I know this post is old but I have something to add for your concideration.
You have recieved some very good advice to date and after reviewing all that has been mentioned as the posts progressed I noticed 2 additional images #3347/#3352 of broken bend test coupons which cause me to write.
It's not hard to pass a bend test.
But it is hard to understand how the effects of welding prevent that from happening.
Those coupons showed a weld that in simples terms wouldn't stretch evenly. The images suggested a lacking of uniform ductility in areas of the weld and HAZ areas if I correctly matched the # of the pics to what it is I'm attempting to offer in explaning my thoughts.
Ductility, the ability of a metal to bend and stretch without breaking is all the bend test shows? I may be over simplifying things?
If an origional unwelded strip of metal was bent and passed what ever die was used to make the bend, then the test is to get a welded coupon to do the same is it not? Done right there is a good chance it will.
My simple answer, your coupons broke because some areas of the weld metal, adjacent HAZ material changed, became tougher, harder and seemingly less ductile during the welding process? Why? With no offense intended, because of how they were welded.
The process of GTAW seemingly is at your disposal as well a lab of equipment. I think your problem is one as previously mentioned a welding skills issue. You do show a very good understanding of what you need to do, a good degree of skill to do the task, however as some welders have mentioned you have a few issues and while small they need to be addressed. So really it then becomes one of education as to why your not passing the bend test if seemingly you should by your skill level in handling a torch?
If you took that torch head, clamping it into a track radiograph in-place of a cutting torch, you would have a mechanical device that would allow for consistent speed and a consistant and variable way of arc length control. Do that on a few samples and when your left with adding rod as the variable you'll see quick enough the problem and my thinking on the matter.
I think your arc length, the distance your Tungsten electrode is held off the plate, is excessive when welding. For a number of reasons this happens, but to create visibility comes to mind as the #1 reason. Your WPS indicates a 3/4" nozzle, 1/8" tungsten.
I'll hope to explain the basis of this opinion by saying the greater the arc length held when doing the weld, the greater the voltage (yes that depends on power source), typically lowering amperage ( again power source), causing your weld and base metal to be less heated, and the filler metal/ weld surface more fluid. The filler metal is liquid and runny rather then thicker and sluggish.
Don't missunderstand my meaning, it's heating, but more broadly on the surface rather than imparting into the base material.
Take a flash light, turn it on against a wall, then slowly pull it away.
I'm sure there is something educational in that to be learned? But if the light was a flame, what works better for setting the wall on fire? I think close/tighter to the wall but this is of course my opinion on how I see your welding issue. Closer it it the less of a flame required?
Your bead shows a lot of fine ripples, with a slight spreading of the weld pool. That suggests a steady consistant travel speed. A low current and longer arc slows the process down while building enough heat to get things to melt and flow with some manipulation. Add a drop, blend it out.
The appearance of a weld tells a bit of information but the more information a person has the less they guess when offering solutions.
You don't mention the degree of angle your tungsten is ground to a point. It as well this has an effect on arc length/arc column pattern profiles, penetration profiles, amperage required, effecting voltage as well. A simple variable that effects heat input, heat loss, speed of travel, penetration profiles. No one mentioned torch inclination, that can cause excessive distances in arc length. Tilt the torch, tilt your head? To often the torch wins.
Throw in the variable of an .045" Rod size to the ability of the welder to feed it preventing over heating or high interpass temps while welding, then heat and cooling of the weld pool and debate of the subject of ductility is again worthy of mention.
They say a large single pass is assumed to be more ductile a deposit than a multi pass weld. In theory, a hot wet thinly welded root pass followed by a heavy fill and two thin cover passes could be a recipe for trouble. Then again, a root that is placed with the thought of good fusion would cool slower, and fill slightly more penetrating less, following this thinking more is less and less is better in current. Your not trying to penetrate through the depths of unbeveled material in the quest for good fusion. It hold little value but to extend the HAZ. I don't think the problem was as much excessive amperage, but the appearance of it being too excessive because voltage is flaring with arc length and the filler rod not chilling the puddle enough.
Adding a filler rod and the speed at which a welder can do it is a reflection on his abilities and makes a difference in how hot the weld pool gets, how quickly it cools. In concideration of all the other variables, the choice of rod size has an effect on the heating and cooling of the weld pool so this does offer the possibilty of affecting the mechanical properties, changing or resulting from welding.
Your using .045" diameter filler. Got to feed a lot of that wire size to consistantly build a weld, fill a joint and control heat in put in the process.
I hope my offering and explaination are of some value?
But passing a bend test isn't rocket science or that hard, but again, understanding why it failed is.
When I ask myself why a weld isn't ductile enough to pass a bend test I have to think what causes toughness and hardness?
So...what was the problem? Welding Technique? Maybe...but it is an old post and with my spin on it may hold a degree of validity?
But I really don't know, I didn't see a reply explaining to me what the answer was? So I offer this as my guess for your concideration if your reading my late posted reply.
I also think, done properly the root should pass the bending.
After all, that metal just has to be ductile enough to stretch as the root doesn't have a notching effect with the over lapping/collar design of the joint surfaces and seemingly will support it self as it's pressed through the die if properly centered.
I'll offer up in thinking differently,somewhere between between penetration and fusion is melting a solid to a liquid. A root pass failure could result from excessive penetration into the base materials and with poor mechanical characteristics in this HAZ area having resulted from excessive arc length heating of adjoining material surfaces, on cooling, this material might be displaying increased toughness, weld hardness with less ductility or uneven ductility through the material. Some metal stretches, some holds firm, some gives excessivly while some resists and "we" wonder why it broke? Poor welding technique will cause this and elevated arc lengths fall under techniques.
Controlling the arc length when welding allows for the welding current to be stabilized to a value for control purposes when welding.
The heating, melting, filling, and cooling cycles a spectrum of characteristics to the metal so if variations in arc length occur, one could accept the outcome may not be as desireable or consistant and that will effect a bend tests out come.
If you don't concider the value of all the variables making up this relationship, you limit your ability to control the out come.
From a welding perspective you don't see it mentioned but arc length is a perception of depth skill. Most welders would say the number one to learn how to control, as well the variable that practice over comes in difficulty.
There is lots that can be said when discussing this problem. I only offer up my simple theory of arc length for concideration.
It's educational value however limited, hopefully offers a simple solution at it's basic level.
Your bent broken coupons while disappointing, keep showing improvement! Things are looking up.
I would hazard to guess that with a small gauged press you might hear a change to the sound of the coupons pressing into the die, I liked that part on test day. I called it the pinch a turd spectrum. Aaaaaggggggggggggggghhhhhhhhhhhh oh my fire truck that hurt... Or... sploooooosh sweet relief. Looking at the weld, it's size, shape and profile, it was never so much a question would they pass, but how they would sound. Some days some get lucky.
That's a bend test for you.
And in giving a shout out to those dedicated gentlemen who post regularly to the site, thank you guys. It's alot of work typing a response and some of you have such great minds with a much clearer ability to simply put to print what some struggle to express.
With understanding, please be gentle. I'm of a simpler mind.
Correct me if you find me to be off in my thinking, we are here to help and I have been wrong before.
Michael... you must have welded up quite a few by now and I'm left to assume almost a year later, your getting a lot of practice in? As far as things go, has practice made perfect? Have you been sucessfull in your attempts in bending those coupons?
The information you posted was regardless of the out come, wonderful to digest.
I liked the pictures.
I like pdfs.
I like learning from others.
Three strikes I'm out.