PDA

View Full Version : Nicopress fitting failure report



Chad Jensen
12-27-2011, 08:57 AM
This is a report that was submitted to us by an NTSB investigator who needs to remain anonymous as an impartial investigator. Pictures and words are not mine so I can't offer any answers to questions. This is simply a passing of information to anyone using nicopress fittings to use the proper tools.



Thankfully, this was a known-circumstance, no-injury accident, and we were hoping to share the lessons learned from it with an interested audience. Your members are likely to be involved in the building of flight control systems, or the re-rigging of those systems in antique aircraft, and that’s why we sought you out.

The NTSB Accident Number is ERA11CA375 (http://dms.ntsb.gov/aviation/AccidentReports/qskvhq55mztre3q2v52mvt451/P12272011120000.pdf). You can go to our website and read the Factual Report as well as the Board’s finding of probable cause. The public docket which contains the documents and photographs used to support the investigation can also be found there. I have attached a number of those items for your convenience.

Because my position requires that I be an impartial investigator, I am reluctant to self-publish on a website or discussion forum. It am not, however, precluded from sharing the facts and circumstances with you and your members. From there, you can draw your own conclusions, or start a discussion or debate as the situation dictates.133113321333

Mike Switzer
12-27-2011, 10:32 AM
Did the fitting fail or did the cable break? I can't tell from the pictures. We use similar crimp sleeves for fencing & vineyard wire splices, and if you use the wrong tool it will not crimp properly & the wire will pull out under load.

Chad Jensen
12-27-2011, 10:35 AM
Fitting failed, cable slid through. That's all I know though...

Bill Berson
12-27-2011, 01:22 PM
Ultralights often use two copper sleeves instead of one.

But in his ultralight book, Lack Lambie said using two sleeves will weaken the cable slightly.
I never understood this. Anybody have the facts on using two sleeves?

Frank Giger
12-27-2011, 06:23 PM
I can't really tell, but it looks like he didn't crimp them properly, and didn't use a go/no go gauge to ensure they were tight enough.

Neil
12-27-2011, 11:33 PM
Don't know in this particular case but I see this type of failure from time to time when a builder tries to make up cables with the "Econo Swagers" that get all their clamp force from a couple of bolts. There is no good substitute for the proper lever type swaging tool.

Dana
12-28-2011, 06:26 AM
I don't think there is a problem with the clamp bolt swagers if used properly.

Some years ago Quicksilver had a problem with a lot of improperly swaged wing cables at the factory. Seems 1/8" cables had been swaged with a 5/32" tool; the swage looked OK but wouldn't pass a go/no go check.

dewi8095
12-28-2011, 07:09 AM
I can't really tell, but it looks like he didn't crimp them properly, and didn't use a go/no go gauge to ensure they were tight enough.

What is a go/no go guage and how does it work?

Don

Hangar10
12-28-2011, 10:02 AM
I've built a few of these cables, but by no means am I an expert on the subject. I'd just like to share a few things that I've discovered.

For the first cables I constructed I borrowed a pair of Nicopress crimpers and cable cutters from a fellow EAA member. Here you can see the materials and tools I used to construct some turnbuckle ends.
http://www.kitlog.com/photos/project_687/full/FP02012010A0004G.jpg

If used properly (as illustrated in Tony Bengelis' books) they do a nice job.
http://www.kitlog.com/photos/project_687/full/FP02012010A0004H.jpg

The Nicopress tool worked great, but I wasn't the only one in line to borrow them, so I looked around for a pair of my own. There are several different types of nicopress tools... the type that are capable of 3/32", 1/8", 5/32", etc. are not cheap. Somewhere in the neighborhood of $300 was not unusual, even for a used set. So I chose to try the $28 type that Home Depot sells. These are typically used for making dog runs and leads, or for fencing, but I know of others that have used these, and if properly adjusted and monitored I don't see how the end result is any different than the more expensive tools. The authentic Nicopress tool was longer (more leverage when crimping) and the finish and operation were a bit smoother, but otherwise either tool is capable of putting the squeeze on copper sleeves.
http://forums.matronics.com/files/p1100267_151.jpg

Prior to making my test crimps, I referred to some information I found on the Pietenpol forums.

Size - Range (Chris Bopka posted as found in EAA Aircraft Volume One File Number 6 on Aircraft Homebuilding Tips, page 34)

1/16" sleeve - .190-.195" go dimension
3/32" sleeve - .255-.265" go dimension
1/8" sleeve - .343-.353" go dimension
5/32" sleeve - .380-.390" go dimension

Using the Home Depot tool straight off the shelf, my first crimp on a 1/8" cable/sleeve combo was .318" (way too tight). After a minor adjustment, my second crimp was .355" (slightly too big). Another adjustment and I was able to squeeze a .345" crimp. I grabbed another sleeve and proceeded to squeeze 3 more crimps right around .345" (+/- .0005") as measured with my calipers. In other words, plenty good and on par with the more expensive tool.
http://forums.matronics.com/files/p1100265_676.jpg

Hope this helps!

rwanttaja
12-28-2011, 11:30 AM
What is a go/no go guage and how does it work?

Don
A go/no-go gauge is a piece of metal with a slot of a particular width cut in it. After you get done crimping your nicopress sleeve, you attempt to slide it into the gauge. If it won't go in, the sleeve has not be compressed sufficiently. You could use a micrometer the same way. Dimensions are at the bottom of this diagram:
1335
One problem with some bolt-type swages is the *width* of the swaged area. One unit I owned didn't swage the entire length of a 3/32 sleeve, you had to do multiple impressions like 1/8" and higher.

The above diagram is from my Shopsheet page:

http://www.wanttaja.com/shopsheets/index.html

Ron Wanttaja

dewi8095
12-28-2011, 03:12 PM
Thanks, Ron.

Don

Racegunz
12-28-2011, 07:13 PM
After reading this thread I just had to go and try to destroy one of the first cables I made with my cable swager, (it was loose according to the guage) I clamped the cable end to my welding table and put a 10" long drift punch in the eyelet and levered it under the tubing on the extremely heavy table. I literally stood on the drift punch and bounced on it until it bent some, this is way more tension than you could pull on a stick or push on a rudder pedal. It held, didn't even look like it was stressed. When I get back to work I'll do a more scientific test with our hydraulic ram press and let you know when the same cable fails (if it does). Long story, but I have new confidence in my cables. I can only imagine how bad this crimp must have been to be pulled out like it was.

bwilson4web
12-30-2011, 11:54 AM
I have a 1940 edition of "Aircraft Maintenance" by Daniel J. Brimm, Jr. and H. Edward Boggess. They don't mention nicropress fittings but instead have a section titled "HOW TO MAKE A WRAPPED TERMINAL" (pp. 377.) The short version is they put the thimble in a 'splicing clamp' and run the cut end 21 times the cable diameter. This is wrapped in soft steel wire and soldered with three, 7D wrappings, and an 'inspection' gap. Then everything is soldered. Soldering cable, especially if it has any plastic coating or core, is not going to be practical. Also, the labor per termination is going to be higher than pressing a slieve. But is there there a strength problem with the old style, wrap-and-solder? This is followed by sections "HOW TO MAKE A FIVE-TUCK NAVY SPLICE" and "HOW TO MAKE A ROEBLING ROLL SPLICE" . . . Thanks, Bob Wilson

Dana
12-30-2011, 06:04 PM
My 1941 Taylorcraft still had the old woven eye splices on some of the control cables. Properly done they're very strong, but they're also very labor intensive and require a high degree of skill from the mechanic. It was state of the art back then... Nicopress didn't exist yet.

P.S. I have the same book... and the companion "Aircraft Engine Maintenance" volume.

Tom Downey
12-30-2011, 06:39 PM
Inspect cables old enough to have woven splices, very well they are subject to corrosion and wear, AC 43-13 has a very good inspection criteria for this.

Dana
04-27-2012, 05:08 PM
I know this is an old thread but there is a discrepancy between Hangar10's checking figures and Ron W.'s ShopSheet figures... which is correct? I need to check the swages I did on new 3/32" rudder cables for my Kolb.

Hangar10 says:
1/16" sleeve - .190-.195" go dimension
3/32" sleeve - .255-.265" go dimension
1/8" sleeve - .343-.353" go dimension
5/32" sleeve - .380-.390" go dimension

but Ron W says:

1/16" .1908 max
3/32" .2674 max
1/8" .3532 max
5/32" .3965 max

RV8505
04-27-2012, 07:19 PM
As A Tech Counsler and Airline Mechanic I discourage the use of home made cables! It is bad genral practice to use uncalibrated tooling to crimp your fitting and then not load test it on a calibrated load testor. This is just an example of why I discourage it.

For example:

Moreover, completed cable assemblies in certified aircraft are required to be tested prior to sale or installation. That test load or proof requires that the cable be pulled from the attachment pin, eye, or other fitting to at least 60% of ultimate cable load, or at least 1200# (WITHOUT SLIPPAGE OF THE FITTING!). ALL FAA-PMA cables are produced to this standard, which is difficult if not impossible to meet in a field environment where a simple nicopress tool is used to fabricate the cables.
(Classic Aero LLC and Univair both produce cables to this OEM standard)

The point I want to reinforce, is that you can make cheap cables for about $25-28 each, And, the untested failure rate is relatively high. In the alternative one can buy certified cables for twice to three times the home-made cost which have been pull tested for slippage to more than 1200#- and you will be assured of proper, long-term, safe operations without cable failure.
Is the loss of control in your aircraft worth the additional $300-400? For most of us it is.

Racegunz
04-27-2012, 07:50 PM
As A Tech Counsler and Airline Mechanic I discourage the use of home made cables! It is bad genral practice to use uncalibrated tooling to crimp your fitting and then not load tested on a calibrated load testor. This is just an example of why I discourage it.


, And, the untested failure rate is relatively high. .

Really? relatively high? until this thread I had not heard of a homemade cable failing and this threads example was just a ridiculously bad job. What is relatively high? do you have any NTSB reports verifying your stance?

RV8505
04-27-2012, 08:24 PM
Really! All it takes is one to ruin your day! By not using the proper equipment your rolling the dice and gambeling with your life and the life of your passengers. Here is one example, google the internet you won't have to go far.

On August 21, 2011, at 1514 central daylight time, a Hendryx/Wiley Super Rebel experimental amateur-built single-engine airplane, N540AF, sustained substantial damage during a forced landing following an in-flight failure of the aileron cable near La Porte City, Iowa. The private pilot, who was the sole occupant and registered owner, sustained minor injuries. Visual meteorological conditions prevailed, and no flight plan was filed for the 14 Code of Federal Regulations Part 91 personal flight. The flight departed the Nichols Airport, La Porte City, Iowa, approximately 1425.

According to the pilot, the accident flight was the third flight into the Phase 1 flight test program. While maneuvering the airplane at 2,500 feet above ground level through a series of turns, the pilot noticed the control stick force increase, and then he heard a bang. Following the bang sound, the pilot had no aileron control authority. The pilot maintained some control via rudder inputs and executed a forced landing to a field. During the forced landing, the airplane nosed over and came to rest inverted.

A postaccident examination of the airplane revealed the firewall and fuselage were buckled. One of the aileron cable swages failed, which resulted in the loss of aileron authority. According to a Federal Aviation Administration inspector who examined the airplane, the cable swage failed due to an improper crimp. The cable was manufactured by the builder during the airplane build process.

From the NTSB Docket
Steve Johnson

Frank Giger
04-28-2012, 12:52 AM
That's why EVERY swag has to be checked with a go/no go gauge, which takes just a second to do.

For my aircraft, buying pre-crimped cables is changing one problem for another - the drag/anti-drag cables in the wings, for example, are tensioned in situ to ensure both tightness and to keep the wing square. Using pre-measured cables would mean introducing a turnbuckle, which wouldn't be seen again once the wing is covered.

The fewer parts to break or loosen the fewer points of failure.

I got a friend of mine with the right tools to take one of my oopsy cables (nothing to do with the swags) and take it to failure. The cable failed before the swag, breaking well away from it, which made me pretty happy.

Racegunz
04-28-2012, 06:56 AM
"Here is one example, google the internet you won't have to go far. "
So you googled it just now as well, so you saw the Fly-baby failure as well where it was "apparently" a tension differential that caused failure and not a bad nicopress job. I'm not saying anything other than where is the stacks of proof that properly made "at home" cables have a relatively high failure rate. I'm not buying it, I have a guage and use the parts from an aircraft supplier. I have tested mine and can say the rudder/elevator horn will fail before the cable or nicopress do. Perhaps a relatively high means there are a few reports of cable failure in experimentals and only a couple in GA? I might buy that. Just not buying into the bold lettered DON'T MAKE CABLES AT HOME THEY WILL FAIL AND KILL YOU, THINK OF THE CHILDREN!!!!!!! know what I mean.

RV8505
04-28-2012, 08:43 AM
"Here is one example, google the internet you won't have to go far. "
So you googled it just now as well, so you saw the Fly-baby failure as well where it was "apparently" a tension differential that caused failure and not a bad nicopress job. I'm not saying anything other than where is the stacks of proof that properly made "at home" cables have a relatively high failure rate. I'm not buying it, I have a guage and use the parts from an aircraft supplier. I have tested mine and can say the rudder/elevator horn will fail before the cable or nicopress do. Perhaps a relatively high means there are a few reports of cable failure in experimentals and only a couple in GA? I might buy that. Just not buying into the bold lettered DON'T MAKE CABLES AT HOME THEY WILL FAIL AND KILL YOU, THINK OF THE CHILDREN!!!!!!! know what I mean.


No I don't! Well, All I have to say is that speaking from years of experience making cables I know what I know. I didn't say homemade cables would fail. I merely said that using the proper tools would remove the probability of a a failure. There is nothing wrong with making homemade cables if you have acess to the proper calibrated tools and testers. I reallly don't have the time to search for stacks of proof so let's let's just say some rely on time proven quality practices and some rely on luck.

rwanttaja
04-28-2012, 09:39 AM
No I don't! Well, All I have to say is that speaking from years of experience making cables I know what I know. I didn't say homemade cables would fail. I merely said that using the proper tools would remove the probability of a a failure. There is nothing wrong with making homemade cables if you have acess to the proper calibrated tools and testers. I reallly don't have the time to search for stacks of proof...
I do.

I've got a 13-year database of homebuilt accidents (1998 through 2010, inclusive). Over that 13 years, there were two accidents (LAX99LA074 and SEA00LA128) attributed to improperly-made nicopress fittings. Two accidents out of over 2,700, in a time period where 13,000 new homebuilts were added to the rolls.

You're absolutely right this this is a critical operation, and users should use proper tools in a proper fashion. I believe the evidence indicates that the skills required are within the capabilities of an average homebuilder.

Fabrication of cables using nicopress fittings is an approved method for both homebuilts and certified aircraft. AC43-13 does recommend (repeat, recommend) load testing. A good idea, but I think the evidence shows most builders can make safe-enough fittings reliably.

Ron Wanttaja

RV8505
04-28-2012, 10:46 AM
It's all in how you peel the onion Ron. I have found several accidents besides the ones you quoted using different search terms. Check the NTSB site http://www.ntsb.gov/aviationquery/index.aspx and try cable failure, Nicopress, Nicropress, Cable End, Eye End, and any other applicable search term you can think of. A cable failure is a failure regardless of homebuilt or production as I know a lot of A&P mechanics make cables improperly as well. What is an acceptable rate of failure? Even a 2 in 13,000 chance is still unacceptable. One aircraft in particular (Zenair) had about 950 hours before the nicopress failed. I don't think we could even imagine the sheer terror of a elevator cable failure. At a minimum cables should be load tested. Go down to a airport with your cables flip the Mechanic a few bucks and get them tested.

Racegunz
04-28-2012, 12:40 PM
Thanks for the voice of reason Ron. Proper tools and attention to detail, that is what the message should be, not plop down the money for it. I'd go certified if I was made of money,...... maybe.
I'd much rather learn the proper way to build/make things, that is what made this country great, not relying on someone else's work. RV8505 I meant no disrespect, but you did bold print your post, and I thank you for putting the NTSB link there, I tried to use their database without success the other night. Let's face it flying is full of risks but the prevalent accident factors are not mechanical failures, it's pilot error. So a few more facts and a bit less drama and fear, that would be helpful.
I got fired up and went and tested a poorly swaged cable, it does guage correct though.


http://www.youtube.com/watch?v=EkuPwZFUL4U

RV8505
04-29-2012, 12:08 AM
"I'd much rather learn the proper way to build/make things, that is what made this country great, not relying on someone else's work." "So a few more facts and a bit less drama and fear, that would be helpful"

What??????????????

The backyard video really proves nothing other than you can lift the back of your truck. If you would just pick up and read the 43.13 and the NTSB data base you would see I am correct. Everything writen in the 43.13 is Factual and in there for a reason and written with blood. It recomends cable testing and I recomend it as well. Wouldn't that be part of learning things the proper way? Your truck proably weighs 5000 lbs and lastime I checked the majority of the weight is in the front. Just because you lifted half of the truck doesn't mean you lifted half of the weight. Further, what you have created is a basket hitch and each leg ( Each fitting ) is carrying half of the load lifted. So really don't you have any idea what type of load your putting on that cable. It will probably not fail but your methods are unsound. This is what a cable tester looks like http://www.atitools.com/ATICatalog2010PDFs/ATI_Aircraft_Cable_Swaging_2010.pdf Page 4. I have a couple of A&P instructor friends at Vinceness University A&P school in Indianpolis that I worked with at Britt Airways if you are old enough to remember Britt. I am not sure if they have a tester or swagger but I could call them and ask if you could come over and test them on their machine if they have one.

Racegunz
04-29-2012, 06:41 AM
No it proves that even a sloppy job will sustain a continuous load more than you could hold with your control stick. I did say it wasn't scientific, didn't I? and yes the cables ends shared the load I know that. I don't have a problem with testing them witha proper tester and would love to. I put that together in a few minutes mostly to show that the ones that failed must have been a miserable job. I have a 4313 and have no problem with it, I also have confidence in my workmanship. I would love to go to your friends and test some of my cables, now that would be the most helpful thing.

nrpetersen
04-29-2012, 08:59 AM
I thought it was kind of a clever idea to lift a trailer attachment point w an engine lift. :) If these were 1/8 inch cables, the breaking strength should be about 1500 lbs in each strand.

rwanttaja
04-29-2012, 10:43 AM
It's all in how you peel the onion Ron. I have found several accidents besides the ones you quoted using different search terms. Check the NTSB site http://www.ntsb.gov/aviationquery/index.aspx and try cable failure, Nicopress, Nicropress, Cable End, Eye End, and any other applicable search term you can think of.

OK...let's do just that.

During the period from 1/1/1970 to 4/15/2012, running the NTSB database with the "amateur-built" flag set to "Yes":

There were six Experimental amateur-built aircraft accidents that were attributed to Nicopress failures (plus one ultralight, which are not Experimental Amateur-Built aircraft).

There was one case where the failure was a "Nicropress" fitting (plus two ultralights).

There were two instances where references were made to failures of a cable end. One was on a Sukhoi (not an Experimental Amateur-Built aircraft) and the other referred to a throttle cable (not a stress-bearing item).

There was one accident (FTW01LA034) that referenced "Cable Eye". In this case, it had nothing to do with cable fabrication or cable strength.

There was one accident (NYC00LA134) that references "Eye End." This case also did not involve cable fabrication, but improper adjustment of a turnbuckle.

So in that forty-two year period, there were a grand total of seven Experimental Amateur-Built aircraft accidents due to the failure of "home-made" cables. That's about 0.1% of the total accidents. That's an average of one accident every six years. Heck, the incident last month was the first case since 2000.

When you read those Nicopress-failure accidents, one thing seems to stand out: The NTSB investigator didn't have to look far to find the problem. They didn't need a testing lab to ascertain that the builder made the Nicopress fittings wrong. It was apparently visually obvious that the builder made the fittings incorrectly. In one Fly Baby accident, the sleeves had been compressed with a pair of pliers. Quoting another report (CHI83LA256): "WHEN THE PILOT/ BUILDER WAS QUESTIONED ABOUT THE INSTALLATION, HE APPARENTLY DID NOT KNOW THE PROPER INSTALLATION PROCEDURES."

So the problems were not subtle...they probably should have been caught by a tech counselor or another builder. But too many builders don't avail themselves of EAA's services, or insist on going their own way. Four out of those seven cases occurred prior to 1985, which I think pre-dates the tech counselor program.


A cable failure is a failure regardless of homebuilt or production as I know a lot of A&P mechanics make cables improperly as well.

Allow me to quote your initial posting on this subject:


As A Tech Counsler and Airline Mechanic I discourage the use of home made cables!

Few people would consider a cable made by an A&P as a "home made cable."

Ironically, there was one Fly Baby accident (ATL99LA092-- non-fatal, fortunately, the pilot had a chute) that happened because the builder did NOT use "home made cables."


What is an acceptable rate of failure? Even a 2 in 13,000 chance is still unacceptable.

Flight involves risk management, as does building or maintaining an aircraft. The question is, what do you do to minimize the risk at an acceptable cost? Two accidents out of 13,000 aircraft is a 0.016% failure rate...ridiculously low.

That 2 out of 13,000 case was over a 13-year period. Over that same period, there were two additional cases where the builder hooked up the controls in reverse (SEA00LA107 and CHI06CA039). Does this mean we should tell people not to do "home made" control connections? Should they defer critical work like that to licensed A&P mechanics? Is that risk *truly* unacceptable?

During the same time period, there were 96 homebuilt accidents caused by faulty fuel systems. Seems that, if a builder is going to pay to have someone check their work, it'd be a lot better if they had someone go over their fuel systems than tug on a few cables.

Or, for that matter, maybe they should spend the money on some refresher training. As has been posted, the majority of the accidents are due to pilot error, not mechanical issues. If you want builders to spend $200 for increased safety, you get more "bang for the buck" from an hour with an instructor than a formal test to detect errors which are usually visually apparent.


As A Tech Counsler and Airline Mechanic I discourage the use of home made cables!

And, as a run-of-the mill EAA member, I disagree. :-)

Ron Wanttaja

Racegunz
04-29-2012, 01:55 PM
Now that's peeling the onion! Ron. seriously, thanks for taking the time to put that together. who would have guessed the fuel system would be a difficult item to do properly?? I would like to see that broken down to ,pump versus gravity fed systems, and how many were new or neglect of the system that caused failure. Well I have had a great week-end of fabric covering and stitching, so off to do more ploybrush/taping.

Racegunz
04-29-2012, 01:57 PM
I thought it was kind of a clever idea to lift a trailer attachment point w an engine lift. :) If these were 1/8 inch cables, the breaking strength should be about 1500 lbs in each strand.

Field expedient method:D, the truck was just the heaviest thing I had to lift.

rwanttaja
04-29-2012, 04:16 PM
Now that's peeling the onion! Ron. seriously, thanks for taking the time to put that together. who would have guessed the fuel system would be a difficult item to do properly?? I would like to see that broken down to ,pump versus gravity fed systems, and how many were new or neglect of the system that caused failure. Well I have had a great week-end of fabric covering and stitching, so off to do more ploybrush/taping.

Well, the attached is probably a good start. Note that *all* accidents that had fuel system problems are included here, not just the ones where the fuel-system problem actually caused the accident. In a lot of cases, there's a pre-flight error (such as leaving the fuel cap off) which is compounded by some issues with the fuel system.

If you want to read the full report for a given accident, go to:

http://www.ntsb.gov/aviationquery/index.aspx

and enter the NTSB number under "Accident Number" in the NTSB section.

Ron Wanttaja

RV8505
04-30-2012, 12:55 AM
OK...let's do just that.

During the period from 1/1/1970 to 4/15/2012, running the NTSB database with the "amateur-built" flag set to "Yes":

There were six Experimental amateur-built aircraft accidents that were attributed to Nicopress failures (plus one ultralight, which are not Experimental Amateur-Built aircraft).

There was one case where the failure was a "Nicropress" fitting (plus two ultralights).

There were two instances where references were made to failures of a cable end. One was on a Sukhoi (not an Experimental Amateur-Built aircraft) and the other referred to a throttle cable (not a stress-bearing item).

There was one accident (FTW01LA034) that referenced "Cable Eye". In this case, it had nothing to do with cable fabrication or cable strength.

There was one accident (NYC00LA134) that references "Eye End." This case also did not involve cable fabrication, but improper adjustment of a turnbuckle.

So in that forty-two year period, there were a grand total of seven Experimental Amateur-Built aircraft accidents due to the failure of "home-made" cables. That's about 0.1% of the total accidents. That's an average of one accident every six years. Heck, the incident last month was the first case since 2000.

When you read those Nicopress-failure accidents, one thing seems to stand out: The NTSB investigator didn't have to look far to find the problem. They didn't need a testing lab to ascertain that the builder made the Nicopress fittings wrong. It was apparently visually obvious that the builder made the fittings incorrectly. In one Fly Baby accident, the sleeves had been compressed with a pair of pliers. Quoting another report (CHI83LA256): "WHEN THE PILOT/ BUILDER WAS QUESTIONED ABOUT THE INSTALLATION, HE APPARENTLY DID NOT KNOW THE PROPER INSTALLATION PROCEDURES."

So the problems were not subtle...they probably should have been caught by a tech counselor or another builder. But too many builders don't avail themselves of EAA's services, or insist on going their own way. Four out of those seven cases occurred prior to 1985, which I think pre-dates the tech counselor program.



Allow me to quote your initial posting on this subject:



Few people would consider a cable made by an A&P as a "home made cable."

Ironically, there was one Fly Baby accident (ATL99LA092-- non-fatal, fortunately, the pilot had a chute) that happened because the builder did NOT use "home made cables."



Flight involves risk management, as does building or maintaining an aircraft. The question is, what do you do to minimize the risk at an acceptable cost? Two accidents out of 13,000 aircraft is a 0.016% failure rate...ridiculously low.

That 2 out of 13,000 case was over a 13-year period. Over that same period, there were two additional cases where the builder hooked up the controls in reverse (SEA00LA107 and CHI06CA039). Does this mean we should tell people not to do "home made" control connections? Should they defer critical work like that to licensed A&P mechanics? Is that risk *truly* unacceptable?

During the same time period, there were 96 homebuilt accidents caused by faulty fuel systems. Seems that, if a builder is going to pay to have someone check their work, it'd be a lot better if they had someone go over their fuel systems than tug on a few cables.

Or, for that matter, maybe they should spend the money on some refresher training. As has been posted, the majority of the accidents are due to pilot error, not mechanical issues. If you want builders to spend $200 for increased safety, you get more "bang for the buck" from an hour with an instructor than a formal test to detect errors which are usually visually apparent.



And, as a run-of-the mill EAA member, I disagree. :-)

Ron Wanttaja


You lost your argument with me with in your first sentence Ron. Regardless of a hombuilt, production or ultralite a cable end failure is a failure. Unless we can get down to the individual origin of how each cable was constructed the data is simply not availiable. You massaged the data pretty good but this is how I could do it. If we exclude all of the kits with manufacturer provided cables ( Of course we have no Idea how they were made but you would want to believe they were tested. ) from the gross experimental aircraft total. That would remove thousands of aircraft and would definitely move your numbers in the wrong direction. Then we could identify all of the kits that require the builders to manufacture their own cables and then break it down to builders that actually made their own cables and then compare those numbers to production aircraft and on and on. Like I said there are many ways to slice a onion.

I guess we'll have to agree to disagree. I see your argument just as flawed as you do mine. Just because a A&P or even a aircraft manufacturer constructs a cable using amateur practices doesn't make it a professional cable. Other than A&P school many mechanics or have probably never done it. I believe you would have a better chance of getting a bad cable out of a annual on a Cessna than out of a kit manufacturer. Anyway, at the end of the day, anything you swedege on a cable should be load tested at a minimum. The 43.13 recomends load testing and I would not advise anything less. Period. Over and out.

Frank Giger
04-30-2012, 04:35 AM
"I guess we'll have to agree to disagree."

Yep.

Of course I suspect that in person we'd all come to consensus pretty quickly on the importance of ensuring cables are swagged and installed properly and have smiles all around.

That's the weird thing about homebuilding - we all have our areas of special interest that we think are super critical. Nature of the beast, as building an aircraft in one's back yard boils down to determining the level of risks we're willing to take - sometimes a very narrow margin, sometimes, sadly, far too wide.

Your bugaboo is wires, with good reasons; it's an area where nothing less than commercially produced cables is acceptable for your aircraft. That's great.

For me it's the fuel system and restraints! If one craps out, in my mind, the other better by-God work far beyond expectation. I swag my own cables using acceptable practices and check each one with a go/no-go gauge, having accepted the risk as well within tolerable limits.

Bill Berson
04-30-2012, 09:01 AM
It might be helpful if someone would suggest a quick and simple method to load test a cable at home.
Otherwise, I would not bother, since the cable loads are quite low on control cables for homebuilts. Structural wing lift cables are indeed another matter and a method to test would be nice.

Racegunz
04-30-2012, 10:39 AM
I agree with Bill a control cable has much less asked from it than a wing support cable, I would like to add that testing ALL the cables to a much higher load than they will incur is not a good practice, why subject the cable to near failing stresses??? Would you want your wing tested just to the non-permanent defromatiom limit? I know I wouldn't! Modern manufacturing has learned this lesson ,the process is proven and tested, the materials are quality controlled so that the end result is a known good part, can there be flaws? sure ,but it is likely in the process was not followed, and the 4313 says a properly done nicopress is acceptable.
I'm still willing and wanting to test one of my "homebuilt" cables with the proper tester, although I'll be testing MY tools and process the cable will not go in my plane, so Rv8505 have you been in touch with your friends at Vincennes?

RV8505
04-30-2012, 01:24 PM
Do you live close to Indianpolis? I have a call into my freind Ed Brigaman at the Vincenes University A&P school. If I get the ok, I'll contact you by PM. I have been out of Indiana over 10 years now. When do you go bring some of your wife's nail polish and put a little on the nicopress where the cable meets the nicopress and let it dry. After you pull it out of the machine if you see any silver between the nicopress and the cable that is a fail. If all you see is red nail polish that is a pass.

Racegunz
04-30-2012, 02:12 PM
I live about 40 minutes from Indy, not sure the wife would be thrilled to give up some nail polish though.:D

Eric Witherspoon
04-30-2012, 02:59 PM
As far as testing goes, I am building a design now that uses structural cables. In the builder's guide, they address testing. I haven't gotten that far yet, so I haven't done any yet.

But they do not require use of a load cell or the like. If you can come up with a big enough structure to attach your cables to (such as a hangar), and weights (could be 1/4, 1/5, or some other fraction of the total load to be applied), and a lever arm with a known mechanical advantage (attach the cable 1 foot from the pivot point, attach the weight 4 feet from the pivot point to multiply the load on the cable relative to the weight applied). Only thing spooky about this as I write it, is how to apply the weight such that I'm not around it if the cable lets go. Anyway, you can calibrate weight pretty easily. And you can calibrate the mechanical advantage of a lever arm pretty easily. Given that, I would think it wouldn't be too tough to place a known load on a cable. Someone has done this - let us know what you did.

To Racegunz point - I believe these cables are tested to something on the order of 50% of rated load, and then the airplane is designed to only apply (at worst case) something somewhat under this tested load, then, being structural, there's 2 cables at each of these locations. So bolts are going to fail, spars are going to fail, or other really bad stuff is going to happen way before the cables are even challenged.

rwanttaja
04-30-2012, 04:16 PM
It might be helpful if someone would suggest a quick and simple method to load test a cable at home.
Otherwise, I would not bother, since the cable loads are quite low on control cables for homebuilts. Structural wing lift cables are indeed another matter and a method to test would be nice.
Fly Babies have ~8 structural cables per aircraft, and few (if any) have been strength tested. The only ones that have failed have used commercially-manufactured solid tie rods or improper tools/technique.

The basic question... IF:

1. The cable is made with the proper tool, and

2. The tool is used properly, and

3. The proper procedure is performed, and

4. The nicopress sleeves are checked with a gauge and meet the requirements of AC43-13,

THEN:

What is the chance the cable will fail at 60% rated load?

In my opinion, pretty darn low.

If you take a look at the cases where nicopress fittings DID fail, the failed workmanship is obvious. Some examples:

CHI07LA104: "Three of the sleeves from the four outboard aileron cables had crimp dimensions consistent with a properly installed sleeve, as specified in the control cable engineering drawings. However, the remaining three sleeves had crimp dimensions that did not conform to manufacturer's specifications. These nonconforming sleeves also had tooling markings that were not consistent with the use of the specified installation tool."

ATL83FU003: "THE NICOPRESS SLEEVE (SWEDGE FITTING) ON ONE OF THE CABLES HAD BEEN INSTALLED OVER THE PLASTIC COATING ON THE CABLE, RATHER THAN DIRECTLY ON THE CABLE."

ATL00LA015: "During the examination of the rudder cable assembly a nicopress sleeve was found to be not adequately crimped and the sleeve was slightly larger than the gauge dimension. It was determined that one crimped tool was not set correctly to specification on the elevator and rudder cables. After load testing by the manufacturer several cables with oversize nicopress sleeves crimped by the same tool to over 1100 pounds., (required by AC 43. 13-1b for 1/8 cable) which resulted in the cables failing before the nicopress sleeves."

For all three of these cases, the fact that the fitting was not made correctly should have either been visually obvious, or apparent with the application of the go/no-go gauge. In any case, just simply following instructions would probably have prevented an accident.

I'm all for taking precautions... a load test is a nice-to-have. But every accident I've looked at, the sleeve failed because of improper technique or tools.

Ron Wanttaja

steveinindy
04-30-2012, 10:48 PM
the truck was just the heaviest thing I had to lift.

If you had asked, I could have given you my ex-wife's address.


Note that *all* accidents that had fuel system problems are included here, not just the ones where the fuel-system problem actually caused the accident. In a lot of cases, there's a pre-flight error (such as leaving the fuel cap off) which is compounded by some issues with the fuel system.

Thank you for the report. I would tend to point out one thing though: The NTSB does a rather superficial investigation of a lot of experimental crashes which may skew the data somewhat. It's one of the issues that the EAA and AOPA really need to work with the FAA and NTSB to address.

Racegunz
05-01-2012, 07:04 AM
Indiana girls tend on the heavy side, but I thought you were from out of this "backwater" , I was wondering when you were gonna chime in, don't have a cable tester do ya? :D My curiosity is peaked! I just want to see where my cables fail. I agree with the NTSB comment I know of a fatal lightsport trike crash that the FAA/NTSB wouldn't or didn't even investigate they left that to the locals??? I don't read too many of the accident reports though, after awhile I start to get apprehensive about flying, kinda like in highschool the film Blood on the Highways didn't want to drive much after that either.

Mike Switzer
05-01-2012, 07:56 AM
I wonder if Steve's ex was a Woodsie? :D