How to tell if your forge weld will hold

[Work With Nature]

Been googling this question now for the past hour and can not get a proper answer. Also did ask this question on a different thread here on this forum, but it is not related to the original question of that thread and not any wiser.

I am wondering how do beginners like me know if our efforts have paid of regarding to doing forge welding and having them hold well enough, to not come apart later on. It is probably obvious that visual cracks will not work, but what about lines in the welds? I have now done three welds in total. They look and feel solid enough but them lines have me worried. First picture is my attempt at damascus (two folds in total), second is my attempt at a hatchet. The hatched has a spring steel insert and as many of you know it is very hard to move under the hand held hammer. Don't want to waist 5 hours of my time if it just comes apart later on. Am not that mad. If anyone could let me know would appreciate it. Also would like to invite people to post some pictures of their failed attempts and what a good weld may look like. I think to have some pics would be good and some detailed info on how to recognize a good from a bad one.  

Cheers David. 

[Hotshoein4 (Mark)]

When I learned to forge weld, I welded rings out of mild steel. 5/16"x1/4". If the weld stuck, you could hold the ring on your finger, weld down, and tap it with a hammer lightly and it would have a nice ring to it. If the weld was bad, it would be a dull thunk sound. Just lots of practice and luck. 

[swedefiddle]

Good Morning,

If you are welding a chain link, test by bouncing it on the end of the link. If it has a high pitched ring and bounces, it is welded. As for welding a billett, look for inclusions. It will show up as a line or crack in your billett. Quite often you can't see the inclusion until you are almost finished finishing.

If you play the game, it comes with a price called making "Test Pieces", or altering the intended size! "Unintended Consequences".

Neil

[ThomasPowers]

For a standard billet I usually consider forging it on the edge rather than the flat to be the "test".  However you can shear welds later that had been good before with improper technique like forging too cold especially when some layers are much softer under the hammer than others.  Polishing and then etching and microscopic examination works a treat for formal investigation.  Me I forge a billet on the edge and see that it holds together that way.

[Work With Nature]

Thanks everyone for getting back to me on it. So did I understand correctly, them lines are inclusions??? Could the lines be from carbon trying to leave the steel?  

[ThomasPowers]

NO!  carbon leaving the steel is known as carbon dioxide and carbon monoxide. In a properly made forge weld carbon can migrate across the weld from the higher carbon layers to the lower carbon layers. Not usually an issue if the layers are thick and only welded a few times. However when the layers are thin and with multiple welding cycles you get a more even carbon content. (one suggestions was that after 4 welding cycles with thin layers of alloys that do not block carbon migration , pure Nickel does block it for example, that the carbon content will equalize.  This was given by the Damascus Research Group at University of Il, Carbondale, Daryl Meier; IIRC)

[Work With Nature]

OK the carbon does not leave the welds but migrates and equalizes out eventually across the steels.  Am gona give them tests a go. What do you think them lines are? Anyone.

[JHCC]

Another way to test a weld is to place your workpiece on the anvil while it is still glowing. As the anvil absorbs some of the heat of the workpiece, the latter will change color from yellow to orange to red etc. If the weld is successful (that is, if you now have one continuous piece), the color will change evenly across the workpiece. If the weld did not take, then you will see a distinct line between the two sections, as the lower one loses heat faster than the upper.

[Work With Nature]

JHCC,

Will take a closer look but am starting to get the feeling that it actually took well enough at the back area. Going by what you mention that is exactly how it looks on a picture I took. These are great tips. 

[FoxFire]

When curious about a forge weld I sometimes quench or hammer on them below normal forging temps. Basically all the things you're not supposed to do. From the point that I've "decided" that it is a good weld (and I'm ready to move on) I generally treat the weld as I would any other piece of steel/iron.

 

 

[Work With Nature]

Thanks FoxFire,

Makes a lot of sense. 

Just to update the weld actually came apart on me when I tried to drift the eye wider. Did it at a red heat. In the bottom photo of my first post it is the weld bellow. It was interesting that it was the underside when I did the welds, so turning the billet is important. You live and you learn. Dropping the darn thing and as I recently found out, not removing the scale with a wire brush at orange heat was a mistake too. Well never mind will keep at it. 

[rockstar.esq]

 

Another way to test a weld is to place your workpiece on the anvil while it is still glowing. As the anvil absorbs some of the heat of the workpiece, the latter will change color from yellow to orange to red etc. If the weld is successful (that is, if you now have one continuous piece), the color will change evenly across the workpiece. If the weld did not take, then you will see a distinct line between the two sections, as the lower one loses heat faster than the upper.

JHCC, that's a really useful tip.

WWN,

A while back someone posted a video of a guy who did some destructive testing on forge welds to determine what was good, bad, or indifferent.

The main lesson was that a "one and done" forge weld was basically a tack weld with limited strength.  When he repeated the forge welding process on his first weld, the joint was substantially stronger, but the weld joint failed before the parent material.  When he did three forge welding passes on the joint, the resulting weld was stronger than the parent stock.

Some of the old blacksmith books I have recommend three "steps" for forge welding.  While they might call the steps "tack", "weld', and "dress", they do all of the work at welding temperatures which would be consistent with what this guy was doing.

 

[JNewman]

I have some tongs i make for a customer I used to forge weld the reins on.  The first time I did it I did the destructive test thing on 2  welds  and bent the weld area 180 degrees cold.  I figured this was more abuse than the tongs would see in use.  2 heats was usually  enough.  More than  that and you risk forging the reins undersized,  Which then have to be upset back to full sized, which is not a fun job. 

Reliable forge welding requires practice.  It has been quite a while since I welded reins on,  I would probably have to do a couple practice ones to do a good job now. 

[anvil]

The extreme test for a weld is to quench while it still has color in it.  It will show you good or bad.

When working a forge weld, a good practice is to work it at a yellow, down to a red.

 

Of course that's the proper temp to forge most anything.

 

When it's yellow it's mellow, when it's red, it's dead is a good rule to follow.

[Steve Sells]

 

When working a forge weld, a good practice is to work it at a yellow, down to a red.

this is only true with some alloys, because there are alloys that will shatter doing that not every a steel works at the same temperatures, Sorry

[anvil]

Actually if you look at the recommended forging temps in say "the heat treaters guide companion" you will find most recommend forging from a yellow and not below a red.

  But not all.

Always check yer spec sheet first.

[Frosty]

Old heat treater's guides tend to be for alloys that are covered in a couple pages of modern guides, even then they list different forging temps. NO modern guide recommends by color. Different eyes see different colors under the exact same conditions, it's why learning to judge by color is so intimidating by the book, while someone showing you by example is a snap.

Taking a color photography class is an excellent way to learn how different: color, light level, shading, etc. is for individuals. 

Frosty The Lucky.

 

[anvil]

"the heat treaters guide companion" is a modern free apk for Android and iphone, not an old source.

 

I should qualify and state that forging from a light yellow to a light red is the "color range" for many tool steels. Prolly most of the common steels used by both black and knife Smith's.

To quote from the uddaholm color charts(a rather contemporary company who manufactures tool steels):

Yellow = 1929 degrees

Light red = 1600 degrees.

 

I agree that different eyes see different colors but the same old eyes in the same old shop with a constant light source can easily associate colors to temps.

 

I believe Frank Turley did an experiment once concerning his eye for color vs a thermostat and was within 10% of the thermostat reading.  Seems the coefficient of error for the electrical device was 10%.

Just because our esteemed knife makers and others use their high tech toys, which is a cool deal, doesn't mean that we old luddites aren't or can't be as precise via forge and coal.

 

And reading colors in your own setting is a technique that, I believe, needs be not lost on the new guys.

 

 

 

[Iron Poet]

I primarily use sound and sight to determine if a weld is good. Typically if there isn't any demarcation  between the two pieces then it's probably good, because if it's welded at the surface it's probably welded inside. Sound is also very important because if the weld is bad the metal will sound dead, if you correctly weld 1/4" rod together for instance, it should sound exactly the same as a normal length of the same material.

I forge weld a ton of 20 gauge sheet and 1/8" and smaller for flowers and such, just practice and you'll get it eventually.

[ThomasPowers]

Some high alloys steels will cottage cheese if forged at a good forging temperature for other steels.  Very sad to pay good money for a high alloy steel for tools and have it crumble on you.  Since they tend to be hard under the hammer at forging temps there is an "instinct" to raise the temp to make forging easier---this way lies badness to misquote Dune.

[anvil]

I agree.

Shows to go if you do it wrong, no matter, it will be wrong.

Proper setup, proper tools, proper job.

By eye, tempil stick or thermocouple, know your tools and it will work.

 

However, how do you explain the sheer joy of making say a simple cold chisel out of coil spring in 1980, running miles of work with it, dressing it a half dozen times and still using it as a primary tool when you are 70?

All ya got to do is learn to heat treat by eye.  And just be a simple blacksmith.

Choices, choices.  

[lyuv]

My 2 cents for the original subject of telling if a weld took - a method I "discovered" the hard way: Tried to weld two plates. After they SEEMED to be welded, I ground the edge of the billet flat. Then procceeded to thin the billet by heating and hammering.

Obviously, different steels have different maleability at the same temperature. So one of the plates widened a little more than the other. and the previously flat edge, now had a "step". If the plates where welded, the edge would be flat or a little slanted.

[Work With Nature]

 

I primarily use sound and sight to determine if a weld is good. Typically if there isn't any demarcation  between the two pieces then it's probably good.

 

 

First of thanks guys for all the helpful info. 

Hi Iron Poet would it be alright if you explain the demarcation a little further. Did hit the "axe" and it made a great thud  Better luck next time. Cool flower BTW.

 

Obviously, different steels have different maleability at the same temperature. So one of the plates widened a little more than the other. and the previously flat edge, now had a "step". If the plates where welded, the edge would be flat or a little slanted.

Hi lyuv, Just to see if I got you right, the previously cleaned edge stays more or less in the same plane without one metal moving beyond the edge of the other. 

[lyuv]

 

Hi lyuv, Just to see if I got you right, the previously cleaned edge stays more or less in the same plane without one metal moving beyond the edge of the other. 

exactly what I ment.