steam engine

[plasma]

 Hi I'm working on a project to make pulses of steam. At present I've got mains voltage double to 640V charging up 30uF 250A caps. Its shorted through a salt water solution.
At present it just making a li5le amount of gas, but will build up the Farads.

Put a 9Vac source across 330ohm resistor and the salt gap, measured 87ohms for the salt gap. I moved the electrodes closer and now get sprays of water out after I short the gap.

[Zipdox]

Holy crap that's dangerous. Using high voltage with salty water to create steam is one of the craziest ideas I've ever heard. I suggest you use a regular resistive heater instead.

Also that's not going to work very well since you need to heat up the entire water mass to 100 degrees before any water can start boiling. Read up on how smoke machines work.

My suggestion is heating a mass of metal and pulsing the water instead with a pulsed pump or something.

[plasma]

Was thinking of using a desil injector, I think 1mL of water takes about 10kJ to electrolysis, 10mL should take 4kJ to heat to 100degrees.

Still testing it at the moment.
Thanks for the feedback.

[klugesmith]

>> 10mL should take 4kJ to heat to 100degrees.

At that point none of the water has vaporized.   You need an additional 2.5 kJ for each mL of water turned to vapor.

I made a water resistor for calmly discharging a 10 kJ 22 kV pulse duty capacitor.  About 150 mL in a plastic bottle with two sheet-copper electrodes, and enough copper sulfate to get about 100 ohms.  When connected to capacitor charged to 1 kJ, there's no sound.  Kilovoltmeter reading drops to zero, the water warms up slightly, and one of the electrodes turns black.

[HiVi]

What salt do you use?

Maybe use only spark gap in distilled water so you do not have any byproducts from salts?

If such high voltage is not possible, resistive heater is also good idea. With preheated water and insulated container.

What is the idea/application? Steam engine as in thread subject?

[plasma]

@HiVi
NaCl at the moment, but ammonium sulphate it disovoles better ,lower ohms.
Destilled water could work, but I'm trying to keep the voltage lower to minimize surface tracking.
I'm using the water as reaction mass for jet type thing, well what started the idea.
A resistance heater would defiantly work, if the above doesn't work I'll fall back to that, but the pressure for size will be lower.

[klugesmith]

Look up resistojet technology, used in space for more than 50 years when there's electric power to spare.

Here's a contemporary engine that uses vaporized water as a propellant, with Isp around 100s, at a scale suitable for cubesats.
https://aurorapt.fi/2020site/wp-content/uploads/2020/08/AURORA_Single_Resistojet_Thruster_V3.pdf

[plasma]

Good idea about resistance heating, if I pre heat the water to 70 degrees, I would knock a lot of the energy needed.
Installed more capacitor, making 45uF 9J it arcs inside the water between the electrodes with a blue colour.
Going to look at Scr the screwdriver switch will be maxed out soon.

[plasma]

Tried different spacings for the electrodes, settled on 2mm, it is more reactive with a little moisture between electrodes rather than complete submerged.
Trying to get the resistance as low as possible, the 2mm gap is around 25ohms, it quickly climbs when expanded, making a punny crack when shorted.
Next couple of days should get some more capacitor taking it up to 60uF.

[plasma]

In a couple of weeks I should be getting https://www.digikey.co.nz/product-detail/en/ixys/CS30-16IO1/CS30-16IO1-ND/1651204 Scr its rated at 1.6kV 50A avg, I think a lot of energy is wasted in the mechanical switch.
Got some more capacitor taking it up to 60uF. At present it takes 4 shots to remove all the water in the salt gap, most of it gets blown out.
Later when I've sorted out the water injection I'll get a SSR traic and 555 to switch off the charging circuit then switch on the Scr.
This Wednesday I'll pickup two more capacitor.
Tried putting a magnet under the salt gap, it wasn't as energetic.

[plasma]

I tripled main voltage up and have three capacitor in series ,at 20uF it now only takes two shots to remove all the water.
The DMM reads 510V across one of two 1W 10Mohm voltage divider.

[plasma]

Expanded the gap between electrodes to 5mm, it makes a 0.5mL volume.
NaCl solution was put in the gap, measuring the voltage across the 47ohm gave 9.28V from 10.4V, about 6ohm for the salt gap.
NH4SO4 solution gave 9.6V ,4ohms.
The ammonium sulphate solution gave a burn smell when fired.
At higher resistance solution give larger bangs.

1000V cap bank into 4 ohms might be to much for solid state components, might have to think about mechanical.

[Twospoons]

1000V cap bank into 4 ohms might be to much for solid state components, might have to think about mechanical.

You have a way to go before abandoning solid state switching - all that changes is you have to be a bit more rigorous with your design.
Think hockey-puck thyristors, with appropriate drive.

[plasma]

I'll check them out thanks. There's a problem with Scr the cap bank after a shot normal has 20-140V still on it, which might stop it switching off. Do you think if I short the cap bank after the voltage has dropped to 100V?

[davekni]

Thyristors switch off when current reaches close to zero.  Voltage doesn't matter.

The issue that will need care with thyristors is the parameter called "critical rate of rise of on-state current".  It's often about the same for large thyristors as for small, in the 50-150A/us range.  It may require an inductor in series to limit current rise rate.

[plasma]

The Scr I linked above has 150A/us and a holding current of 100mA.
It says the gate current is 55mA max,min 15mA at 6V if I keep it at 40mA will it work(haven't used one before)
The voltage on the cap bank I was meaning it can still supply holding current if the resistance is less than 1kohm.
At 100A/us does that mean in 4us it will be at 400A, afte how do you size the inductor as after 1Tc more current will flow?
Thanks

[davekni]

The CS30-16io1 part looks great for your use, unless you are going to fire very rapid repeats.  What is your eventual intended pulse frequency?

Holding current is 100mA or less per spec., so could be say 50mA.  As long as your charging circuit is switched off, or is charging with well less than 50mA, then the thyristor will turn off after discharging the capacitor.

Gate trigger current is specified at 55mA max (80mA max at -40C).  That's the worst-case needed to trigger.  So, you need to supply at least 55mA of gate current.  200mA would be fine, or even 2A for a short trigger pulse.  Higher gate currents help a bit with current rise rate handling.

Current rise rate is voltage/inductance.  For 1000V, 7uH would give 1000V/7uH = 143A/us.  7uH is reasonable to achieve with an air-core coil of wire.  Then you don't need to worry about core saturation at 250A peak current.  If you need to use a smaller inductor for some reason, then high gate current may allow operation up to 300A/us (3.3uH inductor).

Surge current rating is 340A for 10ms with 150C junction.  For your short pulses, currents up to 700A (or even more depending on frequency) should be fine.  1000V/4ohms = 250A, so no issue for CS30-16io1.

[plasma]

I'm aim for 1ms pulse every second. At the moment once the cap bank is charged I unplug the charging circuit, but do plan to use a traic to switch it off then trigger the Src ,rinse repeat, but that's going to wait tell I get a injector for the salt water.
Checked out foumlas, I think you use I'll=1/0.000007*(1000*0.000001) =143A
Nice post cheers.

[plasma]

Changed the salt gap to this, I'll add a top with a small hole to convert the pressure to velocity at a later stage.
I'm going to have to minimise the salt gap electrodes, thinking maybe 0.25mL with two shots a second.
The power supply will be a 4-8kW petrol generator, well the boundary I'm working in
Having minimal water in there makes a more reactive shot, I'm guessing the energy is more constructed.
Added 136mH between the charging circuit and cap bank, I started blowing diodes when firing.
Selected a better Scr 1200Apk in case the resistance of the gap drops to 1ohm.

[davekni]

Why 4-8kW generator?  Your 20uF cap charged to 1000V is 10J energy.  Twice per second is 20W.  Even if charging circuitry is inefficient, 4-8kW seems like an over-kill.  A battery and flyback inverter might be more practical for portable use.

A charging series-inductor is a great idea.  That can double efficiency compared to resistive charging.  However, it can double voltage.  If starting at 0V on the capacitor, connecting to a DC supply through an inductor makes a series-resonant circuit.  Analog simulation (LTSpice or one of the other free tools) would be a great way to experiment with charging circuits without frying parts.