Launching with SRBs but not lighting centre engine

[ondofpond]

Done a search, can't get a thread on this specific topic. Been a KSP player for years, like efficiency and since career mode I build for cost, every time, even if I don't need to with excess funds. I'm not bad, but not perhaps the most discerning of method in my builds.

 

So - using KSP Engineer and looking at how much Delta-V I can get for my buck. On build using a Swivel for a two-stage centre, I put two or four BACC SRBs on, but stage so that I launch off the pad with SRBs only. Obviously ensuring I have enough TWR for it to be viable. If I have four SRBs and enough TWR I might light only two SRBs on the pad, drop and light the next two (performing a struggling gravity turn with risk of impact when dropping) before finally lighting my Swivel.

Obviously this makes for a slow initial ascent. I understand the thick atmosphere thing at the pad, get going quickly blah blah, but when I compare Delta-V stats with Atmosphere in Engineer, doing what I say above yields more Delta-V than if I light the Swivel on the pad, or light all four SRBs on the pad if using four. Again, TWR is considered and a slower start seems to be better on final Delta-V against cost.

I'm not arguing the virtues of using SRBs at all vs liquid side-booster stages, to me that's a no-brainer on cost (I can end up with 2.5m Skipper build with no Mainsail side boosters, just six Kickback SRBs using the above method, lighting four on the pad and lighting the other two before finally lighting the Skipper, when payload makes it viable).

What I'm wondering is, am I wrong to be purely looking at Delta-V values in Engineer in these builds? What might I not be considering that is possibly making the Delta-V figures either erroneous or a bad idea for another reason I'm not considering?

 

Edited March 30, 2017 by ondofpond

[Val]

This seems gameplay related. Moved from KSP Discussion

[cfds]

You may have to consider gravity and drag losses:

For every second you thrust against gravity you lose 9.81 m/s deltaV, so when you have low TWR you spend a lot of time thrusting vertically and so lose a lot of fuel to gravity. But if you go too fast you waste too much fuel on overcoming air-resistance.

[Wanderfound]

You'd get better ΔV if you built it so that it maintained the desired TWR (generally about 2 for a Kerbin launch) with all engines firing full throttle at once.

Throttling down or leaving engines off wastes ΔV by effectively increasing payload; the unused engine capacity becomes dead mass. Accelerating at more than TWR2 or thereabouts wastes ΔV to drag losses (this is variable depending on streamlining and flight profile, though).

Choose your core engine, add boosters until launch TWR gets to 1.5 or so, then stop. Ideally you want each stage to start at about TWR1.5 and end around TWR2.5. 

Anything else and you're carrying dead mass. The only exception to that is if the engine involved is ultra-inefficient at sea level (LV-N etc).

You can also easily squeeze a bit of extra ΔV out of that sort of design by activating fuel flow on the decouplers and sticking a small drop tank on top of each SRB. Do that, and you can burn the core engine from launch but still have it pretty much full when you discard the SRBs.

Edited March 30, 2017 by Wanderfound

[ondofpond]

11 minutes ago, Wanderfound said:

You'd get better ΔV if you built it so that it maintained the desired TWR (generally about 2 for a Kerbin launch) with all engines firing full throttle at once.

Throttling down or leaving engines off wastes ΔV by effectively increasing payload; the unused engine capacity becomes dead mass. Accelerating at more than TWR2 or thereabouts wastes ΔV to drag losses (this is variable depending on streamlining and flight profile, though).

Choose your core engine, add boosters until launch TWR gets to 1.5 or so, then stop. Ideally you want each stage to start at about TWR1.5 and end around TWR2.5. 

Anything else and you're carrying dead mass. The only exception to that is if the engine involved is ultra-inefficient at sea level (LV-N etc).

You can also easily squeeze a bit of extra ΔV out of that sort of design by activating fuel flow on the decouplers and sticking a small drop tank on top of each SRB. Do that, and you can burn the core engine from launch but still have it pretty much full when you discard the SRBs.

That's the sort of thing I'm talking about. Thank you.

I'm still not terribly clear on how the Engineer "Atmosphere" thing works. Does it calculate the whole atmospheric effect as your prospective rocket is climbing (presumably at full throttle)? Or does it only do what's at the Altitude/Mach sliders? Which means it's not much good? Dunno.

And so, when I'm looking at Delta-V maps of the Kerbal system, do I want to be looking at Engineer with "Atmosphere" on or off? Obviously any map is good for planets with no atmosphere. Did I read that "Atmosphere" Delta-V for Kerbal orbit is 3200? 4500 without "Atmosphere"?

Dunno - sorry for noob questions. If you know a good tutorial that deals with Engineer and these questions please point me at it.

 

[Wanderfound]

KER Atmosphere ΔV readouts just adjusts the isp in accord with the altitude you set. It doesn't account for gravity or drag losses, or the change in isp as the rocket ascends.

I usually just leave it on vacuum setting, and mentally discount a bit for the atmospheric losses. If you've got a consistent TWR of around 2 and 3,500m/s in the tanks, you should just make orbit unless you're trying to do something daft like using a Poodle as a launch engine.

Edited March 30, 2017 by Wanderfound

[Streetwind]

4 hours ago, Wanderfound said:

You'd get better ΔV if you built it so that it maintained the desired TWR (generally about 2 for a Kerbin launch) with all engines firing full throttle at once.

Throttling down or leaving engines off wastes ΔV by effectively increasing payload; the unused engine capacity becomes dead mass. Accelerating at more than TWR2 or thereabouts wastes ΔV to drag losses (this is variable depending on streamlining and flight profile, though).

While this is technically correct, it doesn't always apply. Just because you have an engine nozzle exposed doesn't mean that the rocket is automatically designed to, or better off to, fire it. For example, you can mount another stage onto the sides of a rocket instead of below - this can help you get around launchpad height limits in career mode. The dV advantage from staging is generally much larger than the dV advantage from launching with extra thrust, or that from reducing engine mass to eliminate unnecessary extra thrust.

That means that there's a possibility that if @ondofpond has built a rocket that for all intents and purposes has a solid first stage and a liquid second stage, lighting both at once on the pad is not advantageous, but rather a mistake.

(Rockets with liquid fuel crossfeed through multiple parallel firing stages, such as asparagus designs, are a special case. Those definitely should always follow your advice, and never leave a single engine offline on the pad.)

 

Of course, you can argue that the Swivel isn't the best upper stage engine But perhaps his rocket is larger than what a Terrier can support.

[bewing]

Below 240 m/s, your ship is not experiencing much drag at all. However, if you are just doing typical boosting, it takes a very long time to climb the first 2km. And as said above, you are losing 9.8 m/s of dV every second of that time.

So I know you said blah blah to the whole "get off the pad quickly" thing -- but the faster you get that very first 200m/s into your ship, the better. Even if it involves burning your central engine for a little while and then shutting it down.

 

[AeroGav]

15 minutes ago, bewing said:

Below 240 m/s, your ship is not experiencing much drag at all. However, if you are just doing typical boosting, it takes a very long time to climb the first 2km. And as said above, you are losing 9.8 m/s of dV every second of that time.

So I know you said blah blah to the whole "get off the pad quickly" thing -- but the faster you get that very first 200m/s into your ship, the better. Even if it involves burning your central engine for a little while and then shutting it down.

 

Yeah i used to do that a lot.    Throttle back at "Max Q". 

Would the OP be better off replacing the swivel with a true vacuum engine?   If you're not going to use it off the pad, might as well take a vacuum optimised motor.  Though there probably isn't a good option for him this early in the game , other than a cluster of terriers 

[Mystique]

57 minutes ago, bewing said:

Below 240 m/s, your ship is not experiencing much drag at all. However, if you are just doing typical boosting, it takes a very long time to climb the first 2km. And as said above, you are losing 9.8 m/s of dV every second of that time.

So I know you said blah blah to the whole "get off the pad quickly" thing -- but the faster you get that very first 200m/s into your ship, the better. Even if it involves burning your central engine for a little while and then shutting it down.

 

See no problem using SRBs as 1st stage, if they provide required thrust to avoid extra gravity losses (I do it often, especially with smaller rockets, which need 2-3 SRBs+a sneeze to get to LKO). The only possible issue is control - as SRBs lose mass and gain TWR, it becomes quite hard to control rocket with winglets&reaction wheels, if such problem occurs, having some LFO engine with good gimbal range is a good idea. On the other hand, I use SRBs as TWR-booster only for large rockets and/or when available LFO engines can't provide enough thrust.

[Streetwind]

1 hour ago, bewing said:

Below 240 m/s, your ship is not experiencing much drag at all. However, if you are just doing typical boosting, it takes a very long time to climb the first 2km. And as said above, you are losing 9.8 m/s of dV every second of that time.

So I know you said blah blah to the whole "get off the pad quickly" thing -- but the faster you get that very first 200m/s into your ship, the better. Even if it involves burning your central engine for a little while and then shutting it down.

Well, to be fair, even above 240 m/s your ship isn't experiencing much drag either Aerodynamic losses are so insignificant compared to the total cost of reaching orbit that, given the choice, more thrust is almost always better unless you really overdo it. I'm totally not opposed to getting off the pad quickly.

The thing is though, if you fire your second stage engine to help the first stage move, you're expending fuel that was meant to push a much lighter rocket - about 1/3rd or less the mass, for most KSP designs. You may save on a couple dozen m/s worth of gravity losses by accelerating faster and turning over sooner and harder, but every m/s you invest now costs you 3 or more m/s down the line. So in order for this to work out in your favor, the first stage boosters that you are helping along must save at least twice as much through this trick as the center engine is expending in pulling it off, and even then you're just breaking even. This also assumes that the center engine is no worse than the first stage engines in atmosphere (in terms of percentage loss of max thrust at sea level), but given how most upper stage engines are statted, this is very rarely the case. So the center stage usually has an even harder time to pull profit out of helping the first stage, and the amount of savings you need to earn with the first stage engines just to break even will be larger.

Sure, it is possible that there are rockets which successfully profit from this trick. I can't confirm or deny either way, I have no math and no examples at hand here (and no KSP to test with either). I just want to point out that I think it's extremely likely that there will also be plenty of rockets that don't benefit at all, or in fact lose from it. It is important to consider the individual rocket at hand when giving advice.

Edited March 30, 2017 by Streetwind

[Kryxal]

If I use 4 SRBs, they're usually attached pairwise, with different thrust values.  If 80% would be a good setting for 4 off the pad, see what a pair at 90 and a pair at 70 would get you.  The greater the spread, the earlier you can jettison the first pair and have your TWR drop back down again.  Just try to avoid it falling TOO far, or kick in the LFO stage at that point at lower thrust.

[Aegolius13]

Even if you were going all SRB, it may also be worth it to light your LF engine at very low throttle,  just to use gimballed thrust.   It doesn't take much to give you decent control. 

Edited March 30, 2017 by Aegolius13

[DarknessHasLost]

9 hours ago, ondofpond said:

Done a search, can't get a thread on this specific topic. Been a KSP player for years, like efficiency and since career mode I build for cost, every time, even if I don't need to with excess funds. I'm not bad, but not perhaps the most discerning of method in my builds.

 

So - using KSP Engineer and looking at how much Delta-V I can get for my buck. On build using a Swivel for a two-stage centre, I put two or four BACC SRBs on, but stage so that I launch off the pad with SRBs only. Obviously ensuring I have enough TWR for it to be viable. If I have four SRBs and enough TWR I might light only two SRBs on the pad, drop and light the next two (performing a struggling gravity turn with risk of impact when dropping) before finally lighting my Swivel.

Obviously this makes for a slow initial ascent. I understand the thick atmosphere thing at the pad, get going quickly blah blah, but when I compare Delta-V stats with Atmosphere in Engineer, doing what I say above yields more Delta-V than if I light the Swivel on the pad, or light all four SRBs on the pad if using four. Again, TWR is considered and a slower start seems to be better on final Delta-V against cost.

I'm not arguing the virtues of using SRBs at all vs liquid side-booster stages, to me that's a no-brainer on cost (I can end up with 2.5m Skipper build with no Mainsail side boosters, just six Kickback SRBs using the above method, lighting four on the pad and lighting the other two before finally lighting the Skipper, when payload makes it viable).

What I'm wondering is, am I wrong to be purely looking at Delta-V values in Engineer in these builds? What might I not be considering that is possibly making the Delta-V figures either erroneous or a bad idea for another reason I'm not considering?

 

considering DV and TWR isn't bad but you still gotta calculate Burn time, Depending on how long it takes to reach LKO your 12k DV could be reduced to 35 DV if it takes you 40 hours to burn it. Ive made plenty of 3300 DV SSTO's that cant break orbit due to the time it takes.

[CSE]

6 hours ago, Streetwind said:

The thing is though, if you fire your second stage engine to help the first stage move, you're expending fuel that was meant to push a much lighter rocket - about 1/3rd or less the mass, for most KSP designs.

Not quite as dramatic as 1/3 of the mass: use the dry mass of the first stage, because it's the momentum in the dry first stage that the rocket forfeits when it drops the stage. The momentum gained by the first stage's propellant is recovered by the rocket when that propellant is ejected through the nozzles. Of course, any mass saving at all is still a good thing.

This leaves us with slightly different reasons for throttling up the core stage later on well-designed KSP rockets:

• Solid rocket boosters have significantly lower I_sp than the liquid core, so burn their fuel first. The momentum is more valuable if stored in the high-I_sp core propellant than in the low-I_sp boosters.
• Liquid rocket engines are rather heavy compared to real life, so the dry mass of LRBs makes a more significant difference to the core stage's performance.

[Reusables]

On 2017. 3. 30. at 5:33 PM, ondofpond said:

If I light the Swivel on the pad

Sorry for being a bit off-topic, but Reliant is better than Swivel. 100 cheaper, 0.25t lighter and produce 12% more thrust for only 10s(3%) less vacuum ISP.

Here is comparison between LF/Ox rocket engines: https://docs.google.com/spreadsheets/d/1FAWGoiUAasr3gukWhhTMjGIQSKcoWXPhvbyM_6E2u3U/htmlview

 

Also it's shown on the sheet that SRBs and Vacuum engines are king of the efficiency. It may be better to have those two, if you are not using combinations like Skipper&Kickback.

Swivel is not a vacuum engine, though. I personally think it's trainer engine for rookie pilots.

Edited March 31, 2017 by Reusables

[Loren Pechtel]

Very low TWR on your lower stages is a bad idea because it increases you gravity losses.  You want to boost quickly.  In vacuum you want to boost as hard as you can (assuming you don't need to haul extra engines for the purpose), on Kerbin I find things get problematic if you go over 2.

Many times I have launched with the center engine offline, though.  I often build a rocket whose main engine is a space engine and lift it high enough with SRBs for the space engine to make sense.  The losses of burning a space engine low in the atmosphere mean it makes more sense to leave it off at first.