Steve Jackson Games Forums

Steve Jackson Games Forums (http://forums.sjgames.com/index.php)
-   GURPS (http://forums.sjgames.com/forumdisplay.php?f=13)
-   -   [Blog] n-Body Politics (http://forums.sjgames.com/showthread.php?t=153002)

Humabout 12-05-2017 12:55 PM

Re: [Blog] n-Body Politics
 
Quote:

Originally Posted by kreios (Post 2140765)
Actually, this is wrong. The most efficient transfer is (depending on the ratio between the two orbits) is either a Hohmann or a bi-elliptic transfer. Both assume instant velocity changes, meaning you can never actually achieve them. However, the higher your acceleration, the closer you can get (and therefore need less dV). Assuming you're able to restart engines (bringing us back to that topic, but which means you can actually to those required two burns), a high-thrust engine is always more efficient than a low-thrust one to change orbits.

At the extreme end, we can look at the real-life example of the SMART-1 mission, which expended ~3.9km/s dV from a geostationary transfer orbit to a low lunar orbit. However, an optimal trajectory (such as produced by a high-thrust system) would only need ~1.5km/s dV for such a transfer. Of course, the ion engine was far more efficient in terms of propulsion (since its Isp is more than four times a chemical rocket's) - but if they would've had a high-thrust system with those numbers available, they would've used it.

Actually, this is correct in the context in which I was speaking.

1) Eric and I are addressing cost efficiency, not dv efficiency - which itself doesn't necessarily translate into mass efficiency or cost efficiency. And what will ultimately determine which engines get used in some hypothetical futurist setting is the combination of cost and demand.

2) Hohhman, bi-elliptic (that ration is 11:1, btw), and other high-thrust transfers are the least time-efficient transfers available. This makes them highly undesirable for most applications of space flight, both because no one wants to sit on a spaceship for a decade while they drift toward Saturn and because no company wants to pay a crew a decade of wages to drift toward Saturn once.

3) The fastest transfers available use constant thrust to attain high velocities over the course of the transfer. This necessarily increases dv, which in turn, increases the fuel costs. It, however, saves time, and thus allows more trips (meaning more revenue) and fewer paid crew hours (meaning lower expenses). This is desirable when maximizing profit, and thus, monetary efficiency.

4) When dealing with brachistochrone transfers, the higher the craft's acceleration, the higher the maximum velocity attained, and thus the more dv required to execute it. Thus, a fusion torch can pull 0.5G, it would need approximately 800 mps - meaning 13 fuel tank systems - to transfer from Earth to Mars, and you would have to refuel there. But, even a regular fusion rocket could pull 0.005G, need only 100 mps and make the trip there and back again on less than 2 fuel tank systems. They use the same type of fuel, so it's pretty plain to see that a fusion rocket is more cost efficient than a fusion torch. Or in other words, the lower-thrust drive is more cost efficient than the higher-thrust drive.

5) High-thrust rocket engines have existed since 1926. They had a high-thrust system available for the SMART-1 mission and chose to use a high-efficiency, low-thrust ion engine instead. It had a higher dv because it went faster, but it had a far lower fuel-mass percentage (3.33%) specifically because the ion engine was more efficient.

kreios 12-05-2017 01:29 PM

Re: [Blog] n-Body Politics
 
Quote:

Originally Posted by Humabout (Post 2140780)
3) The fastest transfers available use constant thrust to attain high velocities over the course of the transfer. This necessarily increases dv, which in turn, increases the fuel costs. It, however, saves time, and thus allows more trips (meaning more revenue) and fewer paid crew hours (meaning lower expenses). This is desirable when maximizing profit, and thus, monetary efficiency.

4) When dealing with brachistochrone transfers, the higher the craft's acceleration, the higher the maximum velocity attained, and thus the more dv required to execute it. Thus, a fusion torch can pull 0.5G, it would need approximately 800 mps - meaning 13 fuel tank systems - to transfer from Earth to Mars, and you would have to refuel there. But, even a regular fusion rocket could pull 0.005G, need only 100 mps and make the trip there and back again on less than 2 fuel tank systems. They use the same type of fuel, so it's pretty plain to see that a fusion rocket is more cost efficient than a fusion torch. Or in other words, the lower-thrust drive is more cost efficient than the higher-thrust drive.

Ah. I believe this is where we miscommunicated. And, I should add, I actually believe one of your assumptions don't necessarily hold. But for this, let's look at some numbers.

Specifically, ignoring orbits, your goal is to move to a target at 1 AU. Your available engines are a fusion rocket (0.005G, v_exh 10290km/s or 350mps/525km/s dV per tank) or a hydrogen-fuelled fusion torch (0.5G, v_exh 1323km/s or 45mps/67.5km/s dV per tank). I'm also ignoring acceleration changes due to fuel expenditure.

A brachistochrone transfer with the fusion rocket will take you just over 40 days (20 each spent accelerating and decelerating). In total, you spend about 175km/s dV for this (1/3rd of a tank).

A brachistochrone transfer with the fusion torch will take 4 days, spending an astronomical ~1750km/s of dV (surprisingly, that's just ~15 tanks). I believe that's where your numbers came from.

However, you can also use the fusion torch to accelerate to 87.5km/s, coast for some time, then decelerate. Total dV spent is 175km/s too (~2.5 tanks), but you spend slightly less than five hours each to accelerate/decelerate, and ten days coasting. In total, you spent the same dV, but cut your transfer time by 75%. In fact, you can achieve the same transit time as with the fusion rocket's constant acceleration by spending just 45km/s dV (2/3rds of a tank)!


Quote:

5) High-thrust rocket engines have existed since 1926. They had a high-thrust system available for the SMART-1 mission and chose to use a high-efficiency, low-thrust ion engine instead. It had a higher dv because it went faster, but it had a far lower fuel-mass percentage (3.33%) specifically because the ion engine was more efficient.
By "these numbers", I meant a high-thrust engine with similar dV. I should have made that clearer; apologies.

kreios 12-11-2017 10:06 AM

Re: [Blog] n-Body Politics
 
Quote:

Originally Posted by The economics of interface travel
We now know what our space and interface propulsion looks like: Fusion torches, probably using water. What remains to be defined is the impact our technology has on the economy - how much does it cost to transport cargo or passengers?

Continue reading...

kreios 12-21-2017 07:32 AM

Re: [Blog] n-Body Politics
 
Quote:

Originally Posted by The economics of interstellar travel
Getting everything into orbit is only the first part in a long journey. For the next one, we’ll have to look at interstellar transport.

Continue reading...

johndallman 12-21-2017 07:41 AM

Re: [Blog] n-Body Politics
 
Quote:

To simplify math, all FTL jumps will deposit the spacecraft into a stable orbit.
It seems better to change that to FTL jumps being able to deposit the spacecraft into a stable orbit, if the dynamics of the target are known with sufficient accuracy. Otherwise you get weirdness with exploratory jumps, arguments over the definition of "stable orbit" and so on.

vicky_molokh 12-21-2017 09:17 AM

Re: [Blog] n-Body Politics
 
Quote:

Originally Posted by ericthered (Post 2144530)
Huh, some interesting decisions here:

I'm not sure why you're using a fusion torch instead of a fusion rocket. the rocket has 4x the delta-v. Do you need the speed?

I'm surprised to find out that using water is cheaper than hydrogen, but the math holds up. Even counting transport costs, water is half the dv for 1/50th of the price, and you only decrease cargo space by 25%, so it makes sense. And you probably don't even need pure water: melted ice from an icy moon like Europa will work.

I'm sad to see you drop the FTL "emerging at rest" mechanism. I was having a lot of fun with it.

A freighter really shouldn't have 15% of its mass dedicated to armor, but spaceships always buts it in, so I can see not rocking the boat.

This has been branched off into a separate thread for the cargo/armour discussion.

kreios 12-21-2017 01:53 PM

Re: [Blog] n-Body Politics
 
I'll answer to the relevant points here.

Quote:

Originally Posted by ericthered (Post 2144530)
I'm not sure why you're using a fusion torch instead of a fusion rocket. the rocket has 4x the delta-v. Do you need the speed?

The acceleration, yes. If you only look at the departure burn, the Fusion Rocket needs more than eight hours thrust to produce the necessary dV (which means even more time spent on waiting for periapsis). The torch needs five minutes of thrust.

Quote:

I'm surprised to find out that using water is cheaper than hydrogen, but the math holds up. Even counting transport costs, water is half the dv for 1/50th of the price, and you only decrease cargo space by 25%, so it makes sense. And you probably don't even need pure water: melted ice from an icy moon like Europa will work.
So was I, but - assuming waystations - hydrogen is far too expensive. If I remember correctly, there's a break-even point somewhere between five and ten jumps, so maybe for exploration craft - but on the other hand, they'd have to deal with the hassle of cryogenic hydrogen.

Quote:

I'm sad to see you drop the FTL "emerging at rest" mechanism. I was having a lot of fun with it.
Hm... maybe I should spend a post looking over that mechanism and computing the dV. It might be worth keeping.

ericthered 12-21-2017 03:08 PM

Re: [Blog] n-Body Politics
 
Quote:

Originally Posted by kreios (Post 2144629)
The acceleration, yes. If you only look at the departure burn, the Fusion Rocket needs more than eight hours thrust to produce the necessary dV (which means even more time spent on waiting for periapsis). The torch needs five minutes of thrust.

hah, every time I did the math I was assuming three engines: that's how many you need to achieve orbit in 8 hours.

Of course now when I look at costs I realize that engines are among the most expensive things in the book, and two systems, the engine and the stardrive, cost around 70% of the ship already.

Quote:

Hm... maybe I should spend a post looking over that mechanism and computing the dV. It might be worth keeping.
I've done some work. The amount of dv for a given orbit scales with the fourth root of mass and the square root of gravitational influence.

The equations got really complex and than all cancelled out. It was stunning. dv where a planet's influence is g and the mass is M is

dv = (GM)^.25 * g^.5 * .001

Where G is the gravitational constant.

doctorevilbrain 12-21-2017 04:09 PM

Re: [Blog] n-Body Politics
 
When I clicked on the continue reading link, it said "significant costs of surging". What's surging?

kreios 12-22-2017 06:30 AM

Re: [Blog] n-Body Politics
 
Quote:

Originally Posted by ericthered (Post 2144651)
Of course now when I look at costs I realize that engines are among the most expensive things in the book, and two systems, the engine and the stardrive, cost around 70% of the ship already.

Extremely so. If you look at the Cargolifter from last post, which costs $8.75M, you'll note that I had designed a ram-rocket variant. Only exchanging the engines means increasing the cost from $8.75M to $60M. Hugely expensive.



Quote:

I've done some work. The amount of dv for a given orbit scales with the fourth root of mass and the square root of gravitational influence.

The equations got really complex and than all cancelled out. It was stunning. dv where a planet's influence is g and the mass is M is

dv = (GM)^.25 * g^.5 * .001

Where G is the gravitational constant.
That really is nice. (I tried reconstructing that formula; I'll probably have to repeat it and expand it in a blog entry, since I got other numbers). One of the issues, of course, is that the orbital math becomes more complicated when considering faster transfer burns.


Quote:

Originally Posted by doctorevilbrain (Post 2144665)
When I clicked on the continue reading link, it said "significant costs of surging". What's surging?

I have updated this passage to "[...] and the arguably significant costs of either coordinating massive numbers of shuttle flights (surging) or in-orbit storage of cargo. Remember - you only have eight hours to load and unload your whole cargo, every hour you waste costs you more than ten thousand dollars."
This hopefully makes it clearer.


All times are GMT -6. The time now is 02:32 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.