mechanically strongest soft solder?

[David Hess]

For tin/lead solders, the common Sn62Pb36Ag2 alloy is suppose to be the strongest.

[johansen]

Stay brite 8 has a tensile of 15,000 psi. Its just silver and tin.

Citation needed.

Horse mouth says 10.6 KSI,

If you google "staybrite 8 tensile strength" you will find half a dozen sources whoch claim 14 or 15k psi, 10,500 shear.

https://www.firstsupply.com/Product/JWHSTABRITE

[T3sl4co1l]

If you google "staybrite 8 tensile strength" you will find half a dozen sources whoch claim 14 or 15k psi, 10,500 shear.

https://www.firstsupply.com/Product/JWHSTABRITE

Do you buy components based on what Mouser, Newark, etc. say in their catalogs?  Same thing: suppliers can be wrong, and often are.  Show me a datasheet -- otherwise it's just hearsay.

And again, no mention of creep, what rate this was measured at.

...Is this just a mechanical thing I don't know?  Are the acceptable standards for evidence just that much looser?  ...Do mechanical engineers science?

I know these things must exist, I see papers about them (if in short supply around this particular range of materials), and I know stuff like aerospace is, or at least can be, researched and simulated and evaluated to extreme degrees, so I don't seem to be asking anything facially unreasonable here.  This makes sense, right?

...Maybe it doesn't make sense, making such a big deal about the correctness of buddy it's just solder.  But then, that was the question, and if you're going to be fine with some random posting on the internet saying it's ca. 10KSI, I'll tell you this: you're equally well off knowing it's 5 or 20 -- all of them are equally wrong.

Tim

[coppercone2]

https://www.rsd.net/assets/prod/742.pdf

it says 14kpsi

[T3sl4co1l]

https://www.rsd.net/assets/prod/742.pdf

it says 14kpsi

[datasheet]
but again I'm dubious about method or how it's actually measured (no standard cited, and not that ASTM standards are generally publicly available anyway).

We're going in circles here.....

Tim

[coppercone2]

its not a random internet source it has harris letterhead on it. again I would be surprised if a primary welding supplier is not doing the tests the same way they test their other joints

[T3sl4co1l]

The source is fine (the sheet, not the supplier).  The content is not -- it's unsupported, a number in relation to nothing else.  If I try to sell you a computer speaker that claims "800W PMPO", are you not the least bit curious what method I applied to arrive at such a ridiculous number?

Only one thing left to do: contact them and inquire what their actual test methods are (and, why they don't put that in the datasheet).

Tim

[coppercone2]

My reasoning is unless there is deception they have some pulling machines to do lap joint tugging for weld rod testing. I suspect they solder two 1 inch area rods together end to end and put it in the machine. I have seen footage some where long ago of silver braze being tested this way, along with welded samples (they have 3-4 different machines with various adapters ETC if they need to test pulling two sheets apart), maybe even in harris. I suspect their technicians will do the same thing over and over again (as close as possible because they don't like changing anything in test labs nor do they like learning). The pull test for rods (i.e. welded rebars, girders) HAS to be good in a company like harris, because they have the certs for construction materials, even probobly bridges etc, which is serious shit. That is occams razor corporate for me.

Now if there is a sneaky salesman doing something else that would be disappointing. I think I have seen copper pipe maybe, for some hVAC testing videos, what they did there is solder two pipes together after one of em was expanded  bit on a mandrel, then attach it to a steel rod with alot of braze, and put it in the pull tester machine

[DiodeDipShit]

Berni has the issue solve.

[Berni]

Also as people above mentioned tensile strength is not the whole story. Once you expose things to a heavy vibration environment you might get a failure due to metal fatigue instead of over stress. Solder is prone to metal fatigue, exotic solder mixes might be even more so.

While designing a bracket to give more surface area to grab onto is pretty easy. Instead of just having a cut off end meet the object you give it a L shape to increase surface area contacting, perhaps add gussets add more area and prevent bending that might cause extra fatigue. Even better if you can design the brackets solder joint to be primarily loaded in sheer force rather than tensile force. If given enough surface area the copper bracket might break before the joint itself does (since copper itself is not a very strong material, also loves to fatigue crack).

Id say if any of that is still too weak, then you are approaching the problem from the wrong direction and need to use a steel bracket attached with a large number of strong high grade screws or something. (In that case id love to know what your product is going trough to have a heat ex changer exposed to such forces, since it sounds like that component is going for a very exiting ride)

[T3sl4co1l]

Huh, there's actually a Wikipedia article on it:
https://en.wikipedia.org/wiki/Solder_fatigue
only discusses the models though, no example parameters for alloys.

This seems interesting,
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4121147/
just scanning it, I don't see a discussion of fatigue limit (stress for number of cycles)...

Or this,
https://www.osti.gov/servlets/purl/1818047
preliminary, so presumably there's a follow-up reporting high-cycle data, hmm...

But yeah, needless to say, solder doesn't do well with static loads, or heavy vibration.  A certain amount of static strain, particularly at elevated temperature, can be relieved by creepage, but constant or repeated creepage eventually leads to crack formation.  High frequencies are better (more strength for given fatigue cycles), as suggested by the time dependency.  Which is again why single numbers don't mean a thing, sure it might be 15KSI, at what rate? Is that a shock load or what? Against what materials, what surface prep, what joint geometry?  All of these matter and the results can vary wildly.  And if you're trying to apply that number to some other geometry, how confident are you that that number will even apply?  A solder joint is not simply a solder joint!

Also, how much and in what ways strain relief happens, probably varies substantially with and without lead (at all, a few percent, ~half..).  And pure tin is known to relieve stress by whiskering; I haven't seen mention of that alongside this [bulk stress, fatigue] yet, but I wouldn't be surprised if that's part of it.  Annealed matte tin is the best to deal with whiskering, IIRC, which is to say, relieved stresses.

Like I said, if you need to fix something in place, using epoxy is likely to fare better; it is free from creepage AFAIK, and apparently can have exceptional fatigue limits (particularly composites like carbon fiber?).  Even if you need to make the joint bigger to handle a lower tensile or adhesion strength (say 1 or 2 KSI peak repetitive loading), as long as you have the area available, sure, slather it up eh?

Tim

[coppercone2]

IDK man for the same size epoxy and solder joint for something small I trust the solder way more

Like if you glue down a stud you can knock it off but a soldered one is damn strong

[DiodeDipShit]

The eternally revolving thread...... Solder fails pull test, take 12......
What is the hang weight of the application, times the pulse rate, times the duration of pull, times factors of solder adherence fatigue fracture, equals ultimate failure.
Is ultimate failure unacceptably shorter than life expectancy, then the drawing board must be revisited.
An angle will tear the joint imperceptibly at first, at the juncture of the right angle, but leading, in shorter time to a catastrophic failure.
Therefore the use of a channel bracket will distribute stress to two junctures, exponentially extending life expectancy.
This requires the use of a clevis pin spanning the opening of the channel and a sturdy eye bolt with a close tolerance engagement of the OD of the Clevis and ID of the eye.
Still, failure will occur given enough cycles of abuse. Has the soldered channel , clevis and eye exceeded life expectancy, so be it.
Has the channel fallen short? Add four bolts close within the the four corners of the channel, aligned with four tapped holes, twice deep of the internal thread diameter of the bolts, into the copper....... If this configuration fails life expectancy, place your head between your legs and kiss your azz goodby.

[jpanhalt]

Like if you glue down a stud you can knock it off but a soldered one is damn strong

I've never tried to solder a wood 2x4 stud.  What flux did you use?  Were you attaching directly to concrete or a wood base?

[T3sl4co1l]

Like if you glue down a stud you can knock it off but a soldered one is damn strong

I've never tried to solder a wood 2x4 stud.  What flux did you use?  Were you attaching directly to concrete or a wood base?

You must think you're such a hit with the ladies.

https://www.pemnet.com/products-overview/products-studs/

I assume something like this... hopefully with a wider base to distribute load better, and, obviously made for soldering.

On a related note, guess I should cross-link this thread as well: https://www.eevblog.com/forum/manufacture/mounting-a-pcb-using-only-surface-mount/

Tim

[jpanhalt]

Steel "studs" used in the USA are not intended for weight bearing and are intended to be screwed, not soldered.  But then, it's CC2, who knows?  Maybe a gallium alloy sticks well to wood studs, and that is what he used.

I have been with one woman for 56 years and have no intention of getting a newer version that can navigate for me.

[DiodeDipShit]

I believe Coppercone is referring to a circuit board terminal stud. . . . . . . See attached file.

[IanB]

I would not  depend  on  solder joint to hold mechanically,

It is only for electrical conduction.

Just because tin/lead solder is very convenient for electrical joins, does not mean it cannot also be used for mechanical joins. It has been used in plumbing applications forever, and still is (minus the lead). Try to pull a plumbing joint apart by force, and see if you can.

When I was young I used soft solder to join various steel parts together in projects, and they held together fine.

[johansen]

Put a lot of vias under anything mechanically surface mounted...

[T3sl4co1l]

Joke being, there's indeed three meanings: stud, the (usually wooden) beams making up walls in conventional wood-frame housing; stud, the threaded metal rod used for fixing things in place (ranging from all-thread e.g. studs in an engine block to weld-on elements used in fabrication, shipbuilding, etc.); and stud, male breeding stock.  Unsure of which one was meant, I chose the obviously-wrong third for hopeful comedic effect.

Tim

[mzzj]

I would not  depend  on  solder joint to hold mechanically,

It is only for electrical conduction.

Just because tin/lead solder is very convenient for electrical joins, does not mean it cannot also be used for mechanical joins. It has been used in plumbing applications forever, and still is (minus the lead). Try to pull a plumbing joint apart by force, and see if you can.

When I was young I used soft solder to join various steel parts together in projects, and they held together fine.

Even more extreme application would be typical double-barrel shotgun. Barrels are soldered to the breech block and have to withstand proof pressures up 20 000PSI.

Although it is known that with some shotguns the barrels start to escape if barrels get really hot during shooting.

[coppercone2]

I drilled and tapped it because I realized there is a place to do it

[tooki]

Joke being, there's indeed three meanings: stud, the (usually wooden) beams making up walls in conventional wood-frame housing; stud, the threaded metal rod used for fixing things in place (ranging from all-thread e.g. studs in an engine block to weld-on elements used in fabrication, shipbuilding, etc.); and stud, male breeding stock.  Unsure of which one was meant, I chose the obviously-wrong third for hopeful comedic effect.

Tim

Don't forget the studs punks and fancy fashion designers rivet to their clothes and accessories!

[T3sl4co1l]

Joke being, there's indeed three meanings: stud, the (usually wooden) beams making up walls in conventional wood-frame housing; stud, the threaded metal rod used for fixing things in place (ranging from all-thread e.g. studs in an engine block to weld-on elements used in fabrication, shipbuilding, etc.); and stud, male breeding stock.  Unsure of which one was meant, I chose the obviously-wrong third for hopeful comedic effect.

Tim

Don't forget the studs punks and fancy fashion designers rivet to their clothes and accessories!

I think even more generally, studs are metal attachments of some sort or another, but I should look up an actual definition at this point.

So there you have it, if they were made out of wood they might be all three at once.

Tim