Induction Heating for aluminum and copper

[badea]

Hello,
i am building an induction heating system without current and voltage feedback, IH working terrifically with iron in general, now i am looking to be able to melt copper and aluminum by IH, but this materials or rows less response than iron,because they have lower resistance than iron
so my question will be :
what is a suitable resonant frequency and design of IH for aluminum and copper melting?
what is a suitable method for current and voltage feedback for IH?

Thanks

 

[mtwieg]

Part of your difficulty with heating copper and aluminum is likely that their thermal conductivity is higher, not just their electrical conductivity. This means your temperature rise will be lower, even when you are delivering the same amount of power.

The higher electrical conductivity means you will have to modify your circuit to most effectively deliver power. The easiest way to do this is to change the number of turns in the coil, or change its size.

 

[badea]

Part of your difficulty with heating copper and aluminum is likely that their thermal conductivity is higher, not just their electrical conductivity. This means your temperature rise will be lower, even when you are delivering the same amount of power.

The higher electrical conductivity means you will have to modify your circuit to most effectively deliver power. The easiest way to do this is to change the number of turns in the coil, or change its size.

Thanks for your valuable reply
i agree with you that changing a coil design may raise delivered power, but i am looking to be able to weld a tube of copper with aluminum one, this means that a working area is small, so if i modified a coil size the effective area will be large respect to welding area.
modifying a coil design maybe a good solution for melting operation

 

[c_mitra]

Both copper and Aluminium have high electrical conductivity; that means induction heating will be less effective (compared to iron). if you increase the operating frequency, there will be less penetration due to skin effect. If you decrease the frequency, you will need higher power.

Welding a tube of Cu with Al is tricky because neither of the metals have a wide softening point where they can be nicely joined. That mean the temperature control is also critical. The joint may not be strong because Cu and Al does not form an alloy.

 

[badea]

Both copper and Aluminium have high electrical conductivity; that means induction heating will be less effective (compared to iron). if you increase the operating frequency, there will be less penetration due to skin effect. If you decrease the frequency, you will need higher power.

Welding a tube of Cu with Al is tricky because neither of the metals have a wide softening point where they can be nicely joined. That mean the temperature control is also critical. The joint may not be strong because Cu and Al does not form an alloy.

Thanks for nice reply,
To solve welding Al with Cu problem, i will add some silver to the welding point, so welding will be stronger

Now my major problem is in low frequencies that i have to put a huge parallel capacitance to achieve resonant frequency, as example if freq is 3kHz and working coil is 2uH so parallel capaciance should be 1400uF!!
any idea to reduce this huge capacitance?

 

[c_mitra]

How you are calculating the optimum frequency? That depends on the overall mass and the power needed. I am not sure about the details.

 

[badea]

How you are calculating the optimum frequency? That depends on the overall mass and the power needed. I am not sure about the details.

Calculating optimum frequency based on L and C Values, i am building a PLL to make current and voltage at same phase.
i am using an working coil voltage feedback to determine if optimum frequency achieved or not
as a rule, voltage on working coil is about 2*pi*Vin, this equation working good till now.

 

[FvM]

3 kHz isn't feasible. Usual induction cookers for steel pots are working in the 20 to 30 kHz range, copper/steel bulk heating needs considerable higher frequencies.

 

[badea]

3 kHz isn't feasible. Usual induction cookers for steel pots are working in the 20 to 30 kHz range, copper/steel bulk heating needs considerable higher frequencies.

For melting applications, frequency should be low, thats because of skin depth effect

 

[Warpspeed]

if freq is 3kHz and working coil is 2uH so parallel capacitance should be 1400uF!!
any idea to reduce this huge capacitance?

Its not just the large capacitance, at resonance there can be hundreds or even thousands of rms circulating reactive amps.

Getting you hands on a suitable capacitor can be difficult. I went through all this myself some time back. The sales rep will say something like, how many hundred per month will you be ordering ?
When you say you only want one, they just laugh and hang up on you.

Only source of these amazing capacitors that are anything like customer friendly for very small orders, is an Israeli company called Celem.
https://www.celem.com/

 

[FvM]

For melting applications, frequency should be low, thats because of skin depth effect

If you already know, why are you asking...

Actually the useful frequency depends completely on your induction oven geometry. High frequencies work with open magnetic path configurations e.g. an induction cooker. I have seen the mentioned low frequency ovens in a steel mill with a donut melting pot and closed magnetic path. In an open magnetic path with air gap, the high conductive metal simply shorts the magnetic field and doesn't absorb much power.

- - - Updated - - -

Presumed 3 kHz is a useful working frequency and you have 2 µH coil inductance, what's the expected coil current? 50 A, 100 A, 200 A or more? What's the power stage DC supply voltage? I would expect a considerable higher inductance to achieve handy current and voltage levels.

 

[c_mitra]

For melting applications, frequency should be low, thats because of skin depth effect...

The low frequency you have mentioned is suitable for a furnace that can melt a couple of tons of iron and is often used for special alloys.

The medium frequency range is mostly useful with medium power (10-100kW) and that is often useful for small furnaces.

I am not sure about the power and frequency useful for induction welding/ brazing. Some modelling and some experiments are badly needed.

 

[badea]

If you already know, why are you asking...

Actually the useful frequency depends completely on your induction oven geometry. High frequencies work with open magnetic path configurations e.g. an induction cooker. I have seen the mentioned low frequency ovens in a steel mill with a donut melting pot and closed magnetic path. In an open magnetic path with air gap, the high conductive metal simply shorts the magnetic field and doesn't absorb much power.

- - - Updated - - -

Presumed 3 kHz is a useful working frequency and you have 2 µH coil inductance, what's the expected coil current? 50 A, 100 A, 200 A or more? What's the power stage DC supply voltage? I would expect a considerable higher inductance to achieve handy current and voltage levels.

Dear FvM:
My main question was about suitable frequency for AL and Cu not for steel.
my dc bus voltage is about 48vdc,10A (just for test) with half bridge topology
with some measurements i got 300Vpp @30khz, when working coil loaded with stell this voltage drops heavily.
now when i load a working coil with Cu or Al, nothing change, just small temperature raising in part, small drop of working coil voltage.

 

[Warpspeed]

It might be worth contacting some induction heater manufacturers, telling them exactly what your process is, and getting them to recommend suitable equipment.

That might at least get you started on narrowing down the required power, frequency, and coil geometry.

As far as building a suitable tank circuit, I have no real idea what the operating Q of these things normally is, but a guess might be between 5 and 7.
So if you need 10Kw into the load, the circulating power and circulating current is going to be multiples of 10Kw.

The impedance of what you use to drive the tank circuit with is important too.
It could be vacuum tubes run at many Kv, or at only 3Khz it might even be GTO SCRs with only a few hundred volts.

So the actual optimum values of Xl and XC can vary rather a lot.
Its really difficult to know where to begin.
Some expert advice is called for here, its a really specialized field.

 

[badea]

It might be worth contacting some induction heater manufacturers, telling them exactly what your process is, and getting them to recommend suitable equipment.

That might at least get you started on narrowing down the required power, frequency, and coil geometry.

As far as building a suitable tank circuit, I have no real idea what the operating Q of these things normally is, but a guess might be between 5 and 7.
So if you need 10Kw into the load, the circulating power and circulating current is going to be multiples of 10Kw.

The impedance of what you use to drive the tank circuit with is important too.
It could be vacuum tubes run at many Kv, or at only 3Khz it might even be GTO SCRs with only a few hundred volts.

So the actual optimum values of Xl and XC can vary rather a lot.
Its really difficult to know where to begin.
Some expert advice is called for here, its a really specialized field.

Dear Tony:
Last year i made some maintenance to 10Kv, 500Kva tubes welding machine, power stage was driven by back to back SCRs and there was high power vacuumed tube for resonant, but nowi think in modern models vacuum tubes replaced by IGBT modules
anyway, contacting IH suppliers looks good step to start.

regards,

 

[FvM]

My main question was about suitable frequency for AL and Cu not for steel. my dc bus voltage is about 48vdc,10A (just for test) with half bridge topology with some measurements i got 300Vpp @30khz, when working coil loaded with stell this voltage drops heavily. now when i load a working coil with Cu or Al, nothing change, just small temperature raising in part, small drop of working coil voltage.

I already answered the question. In my view, the observation confirms my previous comments. A coil geometry that works for steel with 20 to 30 kHz must be operated with a considerable higher frequency for Cu/Al. The only chance to get some energy absorbed in an open path magnet circuit is to utilize skin effect and heat up only a thin layer of the material.

 

[brushhead]

When I was doing Induction heating, Aluminium melting Furnaces did run at about 3kHz so you're in the right ballpark. This easily puts you in the operating area for an IGBT. Normally for a melter you aim for a Q of around 10, but because you are melting Al is may be necessary to aim for a higher Q factor in order to improve the coupling to your charge. This is why most Furnaces have an output transformer such that they can set the coil volts later during commissioning.

I have to say OP that running this without current or voltage feedback is not a particularly good idea!

In general in a parallel tuned circuit your circulating current will be basically input current times the Q. This means that you will have to be very sure that your tank capacitor can withstand that magnitude of kVA(r). Traditionally you water cool the furnace coil and the tank capacitor(s). What kind of power level are you talking about?

 

[badea]

What kind of power level are you talking about?

About 10Kw

 

[c_mitra]

When I was doing Induction heating, Aluminium melting Furnaces did run at about 3kHz so you're in the right ballpark.

At 3 kHz, perhaps you will need a charge (furnace load) of several tons of Al; you will also need a power level of about 100 kW or so (at 3 kHz operating frequency)

I understand the OP wanted to do welding of Cu and Al tubes; we need high power delivered fast into a localized area so that the metal does not melt but gets close to melting point. I suspect that both frequency and power need to be optimized on a case by case basis.

 

[FlapJack]

For an induction cooktop and Cu and Al pans, typical frequency's are 60 to 80 KHz.