Best TWR for space

[federicoaa]

Hi everybody, here's a question I've been giving some thought. This is what I came out with:

TWR is important for lift off. Once in space, this is not important any more.

Well, TWR will define the maximum acceleration of your ship, which means how fast you can change your velocity.

This is relatively important if you don't want to waste minutes to change your orbit (I've heard this is also bad).

So, TWR is defined as F/mg (F is the thrust and m is the mass of the ship).

When in space, with the same force (thrust) you can accelerate a=F/m, which results a=TWR*g

Now, dV = a*t, so if you want to be able to change a given delta v in a given time, your TWR is defined as TWR=dV/(g*t).

You can use this to design your ship, as TWR is easily found using mods like mechjeb.

So for example, if you want to burn 500 m/s dV in 1 minute, you will need a TWR of 0.85

So here is the question, what do you think is the best TWR (or dV-time relationship) for a spaceship? More TWR will reduce the maximum dV you ship can have so there is a tradeoff.

Please also check if my reasoning/math is ok

[Murph]

Somewhere around 2.0 on the pad (which might be, for example, 1.5 at liftoff rising to 2.5 by the end of the stage, as fuel is burnt). Roughly the same from pad to 35,000m. 35km to orbit is really just enough to keep it going up and gaining speed, roughly a minimum of 1.0.

In space, it's no longer important, what matters changes to fuel efficiency. So, for long distance, a single LV-N for small and medium probes, etc. Larger long distance, I'd guess maybe 1 LV-N per 25-50t. Ignore the time that burns take, as long as they are not measured in hours.

[Taki117]

In space it becomes more a question of how long you want to wait to finish a burn. The lower your TWR the more dV you have overall, but the longer it will take to execute the same burn maneuver. For example: I have a ship with a decent TWR (let's say .5) and we want to execute a dV maneuver of 500m/s. This takes us about 45 sec (rough estimate, I really don't feel like doing math right now). To execute the same maneuver with a lower TWR (let's say .1) it will take 5 times longer to execute the same maneuver. Conversely if we have a higher TWR (1.0) it will take half as long.

Now, let's talk about fuel efficiency for a sec. Fuel efficiency is directly related to the specific impulse of the engine (regardless of weight). So, let's say I had a really efficient engine, but it's not very powerful (The LV-N). I plop one of these on a fuel tank and capsule I will have a high dV, but not necessarily a high TWR (they tend to be very heavy). Now, if we take the same ship, but put an engine with a lower specific impulse, but more power (the LV-45) we would get a Higher TWR, but a lower dV. This being said an ideal engine would have an infinitesimally small mass (because mass can't be zero) and an infinitely high specific impulse. This would result in a massive ammount of dV.

Now, at this point you might be wondering, why bother with the LV-N at all? It all comes down to fuel efficiency. Because of it's higher specific impulse it is one of the most fuel efficient engines in the game. The only issue is it's very heavy with a low maximum power. TWR is nothing more than the ratio of the amount of thrust you have to the amount of mass you are trying to move. The higher the TWR the easier it is to move a given mass. So, at this point the Rocket Equation comes into play. The rocket equation basically says for a given amount of fuel burned at a specific rate your velocivy will change by a certain amount. The reason we use the LV-N over the LV-45 is because of that equation. Could you get to Jool using just LV-45s? Sure, but you would need a metric butt-ton of fuel to do it because the LV-45s are not very efficient. By increasing the efficiency (at a TWR trade-off) we can go the same distance with much less fuel.

Tl;Dr: bigger is not always better, and TWR Only affects the length of your burn time.

[federicoaa]

Hi Taki117, I know that there is a trade-off between ISP and TWR, but what I wonder is what is a good trade-off value, and what experienced people do.

[Guest]

On real rockets and spacecraft, it's absolutely abysmal. It's true that low TWR transfer takes more dV than a high TWR one, due to longer burn time, but compared to what you get by using a light, small, efficient engine it's a negligible effect.

[Alchemist]

In space low TWR is not such a problem, but high delta-v maneuvers require some tricks and some extra delta-v (so it may eat half of what you won by putting more efficient engine). It's fully possible to perform interplanetary maneuvers with TWR=0.05 or even slightly less, but for ejections you'll need several cycles of raising apoapsis before finally going to escape trajectory (also mind ejection angle) and orbit insertion might require some breaking even before entering the SOI.

[Taki117]

Hi Taki117, I know that there is a trade-off between ISP and TWR, but what I wonder is what is a good trade-off value, and what experienced people do.

honestly, it's whatever you are comfortable with and how much fuel you want to bring with you to orbit. My long term Laythe mission had an average TWR of about .15 (across 4 ships) but in interplanetary space TWR is far less important. If your burn takes 10 minutes in orbit around a planet or moon, you are going ti wind up way off course, where as a burn farthest from a paletary body of the same length will have little overall effect on your ships position. I am on my phone at the moment, or I would have photographic proof.

[celem]

Im happy with very low twr interplanetary legs provided that theres an atmosphere on the destination. Aerobraking can save you several thousand braking dV and takes a fraction of the (real) time. Otherwise I try and stay above 0.5 in space, just so that im not stuck with the long burns. There are some situations where you only have 15 minutes inside the SOI and need to brake heavy, or you need to perform a swift radial burn for better trajectory. Being stuck with a single nuke pushing 50+ tonnes at those moments can really screw your gameplan.

[rakutenshi]

TWR can be very important in space in some situations, i.e. coming in for a Moho capture, you'll need to be able to burn fast enough to capture before the escape window (sometimes you could have under 2 hours to set everything up and get the burn done. Not really hard, but something to consider. )

[Frederf]

If you go through the math of TWR vs maneuver cost the optimal solution is generally a very small TWR. Spreading a maneuver over minutes or even hours takes sophisticated planning in many cases that few see as worth it. The difference between 30 kN and 1 kN engines in terms of dry mass on a ship of 5+MT is rather tiny if the Isp s are the same. For reasonable space maneuvers I keep TWR between 0.1 and 1.0.

For landers you want to tailor TWR for controllability relative to the surface gravity at destination. Too low will burn excess fuel in the transition as well as make timing the retro burn difficult. High TWR will make touchdown tricky.

[tavert]

For landers you want to tailor TWR for controllability relative to the surface gravity at destination. Too low will burn excess fuel in the transition as well as make timing the retro burn difficult. High TWR will make touchdown tricky.

Shameless plug, I put together some code to optimize this: http://forum.kerbalspaceprogram.com/threads/61659-Wolfram-Web-App-Optimal-Single-stage-Lander-Design-Tool

A big disadvantage of high TWR is the increased engine mass. Delta-V cost for landing approaches an asymptote as TWR increases so you get into diminishing returns.

[Tidus Klein]

ya I also was wonder this because Im actually getting ready to send up the engine module for my grand tour ship,(drops a hab and two rovers on every planet with an atmosphere + bop and minmus) and its heavy as da Butt, with the equivalent of 6 jumbos of fuel i had to use 20 LV-Ns just to get a TWR of .5 and a DV of 11,000 and I haven't even put up the habs/rovers yet!

[Alchemist]

ya I also was wonder this because Im actually getting ready to send up the engine module for my grand tour ship,(drops a hab and two rovers on every planet with an atmosphere + bop and minmus) and its heavy as da Butt, with the equivalent of 6 jumbos of fuel i had to use 20 LV-Ns just to get a TWR of .5 and a DV of 11,000 and I haven't even put up the habs/rovers yet!

1 NERVA for 2 Jumbos (+ some payload) is what I consider usable minimum. But it requires some special maneuver profile and I wouldn't even try LKO-Minmus transfer... well, it can be done, but requires a more complex flight profile than interplanetary ejection.

P.S. I once launched somethign with twice less TWR and extreme wobble to Duna, but that wasn't a pleasant experience. And it wasn't even my design.

Edited December 14, 2013 by Alchemist

[Tidus Klein]

Ya this is why I think the Nerva should have a thrust of 80 not 60.

That way it's still balanced but you can carry more load.

[EtherDragon]

One thing to consider is this:

Instead of relying only on one engine type for your interplanetary stage, use two different engines, with action groups to enable the engines you need for the particular kind of burn:

On the bottom of your 2.5m fuel stack:

Use a TVR-400L Stack Quad Adapter

2x LV-TvT30 Liquid Fuel Engine

and 2x LV-N Atomic Rocket Motor

You end up with an improved TWR, when you need it (All four engines running gives 580kn) and still have plenty of Isp in Vacuum when just running the Nukes (800 Isp, but only 120kn of thrust).

When I have this configuration, I use the full engine stack when performing an ejection with a low Periapsis, otherwise it's all Nukes all the time to maximize Isp.

[arq]

When I design ships, I naturally seem to end up with an acceleration of about 1.5m/s^2 in space (TWR ~0.15). I would very much discourage anything below about 0.08, as at this point your burns are so long that you may be losing significant dV to long burn times. Even with 0.15, you will likely need at least two burns to escape Kerbin (you will need almost 10mins, which is a little too long to do in one orbit).

[Ashery]

...Fuel efficiency is directly related to the specific impulse of the engine (regardless of weight)...

Except to the extent that the mass difference between engines might make the engine with the lower ISP actually have a better fuel to dV conversion rate than it's higher ISP, but heavier, counterpart. Eg, on a one-ton (sans engine) vessel, a 48-7S would have a higher effective efficiency than the LV-909, despite the latter having a pretty significantly higher ISP. The former would constitute a 1.1 ton vessel with a 350ISP (~3.2), while the latter would be a 1.5 ton vessel with a 390ISP (~2.6).

Heavier and more efficient engines might generate more thrust per unit of fuel, but the heavier engine also means that more thrust is required to accelerate the craft some fixed amount. This is primarily noticed on small landers, and is part of the reason I consider the 48-7S to be as overpowered as it is.