I may be having a bit of a wonky moment here, but following on from AlanB's thread on tube notching, I have a question that may, as the title suggests, be a bit thick, but I'm going to ask it anyways!
Basically, if you could get a perfect fit between two tubes/plates/bits of metal in general, would you actually need filler rod if TIGing? IE, could you just set the welder so that you get 100% penetration without filler at all? If not, why not, as the way I am seeing it (probably wrongly) is that the metal is already a perfect fit, so 100% penetration should be possible without having to add filler at all. Off you go then!

That is a valid technique with tig & gas.

For a structural weld the bead gives the joint added strength and so a filler rod would have to be used to add the additional material. For a non structural joint such as an exhaust the filler rod could be missed if the joints were close fitting.

Davie

For Tig it works but this is used mainly for machine 'robot' welding where you can control the heat and duration precisely.

At home in the garage better to put a on a champher and use a filler rod.

Regards Mark

I'd have thought (and have always been told ) that a small amount of filler rod would be needed due to the fact that molten metal will sag due to gravity and a weld pool in itself is usually concave. This would cause an undercut of the parent metals and a stress raiser. All in my opinion of course.

Lap joints and corner joints in thin material up to about 1/8" should weld alright with no filler. In the case of a tube cut at an angle the different thickness will give trouble as the end of the angled tube will be much thinner at one point and thicker at others than the wall of the tube you are welding it to.

As mentioned earlier it's a common process in automated welding...even got it's own name.......autogenous welding..

Having done a bit of welding myself (and thinking back to some "Mechanical Science" theory studied only some twenty years ago), the theory goes something like this:
Weld joints are potentially weak points as any reduction in cross section causes a massive increase in localised stress (stress raisers) This has the effect of magnifying any stresses travelling down the parent metal causing the joint to fail. To avoid this, the weld should at least have the same cross sectional area as the parent metal. This is acheived by either forcing the parent metals together (as in resistance or friction welding) or by adding filler material. Many structural welds have excess material added to the weld area to compensate for any inconsistencies in the weld.



See?

From what i have done it all depends on the thickness of material, type of material and angle and gap of the 2 parts.

Take stainless steel, if no gaps are present on 1.6mm material you can 'RUN' the material in without the use of a filler rod. Doing this takes time due to the fact the longer the weld the hotter the material gets and the faster you have to move the more you do. If not the material sags and a hole appears.

Aluminium at 1.6 mm is pointless as its a very 'saggy' material so a filler rod is needed. HOWEVER, if the 2 parts are perpindicular (how do u spell that?) you can run it in as the sag runs onto the other piece.