a crack may have started. The only way to avoid inadvertant
excursion into a range such as this is by setting the mechanical pitch
stops to limit maximum pitch, then conducting a slow methodical
expansion of the envelope, much as you would with a ground adjustable
unit. The prop as received from NSI was set to allow 34 degrees coarse
pitch, and I had previously operated with 19 degrees pitch on my
Lancair. In retrospect I should have set the mechanical coarse pitch
limit to 20 degrees and started a careful evaluation from there. BTW,
the NSI manual does not include any such cautionary advice and I
STRONGLY RECOMMEND current users exercise EXTREME CAUTION in their test
flight program. Note also that there is no user adjustable fine pitch
stop on the NSI hub, which constrains ones options a bit. In my case the
fine pitch stop was arbitrarily set at 18 degrees by NSI (locked in
during manufacture), too close to my normal cruise setting.
The best course of action is to at least conduct a thorough ground
vibration test on your specific engine and prop combination as is
recommended by Paul. There are companies providing this service at
relatively low cost, and I recall an article a recent issue of either
KitPlanes or Sport Aviation on such testing.
A final piece of evidence concerns a ground prop strike I had
experienced a year ago with my ground adjustable Warp Drive prop. Upon
dis-assembly of the hub, I noticed a long crack in the hub, that I
attributed at the time to the ground strike. In retrospect, I have
re-examined the crack, and it now appears that this was a pre-existing
crack, that I had not caught in my annual inspections. It now seems
entirely possible that this may have been a gradually developing fatigue
crack due to the cumulative effect of the resonance I normally
experienced on low power descents as previously noted over the 350 hours
of operation. I would consequently encourage Warp Drive ground
adjustable users to carefully check for such fatigue cracks if they also
experience similar vibrations in their specific engine/prop/airframe
combination.
BTW, before careless readers jump to conclusions, I did not have a
ground strike with the NSI hub; both the Warp Drive blades and the NSI
hub, were brand new units.
Bottom line is to exercise extreme caution with this hub/propellor
combination. I was extremely lucky to survive uninjured. Others have not
been so fortunate. The second hand information I have received indicates
that the case in litigation involved a Kitfox with a Subaru EA81 which
also threw a blade, on take-off this time with the two occupents
suffering serious injuries. Thus while Paul suggests grounding any
Lycoming/Continental applications of this prop, the evidence seems to
imply a wider scope. In the end analysis, we are all responsible for
making our own risk assessment in the homebuilt movement; I hope this
posting and any subsequent debate helps you make the right safe decision
for your specific application.
Fly safely friends, I expect this post will generate more debate than my
earlier one........................Gary
> Gary Palmer wrote:
> >
> > Hi,
> > Just a very brief note from the TSB safety labs initial dis-assembly
> > and analysis based on a discussion earlier today. Apparently one of the
> > other blades also had an identical fracture which was a hairs breadth
> > away from failing. Consequently this is almost certainly not a materials
> > anomaly, and something either specific to the design or operational
> > environment. Apparently NSI machine a stress relief groove in the collar
> > adjacent to the flange, and there were no specific machining flaws
> > visible.
> >
> > The actual metallurgical micro-analysis may be a few weeks yet, as this
> > project is at the bottom of the priority list in the metallurgy lab. At
> > any rate it looks more and more like a design flaw, probably compounded
> > by the unique stresses encountered in this application.
> >
> > One of the areas that has always concerned me has been the resonance
> > that occurs in Warp Drive blades during a low power descent at around
> > 2200 rpm in my case. I had mentioned this to both NSI and Warp Drive in
> > the past, suggesting that a stiffer blade was called for in this
> > application to move this resonance out of the normal operating range.
> > The response was that one could quickly pass through this range by going
> > to a finer pitch as one descended, and that was certainly true, but
> > nonetheless added to pilot workload, and still left an undesireable
> > added load on a hub mechanism which clearly did not have adequate design
> > safety margins.
> >
> > I will keep you guys advised as I learn more. At this time I don't plan
> > to make another RAH post just yet until I have more details.
> >
> > Comments?...............................Gary
>
>
> >From your description I think you had a case of prop resonance. Then you
> got very bad advice from both Warp and NSI. The thing to do was
> immediately ground the prop until the magnitude of the problem could be
> determined. A bad case of prop resonance can be like control surface
> flutter. The amplification of forces under resonance can quickly exceed
> any practical materials strength. In the dynamics lab I have seen stress
> amplifications of over 1000 times and failures in the blink of an eye.
>
> I am not sure I would go so far as to call it a design error, but it
> clearly is an applications error to allow use on direct drive lyc
> engines without testing for resonance. If the pilot can feel the
> resonance it's WAY too big, and unsafe in my opinion (even for a quick
> passing thru mode).
>
> Many aircraft with the lyc series engines have a "no continuos operation
> rpm range". In the Moony its right where you would like to cruse for
> lower noise and fuel economy. There is a mild resonance problem that has
> been determined to be reasonably save as long as there is no steady
> operation within this rpm range (there is no "feel" back to the pilot of
> any problem when flying this acft. There are a lot of factors that go
> into the amount of stress amplification during resonance that are beyond
> the point here. You seemed to have had case of large amplification.
>
> I think you should post a note to RAH and everywhere else that this
> prop/blade combo may be unsafe with direct drive 4 cyl lyc/cont engines.
> Point out the resonance problem and failure points. There are a couple
> of prop balancing vibration measuring systems available in most of the
> USA and elsewhere that can run an inexpensive analysis that I would
> think should done before the next flight for every user and mandatory
> for any direct drive user of the NSI/Warp Drive CAP. As I understand
> these systems they can pickup this type of resonant vibration.
>
> As I had posted to Gary, Several of us here could not figure out how to
> duplicate the assembly error that causes the damage NSI says was the
> cause, on my prop/hub, and we determined that the crack starting point
> was in line with the forward edge of the blade in normal pitch and also
> the weakest point in the hub due to the pitch cam pin installed there.
> This is one of the points of maximum stress in the case of rotational
> resonance caused by the engine torsional pulses.
>
> The proper thing for both NSI and warp Drive to do is IMMEDIATELY GROUND
> ALL direct drive lyc/cont applications until more testing can be done
> and a fix determined. Otherwise I fear someone will be seriously hurt or
> worse the next time a blade is lost. Gary was extremely lucky.
>
> Gary, feel free to quote all or part of my above comments anywhere to
> anyone at anytime.
>
> Paul Messinger
> EAA Technical Counselor
Douglas
Clearly, Gary is not the first to be lied to by NSI. They strung Jim Rohm
along for years. I wouldn't go near this company's products.
Douglas
"You can fool some of the people some of the time but NSI has been fooling
alot, and now it is catching up with them.
Gary Palmer wrote in article <34293A...@nortel.ca>...
Paul Messinger
Douglas wrote:
>
> Clearly, Gary is not the first to be lied to by NSI. They strung Jim Rohm
> along for years. I wouldn't go near this company's products.
> Douglas
It is important to know that the Warp Drive CAP is the same propeller
with different color blade cuffs. I have seen the warp drive shipping
boxes at NSI. That means that anything said about the NSI cap design
also applies to Warp Drive's CAP.
Paul Messinger
Garfield
On 24 Sep 1997 23:15:42 -0700, in rec.aviation.homebuilt
wil...@nospam.com (Wilson) wrote:
>Several years ago I saw their Subaru conversion at Osh. It was a TERRIFIC tent,
>everything chrome or bright red paint, really beautifully presented.
>
>Then they said that they were taking care of tortional resonance with a sprague
>clutch. And that the clutch was good for 10 million pulses.
>
>Lets see, 2 power pulses per rev, 5000 RPM, 600,000 pulses per hour, means it's
>good for 16.6 hours. Not to mention that a sprague clutch only cuts off half
>the sine wave. If the frequency is resonant without it, it's resonant with it,
>it just has half the input with each pulse. Something's still going to fail.
>
>I went back to that overpriced, underpowered unreliable Lycoming. This stuff is
>the reason The Paul feels the way he does.
Hey Wilson.
Uh, that's spelled The Pall, and I JUST MUST say that there are MANY
both reliable, trustable, technically proficient, soundlyNroundly
conservative guys building Soob engines. I can name at LEAST three (Don
Parham of RFI, Reiner Hoffmann of STRATUS, and Dave Johnson of
REDUCTIONS) that do NOT NO WAY fit the description of "pretty face/bad
design/worse ethics". All three of these guys have LONG track records,
excellent reps, and high quality craftsmanship going into their
creations. Just like you wouldn't wanna lump us all in with The Pall,
please don't lump (even by indirect reference) all Soob builders in with
this one vendor.
Not to get on your case, Wilson. Just wanted to hum a few bars of the
AirSoob theme song. Hee hee.
And just for the record, The Pall was opposed to using automotive
engines IN GENERAL, because he wished to be Pope of the topic, not
because there were bad apples in the trade.
>In article <60cdsh$kto$1...@usenet88.supernews.com>, "Douglas" says...
>>
>> Clearly, Gary is not the first to be lied to by NSI. They strung Jim Rohm
>>along for years. I wouldn't go near this company's products.
>>Douglas
>>"You can fool some of the people some of the time but NSI has been fooling
>>alot, and now it is catching up with them.
Also, that was Jim Rahm, not Rohm. Jim is Steve Rahm's uncle, the
designer of the Vision.
Garfield
P.S. Not tryna be snotty, just crispin up the facts a tad.
Paul Messinger
Wilson wrote:
>
> Several years ago I saw their Subaru conversion at Osh. It was a TERRIFIC tent,
> everything chrome or bright red paint, really beautifully presented.
>
> Then they said that they were taking care of tortional resonance with a sprague
> clutch. And that the clutch was good for 10 million pulses.
>
> Lets see, 2 power pulses per rev, 5000 RPM, 600,000 pulses per hour, means it's
> good for 16.6 hours. Not to mention that a sprague clutch only cuts off half
> the sine wave. If the frequency is resonant without it, it's resonant with it,
> it just has half the input with each pulse. Something's still going to fail.
>
> I went back to that overpriced, underpowered unreliable Lycoming. This stuff is
> the reason The Paul feels the way he does.
There are a few of us "engineers" that could not understand the use of
the sprague clutch by NSI. There are wide spread rumors (at least some
confirmed) of a short life to failure of the clutch. The longest
confirmed life seems to be around 100 hours to the glider mode. The
clutch seems to fail open and the prop is not connected to power any
longer!
The CAP problem is unrelated other than the same co designed it :-) I
bit on one based on the fact Warp Drive was selling them under their
name. Gary will be publishing my latest engineering findings shortly.
I still think that auto conversions can be safe and the way to go for
some. The problem is not usually the long block, its the rest of the add
ons needed to make a safe acft engine. Bruce Frank and Reiner are two of
several who have done it. Unfortunately there are many others who have
not done their homework and given "The Paul" material to expound on.
Paul Messinger
Charles K. Scott
In article <60cvie$s...@drn.zippo.com>
wil...@nospam.com (Wilson) writes:
> I went back to that overpriced, underpowered unreliable Lycoming. This stuff is
> the reason The Paul feels the way he does.
Point of information: It was the prop that failed, not the engine.
While Paul has attacked almost every type of owner built non certified
powertrain in the past, he reserved his most vehement wrath for auto
engines.
Another point, this unit is an over the counter item, not something
designed by a homebuilder and cobbled together with hammered tin and
bailing wire. It is sold as a safe and tested aeromotive product to
unsuspecting customers, this is a different situation from someone
using the OTLAR method of designing his own PSRU.
Please note, I am not attempting to defend NSI.
Corky Scott
Tony P.
Garfield:
And there are dozens, maybe hundreds, of Ford engines flying (Bruce
maintains a partial list). Few of us aspire to a 2000 hour TBO. Since
I can disassemble and rebuild for a few hundred dollars, I'll be doing
it at 500 hours or less, no matter what, just to track wear and
everything else. Likely after the first 100 of course also, for
inspection at an early stage, but I plan on a 500 hour schedule.
Jeepers, that's only every 2 or 3 years even if I fly the paint off of
it, and the cost will be far less than my current annuals on the Arrow
(ignoring AD's).
There is just no NEED for a 2000 hour TBO when you aren't paying for
$10,000+ rebuilds.
--
Tony Pucillo
[I speak only for myself unless I say otherwise. One personality is
quite enough, thank you.]
"Castigat ridendo mores" <laughter succeeds where lecturing won't>
Bruce A. Frank
We have one out there with 1400 hrs on it and still going strong.
--
Bruce A. Frank, Editor "Ford 3.8L Engine and V-6 STOL
BAF...@worldnet.att.net Homebuilt Aircraft Newsletter"
*---------------------------**----*
\(-o-)/ AIRCRAFT PROJECTS CO.
\___/ Manufacturing parts & pieces
/ \ for homebuilt aircraft, TIG welding
O O
Roger D. Mellema
Tony P. wrote:
>
> Garfield:
>
> And there are dozens, maybe hundreds, of Ford engines flying (Bruce
> maintains a partial list). Few of us aspire to a 2000 hour TBO. Since
> I can disassemble and rebuild for a few hundred dollars, I'll be doing
> it at 500 hours or less, no matter what, just to track wear and
> everything else. Likely after the first 100 of course also, for
> inspection at an early stage, but I plan on a 500 hour schedule.
> Jeepers, that's only every 2 or 3 years even if I fly the paint off of
> it, and the cost will be far less than my current annuals on the Arrow
> (ignoring AD's).
>
> There is just no NEED for a 2000 hour TBO when you aren't paying for
> $10,000+ rebuilds.
> --
> Tony Pucillo
>
> [I speak only for myself unless I say otherwise. One personality is
> quite enough, thank you.]
>
> "Castigat ridendo mores" <laughter succeeds where lecturing won't>
All,
My Ford V-6 now has 180 hours on it now (I know - NOT enough!). It
has been working beautifully.
This post is to give my input to the Warp / NSI prop story. I have a 3
blade ground adjustable unit with untapered blades. I now have 90 hours
on it and really like it - no comparison at all with the wood prop that
was initially installed.
I have never had a resonance while in flight. Many pitches have been
used from max rpm (4800) static to extreme cruise.
Full power static runup always produces a 'howl' condition. It is loud
and noticable enough that everyone moves 'way back' when they hear it.
Its onset is at 1800 rpm when the prop is set for cruise and 2100 when
set for climb. It always quits when 20 to 30 mph is reached on
takeoff. It sounds like a 'partially stalled' condition but I wouldn't
bet against flutter.
I really want an MT electric........but can you believe $10,500?
rdm
--
Roger D. Mellema Editor: BD-4 Builders/Owners Newsletter
BD-4 with Ford 3.8 L V-6 2/3 P-51 with 351 Ford V-8
rmel...@who.net
http://www.halcyon.com/www2/rmellema/BDindex.html
Charles K. Scott
In article <60fk2r$3...@drn.zippo.com>
Wilson (wil...@nospam.com) writes:
> That's my problem with _all_ of the big bore conversions. Except for the
> Orenda, they aren't being run to a reasonable TBO in a cell, they're being sold
> with a couple of hundred hours on the prototype engine. It costs a LOT of money
> to run an engine in a cell - 10 GPH, $2/gal, 2000 hrs = $40,000 in gas.
>
> Fix that - run them in a cell - and establish a system so that problems get
> reported and people are told how to fix them (see the soob dizzy magnet post a
> few days ago) and I'll be one of the first to put down my money. But, then,
> that pretty much describes the certification system.
>
> I know some of the soob stuff is getting up into serious hours now (I read about
> a gyro with 1300 hours on it) and so it's starting to look good. But I still
> wonder at claims you can reliably get 2000 hours out of a 2 liter engine at 200
> HP.
>
> For NSI - I'm glad I'm not the only one that was bugged at the sprague clutch.
>
> Wilson (I'm not an engineer, I don't even play one on Sinefeld) Blackey
Wilson, I'm with you on your sentiments but I'm not sure all the facts
are straight here. First, certification doesn't require that an engine
run 2,000 hours at peak power output or any output. I didn't keep the
figures anywhere but several of our certified engine experts will tell
you that total dyno run time for the certification process is only
around 150 to 200 hours. And that's at various power settings, not
just full throttle. And the engines get to be repaired if necessary
during this process, or so I remember.
You don't need to wonder anymore about "claims you can reliably get
2,000 hours out of a 2 liter engine at 200 HP" because no one is
claiming that. There aren't any auto/aircraft engines that do this as
far as I know. The 2 liter engines make about 90 to 110 hp. The 2.3
liter Subaru engines have been shown at 120 to a claimed 197
(Turbocharged). The Ford 3.8 L V-6's make closer to 200 hp. Several
350 cid Chevy V-8's make around 250 to 300 hp with one guy claiming he
reliably gets one horsepower per cubic inch displacement for his 350
and 400 cid engines (he powers scaled WWII fighter replica's with
them).
But speaking of high output, Jim Rahm developed (well his engine guy
developed it but it was Jim's money) a 375 cid V-8 that puts out 420
hp. This is the most highly stressed engine I'm aware of in the
Kitplane industry today. Jim actually did put more than 500 hours on
the engine in the dyno room which is WAY in access of what is required
for certification. His engine, of course, has internals that are
specifically designed to withstand the pressures associated with that
kind of output. He also discovered, along the way, that the block
where the PSRU bolts to could possibly be a weak point in heavy
turbulence so he had the foundary add metal to this area. So any
homebuilder who wants this engine gets the benefit of this meticulous
engineering.
Check out the rpm limits for those big block engines you were referring
to, they aren't excessively high (usually less than 4500). Those
engines are producing less power than they could if power were all the
developers wanted from them. How's that? Take the Legend for
instance. That engine dynoed out at 575 hp but the builder speculated
that by the time he got the engine installed in the airframe with it's
tiny stub exhausts, he was likely making less power than that, he
estimated around 520. They want to go racing at Reno with that engine
and when they do, they will be extracting closer to 1,000 horsepower.
This engine (the Donovan), when run in AA Fueler competition puts out
over 3,000 hp! Of course it's life is measured in mere seconds at that
output. :-)
Corky Scott