Critique please -- Turkish Twist (not a dance)

[Eric Niefert]

I just finished my first Turkish Twist Damascus and thought I would ask you experts what I can do better next time.  I was less than enthused with my results this time, but I tend to be picky. 

 

Here is a brief run down of what I did.  Made a 30 layer billet of 1084 x 1/8 thick and 15N20 by 1/16' thick.  I cut the billet into 6 pieces and added a fresh layer of 1/8' thick 15N20 between each bar of 30.  Rewelded and drew the 3 stacks into a 5/8" bar.  Rounded edges in the press so it wasn't super square.  Used a pneumatic impact to twist the 3 bars while holding the other end in my press.  2 bars twisted clockwise, one twisted CCW.  Then I square them up (mostly) in the press and ground two sides flat for welding.  Fused bars with TIG and forge welded.  I lost a fair amount of the billet due to dlams -- I believe a good amount of my TIG fuse cracked when heating the first time due to welding them with bending stress on the bars (they were 1/8-1/4' off of perfectly matched).  

 

A couple things I think I need to change . . . 

 

I think I need to count the number of twists with each bar so the twist rate is near identical and the pattern blends better row to row.

I think I need to press each bar after twisting to a perfect square and sand off every bit of scale -- even on the sides not being forge welded

I think less layers and maybe a thicker center will stand out better.

 

Anything else?  Thanks! 

[Alveprins]

3 hours ago, Eric Niefert said:

I think I need to count the number of twists with each bar so the twist rate is near identical and the pattern blends better row to row.

I think I need to press each bar after twisting to a perfect square and sand off every bit of scale -- even on the sides not being forge welded

I think less layers and maybe a thicker center will stand out better.

 

In regards to counting the twists...

 

In my experience the bar twist faster in the center of the bar than at the ends. This I assume is a result of more rapid cooling of the steel at contact-points, combined with certain effects of momentum, the latter of which I have not myself bothered investigating any further.

Also, the bar will naturally twist more easily at it's thinnest point. It is therefore imperative to keep the bar as uniform as possible - in regards to it's thickness.

 

I twist my bars by hand, locking the bar in a vise and using a set of modified pipe wrench (modified as in welded on a 2nd handle on the head, so I can twist using both hands/arms.).

 

Instead of counting the times I twist, I look closely at the twisted bar, and lock it down in the vise at different lengths - and tighten up the spots where the twist is not so tight. I continue this until I feel I have a pretty even tightness of the twist.

Then with the next bar I try to copy the appearance of the first as well as I can.

the tighter the twist, the less obvious any inconsistencies are going to be. The bolder the twist, the bigger the gap between twists that don't exactly match up.

 

Sincerely,

Alveprins. 

[Alan Longmire]

I twist by hand as well.  I do count my twists, and this is even more important if you really round the bar.  I prefer lower layer counts, usually 7 layers of 1084/15N20, twisted 5 to 7 ridges per inch (that's around 3 ridges per cm for you metric folks).  I use 1/4" thick 1084 and 0.0625" 15N20.  The thicker 1084 will be much the same as the 15N20 by the time it's all forged down. Hint: make sure the 15N20 is the center layer if you want silver stars, or 1084 if you want dark stars.  

 

I forge these billets down to around 3/8" / ~1cm square before twisting.  When twisting, assume you'll scrap the last few inches off both ends, and if you have an acetylene torch use it to keep the twists even.  

 

Not having a press, I forge them mostly square after twisting, then grind the mating surfaces flat.  I don't bother with the non-mating surfaces, because most of that will be ground away after forging. 

 

Grinding is what makes a twist pattern shine.  I assume you've seen this:

 

 

This is from Niels' thread up in Video and Multimedia, and it's a CGI of how the pattern in a twist changes as you go deeper.  The top is full thickness, and each step down represents a 5% reduction in thickness via grinding such that the bottom bar is the very center of the billet.  

 

 

There are many ways to make these billets.  Ric Furrer and Owen Bush make great big ones about 2" square and twist them up as tight as possible with pipe threaders, then forge them out to lengthen the apparent number of ridges per inch.  Economy of scale and all that.  

 

 

[Eric Niefert]

Thanks Gents. So too many layers and maybe too many twists too. A couple more pics below. 

[Alan Longmire]

I think the main thing is not enough stock removal after welding up the blade.  You didn't get down into the interesting parts of the billets.  Forge the next one and leave it about 3/8" - 7/16" thick, and grind it down to final thickness.  It'll show a lot more pattern activity.

Layers are fun to play with too.  I'm a bit of a masochist when I make pattern-weld, so I like to throw some wrought iron in the core layers.  When it works, it looks like a few hundred-layer bars got into the otherwise five-bar twist.  It does exponentially increase the risk of shearing while twisting, though.  

[Eric Niefert]

Thanks. The billet I have now is 5/16-3/8” thick now. Started at 1/2-5/8” thick so yeah, I suppose I have some grinding to do. 
 

you’re braver than me with the wrought iron addition. Yikes

[Niels Provos]

I don't want to hijack your thread but here is another conceptual view of how this comes together. This illustration simulates a fuller but when you move closer to the right the steel has not been removed and the pattern looks closer to what you have. As Alan mentioned, not enough removal of material, so you end up with the pattern that's right at the surface of the twisted bars.

 

[Joshua States]

Another suggestion: build yourself a set of squaring dies for your press. This will help getting the bars square to start. I take my bars for this pattern down to either 3/4” or 5/8” before twisting. Then I put the 1/2” dies in and make two ends to fit in my impact wrench socket. 
 

Heat half the bar for twisting. The cold end gets locked in the press and the hot end goes in the impact wrench. I find this makes for more uniform twisting and less heat loss to the press dies

[Joshua States]

My squaring dies: