Can Loctite 638 really be this good?

I was asked by a ceramicist friend if I could machine an external flange from 6mm steel plate to fit his clay extruder:

The idea is to bore out the plate to slip over the tube and fix it in place. Welding is the obvious way, but I have only a stick welder and little skill with it - I worry that I would blow through the ~2mm tube trying to weld 2mm to 6mm.

I suggested silver solder, and after having quickly got past the usual 'Him:but will silver solder be strong enough? Me: yes.' conversation I came across the specs for Loctite 638.

Henkel give a tensile shear strength of >25N/mm^2 for a joint up to 0.25mm. The tube is about 100mm external diameter, Pi is 3, the plate is nominally 6mm, so that's roughly 1800 mm^2 surface area. If I've got the arithmetic right the joint should fail at somewhere around 1800 x 25 = 45,000 Newtons, or about 4.5 tonnes force.

Can this be right? I doubt that the arm-powered force on the piston which drives the clay can be much more than 5,000 N which would give quite a good safety margin.

It seems crazy that glue can be that strong , but maybe?

Robin.

Edited By Robin Graham on 31/03/2021 00:53:10

Looking at the ammount of pitting corrosion on the tube Loctite is not the answer.

Machine the flange with a 45 deg chamfer 2 to 3 mm on the bore.

Carefully Swage the end of the pipe with a ball pein hammer to the chamfer.

Eric

Another answer is to approach a local garage and get it welded. I don't imagine that a full weld would be required, just a series of long tack welds.

Brian

Edited By Brian H on 31/03/2021 08:29:30

I would say future corrosion will be the death of a repair anyway so OV's idea + epoxy is worth a try.

To the uninitiated The term silver solder probably suggests a relatively weak join as per ' soft ' solder, which may be off putting. If the term brazing was used, might give more confidence and as we all know can provide a remarkably strong joint.

However, I do agree with previous contributors that due to the relatively small surface area and corrosion if professional welding is not an option the epoxy route might be the one to consider.

Regards

I think the proper way of fixing the plate is to braze on a fairly thick walled flange to the cylinder, it could be bent up from 25 mm angle iron, and then bolt the new plate to the face of that. I really don't think Loctite would hold it.

Brian

As I have repeatedly commented on this forum:

Loctite 638 [*] works by hardening to a material not unlike Perspex ... which fills the voids between the two surfaces and then then acts in shear.

[ it is not a particularly good adhesive, and has quite low peel-strength ]

The exact mechanics of a particular joint will vary, but : provided that the 638 cures, the presence of a rough surface is no great detriment ... the ‘balls’ of resin are in compression when the joint is loaded.

If you can get clean metal surfaces [albeit with some pitting] on both components, and a sliding fit ... I predict that 338 would work nicely.

A wipe over with Copper Sulphate solution, allowed to dry before assembly, would provide a useful catalyst to accelerate curing.

MichaelG.

.

[*] at least in the original formula, which is the only one I have used

.

Edit:

I have assumed that the 6mm plate will be evenly loaded, not subjected to localised point loads that might peel the joint.

Edited By Michael Gilligan on 31/03/2021 09:42:09

Would JB Weld do it? Thinking you could mechanically key the two together to some extent by abrading the mating surfaces, then applying the JB Weld and forming a fillet around the bottom with the excess?

Some years ago, I fabricated backplates to carry ER25 and ER32 chucks for use on my big lathe.

The mating surfaces were smeared with Loctite, and the setscrews lightly nipped so that the collet chuck could be tapped about until the taper bore clocked true (Within 0.0005" )

Later, having obtained and machined castings, I removed the setscrews and hit the chuck backplate with a copper mallet thinking to dislodge it.

When it looked as if the increasing impacts were going to do damage, I stopped. The original set up still remains!

It is very likely that the crown wheel in your car is secured to the diff carrier with Loctite, or some similar anaerobic.

Howard

Howard,

This is all very interesting but what is being considered in this thread is no more than a butt joint of badly corroded

2 mm wall section material straight onto 6 mm plate. This item is a clay extruder and we are given no guidance at all on what pressures this device creates as it is extruding.

I stand by my recommendation having just had to fit securing pins down the glued joint faces of helical reduction gears on their spigots in the gearbox on my Dore-Westbury mill.. One had failed in shear after nearly 30 years of service from build

OK, a long time maybe, but avoidable and certainly inconvenient when it happened. Had I been more experienced when I built it I would have reinforced the joints at the time. As good as Loctite is, I really can't believe a joint of that sort would hold together with any serious pressure behind it.

Regards

Brian

.

Sorry, Brian ... I beg to differ !!

Robin wrote: The idea is to bore out the plate to slip over the tube and fix it in place.

MichaelG.

Hello Michael,

Indeed you are correct in which case I stand duly chasticed and toppled from my soap box. Mea culpa!

It shows the power of the picture to divert one from the description.

Regards Brian

Thanks for replies.

Brian - I'm sorry if the pic was misleading! I did in fact give an estimate of the force involved (about 5000 N) which would give a shear of roughly 2.8N/mm^2 for this joint - well within Henkel's claimed 25N/mm^2. The reason for my post was that I was surprised by my calculation, and thought that I might have got the decimal point in the wrong place, or that the Henkel number was under lab conditions and someone with real world experience would say I should derate that by a factor of 100 maybe. It seems not.

Encouraged by anecdotal evidence and MichaelG's bold prediction that 638 (I assume that 338 was a typo Michael?) I'll give it a go and report back. The worst that can happen is that it fails and I have to resort to 'brazing' - I take your point Mike - or welding.

Some interesting (to me at least) things I learned in my researches for this job are:

• If I soldered with Silver Flo 55 the joint would (theoretically) be good for 25 tonnes - the tube would burst first.
• Epoxy resins (including JB weld) are weaker in shear than cyanoacrylates .
• There is a significant difference between the tensile and compressive strengths on a joint of this type. In this case it's tensile, which is weaker than a compressive joint. But Henkel specify 25N/mm^2 tensile, so it might work!

Robin

You might struggle to get anything to stick to that poxy old bit of pipe. Appears to be more 'oles than metal. If you are going to the effort to do the flange it seems a new length of pipe might be in order?

.

It was indeed a typo, Robin

... my apologies for that.

MichaelG.

.

Edit: __ The standard ‘Pin and Collar’ test is briefly described here:

http://www.adhesivestoolkit.com/Docs/test/MECHANICAL%20TEST%20METHOD%201%20-%20Continued.xtp

The ISO and ASTM standards are technically similar, but ISO uses Metric units

Edited By Michael Gilligan on 01/04/2021 07:36:53

Hello Robin,

I should apologise to you really, I made the classic mistake of not reading the question fully and then answering what I imagined it said ! I wish you well in whatever route you chose to follow and will be interested in the outcome.

Best wishes

Brian

Very noble of you to make two apologies, Brian

... for the simple mistake that we surely all make from time-to-time.

MichaelG.

Hi MichaelG, I did wonder if you had made a typo, but I searched for Loctite 338, as it wasn't in my Loctite booklet, which is several years old now, and found Bonderite, so connecting it with your paragraph;

"If you can get clean metal surfaces [albeit with some pitting] on both components, and a sliding fit ... I predict that 338 would work nicely."

I read that you meant 338, as it is a Henkel product.

Regards Nick.

Edited By Nicholas Farr on 01/04/2021 15:11:14

.

... and I felt obliged to confirm Robin’s assumption that I had made a typographical error

I should have written 638 ... which was the product under discussion.

I did not mean 338 ... which is a product that I have never used.

Yes, it was a simple typo.

MichaelG.

Is this too simple ? just run it past Henkel's technical support people and follow their reccomendation. I have always found them extremely helpful, no matter how small the scale of the request.