Looking for info on TWR - gravity losses - steering losses

[Norcalplanner]

So I've played this game for a year now, and feel like I have a pretty good idea what I'm doing. Nevertheless, I feel like I need some additional education.

For my entry in the Fastest small step for a Kerbal challenge (see here) I was under the impression that there was some value in keeping a higher TWR early in the burn. I kept my TWR above 1 for the first 2,500 m/s worth of ejection burn, and kept it above .75 for another 2,000 m/s on top of that before stepping down to .3 or so. I departed from a 200 km LKO, thinking that I would be minimizing gravity losses compared to the more Kerbal method of pointing directly at the target. The whole point with this challenge is to get there as soon as possible - the craft had around 35 km/s of delta V. I also needed to match a fairly precise ejection angle which was not parallel to Kerbin's orbit.

So to get to my questions - is there any good thread discussion or information source for weighing these various factors? Higher TWR over a shorter period of time to take greater advantage of the Oberth effect and reduce gravity losses vs. lower TWR over a longer period of time (and likely at a higher Isp), even though there will be greater gravity and steering losses? Is there a chart somewhere which shows Kerbin gravity plotted by altitude or discusses gravity and steering losses? And with the way the game figures velocity and gravity, is there any advantage to trying to complete most of the burn before departing Kerbin's SOI when doing this sort of speed run?

Thanks in advance for any help.

[AlexinTokyo]

If you are going for a ballistic (i.e. unpowered) transfer, then instantaneous impulse gives the theoretical maximum in terms of efficiency. Note that in-game maneuver nodes (as well as the transfer window planning tools) all assume instantaneous impulse, in order to keep the calculations manageable.

Since you can't (legitimately) achieve instantaneous impulse, as it would require an infinite thrust to weight ratio, the highest practically achievable efficiency would be given by the highest practically achievable TWR (always assuming that acceleration at that rate doesn't rip your ship apart, which has a serious negative impact on efficiency). I_sp really only enters into consideration in determining whether the engine configuration for high TWR can actually provide the requisite ÃŽâ€v.

However, the fastest transfer possible may not be a ballistic one. My understanding is that the fastest possible transfer IRL would be one of continuous acceleration until the half-way point of the journey, followed by continual deceleration until arrival. With this class of transfer, I_sp is likely to be much more important as you are unlikely to be able to carry enough fuel with anything other than non-traditional propulsion. In this case, TWR is less important, as you are applying the impulse over the course of the entire journey. Of course, planning a transfer of this type in KSP is a non-trivial exercise, as the maneuver-node system assumes ballistic transfers.

In fact, in stock, 0.90.0 KSP the fastest transfer to a body with an atmosphere would likely be continual acceleration until arrival, followed by an extremely low-altitude aerobrake. Since shock heating and aerodynamic stresses are not modeled at all by the stock game (currently), it is possible to gain almost infinite impulse from aerobraking at negligible risk (the only risk being accidentally lithobreaking instead). This might not work at Duna, as the atmosphere is so thin, but would almost certainly be feasible at Eve.

[Foxster]

I'd be interested in anything on this too. It is a black art to me as I don't understand the maths or mechanics.

I have a feeling that things like the Oberth effect and some of the other losses are fairly marginal i.e. on a 35-45kdV craft they might get lost in the noise as they could be 10s of dV.

If you are in a lower orbit to take advantage of Oberth you might need to waste some time swinging around the planet to get the planet out of the way for a burn, whereas at a higher altitude the arc of Kerbin is less and so easier to miss and you can go straight for a burn...but it took a little longer to get to the higher altitude...argh, I dunno.

What does seem to help on these ballistic crossings is lots of adjustment. I was initially using one node and burn to leave Kerbin's SOI and then another to get up to max speed, with one correction burn to get a closer Pe at halfway through the coasting. Seems better though to keep adjusting through the burn. It does mean more hands on though. I guess that's as a result of instantaneous impulse (whatever that is) not being possible as said above.

I suppose one of the big questions is as you said, at what point do you stop needing higher TWR? Obviously you need >1 to be able to launch and then >1.2 for efficiency in the atmos. When though is it OK to start using a TWR of, say, 0.1 with high ISP engines? To leave orbit is it better to have greater TWR until you are closer to aligned with the target? How would you even calculate that? And does it really matter?

[John FX]

The whole point with this challenge is to get there as soon as possible - the craft had around 35 km/s of delta V. I also needed to match a fairly precise ejection angle which was not parallel to Kerbin's orbit.

is there any advantage to trying to complete most of the burn before departing Kerbin's SOI when doing this sort of speed run?

If you need to be there as fast as possible after starting your burn then do it from a high orbit.

If you need precision on your ejection angle have the highest TWR you can.

If you need to be there fast accelerate as fast as possible, meaning don't wait to do any burns if you can accurately burn earlier.

If you need to be there fast get to your cruising speed as soon as you can.

If you have 35KDv+ then the gains from not doing this are negligable

[Foxster]

I've been playing with the various trajectory plotting tools - Mechjeb porkchop, Kerbal alarm clock transfer window alarm, Transfer Window Planner mod and its online equivalent. With my current level of understanding, the fastest possible non-ion one-ship ballistic trip to Jool is around 60 days, plus more time for burns and Jool orbit establishment (after aerobraking).

This is based on the ship with the most dV I could build and launch in one go having a dV on the pad of about 45kdV. It might be possible to get a bit more but its getting very massive and the returns are diminishing. This means about 40kdV available at a 100km orbit. You need some spare dV for course corrections and orbit establishment. That means having about 37kdV at most to play with and perhaps less. I've tended to err on the safe side and flown with 35kdV.

There are only so many windows and they come around again. Looking 10 years ahead for Jool, the best transfer time with 37kdV is around 62 days. Might find something a little better but its in that ballpark. The only savings then to be made seem to be in leaving Kerbin orbit and burning until the fuel safety margin is reached but I can see that saving a few hours at best if optimised.

Unless there is some major thing I'm missing here? Is there a way to do this with one non-ion ship that will shave days off the trip?

Edited March 31, 2015 by Foxster

[OhioBob]

So to get to my questions - is there any good thread discussion or information source for weighing these various factors? Higher TWR over a shorter period of time to take greater advantage of the Oberth effect and reduce gravity losses vs. lower TWR over a longer period of time (and likely at a higher Isp), even though there will be greater gravity and steering losses?

I don't know if the following is what you're looking for, but I just performed an experiment in which I simulated an interplanetary injection from low Kerbin orbit using different thrust-to-weight ratios. I assumed starting from a 70 km circular orbit and maintained the thrust vector pointed in the prograde direction throughout the duration of the burn. I stopped the burn when the hyperbolic excess velocity reached 1,000 m/s, which is about the median for a trip to Duna or Eve (a little less for Duna and a little more for Eve). I assumed a specific impulse of 390 seconds, though Isp really doesn't make that much difference. Below are the results:

[TABLE=width: 300]

[TR]

[TD=align: center]TWR[/TD]

[TD=align: center]ÃŽâ€v (m/s)[/TD]

[/TR]

[TR]

[TD=align: center]∞[/TD]

[TD=align: center]1101.5[/TD]

[/TR]

[TR]

[TD=align: center]2.00[/TD]

[TD=align: center]1102.6[/TD]

[/TR]

[TR]

[TD=align: center]1.50[/TD]

[TD=align: center]1103.2[/TD]

[/TR]

[TR]

[TD=align: center]1.00[/TD]

[TD=align: center]1104.9[/TD]

[/TR]

[TR]

[TD=align: center]0.75[/TD]

[TD=align: center]1107.3[/TD]

[/TR]

[TR]

[TD=align: center]0.50[/TD]

[TD=align: center]1113.8[/TD]

[/TR]

[TR]

[TD=align: center]0.30[/TD]

[TD=align: center]1133.6[/TD]

[/TR]

[TR]

[TD=align: center]0.20[/TD]

[TD=align: center]1167.9[/TD]

[/TR]

[TR]

[TD=align: center]0.15[/TD]

[TD=align: center]1209.2[/TD]

[/TR]

[TR]

[TD=align: center]0.10[/TD]

[TD=align: center]1301.7[/TD]

[/TR]

[/TABLE]

A TWR of ∞ is, of course, a theoretical instantaneous burn. As you can see, for TWRs greater than about 0.50, the losses are very small. However, for extremely small TWRs, the losses start to become significant.

Another problem with small TWRs is the issue of obtaining a good intercept with the destination planet. Low-thrust burns are so long in duration that there can be considerable error between the planned trajectory, based on the maneuver node, and the actual trajectory, based on the injection burn. After performing a high-thrust burn it is typical that only a small correction is required to fine tune the trajectory. Low-thrust burns can require rather significant course corrections, which must be added on top of the losses we see in the table above.

[Norcalplanner]

OhioBob, thanks for that info. That's the sort of thing I'm looking for. Now the question in my mind is how large the gravity losses are when a craft is 500km or more away from Kerbin and still burning.