Flight Question How can I find launch windows?

[Xavier]

I launch at the wrong times apparently, because my escape vector is too high.
How can I calculate a launch window to other planets? Is there an MFD?

 

[Tommy]

IMFD and TransX can both be used to find launch windows. If you search the forum using "launch window" you'll find some links to websites that have launch windows.

If you use IMFD, get the "IMFD FUll Manual/Playbacks" from Orbithanger - it has a step by step process to find launch windows.

 

[statickid]

yeah it just depends on where you are going. you can get a launch window for the ISS by just watching the map and guessing after a few tries you will know when to launch intuitively. It won't be nearly as complicated to to interplanetary transfers if you just launch into a nice circle orbit before making your interplanetary burn then you can do all your planning based off the circle. if you want to launch straight into the transfer it's a little more sensitive and ramps up the difficulty but learning the MFD's it's not too much trouble, it is the most difficult to do though, IMO

 

[Tommy]

It won't be nearly as complicated to to interplanetary transfers if you just launch into a nice circle orbit before making your interplanetary burn then you can do all your planning based off the circle.

Whether you are in orbit or not doesn't have anything to do with whether or not you are using a good launch window. Until you have selected the launch window you can't predict the ejection plane - so your nice, circular orbit will most likely NOT be aligned with the ejection plane. Making the plane change from your parking orbit's plane to the ejection plane can cost more dV than the transfer burn, because the plane change will happen in low orbit where plane changes are most expensive. If you eject in a poor plane, you will have a much larger plane change that will need to made enroute - unless you get lucky enough that the intercept point coincides with a node between your ejection plane and the target's orbital plane.

What ever tool or method you use to find a window, efficiency requires you to launch into the correct orbital plane - the ejection plane, and to do so at a time when the energy required for the transfer is at a minimum.

Sure, some vessels have enough delta-V to compensate for poor planning, but an interplanetary flight should be planned BEFORE launch.

 

[Wishbone]

Also heartily recommend playing with Arrowstar's Trajectory Optimization Tool (available at OHM).

 

[Axertan]

I use Cosmic Train Schedule for determining my launch windows to other planets. Those tables are for Hohmann Transfer Orbits tho'.

 

[statickid]

Whether you are in orbit or not doesn't have anything to do with whether or not you are using a good launch window. Until you have selected the launch window you can't predict the ejection plane - so your nice, circular orbit will most likely NOT be aligned with the ejection plane. Making the plane change from your parking orbit's plane to the ejection plane can cost more dV than the transfer burn, because the plane change will happen in low orbit where plane changes are most expensive. If you eject in a poor plane, you will have a much larger plane change that will need to made enroute - unless you get lucky enough that the intercept point coincides with a node between your ejection plane and the target's orbital plane.

What ever tool or method you use to find a window, efficiency requires you to launch into the correct orbital plane - the ejection plane, and to do so at a time when the energy required for the transfer is at a minimum.

Sure, some vessels have enough delta-V to compensate for poor planning, but an interplanetary flight should be planned BEFORE launch.

I just do off-plane intercepts- no problem

yes, you will have to launch at the right time of year

 

[garyw]

This Spreadsheet comes in handy for giving you dates for launch windows. Once you have the date you can mess with IMFD or TransX for the perfect setup.

 

[statickid]

who said the circular orbit wasn't aligned with an ejection plane anyways? :thumbup:

 

[Tommy]

I just do off-plane intercepts- no problem

This still has nothing to do with whether or not your parking orbit is aligned with the ejection plane. If it's not, you will either need to make a plane change before the ejection, or incorporate it into the ejection burn. Either way, it's very fuel expensive. Besides, the OP was most likely using TransferMFD, which doesn't calculate off-plane transfers.

who said the circular orbit wasn't aligned with an ejection plane anyways?

The chances that any orbit you fly into are going to turn out to be the same as the ejection plane is very small.
You won't know what the transfer plane is until you plan the course. This is even more true with off-plane transfers.

Let me break this down for clarity (for the newer orbinauts). For any interplanetary transfer, there are at least three planes that need to be considered. For an Off-plane transfer four planes need to be accounted for.

The first plane is the Source Plane. This is the orbital plane of the body you are starting from. The second is the Target Plane, and is the orbital plane of the destination body.

Now, place an imaginary point at the position you will leave the source body. This point is going to be the Source plane. Now, place another point at the position where you will intercept the target. This point will be on the Target plane (obviously). Last, place a third point at the center of the Reference body (the body both the Source and Target orbit - for an Earth - Mars transfer that is the Sun). These three points define a plane - this is the Transfer Plane. If your ejection point and intercept points both coincide with nodes between the Source and Target planes - this will be identical to the Source plane. Most times you won't have nodal alignment at both ends - so the Transfer plane will be different from both the Source and Target planes. This means there is a relative inclination, and you need to change your plane (with regards to the Reference body) during the ejection.

The last plane to consider is the Ejection plane. Unlike the other planes, this one is referenced to the Source body - not the Reference body. The Ejection plane (even if referenced to the Reference body) will rarely be the same as the Transfer plane because the orbital velocity (of the parking orbit) is used to help make this plane change from the Source plane to the Target Plane. The ejection plane is determined by a combination of The Source plane, the Transfer plane, and the vessels velocity as it leaves the Source body's SOI.

In other words, there is no way to determine the Ejection plane until the transfer has been calculated. Sure, you might get lucky and launch into the correct Ejection plane - but it's not very likely.

If your Ejection plane is incorrect, you will need to make much larger mid-course corrections - assuming you don't miss the target altogether.

 

[jthill]

the ejection plane

There really isn't any singular eject orbit plane. There's a singular eject vector that puts you on your transfer orbit, but how you get to it ... well, TransX will show you how to get there from a huge range of LEO planes. It's the "Ej Orientation" parameter in the Escape stage.

And since you're going to be raising ApA rather a lot anyway, doing it in stages to stop by a nice low-cost plane change node on the way isn't much trouble.

 

[Tommy]

There really isn't any singular eject orbit plane. There's a singular eject vector

A vector is a line, and lines are on planes. In fact, the ejection isn't instantaneous and the vessel travels through an arc during the ejection burn - and an arc defines a singular plane. I suppose if you really want to nit-pick you can call it an optimal ejection plane.

TransX will show you how to get there from a huge range of LEO planes. It's the "Ej Orientation" parameter in the Escape stage.

IMFD can do this as well, but what you are doing is incorporating a plane change, or adding outward velocity, into the ejection burn - and that's less efficient than an "in plane" prograde burn. It will result in a less efficient transfer.

For ANY transfer, there is one "parking orbit" plane and ejection point that will work most efficiently and accurately. If you aren't in that plane before you eject, you WILL pay for it in corrections - or in a less efficient transfer orbit.

Not to mention that it's easier to be in the right plane - especially for a newer orbinaut. It makes for a simpler solution as well as a more efficient solution.

Trust me - NASA chooses it's ejection plane and point VERY carefully when launching probes to other planets.

 

[statickid]

doing off plane transfer uses no more fuel than the other type

 

[jthill]

lines are on planes.

The eject vector is on an infinite number of planes. That's the point. They're all good. No plane is optimal for all starting points. NASA picks the one that's cheapest to reach from Cape Canaveral From the ISS, it'll be cheapest to reach another, and from the Moon, a third.

Plane change costs don't have to be nearly so expensive as you make it sound. Changing planes at low nodes is expensive, but "don't do that, then" works well here too. You might notice that the one-burn transfer we're discussing starts out with a low-node plane change. There's a whole lot of space to play in, here. There isn't just one right way to do it.

 

[Tommy]

Off plane is rarely the most efficient transfer - usually only when the ejection and intercept points align with the nodes between the source and target planes (in which case the transfer is actually a single plane Source plane transfer), or are VERY close to the nodes. Sometimes it is the only transfer available - this can sometimes be the case with an Earth - Mars trip.

How much less efficient one type of transfer is compared to another will vary with each window, as will the difference in efficiency between the various types of transfer.

For an Earth - Jupiter flight using the 1976 Voyager window, a Target plane transfer is about 7% more efficient than an off-plane transfer (and also sets up a better approach for a slingshot to Saturn if you plan to continue past Jupiter).

I'll add that the difference in efficiency is more pronounced on outbound trips than it is inbound, and the farther from the nodes the transfer takes place, the less efficient an off-plane transfer is.

I'll also add that a "standard" two plane transfer (which starts on the source plane, then changes to the target plane at the node) is often less efficient than either a Source plane or Target plane transfer.

---------- Post added at 03:18 AM ---------- Previous post was at 02:43 AM ----------

The eject vector is on an infinite number of planes.

No. A point is on an infinite number of planes. A vector is NOT a point - it is a line, and lines exist on a finite number of planes. In reality, a vessel doesn't eject in a straight line (we use a vector in calculations because it's simpler, and close enough), rather it travels in a curve - and curves exist in one plane only.

When you change your ejection orientation you are changing the transfer plane. Not by enough to see on an MFD screen, but enough to make a difference in efficiency.

For any interplanetary transfer, there is exactly one optimal plane on which you should leave the source. This will vary depending on the type of transfer (Off-plane, target plane, etc), but for any specific transfer (ie, window and transfer type) there is ONE optimal departure plane. For any given departure plane there is ONE optimal ejection plane (although either prograde or retrograde will work).

The reason you would use a different ejection plane if you left from the ISS (as opposed to KSC) is because it is cheaper to modify the transfer plane slightly than it is to modify your Earth orbit by several degrees. Even if you raise your Apo, a plane change of several degrees will still be more expensive than changing the departure plane by half a degree. And that will only work well if the node coincides with the ejection point.

If you are leaving from KSC, you can launch directly into a plane much closer to the optimal ejection plane and eject into a plane much closer to the optimal departure plane without needing to make a plane change in LEO. If you launch from a low enough latitude - you can launch into the optimal ejection plane and eject into the optimal departure plane - Nasa launches from KSC because the constraints of geography, safety, politics, and weather patterns make a more southern launch site inadvisable.

In other words, sure - you can use any ejection vector you want - but you'll pay for it in reduced efficiency of the transfer as a whole. The difference may be small, but in "real life" vessels with realistic Delta-V budgets, every gram of propellant counts.

 

[orbekler]

...For any interplanetary transfer, there is exactly one optimal plane on which you should leave the source. This will vary depending on the type of transfer (Off-plane, target plane, etc), but for any specific transfer (ie, window and transfer type) there is ONE optimal departure plane. For any given departure plane there is ONE optimal ejection plane (although either prograde or retrograde will work)....

This is a VERY interesting issue. I'd like to ask you if have have an idea of the algorithm IMFD uses to find TIn. Let's take for example the classical Earth-Moon trip, Off-Plane. When you target, you'll see that TIn assumes always, does not matter wher you are, if on surface or in any orbit, a value between 390ks and 460ks. I may be wrong, but to me, normally that value has nothing to do with the FIRST BEST window available, because if you tune TIn manually starting from 0, you should find a better, more efficient time of intercept. It seems more to me a default or failsafe value. What do you think about?
Just another small thing: while an aligned plane Eject should be just a little more efficient than an Off-Plane, it seems that iV value is far greater in an Off-Plane, something that normally is not taken into account in Trans-X.
In these pictures (EIn~0 for both eject), dV value in Off-Plane is a little better, probably due to my poor and rushed settings :lol:

 

[Wishbone]

IMFD _does not_ optimize your transfer.

 

[orbekler]

IMFD _does not_ optimize your transfer.

That's why I was asking about that "variable" default TIn value - it's rather important for Surface Launch.

 

[Tommy]

I may be wrong, but to me, normally that value has nothing to do with the FIRST BEST window available, because if you tune TIn manually starting from 0, you should find a better, more efficient time of intercept. It seems more to me a default or failsafe value. What do you think about?

It's more of a FIRST REASONABLE window, but it won't be the most efficient. I'm not sure exactly how th algorith functions, treat it as a "ballpark" figure.

Also, IMFD is designed for interplanetary trips. It's capable of a lunar trip - but that's not it's strong point and requires a different technique. In an interplanetary trip Target Intercept isn't calculating a course for your vessel so much as it's planning a course for the source planet (src is the body you are leaving - not your vessel) - and the course is plotted before launch. That means Surface Launch can get the TIn and calculate the departure plane.

For a lunar trip, the src is the vessel, and can't be plotted until you are in orbit. That means Surface Launch can't find out the TIn from Target Intercept, so you have to provide that information yourself.

Just another small thing: while an aligned plane Eject should be just a little more efficient than an Off-Plane, it seems that iV value is far greater in an Off-Plane,

This is one of the confusing things about IMFD. "iV" has different meanings depending on where it's used. In Burn Vector mode, or Delta Velocity, it means "inward velocity", more or less like it does in TransX.

On Target Intercept's main page, however, "iV" means "inherent velocity", and is basically the relative velocity you will have with your target body at intercept. An off-plane transfer wil have a higher RVel at intercept than an "on plane" transfer. If you know the velocity of your desired orbit around the target body, you can use that and iV to predict the dV required for the orbit insertion.

IMFD is best used to plan a course BEFORE you launch. It calculates the ejection plane based on the departure plane. TransX will allow you to "tailor" the departure plane based on your current plane. While that results in a "less than optimal" trajectory, it will be more efficient overall than altering your orbital plane to match the "optimal" ejection plane.

 

[orbekler]

Tommy, for TIn I suppose only the creator of IMFD can aswer
About iV, it seems you didn't watch the pictures. Forget about Burn Vector inward V: we already discussed the doubt (asked by me) and Jarmonik sealed it. I'm talking about inherent V. If you neglet this component (as TransX do), some people could believe that a "perfect" off-plane (let's say 90°) is spending the same energy as a coplanar transfer. But of course, it's not like that. That's all I wanted to say.