TWR - a general question.

[BoilingOil]

I've seen some references to keeping TWR below certain levels. Some mention trying to keep it as low as 1.15 :$

If you lower the thrust of an engine, aren't you essentially also slowing down the rate at which you get rid of that bulky, heavy load of fuel that you're hauling around? Aren't you essentially making it harder to get away?

I understand that there is a point at which drag is affecting your ship and its ascent negatively. It also heats up your ship, to the point of stuff exploding, which you would want to avoid. That much I get.

But as long as you steer clear of all those kinds of disasters, why would one not use as much thrust as available?

Edited November 5, 2015 by BoilingOil

[Dfthu]

I keep my engine at max at takeoff and throttle down when I have too.

[Vectura]

Well, the faster you go the more drag you have, and it increases much more when going supersonic. If you then need to throttle down to stay subsonic, then you are using an engine that is too powerful, so you could save weight (and maybe cost) by using a smaller/weaker engine.

[Oafman]

I wouldn't launch with a TWR as low as 1.15. It gets hard to keep your ship pointing upwards when it is accelerating so slowly.

I don't know what advantage you might get from a low TWR, beyond the fact that you can generally achieve it with more efficient engines than if you were going for a higher thrust. Perhaps that is what these references you've seen are alluding to?

[Alshain]

1.15 is a little low, but not bad. I try to keep mine around 1.25. It's not a hard rule though. If you need to cram in a little more fuel, it may crawl off the launch pad but as long as it is above 1.0 and you don't turn too quickly you will be fine. I don't recommend that low for a non-gimbal engine though.

Keep in mind the Saturn V Skylab had a TWR of 1.1. I think the Discovery Hubble was about the same.

Edited November 5, 2015 by Alshain

[Snark]

There are two penalties you pay for high TWR.

The first one applies mainly for takeoff from Kerbin: if you go too fast too low, you're wasting lots of fuel fighting aerodynamic drag. High TWR ships have problems with that. You can throttle down for that, of course, but then you're not getting the benefit of your high-TWR engines.

The second: if you have a TWR higher than you need to, it means you've got more engine than you need, and engines are heavy.

Of course, if your TWR is too low for takeoff from a planet, then your gravity losses build up a lot, which is bad.

That's why TWR is a balancing act. For takeoff from Kerbin, I like to keep my launchpad TWR at around 1.5 (which is a bit on the high side). Any higher than that and I'm fighting drag.

For launches where aero drag isn't so much of an issue (e.g. takeoff from anywhere but Kerbin or Eve), then higher TWR is fine (and usually easy to obtain with small engines, since most places have much lower gravity than Kerbin).

Some mention trying to keep it as low as 1.15

That seems inefficiently low to me, at least for takeoff from a planet's surface. I'd prefer not much less than 1.3.

If you lower the thrust of an engine, aren't you essentially also slowing down the rate at which you get rid of that bulky, heavy load of fuel that you're hauling around?

Actually, that's not the problem. The math works out the same. Imagine I have two ships floating side by side in zero-gravity space at zero velocity. Both have the same Isp, same dry mass, same fuel load, but one has an engine that can give 1000 kN thrust and the other's engine gives just 1 kN thrust.

Both ships will have exactly the same velocity profile in terms of "how fast are you going after burning X kilograms of fuel". That depends only on Isp and ship mass, and doesn't depend on engine thrust at all. (Note that the Tsiolkovsky rocket equation, which is what governs this, only cares about Isp and mass ratio-- there's no term in there for thrust.) The low-thrust ship will take longer time to reach speed, but it will still reach the same speed.

The reason that a too-low TWR causes problems on takeoff is different: it's because you take longer to build up velocity and therefore spend more time fighting gravity, thus higher gravity losses.

If you're building a ship that's purely orbital and never has to lift itself off a planetary surface, in general you want a very low TWR, mainly because gravity loss isn't an issue, and low TWR means you're not wasting mass on big heavy engines.

If you're building a ship that has to take off from a planetary surface, then picking the "right" TWR involves a balancing act between avoiding gravity losses (which wants a higher TWR), versus causing too much drag and/or spending too much mass on engines (which wants a lower TWR).

Edited November 5, 2015 by Snark

[Empiro]

I remember seeing an analysis here that showed that for the lowest $ per ton into orbit, you want somewhat low TWR (something like 1.2-1.3 I think). The reasoning being that engines are expensive and heavy, while fuel and fuel tanks are cheap.

[Alshain]

I remember seeing an analysis here that showed that for the lowest $ per ton into orbit, you want somewhat low TWR (something like 1.2-1.3 I think). The reasoning being that engines are expensive and heavy, while fuel and fuel tanks are cheap.

That is correct, by adding more fuel weight you will be consuming more fuel to get off the ground but you end up with a net gain because in most cases it doesn't consume more extra fuel than you added, even if you factor in the added drag from being lower longer (assumes an aerodynamic rocket). It's also easier to fly if you aren't going TOO fast. To maximize the engine potential, you want a fairly low TWR, but not so low that it is hard to fly.

[BoilingOil]

There are two penalties you pay for high TWR.

Very informative, Snark. Thank you so much. Pity I just gave you rep for another post. Otherwise I would have upped you one for this here, too! +1

[Streetwind]

Pretty much this, yeah. Overburdening the launch stage is "free", so to speak.

What I mean is this: If you add fuel to your craft, it will affect the performance of the stage it is added to, as well as the performance of all stages below that one. However, it will not affect the performance of stages above that one in any way. As such, the launch stage is a great place to add extra fuel, because there is no stage below it. All other stages are above. So if you have TWR leftover in your launch stage, toss in more fuel for a dV boost that doesn't change anything whatseoever about the rest of the vehicle.

Of course, if you want to save money, it can sometimes be more efficient to bump up one of the upper stages, where only a small amount of fuel adds a whole lot of dV. But you always need to weigh that against the chain reaction of performance losses (both TWR and dV) that will trickle down through all the stages below and eat up again some of what which you have so cheaply gained.

And finally, launching with a high TWR often reduces the dV cost to Kerbin orbit. So packing less fuel, and thus less dV, may also be a valid moneysaving strategy, unless you end up overpaying for more engines than you need. It depends a lot on the vessel.

Just, always remember: If you must overburden a stage in your rocket, always do it with the launch stage.

[Invader Jim]

There are two penalties you pay for high TWR.

The first one applies mainly for takeoff from Kerbin: if you go too fast too low, you're wasting lots of fuel fighting aerodynamic drag. High TWR ships have problems with that. You can throttle down for that, of course, but then you're not getting the benefit of your high-TWR engines.

The second: if you have a TWR higher than you need to, it means you've got more engine than you need, and engines are heavy.

Does this apply to SRBs? I frequently reduce the thrust on the SRBs when designing a rocket so my TWR range (according to Kerbal Engineer mod) is reasonable both at the low and high ends. Is it better to use fewer SRBs at 100% than more at a lower %?

[BoilingOil]

What I mean is this: If you add fuel to your craft, it will affect the performance of the stage it is added to, as well as the performance of all stages below that one. However, it will not affect the performance of stages above that one in any way. As such, the launch stage is a great place to add extra fuel, because there is no stage below it. All other stages are above. So if you have TWR leftover in your launch stage, toss in more fuel for a dV boost that doesn't change anything whatseoever about the rest of the vehicle.

Oh my, that's a great point!

Of course, if you want to save money, it can sometimes be more efficient to bump up one of the upper stages, where only a small amount of fuel adds a whole lot of dV. But you always need to weigh that against the chain reaction of performance losses (both TWR and dV) that will trickle down through all the stages below and eat up again some of what which you have so cheaply gained.

Call me 'Scrooge McDuck', I'm so cheap! If I need 120 thrust, and my only choices are: a dozen engines of 10 thrust at 100 coins each, or a single 180 thrust engine at 900 coins, I'll go for the latter option, even if that means hauling a big-ass, overpowered mofo of an engine. And then it needs to be lifted up, so a big-ass tank is required. Advantage of the big engine/tank is, I don't have to figure out how to have a dozen engines work in tandem!

And finally, launching with a high TWR often reduces the dV cost to Kerbin orbit. So packing less fuel, and thus less dV, may also be a valid moneysaving strategy, unless you end up overpaying for more engines than you need. It depends a lot on the vessel.

Just, always remember: If you must overburden a stage in your rocket, always do it with the launch stage.

You brought the point home nicely, thank you! I have some stuff to consider now...

[Red Iron Crown]

Does this apply to SRBs? I frequently reduce the thrust on the SRBs when designing a rocket so my TWR range (according to Kerbal Engineer mod) is reasonable both at the low and high ends. Is it better to use fewer SRBs at 100% than more at a lower %?

Those aren't directly comparable, as more SRBs at a lower % will have more propellant to expend (i.e. will add more delta-V). The better comparison is 100% SRBs vs the same number of SRBs set to a lower percentage; generally speaking it is better to leave them at 100% (you are paying the mass for those engines either way) unless it makes your initial TWR so high that drag/overheating become significant.

[CrashTestDanny]

Looks like your question has been really well covered, but I'm going to put in my half-cent's worth anyhow...

High TWR can be fun. It's also very easy. You stick a big-ass fuel tank on a big-ass engine and get 3500 delta-V and 2.5 TWR out of it, and you're [almost] guaranteed to go to space today. I have been guilty of this. A lot. Like really... I've launched from Kerbin with 4.5 TWR, and vanilla KSP can be pretty forgiving with most parts (maybe more than it should), so if you like BIG (and I do!), load her up! But then the aerospace engineer in me feels compelled to echo what everyone else has said - there is always a price to be paid, and here's what it is. If your TWR is too high, then your cost is not that you paid too much for your engine or the fuel or the aerodynamic drag. Your real cost (and I don't think anyone has said this yet) is that you could have hauled up a bigger payload (that's the economist in me - he's always talking about "opportunity cost":rolleyes:).

So, again as many people have said, your best bet is to set your TWR to something reasonable that will balance aero drag losses against gravity losses and then tweak your delta-V to get exactly to your destination with just enough left over to recover your launch stage. That's all assuming of course that your payload is either stable at the destination your launcher takes it to or has its own propulsion to get beyond that point. BTW - look up Warzouz' Cygnus Recoverable Rockets for a beautiful example of how to do this. His rockets are designed for a specific payload weight and lift that payload to about an 80k orbit, then land back at the KSC with excellent recovery rates and very little waste. So far the ones I have used have been stable and pretty much flawless for their advertised purpose.

Danny

[Gojira1000]

Or if you're me and have a lazy-man's super-heavy booster in the VAB as a sub-assembly, you get fun variations from 1.10 to 2.85 at launch. I should really make more versions in there other than "light and low-tech", "medium-and-never-used" and "stupidly overpowered, excellent!"

[Laie]

But as long as you steer clear of all those kinds of disasters, why would one not use as much thrust as available?

In real life? Because engines are expensive and fuel is dead cheap, at least in comparison. Also, every staging event is chock-full with possible failure modes -- but that's probably already summarised under engines being expensive. Long story short, using fewer stages and running them longer is cheaper / easier / more reliable. We are optimising for cost, not for takeoff mass.

In KSP, however, we don't even have a common benchmark for what constitutes the "best" rocket. Some people pine for the cheapest design, others are willing to pay a premium for a speedy ascent so they can get on with the mission. Others take pride in smart or elegant solutions, or ease of use.

[GoSlash27]

Snark's got it covered. The real cost of high t/w isn't so much the drag penalty as the mass and cost of the engines. There's also a penalty in additional control systems (fins/ reaction wheels/ batteries/ etc) to keep a high t/w launcher pointed the right way.

A high t/w stage can get the job done with minimal DV expenditure, but this doesn't necessarily translate to the lightest, cheapest, or most fuel- efficient way to go about it.

Generally, the important criteria for a booster stage is bucks per tonne and for an upper stage is mass. Neither is optimized by seeking minimum DV expenditure through high acceleration.

I design my boosters to have an initial t/w of 1.4. This tends to maximize the payload fraction. SRBs will go even lower in order to keep the t/w in check throughout the flight (they can't be actively throttled).

Best,

-Slashy

Edited November 5, 2015 by GoSlash27

[KerikBalm]

Just to point out one thing, the optimal ascent to minimize gravity drag and aerodynamic drag is when the force or aerodynamic drag = the force exerted by gravity... ie... at terminal velocity, except going up, instead of falling down.

To maintain terminal velocity in an ascent, you need a TWR of 2:1...

But as you get higher, terminal velocity generally goes up as air density goes down... which would imply higher acceleration is needed to because you don't need to just maintain a certain velocity, but you need to accelerate to ever higher terminal velocities...

But as your rocket burns fuel, TWR goes up.

A stage that starts at 1.5:1 can easily end with a much much higher TWR (as atmospheric pressure decreases, thrust and Isp inrease... and the mass you are pushing decreases).

So when I build SSTO rockets, I aim for 1.3:1 initial TWR.

Then of course minimizing drag is one thing... but if it means taking extra engine weight, it may be better to take a bit more of a hit to drag, and use less mass for engines.

Don't throttle down to stay under a 1.X :1 TWR...but you can keep adding fuel until you hit that TWR.. TWR goes up in flight, but it may be a good idea to throttle back when TWR starts going well over 2:1

[GoSlash27]

Just to point out one thing, the optimal ascent to minimize gravity drag and aerodynamic drag is when the force or aerodynamic drag = the force exerted by gravity... ie... at terminal velocity, except going up, instead of falling down.

To maintain terminal velocity in an ascent, you need a TWR of 2:1...

But as you get higher, terminal velocity generally goes up as air density goes down... which would imply higher acceleration is needed to because you don't need to just maintain a certain velocity, but you need to accelerate to ever higher terminal velocities...

But as your rocket burns fuel, TWR goes up.

A stage that starts at 1.5:1 can easily end with a much much higher TWR (as atmospheric pressure decreases, thrust and Isp inrease... and the mass you are pushing decreases).

So when I build SSTO rockets, I aim for 1.3:1 initial TWR.

Then of course minimizing drag is one thing... but if it means taking extra engine weight, it may be better to take a bit more of a hit to drag, and use less mass for engines.

Don't throttle down to stay under a 1.X :1 TWR...but you can keep adding fuel until you hit that TWR.. TWR goes up in flight, but it may be a good idea to throttle back when TWR starts going well over 2:1

KerikBalm,

Ah, but this is only true for vertical flight and ignores Mach effects on drag.

As you turn prograde and fly more horizontally, the optimal acceleration falls below 2 Gs in proportion to the sine of the pitch angle.

As a practical matter you'll never reach terminal velocity in the vertical boost phase.

And as I said earlier, minimal DV expenditure isn't a useful design criteria anyway. It's really about minimal cash/ fuel/ and stage mass.

Best,

-Slashy

[magnemoe]

I tend to launch at 1.3-1.4, if not more boosters. uses mechjeb launch autopilot and it works well outside of having to adjust parameters to do an more 0.9 burn then you launch an barn, this also require closer to 4km/s dV but you get your barn into orbit.

[uglyduckling81]

Not sure how people are launching with less than a TWR of 1.5? If I go lower my craft generally just sits on the pad thrusting and not moving. Usually I go for 1.6 just to make sure I'm actually going to lift as even 1.5 sometimes just sits there.

It's always been that way for me despite reinstalling the game numerous times on different PCs.

[BoilingOil]

Looks like your question has been really well covered, but I'm going to put in my half-cent's worth anyhow...

Great, thanks!

High TWR can be fun. It's also very easy. You stick a big-ass fuel tank on a big-ass engine and get 3500 delta-V and 2.5 TWR out of it, and you're [almost] guaranteed to go to space today.

Yeah, funb and easy. But not much of a challenge anymore in the end... I've kicked some serious poo into orbit at breakneck speeds myself, so maybe I need to do something new

vanilla KSP can be pretty forgiving with most parts (maybe more than it should)

Agreed! Especially in easy sandbox mode. Load her up, and kick her in the butt. HARD!

Your real cost (and I don't think anyone has said this yet) is that you could have hauled up a bigger payload

BINGO!!

So, again as many people have said, your best bet is to set your TWR to something reasonable that will balance aero drag losses against gravity losses and then tweak your delta-V to get exactly to your destination with just enough left over to recover your launch stage. That's all assuming of course that your payload is either stable at the destination your launcher takes it to or has its own propulsion to get beyond that point. BTW - look up Warzouz' Cygnus Recoverable Rockets for a beautiful example of how to do this. His rockets are designed for a specific payload weight and lift that payload to about an 80k orbit, then land back at the KSC with excellent recovery rates and very little waste. So far the ones I have used have been stable and pretty much flawless for their advertised purpose.

Danny

Ah, I'll look at that link. Thank you a lot, Dan!

- - - Updated - - -

I design my boosters to have an initial t/w of 1.4. This tends to maximize the payload fraction. SRBs will go even lower in order to keep the t/w in check throughout the flight (they can't be actively throttled).

Best,

-Slashy

Thanks for the tip, I'll see how I can use that

Edited November 5, 2015 by BoilingOil

[Red Iron Crown]

Not sure how people are launching with less than a TWR of 1.5? If I go lower my craft generally just sits on the pad thrusting and not moving. Usually I go for 1.6 just to make sure I'm actually going to lift as even 1.5 sometimes just sits there.

It's always been that way for me despite reinstalling the game numerous times on different PCs.

Are you perhaps looking at vacuum TWR instead of sea level TWR? Anything with >1 sea level TWR should lift off.

[BoilingOil]

A stage that starts at 1.5:1 can easily end with a much much higher TWR (as atmospheric pressure decreases, thrust and Isp inrease... and the mass you are pushing decreases).

So when I build SSTO rockets, I aim for 1.3:1 initial TWR.

Marvellous!

- - - Updated - - -

I tend to launch at 1.3-1.4, if not more boosters. uses mechjeb launch autopilot and it works well outside of having to adjust parameters to do an more 0.9 burn then you launch an barn, this also require closer to 4km/s dV but you get your barn into orbit.

Your barn, and then some. Right?

- - - Updated - - -

Not sure how people are launching with less than a TWR of 1.5? If I go lower my craft generally just sits on the pad thrusting and not moving.

That's why I started this thread! I had seen some stuff just sit on the platform, heating the sewers! So I wondered "What low TWR?"

Anyway, you've all been incredibly helpful here, guys. Thank you a million times and then some. If anyone has anything more to add, though, please keep it coming! I'm highly appreciative of all info on the subject

Edited November 5, 2015 by BoilingOil

[uglyduckling81]

Are you perhaps looking at vacuum TWR instead of sea level TWR? Anything with >1 sea level TWR should lift off.

Seems like a weird thing for me to of missed this whole time but I'll check it anyway. Does KER even give the TWR in vacuum in the assembly building?