Vertical Ascent vs. To LXO First

[arkie87]

I calculated your DV expenditure from the mass.

42.6 tonnes on the pad and 10.7 tonnes at burnout. Looks like a mainsail to me. Please correct me if any of this is incorrect.

KER said 47.428 ton, but I summed parts manually in spreadsheet and came out to only 46.670 ton. Either way, 42.6 ton is incorrect.

Similarly, KER said 15.458 ton at burnout, but i calculated 14.67 ton. Don't know where you got 10.7 ton...

Your prograde launch was 4,880 m/sec DV. Minus the correction for overshoot.

I think your number here is incorrect because you used incorrect masses. By my calculations, using ISP 320 (at sea level for mainsail), i get 3633 m/s deltaV in lifter stage and using ISP = 360 (vaccum for mainsail), i get 4087 m/s deltaV. Since i only have approx 7 m/s deltaV left for LKO-to-mun burn, lets be conservative and assume, correct deltaV spent is the total of 4087 m/s...

I think the values you used were incorrect...

For vertical, it was the same plus your stage 2 burn. I couldn't get the type of engine. Looked like an LV-T30, but not sure.

Yes, it is a LV-T30 engine to maximize TWR

You were 4,900 m/sec on a mission that should've only been 4,270 for FAR.

So yeah, that's what I mean by "drag- monster". If you've added over 600 m/sec due to inefficiency, *of course* it'll bury some of the 300+ m/ sec you waste by going vertical.

Inefficient rocket is inefficient no matter which way you point it.

Your conclusions here are misinformed since you used incorrect numbers. See above.

Also the point that KER misled you about how much DV you had actually used, just as it did about your terminal velocity.

-Slashy

I didnt use KER for terminal velocity--only FAR.

I admit KER's numbers seem fishy, when it comes to mass and deltaV though....

[GoSlash27]

I could make the same argument for you that you are just arguing and not listening. We obviously both think we are right... But the difference is, when i argue, i make an argument with math or logic rather than claim the other side is wrong. Claiming the other side is wrong without supporting evidence is pointless...

From my side, it seems you arent listening either. You repeatedly use arguments that i've argued apply to stock aero only, and not FAR, even after ive explained why and how stock and FAR are different. Furthermore, you've never addressed the terminal velocity argument other than claiming its wrong without providing a reason.

Please follow community rules, as per B787_300.

And the fact that my rocket is glowing from re-entry effects on liftoff says nothing. You still haven't refuted my claim that terminal velocity is what matters-- not rerentry effects.

Look, this is really simple...

You *claim* to be "seeking answers", but all the while you're "highlighting advantages", which isn't asking questions, but rather making statements.

This is what they refer to as a "rhetorical question", where you ask a question not wanting an answer, but rather as a pretext for making an argument.

I got sucked into this because I was genuinely trying to answer your question, but you don't seem to be willing to listen to the answer because you're too busy trying to promote your argument.

This is highly exasperating. If you want an answer to a question, then listen when people tell you. If you don't want an answer, then don't ask.

Best,

-Slashy

[arkie87]

Look, this is really simple...

You *claim* to be "seeking answers", but all the while you're "highlighting advantages", which isn't asking questions, but rather making statements.

This is what they refer to as a "rhetorical question", where you ask a question not wanting an answer, but rather as a pretext for making an argument.

I got sucked into this because I was genuinely trying to answer your question, but you don't seem to be willing to listen to the answer because you're too busy trying to promote your argument.

This is highly exasperating. If you want an answer to a question, then listen when people tell you. If you don't want an answer, then don't ask.

Best,

-Slashy

Can you please redo your calculation with the corrected numbers.

I think we are finally close to resolving this. Don't tap out now.

It is exasperating on my side too

Here is my spreadsheet:

[GoSlash27]

KER said 47.428 ton, but I summed parts manually in spreadsheet and came out to only 46.670 ton. Either way, 42.6 ton is incorrect.

Similarly, KER said 15.458 ton at burnout, but i calculated 14.67 ton. Don't know where you got 10.7 ton...

I think your number here is incorrect because you used incorrect masses. By my calculations, using ISP 320 (at sea level for mainsail), i get 3633 m/s deltaV in lifter stage and using ISP = 360 (vaccum for mainsail), i get 4087 m/s deltaV. Since i only have approx 7 m/s deltaV left for LKO-to-mun burn, lets be conservative and assume, correct deltaV spent is the total of 4087 m/s...

I think the values you used were incorrect...

Yes, it is a LV-T30 engine to maximize TWR

Your conclusions here are misinformed since you used incorrect numbers. See above.

I didnt use KER for terminal velocity--only FAR.

I admit KER's numbers seem fishy, when it comes to mass and deltaV though....

3 different answers. I got the figures from the video in the lower right hand corner where it said "total mass".

And you certainly wouldn't use sea level Isp figures for that. They don't apply on Kerbin above 5Km or so. vacuum is more accurate.

You seemed more than happy to use the numbers that add- on generated before, now all of a sudden they're "fishy"??

Dubious,

-Slashy

[arkie87]

Look, this is really simple...

You *claim* to be "seeking answers", but all the while you're "highlighting advantages", which isn't asking questions, but rather making statements.

This is what they refer to as a "rhetorical question", where you ask a question not wanting an answer, but rather as a pretext for making an argument.

I got sucked into this because I was genuinely trying to answer your question, but you don't seem to be willing to listen to the answer because you're too busy trying to promote your argument.

This is highly exasperating. If you want an answer to a question, then listen when people tell you. If you don't want an answer, then don't ask.

Best,

-Slashy

I am making statements that appear to be valid to me based on the math and physics i know. I am seeking mathematical/specific rebuttals if I am wrong. Could you expect any different from a scientist?

I appreciate you genuinely trying to answer the questions (though i would prefer less rude responses sometimes ). Let's finish up these calculations. I think we are close to resolving this...

[arkie87]

3 different answers. I got the figures from the video in the lower right hand corner where it said "total mass".

And you certainly wouldn't use sea level Isp figures for that. They don't apply on Kerbin above 5Km or so. vacuum is more accurate.

You seemed more than happy to use the numbers that add- on generated before, now all of a sudden they're "fishy"??

Dubious,

-Slashy

You can use vacuum ISP. That is what i used, though real value is slightly smaller than that.

I trusted those values before we got into this discussion, but after investigating and using a spreadsheet (as per your recommendation in another thread), those masses dont make sense to me. Maybe the add on has updated part masses or knows something I dont...

Anyway, there are two numbers for mass reported in KER: one is current-stage only value, and one is total value. I think you used stage instead of total. Can you please redo your calculations...

[GoSlash27]

Going by the numbers on that spreadsheet, 4,087 m/sec is reasonably accurate. Not enough variance to make a fuss over, since I only use 3 digits of resolution.

I couldn't tell how many m/sec you removed due to the overshoot, so I can't confirm what the final DV would've been.

For the vertical shot, it would be the 4,087 m/sec plus a little of the second stage, but since the KER numbers are suspect, I don't know how much. It said stage 0 weighed 3.54 tonnes, which is also incorrect according to your spreadsheet.

Best,

-Slashy

Edited December 15, 2014 by GoSlash27

[arkie87]

Going by the numbers on that spreadsheet, 4,087 m/sec is reasonably accurate.

Best,

-Slashy

So then you admit this conclusion is no longer applicable, and my rocket does not produce excess drag...:

I calculated your DV expenditure from the mass.

You were 4,900 m/sec on a mission that should've only been 4,270 for FAR.

So yeah, that's what I mean by "drag- monster". If you've added over 600 m/sec due to inefficiency, *of course* it'll bury some of the 300+ m/ sec you waste by going vertical.

Inefficient rocket is inefficient no matter which way you point it.

-Slashy

If that is the case, then my rocket saves ~100 m/s compared to the nominal value required by FAR for this task....

[arkie87]

Not enough variance to make a fuss over, since I only use 3 digits of resolution.

Best,

-Slashy

Yes, of course. If you were to start arguing that it's actually 4086.7772342343 instead of 4087, i would lose a bit of respect for you

But my main point is that your past conclusion that my craft is inefficient because it creates excess drag is not supported by this deltaV calculation.

Though if you would like to still argue that, I am open to hearing arguments...

[GoSlash27]

I am, actually.

I no longer have any confidence in any numbers you're posting here. How do you know if your spreadsheet is right, or if FAR is right, or if KER is right?

*My* numbers from my test are 100% solid. No questions, no variations.

If you really want to *know* you should open this up for peer review.

Post your craft file and have somebody else who uses FAR confirm whether your rocket is *truly* more efficient than the norm or less.

I'm thinkin' it's a lot less and your numbers are gacked.

Best,

-Slashy

Edited December 15, 2014 by GoSlash27

[arkie87]

For the vertical shot, it would be the 4,087 m/sec plus a little of the second stage, but since the KER numbers are suspect, I don't know how much. It said stage 0 weighed 3.54 tonnes, which is also incorrect according to your spreadsheet.

Best,

-Slashy

This is true. KER says about 70 m/s... but who knows... but it is about right on target... as i did not use too much of the upper stage...

Regardless of the vertical ascent accuracy, your main argument was that the LKO-to-Mun flight was flawed due to it wasting too much energy on atmospheric drag. But given that it comes in below nominal value, that clearly isnt the case, right?

[Kesa]

Well, you seem to be at half the drag of terminal velocity, which is very high, but not dramatically high.

With FAR, Kerbin not rescaled, and enough booster, Kerbin seems to act like a small moon with thin atmosphere, validating the "point at your targeted velocity" strategy, which is different than the verticale ascent. Even though, you proved the verticale way isn't the best with a very poor fly path. The difference would be greater if you flyed correctly a well designed rocket. Since I don't use FAR, I won't do any practical fly, but I will try the Mathematical way.

For now, what I can say is that gravity loss of verticale ascent is : g*time of burn (~820m/s in your video).

Also, your craft is actually using far less dv than usual rockets (it could use even less), that's because of high TWR. But the counterpart is that engines weight more compared to fuel and your rocket has less DV. There could really be a valid tradeoff. Anyway, this tradeoff is not between verticle and horizontal ascent, but between direct High TWR ascent and orbital insertion low TWR ascent

Edited December 15, 2014 by Kesa

[arkie87]

I am, actually.

I no longer have any confidence in any numbers you're posting here. How do you know if your spreadsheet is right, or if FAR is right, or if KER is right?

*My* numbers from my test are 100% solid. No questions, no variations.

I assume KER is either using updated part masses, or is including weight of RCS fuel in command pod, or something else i am not aware of.... either way, the numbers are not far off...

In any case, you can assemble the same craft that you see in the video and/or the spreadsheet and check the numbers for the LKO-to-Mun case assuming that i burned out the first stage completely (which is approximately true), and let me know what deltaV you get...

[GoSlash27]

I am making statements that appear to be valid to me based on the math and physics i know. I am seeking mathematical/specific rebuttals if I am wrong. Could you expect any different from a scientist?

Going into an experiment with a bias towards a specific outcome and coloring the results isn't "science", it's "lobbying".

I don't personally have a dog in this fight one way or the other. If somebody tells me that launching rockets down the toilet will get my payload to orbit more efficiently and I can confirm it, then I'll happily launch rockets down toilets.

But you're not testing this theory, you're *selling* it. And from everything I can tell (including my own testing), it's wrong.

Best,

-Slashy

[arkie87]

Well, you seem to be at half the drag of terminal velocity, which is very high, but not dramatically high.

It might be high enough to cause aerodynamic disassembly, but I have that disabled

With FAR, Kerbin not rescaled, and enough booster, Kerbin seems to act like a small moon with thin atmosphere, validating the "point at your targeted velocity" strategy, which is different than the verticale ascent. Even though, you proved the verticale way isn't the best with a very poor fly path. The difference would be greater if you flyed correctly a well designed rocket. Since I don't use FAR, I won't do any practical fly, but I will try the Mathematical way.

I'm not sure if GoSlashy would agree with this, and Im not sure I do either... I think it depends on the case...

For now, what I can say is that gravity loss of verticale ascent is : g*time of burn (~820m/s in your video).

Yes, a similar calculation can be done for LKO-to-Mun ascent, and i think you will find it is significant, though not as large since eventually centripetal forces reduce effective g forces... meanwhile, vertical ascent does not get far enough away from the planet to significantly decrease parasitic gravity losses.

Also, your craft is actually using far less dv than usual rockets (it could use even less), that's because of high TWR. But the counterpart is that engines weight more compared to fuel and your rocket has less DV. There could really be a valid tradeoff. Anyway, this tradeoff is not between verticle and horizontal ascent, but between direct High TWR ascent and orbital insertion low TWR ascent

I agree 100% and this is my point. In practice, you could and maybe should substitute mainsail for a skipper, in order to get more deltaV from the lighter engine at expense of TWR (though reduced TWR also costs deltaV).

EDIT: by the way, the only reason i group it as "LKO-to-Mun" and "vertical ascent" and not low TWR vs high TWR is because with low TWR, LKO-to-mun is easy to perform due to use of thrust-vectoring engines, which allow for much more control during ascent and gravity turn; high TWR seem to lend themselves to favor boosters, which do not have thrust vectoring, and would be harder to steer into a gravity turn, thus, they favor vertical ascent.

Edited December 15, 2014 by arkie87

[arkie87]

Going into an experiment with a bias towards a specific outcome and coloring the results isn't "science", it's "lobbying".

I don't personally have a dog in this fight one way or the other. If somebody tells me that launching rockets down the toilet will get my payload to orbit more efficiently and I can confirm it, then I'll happily launch rockets down toilets.

I am testing the theory, with videos and math-- as much as you expect me to--and I'm trying to sell it, since it seems correct to me. How else would you like me to "test it"???

Anyway, why are we talking about this and not the issue at hand????

But you're not testing this theory, you're *selling* it. And from everything I can tell (including my own testing), it's wrong.

Best,

-Slashy

Your own testing is with stock aerodynamics, not FAR. Install FAR and test it if that's what you need to do...

And once again, you assert it's wrong without giving a reason, even though we have clear and present evidence that it actually reduces deltaV requirements than the nominal ones required by FAR, and not vice versa...

Edited December 15, 2014 by arkie87

[GoSlash27]

I assume KER is either using updated part masses, or is including weight of RCS fuel in command pod, or something else i am not aware of.... either way, the numbers are not far off...

In any case, you can assemble the same craft that you see in the video and/or the spreadsheet and check the numbers for the LKO-to-Mun case assuming that i burned out the first stage completely (which is approximately true), and let me know what deltaV you get...

This wouldn't answer anything, since I'm not running FAR. You really need somebody who uses FAR to test it.

Best,

-Slashy

Edited December 15, 2014 by GoSlash27

[GoSlash27]

Kerbin seems to act like a small moon with thin atmosphere, validating the "point at your targeted velocity" strategy, which is different than the verticale ascent.

Actually, this was a point I confirmed today with the folks at FAR: Kerbin "acts" as it has always acted. FAR doesn't change any of it's properties. Same gravity, rotation, radius, surface pressure, and scale altitude.

Kerbin doesn't act any different at all. It's your rocket that acts different.

This is a critical point.

Best,

-Slashy

Edited December 15, 2014 by GoSlash27

[GoSlash27]

I am testing the theory, with videos and math-- as much as you expect me to--and I'm trying to sell it, since it seems correct to me. How else would you like me to "test it"???

Anyway, why are we talking about this and not the issue at hand????

Your own testing is with stock aerodynamics, not FAR. Install FAR and test it if that's what you need to do...

And once again, you assert it's wrong without giving a reason, even though we have clear and present evidence that it actually reduces deltaV requirements than the nominal ones required by FAR, and not vice versa...

#1, we don't have any "clear and present evidence" of anything at all, other than you don't know what your numbers are, but are attempting to draw conclusions from them anyway. Specifically, the conclusions you prefer. That's just plain bad science.

#2, I have already given you a clear and detailed explanation of why your reasoning is flawed, but it's in one of the several other threads you've posted on this topic.

Edited December 15, 2014 by GoSlash27

[arkie87]

This wouldn't answer anything, since I'm not running FAR. You really need somebody who uses FAR to test it.

Best,

-Slashy

Yes, and that clearly isnt you, so i suppose there is no point in continuing this debate further, sad to say

[Kesa]

Yellow arrow is trust/mass, red is grvitationnal acceleration (wheight/mass), green is acceleration (sum of the two other)

If you are accelerating upward (first picture), your acceleration will be :

a = g(TWR - 1)

therefore, gravity loss per second is :

trust acceleration - real accelleration = g * TWR - g*(TWR - 1) = g

which represents 1/TWR of your current dv expend (loss/s divided by trust). You're fighting against gravity.

For example, a starter stage with a low TWR of 1.1 will waste 10/11 of its DV fighting gravity, while 2 TWR starter stage will only waste half of dv (but more du to drag, if any), and craft with TWR = 6 will waste 1/6 of their fuel, and so on.

If you're accelerating in such a way that your verticale speed is constant (pic 2), using the theorem of pythagore, you have :

a = g*sqrt(TWR^2 - 1)

Therefore gravity loss per second is :

g * (TWR - sqrt(TWR^2 - 1)) < g :

Which represents 1 - sqrt(1 - TWR^-2) of your dv : you're less fighting with gravity.

For examples, a mid stage booster with TWR = 1.2, it is 45% of dv (you need far lower TWR to achieve same acceleration as verticale ascend). For TWR = 1.4, it is as low as 29% of the dv , if TWR = 2, ~13% of your dv, making it competitive with the 6 TWR monster on verticale ascend.

One can also prove that

1/(2*TWR^2) < 1 - sqrt(1 - TWR^-2) < 1/(2*TWR^2) + 1/(8*TWR^4)

So loss fraction = 1/(2*TWR^2) is a good approximation.

So loss fraction not only is way better for inclined burn, it also decreases much faster as TWR grows, (therefore requiring lower TWR)

Finally, when you build up horizontale speed, you are granted a centrifuge force (blue arrow, pic 3). I won't express it with gravitational parm, but as it grows like a square, if we call vh the horizontale velocity and v0 the orbital velocity at current altitude (eg 2300m/s for kerbin), we can write it :

g * (vh^2/v0^2)

Loss fraction may be calculated as above by replacing the gravity by gravity minus this force, effectively multiplying TWR in above formula by 1/(1-v^2/v0^2), which is 9/8 over kerbin at 800 m/s, 4/3 at 1150m/s, up to infite when reaching orbital velocity (you're no more suffering gravity loss).

To sum up gravity loss are by far greater while accelerating vertically. Even with non trust-vectoring engines, you will save huge amount of mass by adding some control surfaces to make an early enough gravity turn. If you like great acclerations, you should ask if a TWR > 2 (arbitrary high but resonable value) worth it : will you really lose less than 13% of DV by adding more boosters?

Of course, this is for a heavy body ignoring drag which is reasonable at some altitudes and speeds, and which is always reasonable for kerbin with FAR, since it seems to make drag comparable to stock Duna. If you are going out of a mun with high TWR and low dv requirement, any ascent, may it be verticale, will cost very few dv compared to optimal path.

Edit : fixed the image. btw how do you make spoiler?

Edited December 15, 2014 by Kesa

[arkie87]

FYI: Your imgur gallery didnt work.

Very interesting work. I need to review in detail. Looks promising/poignant though. Thanks for posting!

[Kesa]

Actually, this was a point I confirmed today with the folks at FAR: Kerbin "acts" as it has always acted. FAR doesn't change any of it's properties. Same gravity, rotation, radius, surface pressure, and scale altitude.

Kerbin doesn't act any different at all. It's your rocket that acts different.

This is a critical point.

Best,

-Slashy

Only a matter of point of view, since we could say this is the kerbin air that acts differently on the craft

If that kind of subtle nuancing really does matter, you might want to see :

Anyway I only meant that with far installed, launching from kerbin seems as hard as launching from Tylo with a duna atmosphere.

[Superfluous J]

Arkie, your video demonstration is not valid, because getting 2 apoapses at the Mun without getting the Periapses at the same place means your orbits have different energies.

Create a ship and fly it vertically in a reliable way (ie, you can hit it every time reliably after you're done perfecting your process) to get into - say - a 30km orbit over Mun from KSC with a certain payload. You don't have to use Procedural Fairings if you don't want to, but you should build the payload to reasonably fit under such fairings. I say, make a lander that can return to Kerbin on its own from a 30km orbit (from that orbit, land, take off, and return to Kerbin).

Once you've done that and posted the stats and craft file, I'll take that craft file to get the payload, and then build my own craft to get into that 30km orbit with a smaller, cheaper ship. It will use less fuel, and be cheaper. It may only save 10% of the total dV but because it's smaller and cheaper it will save a lot more than 10% of the fuel and money.

Until you can confidently state the same (and pull it off) your entire premise is moot. Saving dV isn't enough. "Only" losing a little dV isn't enough. You need to save FUEL and/or MONEY or consider the extra FUEL or MONEY expenditure to be okay.

And that will always be an opinion. The fact will remain that getting into orbit first with a craft designed to do so will always save fuel and money over a craft designed to burn straight up.

[arkie87]

Arkie, your video demonstration is not valid, because getting 2 apoapses at the Mun without getting the Periapses at the same place means your orbits have different energies.

Create a ship and fly it vertically in a reliable way (ie, you can hit it every time reliably after you're done perfecting your process) to get into - say - a 30km orbit over Mun from KSC with a certain payload. You don't have to use Procedural Fairings if you don't want to, but you should build the payload to reasonably fit under such fairings. I say, make a lander that can return to Kerbin on its own from a 30km orbit (from that orbit, land, take off, and return to Kerbin).

Once you've done that and posted the stats and craft file, I'll take that craft file to get the payload, and then build my own craft to get into that 30km orbit with a smaller, cheaper ship. It will use less fuel, and be cheaper. It may only save 10% of the total dV but because it's smaller and cheaper it will save a lot more than 10% of the fuel and money.

Until you can confidently state the same (and pull it off) your entire premise is moot. Saving dV isn't enough. "Only" losing a little dV isn't enough. You need to save FUEL and/or MONEY or consider the extra FUEL or MONEY expenditure to be okay.

And that will always be an opinion. The fact will remain that getting into orbit first with a craft designed to do so will always save fuel and money over a craft designed to burn straight up.

I like this approach, since the only thing that matters is in career mode is the Kerbucks required to get a certain package into orbit-- and not its deltaV, efficiency, etc...

I think this is the crux of my argument, so I'm glad you proposed this test.

My proposal is as follows:

Build the cheapest craft-- in terms of Kerbucks-- to get the upper stage into orbit around the Mun with FAR aerodynamics installed (if you dont use FAR, you will lose).

Upper stage consists of:

Mk16 parachute, mk1 command pod, TR-18A stack decoupler, FL-T400 fuel tank, and LV-T30.

I will try tonight. I look forward to see your results.