Welding; How much undercut?

Some of the welding manuals I have say that a certain amount of undercut is
inevitable. Certainly in examining the fuselages of some factory aircraft
I've seen some shockingly large errors like this in the finished welds. On
many of my practice wleds, particularly on tubing clusters, an otherwise
perfect weld has a tiny bit of undercut. I'm talking barely visible here,
only just able to catch the end of a rod, for instance. At a guess i'd say
no more than .02 inches, maybe a bit more. Is this an acceptable situation?
If anyone's interested, maybe I can post some pics of one on a wesite but I
don't think the undercut will show up on a pic, it's that little...

Fortunat1 wrote:
Some of the welding manuals I have say that a certain amount of undercut is
inevitable. Certainly in examining the fuselages of some factory aircraft
I've seen some shockingly large errors like this in the finished welds. On
many of my practice wleds, particularly on tubing clusters, an otherwise
perfect weld has a tiny bit of undercut. I'm talking barely visible here,
only just able to catch the end of a rod, for instance. At a guess i'd say
no more than .02 inches, maybe a bit more. Is this an acceptable situation?
If anyone's interested, maybe I can post some pics of one on a wesite but I
don't think the undercut will show up on a pic, it's that little...

A textbook on welding MAY have excerpts from various welding codes on
how much undercut is allowable, here's one such:

The AWS D1.5 Code requirement for undercut is:
· Undercut shall be no more than .01 inches deep when the weld is
transverse to tensile stress. (Example: If a cross-frame angle is
welded into the web of a beam then the allowable undercut along the
edge of the weld touching the beam is .01 inch)
· Undercut shall be no more than 1/32 inch deep for all other cases

"flybynightkarmarepair" wrote in
oups.com:


Fortunat1 wrote:
Some of the welding manuals I have say that a certain amount of
undercut
is
inevitable. Certainly in examining the fuselages of some factory
aircraft I've seen some shockingly large errors like this in the
finished welds. On many of my practice wleds, particularly on tubing
clusters, an otherwise perfect weld has a tiny bit of undercut. I'm
talking barely visible here, only just able to catch the end of a
rod, for instance. At a guess i'd say no more than .02 inches, maybe
a bit more. Is this an acceptable situatio
n?
If anyone's interested, maybe I can post some pics of one on a wesite
but
I
don't think the undercut will show up on a pic, it's that little...

A textbook on welding MAY have excerpts from various welding codes on
how much undercut is allowable, here's one such:

The AWS D1.5 Code requirement for undercut is:
· Undercut shall be no more than .01 inches deep when the weld is
transverse to tensile stress. (Example: If a cross-frame angle is
welded into the web of a beam then the allowable undercut along the
edge of the weld touching the beam is .01 inch)
· Undercut shall be no more than 1/32 inch deep for all other cases


Thanks, that's what i needed to know OK. But from a practical
standpoint, is it normal to have such an undercut on a weld? I would
imagine undercut would be a stress riser, for starters.

edge of the weld touching the beam is .01 inch)
· Undercut shall be no more than 1/32 inch deep for all other cases


Thanks, that's what i needed to know OK. But from a practical
standpoint, is it normal to have such an undercut on a weld? I would
imagine undercut would be a stress riser, for starters.


I was hoping someone would come along with some solid data on this one, but
if you are talking about an aircraft weld, I hesitate to think ANY undercut
should be acceptable.

Are you welding tubing or plate? Aluminum or steel? Gas, stick, mig or tig?

Usually, undercut is caused by a combination of factors. Weld speed, deposit
rate and weld position, are common factors.

"Maxwell" wrote in
:


edge of the weld touching the beam is .01 inch)
· Undercut shall be no more than 1/32 inch deep for all other cases


Thanks, that's what i needed to know OK. But from a practical
standpoint, is it normal to have such an undercut on a weld? I would
imagine undercut would be a stress riser, for starters.


I was hoping someone would come along with some solid data on this
one, but if you are talking about an aircraft weld, I hesitate to
think ANY undercut should be acceptable.

Are you welding tubing or plate? Aluminum or steel? Gas, stick, mig or
tig?

Usually, undercut is caused by a combination of factors. Weld speed,
deposit rate and weld position, are common factors.


Yeah, for the most part, I can control the puddle quite well and my
welds are tidy enough. It's Gas, I'm talking steel tubing mostly and
it's in clusters, mostly in the "armpit" where there's an acute angle
between the tubes. I've been practicing in anticipation of finish
welding my already tacked fuselage.
If the undercutting is a problem, I think it must be in the way I'm
viewing the puddle. that is to say I'm consciously getting just a bit
too much penetration in those ares where the bead is more concave. I'm
afraid if I don't do this I'll end up with a cold weld.I'm feeding rod
in at a good rate, I think, and I also think I have the tip size and
pressures about right (it's a little difficult to get the bead started
in the tighter angles) and the weld runs fairly quicly once I'm out of
the tight area. The bead is fine once I'm running up the outside of the
longeron.. I've tried varying the angle of the tip in relation to the
work, everything from near vertical to angled well towards the progress
of the weld, but this only seems to give fine control. Less heat than
I'm using makes for a very globby, cold start. Having said all that, I
picked up my torch for the first time in a year and a half last month
and have made huge progress in welding in that space of time, so maybe
I'll get a handle on this through practice. OTOH, i don't want to
continue practicing a mistake!

In a clusters there is a lot of heat build up necessary because so much
metal to draw the heat away. It is a good idea to switch to the next larger
diameter rod when welding clusters than when welding single tube to single
tube junctions. The larger filler rod helps to suck a little of the heat
away from the tube so that the tube doesn't get so hot that undercut forms
as you weld.

Undercut is a stress raiser, but in a standard truss-frame tube fuselage
built with 4130 there is a major overkill situation. The tubes individually
just aren't stress enough to lead to cracks where there is minor undercut.

That said, I do not think you can allow any undercut in the firewall
structure which is holding the weight of the engine, or in the engine mount
itself. I would also avoid it in the landing gear structure.

If you have the tube up to the red heat level (gets that slight "wet" look)
then concentrating the heat on the filler letting the puddle expand outward
to incorporate the tube, you will accomplish a solid weld with no cold lap.
For the record, as much heat as is put into the base metal when using an
oxy/acet torch, cold lap is virtually impossible. Not saying it can't be
done, just not likely.

Fortunat1 wrote:

"Maxwell" wrote in
:


edge of the weld touching the beam is .01 inch)
· Undercut shall be no more than 1/32 inch deep for all other cases


Thanks, that's what i needed to know OK. But from a practical
standpoint, is it normal to have such an undercut on a weld? I would
imagine undercut would be a stress riser, for starters.


I was hoping someone would come along with some solid data on this
one, but if you are talking about an aircraft weld, I hesitate to
think ANY undercut should be acceptable.

Are you welding tubing or plate? Aluminum or steel? Gas, stick, mig or
tig?

Usually, undercut is caused by a combination of factors. Weld speed,
deposit rate and weld position, are common factors.

Yeah, for the most part, I can control the puddle quite well and my
welds are tidy enough. It's Gas, I'm talking steel tubing mostly and
it's in clusters, mostly in the "armpit" where there's an acute angle
between the tubes. I've been practicing in anticipation of finish
welding my already tacked fuselage.
If the undercutting is a problem, I think it must be in the way I'm
viewing the puddle. that is to say I'm consciously getting just a bit
too much penetration in those ares where the bead is more concave. I'm
afraid if I don't do this I'll end up with a cold weld.I'm feeding rod
in at a good rate, I think, and I also think I have the tip size and
pressures about right (it's a little difficult to get the bead started
in the tighter angles) and the weld runs fairly quicly once I'm out of
the tight area. The bead is fine once I'm running up the outside of the
longeron.. I've tried varying the angle of the tip in relation to the
work, everything from near vertical to angled well towards the progress
of the weld, but this only seems to give fine control. Less heat than
I'm using makes for a very globby, cold start. Having said all that, I
picked up my torch for the first time in a year and a half last month
and have made huge progress in welding in that space of time, so maybe
I'll get a handle on this through practice. OTOH, i don't want to
continue practicing a mistake!

--
Bruce A. Frank, Editor "Ford 3.8/4.2L Engine and V-6 STOL
Homebuilt Aircraft Newsletter"
| Publishing interesting material|
| on all aspects of alternative |
| engines and homebuilt aircraft.|
*------------------------------**----*
\(-o-)/ AIRCRAFT PROJECTS CO.
\___/ Manufacturing parts & pieces
/ \ for homebuilt aircraft,
0 0 TIG welding

While trying to find the time to finish mine.

"Bruce A. Frank" wrote in
:

In a clusters there is a lot of heat build up necessary because so
much metal to draw the heat away. It is a good idea to switch to the
next larger diameter rod when welding clusters than when welding
single tube to single tube junctions. The larger filler rod helps to
suck a little of the heat away from the tube so that the tube doesn't
get so hot that undercut forms as you weld.

Undercut is a stress raiser, but in a standard truss-frame tube
fuselage built with 4130 there is a major overkill situation. The
tubes individually just aren't stress enough to lead to cracks where
there is minor undercut.

That said, I do not think you can allow any undercut in the firewall
structure which is holding the weight of the engine, or in the engine
mount itself. I would also avoid it in the landing gear structure.

If you have the tube up to the red heat level (gets that slight "wet"
look) then concentrating the heat on the filler letting the puddle
expand outward to incorporate the tube, you will accomplish a solid
weld with no cold lap. For the record, as much heat as is put into the
base metal when using an oxy/acet torch, cold lap is virtually
impossible. Not saying it can't be done, just not likely.


Thank you Bruce! As always, you are the man..

I went out and did a bit more. I had my welds pretty good a few years
ago and I went out determined to recapture that. Increasing the feed of
filler worked wonders and all signs of undercut are gone..
I think I was afraid of not getting the penetration I needed and was a
little too agressive with the puddle as well. At the moment,I'm getting
a nice bright glow from the parent edges just before the puddle runs
over them so I'm confident there's good penetration (confirmed with the
hacksaw) wheras before I was looking for a virtual lava river which was
obviously causing the undercut.

Welds are pretty good, but I am having one further problem; After a
minute or tow, my torch (a meco) often tends to go oxy, even though I
had it set right in the first place. Not a big problem once I spot it
early, but annoying al the same. It seems to stabilise after a while.
Could be that my acetylene tank is just running down, but I thought that
the regulators would cover that almost until the gas ran out.. Maybe I
need new regulators..

"Fortuant1" wrote

Thank you Bruce! As always, you are the man..

Yep; it's handy to have a welder man nearby, isn't it? :-)
--
Jim in NC

"Bruce A. Frank" wrote in
:

[Image]

^Here is a picture of the work of a new welder who is worried that he
may not be up to the job of welding his fuselage together.

thanks! That's a good looking weld.. Sometimes I do that well, other
times not..;(



As for the Meco, try tightening the packing gland nut on both the oxy
and acet valves. I find that most torches that won't hold the flame
stable suffer from loose packing nuts.


OK, did that. Seems to have helped!

Fortuant1 wrote:


Welds are pretty good, but I am having one further problem; After a
minute or tow, my torch (a meco) often tends to go oxy, even though I
had it set right in the first place. Not a big problem once I spot it
early, but annoying al the same. It seems to stabilise after a while.
Could be that my acetylene tank is just running down, but I thought that
the regulators would cover that almost until the gas ran out.. Maybe I
need new regulators..


I had the Kent White set and the Medalist acet regulator would hunt and
surge and finally sent it back. Went out and bought another one locally
which seems to work well. His video is highly recommended.

I find the acet needs a few seconds to stablize any time there is an
adjustment on that side.

I have a Meco too and it is great, but seems to pop about as much as any
other torch. Welding 049 with a N-2 tip I had the gas turned way up and
it would still pop. May just be dirty.

Kent also includes and recommends E70 rod but it is terrible rod to gas
weld with and I switched to coat hanger to practice. Went out and got
some RG45 which is wonderful.

I can make more or less airworthy welds now, now working on making them
look nice. Things got a LOT easier once I learned the importance of
using the rod to help control the heat applied to the puddle so now I
almost never get burnaways. Still having to do a lot of stopping and
"adjusting", but about half the time now I can find that sweet spot
where the puddle is just right, the torch angle and heat is just right,
and I just have to move it along feeding the rod and the resulting bead
looks like a tig bead.

John