99.5%? Is it possible?

[qst42know]

I had no idea this was in your library. 8)

I take it atomizer tanks are included in the refinery's you consult for. There is one thing I wasn't clear on.

Is the water recirculated through the high pressure pump, or the tank sized to accommodate the additional water.

If the water is recirculated what sort of filter is used to keep the pm's out and how much material is tied up in the filter?

 

[4metals]

Generally the water is not recirculated, although with a large enough holding tank it could be. The processing lot size determines everything, smaller units will atomize 2 oz of karat per second using 4 nozzles pushing 1.4 gpm each at 5000 psi. So to atomize 1000 oz, you will generate about 50 gallons of water, that requires a tank holding about 75 gallons because if you don't allow a few inches of water in your receiver to start, you will be spending time scraping gold off the bottom!

The tanks I like to use will accommodate the full volume water needed, then it can settle overnight and be pumped down in the morning. The water could be re-used for the next lot if you use a shallow well pump to feed the pressure pump. If you need to process immediately all but 2 ounces of a 1000 ounce lot settle within 5 minutes. Within 24 hours it's all down.

I prefer not to filter it because the super fine particles work their way into the fiber of the filter media requiring incineration and all the associated processing and accounting.

 

[qst42know]

Thinking small scale, is an atomizer rig efficient enough in cooling a pour you could do so in a poly drum?

 

[4metals]

No firsthand experience, I like stainless. I guess with enough depth before you start you will be OK.

 

[HAuCl4]

4metals: I understand you melt with an induction furnace. How do you heat the ceramic nozzle/crucible with a hole for the trickling molten metal or did you put a ceramic tap on the bottom of the induction crucible?.

Thanks for posting the exact specs of the atomizer.

 

[4metals]

The orifice (tundish) is mounted in an induction heated head. Small melts op to 1000 oz can be melted in the tundish which is plugged with a removable rod to start the flow, larger melts are accommodated by pouring at a controlled rate into the heated tundish.

I suppose a small unit could be fabricated from an electric kiln, with a hole in the bottom to drain into the spray and a top that opened to feed it. This would suffice for karat gold and smaller melts, feeding from the top would be possible I guess.

 

[HAuCl4]

http://en.wikipedia.org/wiki/Tundish

In modern use, a tundish is a broad, open container with one or more holes in the bottom, used in various industrial processes. In metal casting, the tundish is used to feed molten metal into an ingot mould so as to avoid splashing and give a smoother flow.

The tundish allows a reservoir of metal to feed the casting machine while ladles are switched, thus acting as a buffer of hot metal, as well as smoothing out flow, regulating metal feed to the moulds and cleaning the metal.

Metallic remains left inside a tundish are known as tundish skulls[1] and need to be removed, typically by mechanical means (scraping, cutting).

 

[4metals]

A properly heated tundish leaves no skulls!!!

 

[HAuCl4]

What is the material the "tundish" you use made of and what do you use as a rod/plunger to close the drip hole?. Any chance of a quick hand drawn diagram or a picture?.

Can Platinum scrap be "nuked" with this type of atomizer in your opinion, or melting point is too high?. :shock:

 

[4metals]

The tundish is a continuous casting crucible which have the hole and the plunger. Off the shelf stuff, all sizes available http://www.remet.com/specialty_metal_powder_production.php

Platinum is difficult and I haven't worked with it as a high grade metal with this application, however I have "inquarted" platinum in karat scrap 3:1 and atomized it effectively. The chemistry used after atomizing lends itself to PGM recovery nicely.

The long standing issue with platinum is its difficulty in digestion for high grade material, the above inquarting scenario addresses that nicely.

 

[HAuCl4]

Thanks 4metals as usual. 8)

 

[qst42know]

The spray nozzles used for atomizing must create a considerable draft of air. Does this process require any sort of shroud?

 

[4metals]

There is some spray and the top of the receiving tank is sealed with a lid leaving the only opening, which is where the stream of gold is delivered to the high pressure stream.

 

[HAuCl4]

I believe a relatively cheap atomizer can be built using a second-hand SCUBA air compressor attached to a cascade, attached to an air-water tank, attached to the water cleaner nozzles, delivering the spray into the receiving tank. The air cascade would keep the pressure constant for a steady spray.

http://www.northshorecompressor.com/cascade-systems/cascade-systems.html

Instead of tundish (to make it simpler), an electric kiln at around 1200 C, with a graphite crucible with a small hole (or several for a larger multiple nozzle system) in the bottom, fed molten metal from the top, would provide a relatively steady metal drip into the atomizing nozzles.

A little trial and error would be expected for a first time builder however, and a safety warning about the high pressures used and the pipes, tanks, valves, etc to handle the pressures safely. This isn't something to toy with. :shock:

 

[4metals]

The crucible with the hole in it is exactly what a tundish is. It's pretty simple already. I happen to know a bit about scuba diving, I've been doing it for 40 years, and there is no way a cascade system will be cheaper than a 5000 psi power washer.

However 4, 1250 psi washers will be cheaper than 1 bigger one and the plumbing will not need to withstand the same 5000 psi so it can be threaded pipe. Just run the pipes so they form the same 4 jet configuration and make sure all the 4 washer can be turned on at the same time.

 

[HAuCl4]

Generally the water is not recirculated, although with a large enough holding tank it could be. The processing lot size determines everything, smaller units will atomize 2 oz of karat per second using 4 nozzles pushing 1.4 gpm each at 5000 psi. So to atomize 1000 oz, you will generate about 50 gallons of water, that requires a tank holding about 75 gallons because if you don't allow a few inches of water in your receiver to start, you will be spending time scraping gold off the bottom!

The tanks I like to use will accommodate the full volume water needed, then it can settle overnight and be pumped down in the morning. The water could be re-used for the next lot if you use a shallow well pump to feed the pressure pump. If you need to process immediately all but 2 ounces of a 1000 ounce lot settle within 5 minutes. Within 24 hours it's all down.

I prefer not to filter it because the super fine particles work their way into the fiber of the filter media requiring incineration and all the associated processing and accounting.

Obviously we have a confusion. I understood the system operated at 5,000 psi. Now you say that it operates at 1,250 psi?. That would be certainly cheaper. 4 water cleaner systems at 1,250 psi would deliver higher flowrate, but they wouldn't take 5,000 psi of pressure. I'm confused about the specs you quoted above in one of the earlier posts.

In my head I was building a 5,000 - 6,000 psi atomizer, not a 1,250 psi one. :shock:

Here's a link for several 5,000+ psi pressure water washers. Probably a used one is cheaper.

http://www.ultimatewasher.com/5000psi-gas-powered-pressure-washers.htm

 

[4metals]

a 5000 psi pressure source will supply 4 nozzles which each have 1250 psi of pressure and 1.4 gpm of flow (each) The 5000 psi is what is delivered to the manifold before it is broken down to the individual spray nozzles. I suppose you could use 5000 psi per nozzle but you may make powder so fine it will never settle.

 

[HAuCl4]

I see. So the system operates at 1,250 psi of maximum pressure and 1.4 GPM of flowrate per nozzle. Thanks. 8)

I used to teach fluid mechanics at a university 25 years ago, and design protection for pipe systems against water hammer events, but there you go, I still managed to get confused about pressure and flowrate. :lol:

IMHO a 5,000 psi pressure source would destroy equipment designed for 1,250 psi unless the designer used a very generous safety factor. Pressure at the manifold is likely a little over 1,250 psi and the manifold simply distributes flowrate. A technicality over which I'm not ready to start world war III with you!.

 

[4metals]

Well I never taught fluid mechanics so I'll defer to your wisdom. I do know that I've personally seen manifolds supplied with 5000 psi of pressure with 4 individual 1250 psi nozzles at 1.4 gpm coming off to form the 4 nozzle arrangement to atomize the metals.

It used butt welded schedule 80 pipe fittings for a manifold. I was told that these fittings are in excess of the requisite 5000 psi and the welds were x-rayed to assure they could take the pressure.

I assume, and while assuming can get one into trouble in this case I feel relatively safe, that the maximum pressure the manifold sees it the output of the pressure washer. That pressure is vented in 4 places with the nozzles each rated for 1250 psi. So the manifold can take 5000 psi but I doubt it ever sees close to that.

 

[HAuCl4]

I see. The pressure inside the manifold depends on the pump and the flowrate. The pump (probably a reciprocating pump), if one attaches a wide hose to it (with no backup resistance to flow), the pressure delivered is basically zero (the reading of a gage at the manifold at atmosferic pressure). If one restricts the flow with nozzles, or a valve, or whatever, then the pressure rises to the maximum limit the pump is capable of (most have a bleed off safety valve to prevent self-destruction, or a shut down mechanism).

If what you saw was a max 5,000 psi machine attached to the manifold, then the pressure in the manifold (and each nozzle) could get close to that. If however a 1,250 psi machine (or four of them in parallel) were attached, then the pressure in the manifold could be close to 1,250 psi.

The way nozzles work is that, for a certain flowrate, a certain pressure drop across is needed. This is the 1,250 psi drop at 1.4 GPM that you quote. The pressure at the discharge point of the nozzle is zero (atmosferic pressure), all the energy has been converted to fluid kinetic energy. This speed is what cuts the metal. In the inside of the nozzle, and towards the manifold, the pressure is 1,250 psi only if the flowrate is 1.4 (for that quoted nozzle). If one forces more flowrate through the nozzle, the pressure drop is bigger. If you put a 6gpm nozzle, the pressure drop would be smaller.

It's not clear to me just yet what the atomizing pressure is. Was there a pressure gage at the manifold?. What was the pump used 5,000 psi pump or 4 pumps of 1,250 psi?. :?: Of course the lower the pressure, the cheaper the system. Also the less dangerous.

If the flow rate at 1.4 GPM measurement is accurate, then the most likely pressure in the manifold is 1,250 psi. If you put 5,000 psi to that same nozzle, then the measured flowrate will be bigger.