Has anyone seen a publication be it a book or university paper that
describes in detail how to heat metal for bending purposes.
I was just informed by the shop forman that the torch has to be moved in
line with the stress the part is to carry, for example a beam would be
heated by moving the torch back and forth along the long axis. In order for
shrinking to take place.
If the torch is moved across the flange apparently the shrinking effect will
not take place; due to the molecues in the metal by incorrectly aligned.
They even brought a photocopied paper showing the correct way to move the
torch, But I did notice the top of the page was describing the method I was
using but they do not have the previous page.
Apparently there are different recommendations of how to apply the heat to a
beam for depending on its end use; staticly loaded or dynamically loaded
So if anyone has seen a publication like this please send the info, I am not
worried about the cost so long as it is available.
I am rolling on the floor with laughter. It is as good as when a foreman
challenged me about how I could measure temperature with a magnet or that
when you heat steel the water rises out onto the surface.
I woud like to align some molecules myself... I might be up for a Nobel
science award if I can do it right.
Most code doesn't like heat shrinking beams and have quite low limits on
how hot you can get the material. The problem is that you cannot get a good
red heat with the limits required.
If you e-mail me I might be pursuaded to send you my snail mail
something I have found useful. (TWI stuff ) I hope the other stuff I send
was siilarly so.
Yes you are correct about the procedure handbook. I have a few of copies one
even from back in the forties great reading.
The one thing missing is a picture showing the movement of the torch and the
explanation that the molecules align with the torch movement and provide a
greater shrinkage force.
Though the technique I was using has work before my supervisor claims it is
impossible because they tried it and it failed.
One thing I did notice today was the difference in the stress lines formed
in the mill scale of the beam. I plan on measuring this tomorrow if the shop
is open to record the difference for future use.
As I mentioned to Randy I plan on conducting a few experiments with a couple
of six inch beams In order to have a comparison of methods, might prove
I have downloaded all the files on distortion from the TWI and saved them in
word format if anyone would like a copy I will send it to them via email. I
will transfer the format to plain text or pdf as needed.
Now back to searching for that elusive publication.
Have you got the TWI booklet Control of Distortion in Welded Fabrications?
Mine is a 1982 reprint. I am sure there is a later edition.
A classic mistake on large beams is to just heat the beam flange for a
length of say six inches but not put any heat in the web. If you do shrink
in this manner you will see the web buckle as the flange shortens. One
really has to heat a full pie section which includes the flange and down the
web to the opposite flange at least half way.
Molecules don't align as you heat... they vibrate! We are dealing with
expansive forces that cause plastic flow of the red areas. When those red
areas go black they contract bending the beam. It is important to
understand that when things take a bend yield point has been reached
somewhere on the structure.
People often do not heat a pie section concentrating on the beam flange.
Little will happen if you only heat the flange since it would have to buckle
the web which takes massive forces. I have seen buckles of 1/4 inch when a
two foot square is put against the web. The beam was cambered but the web
If you are looking at stress lines forming on the mill scale I will bet
dollars to doughnuts you are not heating the web????
I like flame shrinking beams but it is not very practical in dollar
cost. A better way is to cut the beams to length, load them up and send
them to a shop that cambers them with a press.
Are you using a air /water spray, wet rags or just letting the unit cool in
air? the most effective shrinking is done with two people, one on the torch
and one with the spray.
I cannot see how someone "aligns" molecules with a torch. Maybe change
shape of the grain structure but not aligning molecules.
Also if you don't know, heating with acetylene is best because the heat
is so concentrated but there are always flow rate problems. MAAP gas is
next best and then comes propane. I am always using oxy-propane but on the
few occasions I used MAAP I noticed the difference.
You could also e-mail the guys who put up that section on distortion
control. they might have some additional ideas. There is a credit at the
I searched the TWI for the booklet you mention but I must be going blind as
I cannot find it.
Thanks for the info you provided at least now I know I am not out in left
field. I did heat the web, first actually as they requested. Then the
This was the first time I ever noticed this type of stress line showing up
in the area heated. Probably never see it again.
This shop will not allow the use of a cooling mist on the steel as they are
worried about hardening the steel. Which makes for a very long process of
cambering any beams.
Your right about the molecules as they are heated they move around more; but
this area has a few old wives tails that are good for a laugh or two.
OLD WIVES TAILS
For example the reason for preheating steel "is to remove the water that the
steel has soaked up, Thicker section soak up more water hence require more
7018 cannot be used for open root welding
7018 is not suitable for vertical welding
7018 has to be pushed for max penetration
Visual inspection of a weld tells you nothing
Weave beads 2" wide are stronger than a stringer bead
100 amps with a 1/8" SMAW electrode gives you a 1/4" penetration, which is
why slag removal is not that critical in multipass welding.
Slag on the weld protects the weld from corrosion
SMAW gets more penetration than GMAW of FCAW
1/8" tacks with porosity are best as the bubbles help mix the tack and the
weld metals together.
6011 is stronger than 7018, which is why it is used it is used in root
Selfshielded fluxcore wire cannot be used in building fabrication
Metalcore wire cannot get porosity in the weld bead
The older the welding machine the better it runs over thirty years is the
Its impossible to start a 7018 rod that is at room temperature
Cast iron cannot be welded
These are just a few I can remember off hand
He used rich text format which will be quarantined by the dropbox
bot. I'm sure that when Steve gets a chance he'll check it out and
place it in the dropbox (probably convert it to standard txt in the
process). The reason for the restrictions is that people try to put
viruses in the dropbox so Steve restricts automatic files to the safe
I compressed the file before sending from the looks of things in my folder.
I just converted to plain text format but all the pictures were lost in the
May have to fire up the other computer and convert the file to pdf format.
John (or anyone else), a plea for clarification on a couple of the old
wives' tales -- there are a couple that I thought were true at least to some
If they meant that the slag was supposed to stay on indefinitely to prevent
corrosion -- sort of like an alternative to galvanizing -- then I certainly
see that this is ridiculous. But in terms of the welding process itself,
wouldn't it be correct to say that slag protects the molten metal from --
well, maybe corrosion is not the right word, but from the nitrogen/oxygen in
the air that would otherwise result in a weaker/more brittle weld? If the
former was meant, perhaps it could have started from a misunderstanding of
Okay, on this one I must have been taught the same old wives' tale -- in
fact, I am pretty sure I read something like this here. Of course, it would
have to be qualified: "all things being equal." Obviously one could do SMAW
in a way that would give less penetration than another approach with GMAW.
But I thought that, in general, SMAW could penetrate more than either GMAW
or FCAW?? Or am I just confusing that with SMAW, especially 6010/6011, being
more able to deal with rust & dirt?
Thanks to all for continuing my education ... just when I think I've learned
something, I find out there's an exception or another way to look at it or
(( Slag is for shielding the weld while it is in the molten state, once the
weld has coolled below approx 900 deg then the slag can be removed. Thought
the cooler the slag when being removed the better; nothing worse than hot
slag rolling down the inside of your shirt. Especially if it makes it into
Slag left on a weld for the entire life of the weld actually traps moisture
and along with its acidic nature helps speed up the corroding process))
((Amperage is more of a factor for penetration than the proccess; also the
size and shape of the arc cone has an effect.
When comparing directly, throw in a variety of shielding gases and things
can take a twist. Some gases provide a wide bead with shallow penetration
and others a narrower bead with deaper penetration.
Best if seen in a book so side by side comparisons can be made.
Pushing or pulling the electrode effects penetration
PULL "slight increase"
PUSHING " reduces penetration
And of course the polarity affects penetration))
This is one of those things that many variables have to be taken into
account For example I have been told many times that a 1/8" 7018 at 100 amps
gets 1/4" penetration or double that of MIG "GMAW" even when in spray
transfer even after blowing a hole through a 1/4" plate. His response was
pushing doubles the penetration and accused me of cheating.
The best thing I found was to run several beads side by side with a couple
kinds of electrodes and a couple of different processes all at the same
amperage or at least the minimum amperage the electrode can run. For example
120 to 125 amps and select as many 1/8 electrodes you can find. Set you wire
feeder up to run at 125 amps as well.
After running all the combinations you can saw the weld specimen into a
couple pieces across the weld.
Sand the welds to a nice smooth finish around 600 grit. Etch the ends of the
plate with a mild acid; this will make the welds stand out.
Now take a close look at the weld profiles with a magnifying glass or borrow
a microscope 100x to 300x is more than enough.
You can actually measure the weld penetration with a micrometer to see just
how different each electrode is for penetration
With a 7018, the slag definitely protects the molten puddle 'till it
freezes (and helps to support it too, particularly in vertical work).
The slag on a 6010 doesn't do much at all to protect the puddle, if you
look at it closely you'll see that it's pretty spread out with
considerable exposed weld.
Could also be that someone just used the word "corrosion" when
"contamination" might have been the better word.. and it could also be
that a thought extension was made- "slag protects weld.. slag good,
leave slag on weld.."
Regardless, it sounds funny, and funny sounding ideas are sometimes a
pain to the new guy with good intentions..
All things being equal, a wire feed will give better penetration than a
stick. Penetration is acheived with a fast and hot travel, the wirefeed
naturally allows this better than the stick. "Go nice and slow to get
good penetration" is a contradiction in terms which I've heard before..
You can build a special case for 6010, such as vertical work on a very
heavy piece, but I think the same case could be built for the wirefeed.
Metalmangling Mike and I discussed penetration to some extent and I have
come around to his view on some penetration issues; it's not that
important and, no matter what, there's times when it's detrimental,
sometimes you want minimal parent metal admixture in the weld.
I'm all for root pentration, it's possible to save time and material on,
say, a lap or tee joint in 3/16" stock by getting good root penetration
and depositing a narrower bead.. but if you're making multiple passes on
a heavy piece you should worry more about joint preperation and fitting
so that exceptional penetration is unnecessary.
Once you've made a good root, penetration is pretty much a waste because
you're melting metal that's already been deposited or doesn't need to be
melted, better to melt the minimal amount of what's already there and
put the heat into what your depositing. If you don't need to use xx18
that's where a 7014 or 24 really shines, fairly shallow penetration with
very high deposition rates.
That's pretty much how I look at penetration now- if you're single
passing stuff that's thin enough that you can melt the root to a
considerable fraction of the parent metal thickness then it's a good
thing.. but on heavy work, after the root pass, it's a waste of heat
that could have melted filler material.