Conclusions on Aluminum Adhesive Bonding Tests

[MadRocketScientist]

That doesn't surprise me at all, I think there is a very nasty corrosion tendency when those two materials are touching.

Shannon.

 

[GESchwarz]

Well, I’ve finally had the chance to get around to doing the testing on the bonding of PVC foam to aluminum. I am comfortable in recommending the use of foam bonded to aluminum in applications where stress concentrations are known to be low, or otherwise in areas adjacent or away from major load paths. I would call these secondary structure, such as control surface ribs. Fin, stabilizer and wing ribs may be made of foam or have foam as the rib caps only after an equivalency of strength has been determined between the foam and the aluminum structure it is replacing.

I have completed some 45 equivalency pull tests between Divinycell H100 foam and standard riveted aluminum technology. The rivets used were all flush –3 and –4’s. The material included .016, .020, .025, and .032. All of the aluminum was 6061 T6 as opposed to 2024 T3 because the 6061 poses less of a long-term risk of corrosion, which is detrimental to adhesive bond joints.

The foam test sample as shown in the photographs had a trapezoidal cross section; 0.9” wide at the base and 0.4” across the top, with a height of 0.5”. This shape would be used as a rib cap, bonded between a rib web and the skin. With the 0.9” face mating to the inside of the skin, and the rib cap bonded inside a slot cut 0.4” deep into the center of the 0.4” wide face.

Five foam samples of this same cross section, but of different lengths, were pitted against various combinations of aluminum thicknesses and various numbers of –3 flush rivets. All holes were dimpled. A total of 45 pull tests were performed.

The result of these test showed that this foam/adhesive joint was consistently stronger than a rivet spacing between 1.2” to 1.5”, but with 3 of the 5 samples besting a rivet spacing of 1.34”. In all cases the rivets tore through the aluminum as opposed to the rivets shearing.

Proseal and Methacrylate were used as adhesives on these joints. In all cases the failures occurred in the foam, not the adhesive, which is good, because with so much concern about the quality of the bond joints, this failure mode is highly consistent and therefore can be used to make reliable engineering decisions about rib spacing and equivalency to conventional riveted construction.

Earlier lap shear strength tests of adhesive joints between two pieces of .040” and greater aluminum, demonstrated strength far greater that what this foam is capable of, shearing multiple –4 universal rivets.

It’s up to the designer to determine what schedule of foam, bond joint width, and rib spacing is of equal or greater strength to conventional riveted construction.

It is debatable as to which way is quicker to build with, as the bonding process does require it’s own special processes and setups which include some form of clamp and hold duration. The one clear benefit with bonded foam is the nice, smooth aerodynamic skin surface that is attainable.

 

[BBerson]

There is an interesting discussion about aluminum bond failure going on now in this helicopter forum: AW139 lost tail taxying DOH - PPRuNe Forums

A tailboom fell off an AW139 helicopter while taxiing, no one hurt. Major damage however, thought to be caused by aluminum bond failure.

You can read the entire rather long thread, or just the comments from Blakmax, who claims to be a retired aircraft bonding and failure analysis expert.

 

[blakmax]

Guys

I am the "blakmax" referred to in the last posting. I have signed in because I was alerted to your forum. Please read the AW139 reference above, because the test methods you are using give you a false sense of security which is unwarranted. Please do NOT ever rely on just an abrade and solvent clean for surface preparation of metals, because I can absolutely guarantee that if you use just a sovent clean and abrade, your bonds will fail over time. Rather than repeat myself, please read the link and understand the mechanisms of adhesion and bond degradation and failure outlined there.

I am prepared to revisit this forum to help as I did in PPRuNE. Someone has to stop people from bad bonding practices. Please comment after you have read that thread.

Regards

blakmax

 

[bmcj]

Welcome and thank you Blakmax. The original PPRuNe thread seems to have been greatly diluted by the other posts. For the benefit of all here, would it be a great deal of trouble for you to copy the highlights and important points of your posts there and paste them in here as a whole?

 

[Topaz]

Blakmax, welcome to the forum. We're happy to have your expertise here! Adhesive bonding comes up on this forum with some regularity, and your input would be very welcome. I've read through the PPRuNe thread (through page nine, at least) and you raise some interesting points.

(For others inclined to do the same, Blakmax's comments start in earnest on page five of that thread. Prior to that the discussion is pretty much exclusively pilots speculating on the accident and the actual suitability of composites for aircraft construction which, while always entertaining, isn't very enlightening.)

One of your earlier comments raised an eyebrow for me, where you state that peel plies are not a sufficient means of ensuring a reliable bond between composite panels. Having looked through your responses and your website, I can see that you have quite a serious level of experience with the subject matter, but I've got a few thousand VariEZ's, LongEZ's, Cozy's, and other similarly built aircraft out there flying safely that say differently. Disbonding hasn't ever been a problem in those structures in over thirty years of service, and they use peel ply extensively to prepare composite panels for bonding to one another. If I'm not mistaken, peel ply is also used fairly extensively in composite sailplane manufacture, which industry has an even longer record of service.

Have I misread your statement or misunderstood it? If peel ply also isn't suitable for composite-to-composite bonds in aircraft such as we discuss here, what would be a better, still homebuilder-friendly method?

 

[bmcj]

Hi Topaz. I've heard that some peel-ply has some sort of coating which remains on the structure after the peel-ply is removed and causes some issues with bonding. This is only the case with some peel-ply, but not all. I forget what the treatment/sizing/contaminant is and I forget which brands have it and which don't, but it something that needs to be addressed if using peel-ply.

 

[Topaz]

Hi Topaz. I've heard that some peel-ply has some sort of coating which remains on the structure after the peel-ply is removed and causes some issues with bonding. This is only the case with some peel-ply, but not all. I forget what the treatment/sizing/contaminant is and I forget which brands have it and which don't, but it something that needs to be addressed if using peel-ply.

Yes, I've read that as well, which is why all of Rutan's A/C plans specify a particular type of dacron tape for use as peel-ply.

Still, if you read Blakmax's post on that other thread, his basic problem is not one of contamination, but of not producing a "chemically active" surface for bonding. I'd like to hear more about that.

Various hand-layup composite homebuilt aircraft have been flying safely for literally decades with peel-ply as a integral part of their construction process. I'm not doubting Blakmax's credentials or opinions at all, but there's something amiss here - those airplanes have a very significant service life that says their construction processes are adequate to the task. A blanket statement that says otherwise is a contradiction that I'd like to understand more fully, especially since I will be using Rutan-style construction methods almost certainly on any airplane I build - including the use of peel-ply.

 

[Richard Schubert]

There was a presentation at Oshkosh 2005 on bonding that covered some of these issues including cautions on peel ply:


http://www.cozybuilders.org/Oshkosh_Presentations/AbarisCompositeBondingOshkosh2005.pdf

My question for Blakmax is, are there any preparation methods for aluminum bonding that might be suitable for use by homebuilders?

 

[MadRocketScientist]

Welcome and thank you Blakmax. The original PPRuNe thread seems to have been greatly diluted by the other posts. For the benefit of all here, would it be a great deal of trouble for you to copy the highlights and important points of your posts there and paste them in here as a whole?

I second that, I am very interested in the bonding of aluminium as the CriCri I am building uses adhesives to hold the wing and tail skins on. That is something I don't want to get wrong. I was under the impression that you always needed to prime the aluminium or at least with the adhesive systems commonly used on the CriCri.

Shannon.

 

[blakmax]

With reference to the peel ply issue. We found out about this when we could not get strain gauges to stick to a composite specimen. Examination under SEM found extensive contamination with silicone. We then examined composite patches made with the same peel ply and found the bonding surface contaminated again. We ripped off the first ply and proved that the silicone was not part of the resin system, so it has to come from the peel ply. The details are in [FONT=&quot]Hart-Smith, L.J., [/FONT][FONT=&quot]Redmond[/FONT][FONT=&quot], G, DAVIS, M.J., The Curse of the Nylon Peel Ply, 41st. Int SAMPE Symp. and Exhib., Anaheim, 25-28 Mar 1996.[/FONT]

The problem I have with peel plies is that they are essentially based on the theory that you only need a clean surface to bond. I stress that it is essential that you get chemical reactions between the adhesive and the surface. A clean surfaces is a necessary but not sufficient condition for adhesion. The surface must be chemically active as well, so some form of abrasion is necessary. If the surface is contaminated with silicone, then abrasion simply spreads and embeds the contamination. I would use a heat-set uncoated peel ply and abrae the surface after removing the peel ply. Blow the dust off or dry wipe. Don't solvent clean after abrasion because that simply dissolves any residual contamination and spreads it over the surface. The best method for abrasion is a LIGHT grit blast but be careful to contain the dust and at all costs do not let the dust get into fuel systems because it will cause engine failures.

As for the years of experience, it may be that the bond stresses are so low that you get away with a weak bond. However, some where you may encounter higher loads and there is no guarantee that the weak bond will be adequate. I have personally examined the bond surfaces from a Lancair crash (fuel starvation not structural cause) and every bond surface exhibited interfacial failure with the pattern of the peel ply replicated on the surface of the adhesive.

With regard to preparation of aluminium, the best option for a low production environment would be Boegel which is sold as AC130 and follow the instructions.

Regards

blakmax

 

[Topaz]

...The problem I have with peel plies is that they are essentially based on the theory that you only need a clean surface to bond. I stress that it is essential that you get chemical reactions between the adhesive and the surface. A clean surfaces is a necessary but not sufficient condition for adhesion. The surface must be chemically active as well, so some form of abrasion is necessary. If the surface is contaminated with silicone, then abrasion simply spreads and embeds the contamination. I would use a heat-set uncoated peel ply and abrae the surface after removing the peel ply. Blow the dust off or dry wipe. Don't solvent clean after abrasion because that simply dissolves any residual contamination and spreads it over the surface. The best method for abrasion is a LIGHT grit blast but be careful to contain the dust and at all costs do not let the dust get into fuel systems because it will cause engine failures.

Assuming that the peel-ply tape used was one of the varieties that does not contaminate the bonding surface (they're reputed to exist, and the original Rutan Aircraft Factory [RAF] plans specify a particular tape for that reason), would a light grit blast or mechanical sanding of a peel-ply-prepared surface then be acceptable to promote a better bond? Given homebuilder manufacturing processes, building some of these hand-layup structures would be quite a bit more difficult without the use of peel-ply.

As for the years of experience, it may be that the bond stresses are so low that you get away with a weak bond. However, some where you may encounter higher loads and there is no guarantee that the weak bond will be adequate. I have personally examined the bond surfaces from a Lancair crash (fuel starvation not structural cause) and every bond surface exhibited interfacial failure with the pattern of the peel ply replicated on the surface of the adhesive.

I can't argue with your experience and observations, and yet these aircraft have an admirable structural safety record. Apparently, the bond joint areas were either designed with this in mind or it's a byproduct of other design factors. Neither Lancairs, nor any of the aircraft using RAF methods, seem to be falling from the sky with any regularity, particularly for structural reasons. In every case of which I'm aware (admittedly, a small sample) if there was a structural failure it was due either to overload or the builder incorrectly assembling the structure, with deviations from the plans and/or instructions.

Sounds like we can improve our processes for more certainty, though. There's always room for that.

 

[bmcj]

The problem I have with peel plies is that they are essentially based on the theory that you only need a clean surface to bond. I stress that it is essential that you get chemical reactions between the adhesive and the surface. A clean surfaces is a necessary but not sufficient condition for adhesion. The surface must be chemically active as well, so some form of abrasion is necessary. If the surface is contaminated with silicone, then abrasion simply spreads and embeds the contamination. I would use a heat-set uncoated peel ply and abrade the surface after removing the peel ply. Blow the dust off or dry wipe. Don't solvent clean after abrasion because that simply dissolves any residual contamination and spreads it over the surface. The best method for abrasion is a LIGHT grit blast but be careful to contain the dust and at all costs do not let the dust get into fuel systems because it will cause engine failures.

One of the links I followed indicated that one of the functions of peel-ply was to leave small jagged mounds of resin which acted as a roughened surface for the next layer to adhere to. Given that interpretation, wouldn't abrasion remove those mounds?

 

[orion]

With reference to the peel ply issue. We found out about this when we could not get strain gauges to stick to a composite specimen. Examination under SEM found extensive contamination with silicone. We then examined composite patches made with the same peel ply and found the bonding surface contaminated again. We ripped off the first ply and proved that the silicone was not part of the resin system, so it has to come from the peel ply.

Thank you for this bit of data!!!!

Some time ago we learned that the Boeing Company decertified a wide variety of peel ply products they were using due to left over contamination of the bond. We have not been able to confirm this but my supplier assures me that they carry one of only two or three that remained OK to use.

 

[Richard Schubert]

Originally Posted by blakmax
With regard to preparation of aluminium, the best option for a low production environment would be Boegel which is sold as AC130 and follow the instructions

Thanks Blakmax :ban:

They have it at Pacific Coast Composites!!

Pacific Coast Composites

 

[Topaz]

...but my supplier assures me that they carry one of only two or three that remained OK to use.

Would you happen to know the brand/product name for that? I'd like to compare it to what Aircraft Spruce is selling. Supposedly that's the material that was called out in the RAF plans, but a little comparison never hurts...

Thanks.

 

[blakmax]

bmcj,

One of the links I followed indicated that one of the functions of peel-ply was to leave small jagged mounds of resin which acted as a roughened surface for the next layer to adhere to. Given that interpretation, wouldn't abrasion remove those mounds?

I can confirm that there are small mounds which occur at the areas where the weave passes over and under each strand. The resin rich area there fractures leaving a rough surface which actually does bond because you expose a chemically active surface by fracturing the resin.

The improvement in bond strength provided by grit blasting has been known for many years. John Hart-Smith published results which clearly showed the improvements way back in 1984 in the Douglas Service magazine. I have an electronic copy if anyone wants to PM me.

Regards

blakmax

 

[GESchwarz]

Thank you Topaz and Blakmax for your contributions to this aluminum bonding thread.

As I’ve indicated a bit earlier, I found that because chemical surface prep brings with it additional process variables which can go wrong thereby actually causing a weaker bond rather than the durable bond that was desired.

So that’s why I've decided to use 6061 T6, which is of course less vulnerable to corrosion and perhaps less in need of an etch or anodize. For the purpose of getting back some or all of the strength lost by not using 2024 T3, I’m using a closer rivet spacing, wherever rivets are used. In my experience failure of aluminum structures typically occur at the fasteners, not at the material itself. Furthermore, and this goes for bonded joints too, failure is always preceeded by yielding which results in a tear at the rivet or peel at the adhesive bond.

Blakmax, I’d like to know more about what you know about surface prep and the real need for it on 6061 versus 2024.

It was a bit tough navigating through all the chaff in that PPRuNe thread. It would be nice have that filtered out. I really do admire the Englishman’s sense of humor but there comes a time to get on with the subject.

Regarding the discussion on bonding composites, if you folks think you have much more to discuss on that subject, I’m wondering if a different thread might be a better place for that.

Thank you,

 

[GESchwarz]

Blakmax,

I’ll definitely look into that Pacific Coast product. What is it about this product that makes it your choice for the homebuilder? Do you have any particular adhesive recommendations?


Oh and welcome to this site, I hope you stick around.

Another piece of wisdom I've picked up on this bonding subject is that good design with adhesives involves making oversized joints. Stay conservative. It's like any other joining method, including welding...all joints are stress risers, so the secret is to oversize them so as to spread that stress over a larger area. The goal is to make the joint stronger than the surrounding material.

 

[MadRocketScientist]

Does the AC-130 product have a best case adhesive to go with it or is it designed to be used with various other adhesives?

Shannon.