Computer-less Space Navigation

[Fred Brackin]

So let's say you've got this pulp science Space setting where rockets go "Whoooosh!" but "Computer" is a job for math-savvy human being and not a machine. Besides vintage science fiction the IW line of Lucifer-5 would qualify.

So how do you navigate and how long does it take and how difficult is it with nothing but mechanical calculating machines?

I have some idea of "How?" ( three dimensional triangulation based on stellar observations as just one possibility) but I'd be totally guessing about how difficult and how long the calculations took.

I can vaguely remember mechanical adding machines from when I was small but that's not much help. They only did simple math for one thing.

Ideas about e complexi9ty of the navigational math would be at least as much help as details about mechanical calculators.

so, anybody know anything?

[DemiBenson]

Simple questions like:

• * What angle to apply fixed thrust to get towards Planet X
  * Do I have enough fuel to get me there?
  * How fast is that other ship going?

Those are all easily doable with a slide rule and proper lookup tables. Every ship should have lookup tables for all the expected tasks. Call it 5-15 minutes for ballpark figures, of which most is using instruments to find values to slur rule and lookup. More time gives you more precise results, with more instrument use to find the Nth decimal or get a wider baseline/parallax, etc.

The value of analog mechanical calculators is they are built with the lookup tables and slide ruling internally and invisibly. Results will be quick (a few minutes) once measurements are input. It's instrument time that determines how long it takes, and how optimized you want to get.

If this were my game, I'd say Navigation -4 for 10 minutes, +0 for 1 hour, +4 for 8 hours. Totally winging it, mind you.

[vicky_molokh]

Calculation machines may or may not be necessary. Scroll down to Gordon Cooper for an example.

[johndallman]

There's an instrument that you need in addition to telescopes and calculating facilities, which is an inertial platform. That tells you how much acceleration you've actually done, and in which direction. The V-2 had the first of these to be used seriously on a rocket, so they're reasonable for Rocket & Raygun SF.

The calculations will be very laborious if you have to do them with nothing but yourself and a mechanical calculator. Those could add, subtract and multiply, but division was normally beyond them, let alone trigonometry and roots. The solution to this is the books of tables. They were produced for celestial navigation on Earth, and gave you logs, trig functions, and so on, plus shortcut formulae with pre-calculated constants to plug in depending on the time of year and the like.

For space, the books of tables will be bigger - ships on regular routes may save weight by only taking what they expect to need - and part of navigation help by radio if you get lost may well be sending you the constants you need. With the tables, the job will be small enough to make it feasible for ships on routine flights with skill-12 navigators. If it's too hard for them, someone will devise a better set of tables.

As an example, the first ICBMs didn't have microprocessors, but just steered to make the outputs from the inertial platform match a pre-calculated course, "flying a wire." That meant they had fixed targets and no flexibility. With four-bit microprocessors and a lot of pre-calculation, re-targeting to pre-calculated targets was possible. SLBMs made it much harder, because you didn't have a fixed launch site, unlike a land-based missile from a silo, but it was done, before microprocessors. Pre-calculation can do a lot for ballistics problems.

[Anthony]

Originally Posted by Fred Brackin

So how do you navigate and how long does it take and how difficult is it with nothing but mechanical calculating machines?

Depends on how much surplus delta-V you have, and how accurate your instruments are. Hand-calculating the correct numbers for a simple path such as a Hohmann transfer orbit is not particularly difficult, but even a very small measurement error means you miss the target by a lot, hitting Mars from Earth (without mid-course correction) is equivalent to getting a hole in one from about five miles away.

[Varyon]

Spaceships 7, page 24, has some guidelines under the No Computers design switch.

[GoldenMonkey]

What is the scale here -- Earth to LEO/Moon? Inner planets? Interstellar? Intergalactic?

Is fuel a consideration, like do they have reactionless drives or something? Or does each launch need to be an optimized/measured-to-the-milliliter affair?

The answer to these will impact the margin of error acceptable for navigating.

[malloyd]

Originally Posted by Anthony

Depends on how much surplus delta-V you have, and how accurate your instruments are. Hand-calculating the correct numbers for a simple path such as a Hohmann transfer orbit is not particularly difficult.

This is more or less true for everything. The basic math for interplanetary or interstellar navigation is not particularly hard, the large scale number crunching demand comes in because we don't have a generalized solution for the inverse square interaction of 3 bodies, let alone n of them. This means perturbations for all the objects in the universe need to be computed numerically - i.e. by lots and lots of repeated iterations.

Anybody with a suitable undergraduate degree can probably calculate everything you'd need to for almost any interplanetary or interstellar mission by hand in an afternoon if you only care about the end points (i.e. you aren't trying to do a flyby of multiple planets with no course corrections) and can live with an answer that is off by a percent or two. Mechanical computers that will give you at least that good in a few minutes are not particularly difficult - it's actually much the same problem as firing a shell from one ship to another, and mechanical computers for that are on the drawing board by the beginning of TL6, and shipboard by the middle of it.

If you have lots of surplus delta-V, and you are going to burn some of it near the destination anyway, that's more than good enough. If you are trying for a single burn solution you will probably miss. If you are horribly fuel limited, like, say, our actual spacecraft, allowing this much slop is going to mean missions that you might be (barely) able to do with better computations, particularly the ones that really need a gravity assist to make work at all, are too risky to attempt, but the ones you can do ought to be fine.

[Anthony]

Note that there are various rocket flight simulators that will allow you to try this by hand, so you can fairly easily see that if you have plenty of delta-V, it's really not that hard to get there.

[jeff_wilson]

Charles Stross' Laundry stories and THE LAUNDRY FILES rpg based on them include descriptions of numerous historical mechanical calculating devices, there employed because they are fast enough to get the job done, but too slow to do anything secretly if subverted.