I'd start by tightening up your motions on your weave. The undercut is coming from not pausing long enough to fill and traveling to far forward with each cycle. Seems I remember having this conversation before with you possibly.

Try traveling only 1/2 to 1/3 as far forward with each side to side motion. You'll need to speed up a bit to get things to even out some. Again I don't think you are quite able to read the puddle yet and so aren't actually able to react to what is happening.

 

What he said /\


When I first started welding and couldn't yet read the puddle yet, I had instructor tell me to count out the hold on the edges. Your instinct is to pull away and move too quickly. Do a pass where you hold a long 2 count on each side all the way up, no matter what you think you're feeling or seeing.

Letting your angle go nuts can hurt too. With your side to side you wanna swivel a bit as you're swinging through the middle so your rod is pointing "at" the toe line on each side. But I'd concentrate on getting consistency on the timing first.

 

Yes sir, that was me. I was doing vert lap joints. I'll work on it.
thx, Gizzmo

 

I ran out of 3/8"plate so these are in 1/2" plate
I'm going to try and mill off the backer to take a look at the root.
Any better?
Gizzmo

 

Second attempt is a lot better.

When I do vertical up it's always been open root with 6010, but that's besides the point. And I do it on pipe, so it's actually a 5G... but anyways.

I would say ideally, you should be doing it in 3 passes. Root, fill, cap. The single fill weave should be laying down enough metal to get you just under flush... about 1/16th to 3/32.. maybe not as much as 3/32 but in that range. Tighten your weave up so you're keeping the tip of the rod in the top 3rd of the puddle (you know what I mean when I say puddle, right? Not being rude, just trying to clarify). I usually pause on the sides long enough to say "One Mississippi" in my head and zip over to the other side for another pause. You don't really want to "zip" or "whip" across the center, but don't hover or hesitate either. Just a straight shot, edge-hold, edge-hold, and repeat.

It looks like you've done a pretty good job of it on the thicker plate. But you aren't pausing quite long enough on the edges to let the metal fill in. Watch the edge of the rod where it meets the bevel and you'll see it undercut, and then fill in. This happens fast so don't dawdle. Once you see it fill in, weave on over to the other edge and repeat. I find a very tight Z weave is easier to do for me in this situation. Other people do a U motion. The only time I do that is when I tie in to the crater. It helps make it look seamless. If you start a Z weave at a tie in you'll have a low spot where you just cut across the gap instead of filling in the entire crater.

Well that was a long post for a little bit of information. Listen to DSW. Just pause longer on the edges. If you're reading the puddle you can see it fill in and know it's time to move. Keep your eyes open and don't get distracted by the arc.

 

The pic of the fill with the stringers was close. A properly executed weave would be sufficient too. you should fill the gap as much as you can and as smooth as u can. Make sure the top edges of the joint are still showing but not touched. I have heard of this last pass before the cap being called the "witness". Hopefully I remembered this and am not using the term wrong. You should learn how to make these passes smooth. But if your practicing, you've come this far, you should practice a perfect cap. so grind the fill smooth. Now with about 1/16th of the edge showing (1/16th witness) evenly and your plate has cooled down so u can touch it with bare hands and your amps bumped down a few, your ready to cap. Start with building a shelf. Moving the rod slowly you will basically be doing small horizontal welds, this is an important stage for eliminating undercut. Make your first horizontal weld with this shelf and pull it over the edge of the virgin edge of the original joint (witness) half a rod width, hold it there until it fills and the puddle catches up to the rid and fills, and a moment later move quickly across the middle up another half a rod width or so. Just paint its way up there. Hope that helps.

 

so grind the fill smooth.

If he is getting ready for tests. The grinder is out! I've taken hundreds of plate tests, never once was I allowed to use a grinder. On pipe tests yes.

 

Well I "thought" I knew how to read the puddle! I used stringers for the fill that time just to give it a try but liked the weave better as it seemed to be flatter. Thanks for all the great replies and suggestions.
To clarify- I'm practicing for take a test for "vertical-Groove Weld - on 3/8"plate" using 1/8 7018 for all passes. I'll also be testing in the 4G (overhead position). Beyond just passing the test I really want to learn how to do this correctly and your suggestion have helped a great deal and I appreciate it- Thanks!
I've adopted a rod angle of straight in to even slightly down for all passes. I seem to get a better look at what I'm doing and it keeps the slag behind the puddle. My viewing angle is from below the rod and just off to one side. I have been trying to prop were I can but there is a lot of "one hand welding" in these welds.
Doug247- I've been trying not to "clip" the corners of the prepared edge on the cap. Also I've been bumping the amps "up" on the cap trying to get it to tie-in better at the edges.I'll give the 1/2 rod overlap of the edge and a drop in amps a try - Thx.

Thanks for all the help. Time to prep more plates!

Gizzmo

 

If I haven't done any vertical up in awhile I have to remind myself to hold on the edges until I think the puddle is going to roll out, then slide across the middle, and do the same thing on the other side. I also have to remind my self to raise my shoulder when getting close to the top of the plate, other wise I have the wrong rod angle.

Reading the puddle is one thing, knowing what to do is another! You have to learn what to do for each situation.

 

Use something like vice grips on the work to prop yourself. And like CEP said, don't forget to raise your shoulder as you progress.

A lot of Inspectors won't let you use a grinder during a test.. or tell you to keep it to a bare minimum. They want to see what you can do without it. The only time I was allowed to use one (to grind, not wire wheel) was a 6G pipe test to grind down the 6010 root. After that I was only allowed to use the wire wheel. Never taken a structural test so I couldn't tell you, but keep working on it without using one. It will only make you better.

You're using 1/8 7018? I don't remember the amp range I used back in school... I know on a 5G I ran it around 85-88 on the cap, with 3/32 7018.

Experiment, and write down your best settings so you don't forget them. You are allowed to whip out a small notepad during a test.

And when you get to overhead... blah. I hate it. Seems a bit easier than vertical up to most people though. Just don't be afraid of the fire, wear leathers, and tape your gloves closed.

 

Yes vice grips come in very handy when making certain welds.

 

Yeah, I use all kinds of stuff to prop myself up.

A wrote a bit about it in this blog post...kind of how I overcome some of my physical shortcomings lol: http://oppositelock.jalopnik.com/a-steady-hand-and-a-good-eye-990215275

 

Today's practice-
I tried to keep your suggestions in mind today. Things went well and without any issues. my caps still need something but they look better to me anyway. As always please feel free to offer suggestions-I'm all ears!
These are on 1/2" plate , 1/8 7018 at about 110 amps.

Gizzmo

 

I also did a 3/8" set today.
Same settings

 

Still gotta pause more on the edges, everything else looks good. Except your bevel. What angle are you beveling at? Looks way too narrow. I think the average is 30-37.5 degrees.

On your weave, stop just inside of the bevel and pause. You'll watch the puddle roll over the edge. Keep your eyes open, to prevent those gaps where you didn't hit the edges. I make sure my rod touches the edge of the bevel, to know I'm over far enough.

Getting better each time.

 

These are right out of an AWS D1.1 Structural Welding Code book.

 

Because I didn't want to take any chances I milled those edges. They measure exactly 47 degrees (23.5 deg each side).I figured that was close enough for practice. The gap is 1/4" with 1/4" x 1" backer. I know the edges aren't just right yet. I do have to learn to let it "spill over" a bit. I tend to try and preserve the edges and I think that's my problem here. I feel like I'm closing in on being able to pass a test. I'm still hunting around some on just how much to fill on each pass. I'll continue to work on it.
Here's a picture of the back of the root after milling off the backer on the 1/2" plates.

Thanks again for the help,
Gizzmo

 

By-the-way, I was just wondering if it's permited to have the backer plate run long past the ends of the test plates? Anyone know?

Gizzmo

 

Typically the beveled pieces are 6 inches, and then the backer is 8 inches long. Gives you plenty of run off tab to work with.

 

backer bar can extend past on both ends of coupons. makes for a good place to fire up when starting and allows you to fill at end preventing the notch effect.

 

I've seen them both ways. Also taken tests where you had to carbon arc off the backer plate, clean the test plate up with a grinder and weld the back side too.

 

I have had this as well. I took a test where it had no run-off tab as well.

 

Looking way better. As said pause longer on the edges, longer than u think you need to. Also try making your motions like a slight U, going up a little on the sides (|__|). Its all part of learning how to read a puddle. Find a good video and study it, you can focus on what's going on without having to do the work.

 

Excelent, because that's the way I've been doing it! Thx
Cep - Is doing 1" thick plate much different than 1/2" ?Unlimited sounds like a good thing to have!
Doug247 - I'll really have to try some different things to fix these cap passes. It takes so long to prep the plates I get nervous taking chances with the unknown. I didi try that once but I let the spacing grow too much between each side-2-side motion.
I thought I had it when I made the fill pass pictured here but I let the cap get away from me.
Gizzmo

 

Around here all you see is 1-inch tests for structural welders.

If it takes you to long to prep your plates you need an Ellis saw. Only takes a few minutes to have them ready.

 

CEP- Nice tool! I'll have to keep it in mind. My instructor tells me the plates go out for testing with the backer on. Thats why he uses the 3" backer's. He also said there is no tail on it to start the arc. Todays plates are done with no tail on the backer.
Here's today's practice. I'm happier with this one over the past tries.I think I crossed the middle too fast and it created a bit of a valley but the undercut on the edges is gone. What do you think?

Gizzmo

 

I prefer my backing strip to extend an inch past the edges of the plate. When I do a vertical up plate test I will build up weld on the lower part of the backing strip that is extended out past the plate before putting on my cap. Build up weld where the edge of the plates meet the backing strip. This gives you a good base of solid metal to start laying your cap on. I do this because I weave my cap real hot with a real heavy puddle.

 

I don't know why but it just looks cold to me. Maybe practice a little more on your puddle control, your undercut issues are looking a lot better now.

that's my .02 :]

 

I would turn up at least to 125 amps. If you burn hotter and weave a little wider, your cap will smooth out, and you are less likely to undercut believe it or not. Just make sure you allow the piece to cool in between passes.