To get specific answers I think you're going to need to post more details,
such as the material in question.
But in the mean time: Let's compare the bars to an ice cubes. A large ice
cube and a small ice cube are made from the same material and both melt at
the same temperature. However the large ice cube needs more heat (BTU's not
degrees) to melt. Likewise "large bars" require more heat than "small bars"
to heat up the center. Small bar don't require preheat, because the weld
heat alone is sufficient to heat the center. So you ask "Why do you need to
heat up the center?"
If the center were cold during welding then the cold center could quickly
quench the weld area. This is know as "self quenching" The quenching could
cause cracking, small hard grains, martensite, and distortion. In general
quenching is bad, when not expected.
On the other hand plain low carbon steel is considered non-heat treatable.
Therefore it does not require preheat treatment.
Preheat reduces the rate at which the weld area cools. This allows the
hydrogen trapped in the weld metal to migrate out. Richard might chime in
on this subject since he has done some research on the subject. Hydrogen
trapped in the structure will be a crack site in the future.
A geat example I saw several years ago was a 2.5 inch plate six inches
wide that was being rolled when it fractured suddenly. The material was
regular old A36. Nothing could be softer than that stuff. I was told that
the steel was "defective" and the supplier had been called. I took a closer
look at the fracture and it appeared that someone had used the plate as a
work table of some sort. There was a tack weld about a quarter inch long
that had been ground flat before the plate had been cut in strips for
The sudden heating in a small location and then the rapid cooling from
the massive heat sink created a site for fracture. It might have been
martensite or a fracture caused by entrapped hydrogen. I am not sure. I do
have pictures. If enough people are interested I could post in the dropbox.
As a general rule you should preheat any material over one inch thick.
I have also been ordered on many occasions to tack weld at least for a
length of one inch on large weldments. The weld can be very small in cross
section but must be one inch long.
Hi "Chopster", Randy
Will do my best...
Point which needs to be mentioned first-off - all "cold-cracking" you
will encounter in commercial practice using steels will involve
hydrogen in embrittling the metal. One part per million by mass of
hydrogen in steel (not austenitic stainless steel - that is highly
immune!) will give you a susceptibility to hydrogen cracking. By
susceptibility we mean it can hydrogen crack - that doesn't necessarily
mean it will if other things are OK.
The reason the section has to be big in order to have to start worrying
about avoiding cold-cracking for a "weldable" steel:
* it takes time for weld hydrogen to escape the section - if it were
3mm thick it would be gone in half an hour - much reduced in
concentration in minutes - too fast for cracking to have a chance. In
thicker sections it can be days and weeks.
* you need static stress to burst by cold-cracking and a thin section
simply cannot hold this. Thinner sections warp. Distortion control is
a major issue in thinner sections. Everyone who welds has seen that!
In thicker sections, you self-restrain against shrinkage stresses,
often with static stresses up at the yield point (!). So then your
issue becomes then this business of cold cracking.
So to a good approximation, if your component is smaller section and
bending around as you weld, you are not going to be having worries
about hydrogen cracking.
Randy Zimmerman wrote:
good info Richard, thanks
a few related question,
what is considered thin and thick section?
would you say anything under 1/4 " be considered thin?
by allowing the welded material sit for days before stress applied,
would it releave hydrogen from the metal?
will differnet welding procedure effect this hydrogen embritlement ?
example TIg would have less vs. MIg (metal heated for a longer time
more in TIG) same goes for HAZ?
I've heared people tlak about hydrogen embrittlemnet in stick welding,
but some one told me at welding school that it is almost does not
apply to MIG and TIG.is it true?
would post heating releave the chance of cold cracking? example
material which is geting powder coated usually heated to 400F for at
least 10 minutes?
what about parts that are getting welded on both side? they are pretty
much pre heated should that rule out hyd. emb.
On 22 Feb 2005 00:53:56 -0800, firstname.lastname@example.org wrote:
That's where the streetwise guys need to contribute? What do you
really know to be the case? Is this guy's guess right - that regarding
hydrogen-induced cold-cracking, anything 1/4inch (6mm) or less is
"thin" and doesn't need preheat or other precautions?
In thick sections, the "frozen-in" stress from contraction on cooling
is very very high, as I mentioned. Externally applied stress will make
not-a-lot of difference. So no, you can't escape the danger of
hydrogen cold-cracking that way.
About the only place where external stress does seem to make some
difference is in pipelaying - if you are sort of "dangling" the pipe
down into the trench, slinging off the newly attached pipe which has
just been root-runned into place. The underside is going to be in a
lot of tension. Seen a pick from the 1970's of this as a practice and
an analysis of it.
Sound familiar to any of you practiioners?
Yes. Solid wire MIG gives very low H. I think in the USA it is used
for that reason (because there isn't much of the TMCP-HSLA steel made
in Germany and Japan(?)). Lots of smaller runs on big structures like
say arc-furnace roof structures - more runs but no need for preheat and
other expensive and troublesome procedures (do you really want to be
standing in the middle of a big structure at 100C (b.pt. of water)
doing a weld? Not really!) - so better-off overall.
Assume TIG gives even lower H, but it isn't used for large-scale fab.
as limited in power and slow anyway compared to MIG.
Well yes, that is pretty much so, I believe. All but basic stick
electrodes need some moisture in them, giving a steam blanket over the
weld. Air getting to the weld pool would give an irredemably rubbish
weld. With hydrogen/steam blanket, need to take anti-H-cracking
precautions in thicker sections, but at least the weld is good. Basic
electrodes can be baked down to an H-level about 3 to 5 times higher
then the H level from solid wire MIG.
But flux can be very useful in controlling the weld. In
flux-cored-wire welding (like MIG but has a core of flux in hte wire),
the shielding gas does all the shielding, so don't need moisture so is
baked down to very low H level in manufacture. And being inside the
wire, does not reabsorb moisture. FCW down at H level of well-baked
Post-heating is good, but usually expensive.
But do you really need it on your components? Are they thick and
Preheat is reasonably cheap and good - heat with blowlamps or electric
blankets to a specified temperature, then you take away heat and do
your weld immediately. Welded both sides doesn't make much difference
does it - you experienced weldors out there...?
Ok I have posted that fracture of a plate strip in the drop box. It is
listed uder "fracture"
As shown the tack is quite small and the metal obviously was molten and in a
very short period of time quenched down to ambient temperature.
If in doubt PREHEAT!!!!
According to Mr Dyson's calculator:
Shows suggestion of 1/4inch (6.3mm) for largest plate thickness butt
weld without needing preheat is spot-on...
Setting the "worst" conditions I could set for "normal" commercial
- plate had 0.22%C (highest encountered for plate steel) and 1.5%Mn
(also top), but unalloyed (typical for eg. pipeline) -> 0.47 Carbon
- SMA using higher-H electrode - calculator says ">15mmlH2/100gFe" for
rutile and cellulosic - though cellulosics go way above that - said to
be up to 100mlH2/100gFe
- fast weld with low heat input 0.7kJ/mm of arc about 250mm/min weld
speed at 70A arc (arc at about 40V due to high H?) -> 0.56kJ/mm heat
per length going into metal of weld (rest eg. radiated away)
No preheat needed to 14mm of combined thickness - which for a butt weld
means half that in plate thickness - 7mm
My Dyson migh like to correct me, but I think a cellulosic with a
healthy hydrogen level is going to need a little bit more of
So the suggestion of 1/4inch (6.3mm) for largest plate thickness butt
weld without needing preheat is spot-on, given the guidance of Mr
John Dyson wrote:
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.