drag and weight question

[miklkit]

I'm trying to decide which is the best fuel tanks to use.  The engines are NERVs and Rapiers.  Is a stack of tanks for each engine better overall or is a cluster of 4 engines on a rocket tank better?

I'm trying both methods and they both work, and the rocket tank looks cleaner, but I can't tell any difference in performance.

[Streetwind]

If you mount a tank that contains LF+Ox to an engine that consumes only LF, then that engine will run out of fuel once all the liquid fuel is gone, with the tank still being roughly half filled with oxidizer.

This not only means that you're carrying around the mass of the useless oxidizer, which requires more propellant expenditure to accelerate, and more lift to keep it airborne, which requires more wing, which creates more drag, and also add mass, which requires more propellant expenditure... yada yada; but also, only half the tank's actual volume will be dedicated to carrying fuel you can use. Which means you need to add more tanks to get the total amount of fuel you need. All of which, again, only have half of their volume dedicated to useful fuel, and the rest to useless oxidizer, which you have to carry along, see above.

But even if you think "hey, I can remove the oxidizer in the hangar, no problem", it's still not equal to using pure liquid fuel tanks. Because the fact that you need twice as many tanks still applies. And tanks have dry mass. Meaning, they have a mass even when empty. A base mass that gets added to the mass of fuel inside them. So if you carry twice as many tanks, you're carrying twice as much dry mass. (This, by the way, is why rockets stage: to get rid of performance-killing dry mass.)

Now, RAPIERs can be used in closed cycle, where they consume oxidizer. A typical spaceplane will want to carry some oxidizer for the moment when the atmosphere gets too thin to support airbreathing combustion. But you should strive to bring only exactly as much as you need to successfully fly the mission.

[Zhetaan]

15 hours ago, miklkit said:

I'm trying to decide which is the best fuel tanks to use.  The engines are NERVs and Rapiers.  Is a stack of tanks for each engine better overall or is a cluster of 4 engines on a rocket tank better?

I'm trying both methods and they both work, and the rocket tank looks cleaner, but I can't tell any difference in performance.

I'll second @Streetwind's post, at least for planes.

However, nothing in your post says specifically that you're flying a plane; just that you're flying in an atmosphere.  

Nevertheless, the logic holds.  The only really good reason to take a tank full of LFO is if you intend to use it, and if you're using closed-cycle Rapiers when you have Nervs available ... well, why would you do that?  Pushing through the upper atmosphere might be a good reason, but you don't need a big rocket tank of LFO for that.

Depending on how much oxidiser you need, you may be best-positioned to use a Mark 3 liquid fuel fuselage with a Mk3 to 2.5m Adapter.  It still has a lot of LFO--likely more than you'll want--but you'd need the adapter anyway for the cleaner look.

Of course, you can always try a mod; there are plenty of rocket-profile LF tanks out there, since people correctly saw that it made no sense for the Nerv to be a fantastically-efficient LF-only space engine and not have any good rocket-profile LF-only tanks to go with it.  Some of these are just fuel-switching mods that will make a stock LFO tank into an LF-only tank without the need to add extra parts.  It's worth a look: without mod tanks, the result is often that you either get a Mark 3 fuselage assembly that adapts down to 1.25m for the Nervs (and the adapters almost all have LFO, so watch out), or else you get clusters of Mark 1 fuselages because they have the same 1.25m profile as the mid-diameter rocket parts (small in KSP would be the .625m parts, like the Oscar-B).

In fairness to the developers, the Nerv originally was an LFO engine like the other rocket engines (it wasn't realistic in its function; it just had great efficiency), so when it was introduced, there was no need for new tanks.  They corrected that later, but never added new tanks to go with it.

[miklkit]

Ok, pictures needed.  Yes they are spaceplanes and I'm trying to find out if I'm going down a rabbit hole or not.

This is an early design after lots of development, and it has gone to Duna and back with fuel left over.  But that monoplane design is not the best.

This biplane works well and also has good range, but I'm unsure if that big rocket tank is a good idea or not. 

Yes the balance of different fuels is a pretty basic part of design.  The faster it gets out of the atmosphere, the less Ox it needs, and that is my goal.

Over time this plane's performance increased to the point I needed to cut back on the Ox it carries.

[18Watt]

In general I’ve found keeping the number of fuselages to a minimum helps with performance.  If you can keep everything in a single fuselage stack, that is usually ideal.

One reason to add additional fuselage stacks is to get extra intakes.  One shock-cone intake will feed 3.5 RAPIERS.  (2 will feed 7 Rapiers..).  So if you have more than 3 Rapiers, you likely need at least two intakes.

Your goal should be to get the highest velocity you can in air-breathing mode.  The extra drag from additional fuselage stacks is counter-productive.  That can be overcome by additional engines and weight, but that decreases your useful payload.

Anyway, my suggestion is to reduce the number of fuselage stacks if possible.

[miklkit]

Ah, so one big rocket tank with 4 rapiers is more efficient than 4 MK1 stacks?  That is where I'm heading, but for clean looks more than anything else.  All those stacks look cluttered. 

Performance isn't an issue as the last screenie shows.  That one has finished burning and is coasting up to space.  They can go over 1400 m/s on air and the best fuel savings are found in the most efficient ascent profile.  Still working on that.

I've settled on 8 rapiers and 5 nervs with 4 intakes.  Now to figure out the best combination while staying away from MK2 parts.  They are kinda heavy.

[18Watt]

9 minutes ago, miklkit said:

I've settled on 8 rapiers and 5 nervs with 4 intakes. 

You might consider that you only need 3 intakes to power up to 10 Rapiers.

Also, if you can reduce your Rapiers to 7, then you’d only need 2 intakes.

5 Nukes is a lot, they are really heavy.

13 minutes ago, miklkit said:

They can go over 1400 m/s on air and the best fuel savings are found in the most efficient ascent profile.

By trimming your drag down, you should be able to get over 1500 m/s., which is usually my target speed.  1600 m/s + is also possible.   At those speeds heat does become a concern however.
 

 

[miklkit]

I like even numbers, and odd numbers end up asymmetrical anyway.  Yuck.

I use landing on the Mun as the benchmark.  Some of them can't overcome the gravity with only 4 NERVs and playing with the rapiers makes things get hectic really fast.  One has 6 NERVS and it needs them.

In another thread some people were talking about the best flight path for SSTOs and they seemed to think 1300 m/s was about the max before going closed cycle.  I was timing it so they only get to 1300 for a long time and am only now letting them run up some more speed.  And yes heat is a problem. 

[18Watt]

26 minutes ago, miklkit said:

I like even numbers

Well, I can’t argue with that, me too!

 I didn’t realize landing on Mun was a mission parameter.   Yes, two nukes isn’t enough for that.  Using the Rapiers there would require a lot more oxygen.

You might find it interesting to look at profiles that get you to 1500-1600 m/s.  Finding parts that can withstand the heat becomes tricky.  The Shock-Cone intake works well, but not all of the nose cones will handle that heat.

[miklkit]

I've got a bunch of different MK3s.  There are ore carriers, cargo carriers, rover carriers, and passenger carriers.  I just tried to build a different style passenger ship and it ended up pretty much like the one in the 2nd screenie landed in the grass.  That one is a rover carrier.  I'm stuck in a rut.

I took it for a spin and let it stay down in the air, and it got over 1500, but it used a lot of OX getting up to orbit.  Oddly the hottest things were 3 of the NERVs.  They darn near blew!  Methinks I'll hold the speed slower and climb steeper next time.  I also relocated some of the engines so they aren't cooking each other.

[Zacspace]

2 hours ago, miklkit said:

In another thread some people were talking about the best flight path for SSTOs and they seemed to think 1300 m/s was about the max before going closed cycle.

I remember having a pretty confused exchange about this with you in another thread, it might have been me. To clarify, you go as fast as you can on open cycle. The limit's when you blow up, and even then there's tricks you can use to go a little faster.

I tried to find a tread about high-efficiency SSTOs to help explain how they work. The best I could find is this (where you'll find both me and our friend 18watt). It's unfortunately pretty light on documentation, but it has a lot of pictures of highly min-maxed SSTOs and some of them being flown.

I'm dropping a link to one of my own SSTOs which has what I hope are pretty easy to follow flight instructions on its kerbalx page. It uses a lot of the same construction techniques and a similar ascent profile to the crazy high-performance SSTOs you'll see in the thread I linked above. Hopefully if you try it out and take it apart to see what makes it tick it'll help you out.

[18Watt]

9 hours ago, Zacspace said:

I'm dropping a link to one of my own SSTOs

I just downloaded that ship, because after looking at screenshots of it I had a few questions.  After loading it, I was able to figure out most of my questions, like where the intake is.

Fantastic design!  On my first test run with it I managed to get to 1,670 m/s on open cycle, which is fantastic!  I normally consider 1,500 to be adequate, but faster is always better.  Also, the compact rover design is great.  I also liked the offset structural panels connected to the hydraulic cylinders, brilliant!

Heres the question I wasn’t able to figure out:  What are the Rapiers attached to?  They don’t like to surface-mount to the back of large tanks.  I’ve used engine plates, cubic octagonal struts, and even short 1.25 m fuel tanks (which will surface mount to the back of tanks).  But it looks like you were able to tack them on to the back of the tank without using any of those tricks.  Just curious if you remember how you attached them..

[Zacspace]

2 hours ago, 18Watt said:

What are the Rapiers attached to?  They don’t like to surface-mount to the back of large tanks.

They do with the mod Editor Extensions Redux. I actually forgot I was even using a mod, since it doesn't cause craft incompatibility with stock games. I personally find it indispensable for technical builds, but now that I'm thinking about it, between the mod and my use of both DLCs this plane probably isn't the best teaching tool. Maybe I should build an SSTO just for that purpose.

2 hours ago, 18Watt said:

I also liked the offset structural panels connected to the hydraulic cylinders, brilliant!

I spent a lot of time trying to come up with a more elegant/realistic solution for that. Nothing else worked as consistently as the floating squares in the end though.

[miklkit]

Been doing lots of reading.   On first glance I noticed on those optimized SSTOs is that everyone likes stacked strakes.  Dunno why, but can use them to balance fuel types as well as weight and aero balance. 

Never DLed anything and don't know how to install anything.  Hope it comes with instructions.  Those ascent instructions are interesting.  I may be wasting fuel by keeping the Rapiers on too long.  Normally I'm done burning by 38000-40000 m and just coast up to 80 from there.

[Zacspace]

16 minutes ago, miklkit said:

On first glance I noticed on those optimized SSTOs is that everyone likes stacked strakes.  Dunno why, but can use them to balance fuel types as well as weight and aero balance.

The strakes pull double duty as fuel tank and wing, so 2 parts for the "price" (in terms of mass and drag) of one. They use them over the shuttle wings I think because they have a better fuel/weight ratio

18 minutes ago, miklkit said:

Never DLed anything and don't know how to install anything.  Hope it comes with instructions.

Put downloaded craft files in "Kerbal Space Program/saves/{your save}/SPH" Then it should show up alongside your own planes in the spaceplane hangar. Hopefully you're on PC since I don't think console players can use downloaded craft files.

[18Watt]

1 hour ago, miklkit said:

everyone likes stacked strakes.  Dunno why

I’m not sure either.  As far as I know there is no real advantage to using a bunch of strakes to replace one wing.  But I could be wrong.

I normally put some effort into keeping my part count as low as possible, to help my computer keep up.  So using a bunch of strakes instead of one wing section is counter-productive for me.  One forum member who I’ve seen use strakes heavily and creatively is @camacju, some of his creations have insane performance.  So perhaps I should look into using strakes more..

I do use strakes to add ‘just a little more lift.’  Sometimes I end up needing a little more lift, but don’t need another full set of wings.  The strakes are easy to tack on.  Although if I need a lot more wing, I’m not ashamed to just add another full set of wings.  I’ve ended up with a lot of bi-plane SSTOs.  A bi-plane SSTO is brutally ugly, not elegant looking at all.  But if it works…

1 hour ago, Zacspace said:

They do with the mod Editor Extensions Redux.

Hah!  I’m laughing at the image of myself, scratching my head trying to figure out how to stick the Rapiers to a fuel tank!  I ended up using an engine plate.  By the way, your craft did load in my stock install.  I got one error message, something about ‘module waterfall’ something something not available.  Not sure if that’s related to the mod you mentioned or not, but the craft did load just fine, and seemed to function properly.  I took note of your landing gear settings because the landing gear seemed to work extremely well.  You went a different direction than I usually go with gear settings, so I’m taking notes on your approach, which again worked really well.  Oh, I do play 100% stock with MH and BG.  I think the only MH or BG parts in your ship were either on the rover, or related to rover deployment, so your ship should work fine in a 100% stock game.

[Lt_Duckweed]

37 minutes ago, 18Watt said:

I’m not sure either.  As far as I know there is no real advantage to using a bunch of strakes to replace one wing.  But I could be wrong.

One big-s wing is .5 tons, 5 wing area, and 300 fuel storage.

5 big-s strakes is .5 tons, 5 wing area, and 500 fuel storage.

In other words, each pair of big-s wings that you replace with 10 strakes saves you the need for either: 1 mk1 tank (250kg dry mass) or 8 mk0 tanks (200kg dry mass)

 

Another way to think about it, each strake replaces 2 mk0 tanks (the absolute best liquid fuel tank) so it is essentially 2 mk0 tanks (50kg) + a half mass wing (50gk per area instead of 100kg per area)

Edited March 3, 2022 by Lt_Duckweed

[Lt_Duckweed]

Not to toot my own horn too much, but if you want to learn nor about how lift and drag is calculated, and how wings work in ksp, I highly recommend you check out my 2 videos:

 

Then, you will want to check out F00FlGHTER's ssto tutorial: 

Between the 3 videos, you will get a pretty solid understanding of how the KSP aero model works, how to optimize for it, and how to implement those optimizations into a spaceplane ssto.

Edited March 3, 2022 by Lt_Duckweed

[miklkit]

Lots of information.  Making small changes one at a time.

[18Watt]

5 hours ago, Lt_Duckweed said:

I highly recommend you check out my 2 videos:

Awesome!  I’ve been playing since 0.18, about 10 years I guess.  I realize the atmospheric model has changed a few times, but I’ve never seen a better explanation of how drag and lift is handled in KSP.  Very nicely done videos!

[miklkit]

Been testing ideas on a new plane with mixed results.   One question is how to angle the wings properly.  I think I got them too angled.   It flies ok like this, but differently.  It needs lots of altitude to do its best.  This seems to be my MK3 basic design now.

[18Watt]

Turn off angle snap to rotate things less than 5 degrees at a time.

Angle snap is the button that normally looks like a hexagon in the lower left of the SPH or VAB screen.  If you click on it, the hexagon will change to a circle with a dot inside, meaning angle snap is no longer active.

With angle snap inactive, you can adjust the incidence of your wings more precisely.  I usually get the best results with incidence angles of 2 - 3.5 degrees.  Or even no incidence at all.  You may need to try different values to see what gives you good performance.

I can’t really tell from your screenshot what angle your wings are at, but it does look like more incidence than I would normally use.

Edit- I have to mention several rear-facing nodes you have left uncovered (it looks like).  The ‘stinger’, or tail section looks bare, as well as two 1.25 m stacks next to the fuselage.  I do recommend adding rear-facing nose cones to those nodes.  The added weight is negligible, but the reduced drag will be a big help, with compounding gains.  You’ll get past the supersonic drag hump faster, and probably be able to reach a higher speed on Open-Cycle too, all of which will either reduce fuel burn or increase payload, however you want to look at it.

Edited March 4, 2022 by 18Watt

[Zacspace]

3 hours ago, miklkit said:

One question is how to angle the wings properly.  I think I got them too angled.

2 hours ago, 18Watt said:

With angle snap inactive, you can adjust the incidence of your wings more precisely.

I gotta shout out Editor Extension Redux again here. It lets you select a bunch of different angles to snap to, even just 1 degree at a time,  instead of just being snap or don't snap.

On all my good SSTOs the wings have either 1 or 2 degrees of incidence. The one I linked earlier had 1. The sweet spot for wing incidence should keep your fuselage facing directly into the airstream to minimize drag while still producing enough lift with your wings to ascend at an agreeable rate.

[miklkit]

How do you find out what the actual angle is in degrees?  I read somewhere (or saw it in a video) that holding shift while clicking on the adjuster gives it fine angle adjustments and that is what it is set to.  It does "feel" pretty steep.  It seems to help at low altitude and speed but faster and higher it feels draggy.  Dunno how to make all wings the same without snap to.  Guess I will have to look into that editor gizmo.

 

@18watt  Those are NERVs that have the small cone in them.  There are 5.  2 on each side and 1 in the tail.  The Rapiers have cones in them also. 

[camacju]

On 3/2/2022 at 2:19 PM, 18Watt said:

5 Nukes is a lot, they are really heavy.

5 nukes for 8 rapiers is actually a little low. I tend to use between 2/3 and 1 nuclear engine per rapier, so I would suggest adding another Nerv. The reason is that you don't have much TWR on the Nerv so you do need quite a few for a large craft.

On 3/3/2022 at 11:08 AM, miklkit said:

On first glance I noticed on those optimized SSTOs is that everyone likes stacked strakes

The Big-S wing parts in general have good heat tolerance plus a "free" fuel tank (zero extra dry mass). The strake in particular is useful because, as Duckweed pointed out, it has a better fuel capacity per mass than the large delta wing, but also because you can fine-tune wing area more easily (increments of 1 wing area rather than 5). 

5 hours ago, 18Watt said:

Turn off angle snap to rotate things less than 5 degrees at a time.

Smooth angle snap isn't really necessary unless you are squeezing out the very last drop of performance from a design, and then as you go further down the optimization rabbit hole, you begin to find that lower wing incidence can perform better in some situations. So I think just a 5 degree incidence is perfectly fine until you go very far into optimization land.

(If you're not doing some funny stuff with fairings as your craft's hull, you always want a 5 degree incidence, since that gives the best lift to drag ratio. If you are doing funny stuff with fairings then technically -22 degree wing incidence is "best" but that's probably outside the scope of this thread)