On Fri, 09 Jul 2004 03:31:56 -0400, Gary Coffman
wrote:

On Mon, 05 Jul 2004 11:44:08 +1200, Eric Stevens wrote:
There is NO weld technique which produces a weld with metallurgy
identical to the the parent metals. ANY weld technique leads to a
discontinuity in material properties in or around the weld zone which
ALWAYS results in a propensity for the welded structure to fail in or
around the weld zone rather than the parent metal.

Incorrect. Consult a good welding text such as "Modern Welding" by
Althouse and Turnquist (the most widely used, and most authoritative,
welding textbook).

I've already answered this but you deserve a better response than a
mere battle of authorities.

It is true that fusion welding produces a HAZ (Heat Affected Zone) around
the actual weld joint. This can significantly alter the properties of *some*
materials, namely medium and high carbon steels, some alloy steels, and
some aluminum alloys. But *part of the welding process* in those cases is
post-weld heat treatment to restore those properties to their original pre-
weld state. In other words, you haven't completed the welding process for
those materials until you've done the post heat treatment.

For materials such as mild steel, the most commonly welded material,
there is no such concern. The HAZ doesn't affect the material properties.
That's because mild steel has too little carbon in the solid solution to
produce the phase changes that could alter its crystaline structure.
A *competent* welder will also choose an appropriate alloy filler material
so that the fusion zone won't have different properties from the parents
either.

It is well to note too that different welding techniques produce differing
size HAZ. TIG welding produces less than arc, MIG produces less than
either, and exotic techniques such as laser welding produce practically
none at all.

OK, you have partly acknowledged my main point, that there is a
discontinuity around the weld. In this case it is the 'heat affected
zone'(HAZ). Now, not all HAZs need post-weld heat treatment, but HAZs
still exist.

The metallurgy of the HAZ is visibly affected as can be seen in the
micrographs in http://www.nuvonyx.com/catalog2/welding.html
The variation of the mechanical properties of the material in the
immediate vicinity of the weld can be seen in diagrams at the same
site.

More information is available at the site of The Welding Institute at
http://www.twi.co.uk/j32k/protected/band_3/jk48.html

It is a common view that the resulting weld will be strong if one uses
a weld material that is stronger than the parent material. However
fusion welding is a casting process and entails a puddle of molten
material cooling down and solidifying. The resulting shrinkage
requires local yielding of both the parent and weld material. If the
weld material is too strong it will not yield and all the shrinkage
has to be pulled out of the parent plate. Even if this does not cause
cracking it leaves high residual stresses.

Someone will now say that you can relieve the stresses by heat
treatment. That is perfectly correct but this further modifies the
metallurgy of the weld zone and some of this may be adverse. Grain
growth and embrittlement is a particular problem.

The reason why any discontinuity in properties at a weld results in a
local weakness is that the mechanisms of failure are complex. Even
ignoring any geometrical disturbances caused by the original welding,
deformation will occur in the metal when stresses are applied.Tension
applied to a plate will cause it to become slightly thinner and
narrower. At low stresses the deforemation is elastic and the plate
will return to its original shape if the load is removed. If the
applied load becomes sufficiently high yielding (i.e. plastic
ceformation) will occur in the material. If there are local variations
in the material properties (e.g. at a weld) some parts will yield
before others. This will result in a local redistribution of the
stresses and the stronger parts of the material will end up carrying
more load than they would if they had been surrounded by a material
with a higher yield stress.

Now it gets really complicated. The local material is subject to a
mixture of tensile and shear stresses. The interaction of these gives
rise to 'principal stresses'. there is a principal tensile stress and
a principal shear stress which acts at right angles to the principal
shear stress. With some meterials it is the principal tensile stress
which causes the failure and with others it is the principal shear
stress which causes the failure.

The metallurgical discontinuities associated with a weld give rise to
local disturbances in the stress pattern. These give rise to local
variations in the principal stresses. Some will be higher than in the
undisturbed parent plate. Others will be lower. Some parts of the weld
zone will be less likely to fail than the parent plate but there will
always be parts more likely to fail than the parent plate. It is at
these latter locations that failure will commence.

That's why I say a weld is never as strong as the parent material.

For materials such as mild steel, the most commonly welded material,
there is no such concern. The HAZ doesn't affect the material properties.
That's because mild steel has too little carbon in the solid solution to
produce the phase changes that could alter its crystaline structure.
A *competent* welder will also choose an appropriate alloy filler material
so that the fusion zone won't have different properties from the parents
either.

It is well to note too that different welding techniques produce differing
size HAZ. TIG welding produces less than arc, MIG produces less than
either, and exotic techniques such as laser welding produce practically
none at all.

Now you are postulating *cold welding* for the gage blocks, and that
produces *no HAZ at all*.

Unless the weld is perfect, it still leaves an interface detectable by
microscopy.

So the material properties surrounding the
weld joint would not be altered *at all*.

Its not just the properties 'around' but the properties 'at' which
matter.

Of course cold welding isn't
what's actually happening when you wring gage blocks together, but
if it were, you'd still be wrong.





Eric Stevens