Adhesive replacing rivets or bolts

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

12notes

Well-Known Member
Lifetime Supporter
Log Member
Joined
Aug 27, 2014
Messages
1,208
Location
Louisville, KY
I'm doing initial calculations on a "black aluminum" approach to my turtledeck and canopy, and have made some approximate calculations for replacing rivets and/or bolts with adhesive. This is probably completely unnecessary, and I would be better off just using the same number of rivets on the composite, but now it will bug me if I don't know.

What I've read, and I can't help but feel is incomplete, indicates that, given even distributions of the n rivets/bolts, if the tensile, shear, and peel strengths of the n * rivets and n* bolts equals the same respective strengths X bond area of the adhesive, then the bond should be equivalent. Is this accurate, or at least good enough to be within, say 10%? Are there other factors that need to be taken into account in the calculations?

I'm also assuming the adhesive is appropriate for the materials, conditions, and exposure of the joint. 3M DP420 was what I had in mind specifically, but I'm really curious about this in a general sense rather than specific calculations based on that. I'm not asking anyone to do my math homework for me.

I'm only looking for a good approximation, so factors which change the strength 1% ish, while they are a welcome addition to the post, are essentially statistical noise in this phase of my design. This is a first pass of answering "Can it be done with a FOS of 3 with an acceptable weight gain?" rather than "What's the optimum way to do it for a FOS of 1.5?" If it passes this sniff test, then I'll move to the second question. The turtledeck redesign is mainly for looks and to add a lip in front to easier seal the canopy opening - I do not expect any weight savings or strength advantages here. Ultimately, I'm more likely to use rivets, but am interested in this as an alternative.

The other item I couldn't easily find information on how the differing rates of thermal expansion of different materials affect the whole, specifically an adhesive joint as opposed to one that is riveted/bolted. Let's say I put a carbon fiber turtledeck on an aluminum skin fuselage. Would the ever cause a noticeable problem on a small plane, or is this not a significant factor? The number of planes that seem to just slap one on would indicate it's not significant, but maybe there's way more math than it appears behind some of these modifications, or there's an internal hat section longeron on the bottom I can't see. The plane I'm building is only 13 feet long, turtle deck about 4', my gut says that it wouldn't move enough to matter, but actual experience beats my gut every time.

Would an adhesive be better or worse for thermal expansion than a line of rivets? It seems that the rivets would concentrate the load on their shafts and try to open their own holes up, but the adhesive would distribute the load better, especially if it has a little bit of flex in it.

EDIT: I forgot to mention that a layer of kevlar/fiberglass would be at the bottom of the carbon to prevent galvanic corrosion.
 
Last edited:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,050
Location
Saline Michigan
Bonding graphite fiber parts to aluminum parts is not done much, and for good reasons.

Graphite fiber in contact with aluminum or magnesium alloy makes for galvanic corrosion. A layer of 2 oz glass cloth laminated on in the contact region will electrically isolate graphite from aluminum... at least until the glass layer wears out.

Bonding to aluminum is generally not done in homebuilts because the processes to make the aluminum clean enough, keep it that way until bonded, and then prevent corrosion along the edges of the bond lines are usually beyond what can be managed in a home shop.

If you can achieve good durable bonds, you should check that the live loads plus loads due to thermal expansion do not use up your factor of safety in the bonds and part. Graphite laminates have a small negative thermal expansion along the length of the fibers, while they have somewhat larger thermal expansion than aluminum across the fibers.

Billski
 

12notes

Well-Known Member
Lifetime Supporter
Log Member
Joined
Aug 27, 2014
Messages
1,208
Location
Louisville, KY
Oops, sorry, I forgot to add that the bottom layer was planned to be kevlar/fiberglass to avoid the galvanic corrosion issue. Also, if the adhesive is non-conductive, wouldn't that add a barrier to the galvanic corrosion as well.

What do you mean by the fiberglass layer "wearing out"? Wouldn't that require some sort of movement in the joint that would be damaging to any material? Maybe I'm just not understanding how you're using the phrase. Do the electrical properties change over time?
 
Last edited:

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
1,706
Location
Pocahontas MS
1st Q: is it structural? If it's not, then you have the option to protect the aluminum and pretend that the turtle deck is just another canopy. Then stuff like Sikaflex becomes an option.

If it *is* structural, then you should talk to an aero engineer who's not the ghost of Jim Bede. ;-)
 

12notes

Well-Known Member
Lifetime Supporter
Log Member
Joined
Aug 27, 2014
Messages
1,208
Location
Louisville, KY
The particular adhesive I was looking at does have a complicated preparation technique for maximum shear/peel strength on aluminum, but also lists MEK/abrade/MEK as an acceptable method with lower, but still considerable strength. So application would be possible in a home shop, if not advisable for other reasons.

Still wondering about my main question, though - is the calculation for replacing rivets and or bolts with adhesive as simple for a ballpark estimate as I described above, or are there other large factors at play?
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,050
Location
Saline Michigan
Oops, sorry, I forgot to add that the bottom layer was planned to be kevlar/fiberglass to avoid the galvanic corrosion issue. Also, if the adhesive is non-conductive, wouldn't that add a barrier to the galvanic corrosion as well.

What do you mean by the fiberglass layer "wearing out"? Wouldn't that require some sort of movement in the joint that would be damaging to any material? Maybe I'm just not understanding how you're using the phrase. Do the electrical properties change over time?
Several issues. First is that bonding composites to aluminum alloys has been a nice idea that does not work for very long. The composite-aluminum bonds tend to fail over time for several reasons I delineated in an earlier post. If you go ahead with bonding, you should included backup mechanical fasteners. If you are building with mechanical fasteners, why put in the weight and effort for bonding? Once the thing is mechanically connected together, well, composites are not great for mechanical connection - bearing strengths of composites are low and so you end up needing to locally thicken and/or add bushings to make them durable. For lightness, composite designs avoid hardpoints.

Now if you say "my turtledeck is not structural" I have to ask you what you are attaching it to. The fuselage tub will bend in response to loads from the vertical and horizontal tails, and the tub will change dimensions when temperatures change. The turtle deck either has to be much softer than the tub under these loads so that it moves with the tub, or it has to float on the tub.

Moving with the tub means mechanical strength at all of the connections and locally "improved" load carrying to stand the preloads (from mechanical connections) and live loads in use.

Floating the turtledeck means that lifting loads (moving air outside, stationary air inside, the turtledeck is inflated by these) are still carried, but that bending, twisting, or thermal expansion loads are largely avoided. Now the surfaces are moving relative to each other. These surfaces can either have a renewable bearing surface or they will wear on each other. This is also the case with most attempts to bond aluminum to composites - they come loose and now you have either the floating or the firmly attached case... Take your pick.

As I pointed out earlier, if you can get a durable bond, great. The field experience, however, is poor with this option.

If I were faced with this, I suspect that I would just make the turtledeck out of the same stuff I made the tub out of...

Billski
 
Last edited:

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
15,000
Location
Memphis, TN
The data sheet I saw had various holding strengths for different materials. Some were quite low. Lots of time you see something bonded it will have a rivet at the ends. It’s to keep it from all peeling apart. Overall strength is ok as long as its stays. You can’t let it start to peel so they add a mechanical fastener. Before I owned my airplane, the fiberglass turtle deck was bonded to the steel with kitty hair mashed around the tubes. The plane got damaged and was taken apart to find that. The plane was only a year old but the stuff came apart easy enough that the covering was really what was holding it all together.

If it’s not structural I would see if the VHB tape would do the job, but any glueing or taping will need surface area and I would want x2.
 

Map

Well-Known Member
Joined
Dec 29, 2020
Messages
72
Location
California
If in doubt about bonding aluminum to composite you can always make some small test pieces and pull them apart to see how strong the bond is with your process. But you might find out that the data will vary too much to be reliable.

I agree with all the comments above that bonding aluminum is not a good idea for a homebuilder. I worked on an airplane (production) project that had bonded aluminum wing skins. The processes for getting the required bonding strengths repeatedly were way beyond what you could do in your garage.
I would say use rivets to join the parts (consider the poor bearing strength of a laminate) and use adhesive as sealant only.
 

12notes

Well-Known Member
Lifetime Supporter
Log Member
Joined
Aug 27, 2014
Messages
1,208
Location
Louisville, KY
The MEK-abrade-MEK prepped shear strength on aluminum was still 3500psi for DP420NS, since the flange area is 1/4 wide and there's 2" rivet spacing, this seems like it would be the equivalent of 1750 psi rivets, well beyond what the 3/32" pop rivets currently do.


So it would seem that if you ignore the construction of the parts and just look at just the thermal expansion differences and attachment issues, it would be an easier task to replace the wing or tail surfaces on an aluminum plane with composite, with their nice defined straight line bolting on points at the spars, than it would be to replace the turtledeck, with it's top centered curve on a cone composite to aluminum attachment.

There are many cases where, lacking experience, "new method" and "bad idea" can look identical. This is falling under the bad idea category.

Because of all the other reasons mentioned here, my plans have been suitably modified from "thinking about, but probably not using adhesives on carbon fiber turtledeck" to "thinking about, but probably not using adhesive as a test on a tiny carbon fiber tailspring fairing, where it won't hit anything if it falls off."

Don't want to kill this discussion, as there's been a lot of good information, so please keep it coming, both for me and anyone else who "innovates" this idea again in the future.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,050
Location
Saline Michigan
The MEK-abrade-MEK prepped shear strength on aluminum was still 3500psi for DP420NS, since the flange area is 1/4 wide and there's 2" rivet spacing, this seems like it would be the equivalent of 1750 psi rivets, well beyond what the 3/32" pop rivets currently do.
The problem is not how the bonded joint does when new, the problem is how the joint ages. Aluminum with a bond line and with atmospheric oxygen and water present will oxidize along the bond line. Everyplace there was two aluminum atoms becomes two aluminum atoms plus three oxygen atoms, taking up much more volume than the original did, which wedges apart the bond line, allowing atmospheric moisture and oxygen to get to new metal, where it oxidizes, and the joint peels itself apart over time. Aluminum is really good at this.
 

Rhino

Well-Known Member
Lifetime Supporter
Joined
Apr 8, 2004
Messages
1,218
Location
KTHA
Why are you even reading Aircraft Design/Aerodynamics/New Technology if that's your attitude?
It was a joke. Maybe you shouldn't lecture people about attitude.
 

BoKu

Pundit
HBA Supporter
Joined
Aug 15, 2013
Messages
2,931
Location
Western US
This is how we attach our CElipse wingtips to an L39 jet. The aluminum blades fit into the slots for the tip tank attachment. The blue stuff is EA9360 high-temperature high-strength epoxy, about $200/quart. The babyshit yellow is a special Boeing system primer we bake on after an acid etch treatment. The gold things are NAS4703 screws plus AN365-1032 nuts and AN960-10L washers 1" OC (24 per blade, 48 per wing, 96 per shipset.) Not shown is the final ply of fiberglass that insulates the carbon from the aluminum. This was all detailed and blessed by my aero structures guy. He figures that under idea conditions either the screws or the glue is enough. At 6g each tip reacts 3000 lbs of lift into the jet. One set has been tested to 8.5g.

20210424_121907.jpg
 

terke

Member
Joined
Dec 4, 2009
Messages
13
Location
Calgary, AB, Canada
'Structural' is a subjective term and changes with every application. To a folded paper airplane Scotch tape is structural. To a Zenith CH 180, capable of pulling +/-8 G, Avex pop rivets are structural, to an RV-7 .125" solid aluminum rivets are structural, and to the Empirestate building 1.5" steel rivets are structural. All of them are appropriate to their various foreseen load conditions.
What you need to do is define what load conditions you are going to see. Billski has pointed out some good points to keep in mind, so have others. Then find an adhesive that would be able to deal with all of those loads. When your affixing method is sufficiently strong to deal with all foreseen scenarios it will be structural for that application.
 

Rhino

Well-Known Member
Lifetime Supporter
Joined
Apr 8, 2004
Messages
1,218
Location
KTHA
At some point you'll also need to determine what happens to adhesives over time, particularly when exposed to environmental conditions. Many adhesives don't hold up well over time. I'm not sure how'd you'd test that without leaving different examples out in the weather for several years. I imagine there's a better way to test that, but developmental testing is not my forte.
 

dog

Well-Known Member
Joined
Dec 29, 2019
Messages
578
Carbon driveshafts have aluminum ends.
Cant think of a more demanding use with
higher expectations.
Not exactly a secret as there are many companys bonding aluminum to carbon,
and adhesives manufacturers brag about there
thousands of formulations for any conceavable
use and there teams of enginers.
Gona be a u tube of buddy whats his name getting results.Yup.just checked.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,050
Location
Saline Michigan
Carbon driveshafts have aluminum ends.
Cant think of a more demanding use with
higher expectations.
Not exactly a secret as there are many companys bonding aluminum to carbon,
and adhesives manufacturers brag about there
thousands of formulations for any conceavable
use and there teams of enginers.
Gona be a u tube of buddy whats his name getting results.Yup.just checked.
Factory processes and materials under better control than we can manage in home shops.
 
Top