induction heating diesel injection lines

[gearhead]

The title says it. I plan to use induction heating on the injection
lines
on the 6-cylinder diesel engine in a pickup truck in anticipation of
converting
it to run on vegetable oil. (Also more immediately it will help this
old
engine run better on diesel while warming up. We've had some record
cold.)
I think I'll use center-tapped coils and drive them push-pull with
some chip like the SG3525.

First, a question.
What frequency should an induction heater like this run? This relates
to the magnetics of steel, obviously. The injection lines use high-
grade steel. I ass-u-me fairly
ductile, because they use flare fittings and must resist work
hardening
from constant vibration.

 

[MooseFET]

The title says it.  I plan to use induction heating on the injection
lines
on the 6-cylinder diesel engine in a pickup truck in anticipation of
converting
it to run on vegetable oil.  (Also more immediately it will help this
old
engine run better on diesel while warming up.  We've had some record
cold.)
I think I'll use center-tapped coils and drive them push-pull with
some chip like the SG3525.

First, a question.
What frequency should an induction heater like this run?  This relates
to the magnetics of steel, obviously.  The injection lines use high-
grade steel.  I ass-u-me fairly
ductile, because they use flare fittings and must resist work
hardening
from constant vibration.

I'd say that you should assume that you will heat with eddy currents.
This suggests 100s of KHz as the running frequency.

 

[Jamie]

The title says it. I plan to use induction heating on the injection
lines
on the 6-cylinder diesel engine in a pickup truck in anticipation of
converting
it to run on vegetable oil. (Also more immediately it will help this
old
engine run better on diesel while warming up. We've had some record
cold.)
I think I'll use center-tapped coils and drive them push-pull with
some chip like the SG3525.

First, a question.
What frequency should an induction heater like this run? This relates
to the magnetics of steel, obviously. The injection lines use high-
grade steel. I ass-u-me fairly
ductile, because they use flare fittings and must resist work
hardening
from constant vibration.

Sounds bizarre, hope you don't have a loose fittings, the arc will
indubitably display some nice side effects.



http://webpages.charter.net/jamie_5"

 

[Tom]

Why not wrap the lines with resistive heater tape? That's cheap and
common and very reliable.

John

Induction heating sounds like more fun though.

The guy on this web page is using about 200 khz.
http://www.richieburnett.co.uk/indheat.html

You won't need as much heat has he got but he also describes several
ways to limit heating.

Tom

 

[gearhead]

Induction heating sounds like more fun though.

The guy on this web page is using about 200 khz.http://www.richieburnett.co.uk/indheat.html

You won't need as much heat has he got but he also describes several
ways to limit heating.

Tom- Hide quoted text -

- Show quoted text -

Thanks for the link.
That guy mentions skin effect. I'd rather go for internal heating
(eddy currents and hysteresis loss).
Check out Tim Williams' induction heater page
http://webpages.charter.net/dawill/tmoranwms/Elec_IndHeat6.html
Scroll down to the oscillator diagram. After a quick check of the
SG3524 datasheet it looks to me like it would run at 5 kHz if you
leave the "frequency control" open.
I'll admit I know nothing about designing an induction heater. I had
fancied that it might not be too complicated, just a matter of driving
the coils. But there could be power factor issues.

Here's the link that got me started thinking about induction heating
http://www.dieselveg.com/injection_pipe_heating.htm

 

[PeterD]

The title says it. I plan to use induction heating on the injection
lines
on the 6-cylinder diesel engine in a pickup truck in anticipation of
converting
it to run on vegetable oil. (Also more immediately it will help this
old
engine run better on diesel while warming up. We've had some record
cold.)
I think I'll use center-tapped coils and drive them push-pull with
some chip like the SG3525.

First, a question.
What frequency should an induction heater like this run? This relates
to the magnetics of steel, obviously. The injection lines use high-
grade steel. I ass-u-me fairly
ductile, because they use flare fittings and must resist work
hardening
from constant vibration.

Consider looking at prior art. There are so many better ways to run
veggie oil that an exotic induction heater seems like more problems
that it is worth.

 

[gearhead]

I'm a newbie at magnetics but I'll give this a shot.
o ...ducks flying books..)

The Goldilocks point for this is driving hard near saturation at the
frequency where the permeability cuts off.
That is if you want the least amount of turns.
(A spin off idea is to use winding heat + induction heat. So more
turns. Perhaps a nichrome wound inductor?)

Your core power follows the steinmetz equation. Lots of flux and use
the highest frequency that doesn't degrade the inductance.
You want to swing near -Bsat to +Bsat and that takes time.
Too high f and you get less eddy heat and an overloaded driver.
(That could be incorrect. The flux intensity is dropping but there's
more frequency flux changes. Sweet spot?. )

Google for steel permeability vs frequency.
I'd just do an experiment to test ur vs f and another experiment for
Bsat.

Perhaps another experiment cause permeability varies with temperature.

D from BC
myrealaddress(at)comic(dot)com
British Columbia
Canada- Hide quoted text -

- Show quoted text -

Yes, maximize B.
Steel saturates at about 1.6 T or something like that.
So with one or two dozen turns we're talking about a few kiloHertz.
Let me ask you (or whoever might know) whether the following equation
is correct for a coil driven with a square wave?
B=E/(N Ae 4 f)
B: Teslas
Ae: m^2

 

[Nico Coesel]

The title says it. I plan to use induction heating on the injection
lines
on the 6-cylinder diesel engine in a pickup truck in anticipation of
converting
it to run on vegetable oil. (Also more immediately it will help this
old
engine run better on diesel while warming up. We've had some record
cold.)
I think I'll use center-tapped coils and drive them push-pull with
some chip like the SG3525.

I'd use resistive heating in the fuel line running towards the fuel
pump. Heating after the fuel pump won't be necessary and also
introduces errors if the engine is controlled electronically. Also
make sure not to heat the fuel too much otherwise too little fuel may
be injected. Room temperature is enough.

 

[gearhead]

E is average voltage.
So I suppose that yeilds Bavg.

But you might want Bpk

Bpk = Epk * 10E8* (pulse time)
      ----------------------- gauss
         N*Ae
Ae cm

1T = 10000gauss

I'm guessing you could try a resonant design if you need a crapload of
supply voltage.

D from BC
myrealaddress(at)comic(dot)com
British Columbia
Canada- Hide quoted text -

- Show quoted text -

I don't get why they call it "peak."
Doesn't that equation just address a flat pulse of voltage E of
duration t
and Bpk equals B at time t
or did I miss something?

 

[gearhead]

No need to look at permeability just yet. The equation for flux in a
voltage-driven coil doesn't include a term for permeability.
I reviewed the theory last night and that equation I gave for a coil
driven with a square wave
Bpk = E / ( 4 f N Ae)
still looks quite correct to me.
So if steel saturates at 1.6 T and I say I want Bpk = 1 Tesla and I
have 20 turns
on an injection line with 8mm diameter, I get f = 16 kHz
Unfortunately, the equation for skin depth _does_ include a term for
permeability,
and steel's high permeability means a very shallow skin depth of only
about .04 mm
at 16 kHz. No way reduce the frequency enough to get eddy currents
_inside_ the tubing, either.
I'd have to drive it at about 10 Hz.
I can hope though, that magnetic hysteresis from the high B provides
some internal heating.
I should stop typing and start experimenting? Perish the thought!

Epk actually means Epp. The peak to peak voltage of the pulse.

Epk suggests a discontinous voltage. IOW..a pulse.

Just E alone suggests DC. This is not a DC circuit.

Bavg.. suggests a ripple. Bavg = /\/\/\/\/\/\/\/\/
Same goes for Eavg.
It's to account for pulse duty.

                                 _______Max
                                / /
                               / /
Bmax, Bpk...suggests a limit. / /
                     Max    _/ /  

That's the B you need when you compare to the saturation level.

Recall I mentioned 'near' saturation. I meant less than saturation.
If you go over the curve you loose magnetizing efficiency.

Ok...I might botch the math and theory but here goes...

Inductor Energy
E = LI^2 Joules

Specific Heat Capacity
Vegatable oil = 1.67 KJ/KgK

Specific heat
E = C*m*(dT)

So if you had 10 amps, 1H and 100% efficiency then
transfering 100 Joules of EM energy to heat energy will raise 1gram of
oil about 60K or 213.15 degrees.

Good enough to cook 1 french fry.

You mean one french fry per second (f^2 s^-1)?
Plenty enough to make my truck happy at idle, just maybe not on the
highway.

 

[Rich Grise]

The title says it. I plan to use induction heating on the injection lines
on the 6-cylinder diesel engine in a pickup truck in anticipation of
converting
it to run on vegetable oil.

Well, I've held my tongue here, but this thread is getting silly.
Induction heating won't buy you anything over plain ol' resistive. Since
deisel fuel and Mazola oil aren't conductors, they won't get inductively
heated - only the fuel line will, and you'd have better luck with that
by just passing a current through it, or lash-up some kind of heating
element; someone mentioned "heat tape", which I notice didn't get much
of a response.

Hey, I might sound contrary, but I think my idea is better. ;-)

Have Fun!
Rich

 

[linnix]

Well, I've held my tongue here, but this thread is getting silly.

I've been holding too, but...

Induction heating won't buy you anything over plain ol' resistive.

Induction heating won't heat up the coil, as heating coil would.

 

[PhattyMo]

Thanks for the link.
That guy mentions skin effect. I'd rather go for internal heating
(eddy currents and hysteresis loss).
Check out Tim Williams' induction heater page
http://webpages.charter.net/dawill/tmoranwms/Elec_IndHeat6.html
Scroll down to the oscillator diagram. After a quick check of the
SG3524 datasheet it looks to me like it would run at 5 kHz if you
leave the "frequency control" open.
I'll admit I know nothing about designing an induction heater. I had
fancied that it might not be too complicated, just a matter of driving
the coils. But there could be power factor issues.

Here's the link that got me started thinking about induction heating
http://www.dieselveg.com/injection_pipe_heating.htm

You can catch Tim,and other smart folks at www.dutchforce.com too. ;-)

 

[gearhead]

The title says it.  I plan to use induction heating on the injection
lines
on the 6-cylinder diesel engine in a pickup truck in anticipation of
converting
it to run on vegetable oil.  (Also more immediately it will help this
old
engine run better on diesel while warming up.  We've had some record
cold.)
I think I'll use center-tapped coils and drive them push-pull with
some chip like the SG3525.

First, a question.
What frequency should an induction heater like this run?  This relates
to the magnetics of steel, obviously.  The injection lines use high-
grade steel.  I ass-u-me fairly
ductile, because they use flare fittings and must resist work
hardening
from constant vibration.

diesel injection line homebrew induction heater project lash-up:


12 volts
|
+-----+-------,
| | |
| 0.1 = 1000 =
| | |
| gnd gnd
|
________|___________
| IR2153 half bridge | 0.68
| driver with BUZ71A |---||----,
| mofsets and pot for| |
| adjustable freq. | |
|____________________| |
| |
| peak voltage |
| detector |
| ,---+----|<----------+
| | | |
| 0.1uF = 1 Meg 12 turns )||
| | | of 14 gauge )||
| +---' stranded copper )||
| | wound on a )||
| | screwdriver shaft )||
| | |
| | |
'-------+--------------------+
|
gnd

I used a screwdriver about the same thickness as the steel injection
lines on my diesel engine.
I adjusted the pot on the driver until I got the highest voltage
reading at the peak detector, then nudged the frequency up a bit.
It ran 140 kHz and showed about 30 volts peak at the coil, and drew
about 3 amps from the battery.
I pointed an infrared thermometer at the coil.
In about three minutes the temperature rose from 80F to 140F.
I kept it running for a while, until the screwdriver got very hot.
The mosfets got hot too, so I scoped the gates in the half bridge.
Bottom mosfet gate waveform not too bad. High side mosfet looked
horrendous.
Waveform at the coil looked real weird too, like somebody took a
machete to a sine wave.
By the time I finished scoping the circuit at different frequencies
and so on, I couldn't touch the screwdriver.
I see the feasability of induction heating and I have an idea for a
much better driver.

 

[[email protected]]

diesel injection line homebrew induction heater project lash-up:

     12 volts
         |
         +-----+-------,
         |     |       |
         | 0.1 =  1000 =
         |     |       |
         |    gnd     gnd
         |
 ________|___________
| IR2153 half bridge |  0.68
| driver with BUZ71A |---||----,
| mofsets and pot for|         |
| adjustable freq.   |         |
|____________________|         |
  |                            |
  |  peak voltage              |
  |    detector                |
  |       ,---+----|<----------+
  |       |   |                |
  | 0.1uF =  1 Meg     12 turns )||
  |       |   |     of 14 gauge )||
  |       +---' stranded copper )||
  |       |          wound on a )||
  |       |   screwdriver shaft )||
  |       |                    |
  |       |                    |
  '-------+--------------------+
                               |
                              gnd

I used a screwdriver about the same thickness as the steel injection
lines on my diesel engine.
I adjusted the pot on the driver until I got the highest voltage
reading at the peak detector, then nudged the frequency up a bit.
It ran 140 kHz and showed about 30 volts peak at the coil, and drew
about 3 amps from the battery.
I pointed an infrared thermometer at the coil.
In about three minutes the temperature rose from 80F to 140F.
I kept it running for a while, until the screwdriver got very hot.
The mosfets got hot too, so I scoped the gates in the half bridge.
Bottom mosfet gate waveform not too bad.  High side mosfet looked
horrendous.
Waveform at the coil looked real weird too, like somebody took a
machete to a sine wave.
By the time I finished scoping the circuit at different frequencies
and so on, I couldn't touch the screwdriver.
I see the feasability of induction heating and I have an idea for a
much better driver.

I have to wonder what your car insurance company would say if your
truck caught fire because you were heating your fuel lines.

Why not biodiesel? It's quite easy to make.

Is your fuel pump located in the fuel tank? If so, you've got a
problem.

Michael

 

[gearhead]

diesel injection line homebrew induction heater project lash-up:

     12 volts
         |
         +-----+-------,
         |     |       |
         | 0.1 =  1000 =
         |     |       |
         |    gnd     gnd
         |
 ________|___________
| IR2153 half bridge |  0.68
| driver with BUZ71A |---||----,
| mofsets and pot for|         |
| adjustable freq.   |         |
|____________________|         |
  |                            |
  |  peak voltage              |
  |    detector                |
  |       ,---+----|<----------+
  |       |   |                |
  | 0.1uF =  1 Meg     12 turns )||
  |       |   |     of 14 gauge )||
  |       +---' stranded copper )||
  |       |          wound on a )||
  |       |   screwdriver shaft )||
  |       |                    |
  |       |                    |
  '-------+--------------------+
                               |
                              gnd

I used a screwdriver about the same thickness as the steel injection
lines on my diesel engine.
I adjusted the pot on the driver until I got the highest voltage
reading at the peak detector, then nudged the frequency up a bit.
It ran 140 kHz and showed about 30 volts peak at the coil, and drew
about 3 amps from the battery.
I pointed an infrared thermometer at the coil.
In about three minutes the temperature rose from 80F to 140F.
I kept it running for a while, until the screwdriver got very hot.
The mosfets got hot too, so I scoped the gates in the half bridge.
Bottom mosfet gate waveform not too bad.  High side mosfet looked
horrendous.
Waveform at the coil looked real weird too, like somebody took a
machete to a sine wave.
By the time I finished scoping the circuit at different frequencies
and so on, I couldn't touch the screwdriver.
I see the feasability of induction heating and I have an idea for a
much better driver.- Hide quoted text -

- Show quoted text -

I now have a driver circuit made solely of discretes that powers a
tank circuit with the heating coil and a high quality capacitor (the
cap doesn't get hot, even carrying over six amps at 90 kHz).
The scope shows a very clean sinewave at the coil, and the screwdriver
gets too hot to touch in a matter of seconds. Well, at 75 watts what
would you expect.
I can fix that by increasing the turns count I think.

 

[gearhead]

Your heat energy is due to copper resistance and steel core
resistance.
I^2Rsteel where I is due to core eddy currents.
I^2Rcopper where I is the inductor current.

You can add turns for more magnetic flux (if inductor current is made
the same and core perm. ur stays constant) but you increase copper
heat.
Decide how much copper heat to have.

This project is a math/physics project.
It's also a cooking project.. Not too much salt..not too much
pepper..just a dash of parsley.
And I'm not a magnetics expert.
Maybe there's a magnetics simulator program out there?? Dunno..

Maybe this will work as a design approach..
1)Select an ac coil current..
2)Take rms of above and run dc current through wire spread all over
the place.
(Don't cause a fire.)
3) Start winding.
4) Stop winding when your coil temperature is at risk of lamination
breakdown. Or whatever toasty level.
Next, AC testing:
5) Find Bsat and -Bsat
6) Find the time it takes to go from -Bsat to +Bsat.. That will
determine your operating frequency.
(I think it begins to be hard on the driver to compensate with more
current to get back up to Bsat at higher frequencies. <<might be
right)
7) Set you driver voltage to get the drive current picked in step 1.
Use frequency from step 6.

Taa daa...Max induction heating. <<maybe.

Note: I've neglected other copper losses in #2.
The above might be full of errors. But..it's gotta be close..

D from BC
myrealaddress(at)comic(dot)com
British Columbia
Canada- Hide quoted text -

- Show quoted text -

Sheesh, didn't you read my last post? I have a prototype that kicks @$
$, all I have to do is get away from the bench now so I can work with
the actual injection line (will have to do this under the hood,
unfortunately) and adjust the number of turns to get optimum power
level. Or I could need to do some other tweaking, like change the
capacitance in the tank circuit. No big deal.
Copper resistance accounts for negligible loss. No detectable warmth
in the wire. I also calculated the skin depth, effective cross
section, yada yada yada. Nada.

Work remains, I have to build a feedback loop with a thermistor to
regulate the temperature. Just a comparator with hysteresis, turn the
shebang off when it's nice and toasty, then turn it on again a few
degrees lower.
I ordered teflon-insulated wire for the coils, some nice mosfets...

 

[Jasen Betts]

On Sat, 24 Jan 2009 14:25:21 -0800 (PST), gearhead



I suspect heating up an injection tube with heat bleeding endlessly
along the metal and the heat capacity of the oil will take more energy
than raising the temperature of a screwdriver.

one end is connected to the cylinder head the other has cold fuel
coming into it, once the head heats up not much heat will be lost that
direction, and the incoming cold fuel will make good use of any heat
leaking out that end.

And...a tube has a lower cross section[1] than a rod, so I think it's
going to take even more changing flux to get a tube hot.
(You should be using an oil filled tube for experiments, not a
screwdriver.)

more of a problem is that the screwdriver was probably made from a
completely different type of steel to the pipe. different magnetic
properties and different conductivity.

So.that's why I suspect you might bump into a copper heat problem if
it's required to 'beef up' the coil and drive harder for higher flux
density at the max.frequency.

(Note vegetable oil and water have roughly the same heat capacity.)

vegetable oil has about half the heat capacity of water.
(as do most other types of oil)

All depends really on how much heat you need and how fast.

yeah, flow rate and heat requirements figures would be useful.