Do you throttle-limit engines that are too big?

[CrashyMcCrashFace]

i.e. On launch sometimes a Skipper is not quite powerful enough and Mainsail is way too powerful. Rather than going with something more complicated with a Skipper and a bunch of other small engines I'll just run the Mainsail but set the throttle limit to limit the TWR where it won't go too fast through the atmosphere. Is this a good practice? Not sure if this is relevant but I always use a fairing even if only slightly needed and usually let MechJeb do the launch for me.

[Foxster]

11 minutes ago, CrashyMcCrashFace said:

i.e. On launch sometimes a Skipper is not quite powerful enough and Mainsail is way too powerful. Rather than going with something more complicated with a Skipper and a bunch of other small engines I'll just run the Mainsail but set the throttle limit to limit the TWR where it won't go too fast through the atmosphere. Is this a good practice? Not sure if this is relevant but I always use a fairing even if only slightly needed and usually let MechJeb do the launch for me.

If it works for you then fine. 

Of course, you are carrying extra mass and spending extra dosh on doing this but if that doesn't worry you then where's the harm?

To add a little umph to a Skipper you could always slap a couple of Thuds on the side. 

Getting to orbit quickly is generally a fairly good way of doing it with less dV, as long as you don't overdo it and burn up in the process. 

 

Edited May 3, 2018 by Foxster

[Human Person]

With higher TWR I usually just go more aggressive on the turn (if aerodynamics allow for it). Otherwise, I still go with full thrust on liftoff and throttle back a little once reaching Mach 1.

Edited May 3, 2018 by Human Person

[bewing]

I prefer to stick with the Skipper and slap a handful of Kickback SRBs on the sides of it.

 

[HebaruSan]

1 hour ago, CrashyMcCrashFace said:

I'll just run the Mainsail but set the throttle limit to limit the TWR where it won't go too fast through the atmosphere.

I add fuel.

[mystifeid]

48 minutes ago, Foxster said:

and usually let MechJeb do the launch for me

If using Mechjeb, try limiting Q instead of setting a throttle limit so that you have a dynamic max thrust.

[Foxster]

10 minutes ago, mystifeid said:

If using Mechjeb, try limiting Q instead of setting a throttle limit so that you have a dynamic max thrust.

I don't remember saying that 

[mystifeid]

3 minutes ago, Foxster said:

I don't remember saying that

Sorry, quoted wrong post! Still cross eyed from too much KSP tonight.

[Superfluous J]

I assume your question is because you think it may not be the best way and you're (sort of) right.

If you want to be cost- and fuel-efficient then the Mainsail is bad. It's expensive and heavy and by throttling it down you're paying for something you don't use.

If you want to get more stuff into orbit quickly (in your play time not game time) or your computer is struggling with part counts, then using a Mainsail is perfectly fine. It's easy and usually works and is only one part.

[CrashyMcCrashFace]

50 minutes ago, 5thHorseman said:

I assume your question is because you think it may not be the best way and you're (sort of) right.

If you want to be cost- and fuel-efficient then the Mainsail is bad. It's expensive and heavy and by throttling it down you're paying for something you don't use.

If you want to get more stuff into orbit quickly (in your play time not game time) or your computer is struggling with part counts, then using a Mainsail is perfectly fine. It's easy and usually works and is only one part.

I've already finished the career mode (unlocked all science) and have enough money to burn but It's more of a roll playing thing. I just don't want my stuff on the verge of burning up and using 2-3x as much power and fuel as needed. It seems like the rocket is fighting the atmosphere and wasting a lot if it goes too fast.

[Vanamonde]

The maximum thrust of an engine is roughly proportional to its mass. So if you're not running it at 100%, then you're carrying extra mass in the form of the unused engine potential. Rather than run it at partial throttle, use more than one smaller engine or, as Bewing suggested, add some SRBs to give it an extra kick right off the pad. 

[Snark]

21 minutes ago, CrashyMcCrashFace said:

I've already finished the career mode (unlocked all science) and have enough money to burn but It's more of a roll playing thing. I just don't want my stuff on the verge of burning up and using 2-3x as much power and fuel as needed. It seems like the rocket is fighting the atmosphere and wasting a lot if it goes too fast.

As long as your launchpad TWR is in the range of around 1.3 to 2.0, you're good.  You probably don't want to go much lower than 1.3 (due to excessive gravity losses), or much higher than 2.0 (too much mass wasted on engines; too much aerodynamic drag loss).  But "1.3 to 2.0" is a pretty broad range and gives a good deal of design flexibility.

My own solution, if I've got a main engine that's almost-but-not-quite powerful enough, is to just strap on some SRBs, as others have suggested.  Personally, I favor the Thumper over Kickbacks, for a few reasons.  First, the Kickbacks are so long and skinny that I often have trouble physically fitting them where I want them to go.  Second, I like to launch rockets at around 2.0 TWR myself, and even a naked Kickback without any other payload only has an inherent TWR of 2.5, meaning that they don't do a great job of lifting my overall TWR since most of their oomph is spent just accelerating themselves.  The Thumper has a significantly higher inherent TWR of around 3.3, so is more useful for lifting my launchpad TWR.

One thing I like to do, when strapping on some radial Thumpers, is to attach some 1.25m fuel tanks on top of them, and turn on the radial decouplers' fuel crossfeed.  This is nice because I can use my central LFO engine off the pad, and it's draining the radial tanks atop the SRBs-- a very simple form of asparagus.  That way, when the SRBs burn out and are jettisoned, the central LFO engine still has close to a full tank.  Gives it legs.

[Curveball Anders]

I throttle back bosters, to match the desired TWR.

 

[Just Jim]

I don't throttle limit an engine, but I have many rockets I'll throttle back a bit right after launch, until I get out of the lower atmosphere, then go back up to 100%.

[Zeiss Ikon]

I mostly use the Twin Boar with a variable number of strap-on Lf/O boosters.  Disclaimer: I'm not a fan of solid rockets; bad memories of the Challenger disaster might have something to do with it.

If you make a booster with a Reliant or Skiff (the latter from Making History) and three 800 1.25 m tanks, add a Sepratron on the nose cone, use crossfeed in the decouplers, and watch your booster tank levels, you can use anywhere from zero to six of them around a Twin Boar and tailor things so the Twin boar runs out of fuel just before you make orbit, hence avoiding adding trash to the part of LKO you use most -- the lowest altitude equatorial band.  I recently launched a crewed Duna flyby with that setup (six boosters, upper stage was a 6400 tank with a Poodle pushing a Mk. 1-3 with a lot of science aboard); with six boosters, it can haul a relatively large payload into LKO.

The only places I've used throttle limiting was in the early days of my career, when I only had a Flea and a Mk. 1 and wanted to keep the G load under control, and when I need to make correction burns long enough to shut down fairly accurately.

[foamyesque]

I make a bigger payload :v

 

Alternatively, some Thumpers or even Hammers, if you're launching something small, will usually bring launch TWRs up to something useful if you're a little short on thrust. I usually don't bother with crossfeeds on the grounds that I *want* my core to burn its fuel so's it has a desired TWR once my boosters are jettisoned.

[Gargamel]

My technique, and it's probably wrong and ineffecient, is to have a decent main engine, but heavy SRB's.   I'll then limit the thrust of the SRB's so at launch I'm in the 1.6-1.8 range.  I then use MJ to keep my acceleration at 20 m/ss, and it will throttle back the main engine as the SRB's get more relative TWR.    Then when they separate, I have more Dv left for the main engine to complete the orbital insertion burn.    This allows me to build lifters that cover a larger range of payload masses, but keep the same ascent profile, so I can launch directly to rendezvous without trial and error.    Of course, this is MJ dependent, flying this by hand is doable, but a little more complicated. 

[CrashyMcCrashFace]

I didn't want to add fuel or payload because the objective is very specific and nothing else is needed.

So I ran some experiments with MechJeb flying to keep things consistent. Throttling by limiting Q vs unlimited. Turns out unlimited uses slightly less fuel, which I'm all for. But this doesn't make sense. Air drag is not linear to speed. IRL aircraft and even cars will get worse millage if they go too fast. I understand the ship is getting out of the atmosphere faster but that doesn't change the amount of atmosphere needed to travel through. Oh well, I guess I'll have to live with it.

In summary, like almost everybody said, I probably won't limit the throttle anymore and just live with a little "speeding".

Thanks everybody.

Edited May 4, 2018 by CrashyMcCrashFace

[bewing]

31 minutes ago, CrashyMcCrashFace said:

Air drag is not linear to speed. IRL aircraft and even cars will get worse millage if they go too fast

This is true ... however, when you start talking about supersonic flight regimes, it gets a lot more complicated. As long as you are below transonic speeds, drag goes pretty closely as v-squared. Once you are transonic it gets a lot worse than that. Once you go supersonic it levels out a lot. In the KSP drag model, in fact, drag drops a significant amount once you break the sound barrier. These effects combine to make it true that you want to go quite fast, fairly quickly, and get supersonic.

 

[CrashyMcCrashFace]

5 minutes ago, bewing said:

This is true ... however, when you start talking about supersonic flight regimes, it gets a lot more complicated. As long as you are below transonic speeds, drag goes pretty closely as v-squared. Once you are transonic it gets a lot worse than that. Once you go supersonic it levels out a lot. In the KSP drag model, in fact, drag drops a significant amount once you break the sound barrier. These effects combine to make it true that you want to go quite fast, fairly quickly, and get supersonic.

 

Thanks. I can accept that and just suspend reality.

Edited May 4, 2018 by CrashyMcCrashFace
suspend not spend

[Curveball Anders]

On 5/4/2018 at 9:53 AM, Gargamel said:

My technique, and it's probably wrong and ineffecient, is to have a decent main engine, but heavy SRB's.   I'll then limit the thrust of the SRB's so at launch I'm in the 1.6-1.8 range.  I then use MJ to keep my acceleration at 20 m/ss, and it will throttle back the main engine as the SRB's get more relative TWR.    Then when they separate, I have more Dv left for the main engine to complete the orbital insertion burn.    This allows me to build lifters that cover a larger range of payload masses, but keep the same ascent profile, so I can launch directly to rendezvous without trial and error.    Of course, this is MJ dependent, flying this by hand is doable, but a little more complicated. 

I do it similar but I set MJ to limit MaxQ so it throttles down the main engine when crossing the sound barrier (-ish).

My target is to set the booster so that the main engine never throttles back fully but a nudge pwr.

On bigger launchers I'm also using hybrid booster/aspargus with tanks on top of the boosters.

 

[Gargamel]

27 minutes ago, Curveball Anders said:

I do it similar but I set MJ to limit MaxQ so it throttles down the main engine when crossing the sound barrier (-ish).

My target is to set the booster so that the main engine never throttles back fully but a nudge pwr.

On bigger launchers I'm also using hybrid booster/aspargus with tanks on top of the boosters.

 

What Q value do you use?  I've tried it a couple times, and almost lost the vehicle.  I just don't know enough of the exact values to trust it (myself really).  I understand the concept in general, enough to know I should be using it. 

[Curveball Anders]

1 hour ago, Gargamel said:

What Q value do you use?  I've tried it a couple times, and almost lost the vehicle.  I just don't know enough of the exact values to trust it (myself really).  I understand the concept in general, enough to know I should be using it. 

18000-ish depending a bit design, but after several relaunches of the same lifter with various profiles I landed about there.

[Gargamel]

10 hours ago, Curveball Anders said:

18000-ish depending a bit design, but after several relaunches of the same lifter with various profiles I landed about there.

Cool, thanks.  I had tried the 20k default value, and my first few attempts were "AHHHH FIRE!!!"

[foamyesque]

On 5/4/2018 at 9:27 AM, CrashyMcCrashFace said:

I didn't want to add fuel or payload because the objective is very specific and nothing else is needed.

So I ran some experiments with MechJeb flying to keep things consistent. Throttling by limiting Q vs unlimited. Turns out unlimited uses slightly less fuel, which I'm all for. But this doesn't make sense. Air drag is not linear to speed. IRL aircraft and even cars will get worse millage if they go too fast. I understand the ship is getting out of the atmosphere faster but that doesn't change the amount of atmosphere needed to travel through. Oh well, I guess I'll have to live with it.

In summary, like almost everybody said, I probably won't limit the throttle anymore and just live with a little "speeding".

Thanks everybody.

 

Air drag isn't the only thing impacting rocket efficiency, though: you also lose energy to gravity, and that's mitigated by having more thrust. Gravity drag on a Kerbin launch usually predominates unless you're launching particularly awkward payloads and/or have tremendous TWR, so staying at full throttle helps.