20 Kilowatt Plasma Blaster Engine Design

From:  Scott Stephens [SMTP:stephens-at-enteract-dot-com]
Sent:  Wednesday, June 24, 1998 1:36 AM
To:  tesla-at-pupman-dot-com
Subject:  20 Kilowatt Plasma Blaster Engine Design

So what about my possibly clever idea for a flame-excited Tesla Coil? Too
boring and lame? Maybe I was too flippant and irrelevant to be taken
seriously by the list. Maybe its ridiculous to think that a microwave
excited plasmoidal bubble wouldn't pinch and/or rupture when magneticaly
compressed in a TC, or when the fuel sprayed in it explodes.

What is the equation for "Pondermotive" force? That is, the compressive (and
expansive) force undergone by a current loop when it is fluxed in a
compressive or expansive magnetic field? I would think its a form of
Ampere's law, F(dynes)=BIL (cgs).

In this (second) design iteration, I come to the conclusion that 1 joule in
a .1mH inductor can put a 50 atmosphere (700psi) squeeze on a 12cm plasma
current loop for 10uSeconds.

I would most likely spray propane-air from a propane torch or bunsen into a
microwaved plasma, around 2cm radius, in around 10 microseconds. This plasma
will be pulsed, oscillating in volume, around 2000 times/second.

Fuel compression energy =1 Joule

Pulse capacitor: E=.5cv^2, 1J+=1.4KV,1uF (maybe could use a cheap, small 10J
photo-flash caps if ESR, pulse width and fuel character allow, and a step-up
winding on inductor)

Inductor aka Tesla Coil (magneto-sonic plasma transducer; pulsed by cap
through triggered spark, Xe strobe tube or SCR)
E=.5 li^2 = 1J=.5*.1mH*(140A)^2,
V= L di/dt = , 1.4KV=.1mH dI(140A)/dt (10 microseconds)
N=(L(9A+10B)^.5)/A; L is 100uH; A is 1" Rad; B is 1" Len; So N = 43 Turns
B= V T 10^8/NA= 1400V*10^-5*10^8/43*PI*2.54^2=1.6Kgauss or .16 Tesla

K notwithstanding, I'll assume a 43:1 step-down ratio, and a pretty low < .1
ohm plasma resistance, and .1uH for a 1" radius plasma-loop inductance:

Plasma I = BNA/L10^8 = 1.6Kgauss*1*20cm^2/.1uH*10^8=3200A (close enough to
the 33:1 xformer assumption (4600A) 
Perfect-coupling xformer step-down estimate 33:1 * 140A = 4.6KA
Realistic .5 K suggests 2000A induced current, so

=1.6Kgauss*2000A*(2*PI*2.54cm)= 51 mega-dynes/cm^2 = 50 atmospheres
Plasma Vol.=4PI(2.54^3cm)/3=68cm^3 or around 70cc

So for 1 Joule + of energy from my capacitor, I put maybe a 700 psi squeeze
on my 70cc plasmoid  in 10 microseconds in a 1.6 Kilogauss field using my
.1mH 43 turn, 2.5cm radius, 2.5cm long air core coil.

If my resonant cavity can accept a ferrite transformer core, and I can have
a K near 1, pulse it around 50uS, and fuel temperature and plumbing
permiting, a very nice compact motor could be constructed, with
semiconductors for realistic effiecency :-)

Propane: .5kg/l; 46MJ/kg

For 10% efficiency (10J/pop), I'll want to burn .22mg and .00044 l per pop;
For 2000 pops/second = 1ml/second propane flow; at 5% explosive limit
mid-range, air volume = 20ml/second; plasma core around 5cc so gas flow rate
is .05 meters/second.

Reactor power is 20 kilowatts. 2 Kilo-watts sustain magnetic oscillations,
and the rest make heat, with a howl like a banshee }:-O

My latest effort has been to search for EM codes to design the resonant
cavity. NEC uses the method of moments technique to calculate fields, which
may not be good for microwave cavities. APLAC (free version) uses FDTD, but
is too limited. And none model ferrite, which probably is too complicated
and material-dependant anyways. And what will the plasma E-eff be as the
field is pulsed?

The image of a metal, bi-conical center, with gas inlets/exhaust ports,
surrounding a ferrite core, all in a TEM 10 cavity comes to mind.

Scott Stephens