Butt joint strength: epoxy versus screw

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

lothman wrote:On your test you did the worst scenario for any adhesive.
I agree. The point of the test was not to illustrate a good way of making a joint. It was to demonstrate as vividly as possible some of the differences in character between epoxy and mechanical fasteners.
So with a tube you could distribute the load to all the adhesive area and also with shear stress. This would lead to much higher strength of the joint because of the load suport to the whole adhesive area.
I agree here also. And not by coincidence, the picture that you've drawn is almost a perfect match to what I described in words regarding my door latch repair:
sandwiched the broken part between two splints made of 0.010" steel plate that I carved from an old feeler gauge.
Thanks for the illustration!

--Rik

lothman
Posts: 959
Joined: Sat Feb 14, 2009 7:00 am
Location: Stuttgart/Germany

Post by lothman »

rjlittlefield wrote:The point of the test was not to illustrate a good way of making a joint. It was to demonstrate as vividly as possible some of the differences in character between epoxy and mechanical fasteners.
Sorry but what can we expect from a poor designed adhesive joint? That does not demonstrate the character of an epoxy fastener. See here for the best of two worlds, glued bolts for climbing :-)

SONYNUT
Posts: 635
Joined: Sat Jan 22, 2011 2:27 pm
Location: Minnesota USA

Post by SONYNUT »

REPAIR WILL ONLY BE AS STRONG AS THE MATERIAL USED
..............................................................................
Just shoot it......

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

lothman wrote:See here for the best of two worlds, glued bolts for climbing :-)
Definitely a lovely application. I'm pretty sure I'd be using epoxy for this job too.
Adhesive bolts work because the glue used with them penetrates the rock where it forms a chemical coupling. With right glue, this molecular linkage can be as strong as the rock itself, meaning that a chunk of the cliff has to break loose for the bolt to pull out. Further, the glue reinforces loosely layered or fractured rock, giving adhesive bolts a decided edge over mechanical anchors.
--Rik

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Here are some follow-on observations about a different 2-part adhesive:

Image

Of course I got started with this stuff because I wanted to bond some plastics, mostly PVC.

But then I noticed that unlike the epoxies I was used to, this stuff is a lot tougher, not nearly so brittle. I liked that aspect, and the quoted bond strength is good, so I've been meaning to try it on metal too. This thread gave me a nice excuse to try that, as a sort of direct comparison with a traditional epoxy.

As the package says, this stuff has a "non-sagging" consistency that prevents forming a nice thin bead like I did with the gray Loctite Weld. So, I ran a quick test of two joints with sort of globby beads, plus one joint in which I filed off the entire bead after the glue had hardened, leaving the worst case of a thin joint with a sharp edge against the aluminum.

Here are the two extreme cases:

Image

Image

After all three joints were broken, they look like this.

Image

Red marks the "up" side of the joint, and the two green arrows mark the only places where there was a significant separation of adhesive from metal, on the filed-clean joint.

Now here's the cool part: even the worst case, with the filed-clean joint, was as good as the best of the Loctite Weld joints that I tested. The joints with intact beads were significantly stronger.

In order, left to right, the joints held (failed) at the following lever arms:

3" (3-1/4")
4-3/4" (5")
2-1/4" (2-1/2")

This corresponds to the values reported earlier of 1.75" (2") and 2" (2.5") for the Loctite Weld joints with intact beads.

I was pretty pleased by these results. This seems like good stuff!

Then, since I was revisiting this thread, I happened to go back and look at Ben Krasnow's video that g4lab referenced, the one at https://www.youtube.com/watch?v=YBQp04glQqc. For other people's benefit, I'll mention that this video is titled "Tools and Tips #1 from Applied Science". It contains 12 minutes of miscellaneous cool ideas starting with a cable tie tool and ending with a specialized drill bit for working with brittle plastics.

Someplace in the middle is a section on adhesives, one of which Krasnow describes as "my favorite adhesive of all time. ... If I had to choose only one adhesive to have in my toolkit it would definitely be this one."

Turns out, he's talking about the same stuff:

Image 00:59 https://www.youtube.com/watch?v=YBQp04glQqc

It's not an epoxy, by the way. As Krasnow mentions in the video, it looks like some sort of poly(methyl methacrylate).

Image

Krasnow is also very specific that you have to get this exact stuff: Devcon, Plastic Welder, cream (not clear).

There are a bunch of other plastic bonders that look different and act different.

I hope this is helpful to somebody else. I'm pretty sure it is to me.

--Rik

Edit: to clarify what the red "up" arrows mean
Last edited by rjlittlefield on Mon Mar 16, 2015 11:52 am, edited 1 time in total.

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Post by steveminchington »

This looks fairly serious stuff, and a lot stronger than the Loctite product, possibly due to being less brittle. It looks like the adhesion to the metal is better. As it contains methacrylic acid, a very toxic substance by the sound of it, that may be etching the surface of the metal giving a better bond.
I wonder if this is available in the UK as I have the perfect job for this, making a camera adapter tube. It will have a sleeve joint between two pieces of aluminium tube, so self supporting to a degree. If not, the Loctite will be adequate for this application.
A useful test though Rik.

lothman
Posts: 959
Joined: Sat Feb 14, 2009 7:00 am
Location: Stuttgart/Germany

Post by lothman »

rjlittlefield wrote:Krasnow is also very specific that you have to get this exact stuff: Devcon, Plastic Welder, cream (not clear).

There are a bunch of other plastic bonders that look different and act different.

I hope this is helpful to somebody else. I'm pretty sure it is to me.
Thank for this test. Methacrylate adhesive seems to be very good for gluing alumina. Another source would be stuff called Araldite. For example the Araldite 2022 Datasheet or the Araldite 2021 what is said to have a high peel strenght, what is needed for your joint ;-)

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

test of J-B Weld

Post by rjlittlefield »

While buying some more Loctite Weld the other day, I happened to pick up a package of J-B Weld also. Results from testing the J-B Weld were interesting, so I thought I'd post them out for posterity.

The test protocol was very similar to what I did before. I cut short pieces of 3/8" aluminum rod using carbide blade on my table saw, butt-jointed them with the adhesive to be tested, let them cure for a few days, then tested them to destruction.

One difference is that this time I had the saw set up a little differently, and I noticed that some of the cut ends had metal flakes stuck to them. So I filed those ends until I had a clean filed surface, before cleaning them with xylene and acetone and then gluing them.

Here is the result, after gluing and breaking. Red arrows mark the side that was facing up while breaking.

Image

These two pieces held with 10 pounds on lever arms of 3" and 2-1/2" respectively, failing at 3-1-4" and 2-3/4". This is significantly stronger than the Loctite Weld material.

One huge difference between the adhesives is that J-B Weld has a much longer pot time. The J-B Weld remained mixable and self-smoothing well beyond 2 hours, versus a few minutes for the Loctite Weld.

What really caught my attention, however, were two aspects not directly related to the adhesive strength.

The first aspect is that, while one joint failed by almost uniformly fracturing the epoxy, the other joint appeared to have large areas of unglued aluminum in it.

I had applied the glue by smearing both ends and then working them together, so it was not believable that large areas could have escaped having glue on them. But when I examined the bare aluminum under a low power scope, it was clear that there really was no glue in those areas. So, what happened??

Well, remember those metal flakes I mentioned? It turns out that the flakes I saw were just "the tip of the iceberg". When I filed off those flakes, leaving a pristine filed surface, what I did not realize was that there were still layers of disturbed aluminum below them. What shows as bare aluminum in the weaker broken joint are those places where the adhesive stuck to the surface of the aluminum better than the aluminum layers stuck to themselves. This caused an aluminum flake to pull off with the epoxy, leaving the flake's back side and the matching metal underneath it showing as bare metal in the final broken joint.

Following are high mag photos of the broken ends.

First, the well-behaved joint, shown on the left side of the picture above. At upper left in this new picture, the pit with a bottom of bare aluminum is one small place where the epoxy separated from the aluminum. Farther to the right, there's a pedestal with file marks showing on the top of the pedestal. That's a corresponding place where epoxy separated from aluminum. Everywhere else is fractured epoxy -- despite the appearance of many shiny crystal facets, which I'll talk about separately.

Image

Then we have two close-up stereo views of one face of the weaker joint. In addition to fractured epoxy, it's pretty clear that the surface of the aluminum itself is flaking apart. In the first picture, it's clear that the machining marks run in different directions on the flake and on the metal below it, further indicating that this was a bit of aluminum that had been separate at some point in time and had then gotten re-bonded with the main part of the rod, before I cleaned up the end by filing.

Image

Image

I had not anticipated this laminar reforming of the surface caused by cutting. It's another lesson for future work.

Now, about those shiny crystal facets, well, those are apparently some sort of strengthening component that the J-B Weld manufacturer puts into their resin. You can see a lot more about those over in the Microscope forum, at J-B Weld "Steel Reinforced Epoxy" resin.

--Rik

Chris S.
Site Admin
Posts: 4042
Joined: Sun Apr 05, 2009 9:55 pm
Location: Ohio, USA

Post by Chris S. »

Very interesting pair of posts (including the related one in the micro gallery). Those stereo pairs really add a lot, in this demonstration. (I'm glad my eyes finally learned to fuse them.)

--Chris

lothman
Posts: 959
Joined: Sat Feb 14, 2009 7:00 am
Location: Stuttgart/Germany

Post by lothman »

are the machining marks on the loose flakes perpendicular to the machining marks on the solid base material?

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

lothman wrote:are the machining marks on the loose flakes perpendicular to the machining marks on the solid base material?
I don't know if they're exactly perpendicular, but the ones shown are pretty close to it.

As I mentioned in the post,
rjlittlefield wrote:it's clear that the machining marks run in different directions on the flake and on the metal below it, further indicating that this was a bit of aluminum that had been separate at some point in time and had then gotten re-bonded with the main part of the rod.
--Rik

phytoplankton
Posts: 37
Joined: Sat May 28, 2011 11:38 am

Post by phytoplankton »

Rik, your most recent finding is interesting. You may have issues because carbide blades are very picky about speed. Being incredibly hard but very brittle, cutters made from carbide usually have cutting edges with more obtuse angles and/or radii/chamfers. As a result carbide cutters have to be run at significantly higher RPM than their high speed steel counterparts.

If lower than recommended speed is used then the carbide tends to push the material away rather than properly shearing it. This can result in "tearing" of the metal surface, introduction of compressive stresses into the mictostructure, thermal cracking, bad surface finish, and a built up edge. A built up edge (BUE) is when some of the metal welds itself to the cutter, and then can weld itself back onto the cut surface on subsequent passes of that tooth. BUE is what I thought of when I saw the perpendicular marks on that little bit of aluminum.

If this is what you are experiencing then you could adjust the rpm, change the diameter of the blade, and/or switch to a high speed steel blade. A dull blade could exhibit similar problems except no amount of adjustment, besides resharpening, would alleviate the issue.


I love how we nitpicked the epoxy to death, but being a machinist I can't help but think we could nitpick the bolts too. Did they break via brittle or ductile fracture? What material are they made out of? How were the threads constructed: cast, single pointed, milled, ground, or rolled? Was it heat treated and how? What is the length of full thread engagement?

Anyway, for this test the threaded fastener is obviously better. But epoxy could easily be the superior if it was a choice of large surface area vs a tiny bolt or short length of thread engagement.

EyeDontKnow
Posts: 17
Joined: Wed Apr 08, 2015 11:37 pm
Location: Los Angeles

Post by EyeDontKnow »

At my workplace, we often use a 3M product called "Scotch-Weld" urethane adhesive.
It's a two-part product, and comes in manageable dual cartridges applied with their hand-held squeeze gun, with replaceable mixing tips.
3 minutes (or so) working time. (it hardens fast)

We tried butt-joining two 1/8 inch aluminum plates, 3 inches wide, end-to-end......and after 8 hours or so, trying to manually breaking the bond with our hand strength, was very difficult (but was possible if you ate Wheaties for breakfast)

I think this was the product.....(I'll verify when I go back to work)

http://solutions.3m.com/wps/portal/3M/e ... 556&rt=rud
"Go back in time to the beginning of the future."

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

phytoplankton wrote:A built up edge (BUE) is when some of the metal welds itself to the cutter, and then can weld itself back onto the cut surface on subsequent passes of that tooth. BUE is what I thought of when I saw the perpendicular marks on that little bit of aluminum.
I agree, and as further evidence I submit the following image of the saw blade after completing a cut.

Image

I was not surprised that some material got re-deposited onto the cut face. What did surprise me was that the re-deposited material was well enough integrated that it faked me out when I thought I had created a pristine filed surface. It's just one more box to be checked, on the list of "What all do I have to do, to be sure that this surface is suitable for gluing?".
Anyway, for this test the threaded fastener is obviously better. But epoxy could easily be the superior if it was a choice of large surface area vs a tiny bolt or short length of thread engagement.
Sure, no disagreement there.

This thread was started so long ago that I'm pretty sure its context is no longer clear. Perhaps it never was.

So to make that explicit: what prompted me to start this particular thread was another thread in which I supported constructing a lens extension by screwing together a bunch of off-the-shelf M42 extension tubes, while another member recommended instead cutting a piece of PVC pipe to the proper length and gluing on a couple of lens mounts.

Based on previous experiences with adhesives, I twitch pretty badly at the thought of recommending to hang either an expensive lens or an expensive camera off the end of a glued butt joint.

I've had a fine time and learned a lot in fleshing out this thread, but that earlier opinion is unchanged. For critical applications I still won't be using an adhesive in preference to a suitable mechanical attachment, and I won't be recommending that anybody else do it either.

Reasonable people can differ on this. In that case I hope this thread is useful in at least helping to identify and illustrate some of the issues involved.

From my standpoint this thread's major value seems to have morphed into a place to accumulate data about various adhesives that I use for situations when mechanical attachments are not feasible, for whatever reason.

--Rik

soldevilla
Posts: 684
Joined: Thu Dec 16, 2010 2:49 pm
Location: Barcelona, more or less

Post by soldevilla »

I think that this test is poorly performed from the beginning. Aluminum is very demanding on the joints with adhesive (and welding) because immediately it is cut, the surface is oxidized, becoming alumina. You're not sticking aluminum from aluminum, but alumina with alumina, and effort of the test does is separate the alumina layer from the aluminum (and re-form alumina in the new area).

Try this: Prepare the mixture of epoxy, put a generous amount on each end to join, and with a sharp steel tip, scratch all you can the surfaces to bond. What you're really doing is removing the layer of alumina, but as it is covered by epoxy, no oxygen can oxidize the aluminum. Do it on both surfaces to be joined and then attach them with pressure, so that the epoxy layer as thin as possible.

Your explain me later the result :D

Post Reply Previous topicNext topic