AW139 lost tail taxying DOH
Crab, while I agree with your comments on the use and damage characteristics of compositie structures, I disagree with people calling the 139 tailboom composite. It is actually constructed from Aluminium, just like the fuselage. It is called composite by many because of the fact that it uses adhesive for the structural joining of parts instead of rivets, in which case we should be calling most, if not all, metal rotor blades composite as well.
The only true composite part of the airframe is from the aft edge of the cockpit doors forward, which is kevlar. The rest of the helicopter structure is Aluminium honeycomb construction. The tailboom construction is quite similar to the forward section of the 412 tailboom, and you should see how they can debond if you don't have Alpine Ejectors fitted!!
I dare say, like the AS365, earlier S-76, and some other types, a heat shield of some sort is needed until a permanent solution is found. I do see that there is a newer tailboom out with different honeycomb material between the metal skins. Will that solve the issues that some customers are having? Only time will tell, but I'm sure they are working on it.
The Old Timer, who said the defects weren't visible? Some people are saying nothing was visible, others are saying that there was something visible before this happened. Hopefully we will all soon know something definite.
The only true composite part of the airframe is from the aft edge of the cockpit doors forward, which is kevlar. The rest of the helicopter structure is Aluminium honeycomb construction. The tailboom construction is quite similar to the forward section of the 412 tailboom, and you should see how they can debond if you don't have Alpine Ejectors fitted!!
I dare say, like the AS365, earlier S-76, and some other types, a heat shield of some sort is needed until a permanent solution is found. I do see that there is a newer tailboom out with different honeycomb material between the metal skins. Will that solve the issues that some customers are having? Only time will tell, but I'm sure they are working on it.
The Old Timer, who said the defects weren't visible? Some people are saying nothing was visible, others are saying that there was something visible before this happened. Hopefully we will all soon know something definite.
Join Date: Jul 2001
Location: ...in view of the 'Southern Cross' ...
Posts: 1,383
Received 0 Likes
on
0 Posts
Mmmm ....
Nooby .....
An opportunity I had to (personally) visually inspect 2 areas of de-bonding on the 139 airframe both of which had the surface skin removed to reveal the "honeycomb" below ... showed a material which was not aluminium honeycomb but which appeared to some sort of "Kevlar" type honeycomb product (or similar) if I am not mistaken (and it has been known for that to happen, on more than one occasion) I would respectfully suggest that this construction is then truly a "composite construction" by definition ie one composed of dissimilar materials! Those areas were ... The Top deck area between the engines normally occupied by the exhaust ducts , and the area of de-bonding on the RHS of the tail-boom exactly where this "incident" damage occurred.
I must acknowledge however to not having been privy to an Engineers course on this Airframe (just a "drivers' conversion!) much to my personal frustration! .... as usual I stand to be corrected as necessary!
Cheers
Nooby .....
An opportunity I had to (personally) visually inspect 2 areas of de-bonding on the 139 airframe both of which had the surface skin removed to reveal the "honeycomb" below ... showed a material which was not aluminium honeycomb but which appeared to some sort of "Kevlar" type honeycomb product (or similar) if I am not mistaken (and it has been known for that to happen, on more than one occasion) I would respectfully suggest that this construction is then truly a "composite construction" by definition ie one composed of dissimilar materials! Those areas were ... The Top deck area between the engines normally occupied by the exhaust ducts , and the area of de-bonding on the RHS of the tail-boom exactly where this "incident" damage occurred.
I must acknowledge however to not having been privy to an Engineers course on this Airframe (just a "drivers' conversion!) much to my personal frustration! .... as usual I stand to be corrected as necessary!
Cheers
spinwing, you are correct, the honeycomb material used is not aluminium. Nomex I think on the new tailbooms? Not sure about the older tailbooms (I'd have to have a look at the BT).
Composite construction for the tailboom yes, but people seem to be implying that the boom skin is made of composite material, which is not correct.
Poor gluing seems to be one of the issues, specially with the top deck area. The other main issue seems to be dissipation of, or protection from, excessive heat. Improvements in the manufacturing processes are definitely needed!
Composite construction for the tailboom yes, but people seem to be implying that the boom skin is made of composite material, which is not correct.
Poor gluing seems to be one of the issues, specially with the top deck area. The other main issue seems to be dissipation of, or protection from, excessive heat. Improvements in the manufacturing processes are definitely needed!
This one above belongs to R44 and Frank already deal with this subject.....
BTW have some info that Robinson is going to perform real life test of
new protective covers with two units flying from UK to East up to middle
of nowhere island....
BTW have some info that Robinson is going to perform real life test of
new protective covers with two units flying from UK to East up to middle
of nowhere island....
Join Date: Aug 2005
Location: All The Places I Shouldnt Be
Posts: 1,009
Likes: 0
Received 0 Likes
on
0 Posts
On the AH1Z I remember they had problems with the exhaust heat causing MAJOR issues to the tailboom and their solution was to create new exhausts directing the heat upwards and away from the boom.
Could the heat from the exhausts be a contributing factor to this problem and could the same solution be applied.
Just wondering.
Ned
Could the heat from the exhausts be a contributing factor to this problem and could the same solution be applied.
Just wondering.
Ned
Join Date: Jul 2008
Location: Australia
Posts: 372
Likes: 0
Received 0 Likes
on
0 Posts
composite construction vs composite material
Noooby is correct. There is a lot of confusion about differentiating between composite construction (adhesive bonded structures) and composite materials (fibres in a plastic matrix). I have never understood why the FAA lumps these two completely separate technologies into one. See for example AC-20-107A, and the entire basis for training in composite repairs. Just because a technician can successfully fabricate a piece of fibre-glass does not mean that he can prepare a metallic surface for sticking that fibre-glass to it.
Adhesive bonding and composite fabrication are two completely different technologies. The design methodology is different, the failure modes are different, testing methods are different. The only commonality is that they start off as a sticky material which hardens to form the final product. The FAA lumps them together for the purposes of certification, yet if adhesive bonded structures were designed on actual strength (rather than an average shear stress basis which is known to be invalid) the certification methodology could be dramatically streamlined, saving heaps in certification costs.
What is totally missing in the FAA system at present is a requirement to certify that the bonding processes (for metals or composites) actually do produce durable bonds which will last for the life of the aircraft and will not result in failures like the one shown in the adjacent photograph.
The way to tell if the AW139 failure is due to bond durability is to inspect the failure surface using the Mark I eye-ball. If the adhesive has fractured leaving a complete coating on both surfaces, then is is a cohesion failure and the failure is due to a design issue or an overload. However, if the adhesive separates at the interface, then this is a bond durability issue which is 99.99% due to the process used to prepare the metal for bonding at the time of manufacture. And just to be clear, humidity and temperature actually only accelerate failure of bonds with poor durability. An absence of humidity and temperature simply delay the same outcome, which will be interfacial failure ata a later time than if humidity and temperature were involved.
Adhesive bonding and composite fabrication are two completely different technologies. The design methodology is different, the failure modes are different, testing methods are different. The only commonality is that they start off as a sticky material which hardens to form the final product. The FAA lumps them together for the purposes of certification, yet if adhesive bonded structures were designed on actual strength (rather than an average shear stress basis which is known to be invalid) the certification methodology could be dramatically streamlined, saving heaps in certification costs.
What is totally missing in the FAA system at present is a requirement to certify that the bonding processes (for metals or composites) actually do produce durable bonds which will last for the life of the aircraft and will not result in failures like the one shown in the adjacent photograph.
The way to tell if the AW139 failure is due to bond durability is to inspect the failure surface using the Mark I eye-ball. If the adhesive has fractured leaving a complete coating on both surfaces, then is is a cohesion failure and the failure is due to a design issue or an overload. However, if the adhesive separates at the interface, then this is a bond durability issue which is 99.99% due to the process used to prepare the metal for bonding at the time of manufacture. And just to be clear, humidity and temperature actually only accelerate failure of bonds with poor durability. An absence of humidity and temperature simply delay the same outcome, which will be interfacial failure ata a later time than if humidity and temperature were involved.
blakmax, haven't seen a tailboom delam on a 139, so don't know about the failure type there. Have seen and repaired the upper deck delams, and the glue all stayed with the skin on the ones I saw. Basically none left on the honeycomb.
Smells like poor quality control at manufacture to me. I'm sure heat has something to do with it, but if the bond joint was 100% to begin with, I don't think we would be seeing these problems.
Smells like poor quality control at manufacture to me. I'm sure heat has something to do with it, but if the bond joint was 100% to begin with, I don't think we would be seeing these problems.
Join Date: Jul 2008
Location: Australia
Posts: 372
Likes: 0
Received 0 Likes
on
0 Posts
Honeycomb disbonds
Hi nooby and spinwing. What you are describing is termed adhesion fillet bond failure. This was a phenomenum which led to in-flight failure of a significant number of rudders on one type of military aircraft. It is quite significant in that the flatwise tensile strength of the sandwich structure (the load necessary to pull the skin vertically off the core) degrades to less than 10% of original strength. see DAVIS, M.J., Chester, R.J., Perl, D.R., Pomerleau, E., Vallerand, M., Honeycomb Bond and Core Durability Issues; Experiences within CREDP Nations, Aging Aircraft Conference, Williamsberg, VA, Aug 31-Sep 02 1998. I can mail you a copy if you PM me your address.
The real problem is that it is very difficult to detect until the problem is significant. If there is still contact between the core and the adhesive, it is possible to transmit sound waves (tap test or ultrasonic) and a false positive results. It is usually only when the skin experiences out of plane loads (aerodynamic, buckling or internal pressure due to heating) that the skin separates from the core and the problem can be found.
The problem typically occurs where water has entered a sandwich panel (often through those stupid injection repairs which never work anyway) and the interface hydrates. If the problem occurs early in the TIS, it is probably because the core was contaminated or not properly dried at the time of production. Nomex core absorbs a large amount of water just sitting on a bench. (I have measured 5% by weight.) If the production environment is not humidity controlled and if the core is not moisture evacuated prior to bonding, then the water will inhibit any chemical reactions at the interface and the bond will be weak.
Regards
Blakmax
The real problem is that it is very difficult to detect until the problem is significant. If there is still contact between the core and the adhesive, it is possible to transmit sound waves (tap test or ultrasonic) and a false positive results. It is usually only when the skin experiences out of plane loads (aerodynamic, buckling or internal pressure due to heating) that the skin separates from the core and the problem can be found.
The problem typically occurs where water has entered a sandwich panel (often through those stupid injection repairs which never work anyway) and the interface hydrates. If the problem occurs early in the TIS, it is probably because the core was contaminated or not properly dried at the time of production. Nomex core absorbs a large amount of water just sitting on a bench. (I have measured 5% by weight.) If the production environment is not humidity controlled and if the core is not moisture evacuated prior to bonding, then the water will inhibit any chemical reactions at the interface and the bond will be weak.
Regards
Blakmax
Aha, Nomex is the water sponge. That explains sucking so much water out of the top deck during the repair!
Agusta BT139-159 is an optional bulletin that details the replacement of some (looks like most, if not all!) tailboom panels. The original panels have nomex honeycomb, the new ones have metallic honeycomb. Anyone with a 00234 tailboom, you have Nomex honeycomb. If you have a 00235 tailboom, the metallic honeycomb is inside. Supposed to have superior adhesion at high temps, and won't be as affected by moisture ingress.
Better go check your tailboom P/N's!
Agusta BT139-159 is an optional bulletin that details the replacement of some (looks like most, if not all!) tailboom panels. The original panels have nomex honeycomb, the new ones have metallic honeycomb. Anyone with a 00234 tailboom, you have Nomex honeycomb. If you have a 00235 tailboom, the metallic honeycomb is inside. Supposed to have superior adhesion at high temps, and won't be as affected by moisture ingress.
Better go check your tailboom P/N's!
Join Date: Jul 2008
Location: Australia
Posts: 372
Likes: 0
Received 0 Likes
on
0 Posts
More on core problems
Hold that thought nooby!
Another problem with honeycomb core is that you can not do an adequate surface preparation process on bare core. To get a durable bond to aluminium, you need a chemically active surface which is treated to develop a resistance to hydration. You can not abrade or chemically treat bare aluminium core material, so it is usually just solvent degreased (if anything is done at all!). Hence, the bond is still susceptible to hydration.
The real solution to hydration is for the manufacturer to use phosphoric anodised core which has an epoxy primer applied as part of the core manufacturing process. ALCORE and Hexcell offer these products. In that case the interface between the aluminium and the primer is appropriately managed, and when the part manufacturer bonds to the core, he performs an epoxy-to-epoxy bond on which is far easier to generate hydration resistance. The studies outlined in the reference in the previous message found that PAA and primed core was considerably less susceptible to adhesion fillet bond failure between the adhesive and core.
So, the question is "Did Augusta use PAA and primed core?" If not, they simply delay the onset of the same problem.
Regards
Blakmax
PS don't you ever sleep?
Another problem with honeycomb core is that you can not do an adequate surface preparation process on bare core. To get a durable bond to aluminium, you need a chemically active surface which is treated to develop a resistance to hydration. You can not abrade or chemically treat bare aluminium core material, so it is usually just solvent degreased (if anything is done at all!). Hence, the bond is still susceptible to hydration.
The real solution to hydration is for the manufacturer to use phosphoric anodised core which has an epoxy primer applied as part of the core manufacturing process. ALCORE and Hexcell offer these products. In that case the interface between the aluminium and the primer is appropriately managed, and when the part manufacturer bonds to the core, he performs an epoxy-to-epoxy bond on which is far easier to generate hydration resistance. The studies outlined in the reference in the previous message found that PAA and primed core was considerably less susceptible to adhesion fillet bond failure between the adhesive and core.
So, the question is "Did Augusta use PAA and primed core?" If not, they simply delay the onset of the same problem.
Regards
Blakmax
PS don't you ever sleep?
Westlands have past experience with debonding items - the TR driveshafts on the Lynx had a titanium plug glued into the steel driveshaft tube and because there were no rivets or bolts, there were no witness marks to show the bonding was letting go.
Sadly this caused the loss of two AAC pilots when the bonding in a TR driveshaft let go at 200' in the hover on an airtest.
It then turned out there were 3 differently modified TR shafts in the fleet and the type that failed was being phased out.
Sadly this caused the loss of two AAC pilots when the bonding in a TR driveshaft let go at 200' in the hover on an airtest.
It then turned out there were 3 differently modified TR shafts in the fleet and the type that failed was being phased out.
Join Date: Jul 2008
Location: Australia
Posts: 372
Likes: 0
Received 0 Likes
on
0 Posts
bond failure
Sadly Crabby, this is what happens when people believe that you only need a clean surface to get a good bond, and short term testing as stipulated by the FARs lets such bonds pass certification.
The organisation I worked for actually listened to reason and after we changed our processes based on a reliable validation test, we had only three failures in bonded repairs since 1992, out of a total exceeding 4000 repairs. In every case, we could determine that technician error (or laziness) was the source of the failures.
As long as we have certification rules based on horns-and-hooves adhesive bonding technology we will see bond failures which may unfortunately result in loss of aircraft and lives where the bond is flight critical.
Even worse, the level of comprehension of forensic data by crash investigators is often very low. I don't mean any disrespect to our wonderful crash investigators, but this really is a specialised area and it probably only constitutes a few percent of their experience, and their knowledge is often hand-me-down based on bad historic information. I have seen examples of this low level of knowledge where one investigator found fatigue markings in a bond failure and concluded that the bond failed by fatigue. Pity that the striations were at the middle of the bond and the rest of the disbond was at the interface. The two features could never be related. I could list several other examples where incorrect conclusions were drawn, but I don't wish to extend the message length too much.
Regards
blakmax
The organisation I worked for actually listened to reason and after we changed our processes based on a reliable validation test, we had only three failures in bonded repairs since 1992, out of a total exceeding 4000 repairs. In every case, we could determine that technician error (or laziness) was the source of the failures.
As long as we have certification rules based on horns-and-hooves adhesive bonding technology we will see bond failures which may unfortunately result in loss of aircraft and lives where the bond is flight critical.
Even worse, the level of comprehension of forensic data by crash investigators is often very low. I don't mean any disrespect to our wonderful crash investigators, but this really is a specialised area and it probably only constitutes a few percent of their experience, and their knowledge is often hand-me-down based on bad historic information. I have seen examples of this low level of knowledge where one investigator found fatigue markings in a bond failure and concluded that the bond failed by fatigue. Pity that the striations were at the middle of the bond and the rest of the disbond was at the interface. The two features could never be related. I could list several other examples where incorrect conclusions were drawn, but I don't wish to extend the message length too much.
Regards
blakmax
Somma Lombardo, Italy
August 26th, 2009
AgustaWestland Statement
To: All Customers and Operators of AW139 Helicopter Model
All AgustaWestland Authorized Service Centres for the AW139
Helicopter Model
Subject: AW139 Tailboom Damage Occurrence
Dear Customer,
The purpose of this Letter is to inform You that on August 25th, 2009 an AW139
aircraft had an occurrence resulting in major damage to the aircraft tail boom while ground taxiing.
Strange name for Failure, (my bold highlight) what about the rest of the tails that are showing serious problems, "occurrences or failures".
Do these machines have HUMS?.
What is the inspection regime to detect failure in-service.
If as some say there is no outward sign of de lamination, how is a pilot to know\inspect for a serviceable craft.
August 26th, 2009
AgustaWestland Statement
To: All Customers and Operators of AW139 Helicopter Model
All AgustaWestland Authorized Service Centres for the AW139
Helicopter Model
Subject: AW139 Tailboom Damage Occurrence
Dear Customer,
The purpose of this Letter is to inform You that on August 25th, 2009 an AW139
aircraft had an occurrence resulting in major damage to the aircraft tail boom while ground taxiing.
Strange name for Failure, (my bold highlight) what about the rest of the tails that are showing serious problems, "occurrences or failures".
Do these machines have HUMS?.
What is the inspection regime to detect failure in-service.
If as some say there is no outward sign of de lamination, how is a pilot to know\inspect for a serviceable craft.
Join Date: Nov 2008
Location: UK
Age: 66
Posts: 919
Likes: 0
Received 0 Likes
on
0 Posts
If as some say there is no outward sign of de lamination, how is a pilot to know\inspect for a serviceable craft.
tailboom skin on other helis
Slightly off topic,
Does the skin of a tailboom on light helis such as AS350 Bell 206 or r44 form a significant part of its structural integrity?
Mickjoebill
Does the skin of a tailboom on light helis such as AS350 Bell 206 or r44 form a significant part of its structural integrity?
Mickjoebill
Join Date: Apr 2003
Location: Germany
Age: 76
Posts: 1,561
Likes: 0
Received 0 Likes
on
0 Posts
Right pedal?
In the picture the tail boom is displaced to the right, which argues that it was giving the nose a push to the left, or is there something here I am overlooking? (I know very little about helicopters.) I would expect to see the boom off to the left for a right yaw input there.
Is the tailboom so weak that it will break before the nosewheel tires skid, assuming that the steering is locked with full yaw input?
Is the tailboom so weak that it will break before the nosewheel tires skid, assuming that the steering is locked with full yaw input?
Join Date: Aug 2006
Location: Below Escape Velocity
Posts: 416
Likes: 0
Received 0 Likes
on
0 Posts
chuks-
The tailrotor on the 139 is a tractor (not your red, Donetsk, Ukraine manufactured tractor... but a tractor nevertheless). Many Bells (206, 407) are pushers. The Huey folk can tell all manner of tales on tailrotor placement and direction of rotation and the like... all rather confusing.
On the 139, the main blades turn counter (anti, as our British friends would have it) clockwise when viewed from above, as do most American, German, and Italian single-rotor helicopters. This would mean that an increasing torque input (due to increased pitch pull on the main blades, which one tends to need to do to set the 139 in motion on the wheels) would require an increase in left pedal (also known in some circles as antitorque), pulling the tailboom to the right.
The clever observer, viewing a sound 139, will also notice that the tail rotor is canted to the left to provide a bit of lift to the tail to counteract the rather aft CofG in the hover. The long-nose versions are a bit more forward CofG by all reports, but I haven't flown one of those just yet.
The 139 has a nosewheel that automatically centers and locks upon liftoff and has to be selected to unlock upon landing. Because of the aft CofG, the nosewheel is the first thing to lift off the ground in the hover, but by the same token, hasn't got an enormous load pinning it to the ground, either. I'm not particularly fond of this locking nosewheel feature. However, I don't pretend to know entirely what happened in this incident, though I watch with interest.
The boom is probably less likely to fail in the other direction because that is when loads from the tailrotor are lower.
French and Russian single-rotor machines tend to have directions of rotations that are reversed.
The tailrotor on the 139 is a tractor (not your red, Donetsk, Ukraine manufactured tractor... but a tractor nevertheless). Many Bells (206, 407) are pushers. The Huey folk can tell all manner of tales on tailrotor placement and direction of rotation and the like... all rather confusing.
On the 139, the main blades turn counter (anti, as our British friends would have it) clockwise when viewed from above, as do most American, German, and Italian single-rotor helicopters. This would mean that an increasing torque input (due to increased pitch pull on the main blades, which one tends to need to do to set the 139 in motion on the wheels) would require an increase in left pedal (also known in some circles as antitorque), pulling the tailboom to the right.
The clever observer, viewing a sound 139, will also notice that the tail rotor is canted to the left to provide a bit of lift to the tail to counteract the rather aft CofG in the hover. The long-nose versions are a bit more forward CofG by all reports, but I haven't flown one of those just yet.
The 139 has a nosewheel that automatically centers and locks upon liftoff and has to be selected to unlock upon landing. Because of the aft CofG, the nosewheel is the first thing to lift off the ground in the hover, but by the same token, hasn't got an enormous load pinning it to the ground, either. I'm not particularly fond of this locking nosewheel feature. However, I don't pretend to know entirely what happened in this incident, though I watch with interest.
The boom is probably less likely to fail in the other direction because that is when loads from the tailrotor are lower.
French and Russian single-rotor machines tend to have directions of rotations that are reversed.
Is anyone seriously under the impression that the nosewheel lock had anything to do with this?
