How do you melt Fe3O4 / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

How do you melt Fe3O4

previous 1 2 3 4 next last

Move Thread

LAN_403

Andy

Fri Sept 20 2013, 09:19PM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Will do.
How far can you push a heat sink with fan cooling, I've got a heatsink that is rated 11C/W and plan to push 25amps through the mosfet with 0.044ohms resistance, it comes out to 302.5C. Will I need a bigger heatsink to keep it below 175C(max of the mosfet) or could a fan work?

Sulaiman

Sat Sept 21 2013, 07:30AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

YOU NEED BETTER HEATSINKING

An IRF540 has Rds(on) of 44 mOhm at Tj = 25 C
if you run at Tj = 175 C that approximately triples to 120 mOhm.

The IRF540 has thermal resistances;
junction to case 1.15 C/W
case to heatsink 0.5 C/w (grease, no insulator)
So thermal resistance, junction to heatsink >= 1.65 C/W

For a single transistor;
If your heatsink is at 100 C, Power <= (175-100)/1.65 = 45.4545 W
With Rds(on) - 120 mOhm, Id <= sqrt(45.4545/0.12) = 19.5 A continuous.

If ambient temperature = 25 C, the heatsink needs to be (100-25)/45.4545 = 1.65 C/W

If there are negligible switching losses ('zvs' etc.) and the transistor is on for 1/2 cycle
then you could double the power, so Id <= 27.5 A
if each transistor has no insulator and it's own 1.65 C/W heatsink at 25 C ambient.

You want to put 4x IRF540 on one 11 C/W heatsink,
so you will need insulators, which increase thermal resistance case to heatsink
suppose new thermal resistance case-heatsink = 1 C/W
junction-heatsink now becomes 2.15 C/W
Pd <= (175 - 100)/2.15 = 34.9 W, Id(cont.) <= 17 A, Id (zvs bridge) <= 24 A
Required heatsink = 1.65/4 = 0.41 C/W

I doubt that even a powerful blower would reduce 11 C/W to 0.41 C/W

For short bursts you can rely on the thermal inertia of the heatsink.

The Tj < 175 C limit is mainly due to the aluminium diffusing into the silicon shortening it's life
you can run much hotter at the expense of lifetime.

P.S. one cpu fan/heatsink per transistor (grease, no insulator) would work,
but you must electrically insulate each heatsink.
(two will be at +voltage, two at opposite switching voltages)

Andy

Sat Sept 21 2013, 08:11AM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Thanks Sulaiman for a very detailed answer

On Monday I should be ordering some Riston boards and developer, got the negative that should fit on a 152mm*152mm board

Ordered but its on back order a

ferrite core for testing, Will be run at 72volt should have 5 turns primary at 50amp with 50 turns secondary if it follows the theory or if I worked it out correct :)

Still not sure how much power the thing will need, but will make the driver board beable to handle 14 of the above boards.

Edit
I think the specific heat capacity of Magnetite is 670joules/kg/C, one site says 0.028m3 weights 56kg a 56cm*40cm(0.0986m3) cylinder might have about 197kg, 132kj/C to get to 1600C will need 211398kj/18000seconds(5 hours) = 11.7kw.
Car battery's = 60amp/hours = 60*72 = 4.3kw , 15kw/4.3 = 4*5 = 20 in parallel,120 total, ,75kw/19hours = 3.9kw to charge the battery's up again.

Bored Chemist

Sat Sept 21 2013, 12:19PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

"How do you melt Fe2O4"
You don't.
Unless you know what you are hoping to melt I don't think you will get very far.

Andy

Sat Sept 21 2013, 05:06PM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Hi Bored Chemist
Fe3O4, iron oxide from black sand, 1597C melting point, updated

Crucible size = 20cm(D)*20cm(H)
Weight of sand = 12.5kg
Time = 720seconds
kW = 18.7
Number of battery's = 6(series)
Recharge amount = 150watts/24hour
Recharge/hour-720seconds = 4.6kW
kWh = 3.74kWh

Andy

Wed Sept 25 2013, 07:21AM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Just wondering about a couple of question.
I won't be able to read the temperature of the furnace, but if the work coil has the right number of turns for the resistance of the Fe3O4, when it melts, the resistance should drop lowering the current on the work coil. I don't know were to begin with measuring the current of the primary, but was hoping to have a bar of LEDs yellow when its been heated and red when the current drops(by like half), with a audio buzzer.

The seconds question was I plan to make steel, with 2% carbon, but if the iron oxide releases oxygen, the carbon will burn with about 1598kJ of energy, if it happens at 1600C, it might push the limit of the crucible 1700C. Is this going to be a problem , I don't won't to add the charcoal latter(or how could you if the thing is sealed with firebricks?)

Thanks

Ash Small

Wed Sept 25 2013, 01:54PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I think most processes use calcium carbonate, which decomposes when heated:

CaCO3(s) â†’ CaO(s) + CO2(g)

It works best if it is ground into a powder, and mixed with the powdered iron oxide.

Andy

Thu Sept 26 2013, 12:46AM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Thank Ash Small

Found some links

It says that about 800C CaCO3 breaks down and CO2 will react with C at 800C to make CO, does that mean the furnace just needs to reach 800C and the reaction of Fe3O4 + CO will make enough heat to melt the iron(1538C).

Edit
The ferrite cores arrived, from 11-15 turns primary and secondary is 10.34volts, with 8.6volt at 7-9 turns. Increase the frequency from 28khz to 43khz makes the voltage drop with the same amount of turns.

This is the core

How many amps do you think it could handle

looking at the datasheet its got Ae/mm2 of 152(is that amp turns), I'm trying to workout weather to parallel them or have them in series, aiming for 720-800volt 7.2amps per h-bridge

My bad wired the 555 up wrong, it now follows when you increase the frequency it increases the volts

Ash Small

Thu Sept 26 2013, 09:47AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

As far as I'm aware, electric furnaces are not used commercially for producing iron or steel (although I think they are used for aluminium). Electric furnaces, whether arc or induction, are only used for re-melting iron and steel produced in blast furnaces.

An exception might be the 'electric blast furnace' used mainly in sweden, but these have now disappeared.

It appears that most of the carbon used in reducing the oxide ore comes from the coke or charcoal used in traditional blast furnaces

I once built a 'puddle furnace', or open hearth furnace using bricks from a couple of night storage heaters, and old vacuum cleaner, which blew air down a length of scaffold tube, and ran it on charcoal and very dry wood.

I chucked in a handlul of rusty old nails, etc. and ended up with 'puddles' of very high quality carbon steel, really hard, which never went rusty. I cut some of the pieces in half, and polished them, and you could see inclusions of carbon in the polished surface. I never got round to 'forging' any of it, though. (I mainly used the furnace to melt and cast aluminium)

I still can't see any advantage in using an electric furnace to produce iron or steel from 'ore', although they are really useful for re-melting, and producing special alloys.

What advantages do you see in using electric furnaces for this purpose over traditional blast furnaces, or even open hearth furnaces?

,

Andy

Thu Sept 26 2013, 06:42PM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Hi Ash Small

It want just be used for Iron oxide, copper/Al, and maybe titanium dioxide. Back in the day, I built a coal furnace to melt Al, the temperature would have only reached red hot and it went through a tonne of coal :), it mainly for expandability.
I don't think it will need multiable runs to make steel, one run with say 2% carbon should make good steel with the carbon content maybe dropping to 1% from 2%.

A supplier of crucibles has got back to me and recommended a Mo crucible for the temperature range and in atmosphere. Hopefully in a month I'll have a driver circuit 1 h-bridge and 6 deepcycle battery's and will test to see if it can make 0.5mm Nickelchrome wire turn red.

Edit
Some values of the ferrite at different frequency's and number of turns
5 turns
7 volts
43khz
10 turns
9.8 volts
43khz
10 turns
11.5 volts
57khz
10 turns
12.0 volts
72khz
5 turns
10.3 volts
90khz
7 turns
11.5 volts
90 khz
8 turns
12.05 volts
90khz

Measured the resistance through the magnetite, at 5cm and 0.5cm, was to high for the meter.
Got this much out of a spade full from the backyard, buying it from the landscaping company in town might have high purity

Might need need a copper coil, when it melts at 1000C approx it should absorb the Fe3O4 and increase in resistance.

Edit
Reread you post Ash Small, it might be cheaper than a Mo crucible to use clay, Zirconium Silicate/Titanium dioxide, with the work coil in the mixture. Maybe the water cooling will keep the clay from turning to powder, but its cheap enough to replace. Maybe wood could be used for the wall lines, as it ablates it should cool down the walls.

previous 1 2 3 4 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob