Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by GESchwarz, Mar 19, 2009.
That's very interesting. Do you have any supporting data that you can refer us to.
Hello, MichaelValentineSmith & Gary S.;
One more trip down the methacrylate (acrylic) adhesive lane - yes, methacrylate adhesives have an excellent reputation for holding up under severe environmental operating conditions, specifically, aluminum truck bodies (assembled using methacrylate adhesives), particularly delivery vehicles.
From the work done by Gary S., and from my experience, I believe the methacrylate adhesives offer a good combination of performance and ease of use. The key, as with any adhesive usage, is good substrate preparation. With methacrylate adhesives, there may be a relaxation of the degree of preparation. However, if I planned to fly an aircraft assembled using adhesives, I would be doubly-**** careful with my substrate cleaning and etching (dealt with in earlier threads).
The methacrylate adhesives are not expensive, and are widely used in aluminum truck body assembly (see: Replacing rivets: Vehicle body manufacturer switches to adhesives ). Aluminum truck bodies undergo service conditions that aircraft are unlikely to experience (operation in all weather condition - from sub-zero to Arizona summers, plus salt spray from de-icing compounds), with minimum maintenance (washed when the advertising on the side becomes unreadable), and constant impact from cargo loading and unloading. And the truck bodies stay together. And corrosion isn't a problem.
So, I don't believe that expensive specialty adhesives are required for home-builders who wish to assemble an aircraft with adhesives. Go with proven products, learn how to clean and chromate-convert aluminum, practice application prior to commencing, and always do some testing.
As always, fly safe,
Here's one product designed to work with Methacrylates and from what i heard so far, it seems very easy to use. I haven't had the chance to try it yet myself but from other accounts it works as advertised:
http://gluguru.com/GlueDotsMSDS/2006GluGuru/Hysol Data Sheets/HPrimer2000.htm
Awesome thread, thanks to all contributors!
Sorry guys, I have been offline with some family medical issues for the last few months.
Orion, I am a bit wary of a "wipe-on" primer for any adhesive. To be effective a primer must come in contact with a chemically active surface so it can react. I note that the proponent claims that the primer contains an etchant and to some extent that may provide the active surface. I would be a lot more comfortable if the surface was solvent degreased and lightly grit blasted before the primer was applied. I strongly recommend that any user undertakes a wedge test (ASTM D3762) but please ignore what the standard says is an acceptable performance. We found that any process which gives a growth rate of less than 0.2 inches in 24 hrs and less than 0.25 inches in 48 hrs exposure gives excellent durability.
Sorry guys I've been out of dentistry now for over 30 years and don't have any real useful information. Even cemented hip replacements are old hat now and osteo-intergration with bone growing into titanium oxides is the current practice.
I think they may still do cemented hip replacements in special cases but it's old technology. The meth cements are still being used in vertebroplasty proceedures however where the glue is injected into crushed vertebrae. If I recall there can be a problem with heat generated as it catalyzes.
I was under the impression it was a successful technique but recent studies seem to show it's not much better than placebo for pain. For our purposes though the stuff bonds to the bone very well and lasts a lifetime.
I'm interested in the GLARE techniques for reinforcing aluminum tube with bonded S glass or CF from the inside. It seems that strength to weight ratios are approaching carbon fiber with much less cost.
I am new to this group and came up on it by googling aluminum bonding. Hopefully someone will read this although the thread is somewhat old. This discussion has been excellent as a friend and I are both building T-51s near each other and are very concerned about the kit adhesive , uralane 5776. Here in south Florida the working time is extremely limited because of heat and humidity, and with the requirement to mix smaller batches as you apply the uralane to the structure and skin overlaps, the large skin sections, 8 feet long on the fuselage and wing outer panels, the application process runs into a time limit. As a result we have been searching for alternative adhesives with a combination of ready to use self mixing tubes and longer working times. We are currently looking at a Lord product, Fusor 112 as a potential replacement, which is much stronger than the uralane, but we have been told of that the bond is so strong that should future repair be required, the process to separate and replace skin sections could actually ruin the other overlap skin as well. Fusor seems to be an acrylic and Lord's tech department says that heating the bonds up can allow separation.
Our question is this, has anyone attempted to separate these test coupons, or any test overlaps of any kind with acrylics to see if it can actually be done without destruction of the joint metals? It seems to me that this would be a requisite for any adhesive used, since hangar rash or worse incidents could necessitate repair without completely rebuilding the plane. The uralane by the way seems agreeable to heating and putty knife separation. Thanks for any info you may pass on.
Just a couple of comments based on mainly experience from 1974 on the prototype Schreder HP 18 sailplane --bonded metal wing skins and testing on dozens of urethane epoxy and other adhesives (the wesbsite "Soaridaho" has an archive and tech section that has some info on metal bonding in this context ) -- we found that the original used adhesive Scotch Weld 2216 (grey goo) was very good in all the tests using T peel coupons and bonded PVC foam and aluminum cubes holding water or avgas BUT the SLIGHTEST peel stresses resulted in it letting go one molecule at a time in the presence of any water . this was found after building the HP 16 sailplane with integral water ballast tanks and finding that the skins detached virtually overnight -- 3M sent a planeload of engineers down to observe and check on surface preparation , mixing etc etc and went away astonished --the material was subsequently withdrawn and no longer used by Boeing even .
We found the best bonding material to be a Hysol epoxy adhesive (can't recall the exact number but it is on SoarIdaho ) that is also used by Lockheed and formerly General Dynamics and I used when bonding fatigue test coupons for F 18 test work (and got the approval under Boeing spec for surface preparation procedure -- DO NOT USE acidic or alkaline etchs prior to bonding because of the creation of "smut' (google it) which destroys the surface energy and bond integrity --scotch brite scrubbing and ajax cleaner are OK --de ionized water and 23 steps plus volan treatment (silane coupling agent) feature in Boeing bonding process -- a water break test is also imperative .
Make a number of test pieces to put aside and test for durability over time and with adverse environmental conditions whatever adhesive you use.
Heat and a putty knife ought to do it for many adhesives. You're going to just have to experiment with the materials you are using.
BUT the SLIGHTEST peel stresses resulted in it letting go one molecule at a time in the presence of any water . this was found after building the HP 16 sailplane with integral water ballast tanks and finding that the skins detached virtually overnight -- 3M sent a planeload of engineers down to observe and check on surface preparation , mixing etc etc and went away astonished.
Had they not run any wedge tests on their adhesive? I would think that what you describe would have been discovered if they had.
Aluminum/aluminum bonding of primary structures is a very picky process if you hope for any long term results. I'd advise anyone contemplating aluminum bonding getting a copy of BAC 5555 or talk to a shop that is Boeing qualified to etch using their process to get an idea of what is involved.
Here is a link to a fairly good paper on the subject:
The standard wedge test is simple and one that just about any home builder can do in their garage. If your test coupons fail this test I'd not consider the method used to prepare them for a primary structure.
Side note: Years ago when I was working on my BD-5, and not knowing any better, I substituted some Hysol adhesive along with the standard Scotch Bright and acetone scrub for the recommended Proseal adhesive in a few places. It was a BIG mistake. The Hysol bonded parts could not be un-bonded for repair (the Proseal was relatively easy to repair). In hind sight if I'd just drilled the rivets and steamed the joints I probably could have removed the Hysol. Then I might have a nice BD-5 setting in the shop - still waiting for a suitable engine.
Thanks for the inputs above. I wondered about the comment on "steaming" the joints, were you saying that that worked on something later on, or that a heat gun would not work, only steaming? We just received our Fusor acrylic glue package and will be testing the repair aspect of it by trying to separate test sheets. Sam
Any adhesive that is easy to remove will be easy to fail in service -- NOT making it easy to pry the parts apart is the object (worry about repair later ) -any heat softening adhesive will also heat soften in service and probably suffer creep so best avoided for all those reasons.
Yes we conducted all the standard tests including the wedge test (also T peel, shock loading, impact , folding etc ) --the debonding of 3M 2216 was news to the manufacturer --years later I came across an engineer in one of our domestic airlines specifying 2216 for a repair scheme - literally as I, and my exlandlord who had worked for Boeing in Seattle new (Rick Ladley on the off chance anyone ever knew him ), walked past his drawing table --Rick knew the score on 2216 and got it 'fixed' on the spot.
The reference to "steaming" my old BD-5 Hysol joints relates to Aircar's observation about the de-bonding of the HP-18 in the presence of water. My untreated BD-5 aluminum/Hysol joints would probably have de-bonded after only a couple of hours in the presence of steam with no work needed from me. The Hysol softened with heat only but still stuck very well. The Proseal came apart rather easily with heat, and cleaned up completely with MEK and a plastic scraper.
A wedge test is basically 2 strips of aluminum bonded along part of their length and after cure a wedge is shoved in the un-bonded end a set distance. The sample is then placed in a test box with high humidity (for exact procedure see the link below). Bare aluminum joints, such as my BD-5 was, will completely de-bond in just a few hours. Properly etched and primed aluminum will show no, or only slight debonding of the sample.
ASTM D3762 - 03(2010) Standard Test Method for Adhesive Bonded Surface Durability of Aluminum (Wedge Test)
I just posted this (below) on the Titan Mustang chat site, thought it might be of interest to someone here. The hat section I mention is a factory formed aluminum extrusion that is basically a u-channel with a flange on each side. It fits around from the backside of each piece of the 1/2" steel tube frame of the fuselage, that is used to give some strength and serve as a jig to attach the aluminum skins, using a double overlap rivet pattern that is also attached to the flanges on each side of the tube, thereby contributing to a monocoque construction technique as well. Original glue supplied with the kit is Uralane, an expensive mix that is time consuming to prepare and has a short working time in high humidity, which is the norm here in south Florida. The company is now looking into Fusor as well. Incidentally, the recommended prep for aluminum is strictly to remove impurities using only isopropyl alchohol, no scuffing required. That was what I did just to verify the bonding effect. Fusor is prepackaged in double tubes, ready to use in the mixing gun. Gun was about $40 online, 7.6oz mix was about the same. Much cheaper online than what I found in town.
Just thought I would relay some info on an initial test done using Fusor
> 112 as a bonding agent for skins. I found out about this product while
> talking to John Hall, who has used this for a few years, and thought it
> should be looked at. I put together a hat section on the fuselage, match
> drilled two pieces of .025 with 8 rivet holes top and 8 bottom, then
> deburred and dimpled (3/32). I prepped the metal according to
> recommendations from Lord's company that makes this stuff, then applied it
> using the nifty double plunger gun that is used for the two-part
> applicator tubes. That was the neat part, getting this stuff mixed
> through the nozzle and applied with a bead was great, quick and easy.
> Spread the bead using a wood stick on both surfaces, including the hat
> section side, clecoed up and let it set. Tech data says cure is 8 hours
> at 70 F. 24 hours later we started a couple of simple tests. Squeezed
> 3/32 flush rivets into the top holes, took
> the set fine. Drilled out the rivets. First, we could not pull the
> outer skin off by pealing. It started to separate at the edge, then the
> metal tore and pealing stopped. Next, after first testing the heat
> setting on the heat gun using an IR sensor, which showed about 250
> degrees F when applied from about 2 inches onto a black powder coated
> piece of steel 1/16" plate, started heating the aluminum while applying a
> peal load with pliers. When the metal started separating we stopped, let
> the metal cool, and again tried to separate in peal, with no luck.
> Finished heating and pealing, looked at the metal and none of the glue
> had debonded from the metal, the separation was strictly within the glue
> itself. With the metal apart, more heat applied to it while scraping
> with a putty knife took much of the glue off, finished cleaning up with
> acetone just fine. Checked a half inch wide thick piece of the glue left
> over on the wooden stick, it
> remained flexible like hard rubber, similar to uralane. The question we
> had was could this glue, which is stronger than uralane, be separated in
> case repairs had to be made, and it looks like not a problem. The tech
> data on this glue can be found on the Lord's site. I plan on checking
> with Titan and continue a few more tests before making a decision to use
> it, but the working time of 70 minutes at 70 degrees, 40 at 90 degrees,
> and having an applicator so easy to use is a big plus, and it's
> manufactured to assemble auto body panels without rivets or bolts and
> provide corrosion proofing at the same time. Storage time is 18 months.
> I know this wasn't as good as a laboratory test but so far it's
I beg to differ. My Moni was built by a very fine craftsman and was done with almost fanatic precision. Eventually, one wing's bonding failed. I was looking for it because there had been some catastrophic, fatal failures. When we took the wings apart, the failed wing took no effort but the other wing was a chisel & hammer job. The only possible difference between them was the relative humidity on the different days they were bonded.
Monnett changed over to riveted wings at least in part because of the failures.
I don't want to be critical, but it's a dangerous thing to say that in the case of the Moni they did not fall out of the sky because some DID.
Thanks for the links to the thesis on bonding Hotwings --still have to wade through it but you always find something new . The
phosphoric oro acid prep was one of the approved chemical treatments from memory but just using an alkali etch is bad news from the electrochemistry . There are certainly a few surprises to be had in bonding --even in composite bonding between high density autoclaved materials when bonded under only contact pressure (they seem to be too non porous to get a good grip if I can put it that way --I saw one meticulously prepared 'conical' splice repair simply pop off under test at the DSTO defence science and technology organization here in Australia -- between autoclaved high modulus carbon laminates -- in a tensile test rig under carefully monitored strain gauged conditions . The failure occurred nowhere near the anticipated loading --in fact during calibration and to stunned silence.
There have been no serious debonds in HP sailplanes using Hysol as far as I know but a few bond failures from ground loops and dropped wings are recorded -- I repaired a set of wings that had been involved in a pretty severe accident caused by a take off with a disconnected ruddervator --enough G to tear off mass balances and shear some rivets-- I think the flexibility of the PVC foam ribs and diaprhrams is a good thing in 'crackstopping' bond lines to some extent also. (and reducing stress raisers ) Your "Icarus" logo seems appropriate on this subject too --'heat softening of wing bonding leading to catastrophic failure' --the first aircraft accident investigation analysis ?
A proper bond should not depend on porosity or surface roughness since bonding is not a mechanical action - it is a molecular one, similar in fashion to bonding to a sheet of glass. Often times porosity or a sanded surface may actually degrade bond strength due to the voids and stress concentrations caused by the surface discontinuities.
If an autoclaved part refused to bond I would first closely examine the materials used in the fabrication - for instance, about five years ago the Boeing Co. pulled the certification status off of about 90% of the market's peel-ply materials. Turns out some of the fiber treatment (that made it a good peel-ply) was coming off the fibers and depositing itself on the laminate, resulting in very poor bonds.
Certain "permanent" parting compounds also may come off the tool and contaminate the laminate surface.
Then, we have also found resins that under certain conditions may not achieve a proper cure, even at high temperature, leaving a somewhat "loose" surface. these are virtually impossible to bond to.
Any debond needs some form of verification to determine cause. Usually it's a function of surface contamination or material unsuitability.
I find this interesting, and a reason for concern. Do you have any more insight on just what types of resin and/or conditions to watch for this phenomena? Is the lack of surface cure due to some kind of chemical equilibrium reaction with the atmosphere or some other mechanism? Is the uncured surface a problem with secondary bonding using the same type resin as used for the lamination or does it only manifest if using a different bonding resin?
We've probably all seen this kind of problem when working with polyester resin but I've not yet run across it with aircraft resin systems, or at least not recognized it when I did.
I chose the Icarus logo for this group after much thought. It just seems appropriate since I like to try to push things just a bit further than the norm. It reminds me what can happen when you do. This hangs on the fridge. Just printed a new one for the office.
The phenomenon has been discussed at Boeing some years back but outside of a couple of poorly cured samples shown during the presentation, I haven't seen a practical example until about a year ago. We received our parts from being cooked and upon demold everything looked fine. The surfaces came out of the mold clean and shiny with no evidence of anything wrong.
We then went to clean the parts prior to the secondary bond (this was a corrugation core to one of our wings) - this process includes an Acetone wipe, which is followed by a light sand (120 grit) to eliminate the tooling sheen on the surface. This is then wiped again with Acetone until the rag comes off clean. The problem though was that the rag was not coming off clean - it kept getting black and the harder we scrubbed the blacker it got. Looking close it was clear that we were into the fibers and it looked like much of the resin was being removed in the wiping process. The interesting part was that this was happening only on one section of the part - the remaining corrugation surfaces being prepped were fine. Also, this was a problem only on the bond surface (which was in the tool trough), not on the adjacent webs.
We never did figure out what went wrong - this never happened before or since. The only thing we can think of is that either the resin was not properly mixed in the prepreg machine or the tool was somehow contaminated in that area, which prevented the resin from achieving a full cure. To check things out we removed a small section of the surface and did a separate bond in order to test the integrity. This is not a highly stressed area so bond tolerances are pretty loose. Turns out the secondary bond was very good so we did end up using the part. But just to make sure, we reinforced the bond over that section with a series of rivets, just in case.
Thanks for the info. I'll tuck it in the back of my mind just in case I run across something similar. Your supposition of a mix problem sounds reasonable. You just never know when someone will let a tank run dry. :emb:
Separate names with a comma.