Quantcast

Conclusions on Aluminum Adhesive Bonding Tests

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
14,001
Location
Port Townsend WA
With all due respect to the experts here, Moni's, Monarai's, Cri-Cri's, and HP-11, HP-14, HP-16, RS-15, and HP-18's are not falling from the sky with debond failures, and many, many examples have very significant service histories. A blanket dismissal of the technique flies in the face of reason - all evidence points to it being perfectly adequate for these relatively low-stress applications.
Monett has abandoned bonding because of problems. I am not sure if there was a fatal accident. And Richard Shreder spoke of a total wing debond on his HP glider at a forum at Oshkosh. The history of bonding is not that good, in my opinion.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,255
Location
Orange County, California
Well, thanks, good to know. I haven't heard any of that, and quite a few Moni's and Monarai's are still in the air with the original wings. The Cri-Cri's still seem to have a good service record, too. And one HP debond is not a trend. If there's a genuine body of evidence to show that these aircraft are having debond problems, then that's another matter entirely.

It's not that I'm unwilling to be pursueded, but I'd like to see more concrete evidence that the method truly is inadequate before it's dismissed out of hand.
 

MadRocketScientist

Well-Known Member
Joined
May 16, 2009
Messages
1,660
Location
Canterbury, New Zealand, Earth, Milky Way Galaxy.
Well, thanks, good to know. I haven't heard any of that, and quite a few Moni's and Monarai's are still in the air with the original wings. The Cri-Cri's still seem to have a good service record, too. And one HP debond is not a trend. If there's a genuine body of evidence to show that these aircraft are having debond problems, then that's another matter entirely.
The CriCri uses Rivets and bonding on its fuselage and only the wings are a foam/Aluminium join without rivets. The fuselage can also be built without any bonding if 1/8" rivets are used instead of 3/32". I have yet to hear of wing debonding but I will keep my ear to the ground.

Personally I don't know if I would trust an aluminium/aluminium adhesive join without safety (chicken:gig:) rivets. I tend to see the process of a homebuilt bonded join to variable to trust completely in that respect.

Shannon.
 

blakmax

Member
Joined
Sep 15, 2009
Messages
13
Location
Brisbane Australia
Gary (and others)

I noted the comment
I'm not sure that your concern about hydration of the bond and delamination due to corrosion is as much of a concern to me. I live in sunny Southern California.
All that your dry southern Californian environment will do is to let your structure survive a little bit longer. I said that hydration is exacerbated by hot, wet environments. I did NOT say that the absence of such environments eliminates the problem. IT ONLY SLOWS IT DOWN.
All epoxy materials absorb moisture from the environment and as someone else pointed out, this occurs on a molecular level. It does not need the presence of condensation to occur. Now I know that SC has a dry environment, but I also know that they have occasional fogs and frequently have morning dew. This indicates that there is humidity, and that is all that is needed to promote moisture absorption and hence hydration. I have personally consulted for an aircraft manufacturer in SC and found a direct correlation between rain on the day components were manufactured and voids caused by moisture absorbed in hot-cure adhesives.

Your argument is akin to accepting that it is OK to be a little bit pregnant.

Let me also deal with the next issue. You can not stop this with sealants, because again it is at a molecular level. Sealants may prevent the direct transmission of fluids, but they don't prevent diffusion of moisture.

Now for primers. Some primers are beneficial, provided that the correct process is exactly followed. It is the process which is the key, not necessarily the primer. Using a primer with the wrong process simply changes the colour of the disbond. Many primers act solely to reduce corrosion and have little beneficial effect on bond durability. The major constituent in AC130 is a silane coupling agent which is effective as a primer and does inhibit corrosion. We have used the coupling agent alone (gamma glycidoxy propyl trimethoxy silane for the chemically minded) without the additives in AC130 for seventeen years with no disbonds when the process is done correctly. That material is available from Dow Corning as Z6040.

If anyone seriously thinks that they don't need to use such chemical processes, then please let me know the address to send the flowers.

Regards

blakmax
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
blakmax, I believe you.

I've seen first hand what Topaz is talking about. On the subject of moisture ingress, whether it be under paint or epoxy is facilitated by any kind of crack, nick or any kind of opening in the epoxy or paint. That moisture gets in there and slowly works it's way under the surface, delaminating as it goes.

On the subject of epoxy, I had concluded a few months back, and I reported it here that the relatively hard epoxies are worthless for bonding metal to metal becasue they debond so easily, and suddenly when the metal yields under load. It just gives up. Methacrylates and Proseal hold tough even as the two parts are pulled appart. I have not tested any epoxies that are as elastomeric as meth and proseal.

I have mixed a batch of 10 parts epoxy with one part Proseal and came up with a fairly elastomeric material. Of course the A and B parts of both compounds must be mixed together before mixing the two compounds. I have no other data about the performance of this invention, but I'm guessing that the results are a compromise between the two, which isn't such a bad thing.

On the subject of "cheater rivets", there is no such thing as a cheater rivet...it's just good design. Every joining method has its Achilles heal. All structures begin to fail at the stress riser. Adhesives are all weakest in peel because peel concentrates all the load at one point. Rivets prevent peel. As long as you can prevent the metal from deforming under load, the load will tend to be distributed over a large area of the bond joint.
 
Last edited:

blakmax

Member
Joined
Sep 15, 2009
Messages
13
Location
Brisbane Australia
Where do I start? Firstly Gary, it is not good practice to make a witches brew of different materials. For example you may well accidentally provide an over-abundance of one curing agent and end up with the other component being undercured. I'd urge extreme caution in arbitrarily modifying adhesive products.

Next to the merits of adhesive bonding. The main problem with adhesive bonding is not the technology, it is the lack of understanding of the mechanisms involved. Structural adhesives which are capable of transmitting adequate loads rely on covalent chemical bonds formed at the interface, and the resistance of those bonds to hydration as previously discussed. Paints, sealants etc. which are NOT intended to transfer loads can rely on weaker chemical bonds such as electrostatic attrative bonds (van der Waal's). This is why paint and sealants can stick but adhesive bonds don't.

Next to the Chicken rivets. There is a reason why test methods for adhesive bonds use short overlap lengths. Because if you used longer overlap lengths, the adhesive will not fail, the metal will. Hence you would test the strength of the metal, not the adhesive. Now just think for a while. If I am making an aircraft, what do I want to be most reliant on; the adhesive or the metal? SO, if the overlap lengths are adequate AND the processes are valid, you should ALWAYS fail the metal and the adhesive strength is no longer an issue.

To understand the overlap length requirements you need to understand the shear stress distribution in an adhesive bond. The common perception is that all of the load is distributed evenly over the adhesive, so that if you double the overlap length of the joint you can carry twice the load. This is total crap. The shear stresses in an adhesive bond are very high at the ends of the joint and decay towards the centre of the joint. The analogy is that of a bath-tub shaped shear stress curve. Provided the overlap length is large enough, the bottom of the bath tub will approach zero shear stress. At that stage, adding more overlap length simply adds to the length of the bottom of the tub and does not make one ounce of difference to the maximum shear stress.

Now lets put a fastener in the joint. Edge distance considerations require a minimum separation between the fastener and the end of the joint of 1.5 times the diameter of the fastener, 1.5d. The size of the shear stress area at the end of the joint (the edge of the bath tub) is usually smaller than 1.5d. Hence, all of the load is transferred before the fastener takes any load.

The ONLY value of fasteners is to demonstrate which designer did not understand adhesive bonding, or to overcome bad processes, in which case the joint must be designed as if there was no adhesive because eventually it will disbond. I have data which shows that a multi-step bonded joint between composite and titanium in a certain military aircraft could carry over 35,000 lb/in if it was bonded. If it was bolted, it could carry only about 28,000 lb/in. If it was bonded and bolted, the joint could carry only about 33,000 lb/in so almost 2000 lb/in was lost by adding the bolts. Worse yet, the bolts carried a total of less than 500lbs because they were located in that part of the bond where the load had already been transferred by the adhesive.

Now to the hardness of the bonding agent. The reason your results indicate that hard bonding material is so poor is in fact a measure of the inadequacy of your processes, not the adhesive. Have a look at your specimens. Did you actually fracture the adhesive or did the adhesive pull off the surface? If the adhesive fractures, then it is an adhesive bond strength issue possibly due to short overlap lengths etc. If the adhesive separated predominantly from the interface, then it is a process deficiency.

In fact, "hard" adhesives have a high shear modulus which means that for the same relative displacement of the joint, they will transfer far more load. If the processing is correct, these adhesives should require a smaller overlap length than for more compliant materials which effectively stretch heaps to transfer the same load.

Now for the fundamental question: why would you want to use processes which have a knwon, long history of producing bond failures so that you can save a few dollars and a few hours? If I was building an aircraft, I would want to have total confidence in my structure. You can not get that level of confidence by taking short cuts or ignoring the (free!) advice of an expert with over 37 years experience in this field.
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
I knew I'd be taking heat for my "witches brew". That one-part Proseal was left over from a batch I had mixed the night before and I had it sitting in the freezer, so it was still in a fluid state. When I was mixing the epoxy the next day I just thought I'd throw in the Proseal so as to not waste it. The resulting witches brew is a little more elastic than the straight epoxy.

I mix my adhesives by weighing the two components on a triple beam scale that reads to 0.1 gram. You can mix Bondo by eye, not so with these adhesives.

I should back away from saying that epoxy has no place in bonding aluminum. It has just been my experience that the quality and variation of epoxy bonds are more difficult to control than that of Proseal and methacrylate. The performance of the latter two was always more reliable.

blakmax,

I very much appreciate your valuable input. This forum is only as good as the truest input and the ability of it's messenger to communicate what he knows. You and Malcolm are apparently the most exerienced in this body of knowledge.

I can only speak from my experience and what I have found to work and not work. I have acquired quite a bit of experience in making and destroying samples and comparing those samples with known technology (rivets) to determine some level of equivalency. I have no experience in the effects of long term environmental exposure on these types of adhesives.

We all appreciate and are indebited to you for your input and are very glad you have joined this forum.

I will be purchasing a quantity of AC130 in the near future.
 
Last edited:

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
Next to the Chicken rivets.......
Very well said but I'd like to add a bit: This subject is very difficult to address sufficiently in the homebuilt arena since the vast majority of builders generally do not have the proper equipment, materials or facilities available to them for proper surface preparation and maintenance, nor for proper bonding procedures. The goal of quite a few of us in this industry is to find a process, or processes, that will enable the average builder to bond his aluminum airplane, without the concerns raised herein. We can talk about various chemicals, cleaning agents, surface prep techniques and whatever, but in reality the likelihood that a builder will be able to properly follow said techniques to the "nth" degree is pretty low.

For this reason we are looking at many different materials and techniques, as well as processes for assembly. The subject of chicken rivets has nothing to do with lap shear strengths - it is being discussed and used primarily for the prevention of the onset of edge peel. True, in general aerospace use, this is not an issue since all bonds are designed and accomplished in such a way that all edges are properly adhered and stable. Application of bonded structures is also done in a way that outside effects do not introduce loading that would lead to this type of failure and in instances where the potential exists, the design is configured to minimize the chance.

In homebuilding however we do not have this luxury since in some instances the classical structures will be subject to loading and environmental conditions where the increased chance of edge peel does happen. Even certified production airframes that use bonding (Grummans) are, from time to time, subject to this type of failure and interestingly enough, the "approved" fix is the use of edge "chicken" rivets.

In the homebuilt arena, when looking at less than ideal bonding processes, service conditions exist that can introduce significant loading or vibratory conditions that could lead to some form of edge deterioration so the use of "chicken" rivets is often recommended just for that extra level of prevention. The bonds themselves are, as you indicated, more than strong enough for the anticipated service however, edge failure modes could lead to the overall bond failure nevertheless. Whether we're talking about service loads, acoustic vibrations, environmental issues or simple hangar rash, all these tend to be much more prevalent and significant in a small airplane, especially when you couple in the issue of the homebuilt environment and potential build quality. And it is this combination of less than ideal conditions that leads us to consider components or assembly techniques that you might not use in the mainstream aerospace sector, but that are advisable here.
 

blakmax

Member
Joined
Sep 15, 2009
Messages
13
Location
Brisbane Australia
Steady guys. Just because you don't have the whiz bang gear it is no excuse for doing nothing at all or doing things we know are wrong. You can still produce durable bonds in a garage with minimal precautions and limited equipment.

The idea of using rivets to prevent peel is in fact misplaced confidence. Peel also occurs at the edges of the bond. If the rivet is 1.5 d from the edge then the peel disbond can propagate some considerable distance before the rivet becomes effective. If the adhesive peels so easily, it is almost certainly due to deficient surface preparation again.

Do you guys have conferences or meetings where you exchange ideas? It may be possible to organise a training session on adhesive bonding at such a venue.

Regards

blakmax
 

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
Well, as you are "down under", we are also spread out all over. In other words, no meeting. But there are many here that provide expertise in this area and several of us have been in associated industries for quite some time so that we feel confident in presenting our views herein.

In my case I've been investigating aluminum bonding in a homebuilt environment for well over fifteen years and although I've found excellent candidate materials and have worked on several one-off and certified programs where aluminum bonding is key to the products, I have yet to find a process that I would find consistent enough to recommend for kit production or scratch-built use. Even the best candidate materials and prep techniques will demonstrate a less than 100% reliability so in my opinion, it's not something that I would feel comfortable recommending just yet. Even the certified production lines I've been involved with (Grumman, Boeing, etc.) have enough failure rates that all out confidence is as of yet unlikely for an amateur built airframe.
 

blakmax

Member
Joined
Sep 15, 2009
Messages
13
Location
Brisbane Australia
Fair call Orion. There are many processes out there which give initially sound structure but which fail in later service. Even from large manufacturers including the ones you mention. In many cases the field level repair procedures produce even worse results.
However, it has been our experience that where the process has been validated using the wedge test and measured against the criteria outlined in [FONT=&quot]DOT/FAA/AR – TN06/07, Apr 2007[/FONT] then durable and reliable bonds result. On this basis we have virtually eliminated adhesive bond failures on bonded repairs to aluminium sandwich structure since 1992. Three bond failures out of over 4000 repairs and every one was caused by technician short cuts.
And we bond to clad or unclad surfaces without a primer and we do not use chicken rivets. These repairs are on high performance supersonic jets, so we have every confidence in bonding technologies. Now these include on-aircraft repairs in a hangar environment, so it is possible to perform adequate reliable adhesive bonds in less than optimum conditions.

I urge people to read [FONT=&quot]DOT/FAA/AR – TN06/07, Apr 2007[/FONT] and if you can't get a copy PM me and I'll send a copy of the final draft.

Regards

blakmax
 

MalcolmW

Well-Known Member
Joined
Jan 21, 2007
Messages
118
Hello, Blakmax, Gary S., and Orion;

I just got back from an overseas trip and missed (until today) the fascinating discussion on adhesive bonding.

I do find the comments made by Blakmax positive and insightful and agree with him for the most part (the October 3 post). I have emphasized many times (on this panel) as to the need for proper surface preparation for adhesive bonding (chromic acid, chromate conversion or phosphate etch) combined with cleanliness. I do agree that abrasive scrubbing and solvent is not the way to go for a reliable bond.

As for the adhesive, my experience tells me that epoxies work well with aluminum, however, the application requirements may be demanding for the home-builder. I worked with epoxy-novolac (phenolic) adhesives for bonding applications for high-performance military aircraft (all-weather) which performed admirably in all climates, including marine applications. This type of adhesive, used properly, would prove stronger than the aluminum test panels. The requirements (elevated temperatures and uniform pressure during the adhesive cure) would be difficult to achieve in a garage environment. So, I recognized the need for an adhesive system that would be more tolerant of that type of working environment, which led to the methacrylate adhesives (used extensively to bond aluminum truck bodies) which Gary investigated.

As for the use of rivets during the bonding process, I believe they offer utility due to their ability to 'locate' pieces during assembly, sort of a built in 'jig' to assure proper positioning and clamping. The use of a fiberglass scrim cloth with adhesives provides a means to apply a uniform (thin) coating and does not offer any additional strength.

ProSeal is a two-component polysulfide system which can produce a range of elastomeric properties ranging from highly elastic to a very stiff polymer (high durometer) material. The system is more tolerant of surface contamination than many adhesive systems. Bede used (or uses) this in the assembly of some aircraft.

I do need to read further on the references offered by blakmax, however, I suspect that we agree on much about the adhesive bonding process (about which I've posted a number of times).

I encourage the use of adhesives in aircraft assembly due to the advantages they offer: Strength, resistance to fatigue, lower weight and possibly lower drag.

Blakmax, thank you for sharing your knowledge and experience with us.

As always, do some testing and fly safe,

MalcolmW
 

fasee1819

New Member
Joined
Oct 6, 2009
Messages
1
Location
roorkee,uttarakand/india
dear sir,

presently i am working on weld bonding of aluminum alloys.adhesive bonding also a part of my project work.i want to study the role of adhesives on aluminum alloys.please tell that what is the standard joint design for aluminum in adhesive bonding(lap joint)/astm standard number,,,

can any one please explain the role of curing temperature on adhesives?
what happen if going on increasing?
 

blakmax

Member
Joined
Sep 15, 2009
Messages
13
Location
Brisbane Australia
Welcome back Malcolm W. Hope you enjoyed the trip. Thanks for the comments.
I note the military experience. Several years ago I was involved in a program involving NADEP-NI which found that the chromic acid etch process (pasajel) would only work if the surface was heated above 140F. Down under we avoid it because of the dangers of hexavalent chromates which are known carcinogens.

I agree that epoxy-phenolic adhesives give better performance and are more tolerant of surface preparation issues than epoxies, but they are I believe condensation polymerisers (rather than addition polymerisers such as epoxies). Hence they give off a byproduct which is usually water. They also require elevated temperature for cure, so the water turns to steam and hence high process pressures are necessary to minimise the void content. They may also be susceptible to voiding if excessive vacuum is applied during cure. Vacuum does not draw out the voids, it makes them bigger due to the low pressure. It is the adhesive which flows out of the joint. The way we have managed this is to apply full vacuum until just before the adhesive gels, then reduce the vacuum to 10 in Hg. That reduces the size of the micro-voids and draws adhesive back into the joint. It also helps if you minimise exposure of the adhesive to the atmosphere during storage and use.

Summary: Good adhesive, but care with the storage and manage the pressure during cure.

Regards

blakmax
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
Boy, we got the three heavy adhesive hitters here all on the same page. I don't know anything about the phenolic adhesives. As far as recommending adhesives for airframe structures, my experience thell me that epoxy, although it is strong, getting a reliable bond is not a sure thing. ProSeal is very reliable, but of low strength, so if you double up on the bond surface area you may be able to compensate. Methacrylates are as strong as epoxy, tough, and the bond reliability is quite high. I feel confident in recommending the latter two for secondary structures. I define as secondary as more lightly loaded nose ribs and control surface ribs, where the geometry of the assembly doesn't present an opportunity for peel. Also skins that are not highly stressed but where laminar flow could be maintained if the surface were smooth, i.e. not deformed by rivets.

Do you guys agree?
 

MalcolmW

Well-Known Member
Joined
Jan 21, 2007
Messages
118
Hello, Blakmax;

re surface preparation: Yes, chromic acid does require an elevated temperature (above 180 F), which puts it out of consideration for the average home builder. As an alternative, the use of alodine will provide a chromate conversion surface (combined with a distilled water rinse and strict cleanliness) for an acceptable substrate.

If methacrylate adhesives are used, then a milder (but still clean) surface preparation using a phosphate based etch, again with a clean water rinse should provide an acceptable bonding surface.

In regards to the epoxy-phenolic adhesives, well, there is a 'trick' or proprietary chemistry which avoids the formation of water in the polymerization (it is a cross-linking reaction). As a result, the adhesive can be 'B-staged' and spread on glass scrim cloth which is passed through chilled rollers to slow the chemical cure and provide controlled thickness. This adhesive also has very good high temperature performance characteristics. Once again, this process is beyond the capabilities of the average home builder.

IF, and this is a big IF, a home builder wants to employ epoxy adhesives for aluminum bonding, I believe the most important factor is substrate preparation, followed by accurate assembly, and lastly by epoxy formulation (thixotropic agents, fillers and curing agents). However, this is not easy using hand tools and a simple work environment.

Thank you for your comments and insight on adhesive bonding - which obviously come from long experience.

Always do some testing and fly safe,

MalcolmW

To fasee1819 - some of your questions have been answered previously in this thread (affect of temperature, lap joint design, surface preparation, etc.). Do a search and you should have a good start for your project.
 

Michealvalentinsmith

Well-Known Member
Joined
Oct 13, 2009
Messages
361
Location
Rainbow beach QLD Australia
As a point of interest, methacrylates are used in dentures - the mouth is one of the most corrosive environments imaginable - and in cementing hip and knee replacement joints.

The bone to metal junctions using methacrylate have been known to hold up for as many as 20 years and the loads though a hip joint over 20 years are enormous. It has been replaced more recently by uncemented joint replacements where the bone actually grows into the titanium for a superior junction over time - but until recently methacrylate was the best available cementing material.

There are newer composite materials available now in dentistry that have superior metal bonding and may be worth investigation - though I doubt such would be available for use in aviation and the quantities required would be extremely cost prohibitive.

From what I have seen of methacrylate I'd have no problem trusting it over prolonged periods as a replacement for rivets.

If there was a failure due to corrosion it would be due to the aluminum under the methacrylate oxidizing - from what I've seen methacrylates are good at preventing this.
 
Top