"Richard Smith" wrote in message news:lybkxng1ym.fsf@void.com...

There's two things I see that I didn't explain well

*
"microvoid coalescence" is the classic fracture mode of a *ductile*
material in overload.

*
as the fillet weld strength is very high, it is easy to ensure the
fillet welded joint is significantly stronger than the sections it is
joining.

Iterating again, using more fancy terms : given fillet welds break by
sudden fracture at a very high stress - and you want to avoid any
sudden fracture - it is easy to arrive at a safe structure by ensuring
that the structural loadings which make the sections (I-beams,
Rectangular Hollow Sections, etc.) bend by plastic yielding are
significantly lower than the structural loadings which if passed to
the fillet weld joints would make then break.
The reason this is easy to ensure by design is that the fillet weld
strength you see is so high in relation to the yield stress behaviour
of the sections being joined. So the strength of the fillet welds
easily well-overmatches the section's plastic deformation limited
load-bearing capacity.
Yet another round of explaining - what you'd see with such a "safe
structure" on increasing load beyond anything sustainable is that the
long sections begin to bend while the welded joints seem unchanged and
not involved.

_________________

I take that to mean that the sections would have permanently deformed and
would be in the region between yield and UTS, gaining strength by
work-hardening when the weld snaps.

I had to straighten some of my used 4" x 8' channel iron. One piece with
slightly smaller dimensions than the rest didn't change at the deflection
calculated for A36 steel and may have been A50. I had to bow it by over a
foot to straighten it. While it was loaded I didn't get close enough to take
measurements.

> ... Yours represents actual service better.

:-)

> ... One piece with slightly smaller dimensions than the rest didn't
> change at the deflection calculated for A36 steel and may have been
> A50. I had to bow it by over a foot to straighten it. While it was
> loaded I didn't get close enough to take measurements.

Some technologists for a steelworks explained about this to me. With
very accurate control of the steelmaking process, they only just meet
the specification yield stress. Resulting in steel which is lovely to
work with in the workshop. Punches, shears, drills, presumably
press-brakes, etc. perfectly.

Cheaper steels drawn in when steel price is high often seem to have a
hardness and yield above the specified minimum.
The punch bangs, etc.

Might seem paradoxical, but you can't use the steel above the design
loads relating to its nominal yield stress, so every advantage is
gained in the workshop and no advantage is lost in service with just
meeting specified yield.
I know from beam bending tests on good quality Structural Hollow
Section that the actual yield stress of eg. an S355 (50 grade) steel
is such a tiny bit above that nominal 355MPa.

For a good quality steel with a higher strength - you'd get it with
good means which cost a bit more than a lower yield steel, like having
more manganese while keeping carbon content low, so you still just
meet specification yield and you still have a steel which is love to
work with in the workshop.

As I understand it.

"Richard Smith" wrote in message news:lyilruh9hx.fsf@void.com...

> ... One piece with slightly smaller dimensions than the rest didn't
> change at the deflection calculated for A36 steel and may have been
> A50. I had to bow it by over a foot to straighten it. While it was
> loaded I didn't get close enough to take measurements.

Some technologists for a steelworks explained about this to me. With
very accurate control of the steelmaking process, they only just meet
the specification yield stress. Resulting in steel which is lovely to
work with in the workshop. Punches, shears, drills, presumably
press-brakes, etc. perfectly.

Cheaper steels drawn in when steel price is high often seem to have a
hardness and yield above the specified minimum.
The punch bangs, etc. ...

-------------------

There is an old process to salvage steel that has inadvertently become
difficult to machine cleanly from random heating called "water anneal". The
steel is heated red, allowed to air-cool until the glow disappears, then
quenched in water. I've used it to soften flame and plasma cut edges that
dulled a bandsaw blade. It seems to improve the machining of steel that
tears out in chunks instead of shearing smoothly, such as hardware store
cold rolled rod and bar.

In the blacksmithing class I learned that hardened steel can be annealed
with good control in a toaster oven by leaving it in for an hour or so. It
may not reach the oxide color of a brief temper. The smith tempers his
custom knives in a similar industrial oven. They would be difficult to heat
evenly otherwise.

My interest, which I haven't fully succeeded at yet, is a knife with a
durable edge and fileable wood saw teeth on the back. I converted a recip
saw blade from rip to crosscut teeth, made a handle and ground a knife edge
on the back, but it doesn't hold the knife edge for long. It does cut wood
quickly and I've also used it on aluminum, to help the tech repair a CAD
design computer when the new power supply wouldn't fit the space from the
old one. He was amused by the mountain-man aspect of fixing high tech with a
handmade knife.

I haven't tried this yet, it seems like something that should be learned
from a master rather than by trial and error. Maybe I should take the
smith's knife making class.
http://islandblacksmith.ca/process/yaki-ire-clay-tempering/

As a kid I practiced building a wattle shelter in the woods using a knife,
ball of string and plastic drop cloth. It was immediately obvious that a
pruning saw would be the better tool to cut withes.