Engine for Starship Drive

Most interstellar voyages will take about two and one-half years, subjective time.
People who expect generation ships aren't thinking clearly enough.
Certainly, ships for such long term voyages are much easier to conceive.
But people won't build such ships, and won't board them,
because people can recognize an eminently stupid idea when they're hit over the head with it.
Faster is better, and nowhere is this principle quite so obvious as in starship design.
It should be taken for granted that starships will be the fastest vehicles which can possibly be built.
 

The most reasonable assumption to make about a starship's operating characteristics
is that of constant acceleration to midcourse, a rapid turnover at that point,
then a constant deceleration to the destination.
It is also reasonable to presume that the constant acceleration will be at about one gravity.
That is chosen for convenience to the crew. To double that would cause extreme discomfort,
and would be extremely deleterious to the health when maintained for two years.
To reduce the acceleration below one gravity would be perceived as wasting time.
 

It takes about one year's acceleration at one g to approach the speed of light closely.
At that velocity, subjective time aboard the ship is nearly stopped.
That is why turnaround from acceleration to deceleration must be extremely rapid.
Then it will take about one year to decelerate from relativistic velocity to rest in the destination system.
So each starship voyage must take at least two years, and won't take much more.
On longer voyages, the extra space will be traversed at relativistic velocity.
That means it all happens in the middle of the trip, during the turn around time.
So the difference will be scarcely perceptible, in fact negligible compared to two and a half years.
But a few months should be allowed for piddling around at each end, at in-system speeds.
So every starship voyage will take about two and a half years.
That should turn out to be an empirical rule. Make sense?
 

So that is the basic constraint to which starship drive design should be compliant.
This approach makes starship drive design seem more tractable,
than gasping in awe at the magnitude of the difficulties from the very beginning.
It's entirely too obvious that it won't be easy to come up with a drive which will provide
a one g acceleration for a couple years at a time. Don't bother saying that.
That's what we're engineers for. Were it easy, any fool could build starships.
 

Now that we know what we need to build, we can start lining out the constraints
we will need to consider to actualize the drive system.
Let us start with the assumption we will use a self-contained reaction drive.
(The only option which offers feasibility is to rely on a big laser at a pole of Mercury,
to deliver energy to the ship in transit; but for right now, let's consider a self-contained drive.)
The first obvious question concerns the fuel problem.
More exactly stated, this is the reaction-mass problem:
a reaction engine has to throw something backward to accelerate the vessel forward.
Dr. Broussard evidently pioneered the observation that hydrogen could be gathered up
along the way, out of nominally empty interstellar space,
when he designed the ramjet principle which bears his name.
This concept may turn out to be crucial in making interstellar flight possible,
for it relieves much of the burden of carrying really silly amounts of fuel on board.
The first problem with the magnetic ram scoop is that it only picks up ionized hydrogen.
Neutral atoms or molecules ignore magnetic fields.
Interstellar gas is pretty thin stuff to start with,
so we can't afford to let all the neutral gas just slip by us.
In principle, we can enrich the ionic population ahead of the scoop with an electromagnetic beam
of some sort, ionizing the gas with a laser or RF maser tuned to ionize hydrogen.
Creating a shock wave ahead of the scoop in this manner will give us a larger proportion
of ionized hydrogen which can be gathered into the scoop, and a small amount
of the atomic gas will in effect be entrained by our captive protons rushing in.
 

Tentatively, we can count on the magnetic ram scoop to pull in hydrogen,
all the way up to relativistic speeds. Now we can consider the energetics of the drive.
My preference is always for the proton-boron reaction to provide fusion.
The basic reason for this is because it's cleaner than other fusion reactions.
It doesn't yield free neutrons, and neutrons pollute.
A relativistic space ship has lots of radiation problems already,
without the drive even being considered.
So it is vastly preferable to minimize the extent to which the passengers
are getting roasted by their own engines, in addition to the hard stuff coming in from the front.
We should skip the neutrons, fast or slow or in between, from the drive if we can.
So the transported component of the actual nuclear fuel should consist of solid boron.
(Some science fiction writer of the fifties must have worked out this solution,
because I have a vague memory of wondering why boron had been selected as a fuel.)
 

The actual fusion reaction occurs in the jet behind the vessel.
Again, this is for the sake of cleanliness, because the heat of the proton-boron reaction
will ignite the proton-proton reaction, which isn't free of neutron pollution.
The binding magnetic field can be extended well back into the jet,
by spiraling beams of electrons around its positiively charged core.
The protons thus contained will be performing their helical dance around
the confining magnetic lines of force, at their cyclotron frequency.
The accelerated boron ions zipping through this space are so fast,
they make the whirling protons look like they are standing still.
Popping a pair of them, they turn into helium and bright light.
This excites the neighborhood protons so much they blunder into each other.
When they do that, they turn into deuterium, bright light, and a positron.
The positron looks up an electron, and together they make bright light.
Other secondary reactions occur, all giving bright light.
The ship, as a result of all this, gets an urge to move to a quieter neighborhood.
 

The fusion of the boron ions with protons is largely serving as the ignition source
for the dominant proton fusion, which is much more plentiful due to the vast excess
of protons in the jet. It is not strictly catalysis, but a necessary heat source.
The boron is accelerated by a linac for effectively continuous output.
Perhaps a Cockcroft-Walton generator would provide electrostatic injection into the linac.
The net consumption, after scavenging and recycling unionized boron from the vapor phase,
of elemental boron powder and rod, might hopefully be measured in grams per minute,
rather than in grams per second, to drive the big ship at one gravity acceleration.
After laser vaporization and ionization in an arc, the ions are split off by a magnetic field,
then injected electrostatically through a large potential difference into the linear accelerator.
Hot switchable redundancy in these preliminary stages would be a method to provide
continuous online operation of the drive, and cheap at the price.
Cavity resonators at the linac segments reduce energy needs for the accelerator.
Reducing the load on power electronic components, they reduce vulnerability to device failure.
 

Main shipboard power including drive power comes by conversion of radiant energy
from the fusion reaction in the jet, which will very likely make vast excess available.
Vast excess of available energy is splendid when it leaves wide margin for
unforseen contingincies, which might arise on exploration voyages.
Obviously standby power sources must be available, for possible drive maintenance.
Every starship crew is a long way from help, so foresight is essential in ship design.
A starship is big and costly, and is a fraction of the total human energy not negligible.
Not to stress the point, but our species can ill afford to write off any ship or crew.
At the same time, no ship leaving the system will be back tomorrow.
There will be no experimental test flight of a prototype vehicle.
Since a relativistic shakedown cruise will last a minimum of four years Earth time,
there is simply no point in sending a starship to an empty point in space,
just to keep it in the neighborhood, for it will not be any more reachable for rescue
by being light-months out from the system rather than light-years out.
Radio and lightbeam contact can be reliably maintained for nearly the first and last years
of a starship's voyage. It makes most sense to send the first starship to a star system
on its maiden voyage, rather than on a dangerous jaunt inside the Oort cloud.
(There are very good reasons for working inside the Oort cloud, and the Kuiper belt.
There are no good reasons for ever building any more slow spaceships.)
 

The best way Nature shows us to protect against bombardment by charged particles
is with a magnetic field. This concept can be extended to our starship design.
We are using all our might and main to draw every charged particle within reach
of our far-flung magnetic ram scoop field, to the vicinity of our ship.
Simply by reversing the field polarity in the most immediate vicinity of the ship,
forming a shock cone extending far into space beyond our bow,
we can confine the streaming protons to a laminar sheath surrounding the ship proper.
Tapering off the internal protective field after the stern of the vessel,
we squeeze down this column of protons into the reaction zone,
which serves the purpose of a nozzle throat for our engine, where we initiate fusion.
When we design with hot stuff, we substitute fields and shock configurations for engine linings,
and we substitute streams of electrons for magnet windings whenever we can.
In every possible instance, we substitute dynamic plasma configurations for solid structures.
That is the general design principle for fusion engines.
Plasma structures can take the heat lots better than solid components.
They have additional advantages: they can be shed in an instant should they become unstable,
becoming mere knots of turbulence far back in the wake,
and it takes little more time to rebuild them out of fairly empty space.
A dynamically-structured engine is disposable as a paper towel.
 

Unwanted absorbtion of radiant heat can be reduced by interposing a mirror of colder plasma.
When it is essential that some solid surface be exposed to high radiant heat,
it should be the best material we can get to move the absorbed heat elsewhere.
The best is a heat pipe surface of C12 diamond with capillary channels of He3;
often less than the best may be acceptable.
A general preference is the use of ceramics made of lighter elements,
over the use of metals in general structural components.
They are very much stronger and harder and take more heat than metals.
Also, after radiation damage, they don't hold such poisonous residual by-products.
This selection of materials can easily save half the weight of any arbitrary structure.
The structure will survive stresses which would leave a steel structure crushed or melted.
Finally, it seems psychologically easier, living surrounded by ceramic rather than metal.
 

Another useful principle in starship design is the elimination of signal wires.
Signals are more efficiently transferred by optical fibers or waveguides.
Either alternative will carry orders of magnitude more bandwidth than wires.
Wires are just for distributing small amounts of power.
For massive transfers of power, superconductive channels are much more efficient,
or again waveguides (in certain low-loss modes.)
Wires are heavy, and may be systematically replaced to great advantage.
 

These rules of thumb for replacing metals with better materials are quite general.
They apply to all other aspects of spaceship design as well as to strictly drive design.
In fact, they are generally applicable to all other aspects of human construction,
and you may expect glasses, ceramics, and composites from solar furnaces to begin
replacing other materials as well as metals, such as wood, concrete, plastics and asphalt,
when the solar revolution tools up for large-scale production.
 

Whithertowards?

rev 980303