Ideal thrust per weight Ratio of rocket for Lift Off stage experiment.

[Moonfrog]

I conducted some experiments with test rockets to try and find a good TWR for lift off. I use the mod kerbal engineer to calculate the TWR.

I built several rockets with 3000 DV roughly. admittedly with about 50m/s of DV of error. I ranged from TWR of 4 to 1.2.

In order to measure its effectiveness I used the final Apoaphahis after fuel ran out.

What I found was that a TWR of 1.44 reached 212,000 m. Using the same DV, with a TWR higher or lower than this reduced the distance it travelled altogether from lift off.

Just wondering If anyone knew why this was the case and if they had different results with a different TWR, maybe got better results.

[Dispatcher]

Tests will vary, depending on your payload, engine(s) used, tonnage of full fuel tanks, drag, efficiency of design (and for keeping a ship together, perhaps strutting), piloting style; things like that.

If you want to see a TWR chart, visit the KSP Wiki. Or, the same info and my own crude testing is on my chart linked in the sig line. I'm sure others will have helpful suggestions for you.

[shand]

ideal thrust to weight depends on the current conditions.

THEORETICAL IDEAL is

On the pad: infinite.

During in atmosphere flight: 1+ whatever is needed to counteract air resistance + whatever is needed to accelerate so as to maintain optimum speed (generally speaking the opposite of terminal velocity - so terminal velocity going upwards)

During exoatmospheric flight: infinite.

of course. thats assuming the extra T/W doesnt cost you anything. in real terms i agree with your findings: higher the better til (-)terminal velocity, then 1.4 all the way up. obviously you have to compromise at times with a lower ignition T/W and throttling down as you get to the end of a stage. above (i think its around) 20k 2 t/w is good, and above 40k its largely irrelivent, i tend to have 0.9 on ignition for my final stage.

My tip would be to try the same experiment again using your 4 T/W rocket and follow my theoretical ideals list. kerbal engineer gives atmospheric efficency - try to keep that between 95-105% by throttling and you'll get a good idea of whats optimum.

[Alexandicity]

Well, it's not quite a linear response. You could have a huge TWR at liftoff but that would be bad, because you'd accelerate too fast too soon. Conversely, too low and you're in the gravity field longer than necessary. If the atmosphere was constant, there would probably be a value, maybe like the 1.44 you found, that is optimal.

However, the atmosphere isn't constant, and the fact that your craft becomes lighter as fuel is burnt (increasing TWR), means there isn't a single TWR that is good for the full ascent. I'm not sure, since I don't usually do TWR calcs, but I'd guess that you want a high TWR at first, but have it drop fairly quickly to something lower. Maybe someone else more familiar with TWR optimisation can confirm!

Also worth noting that your drag - which will also affect max altitude - is likely to vary between the various test crafts you built (unless you compensated for that somehow?).

All in all, however, good to see such experimentation! It's a complex problem but nice to see the methodology being applied! Would you mind posting your data?

[Francesco]

If you're building a single stage to orbit, you can actually have a TWR even lower than 1.44.

You said you did some experiments, but did all of your ships use the same type of engine? ISP differences might skew your results.

The number of stages also matter:

with a simple two stage rocket, your liftoff TWR could be just that, 1.4-1.5, because the first stage is going to burn for quite a long time.

On the other hand, with asparagus staging, you'll probably want every stage TWR to be a little higher, let's say 1.6-1.7, so that you keep the overall TWR of the ship constant throughout the various staging events.

This is just my rule of thumb though, experiment with this yourself

[numerobis]

If you're just trying to minimize deltaV, you'd want infinite TWR on liftoff as shand says, to get you to terminal velocity instantaneously; then TWR just above 2 to maintain terminal velocity in vertical flight. When you start to turn, I don't know anymore what's optimal, but TWR just over 2 still seems to perform well.

Most people want to minimize something else -- mass, part count, cost, etc.

[Moonfrog]

The conditions I did it in, was just put the engine from launch at full thrust to see how high it would go. I did use different engines, but I maintained the DV was roughly 3000. If the ISP was less, it needed more fuel. I don't really understand what atmosphere efficiency is on kerbal engineer mod. It doesn't really explain how its calculated.

I know that drag is a factor, if I go too fast, the drag increases a lot which should slow me down. But it doesn't, instead it seems to use more fuel, as my acceleration doesn't seem to slow down. Likewise if I start of with really slow acceleration have a very low drag, I use up a lot of fuel as well, which I now think is due to being in the effect of gravity for so long.

I experimented with a rocket with a TWR of 6, it travelled to fast so quickly it turned Red on ascent which was quite cool. Unfortunately I didn't write any of the data down, I just sort of built test rocket then modified in game, and looked at the different distance it travelled. Trying to make the distance as far as possible.

The reason behind doing it was to try and build a good Booster section or at least know how to do my designs as DV alone doesn't seem to insure that u get the best out of your fuel. I try that suggestion of keeping atmosphere efficiency at 100% and see if that turns out better.

This was the final rocket I used. It had a command pod mk1, a small parachute on top, the white one mk16 I think its called. An FL T800 fuel tank plus FL T-100 fuel tank. Small stack bi coupler. Two Lv909 engines. This is what gave me 1.44 TWR, and 2984 DV. I didn't modify the throttle just kept it on full. On average it performed the best. The fastest rocket I had was a skipper engine attached with the rockomax adapter just put enough fuel on it to get as close to 3000 as possible. Can't remember exactly how many FL T800's I used. It was a bit wobbly.

[capi3101]

Optimal TWR at launch is 1.2 for SSTO rockets, 1.6-1.7 for asparagus; everything else is somewhere in-between. That's based on my observations anyway.

[numerobis]

I feel like there's confusion about causality between drag and speed in your account. For a fixed spacecraft, with its engines on, drag doesn't cause you to slow down -- rather, increasing speed causes drag to go up.

In vertical flight, the deltaV used to fight gravity goes down with speed, whereas the deltaV used to fight drag goes up with speed. The point at which you minimize the losses is when they are equal, ie at terminal velocity. To maintain terminal velocity you need twice local gravity: once to fight gravity, plus once to fight drag. But the air thins out, so you need a bit extra thrust to keep up with the increasing terminal velocity.

The flight path that minimizes deltaV to orbit is not entirely clear to me, but it certainly is not to fly vertically the whole way.

[Moonfrog]

I only flew vertically for test purposes. drag should slow u down, but it doesn't in kerbal game. Not sure why. Its why drag is minimised on planes and cars for speed. It acts as a force pushing against you, yes in game it just seems to decrease the effectiveness of your delta V. Its impossible to see your terminal velocity with this system because you continue to accelerate beyond terminal velocity. Unless i'm mistaken shouldn't be possible. I am confused about that.

[SRV Ron]

It is a balance between atmospheric drag and the drag of gravity. Something as simple as an engine change can nearly double the performance of a given design when it falls into that sweet zone of performance.

A test suborbiter using Novapunch mode reached 71K altitude using the K2-X engine which had too much acceleration for the weight being launched. Using a smaller RMA-3 gave an altitude of 182K. An LV-909 gave less altitude for a longer lower powered burn.

Play with different engines, stagings, even SRBs for the initial launch phase, to find the sweet spot.

[numerobis]

I only flew vertically for test purposes. drag should slow u down, but it doesn't in kerbal game. Not sure why. Its why drag is minimised on planes and cars for speed. It acts as a force pushing against you, yes in game it just seems to decrease the effectiveness of your delta V. Its impossible to see your terminal velocity with this system because you continue to accelerate beyond terminal velocity. Unless i'm mistaken shouldn't be possible. I am confused about that.

Drag certainly does act as a force pushing against you in KSP; that's why you lose delta V. It's also why there *is* a terminal velocity; otherwise, gravity would be able just to accelerate you forever until you smack into the surface -- as it does on airless bodies. On an airless body, the ideal TWR is infinite: you're just doing a Hohmann transfer. Drag is why this whole question is complicated.

Terminal velocity refers to the speed that you'd be falling in steady state due just to gravity and drag: at terminal velocity gravity pulls you down at roughly -10 m/s^2, and drag pushes you up the same amount. But it's not a hard limit like the speed of light: you can go faster if you have engines. At TWR 2 flying straight up, you can go at terminal velocity: you lose one TWR to gravity, and one TWR to drag, by definition. At TWR 10, you can go three times faster: you still lose one TWR fighting gravity, and the remaining nine fight drag, which grows quadratically in speed. That quadratic relationship is why you want to avoid flying too fast. The atmosphere thins out with altitude, so that terminal velocity roughly triples every 10km of altitude on Kerbin; by 30 km, terminal velocity is higher than orbit velocity.

[ae00505]

I am trying to analyse your situation but I need more information like what else did you change or kept constant while changing the TWR. Also I need to know if you have fired the rocket vertically up or you did give any [B]Pitchover Manoeuvre[/B]? If yes, then what was your [B]kick angle[/B]?

In the straight vertical firing case, when pitch angle is 90 deg. and there is not pitch change, I found that TWR ratio is not giving any peak in culmination altitude. It keeps increasing but the rate is coming down. You can see that in the graph I plotted below. Don't worry about the culmination altitude value as it also depends upon the Mass ratio and specific impulse which I took as some constant value.
[IMG]https://drive.google.com/open?id=0Bwfo4o89dVSHVjBvSEFRYlNWNms[/IMG]

[COLOR="silver"][SIZE=1]- - - Updated - - -[/SIZE][/COLOR]

I got the answer. In my previous msg I shared a graph of culmination altitude Vs TWR, where there is no peak visible. But after reading your query twice I found that you are talking about the burnout altitude not the culmination altitude and yes burnout altitude has an optimal TWR. Any value before and after it will bring down the burnout altitude. I am not able to share the graph picture, if you can teach me how to share the graph, I shall share it with you.

I think there should be an option to share the pictures and vids from your computer rather than sharing only the links. If there is any, I don't know yet, I am new to this forum.

[OhioBob]

You guys are responding to posts that are over two years old.

[Snark]

[quote name='OhioBob']You guys are responding to posts that are over two years old.[/QUOTE]

Gah. Thank you for the heads-up, didn't notice that the thread had been necro'd.

Ae00505: Welcome to the forums! Always glad to see a new KSPer coming online. :). However, a piece of forum etiquette: please don't respond to ancient threads, they're almost never very relevant (due to massive changes in the game over time) and it just causes confusion, not to mention cluttering up the forum UI by pushing the dead thread up to the top of the most-recently-updated list.

[Claw]

Indeed. If there is new insight for this topic, feel free to start a new thread (given all the changes).

Cheers,
-Claw