Help me understand jet engines

[jros83]

Hello Kerbonauts!

I recently have started messing around with atmospheric only aircraft. I play career mode so I am limited in my parts. I have noticed strange things with the basic jet engine, and I don't know if they are normal as far as the game is concerned, normal in real life and I know much less than I thought, or what.

For instance, I have noticed my craft with basic jet engines equipped gain speed in climbs, even steep climbs, regardless of the mass of my craft vs. thrust (well, so far). Now I understand the higher I get the less dense the atmosphere gets, but these are jet engines that require air intake, as well and because of that they never get high enough for less dense air atmosphere to make much a difference. So then why am I always gaining speed in a climb...? I don't abuse the number of intakes I use, as I like to keep things looking believable. I realize in real world examples it's when the pounds of thrust exceeds the weight of the aircraft (such as why F-16s can gain speed in a near vertical climb) but ingame it happens no matter what.

Another weird thing I noticed is that, even at full engine power in a 90 degree dive, my craft always LOSE SPEED. Now it goes without saying that there'd be no justification for that, as the only thing that could cause that to happen was if I started from so high up in much thinner atmosphere that I would be slowing down to terminal velocity, and it goes without saying that is not the case. So why am I LOSING speed in full powered dives?

I am new to KSP so I'm sure there's easy explanations lurking somewhere for this that I am just not getting. Thank you in advance.

EDIT: also, if anyone could tell me where I find my signature block in my settings, I would appreciate it. I can't find it anywhere.

Edited August 15, 2014 by jros83

[robopilot99]

The jet engines are massivly OP it's really easy to make a craft with a thrust-to-weight of greater then 1. As for the slowing down. Kerbin's apnosphere gets thick quick. Even the difrence from 10,000 to 1,000 is substanshal. the root of your problem is almost certainly apnospheric drag.

[Laie]

The atmosphere in KSP is thick. Pea soup, some say. Your speed correlates strongly with altitude because altitude means less air pressure which in turn means less drag. If you want to fly fast, you need to fly high. And if you get down low, you will slow down. It takes a LOT of thrust to make 250m/s near sea level; but at 10km altitude, it pretty much happens on it's own.

The thrust of jet engines correlates with airspeed, regardless of altitude (provided they have enough air, of course). They have most thrust when standing still, and zero thrust at 1000m/s. For practical purposes, they work well until 300m/s and reasonably until 600m/s, then it goes quickly downhill. To the best of my knowledge, real-life jet engines show different behavior.

Things don't have to be that way: there are mods that change this to a more realistic (or at any rate, more credible) behavior. They're called NEAR and FAR -- the FAR fanboys are sure to be around shortly.

Oh, and congratulations on having built a working airplane. It took me a lot of tries to come up with something that wouldn't crash within moments.

Edited August 15, 2014 by Laie

[Wanderfound]

Things don't have to be that way: there are mods that change this to a more realistic (or at any rate, more credible) behavior. They're called NEAR and FAR -- the FAR fanboys are sure to be around shortly.

That would be me.

And, yeah; the treatment of atmospheric craft in stock is ridiculous. That's why they don't make sense. Install FAR or NEAR if you want them to behave sensibly.

[Murph]

The jet engines are massivly OP it's really easy to make a craft with a thrust-to-weight of greater then 1.

That really depends on what you call overpowered. Their thrust ratings are on the upper end of real world mainstream jet engines, but certainly nothing exceptional. The RR RB211 and Olympus 593 produce somewhere in the 200kN range. If you strapped a couple of those RR engines onto a 10-20t aircraft in the real world, it would happily do a vertical climb on them.

If you're only going to have 2 jet engines, a basic/subsonic one, and an advanced/supersonic one, KSP currently has approximately the correct choice, as less powerful would inhibit building larger aircraft, or building mach 3+ aircraft.

[Wanderfound]

That really depends on what you call overpowered. Their thrust ratings are on the upper end of real world mainstream jet engines, but certainly nothing exceptional. The RR RB211 and Olympus 593 produce somewhere in the 200kN range. If you strapped a couple of those RR engines onto a 10-20t aircraft in the real world, it would happily do a vertical climb on them.

If you're only going to have 2 jet engines, a basic/subsonic one, and an advanced/supersonic one, KSP currently has approximately the correct choice, as less powerful would inhibit building larger aircraft, or building mach 3+ aircraft.

Under stock aero, a single-turbojet basic plane can do better than Mach 6. That's scramjet territory. If you're having fun with it, that's all cool, but it's a long way from realistic.

[Murph]

Under stock aero, a single-turbojet basic plane can do better than Mach 6. That's scramjet territory. If you're having fun with it, that's all cool, but it's a long way from realistic.

That's more a problem with the aerodynamic simulation, rather than the engines themselves being overpowered. If memory serves, the SR-71's ramjet design (which is 50 years old now) was basically good for mach 6, if the materials could be adjusted to cope with the additional thermal load.

[Jouni]

The biggest problem with KSP jet engines is the fuel usage bug. The game includes both fuel and intake air in the mass flow rate, when it should only include the fuel. As a result, jet engines consume around 15x less fuel than they should, raising their actual Isp figures to tens of thousands.

[jros83]

Thanks for the replies all.

And Laie, I used to have a terrible time making atmospheric craft until I started always building with the center of mass/center of lift/center of thrust marks ticked on. Once I got a grasp on how they all interact, building stable atmospheric craft became markedly easier.

Also, and someone please tell me if I'm wrong, but I've noticed installing an SAS module and flying with T and R turned on helps a lot too. Especially helpful with takeoff, as without them I wobble and dip and crash into the runway half the time.

Hmm maybe I'm using the wrong name for the module. All cockpits have SAS, I know. What I'm talking about is the ones like the gyroscopic one and the next model up from that (can't open KSP right now to take a look, but I'm sure you guys know what I mean. Should those be used on planes and should I be flying around with T and R always checked, or am I gaming the system that way?

Edited August 16, 2014 by jros83

[Pecan]

...Should those be used on planes and should I be flying around with T and R always checked, or am I gaming the system that way?

It's definitely not 'gaming' the system, you're just using the equipment you've fitted.

However, R (RCS) isn't going to be doing anything at all unless you've fitted RCS thrusters. If you have fitted those it'll be using-up the RCS fuel (monopropellant) you've installed and/or that is provided with the command module anyway.

So SAS is the only one you can use 'for free' but the torque provided by pods is very low so you don't get much unless you fit a dedicated part; inline reaction wheel, IRW, is the cheap one, otherwise inline advanced SAS module (IAS). I think what you meant was clear and most people just talk about 'SAS'.

If you need to use SAS much though it means your design is wrong, or at least out of trim. Use fine controls (pres capslock, pitch/roll/yaw indicators at the bottom-left of the screen turn blue) and the 'mod' key - alt on windows - along with the normal control keys. Typically you want some pitch-up trim (alt-S) at takeoff but the in-flight settings change with fuel-load - as the centre of mass moves - altitude and speed. Get the trim more or less right and the 'plane should almost fly 'hands free'.

[Wanderfound]

Thanks for the replies all.

And Laie, I used to have a terrible time making atmospheric craft until I started always building with the center of mass/center of life/center of thrust marks ticked on. Once I got a grasp on how they all interact, building stable atmospheric craft became markedly easier.

Also, and someone please tell me if I'm wrong, but I've noticed installing an SAS module and flying with T and R turned on helps a lot too. Especially helpful with takeoff, as without them I wobble and dip and crash into the runway half the time.

Hmm maybe I'm using the wrong name for the module. All cockpits have SAS, I know. What I'm talking about is the ones like the gyroscopic one and the next model up from that (can't open KSP right now to take a look, but I'm sure you guys know what I mean. Should those be used on planes and should I be flying around with T and R always checked, or am I gaming the system that way?

All's fair in love and Kerbal.

A well done plane can get by without torque or RCS, but they do make it a lot easier. The way I look at it, ASAS units on a plane are the equivalent of the fancy computerised aeronautics packages that are used by all modern fighters. Those things are so unstable that they'd be completely unflyable without them. In a modern dogfighter, the computer flies the plane; the pilot just tells it what he'd like it to do.

RCS is handy on aircraft (vectored thrust! Harrier for the win!), but keep it turned off most of the time or you'll waste fuel. Just flick it on and off as and when needed.

Incidentally, if you like RCS, start playing around with the Vernors. Each one has the thrust of eleven linear RCS ports. They're very handy for making unruly aircraft behave, or making better aircraft do ludicrous tricks. Useful for building Munar VTOLs, too.

[Murph]

All's fair in love and Kerbal.

A well done plane can get by without torque or RCS, but they do make it a lot easier. The way I look at it, ASAS units on a plane are the equivalent of the fancy computerised aeronautics packages that are used by all modern fighters. Those things are so unstable that they'd be completely unflyable without them. In a modern dogfighter, the computer flies the plane; the pilot just tells it what he'd like it to do.

RCS is handy on aircraft (vectored thrust! Harrier for the win!), but keep it turned off most of the time or you'll waste fuel. Just flick it on and off as and when needed.

Incidentally, if you like RCS, start playing around with the Vernors. Each one has the thrust of eleven linear RCS ports. They're very handy for making unruly aircraft behave, or making better aircraft do ludicrous tricks. Useful for building Munar VTOLs, too.

It sounds like you probably know this, Wanderfound, but for the benefit of others, the instability of modern jet fighters is actually a deliberate design choice, to give them the ultimate aerobatic/dogfighting ability (i.e. when needed, they will go more or less instantly from straight and level into a max-g manoeuvre). The avionics let the pilot take it right to the ragged edge of the envelope, hopefully keeping it just on the controllable side of the line (and Martin-Baker takes care of the other side of the line…).

The way I see it, is that SAS should be viewed as a standard and non-cheaty feature of planes in KSP. All of the standard cockpits have the full SAS system included. The only difference (after the SAS system was reworked a few versions ago) between the cockpits, the IRW, and the IAS is the amount of torque and cosmetic appearance. If the cockpit alone provides sufficient torque, you don't gain anything by having additional SAS units on the plane, but they can be added as and when required. The 3 standard cockpits provide 10 torque each, which I personally reckon is plenty for most 10-20t planes. The IRW provides 8, and the IAS 15.

If you need more than 1.0 torque per ton to make a plane controllable, there's a good chance that the design has fundamental issues which should be fixed other than by adding more torque, and more than 0.5 per ton should have you thinking about whether you've got sufficient control surfaces and the CoM vs CoL vs CoT reasonable. Too much torque can also be a bad thing, as it can allow you to push the plane beyond the envelope and cause a spin. No amount of torque will properly fix a plane which has a fundamentally bad CoM vs CoL vs CoT and goes into a spin at the slightest control input. When the SAS is active, you can watch the pitch/yaw/roll indicators on the bottom left of the screen to see how hard it's having to work â€â€.if any of them are constantly at or close to max during straight and level flight or gentle climb/descent, there's a design problem with the plane. Of course, as with the real world, it's a valid choice to design a plane to be on the edge of instability, for maximum responsiveness, and there is no 100% correct way to design them (but plenty of ways to create smoking holes in the ground…).

As for RCS, it's a useful feature, a requirement for orbital docking (more or less), but there's something wrong with the plane if it's uncontrollable in simple atmospheric flight without it turned on. Keep it switched off, other than when required, or you'll rapidly run out of RCS fuel and not have any when you do need it.

[cantab]

Re Kerbin's atmosphere, it's effectively very dense (because of the screwy drag model) but it does drop off quickly with altitude. At 5 km the pressue and thus drag is down to 37% sea level. By 10 km it's 14%.