[exnailpounder]

I see in the original pictures a complete lack of root penetration. I also saw a comment that you can't get root penetration without the weld dropping out.

That's nonsense. I see in the broken weld picture you didn't even nick the edge of the tube. You're building a weld bridge, not making a weld joint. The root must break the edge of that tube if you want strength in the joint.

Steve S

That comment about full pen was me and the truth is you can't get complete pen on a filet. You can cut and etch any filet and see what penetration you get and they are never all the way through, now on a gapped butt weld you can get full pen. If you did get complete pen, that means you would be all the way through 3/16" of wall, into the inside of the tube and then you would have an overheated mess. I suggested the OP was dipping in the keyhole and not letting the puddles join. You have to dip in the keyhole on a butt joint cause thats all you have to dip in. I guess I had typing palsey when I wrote that...alcohol was involved.

[Olivero]

Thanks,


I've only got 230 Amps, so that will have to do.

In the past I've ended up with concave welds once I got things hot enough, needing a second pass.

I was hoping for small neat welds

You can still get "small neat welds" if you want. Just bevel the joint prior to welding, I have a Dynasty 200 machine, I build a fair amount of stuff in commercial use and some things are thicker than what 200 A can handle so you BEVEL it to make the thickness less right in the joint, just pick up a grinder and grind it in the corner and bring it in 1/8" or so on both sides or 3/16" on one side and weld the piss out of it and make sure you get that joint and you can have some really pretty welds by the end of it that won't have to be super big and ugly.

Don't even bother with the second pass. Not necessary if you bevel and it double pass looks like shit IMO.... Unless its pipe, then It's good.

[Otto Nobedder]

I see in the original pictures a complete lack of root penetration. I also saw a comment that you can't get root penetration without the weld dropping out.

That's nonsense. I see in the broken weld picture you didn't even nick the edge of the tube. You're building a weld bridge, not making a weld joint. The root must break the edge of that tube if you want strength in the joint.

Steve S

That comment about full pen was me and the truth is you can't get complete pen on a filet. You can cut and etch any filet and see what penetration you get and they are never all the way through, now on a gapped butt weld you can get full pen. If you did get complete pen, that means you would be all the way through 3/16" of wall, into the inside of the tube and then you would have an overheated mess. I suggested the OP was dipping in the keyhole and not letting the puddles join. You have to dip in the keyhole on a butt joint cause thats all you have to dip in. I guess I had typing palsey when I wrote that...alcohol was involved.

What I said didn't quite lead to what I meant. I meant penetrating the full depth of the fillet, fusing the corner of the leg , as opposed to bridging over it and leaving the corner intact. I wasn't trying to imply full penetration.
Sloppy language on my part.

Full penetration can be achieved on a fillet through prep work, by beveling the leg to a small land or a knife edge, though it's not always practical or useful. When both sides of a leg are welded, or in this case the tube is welded 100%, full throat depth is all that's required so your weld is as strong as it's size implies.

Steve S

[Gavin Melville]

Could one of make a simple drawing of how a beveled fillet joint might look - I must be thicker than usual this morning.

[Otto Nobedder]

I'll give it a shot. I haven't used MS Paint in a long time. Failing that, maybe a bar-napkin sketch to scan or photograph.

Edit... sketch it is. Can't draw a straight line with my trackball.

Steve S

[Gavin Melville]

A photo of a bar coaster would be fine. Ms paint is just not the tool - for anything.

[Otto Nobedder]

That took a minute... I forgot I needed to reinstall the drivers for my printer/scanner.

Here's a crude sketch, showing a tee joint, and a pipe-to-plate. I've drawn it with a land, rather than a knife-edge, since you're working with aluminum. A knife-edge would just blow out. The bevel will need to be at least 45* up to about 60, depending on the thickness of the material, so you can get in to the corner.

Working with Stainless would require purging the back side of the tee or the inside of the pipe/tube for the first pass, since the material is thinned out enough for full penetration.

Steve S

Scan.jpg (23.93 KiB) Viewed 1780 times

[Gavin Melville]

Ok, got that. I'll be at work in an hour and can try it then. I didn't think of that, on the surface it appears to make the problem worse.

[Otto Nobedder]

It's not really needed for what you're doing. It's seldom needed unless a procedure calls for it. The trick is to get the tungsten very close when you start your arc, and get both sides of the joint to puddle all the way to the corner. You may have to manipulate the torch, as the arc will favor one side. Once you're molten to the corner on both pieces, add your first dab of rod, drawing the torch back a bit as the puddle will swell rather quickly. Advance to the front edge of the puddle, and don't add rod until you see the previous addition start to flow into the corner. Lather, rinse, repeat.

Steve S

[Gavin Melville]

Thanks Otto,

Fillet joints have always been my nightmare. All other joints, and "open" fillet welds are fine. Once I get to 90 deg it just doesn't work so well. This thread has given me some good ideas.

[exnailpounder]

I see in the original pictures a complete lack of root penetration. I also saw a comment that you can't get root penetration without the weld dropping out.

That's nonsense. I see in the broken weld picture you didn't even nick the edge of the tube. You're building a weld bridge, not making a weld joint. The root must break the edge of that tube if you want strength in the joint.

Steve S

That comment about full pen was me and the truth is you can't get complete pen on a filet. You can cut and etch any filet and see what penetration you get and they are never all the way through, now on a gapped butt weld you can get full pen. If you did get complete pen, that means you would be all the way through 3/16" of wall, into the inside of the tube and then you would have an overheated mess. I suggested the OP was dipping in the keyhole and not letting the puddles join. You have to dip in the keyhole on a butt joint cause thats all you have to dip in. I guess I had typing palsey when I wrote that...alcohol was involved.

What I said didn't quite lead to what I meant. I meant penetrating the full depth of the fillet, fusing the corner of the leg , as opposed to bridging over it and leaving the corner intact. I wasn't trying to imply full penetration.
Sloppy language on my part.

Full penetration can be achieved on a fillet through prep work, by beveling the leg to a small land or a knife edge, though it's not always practical or useful. When both sides of a leg are welded, or in this case the tube is welded 100%, full throat depth is all that's required so your weld is as strong as it's size implies.

Steve S

Out of admiration...I won't make you apologize

[exnailpounder]

That took a minute... I forgot I needed to reinstall the drivers for my printer/scanner.

Here's a crude sketch, showing a tee joint, and a pipe-to-plate. I've drawn it with a land, rather than a knife-edge, since you're working with aluminum. A knife-edge would just blow out. The bevel will need to be at least 45* up to about 60, depending on the thickness of the material, so you can get in to the corner.

Working with Stainless would require purging the back side of the tee or the inside of the pipe/tube for the first pass, since the material is thinned out enough for full penetration.

Steve S

Scan.jpg

Ding dang...you can get a full pen weld on that bad boy.

[Gavin Melville]

Some progress. Back to basics - here is a butt joint welded the same as above, .1 in thick 6061, 100 amps (actually 70 was enough), 120 Hz. I assume that is OK, certainly strong enough.

The fillet joints are still eluding me, will post back when I've got one etched.

It might be the 6261 aluminium that is making things hard - fillet joints with the SHS shown here are better. It's thinner, and also 6061.

I'd like to see a fillet joint snapped or etched - anybody got one ?

[Otto Nobedder]

I will try to do a fillet cut and etch tomorrow, as time allows. I have a bit of material set aside, as I wanted to do one today, but I'm pressed pretty tight until after Friday.

The material will be 1/4" 5053, and I will likely weld it with 4043 or 4943, for a good contrast in the etch.

Steve S

[exnailpounder]

20170426_074100.jpg (51.18 KiB) Viewed 1681 times

20170426_080658.jpg (52.72 KiB) Viewed 1681 times

This doesn't show much but I did the cut and etch. One is tapered like Steve's drawing , the other is just a normal filet. The etching doesn't make the welds stand out really good but both ways get excellent penetration when you allow the 2 puddles to meet and then dip but but the regular fillet has much better penetration into the base metal. I think that is because with the taper, I had to move a little faster to keep from melting back that taper and it actually turns out to be a welding thick to thin scenario and you get a little less penetration on the ticker metal because you don't want to melt you thinner metal and you have to move out quicker. If you dip in the keyhole you are bridging the weld. I don't have any dye penetrant but I suspect you would be able o see the line where the two pieces meet on the normal filet much more than the tapered one.
Unknown base metal .125, 3/32" 4043 filler, 140 amps, pedal, 18cfh, 3/32" 2% lanthanated tungsten (micro balled)

[Otto Nobedder]

Nice, Jeff, and a great analysis on the base penetration. I think this could be limited with torch angle adjustments, but it may be finer control than a blind shaky guy like me can manage. I won't get to do a cut/etch before Saturday, as there's a priority job right now that's meant to be done Friday, and I always get the priority jobs...

What did you etch with?

Steve S

[exnailpounder]

Nice, Jeff, and a great analysis on the base penetration. I think this could be limited with torch angle adjustments, but it may be finer control than a blind shaky guy like me can manage. I won't get to do a cut/etch before Saturday, as there's a priority job right now that's meant to be done Friday, and I always get the priority jobs...

What did you etch with?

Steve S

I used Naval Jelly and when I thought that didn't do a good job I used oxalic acid( Barkeepers Friend) Not ideal but better than nothing and the details show.

[Otto Nobedder]

Nice, Jeff, and a great analysis on the base penetration. I think this could be limited with torch angle adjustments, but it may be finer control than a blind shaky guy like me can manage. I won't get to do a cut/etch before Saturday, as there's a priority job right now that's meant to be done Friday, and I always get the priority jobs...

What did you etch with?

Steve S

I used Naval Jelly and when I thought that didn't do a good job I used oxalic acid( Barkeepers Friend) Not ideal but better than nothing and the details show.

First thing I'll try is Alubabrite. It quickly etches clean aluminum, but I'm not sure it will contrast the weld nugget. We'll see. I know it works a treat on stainless steel that's had the CrO2 recently buffed off.

Also, I'll be welding known material. 5052 base with 4943 filler.

Will report results, of course.

Steve S

[Gavin Melville]

Don't laugh, but Drain Cleaner works well.

Just Sodium Hydroxide. Don't breath the vapour.

[cj737]

Aerosol oven cleaner too...

[Otto Nobedder]

Both those suggestions make sense. Alkaline, rather than acid, might reduce raw aluminum better.
We shall see.

Steve S

[Poland308]

New brite coil cleaner from any hvac supply house. It's alkaline and concentrated. They sell gallons for about $30. It's water soluble. I've seen it eat the aluminum fins off of a condenser coil if left on too long.

[pgk]

We use that to clean AC coils, works!

[Otto Nobedder]

I have a slow-developing etch underway. The oven cleaner appears to be the ticket. If I decide to do this regularly, I'll by a tub of lye.

Steve S

[LtBadd]

I have a slow-developing etch underway. The oven cleaner appears to be the ticket. If I decide to do this regularly, I'll by a tub of lye.

Steve S

Look forward to seeing this, doing a little reading about lye, when in contact with aluminum produces hydrogen gas.

Steve I've haven't done any etching but am interested in the process, any details are appreciated, maybe even a thread dedicated to this?