olnico

Hello guys.

I am about to start a series of experiment on glue joints resistance.
The idea is to try different types of glue on the same item and load it till the bond shears.
I though about using a block of aluminium, but after more reflection, I decided about using my new PYCABS servo bracket.
The reason behind that is that PYCABS is extremely glue friendly and that the servo bracket has got an open honeycomb bonding surface. This means that the material bonding capability does not enter into the equation, but only the glue joint strength.

The experiment protocol is the following:
1. Use one of my strongest workshop table and glue the bracket on the hard wood surface.
2. The surface is clean and has got a series of 10 tiny spherical holes to ensure that glue embeds into the wood.
3. The bracket is glued and maintained with a clamp at moderate pressure til the complete bond time specified by the manufacturer.
4. The bracket is loaded through a steel wire that holds fitness weights.
5. The test is conducted till rupture of the bond
6. The weight at which the bond is braking is divided by the bonding surface. this gives a shear strength in lbs/ sq.in

I will make a table with all kinds of glue, using always the same protocol.

Here is a list of the glues I intend to test:
1. Z-poxy 5 min
2. Hysol E-20HP
3. Hysol E-20NS
4. Hysol 9462
5. CA gel
6. Gorilla glue
7. Clear silicon
8. Wood vinylic glue

Jgwright

oli

You have missed what I reckon is the best glue for plastic parts, 2 part pink and green acrylic glue. It sets in record time but does get hot when setting and gives off a strong smell.

http://www.glueonline.co.uk/shop/ADL...___Green_.html

The only doubt is whether it glues as well to wood as it does metal and plastic. Might be interesting to try a fibreglass bond as well.

John

olnico

Here is the first test conducted with 5 min Z-poxy.

The hard wood surface is traced with the bracket shape and cleaned.



A diamond point rotary tool is used to imprint the wood.



10 dots are made in the wood with this tool:



The PYCABS servo bracket is saturated with glue:



The hard wood beam is saturated with glue as well:



The servo bracket is maintained in position for 16 hours with a pump clamp:



The servo bracket is then progressively loaded. 5 kgs:



10 kgs:



15 kgs:



20 kgs:



25 kgs ( sorry, the 5 kgs weight is on the other side of the load ):



The glue bond sheared at 25 kgs, a few seconds after I took the picture:



The glue took partly some wood away, and some glue sheared off the bracket:

olnico

This first experiment shows a shear resistance for 5 min Z-poxy of 25 kgs.
The bracket used is the Futaba standard size +.
http://www.rcuniverse.com/forum/rc-j...o-bracket.html

The surface of the bracket is 5.165 sq cm.



The shear strength of the glue joint in these conditions is 25kgs for 5.165 cm2 or 4,84 kgs/cm2 or 68.8 PSI.

olnico

I will keep on posting other measurements here.
When I have enough data, I will make a big table.
Any suggestion about modeller's glue to test is welcome here...

mr_matt

My Feedback: (10)

Great test Oli. How do you know if the wood broke first? In turn increasing the shear stress on the remaining glue joint?

olnico

I saw it braking. The glue joint gave up first from the top, then the wood came away a bit below. The picture of the broken bond is pretty descriptive I guess.
The bond broke just as I was starting the camera. I could not make a movie of this...




Bare in mind that this is 5 min epoxy. I was very impressed with the strength of that glue joint.
Actually loading this tiny bit of plastic with 25 kgs of steel is pretty scary. Mind your toes...

https://vimeo.com/104808816

dbsonic

My Feedback: (3)

Great test. It would also be interesting to see if 30 min is stronger than 5min epoxy(it is supposed to be).

Vettster

Very interesting thread... Though I think you could load up more weight if the grain of the wood was going the other way. Though all other test will be done in the same direction... My guess is your going to see a lot of them fail only because the wood let go first and not giving an accurate Shear load test for each glue. Perhaps using the block of aluminium would have been a true test of the glue, and not the wood..

Very good non the less and thanks for taking the time to do this.

mort78

My Feedback: (1)

Great test Oli,

That 5 min epoxy did hold up well, for sure. I agree with Vettster though, you might have issues with the wood giving away first when you use some glues that have more time to penetrate. But I guess these servo mounts will generally be bonded to a servo hatch which would be made of wood, so the test shows what would give up first under pressure! Perhaps you could try with your nice carbon servo covers?
But I guess we'll soon see

I have a run to Dubai coming up on the 9th Oli. I get in at 2055 local time and pretty much stay for 24hrs, if you're around?

olnico

Yes, choosing hard wood over aluminium was all about where the glue joint is going to be made. It is fairly rare to bond stuff directly on aluminum in our models.
I can flip the wood grain for sure, or also use a plywood plate. At the end of the day I want to test the joint resistance, not the adherence resistance.
So if I see that wood gives up first in my tests with stronger glues, I will change the holding surface.

dbsonic

My Feedback: (3)

Agree, typical bonds are to wood and fiberglass mainly. So wood is a practical choice.

olnico

Thanks. Yes, next tests will be with E-20HP then E-20NS.
Both are 30 min epoxy.

Melchizedek

Oli, would you add " System 3 T88 epoxy " to your test.

I have been using Hysol. But is so gosh darn expensive for such small quantities. I love the stuff but costs too much.
I am now using T88 and would like to know how it compares to hysol 9462.

Thanks,

Kevin

mario tavarez14

My Feedback: (37)

what about 6-10 glue strong like hysol and less money

olnico

Sure guys.
I will source these glues and try them.

olnico

Will try that as well John. Thanks for the tip!

FBaity

My Feedback: (2)

Excellent information...keep it coming.

dszabo2

My Feedback: (1)

Olnico

Nice thread!

The last glue listed is "Wood vinylic glue", I assume you are going to use a standard woodworking glue like Titebond I or II ??

Here is a link to an article by Fine Woodworking magazine comparing glue strengths. It will be interesting to see how your results compare,

http://www.titebond.com/Libraries/Ne..._FWW.sflb.ashx

Dave

Cratus

My Feedback: (49)

Very interesting topic...broad application in the hobby. Thanks for your time and willingness to share your knowledge.

Zor

Hello All,

The pictures at this thread beginning do not show shearing strength because there is a moment arm.
The weight is not at the glued joint but a short distance away.

In most model construction the resistance to shearing is not the main concern.
The main concern is the ability of the joint to hold and any break to occur in the structure and not at the joint.

We are not really concerned with the pieces used in this test as illustrated.
We are concerned wth the integrity of the whole structure ( frame ) of the model in an abnormal landing.

The structure and finish is the main concern regardless of the glue ( cement , adhesive ) used.
A finish that provides strength and reduce forces at the joint is the way to go. Such is provided by a fabric finish where the fabic is fastened ( glued ) to all the structure of the model.

Another factor of importance is the area and shape of any glued joint. Filleting all joints to increase the binding area is highly beneficial. Many joints can have five ( 5 ) times the gluing area by filleting. The glue weight increase is absolutely negligeable versus the total weight of a model.

Penetration of the gluing adhesive into the two parts being joined is of secondary importance since the breakage will simply occur just beyond the degree of penetration.

Of course the design of a model structure enters into play.
A model can be designed aiming at replacement ( a new sale ) for the slighest improper landing or the structure can be designed for minimum damage in a high speed nose-in into the ground.

It will be interesting to see the results of the tests initiated earlier in this thread.
I think that the strength of the adhesive ( glue ) is not the main factor in crash-proofing a model airplane.

Just my feelings based on my experience. There is no doubt that some objections to this posting will occur.

Still it will be nice to see the tests being made with other types of adhesive.

Zor

Jack Diaz

My Feedback: (1)

Thanks Oli. Always providing great information !

Zor has interesting points too. Thanks.

For me, the most relevant factor in choosing a glue is its deterioration over time.
With the exception of V-poxy (or Hysol equivalent), most epoxy glues become like old chewing gum in a few years.


Jack

Ron Stahl

My Feedback: (12)

I did a lot of adhesive comparisons at work and while I cannot share the data I can tell you that 3m DP460NS one hands down in the all around testing. It is not cheap compared to others but was the best overall for bonding other than believe it or not Pacer brand CA's. West 610 is way cheaper but hard to get inside some fuses since you have to use a full sized caulking gun to dispense it.

chuckk2

An easily available and inexpensive glue in my area is E6000. (W Mart, Hobby Lobby, etc.)
Once setup it seems to be stronger than the wood it's bonded to.
It's similar to Goop and Uhu glues, but seems to have a stronger bond, and gets a bit harder.
I'd test it on plastics before real use. I've had no problems,
but you never know.

tucson

My Feedback: (1)

Years ago I built a full size Christen Eagle.
I had to build and glue all wing ribs.
I used a formaldehyde based glue.
For each batch of glue I mixed (powder and water)I glued two pieces of the wood together on end(1\4 inch overlapping each other)when the glue dried I placed the glued joint in a vice. Tightened the vice till the wood broke. If the wood broke at the glue joint first the rib was unusable. It always broke outside the joint for me. I wonder why modelers have not used this glue in there models?
Stan