Acceleration in atmosphere question.

[mellester]

Got a question for you all. I design my stages to not exceed 22 m/s*s of acceleration below 40 km. Even above 40 km it is usually not needed unless I aim for a high kerbin orbit.
 I noticed that with higher acceleration I get reentry effects during ascend. Anyone have a different opinion.

[SSgt Baloo]

I, too, am interested in the answers to this question. Posting to be sure I'm subscribed to this discussion.

[Fearless Son]

That is pretty normal.  "Reentry" effects are just a specific circumstance of more generalized atmospheric friction.  If you start to see those effects on your craft, it means that you are going fast enough that drag is sapping some of your kinetic energy.  This can be a good thing if you are aerobreaking, a bad thing if your ship cannot take the stress of that much friction, and of course a complication if you are trying to ascend because it means that some of your thrust is being spent fighting drag instead of ascending.  

[Eric S]

That limit on acceleration would produce the desired results prior to 1.0 when we had an atmosphere that caused a lot more drag than we do currently.  However, since that has changed, you get a lot more actual acceleration out of your 22 m/s, meaning that you will go trans-sonic.

[diomedea]

A good rule of thumb was to never exceed terminal velocity (Vt) during the whole ascent. The math for Vt is actually easier by using ballistic coefficient (Bc), so it goes as Vt = SQRT(2*Bc*g/ρ). Bc actually changes with Mach speed, and g changes with altitude; but none as significantly as ρ (the atmospheric density) that really changes a lot with altitude, and is the prime reason why Vt increases the higher the vessel is. In case anybody wonders, KSP actually shows mach and reentry effects (flames) based on a quite similar equation.

Acceleration has no direct relation with the reentry effects, apart from being the cause for velocity changes with time. Managing acceleration during the whole ascent to always  keep a vessel speed below terminal velocity is generally one of the factors behind a perfectly efficient launch.

[mellester]

1 hour ago, diomedea said:

A good rule of thumb was to never exceed terminal velocity (Vt) during the whole ascent. The math for Vt is actually easier by using ballistic coefficient (Bc), so it goes as Vt = SQRT(2*Bc*g/ρ). Bc actually changes with Mach speed, and g changes with altitude; but none as significantly as ρ (the atmospheric density) that really changes a lot with altitude, and is the prime reason why Vt increases the higher the vessel is. In case anybody wonders, KSP actually shows mach and reentry effects (flames) based on a quite similar equation.

Acceleration has no direct relation with the reentry effects, apart from being the cause for velocity changes with time. Managing acceleration during the whole ascent to always  keep a vessel speed below terminal velocity is generally one of the factors behind a perfectly efficient launch.

not exceeding terminal vel is kind of hard. with a TWR of 1.8 you quickly exceed that velocity quickly . My designs give me a delta v of around 3500 m/s into a 100km orbit.

So I go faster than Vt usually. Anybody using designs with TWR higher than 2.2

[Sovek]

Ahhh, yes, this.

My typical approach is multi-stage and a low T:W off the pad, 1.25 at the most. I usually like my stages to burn for about 2.5-3 minutes, after that, stage and use a lower thrust engine. However, this can be a problem when the payload is very light and the majority of the fuel is in the first stage. Usually my early rockets suffer from this due to lack of engines to choose from.

I'll also lower the thrust output of the engine if it exceeds that 1.2, this can give the rocket expanded capabilities of heavier payloads (and more fuel in the payload) instead of having to strap some boosters to the thing.

[Seaview123]

Wouldn't the solution to this to be just to throttle down during the initial ascent? I noticed this using MechJeb and adjusting the "Q" setting on the Launch Guidance tab. A lower Q (dynamic pressure) setting causes MechJeb to throttle down farther, and results in a lower acceleration through atmo... until the air pressure at higher altitude decreases and full thrust isn't an issue any more.

 

[Fearless Son]

What Seaview123 said.  You want a big thrust-to-weight ratio on your ascent, but you generally want to throttle down once you build up a little speed to keep from exceeding terminal velocity and to not waste too much fuel too early in the launch.  I know solid rocket boosters are not actually throttle-able, but you should not be using them as your only source of thrust on your initial launch.  Think of them instead as helpers to give a bit of extra thrust on the first stage without having to spend tons of liquid fuel building up your speed.  

[Matuchkin]

6 hours ago, Fearless Son said:

"Reentry" effects are just a specific circumstance of more generalized atmospheric friction.

*atmospheric compression.

6 hours ago, mellester said:

 I noticed that with higher acceleration I get reentry effects during ascend.

The speed at which most objects reenter on Kerbin is at two-thousand-something meters per second. In real life, nothing would happen at that velocity except for mild mach effects, but Squad has to induce ultrasonic effects on reentering objects (I think you guys know why). Therefore, reentry effects start taking place at very low speeds. On Kerbin, 3000m/s is synonymous with 10,000m/s on Earth. The reason why that is unusual to everyone is because real-life rockets don't usually start flaing during launch.

Will ask for a mod to improve that problem. Probably Nathankell would agree to take care of that, as he's our awesome realism-guy.

[benjee10]

If you watch some videos of the shuttle's ET separation, you can actually see that the shuttle did in fact experience 'reentry' plasma effects on ascent. Admittedly they're a lot less pronounced than the flaming spears of death we get in KSP, but there is a real-life precedent for it since the atmosphere doesn't make any distinction between whether you're coming or going. 

[FullMetalMachinist]

As I understand it, once you get higher than 10km, the atmospheric pressure drops dramatically, as diomedea says. What that means is that even though you might see flaming effects on ascent, it doesn't necessarily mean that you've gone past terminal velocity, just that things are getting hot. The general rule of thumb that I've seen around the forums is to limit your speed (not acceleration) to less than mach 1 until you get above 10km. After that drag reduces drastically, and you can pretty much go as fast as possible. This is also the reason that a too-high periapsis when you're reentering can be bad, because there's not much drag to slow you down, but there is enough air present to cause you to heat up.

I tend to design my launch stages with a pad TWR of 1.2-1.4, which usually keeps me under mach 1 until about 12km. After that it's off to the races. This does roughly correspond to real life launches. According to the Wikipedia page on Max Q the space shuttle main engines were throttled back until it was at max Q at 11km. After that point the dynamic pressure dropped off as the atmosphere thinned and the shuttle was able to accelerate more rapidly.

[mellester]

4 hours ago, Sovek said:

Ahhh, yes, this.

My typical approach is multi-stage and a low T:W off the pad, 1.25 at the most. I usually like my stages to burn for about 2.5-3 minutes, after that, stage and use a lower thrust engine. However, this can be a problem when the payload is very light and the majority of the fuel is in the first stage. Usually my early rockets suffer from this due to lack of engines to choose from.

I'll also lower the thrust output of the engine if it exceeds that 1.2, this can give the rocket expanded capabilities of heavier payloads (and more fuel in the payload) instead of having to strap some boosters to the thing.

Interesting ill switch my designs for weaker engines and see if I can reduce the amount of fuel I need to get to orbit

[pincushionman]

8 hours ago, mellester said:

not exceeding terminal vel is kind of hard. with a TWR of 1.8 you quickly exceed that velocity quickly . My designs give me a delta v of around 3500 m/s into a 100km orbit.

So I go faster than Vt usually. Anybody using designs with TWR higher than 2.2

You have too much legs. If I recall, under the old souposphere you did want your TWR closer to 2 or so, but in the new system, if you're much above 1.4 you're going to accelerate into a brick wall of drag.

[Streetwind]

5 hours ago, pincushionman said:

You have too much legs. If I recall, under the old souposphere you did want your TWR closer to 2 or so, but in the new system, if you're much above 1.4 you're going to accelerate into a brick wall of drag.

Erm, I've kinda encountered precisely the opposite. Under the old system, you needed to severely limit your acceleration to not exceed terminal velocity. But under the new system, terminal velocity runs away from you very quickly, even if you burn hard. In fact, the current lowest-dV-to-Kerbin-orbit records are set by vehicles with TWRs exceeding 5 IIRC.

[rakutenshi]

I personally like to build my rockets a little bit lighter on the main lift stages as far as TWR goes, but I add SRB kickers aiming to add about 400 dV to it get it off the pad. Once you've reached a decent speed (varies per rocket, but it's usually transsonic-ish) you really don't need a massive TWR, if you're using FAR you should be able to pull up the flight data and look for 'q' .. atmospheric pressure, once that number starts to drop, if you've throttle back your engines to save dV losses to drag, you should be free to open it up.

But then again, I use radial SRBs just because I like watching them fall away. They're not fuel efficient, but they're usually a very very cost effective to give a rocket a kick in the pants off the pad.

[icedown]

My normal rockets start with a vacuum TWR of ~2.0.  I use MJ to limit acceleration through the entire launch to 18m/s². My launch profile is a 65% turn starting at 100m/s and ending level at 45k.  Orbital insertion is usually about 140m/s.  I've found this method to give a pretty solid 3,300 vacuum dV to orbit. When building my ships, I use the liquid engines to build the bulk of the dV needed, then supplement the TWR with SRBs, limiting their thrust so they show a TWR of 2.0 at launch.  This 2.0 is with both the first stage SRB and liquid engines burning together.  This setup allows you to throttle the liquids down with the SRBs running to maintain the 18m/s.  This pattern also keeps the ship pointing along the velocity vector up to around 25k, reducing drag and the chance of a tumble.  Keeping the balance between the engines is very important.  You need to be able to keep the ship at 18m/s for almost all of the profile.  Going over speed by having to much of the thrust as SRBs and/or ending up to slow after they shutdown because the liquids haven't burned off enough propellant to be able to run at 1.8TWR can be equally problematic. 

There are probably much more efficient ways, but I stick to this one because it provides a consistent dV and profile to orbit.  One other benefit that this provides comes when running stage recovery.  I build all of my sections with big expensive engines and parts to cut off about 2,900 dV.  This allows them to drop back in the atmosphere and be recovered relatively close to KSC, providing a pretty good return on those really expensive stages.  I launched a heavy science rover today that cost 323,000 to build and got about 80,000 of that back on the first stage recovery.

This launch profile works equally well with stock and FAR.

Edited January 8, 2016 by icedown

[Nefrums]

Having more TWR is almost always better.

One thing that saves a lot of dV is the initial acceleration of the pad, Getting to ~200 m/s fast helps.  

Also start the gravity turn right of the pad.  Just do an small initial turn to 85-80 right after launch how much depends on you twr. After that just hold prograde.  If you have leveled out to 30-40 degrees by the time you pass 10km the initial turn was right. 

The basic rule about the angle of climb is that it should be as low as possible while you are still be able to keep the AP in front while holding prograde.

On conventional rockets I aim to be at ~700 m/s at 10km going at about 30-35 degree climb.

With space planes I try to reach Mach 5 (~1500m/s) at 10km climbing 6-7 degrees

 

Edited January 8, 2016 by Nefrums

[cantab]

IMHO there's generally no reason to limit TWR on orbital launchers now KSP has vaguely realistic aero. Provided your rocket isn't insane you're best off going full power all the way, or at least until very late when you may throttle back to circularise.

On the other hand engines are heavy, so I tend not to put too much thrust on. A common design for me is a first stage with maybe 1.0 or less atmo TWR, SRBs, and an upper stage with around 0.5 TWR. At launch all engines ignite, the core sometimes being throttled back; by the time the boosters drop the fuel burn and improved thrust with height mean the core can come to full power and manage decent TWR. I've actually made orbit with the final 750-1000 m/s coming at as low as 0.25 TWR, but there's not much margin for error so I prefer more.

[Redhotita1]

Well it's true the fact you need a low TWR to get less aerodynamic drag, but you also need a high TWR to lift off.You have to fight generally two components while lifting off. Gravity and the Atmosphere. If you have a low TWR at lift off (i.e. 1.25) your rocket will take some time to take up some speed, meaning you'll burn alot of fuel just to make few kilometers (that is a common fact for heavy lifters, look at the Saturn V) while lighters rockets tends to escape the first part of the atmosphere quickly, such as the Vega, which has alot of TWR. Usually flights reach about 40m/s^2 I believe.

[John FX]

I tend to get plasma effects around 30km, probably due to my ascent profile and TWR.

Very rarely (usually on a pure solid booster launch) I get effects that last up to 50km during launch...

[mellester]

6 hours ago, icedown said:

Interesting so people have both  high twr and low twr designs. Perhaps higher a TWR gives a good accent for the more steeper turns. As in mechjeb 60% turns . And lower TWR gives a better use if you ascent with mechjebs 40%

[rakutenshi]

Just to point out an additional reason for staging like the Apollo rockets did. real rockets tend to have very little throttle control (assume this is to reduce weight as much as technology limits), so staging is a form of throttle control.

Look at the Saturn V, for example. about half to a third of the total launch mass is in the first stage, by the time the first stage burns out, the effective TWR of the stage has just about doubled, also air pressure has decreased, staging and using a different set of engines allows for total recalculation of ISP And TWR with the new mass, significantly increasing efficiency.

In KSP, rocket engines can be throttled from 0% to 100%, so having one large lift stage with a very high potential TWR is fully acceptable since you can literally adjust the throttle to be whatever TWR would be ideal for your current situation.

[Villageidiot]

21 hours ago, Fearless Son said:

What Seaview123 said.  You want a big thrust-to-weight ratio on your ascent, but you generally want to throttle down once you build up a little speed to keep from exceeding terminal velocity and to not waste too much fuel too early in the launch.  I know solid rocket boosters are not actually throttle-able, but you should not be using them as your only source of thrust on your initial launch.  Think of them instead as helpers to give a bit of extra thrust on the first stage without having to spend tons of liquid fuel building up your speed.  

Keep in mind that you can Rclick and decrease the thrust output on the SRBs. They may not be configurable in flight but I use them extensively to get the initial boost off the launch pad.

[Fearless Son]

14 minutes ago, Villageidiot said:

Keep in mind that you can Rclick and decrease the thrust output on the SRBs. They may not be configurable in flight but I use them extensively to get the initial boost off the launch pad.

Yeah, I used to decrease their output a lot in early career mode to get the most delta-v out of them, but I tend not to do that much these days.  Mostly because I find I want to maximize thrust from them to get my TWR up right off the pad, and if I have to scale their thrust down a bit it means I am using too many for my weight and should take some off.  

Turns out you can have too many boosters.