Gluing Stainless Steel – Some Footnotes

UPDATE – See Final Footnote at Bottom re. the Armstrong Method.

A conversation with a friend included this comment:

Well, Brian, just how strong were these glued joins?

And from e-mail:

Did the join strength you tested confirm the claims of the manufacturers?

Were the joins strong enough to use adhesive applications in lieu of brazing? Well, were they?

We tested the strength of the joins – we tested dozens of joins… Unfortunately, we did not have any sophisticated apparatus to measure join strength – we relied on the “armstrong” method of testing by four able, middle-aged men tugging and pulling in various ways to force join failure. In general, we believe that manufacturer claims are reasonable. Here is what we found:

The lap join – the 1/2 sq-inch plate-to-plate joins were incredibly strong.

Of all the trials, the acrylic joins submerged in water for 10 months all failed with two men pulling together to break the join; these joins were strong enough that one man pulling could not break the join.

All other lap joins tested survived four men pulling to break the join. Wow!

The irregular surface join susceptible to peel – the rod and ring join showed a wide range of strength among the trials tested.

The H8610 acrylic join failed with the one man holding the ring stationary and another man twisting the rod – it took some “heaving” and “grunting”, but the join failed.  This adhesive was the strength standout for the acrylic adhesives, though, and all other acrylic trials failed with considerably less grunting!

The DP420 epoxy required two men twisting the rod to break the join. The E-214HP epoxy required a bit more strength from the two men twisting the rod to break the join than the DP420 epoxy. All other epoxy joins were broken with one man twisting the rod.

So, this is all rather subjective, but the conclusions were apparent to all four of who gathered around the pile of joins…

Some elements of application still need to be investigated to fully evaluate these adhesives. The most significant consideration we did not explore was adhesive thickness. We suspect that there is an optimum join thickness – greater than which and the join will fail by pulling the adhesive apart while leaving adhesive on both substrates, and less than which and the join will fail by pulling the adhesive from a substrate. We suspect that join thickness is less of an issue for epoxies in general. For our trials, we used a binder clip to apply uniform squeezing force on all joins, and the result was a thicker join for a more viscous adhesive. None of the manufacturers could give us guidance on optimum join thickness.

We are satisfied that a lap join can indeed replace a braze or a bolt. The irregular join where peel force is evident may fail in the real world application where the braze currently serves the application adequately; however, the manufacturer will give the epoxy a whirl and see if it survives!

Happy gluing.

Final Footnote – a friend asks,

Say, Brian, what is the Armstrong Method?  It sounds sophisticated, but nothing here is sophisticated…

Well, my good friend, the Armstrong Method is a play on the words “Strong” and “Arm”. When you don’t have a sophisticated tool, you are left with no other choice than brute force methods. You can shy away from that embarrassment by referring proudly to the “Armstrong Method”!

Gluing Stainless Steel – a Problem Revisited

About one year ago, I was drawn into a problem with an acquaintance: is there an adhesive application that can replace brazing for two small pieces of stainless steel? The answer is, “Yes, maybe…” At the time, as my buddy and a few more mutual friends were dragged with me into answering this question, well, we all “strayed into the weeds” before we answered the question. It was an interesting experience, though, and there are things to learn and pass along.

First, just for the record, a description of one of two applications we considered (the second one was a plate to a plate or “lap” join):

Join a 316 stainless steel rod onto an irregularly shaped 316 stainless steel piece with a permanent, mechanically rigid bond. The rod is 10 mm in diameter, and it rests on top of the other piece. The other stainless steel piece is a ring that has been crudely machined to “cup” the rod along a portion of the length – for about 20 mm. The current solution is to braze the pieces.

The dominant force applied to the join is a peel force – the ring wants to twist away from the rod with a variable force as high as 25-30 Kg. The service environment of the assembly is warm, humid ambient air – about 40 C and 95% with a dilute vinegar content in the water vapor at its most extreme state. The assembly will be periodically cleaned in hot water and hot air dried.

So, that was the original requirement. Can a hi-tech adhesive replace the braze?

Conversation drifted to include considerations for precision machining the ring where the rod lies. Adhesive trials tested a multitude of products including epoxies, methacrylates and acrylics, cyanoacrylates (superglues), one-part adhesives, two-part adhesives, two step adhesives (using an activator), VHB (very high bond) adhesives on transfer tape, etc. There are a LOT of adhesive products… None of the “party” of problem solvers except for myself were engineers, though one was a chemist. We downloaded application guides and data sheets.The most promising manufacturers were phoned for advice. We all learned a lot about adhesives.

The manufacturers could have made solving our problem much easier than they did – manufacturers typically had “readers” on the phone to search for terms in their document archives, and read to us from their application guides and data sheets. When we approached the problem with the right amount of knowledge, specificity and “pleading”, the person on the other end of the phone on two occasions forwarded our calls to a product specialist who was a test lab technician in both instances. The manufacturers gave us knowledge, but not insight – the manufacturer representatives we could speak to had no real world experience in a production environment, nor any comprehensive data reporting success gluing stainless steel with their products.

Of all the manufacturers we spoke to directly, Loctite and 3M were the most helpful. Loctite provided us a number of useful publications, introduced us to local distributors, volunteered to send us samples (and they sent us a half dozen different adhesives and dispensing guns), and asked for us to report our experience. 3M also wanted to put product samples in our hands. These two manufacturers were eager to help us, knowing full well that we were trivially small business prospects for them.

Choosing our objective “handful” of adhesives to trial in our garages was challenging. Loctite and 3M would have saved us a lot of time by guiding us more effectively in the product selection process. Neither manufacturer had the confidence in their product application knowledge to recommend the two or three most promising candidate products, and both recommended more than a half dozen products to be certain we would find a “winner”.

After describing our requirements I stated above, Loctite asked us questions that started out like this – in this order:

• Did we want a structural acrylic, a cyano or an epoxy? “We don’t know,” I said.
• What were  the materials to bond? “316 stainless steel.” (I already told you…)
• What were the dominant forces? “30Kg peel force.” “What about sheer force?” they asked. “About the same,” I said. “We don’t know much about peel force performance…” they admitted.
• What was the bonding area? “About 3 sq-cm.” “How many square inches is that,” they asked? “Oh – a little more than 1/2 sq-in,” I replied.
• What was the in-service environment? “Like summer in Houston,” I said.

We had already told them these answers when we described our problem, but they weren’t listening to us, quite yet…  And then:

• What was the allowed handling time – the minimum fixture time? “A minute or two should be plenty of time,” I said. “How about 15 seconds,” they asked? “A little too quick,” I replied. I indicated that I preferred shorter fixture times over longer times.
• What total allowed process time – the maximum curing time? “Overnight – maybe several nights would be OK,” I said. Time is not of the essence!
• What was the scale of production – the number of joins per day? “Between one and a half dozen,” I said. This should have told them we needed 30ml or 50ml containers, and some products come only in 400ml and larger volumes…

And then after narrowing down our choices considerably, finally:

• Did we need to “void-fill” – fill gaps and spaces? “Yes – fill a 1-2 mm void – a 1/16 of an inch,” I said.
• Could we do UV curing or heat curing? “Heat we can do; UV we can’t,” I said, and it should have been obvious that UV curing was not applicable to our application. I have a toaster oven!

Our trial results were interesting. We tried Loctite H4710, H8000, H8500, H8600, H86010, E-20HP, E-30UT, E-60HP and E-214HP products. We would have been interested in trying H4720 in lieu of H4710, but Loctite did not have small 50ml containers of this product in stock. Two Loctite adhesives proved to pass muster, though – H8610 Speedbonder  2-part acrylic and E214-HP 1-part epoxy. All other Loctite adhesives failed under stress for our application. E-214HP may degrade significantly in a humid environment, and H8600 may degrade significantly due to the heat of numerous washings and dryings. My money is on the E-214HP epoxy for the long term solution!

We tried 3M DP420, 4DP20NS, DP460 and DP460NS epoxies – the NS variety are highly vicsous “non-sag” formulas. And we tried DP805, DP810, DP810NS and DP820 acrylics. The NS non-sag formulas sacrificed significant sheer and peel strength for the higher viscosity varieties on our stainless steel test parts. The acrylic products all fell far short of the epoxies for overall bond strength. The 3M DP420 product was a close second to the Loctite E-214HP for overall bond strength.

Surface preparation was performed for all trials. For acrylic adhesives, surfaces were polished with 400 grit alumina paper, cleaned with a detergent, then isopropyl alcohol, and finally rinsed with filtered water and air-dried. For epoxy adhesives, surfaces were cleaned with a detergent, and then with methyl ethyl ketone or MEK, roughed with 80-grit emory paper, and cleaned again with MEK and air-dried.

We tried several of 3M’s F9460-series VHB transfer tape adhesives with disappointing results – surface prepped as if for acrylic. These adhesives never dry – they remain flexible and give slightly under stress which did not meet our requirement for a rigid bond. We also tried Devcon’s acrylic Metal Welder and Lord’s 310 epoxy with appropriate surface prep with disappointing results for our stainless steel adhesive test – these two adhesives were simply below average performers in our field of adhesives we trialled.

The summer heat in a garage degraded the acrylic adhesives across the board – they permanently lost about 10% to 20% of their overall bond strength after a summer. Likewise, long-term water immersion degraded the acrylics across the board with more than 50% of strength lost for the several samples we trialled side by side dry and wet.

We had some trouble with 10:1 mixing applicators – three of the Loctite acrylics demanded really precise 10:1 mixing, and it proved to be a bit difficult to do even with the recommended equipment. None of the other adhesives exposed us to this frustration, but we could do this if we needed to.

It would have been much more helpful to approach our problem with this more practical line of questions from someone with current, practical product experience:

• Do you want a flexible bond (for a high vibration – high motion environment), a strong, durable bond or a permanent, structural bond (from a premium product)?
• What materials are you bonding?
• Do you have any operating environment issues – high heat, high humidity, acid vapor? (those conditions eliminate superglues and most acrylics)
• Do you need to machine the bonded surface? Is the bond exposed to impacts? (points to an epoxy with high compression strength)
• Will you sacrifice sheer strength for peel strength?
• What consistency of adhesive do you think will work best for you – runny like water, gooey like honey, gloppy like jelly, or pasty like peanut butter?
• Do you need a fixture time of seconds, minutes or hours?
• Can you tolerate a cure time of an hour or two? Up to a day? Several days?

This simpler line of questions would have taken us down the product trees directly to the H8610, DP420 and E-214HP products in a quick conversation.  But we had fun experimenting anyway!