True ultimate strength of welds

[JAE]

We have a case where we are trying to "fine tune" a weld of a structural member to a plate such that the weld will fail at a specific load...truly fail.

The reason is we don't want any of the adjoining connections or members to fail first...just the weld.

This is somewhat like seismic design where you strengthen up brittle connections and make sure that more ductile members will go plastic and yield first to avoid a brittle failure.
To do that we use an Ry or Rt factor on member material strengths to ensure that we don't underestimate what our true strength is.
For example we find that A36 steel uses an Ry of 1.5 meaning instead of Fy = 36 ksi it is really 54 ksi.

In our case, we are doing the opposite - trying to undersize a weld such that it fails at a specific range of loading so as not to overstress, and damage, adjoining connections.

So...the question....is there a comparable Rt value for E70 electrodes in welds?
If I use 70 ksi - that is listed as the "minimum" ultimate strength. What is the true ultimate strength?
Or does the base metal let go way before that?

 

[kingnero]

Tensile of the weld metal will be higher than the base material.
To test this, weld a coupon with your desired filler on a higher grade base material and perform a tensile test The weld will fail first, and you'll basically get a stress / strain diagram of rhe weld metal.

 

[JAE]

Thanks!
But I don't have a lab to do that.
There's no published info on this question?
I didn't find any

 

[kingnero]

The MTR of the filler metal might give both yield and tensile, if you can get your hands on a type 3.1 or 3.2 certificate?

 

[Ron]

JAE...the weld profile you choose will have an effect on this. For example, a fillet weld maintains the most pure filler metal whereas complete penetration welds have the most solutioning of base metal and weld metal, thus reducing the filler metal strength to closer to the base metal. Welding process, technique, cooling conditions and electrode selection will also affect the actual failure stress. In short, your premise is do-able but you have a lot of variables to consider as I'm sure you already know!

 

[MJCronin]

Redesign your connection to make it a bolted joint...

Use a Pyrotechnic fastener and possibly a load cell to ensure disengagement exactly as you want it....

Is your joint loaded in tension only ???/

Good enough for NASA, good enough for you .....

https://en.wikipedia.org/wiki/Pyrotechnic_fastener

http://www.specialdevices.com/industrial.htm

Many kinds of frangible (explosive actuated) joint designs and assemblies out there..

https://www.eba-d.com/assets/AIAA-2003-4436-Separation-Joint-Tech.pdf

Tell us more about what you are trying to do ....

MJCronin
Sr. Process Engineer

 

[dhengr]

JAE:
Both the base metal and the welding material MTR’s will show some min. yield and tensile strengths, and the actual value can vary above that level, and be within spec. I would talk with the welding rod supplier to see if they can’t fine tune their values for you. Alternatively, what does this detail really look like? I don’t know many good welders who intentionally undersize welds. So, that means slightly oversized welds, with no defects, good shapes and quality, etc. Can you design it some other way, some other configuration, which will make it act like a fuse (breaking point) irrespective of the welds which are actually applied? You might use a an E60 rod which would give you a weaker weld. The actual failure of the weld is a pretty nebulous value, dependent upon many variables and any imperfections, etc. I’ve always looked at welds two different ways; in the center of the weld nugget (at the throat on a fillet) you probably have the F_y of the weld metal, and that’s basically what we design with, at that min throat, in a matching situation; near the base metal you have a mixing btwn. the two metals; and then some short distance into the base metal you have F_y of the base metal and you should actually check that area for failure/strength too. The upshot is that the weld may not be the best place to try to design for some sort of a failure range of loading. We usually pick welding rod to be slightly overmatched and then design so the weld throat isn’t the failure location.

Again, what does your detail look like, including what you are trying to protect, and what you are trying to do, and how is it loaded and stressed? Think of those fixed moment details where they narrow (hour glass) the top and bot. flgs. of the beam to force the plastic moment to happen at the narrowed location. I can’t think of their proprietary name at the moment, dog bone sections or something. For those particular parts of the joints, you might want to get the mill certs for that particular heat, piece of plate or structural shape and design for that actual yield strength. And, maybe yielding, an elongation or deflection limit should show the start of a failure, or should be the failure criteria.

 

[gtaw]

Look for a CMTR from the electrode manufacturer that lists the "all weld metal" reduced section tensile test results. The actual tensile strength will vary from one heat to another because the chemistry will vary within acceptable range. The electrode classification only specifies the minimum tensile strength of the weld deposit.

Best regards - Al

 

[EdStainless]

Don't try to design a weld to fail. Machine groves into a piece of wrought plate as your failure point.
The problem with welds is that you need to control the strength and the size, which is very difficult.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[glass99]

Don't try to design a weld to fail. Machine groves into a piece of wrought plate as your failure point.

I agree. Everyone is nervous about welds, so there is a lot of variability on the overstrength side. Alternatively you could use A36 plate and E80XX welds so that its governed by base metal failure.

 

The MTR of the filler metal might give both yield and tensile

Which are minimum values and have only a nominal relationship to the properties of a weld deposit. For starters, dilution changes everything.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[kingnero]

Which are minimum values and have only a nominal relationship to the properties of a weld deposit. For starters, dilution changes everything.

Which are actual values if you have a 3.1/3.2 certificate, you left out that part of my statement...
But - true though, I did not take dilution into account.

But I said before:

weld a coupon with your desired filler on a higher grade base material and perform a tensile test. The weld will fail first, and you'll basically get a stress / strain diagram of the weld metal.

which will give very close to reality results.

 

Fair enough kingnero, but the whole exercise is so fraught with variability that I question the entire approach.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[kingnero]

Can't argue there...

 

[JAE]

Thanks all.