It's worth to have TWR in space?

[O Nerd]

Well, i have a question.

Let me propose 2 situations (Both in LKO).

Number 1: A craft has an TWR of 0.11 in space and a Delta-v of 9000+.

Number 2: A craft has an TWR of 5.00 in space and a Delta-v of 5000+.

Which craft will go further? Number one or two? At real, TWR is important in space? I know, it will take longer to do the manuever, but having a lower TWR affects the overall mission?

Edited November 12, 2013 by O Nerd

[Johnno]

Delta-v determines your velocity change, TWR just determines how fast you get to that velocity. So craft 1 will go further, it'll just spend a longer time burning its fuel.

[O Nerd]

Thanks Johnno.

[RSF77]

If I'm not mistaken TWR is supposed to be taken into account when figuring delta v, so delta v would be the total change in velocity that you have left with your current ship regardless. So number 1 would go further, but probably much slower and have very long burns.

I'm not really sure what else it affects in game aside from craft stability maybe, but if I'm not being conservative I usually place extra/bigger engines just to cut down on the burn time. With some bigger ships and more massive fuel tankers nuclear engines can start racking up serious burn times, especially if your are going somewhere other than the mun or minmus.

[Tokay Gris]

Number 2: A craft has an TWR of 5.00 in space and a Delta-v of 5000+.

Just out of curiosity...

You can build a craft with a somewhat useful payload AND a TWR of 5 and 5000 delta-V?

Delta-V of 5 in orbit around Kerbin, I assume.

A TWR of 5 in orbit around Gilly could probably be achieved with an ion engine....

;-)

[Tokay Gris]

Ok, so now a serious answer.

TWR is - as the names says: Thrust to weight ratio. And actually is measured to the weight the vessel has on the surface (namely sea level, I assume) on the body it currently orbits.

Much more important would be a TMR (Thrust to Mass ratio), since that is independent of the orbited body.

It just isn't that important in the game.

And - but that was answered - TWR doesn't matter in space. Actually.

If TWR is low, you just need longer to achieve a certain change in velocity.

My "heavies" sometimes have incredible burn-times. Ejection burn of the famous "Colin Campbell" was about 45 minutes, IIRC.

[vetrox]

-snip-

Never mind

[Franklin]

Keep in mind while TWR is far less important than delta-v in space than during takeoff, and Ship 1 will go farther, you're going to want a healthy TWR for the sake of your sanity during maneuvers. Unless you're using a cheaty mod like MechJeb or something to handle everything for you, you're gonna be waiting a long-ass time to get anything done if you've sacrificed all your TWR for delta-v.

Go ahead and build yourself a deep-space ion craft and get yourself into orbit around Eeloo and back. I guarantee you'll give up when you have to sit through your third multi-day maneuver.

[Tokay Gris]

Go ahead and build yourself a deep-space ion craft and get yourself into orbit around Eeloo and back. I guarantee you'll give up when you have to sit through your third multi-day maneuver.

That sounds like you actually did this.....

;-)

[Franklin]

That sounds like you actually did this.....

;-)

I let the game sit during a long burn for two days before I said **** it and terminated the mission. I'm sure that probe's still making its way there.

[Seret]

Anything with a burn longer than about 5 minutes needs more engines as far as I'm concerned. 10 minutes and it definitely does. The loss of ÃŽâ€v from sticking on more LV-Ns is worth it. I don't have hours and hours to play the game, I want to get things done in a reasonable time.

[Eadrom]

For my big interplanetary ships, or any long burns in general, I am generally comfortable with an acceleration of anything over 3m/s^2. Keeps burns at a reasonable time and allows for precise burns (compared to something with like 30m/s^2 acceleration).

[Vanamonde]

If I'm not mistaken TWR is supposed to be taken into account when figuring delta v

It seems like it should be, but if you look at the math again, you'll find that thrust is not a term in the equation.

[andrewas]

Because thrust has no direct impact on dV - impulse is whats important, which is thrust over time.

That said, a high TWR craft can make more efficient use of its dV since it can apply all the impulse at the optimum time for the maneuver - perigee for breaking orbit, apogee for plane change or de-orbiting, - where a low TWR craft has to either waste impulse or waste time doing endless small burns each time it crosses that optimum point.

[arq]

TWR does have a small effect on transfer range because of the Oberth effect. It is a 'force multiplier' for burning close to periapsis (and some other places, but relating to TWR this is the most relevant). So a craft with a TWR of 2 can make better use of the Oberth effect than a craft with a TWR of 1, which means that if both have 5000m/s of dV then the higher TWR one can go slightly further.

This effect can be somewhat mitigated by 'periapsis kicks,' but Jool is 1900m/s from LKO and by the time you burn half of that you're on escape which obviously prevents any additional kicks, meaning you must finish the burn no matter how long it takes. Still, excepting extreme cases the effect of low TWR is very small (tens of m/s, likely) and the bigger consideration is your sanity.

But high TWRs do make landing and ascending from surfaces more efficient. So remember that if you plan on landing somewhere, dV isn't the only concern.

[arq]

I just did an experiment. I placed a ship in 75x75km LKO with 3837m/s of dV in the tank (57s at max thrust). I waited for MM:SS MET, put the nose 10deg below the horizon, locked SAS, and burned at XX% thrust until empty. I picked the times and angles so that I would be exiting Kerbin almost exactly retrograde (and verified this after the burn). I then recorded the Kerbol periapsis after the burn. Here are the results of my two trials:

17:00 start @ 20% power - Kerbol peri 1.363Gm

18:00 start @ 100% power - Kerbol peri 1.194Gm

Placing a maneuver node shortly after the fast burn, it required about 180m/s to slow down to match the peri of the slow burn. So a TWR difference of a factor of 5 (a huge difference) made a difference of ~4.5% on the net dV. This experiment was imprecise, but should give ballpark results showing that TWR does have a small effect on net dV.

[Dunbaratu]

Ok, so now a serious answer.

TWR is - as the names says: Thrust to weight ratio. And actually is measured to the weight the vessel has on the surface (namely sea level, I assume) on the body it currently orbits.

Much more important would be a TMR (Thrust to Mass ratio), since that is independent of the orbited body.

It just isn't that important in the game.

And - but that was answered - TWR doesn't matter in space. Actually.

If TWR is low, you just need longer to achieve a certain change in velocity.

My "heavies" sometimes have incredible burn-times. Ejection burn of the famous "Colin Campbell" was about 45 minutes, IIRC.

Boredom isn't the only problem that a low TWR can cause. It *does* in fact matter when you have to make a burn to get captured in orbit around your target planet. If it takes you 30 minutes to get the needed delta-V to get captured, but your flyby will zoom past in only 20 minutes, you can't get captured.

[Nao]

I just did an experiment. I placed a ship in 75x75km LKO with 3837m/s of dV in the tank (57s at max thrust). I waited for MM:SS MET, put the nose 10deg below the horizon, locked SAS, and burned at XX% thrust until empty. I picked the times and angles so that I would be exiting Kerbin almost exactly retrograde (and verified this after the burn). I then recorded the Kerbol periapsis after the burn. Here are the results of my two trials:

17:00 start @ 20% power - Kerbol peri 1.363Gm

18:00 start @ 100% power - Kerbol peri 1.194Gm

Placing a maneuver node shortly after the fast burn, it required about 180m/s to slow down to match the peri of the slow burn. So a TWR difference of a factor of 5 (a huge difference) made a difference of ~4.5% on the net dV. This experiment was imprecise, but should give ballpark results showing that TWR does have a small effect on net dV.

Unfortunately This only means that you are loosing dV to steering looses more when burning slowly (not burning at the velocity vector). And this is more of a change in steering error than difference in propulsion.

More precise test would be with mechjeb locked to "prograde" and that would show almost no difference in speed change (periapsis). Unless the low TWR would be really low, and there would be losses to Oberth effect due to burning fuel at higher altitudes. And that can be mitigated by dividing the low TWR burn into several small burns close to Periapsis.

edit: In other words, if you want to get the most from your fuel, never burn at the maneuver node, always pro/retro-grade. (just start around half to 1/3 of burn time before the node).

Edited November 13, 2013 by Nao

[Sirine]

Anything with a burn longer than about 5 minutes needs more engines as far as I'm concerned. 10 minutes and it definitely does. The loss of ÃŽâ€v from sticking on more LV-Ns is worth it. I don't have hours and hours to play the game, I want to get things done in a reasonable time.

Agree. I don't have all days.

[Frederf]

Using a triangle approximation you can estimate the lost deltaV due to finite thrust.

By spreading out an impulse burn at 70km LKO over 2 minutes you may end up say slower than an impulse at the beginning and end of the resulting orbit. This might be 100m/s with a peak speed of 2300m/s. On the average you are traveling at 2250 instead of 2300. This translates into a 23/22.5th increase in deltaV requirement which is about 2%. Those are made up but plausible numbers. Usually packing significant thrust incurs much more than a 4% penalty.

[Umlüx]

long interplanetary burns.....

this is the only time i use MechJebs autopilot. speed the game up to 4x physical acceleration and go watch a movie

[Nao]

Using a triangle approximation you can estimate the lost deltaV due to finite thrust.

By spreading out an impulse burn at 70km LKO over 2 minutes you may end up say slower than an impulse at the beginning and end of the resulting orbit. This might be 100m/s with a peak speed of 2300m/s. On the average you are traveling at 2250 instead of 2300. This translates into a 23/22.5th increase in deltaV requirement which is about 2%. Those are made up but plausible numbers. Usually packing significant thrust incurs much more than a 4% penalty.

I don't think the penalties are that severe in KSP, for example: 2 minutes burn 0,26 TWR vs 13 second burn 2,3 TWR, both 320m/s from 70km orbit the dV difference is only ~0,3m/s that's only 0,1% loss.

In the case of bigger burn: 1561m/s with TWR 1,05 and 9,45 respectively (that's a burn to Moho from 70km Kerbin orbit) the difference is only around 0,68% (counted by energy change).

With higher dV burns will be penalized more but over the 2minutes it shoudn't be more than 1% in most cases.

[Frederf]

And so it turns out for medium thrust bringing extra TWR isn't justified mathematically. I recently did a ion engine transit to Duna and my burn times (if I was to try to do it all at once centered on one maneuver node) would be 30 minutes or more. I think only on that scale do you approach the tipping point between TWR saving dV and mass reduction from low thrust improving dV. I actually wanted to calculated the mathematical optimum setup to find where low thrust started to hurt more than help but I gave up.