Uniformity at low temps

[patrick]

As you may know, most of my forge work over the last few years has been production forging of Mokume. Lately, I've had trouble with over heating. The billets are not melting, but the brass and copper layers are diffusing into each other and resulting in billets that should have 3 colors only having two. My plan to fix this is to build a new forge, which I've been planning on anyway. The dimensions inside will be 20" wide, 24" long and  probably something around 12" high. Roof and walls will be 4" thick kaowool. Floor is hard splits over 2.5" soft fire brick. I'm thinking of a 4 burner arrangement with two on each wall, off set so that the burners don't blow right at their counterparts on the opposite wall. The burners be mounted fairly high in the wall because I don't want them blowing on the work and making hot spots. I'm currently using a 2 burner arrangement made from 2" pipe with a couple of smaller pipes nested in the flame end of the burners. System is forced air/propane. It works OK, but with both burners on one wall, you don't have good uniformity. You can get it if you're running hot as when forging steel, but most of what I do needs to be done at around 1800 F. My thought is to scale down this burner design and go with 4, but I'd like to get the input of others on this.

 

I'm also planning to incorporate at least 2 thermocouples into the forge. I'd like them to be permanently fixed in the walls. With that arrangement, they would need to be able to withstand the temps for forging steel since it do that on occasion. Any recommendations on specific units I should look at and places to purchase same?

 

Thanks.

 

Patrick

[Latticino]

I've used Type K thermocouples up to 2,300 deg. F with pretty good success, though they were relatively large diameter wire (around 10 GA as I recall) and inserted in high alumina ceramic thermocouple tubes for protection (from molten glass in my case, but would also work for flux). 

 

You really should pony up the bigger bucks for either a Type R or Type S thermocouple for these temperatures if you can afford same.  Make sure your digital, or mechanical pyrometer readout is designed for the thermocouple that you choose in any case.

 

Also, for really even heating you might consider a ribbon burner.  They are nice and quiet, and spread the heat out over a relatively large area.  The one drawback you may have is that they are not really well designed for a large range of modulation.  The ones I've seen setup for modulating in the past had some kind of slide gate installed to block off some of the outlet ports, which kind of defeats the purpose.  I believe it may have something to do with maintaining a safe range of velocity of the air/gas mixture exiting the burner ports.

[HWooldridge]

You might think about a muffle design, possibly with the addition of exhaust baffles (similar to a heat treat furnace).  The parts to be heated are separated from the heat source by a partition that more or less encapsulates them but allows the heat to enter through a constricted area.  You can then push a lot of heat into the box without worrying about hot spots and house the thermocouple in the muffled area.

 

Could also run the burners at full fire and dump some of the heat through an exhaust port with a baffle - although that is admittedly more wasteful of fuel than controlling input.

[timgunn1962]

I'd go with the muffle idea and try to avoid direct heating by the burner flame altogether.

 

I tend to use Venturi burners and the idea of trying to balance the mixture (and thereby the flame temperature) on multiple gas mixers seems like way too much effort: I'd go one big one every time.

 

With blown burners, you have a better chance of balancing them, though getting them all to run evenly off one mixer tends to mean more pipework filled with a potentially explosive gas/air mixture. I'd still go one big one.

 

Type K thermocouples look quite good on paper (the tables go up to 1370 degC, 2500 degF), but tend to suffer from "drift" when used above about 1000 degC. 1000 degC is 1832 degF, but note the "about". I'd expect your application at 1800 degF to cause drift in type K, albeit slower than at higher temperatures.

 

Thermocouple types R and S are very stable and make a great cost-no-object choice up to 1300 degC, 2372 degF, but are Platinum-based and priced accordingly. R and S are actually good to around 1760 degC, 3200 degF and there are no cheap options at these higher temperatures.

 

If your temperature readout will accept type N thermocouples, these are an excellent choice up to their limit of 1300 degC, 2372 degF, having been developed as a sort of improved type K without the drift problems.

[teenylittlemetalguy]

Are you sure it is not a material quality issue? Maybe someone is slipping you inferior product?
I am not saying a new forge is a bad idea but it may not solve the issue.

[patrick]

I'm pretty confident it's not a material issue. I've heard of the same thing happening in some damascus billets. Over the last few years I've probably forged something on the order of 4000-5000 lbs of mokume, always supplied to me pre-fused so it is possible it a material issue. My customer, who supplies me the billets for forging to his sizes, is going to send me a piece of what he's seeing and I'll do some metallography on it to confirm the root cause of the problem. I wanted to do a new forge anyway so this is good opportunity for that, even if the problem I'm trying to solve is not temperature related.

[Chinobi]

Look up Eric Fleming forge controls, he did a demo at the ABS hammer-in last October (2014) in Tulare, CA, and it was quite impressive. He had a controller box rigged up to a thermocouple (I forget if it was a type K or something more interesting) and a solenoid system such that once the forge was up to the target temp (which you can dial in to taste yourself) it would pulse on and off briefly and actually kept the temp within a few degrees. Jason Knight used it to forge out and heat treat a large Bowie fighter and dialed in the temps he wanted for each step, and Tom Ferry used it to weld up a large-ish Damascus billet with no flux, setting the initial weld in a hurry because the tank on the forge was running dry and the temp was measurably falling off.

Out of curiosity, could it be a time at temperature issue rather than straight overcooking? Excessive eutectic alloy formation between layers can happen by too much time in the fire as well as too much fire.

[ThomasPowers]

I would wonder about the initial fusing before you received them as well.

[teenylittlemetalguy]

That is a lot of Mokume, would love to see some if you ever felt up to it. Those of us that keep an eye on that subtopic are awfully hungry for advice from some one such as yourself.

[patrick]

To see finished products made from what I've forged look up William Henry Studios. Most of the Mokume they use is supplied by Mike Sakmar, who is my customer. You can also look at his website for images of stuff we've done. Some of the coolest work I've done is to forge large billets that are milled into cell phones. Do a google search for "mokume cell phones" and you'll find lots of images of those. I've forged something on the order of 150 billets just for that product.

 

As far as eutectic formations, that is a function of composition and temperature, not really time at temperature. A eutectic is the one unique composition that has the lowest melting point. The classic example of this is with copper and silver. Not all alloy systems form eutectics. I'm working with copper, red brass and nickel silver. The constituents of those are copper, zinc, nickel and tin. Copper and nickel do not have a eutectic. The copper/zinc  and copper/tin do but those alloys are already combined in the red brass and nickel silver so they will melt rather than form a eutectic. The melting temps of the alloys are as follows:

 

Copper-1984 F

Red Brass-1832 F

Nickel Silver-1870 F

 

Diffusion could be accounting for what is happening and this is very much a time and temperature dependent phenomenon. The billets in question are laminated with all 3 of the alloys listed above and the copper/red brass layers are the ones that seem to be washing out. I find this a bit odd since most of the time overheating of this material will result in actual melting of the brass and it will squeeze out or even splatter if you hit the billet when it is this hot. I've had a few billets where this has happened and if you catch it without hitting and let it cool you can still successfully forge it. Interestingly, I've not had feedback that these billets had the washed out appearance. That makes me think I'm dealing will something that is very time dependent within a particular temperature range. I'm really interested to see the microstruture as that should give me a pretty good insight into what is happening. If I'm getting too hot, I could have both the copper and brass layer melting and mixing together, but I would think that that much melting would have been notceable during forging. 

 

For those interested, feel free to PM me or start a new thread with any questions on mokume. I'll do what I can to answer. 

[Chinobi]

Sorry, terminology error on my part, Ian Ferguson describes basically the same phenomena you are describing both in terms of 'eutectic reactions' and diffusion, but i agree, diffusion is the more closely related mechanic.

 

'overheating' does not necessarily have to be so severe as to cause melting of layers, it only needs to be hotter than the bonding temperature in order for the billet to continue to undergo solid state diffusion.  also per Ferguson's tables the bonding range for copper/brass (no red brass specific entry, sorry =/) is 600-700 deg C, the copper/nickel silver bonding range is also 600-700 deg C, both over a period of 60 minutes. converting roughly to Fahrenheit that correlates to 1100-1290 F give or take.  bonding in an unregulated forge (vs temp controlled digi-kiln or what have you) knocks the time factor off severely by substituting a much higher temperature for a longer time duration.

 

as far as anyone can state from the information provided and no pictures, in my wholly unprofessional opinion, i dont think you are running into a problem actually melting your layers.  you would likely notice if your brass layers have melted or gone to slush either while you pull it out or as soon as you start working on it, and in order for the brass and copper BOTH to be melted your nickel silver will have melted as well, and thats your entire billet in a puddle :)

[patrick]

I have both the Ferguson and Midgett books and they are both very good. Ferguson's work is all based on bonding in an inert atmosphere with pressure applied during heating. He even made special dies to contain the work piece on all sides while pressure was applied. It is a very interesting technique, but not practical for most folks. My customer is doing the bonding and sends me billets ready to forge. He is using a technique much like what is described in Midgett's book with bonding temps around 1800 F. His process is extremely effective provided all the surface are clean.

 

I agree I don't think I'm melting the layers. I think I'm just a few degrees cooler than that. At the point in the process where this is happening, a billet with 88 layers originally 2" thick has been forged to a thickness of 0.375. Assuming the starting stock is uniform in thickness (which I don't know for sure), the layer thickness after forging to 0.375" is only 0.0042". At this point, the billet is patterned via machining and sent back to me for further forging to 0.220" thick. It is during this second forging step that problem is happening. I'm wondering if the combination of temperature, time at temperature and very thin layer thickness are all working together to cause this problem. It is not showing up in other billet sizes or billets in which the starting layer thickness is greater.

[teenylittlemetalguy]

That is a really interesting problem. Can you take the layers thinner and see if it makes it worse?

[teenylittlemetalguy]

By the way I looked up Mokume cell phone and WOW.
I have to say I really like the looks of your materials way more than the stuff from shining wave.

[Frosty]

Patrick: My intuition is telling me that once a threshold is passed the differences in ductility overcomes the strength of the metal or perhaps is magnified causing delamination. By magnified I mean the differential between metals, the thinner the lamina the greater the differences in ductility or elasticity are in relation to each other.

 

Thanks for the search term to look at Tristan, I've been almost dizzy just looking at pictures.

 

Happy new year guys.

 

Frosty The Lucky.