Delta-V Discussion

[GoSlash27]

I'm not knocking the map itself. AFAIK it's just about exactly right. But in my experience so far, it represents a best scenario rather than a worst or average, so I look at it as a rough guide rather than a number to hang my hat on.

A more realistic number IMO would be one where it assumes an average phase angle, average alignment of apoapsis, and a worst case ascending node so you're doing a Hohmann transfer orbit and need to correct at periapsis say 50% for an intercept on the next orbit.

[Superfluous J]

I'm not knocking the map itself. AFAIK it's just about exactly right. But in my experience so far, it represents a best scenario rather than a worst or average, so I look at it as a rough guide rather than a number to hang my hat on.

A more realistic number IMO would be one where it assumes an average phase angle, average alignment of apoapsis, and a worst case ascending node so you're doing a Hohmann transfer orbit and need to correct at periapsis say 50% for an intercept on the next orbit.

Well one thing is... most of us don't do that. At least, I don't. I will do it when going to Moho, Dres or Eeloo because their orbits are quite tilted, they have no atmospheres, and are small enough that Hohmann won't help you much when you're screaming in at multiple km/s anyway. But for hitting Eve, Duna or Jool (or Kerbin on the way back) there's no good reason* to not just aim to hit the planet directly and let the atmosphere take care of most of the slowing down. You can do all that with a couple dozen dV in almost every case.

*Except if you're using a mod like Deadly ReEntry. But if you're going to toss mods into the mix you're rendering any general discussion moot.

[GoSlash27]

Most of us *try* not to do that. Myself included. But in reality unless you're planning on everything going right, the maps are too optimistic. You can get most places that way with less DV than the maps suggest, but you don't build a rocket based on the assumption that everything will go right. That's the point of DV maps.

[Claw]

Most of us *try* not to do that. Myself included. But in reality unless you're planning on everything going right, the maps are too optimistic. You can get most places that way with less DV than the maps suggest, but you don't build a rocket based on the assumption that everything will go right. That's the point of DV maps.

True. But I think I would rather have the optimum chart, then build in MY safety factors.

I would rather have the optimized numbers so I know the minimum requirements to do a straight transfer. Then I can add a margin that I feel is right for the situation. If I start going over those numbers too much, I know I need to modify my profile.

Heck, for probes I don't always build a safety factor at all. Or it'll be a factor based purely on rounding and available fuel tank size without regard to a minimum margin. Kerbaled missions, on the other hand, are another story. Spare dV there for mom and the kids, jettisonable parts, maybe even prepositoned rescue vehicles.

If the dV map included someone elses averages, or fudge factors, or hellacious gravity sling shots and deadly aerobrakes, it would be harder to keep track of all of the assumptions built into the map design. To me that's way more confusing for a new player. The dV map should be a quick, simple look to help in planning.

So to me the point of the dV map should be the plan for everything going right. Then it's up to me to decide where I want to pad the numbers, and how much risk I'm willing to accept.

[GoSlash27]

True. But I think I would rather have the optimum chart, then build in MY safety factors.

Yeah, I gotcha. Whereas I would prefer a chart that shows "realistically feasible" numbers instead of "technically possible".

YMMV

[cantab]

A near-ideal Hohmann transfer is "realistically feasible" if you leave on the right date. IMHO a map should take into account variations from eccentricity and inclination (to my knowledge the former is easy to factor in but the latter is not), but should not start putting in its own safety factors. Conversely it shouldn't give figures based on complicated trajectories like bi-elliptics or gravity assists.

[GoSlash27]

A near-ideal Hohmann transfer is "realistically feasible" if you leave on the right date. IMHO a map should take into account variations from eccentricity and inclination (to my knowledge the former is easy to factor in but the latter is not), but should not start putting in its own safety factors. Conversely it shouldn't give figures based on complicated trajectories like bi-elliptics or gravity assists.

I agree with all of this. The variations of eccentricity and inclination could be included by simply projecting the worst case scenarios. For example, a DV map estimate from Kerbin to Eve would assume a Hohmann transfer from Kerbin's apoapsis to Eve's periapsis and the steepest possible inclination change, which would be the sum of their inclination.

[Claw]

A near-ideal Hohmann transfer is "realistically feasible" if you leave on the right date. IMHO a map should take into account variations from eccentricity and inclination (to my knowledge the former is easy to factor in but the latter is not), but should not start putting in its own safety factors. Conversely it shouldn't give figures based on complicated trajectories like bi-elliptics or gravity assists.

Yes, this is mostly what I said. I'm sorry Slashy only grabbed onto the "optimum" comment. My understanding was that a lot of the dV charts do in fact include the "worst case" inclination change. I would NEVER use a chart that included someone else's fudge/safety factors. And I agree (and think I said) it shouldn't include crazy trajectories.

Edit: Yeah, this...

If the dV map included someone elses averages, or fudge factors, or hellacious gravity sling shots and deadly aerobrakes, it would be harder to keep track of all of the assumptions built into the map design. To me that's way more confusing for a new player. The dV map should be a quick, simple look to help in planning.

Edit again...

After thing about all this, there really is no reason to disagree on what a dV map should look like. A dV map is built off of the assumptions that YOU want. So if you don't like the assumptions, then change it up how you want! There's ne need for me to convince you of "my way."

Edited June 17, 2014 by Claw

[Specialist290]

This may be of interest for y'all: While I was checking for new resources for the Drawing Board, I came across a delta-v map for "simple flight paths" -- basically, rather than relying on transfer windows, he put the map together on the assumption that the player would leave Kerbin's SoI and then plot the transfer from solar orbit. Don't know if it's what you're looking for specifically, but it sounds like it might be relevant to the topic.

Hope it proves useful

[GoSlash27]

This may be of interest for y'all: While I was checking for new resources for the Drawing Board, I came across a delta-v map for "simple flight paths" -- basically, rather than relying on transfer windows, he put the map together on the assumption that the player would leave Kerbin's SoI and then plot the transfer from solar orbit. Don't know if it's what you're looking for specifically, but it sounds like it might be relevant to the topic.

Hope it proves useful

Thanks! These numbers seem like a pretty good ballpark from what I've seen.

[Endersmens]

Any ideas on what i should do now?

[GoSlash27]

Any ideas on what i should do now?

I'm thinkin' you should define the precise circumstances that apply to your chart. Are we talking a burn from LKO for a direct intercept using add-ons? Will we be using a vanilla installation and shooting for an intercept from Kerbal escape at an ascending or descending node instead of a full plane correction at apoapsis/ periapsis? Will we be assuming a missed intercept with a 1 orbit correction?

These scenarios create wildly divergent DV budgets and they can't be factored for. You can't simply take a baseline DV budget and alter it by a certain percentage to cover a different set of assumptions, so step 1 would be to lay out what the assumptions are.

Best,

-Slashy

[Claw]

Yeah, I agree with Slashy. Assumptions are what we had the big debate over. So you're going to have to nail those down before anything else.

And decide what are the important parameters to you. dV and transfer time seem pretty obvious, but what about time for 1 orbit at the planet/moon at the assumed altitude? That's the sort of thing that would effect TAC. So just depends on what you want.

[Endersmens]

Well. A lot of new players use 75-100km for their orbits, i still do. So lets start from there. The burn will take place from a 100km orbit.

Next, Lets shoot for a direct intercept, as most new players (even me, and i'm decently seasoned) dont wanna bounce aroudn the solar system for hours.

As for vanilla, the only thing that would really affect this chart is making sure you have the vanilla solar system, different parts wont really change the delta-v Requirements. They might change fuel requirements though. But that doesn't really matter here.

Lets also assume we are going for a straight burn from 100km orbit with a plane correction in deep space if needed. I still don't really understand how to eject right to neglegate a plane change, and i'm pretty far away from being a newb. Plane changes are easier, even if it means packing more delta-v.

Phase angle is important here, to find the experience given optimum phase angle, for intance, someone could give an example using a phase angle given from a chart, or if you're like me, my jool transfer used the Kerbal alarm clock equation phase angle. (ill check on that later and report back one what it was exactly)

And i guess the placement of the manuever node on the LKO orbit would be important too, although i'm not sure how to document that. maybe angle in relation to the sun?

Finally, lets assume the burn is pretty much straight prograde. That will keep the deep space plane changes similar.

Any other parameters are welcome to be suggested, so feel free to tell me anything you think should be a paramater, or should not be parameter. Also, try to think like a new player when giving these.

[Red Iron Crown]

If you want to correct for different inclinations at the initial burn, you have to add a normal or antinormal component to the burn. I don't bother unless it's a large plane difference, it's easier to correct along the way.

Placement of the maneuver node is measured by "angle to prograde", where prograde is the direction of the planet's orbit around the sun. So you were on the right track by thinking to measure it from the sun, but the convention is 90 degrees different. KER and MechJeb will give you this measurement directly, and I think PreciseNode lets you place it at a specified angle to prograde, too. In stock you can zoom out the map until you can see Kerbin's orbit and eyeball it.

[Endersmens]

Ah. Thank you for clearing that up. I suspect KER will be necesarry to get all the information i need for this chart, good thing i have it. Is there a mod that will tell you your current phase angle to the target planet? I know kerbal alarm clock will tell you which phase angle to use, and it gives two different ones (physical model and equation) So im thinking somewhere in the middle is probably a good average. Does VOID or KER of Mechjeb tell you phase angle? Or does anything else?

[Red Iron Crown]

Both KER and MechJeb can provide phase angle. There's another mod, Protractor, which does it too.

Edited June 25, 2014 by Red Iron Crown
KR? It's KER, stupid phone.

[Endersmens]

*smacks forehead* of course! Protractor! Duh! How could I have forgotten that?

This is feeling much more like a project now.

Thanks for those helping me out! (You'll probably hear that a lot, I like to thank people )

Edited June 25, 2014 by Endersmens

[GoSlash27]

But of course... if you're using Protractor/ MechJeb, etc. it's no longer a vanilla installation. I generally break my intercept into 3 phases for that reason; I have no way to accurately plan the entire burn from LKO.

I wait for a window that looks close (either proper phase or inclination node alignment), shoot for an escape burn from Kerbin SOI, then perform my intercept burn.

Definitely less efficient than a single honk from LKO.

Best,

-Slashy

[Kryxal]

You don't need to do an escape burn then a burn outside the SoI ... just use the calculator to figure out when and take your best guess at where the node should be, then add in the delta-v dictated and see where that puts you, and refine from there. Maneuver nodes are handy, you can figure out the burn without doing it, though it's tough with low TWR...

[Pecan]

But of course... if you're using Protractor/ MechJeb, etc. it's no longer a vanilla installation. I generally break my intercept into 3 phases for that reason; I have no way to accurately plan the entire burn from LKO.

FIFY.

(A dig at not having data available in stock KSP, not a dig at you)

[Endersmens]

Agreed, i know the devs want to keep the game "fun" and i get that but somewhere in the tech tree, a jebulator would be nice. That way it could be "fun" until it gets serious enough to make calculations

[Claw]

Hmm. I think I am inclined to agree about the dV calculations. While it was initially fun to build a spreadsheet and calculate some dV for my ship, it lost some appeal after a while. Especially when the build starts getting complicated.

I did find it fun initially to flail around a bit, attempting to build rockets with enough fuel to get anywhere. But I certainly felt like I needed to understand dV calculations a bit better to actually design with a purpose. But that's just me.

While I proved to myself that I CAN do the calculations, I find it more enjoyable to spend my time on other aspects of my gameplay and let MJ sort out my required math. I think if this data were available in stock, I would uninstall MJ. (Although the auto-ascent is nice after the 10th launch of the day. )

[Endersmens]

Yes, it is nice to hit launch and go make a sandwich. Instead of tedious keeping the rocket on course for 10 minutes (due to 9 fps)

But yes, like I said it is fun to design blindly to a certain extent. Until you have a mission to duna that takes many days of real time and being 6 m/s delta-v short on the return.

[cantab]

You don't need to do an escape burn then a burn outside the SoI ... just use the calculator to figure out when and take your best guess at where the node should be, then add in the delta-v dictated and see where that puts you, and refine from there. Maneuver nodes are handy, you can figure out the burn without doing it, though it's tough with low TWR...

Agreed. You can eyeball the angles reasonably well. Or hold an actual protractor against the screen.