Dusting off the cobwebs

[Mac]

For fauld lame one, I cut out rectangular add-ons that were about the same width as the lame. There was some trouble with cooling cracks, but I eventually got them pretty sound.

For the second lame, I cut add-ons that were a bit wider than the lame in an effort to make the ends of the weld a bit less less prone to notching. I figured the additional thermal mass at the ends of the weld would help to keep me from overheating things. Well..... that part worked OK, but the resulting "step" produced a stresss concentration, and they cracked with wild abandon. I spent a certain amount of time filling the cracks, but to no avail. I figured the best thing to do was to cut off all the overheated and cracked stuff and weld on a larger piece.

The red marker shows the area that had gone all cancerous. By this time, I had ground away the troublesome "step".




This is what those guys looked like when I bent them to 90 degrees in the vise. The one might have survived the rest of the manufacturing process, but the other is pretty nasty. The less said about these the better.



Cut one off and weld the other on, and "Bob's yer uncle". Or so I thought.



Indeed, I was so confident that I took a pic of the filler "rod" manufacturing process.



I eventually got them welded satisfactorily.



For the third lame I had a "cunning plan". I figured that what I needed was for the add-on to overlap the lame a bit, so as to not have a weld running straight out to the edge.



I still ran into trouble. But, this time I was sure it was not the weld as such that was failing.



I subsequently adopted a plan of doing the ends of the welds first, and then heating the adjacent areas so that the weld and any area that seemed subject to tension would cool first. That was not a "cure all", but it was a step in the right direction.

Here is what I have now. The black lines are where I plan to trim. There is almost enough overlap, but I should really add about the same amount of material to the backplate and cullet. Some further welding experiments on strap are called for before I try that.



Mac

[Keegan Ingrassia]

Your fight with the cracks notwithstanding, Mac, I'm always impressed at how you make your welds disappear. I really look forward to trying your method when I have a better heat source.

[marcos.blues]

Really, it's excellent work in welding
Congratulations

[Mac]

Thank you, KI!
Obrigado, Marcos!

I was hoping to use this opportunity to demonstrate how to make a weld disappear, but I am having trouble, and that will have to wait for another time.

Mac

[Mac]

I have not done any experiments yet, but I have been thinking. This sort of cracking is new to me, but I can't recall the last time I welded extra material onto a lame like this. I think the problem something to do with the shape.

I made up a little sketch that illustrates what is happening. We begin with a lame, and that lame is curved in two directions. It's curved along its length, and it is also curved a bit across its width. The add-on is curved to match. In the bottom sketch, the add-on has been tack welded in place, and it lines up nicely with the "flow" of the lame.

The upper sketch shows what it ends up looking like after I have run the weld. The add-on is bent distinctly up at the line of the weld. The amount of bend has been exaggerated in the sketch to make is easy to see and think about.



Now... if we were to take an unwelded lame and try to bend it into that shape, we can all agree that there would be tension along the line of the bend where it reaches the edges. Since my welds are cracking from the edges, I think that this must be the mechanism. As the middle of the lame cools and the metal shrinks, it forces the lame into the shape of the upper sketch, and this causes the (still soft and weak) metal of the weld to seperate in tension. The bending is being driven by shrinking metal in the middle of the lame. This throws the edges into tension as well, and somewhere in between must be a sort of "hinge" in compression. The cracks begin at the edges and propagate inward toward that area of compression until the reach a point where the tension is no longer greater than the metal's strength.

It now remains to devise a procedure that allows us to weld the add-ons without putting the edges into tension. My first thought is to begin by flattening the end of the lame before attempting the weld. If that is successful, the curvature can be restored and continued through the add-on once the weld is cleaned up.

Mac

[Jeremy.G]

Thanks very much for posting your experiences. Good stuff...

[coreythompsonhm]

The copper mass ( not a thin plate) is used during the welding process to distribute the heat and help prevent warping of thin sheet metal plates . It adds mass to the welded material w/o the chance of being welded along with the steel to help make the weld more uniform. Vermiculite won't quite do that. :^o

Knowing Mac, he is using an oxy/acetylene welder. Real easy to keep the surrounding metal up to temp unlike mig. As he is using 1050 and experiencing cracking after the weld is done and cooling, it is cooling too quickly and it will crack in a stress point. As welds at the end of thin sheet are never perfect until ground, this will be its point of stress.

[Kristoffer]

Could you keep the torch on and apply some heat after the weld is complete to make the cooling process slower?

[Mac]

Time for a little experiment.

I made up a "lame" out of the same .059" 1050 stock by just cutting out a rectangular strip and hammering it up a bit from the back.



Then, I cut off a piece about as big as my troublesome ad-ons, and welded it right back on again.



This resulted in the all too familiar crack, as well as a bit of the characteristic back bend.



So far, everything is good. I have reproduced the problem under experimental conditions. If the weld had failed to crack, I would be forced to look to the Supernatural for the source of the trouble.

So... for my next test, I cut off another piece and flattened both of the ends that were going to be welded back together.



Unfortunately, that did not fix the problem. It still cracked.



In both cases, however, it was the first edge to be welded that cracked. The second end appears to have survived the cooling intact. My recollection of yesterday's failures is also that it was the first edge that was more likely to crack. I also see that the cracks extend much farther than my original theory would predict. So, now we are back to the idea of controlling the cooling conditions. A quick google search for 1050 weldability is bringing up all sorts of admonitions about pre and post weld heating. I guess I've got to try that and see what sorts of results I get.

As I said before, this is pretty vexing. When I weld helmet parts together for a "can" construction, there is no trouble. Likewise, welding the cuffs onto hourglass gauntlets is always successful. This cuirasse has to be ready to try back on in seven days, and I haven't got time to fail.

Mac

[Mac]

I'm not ignoring you guy's suggestions about slowing the cooling rate, etc. I'm gathering information and trying to figure out the crux of the problem and how best to solve it with tools at hand.

Thanks all!

Mac

[Duane W]

Mac , this could be tricky as crap to do and a little more clean up, but it might be worth a try to do a heavy tack on both ends, flip the piece over and do a proper weld on the reverse side. Welding from the middle outwards may help as well as long as the speed and material flow are consistant. Other than that, I got nothing.

D-wayne

[Tom B.]

Mac

It looks like the factors involved it this type of crack formation are "effective" carbon content, thickness, & hydrogen availability. These dictate how much preheat / postheat if any is needed.

In your examples of the helmets and gauntlets the carbon content and hydrogen availibility should have been the same as your current project. How much thicker were these projects than this one?

[Kristoffer]

I hope you are certain all material you are using are 1050 so you aren't mixing alloys. When odd things start happening its always good to rule out those simple things that are easy to miss.

[Mac]

I'm getting better results, but I don't feel like I have exactly conquered this problem. Playing the torch on the edge of the lame on either side of the weld as soon I have that end of the bead completed seems to help a lot. This lets the weld do some of its contracting while there are adjacent areas of hotter metal to draw material from . So long as the weld cools faster that the HAZ, things mostly work.

The other things that seem to improve the situation are using a smaller torch tip, and being more meticulous about not letting anything oxidize. This last one is touch with gas.

Mac

[RandallMoffett]

Your understanding of the welding force is very great. Much greater than my own. I have enjoyed learning about it. Best of luck with the fractures. They are truly annoying and I do not wish them on my worst enemies.

RPM

[Ian BB]

Seeing this thread progress is enough for me to get back into armouring. Eternally grateful Mac

[stephenschaeffer]

Mac, you are an inspiration both for your skill and knowledge and for your perseverance.

[coreythompsonhm]

Mac-
I am wondering that if playing the torch over the ends of the weld is helping, that heating the end of the mae to annealing temperatures and a slow cool down in a heat retaining substance as the annealing process usually dictate will help let the old and newly added piece shrink back to its cold state at a consistant rate, possibly preventing or reduce stress fractures.

1050 is finnicky, I hope this won't be more of a struggle than it has already proven.

[Mac]

I took a day off today because I felt bad, so there's no further progress to report. I hope to return to this tomorrow with some tests on the .042 stuff that the back is made of.

Mac

[Mac]

Mac-
I am wondering that if playing the torch over the ends of the weld is helping, that heating the end of the mae to annealing temperatures and a slow cool down in a heat retaining substance as the annealing process usually dictate will help let the old and newly added piece shrink back to its cold state at a consistant rate, possibly preventing or reduce stress fractures.

1050 is finnicky, I hope this won't be more of a struggle than it has already proven.

Corey,

I don't think I can get the weldment out of the clamp and into the vermiculite fast enough to do any good. As far as I can tell the cracks are forming at somewhere between a dull orange and a bright red heat. That's between five and ten seconds after I complete the weld and take the torch away.

Mac

[Mac]

Thank you, Randall, Bowen, and Stephan!

Mac

[coreythompsonhm]

Mac-
I am wondering that if playing the torch over the ends of the weld is helping, that heating the end of the mae to annealing temperatures and a slow cool down in a heat retaining substance as the annealing process usually dictate will help let the old and newly added piece shrink back to its cold state at a consistant rate, possibly preventing or reduce stress fractures.

1050 is finnicky, I hope this won't be more of a struggle than it has already proven.

Corey,

I don't think I can get the weldment out of the clamp and into the vermiculite fast enough to do any good. As far as I can tell the cracks are forming at somewhere between a dull orange and a bright red heat. That's between five and ten seconds after I complete the weld and take the torch away.

Mac

Hmm, that is sort of an odd predicament. I would have predicted between dull red and black, or immediately after black. I will need to ponder this and talk to one of my professional welder friends. They might have a small insight into the matter.

[Sean Powell]

When welding stainless we rarely use the identical filler rod. We use a rod that bonds well to the base material. I'm curious if using 1018 or less 'mild' steel filler rod would be a better choice. The weld wants to be ductile during the cooling process and not recipitate out any carbides. The overall strength of the fauld won't greatly suffer from a small low-carbon area. It may be worth testing to round-out the theory.

Sean

[Signo]

Another way to slow down the cooling process could be of simply keep the torch on the piece, moving it around the weld and increasing the distance slowly. Other than dilatation a piece so thin and small can temper fast enough to harden, so, brittle material + fast cooling = crack. Unfortunately this could mean that the warming up the piece before the weld, and the cooling down after, could last much more than the welding itself. If I would be able to weld with a torch, I would use mild steel in this case, the part is marginal and probably will never be stressed enough to need that bit of carbon , as Sean pointed out, mild steel would give you some forgiveness.

[Arrakis]

Another way to slow down the cooling process could be of simply keep the torch on the piece, moving it around the weld and increasing the distance slowly. Other than dilatation a piece so thin and small can temper fast enough to harden, so, brittle material + fast cooling = crack. Unfortunately this could mean that the warming up the piece before the weld, and the cooling down after, could last much more than the welding itself. If I would be able to weld with a torch, I would use mild steel in this case, the part is marginal and probably will never be stressed enough to need that bit of carbon , as Sean pointed out, mild steel would give you some forgiveness.

I think this is all very good advice, especially preheating and very slow cooling by circling the torch and backing it away over the course of a minute or two.

[Mac]

I got a bit of edge trimming done this morning before running off to a Doctor's appointment. These things can really chew a hole in your day.

The first trim was intentionally a bit conservative, and I've made some lines on the upper edges of the fauld that need further trimming. We are looking for graceful sweeps here.



Mac

[Mac]

When welding stainless we rarely use the identical filler rod. We use a rod that bonds well to the base material. I'm curious if using 1018 or less 'mild' steel filler rod would be a better choice. The weld wants to be ductile during the cooling process and not recipitate out any carbides. The overall strength of the fauld won't greatly suffer from a small low-carbon area. It may be worth testing to round-out the theory.

Sean

Sean,

That's a good avenue of exploration. Dave Rylak also suggested that to me the other day on the phone. I gave it a try with mild steel rod, and had pretty similar results. It's all got me thinking, though. Later today I'll try just running a bead without any filler rod and see what sort of trouble I do or do not have with that.

Mac

[Mac]

Another way to slow down the cooling process could be of simply keep the torch on the piece, moving it around the weld and increasing the distance slowly. Other than dilatation a piece so thin and small can temper fast enough to harden, so, brittle material + fast cooling = crack. Unfortunately this could mean that the warming up the piece before the weld, and the cooling down after, could last much more than the welding itself. If I would be able to weld with a torch, I would use mild steel in this case, the part is marginal and probably will never be stressed enough to need that bit of carbon , as Sean pointed out, mild steel would give you some forgiveness.

Signo,

I gave that a try a few days ago. Just keeping the weld hot seems only to delay the cracking.

I've had better luck playing the torch to either side of the weld, so that the weld can draw material from someplace hotter and softer as it shrinks.

Mac

[Mac]

I just did two little experiments. The first one demonstrates that the nature of the filler rod is not the trouble. I ran five "welds" across this plate, using no filler at all. I let things cool in between "welds". I started at one edge and went right across to the other. There was no pre or post heating. This was intended to produce "worst case" results, and I was not disappointed.

Three of the "welds" cracked badly. The one that was about 3/4" from the edge only cracked at the end. The one that was about 3/8" from the edge did not crack at all.

So... it's not about the rod.

But that leads us to wonder about the last case, where there was no cracking at all.

I am presuming that in this last case, there is not enough material outboard of the weld to put up a fight. That material ends up heated to nearly welding temperature, and cools at about the same time as the weld. If the weld needs to suck in some more material, it just takes it from the narrow band of still hot material. That material gets deformed, but the amount is so small as to be inconsequential. This is consistent with my experience earlier, where the narrower add-ons for the breastplate welded without much trouble.




The other experiment involved a different shape for the add-on. I figured that the narrowness of the ends of this add-on would partake of the nature of the last "weld" in the other test plate. That is to say, if it's narrow enough, it's not strong enough to crack the weld.



This worked, but it was difficult to control the weld puddle at the edges. I used no post-heating in this weld, but just went from the center to the edges, and repeated that on the other side.



Mac

[Kristoffer]

Interesting to see the behaviour for different conditions. What happens if the weld is longer, for example if you cut the plates diagonally? Perhaps cutting the edges to a V shape << would be easier then the rounded form?

[Cet]

The bead nearest the edge on ( the one that didn't crack) was that welded first or last?

[Mac]

Dave,

The one by the end was the last one. I was inspired by the limited cracking of the second one (which was welded first) and thought I'd make one more that was even closer to the end. In any case, everything was a black heat before I began each weld.

Mac

[Mac]

Interesting to see the behaviour for different conditions. What happens if the weld is longer, for example if you cut the plates diagonally? Perhaps cutting the edges to a V shape << would be easier then the rounded form?

While I think any of those would work as well, I suspect that none of them would work any better. So long as the welds approach the edges a steep angle I don't think it much matters if it's a line or a curve. The rounded thing is an attractive option because it would be easy to prepare the old lames by just rounding the corners.

I guess I need to practice controlling my weld puddle. I'll go try some more of those right now.

Mac

[Mac]

The sides of the backplate and cullet are all covered with lines. Here's the code. The red lines represent where the parts came together six (?) years ago. They have been touched up recently. The green lines represent how far the cuirasse would close last week. The black line is where the newly added to breast and fauld come to on the back when in the same position as it was last week. That is to say, the difference between the green and black lines is the stuff I just welded on.

As it stands, there is almost enough overlap to go with, but I really want to increase that overlap a bit by adding to the back and cullet. More overlap will make the armor easier to put on, and will serve as a bit of a hedge against further weight gain.



Using my conclusions from the welding experiments, I decided to use a rounded add-on form the waist lame.



Clamped up...



...and welded. This is not a pretty weld, but it is sound and it will clean up OK...



...with some grinding and hot hammering. Those low areas you can see in the upper part get picked out from behind, and driven up to match the front surface.



Before I quit this evening, I reassembled some of the parts and coaxed them together with clamps and heat. This is a very powerful technique, but I often wish I had bigger clamps.



Tomorrow I plan to get the add-ons welded onto the cullet and perhaps the back as well.

Mac

[Mac]

This is the general shape I have settled on for the add-ons, and it is working well.



Here's the third fauld lame with one of the add-ons clamped.......and tacked. Note the one of the tack welds has failed. This is why it is important to always remelt your tacks when you run the weld.



I weld from the center to an edge. When I reach the edge, I play the flame over the outboard material a bit until it is noticeably hotter than the weld before moving on the the other half of the weld or taking the flame away.



On the back side, I mark out a path in welder's chalk so I can see where I am supposed to be. If everything went well in the front side weld, there is little to do here except to flow it out smooth. I keep the filler rod ready, though, in case things get too hot and I need to cool the puddle, which happened a couple of times here.



When I packed it in this evening I had all three cullet lames done, and one of the back lames as well.



Tomorrow I will get the rest of the back added to and then make things working well enough for a fitting on Saturday.

Mac