How far can you get on SWAGs - before you have to do some math?

[monophonic]

22 hours ago, GoSlash27 said:

I'm a weenie when it comes to this. I never build a rocket without doing the math first. On the one hand, it removes all the fun and suspense of wondering whether it's going to work (of course it is ), but on the other hand it allows me to complete missions I otherwise couldn't do. And for me (and I know I'm in the minority here)... I actually enjoy the engineering side of it.

 I don't use any mods to help me with this, just some spreadsheets I knocked together.

Best,

-Slashy

We should start a club.

Or maybe a support group. I made the first version of my dV spreadsheet before I launched my first rocket in the old (0.18.3) demo. 

[Wallygator]

Yes. KER and a DV map ^^^ is really all you need to have a good time. Playing with maths is quite fulfilling, but not a requirement to enjoying KSP to its upmost.

[panzer1b]

Well the only math i use often is dV calculation (i prefer doing it manually so i use the ISP*g*ln(full/empty) formula, but thats rather simple.  That and i compare it to dV charts to know where i can go with a given vessel, or reverse the formula to calculate my required fuel ratio if i am intentionally developing a vessel for a particular vessel.

 

Aside from this, there is not much i use math for except maybee getting the kerbin sea level TWR by comparing gravity *mass to the total thrust, easily done in head.  All else is purely trial/error.

Edited December 14, 2015 by panzer1b

[herbal space program]

I got as far as landing on and returning from all the major bodies without ever calculating anything. Heck, I even did my Jool 5 mission without explicitly calculating the dV of anything. After a while, you just get an intuitive idea of how far you can go with what configuration. Although I did mess up going to Moho and had to send a rescue mission. Once I started getting into pushing the limits of different designs though, I started doing math both to calculate dV and to figure out how to set up encounters for multiple gravity assists. It was all just Excel spreadsheets though. I never bothered with Mechjeb or KER, as the formulas involved are pretty simple.

[numerobis]

In 1.0.5 I don't really have time to design properly so all I ever do is guess.

That has something to do with why Valentina and four tourists are in a highly elliptical orbit with Pe of 300km and Ap of the Mun...

[MaxL_1023]

What I used to do was take the starting and ending TWR for each Kerbin launch stage, subtract 1, and average them. This gives you a linear approximation with a good over-design margin when you are spending time burning horizontally higher up.

As Jeb says:

"When in doubt, use more boosters. Put boosters on boosters. Use tweakscale to make bigger boosters. Then, use KJR so they don't fly apart. Put a command seat on the final stage, preferably also a booster."

[Kerbol Macrosystems]

5 hours ago, herbal space program said:

and to figure out how to set up encounters for multiple gravity assists.

What's the equation for this look like? I haven't been able to work it out. 

14 minutes ago, MaxL_1023 said:

"When in doubt, use more boosters. Put boosters on boosters. Use tweakscale to make bigger boosters. Then, use KJR so they don't fly apart. Put a command seat on the final stage, preferably also a booster."

Almost as memorable as "Ad Astra per asparagus."

[herbal space program]

38 minutes ago, MaxL_1023 said:

What I used to do was take the starting and ending TWR for each Kerbin launch stage, subtract 1, and average them. This gives you a linear approximation with a good over-design margin when you are spending time burning horizontally higher up.

 

That was my back-of-the-envelope method as well, and as you said it generally underestimates your actual dV, so it's pretty safe. If you have Excel though, plugging the rocket equation into it is pretty quick, and then your spreadsheet is faster than even doing that on a calculator.

[swjr-swis]

I have yet to open a spreadsheet for KSP in years of playing, and I think I will stubbornly keep refusing to do so.

Before I started to play KSP 'for realz', it was purely fun to watch what contraptions I could throw at the sky and how it would turn out, so I didn't care about numbers at all. Once I got a taste for career mode, I did eventually get KER to have more numbers readily on hand to help in getting it right the first time. That's as far as I expect to go.

It's my years-long experience with 'Spreadsheets in Space', aka EVE Online, that turned me completely off of ever again playing a game where more of my time goes into calculating how to play most efficiently, instead of actually enjoying the game itself. I refuse to fall into that pit again.

To me, as much as KSP adds in educational terms, it is still a game, to be played for entertainment foremost. Spreadsheets are for workstuff. What I can't calculate or estimate in my head, or read from a mod-added display, I take care of the fun/Kerbal way - with brute force (aka enough boosters and dV to reach the next star system and bring back a new planet).

[herbal space program]

It's not an equation exactly. What I do is make spreadsheets with columns containing the periodic return times of the target body to some particular spot, generally wherever my ship's orbit intersects the target's, then I make another column next to that listing the return times of my ship to the same spot on its current trajectory. I usually line things up so that  the target body is in the spot where I want to encounter it, then place a node for my ship at the same spot. From that position, I can calculate exactly when in the future, ad infinitum,  both my ship and the target body will return to that spot, expressed as multiples of the target's orbital period for the first column and multiples of my period offset by the initial approach time in the second. I then scan down the two columns to try to find two numbers that are particularly close. When I do, I adjust the orbital period of my ship at the encounter point either up or down so that it will eventually return to that spot at the same time as the target body. This requires solving a pretty simple equation, i.e. time to encounter = Pship*X+ Ti  = Ptarget*Y where Pship and Ptarget are the respective orbital periods, Ti is the initial approach time of the ship to the intersect, and X and Y are any integers small enough that you won't fall asleep in front of  your computer during the time warp. In practice, it amounts to subtracting the initial approach time from some integer multiple of the target body's orbital period, then dividing that number by whatever integer gives the value closest to your ship's current orbital period. If you start with a target return time that is close to one of your ship's return times, the new period should require only a relatively small correction. Unless the two orbits are in some kind of exact phase relationship, the longer you are willing to wait the less energy you will generally need to expend to eventually reach an encounter.

I'm currently using this method to calculate repeated encounters in an attempt to reach Moho capture from LKO for less than 1700 m/s total dV, a feat that I can now say with confidence is harder than anything else I've ever attempted in this game. This involves orchestrating a whole bunch of consecutive Moho encounters to gradually walk the ship through gravity assists down a ladder of resonant energy levels, eventually leading to an encounter that requires only a small burn to achieve capture. The above procedure is what I use for calculating initial encounters or encounters at different orbital intersects. For repeated encounters at the same spot, I've made tables of resonant orbital periods, e.g. 3:2, 3:4, 2:1 etc., arranged in order of decreasing energy. Using these, you can plan your encounter so that it exactly drops you to an orbit that will re-encounter the target body some number of orbits later. In this way, with careful enough planning, its possible to coast a really long way up or down with only tiny correction burns.

[Terwin]

12 hours ago, swjr-swis said:

I have yet to open a spreadsheet for KSP in years of playing, and I think I will stubbornly keep refusing to do so.

I don't know if it counts, but I used a spread sheet to note down the levels of the various USI resources on Minmus to determine where the best place to put a base would be...  (twice, as I did it in 1.0.4 as well as 1.0.5)

(easier to read than my hand-writing and easier to keep lined up than using notepad)

Will probably do something similar on Ike and Gilly once I have scanned them.  No calculations though, just noting down the numbers to identify which location is not missing essential resources.

 

Edited December 16, 2015 by Terwin

[katateochi]

If you enjoy calculating it yourself then do it, but I prefer to outsource all the maths to mods and other services.  All I do is subtract dV values (taken from the dV map) from the available dV of my craft/stage (as calculated by mechjeb).  
For planetary transfers I make use of these two tools; http://alexmoon.github.io/ksp/ & http://ksp.olex.biz/ (and/or use the protractor mod)
There are other calculation tools for various things listed here - http://wiki.kerbalspaceprogram.com/wiki/Calculation_tools 

 

[Kerbol Macrosystems]

14 hours ago, herbal space program said:

It's not an equation exactly...

Thank you!

[Geschosskopf]

On ‎12‎/‎13‎/‎2015‎ ‎4‎:‎03‎:‎40‎, JoeSchmuckatelli said:

At what point - if ever - do I need to start really doing some number crunching before a mission to have a chance of success?  Also - are there resources out there that can help an interested guy with a BA figure out how to run the numbers?

99.9% of the time, the only math I ever do is simple addition of the dV numbers shown by KER or MJ's dV readout, to make sure I have as much as shown on the dV maps.

The most complex math I ever do is figuring out the dark time of solar-powered satellites, which requires a bit of trig and occasionally some differential equations if they're in elliptical orbits.

[OhioBob]

Where is JoeSchmuckatelli by the way?  Hasn't replied since the opening post.

[JoeSchmuckatelli]

I've been reading this and other posts.  I wrote this before I knew about maneuver nodes.  I'd been doing it the hard way, but now I think I understand

[OhioBob]

58 minutes ago, JoeSchmuckatelli said:

I wrote this before I knew about maneuver nodes.  I'd been doing it the hard way, but now I think I understand

When I first got KSP, I played for several weeks before I learned about maneuver nodes.  Life definitely got much easier after that.

[Superfluous J]

If you just learned about Maneuver nodes, you probably don't know that you can drag them around. Most people (myself included) are shocked when they learn this.

[JoeSchmuckatelli]

I just learned about moving them last night and started playing with them this evening.  Much rejoicing.

[herbal space program]

15 hours ago, JoeSchmuckatelli said:

I just learned about moving them last night and started playing with them this evening.  Much rejoicing.

Soon you can join the ranks of those exasperated by their refusal to behave the way you want them to!  Seriously, though, you can get amazingly far just twiddling around with maneuver nodes. With a little experience and a table of Hohman transfer phase angles for bodies outside the Kerbolar system (or using the online calculators accessible through the KSP Wiki), you can go just about anywhere.

[OhioBob]

Just learning about maneuver nodes made a world of difference in the ease of game play.  If you haven't already done so, it's worth your time to read the Wiki article.  Although you can do much with maneuver nodes, after a while I started to become frustrated using them due to my inability to place them exactly where I wanted and my inability to dial in the exact amount of velocity change that I wanted.  I then tried a mod called Precise Node and loved it.  I now consider Precise Node an indispensable addition (along with KER and KAL).  I don't necessarily recommend that you get it immediately (I think it's good to learn the stock way of doing things before adding a bunch of mods), but it's something you might consider in the future.

[Caelib]

You can run the numbers or you can just guess/over-estimate.  If you don't use Kerbal Engineer, I would highly recommend it ... removes a lot of the guesswork around TWR and delta-V.

[InterCity]

SWAG has never failed me (not after a bit of tweaking, anyway  ). Then again, I never let swag do anything more important than a minimus landing.

[Inverurie Jones]

I just do everything by accident.

'Well, we overshot Minmus, boys, but at least we now know how to reach Duna...'