Plane veers to the left. Stumped.

[jsisidore]

1 hour ago, Daveroski said:

Despite being lumpy and camel like in it's gait on the ground it does lift of rather easily.

I had to stick a big SAS behind your huge tank of Monoprop in the boot and a smaller one behind the Shielded Docking port on the front.
I also moved all the wings back a bit so that your COL was behind your COM. As you drain fuel from that big front tank your COM falls backward. If you use that fuel and then come back into an atmosphere you will be wondering why it is flipping oven and over on the way down. Having fuel up front is good.
I think your ratio of Nirvs to Rapiers is a bit too high to go hypersonic in this craft. I would lose a couple of Nirvs on each wing and replace them with rapiers making it 8 Rapiers to 4 Nirvs. Then you might be able to fill up those mid tanks on each wing so the extra Rapiers could use the fuel.

Oh.. and I found it took off better if I disabled the steering in the rear and front wheels.

Thanks! I'm eager to see how it will behave with these mods.

Although:

1. Since when heavy gears have steering? I must have missed that. That means I can finally use them for the front steering. But disabling all steering is radical.
2. Bigger SAS makes difference? Embarrassing to discover it only now. It is "inside" the cockpit then?
3. I thought it was a rule of thumb to set the col slightly ahead of com.

Engines wise, I was unable to escape Kerbin's gravity with only 6 nervs, although I had a different design the twr was more or less the same. Keeping in mind how "camel like" and "lumpy" it is I have to rely on raw power.

[Tex_NL]

3 minutes ago, jsisidore said:

Keeping in mind how "camel like" and "lumpy" it is I have to rely on raw power.

If you want raw power you do NOT want NERVA's. Plenty of stock engines provide a lot more thrust for a lot less weight.

[jsisidore]

55 minutes ago, Daveroski said:

 

Oh my.. I almost forgot.. The Elevons.

They are trying to turn the craft because the craft is moving slowly down the runway.
The runway cambers slightly towards the hangar making your craft want to go right.

Fascinating. Do devs know this?

1 minute ago, Tex_NL said:

If you want raw power you do NOT want NERVA's. Plenty of stock engines provide a lot more thrust for a lot less weight.

I need buckets of delta-v

[jsisidore]

1 hour ago, Daveroski said:

I think your ratio of Nirvs to Rapiers is a bit too high to go hypersonic in this craft. I would lose a couple of Nirvs on each wing and replace them with rapiers making it 8 Rapiers to 4 Nirvs. Then you might be able to fill up those mid tanks on each wing so the extra Rapiers could use the fuel.

Truth be told I'm struggling with calculating how much thrust I need to escape kerbin's gravity, "do I need more fuel and more jets" or "do I need more heavy nukes and subsequently less fuel". I was never good at mathematical planning.

[jsisidore]

Back with results.

Oh wow the big reaction wheel is a game changer! Plane still veers to the right but thanks to sas I don't have to lift a finger.

No, heavy gears have no steering, you must have meant the "stabilizer" gear at the back. 

It does lift quite well, I usually go all the way to the end of the runway, must be the frontal wings.

 

[Daveroski]

31 minutes ago, jsisidore said:

Thanks! I'm eager to see how it will behave with these mods.

Although:

1. Since when heavy gears have steering? I must have missed that. That means I can finally use them for the front steering. But disabling all steering is radical.
2. Bigger SAS makes difference? Embarrassing to discover it only now. It is "inside" the cockpit then?
3. I thought it was a rule of thumb to set the col slightly ahead of com.

Engines wise, I was unable to escape Kerbin's gravity with only 6 nervs, although I had a different design the twr was more or less the same. Keeping in mind how "camel like" and "lumpy" it is I have to rely on raw power.

1. The small one in front and the small one in back.

2. No, I try to avoid any clipping in all my builds. I put it where your large monoprop tank is and put your tank back onto that.

3. You are putting your cart ahead of your horse there Ginger.. as a rule things tend to work much better if you can see all of the little blue ball on the COL indicator but it should be aft of your COM.

1 minute ago, jsisidore said:

Truth be told I'm struggling with calculating how much thrust I need to escape kerbin's gravity, "do I need more fuel and more jets" or "do I need more heavy nukes and subsequently less fuel". I was never good at mathematical planning.

When you build a Spaceplane, no matter what size it is, your first objectives are building one that can get into orbit and get back again.

Soooo.. getting To orbit you only need enough oxidiser to get to beyond the atmosphere  to a point before your AP where the Nirvs can have enough time to circularise before you run out of sky.

A good test for this is an ore tank or two. (grab an ore tank as your first item and engineer's report will tell you how much your craft , the ore tank weighs. Add ore, add weight)

Getting back again means moving any and all remaining fuel to the front of the ship so that it is front heavy and so doesn't feel as needy about going down tail first. This is why a lot of people experience flipping on re-entry.

When you design a plane and you are testing it, Be a Beast! make it burn a bit more than you usually would when hitting the thick stuff.

When you are designing it remember the quirks of KSP. Open nodes cause more drag than they have any right to. Adding the weight of a nosecone to the back of an engine and then using the slidey too to slide it inside the engine so that you can't actually see it any more.(I hate clipping but reserve the right to be a hypocrite when circumstances are in my favour) It will bring the drag down to a more manageable range and some of those craft you were scratching your head over might actually perform a bit better.

Any road up.. you don't want to have to read the book that this is turning into... so crack on!

Chocks away Ginger!

 

[jsisidore]

6 minutes ago, Daveroski said:

1. The small one in front and the small one in back.

2. No, I try to avoid any clipping in all my builds. I put it where your large monoprop tank is and put your tank back onto that.

3. You are putting your cart ahead of your horse there Ginger.. as a rule things tend to work much better if you can see all of the little blue ball on the COL indicator but it should be aft of your COM.

When you build a Spaceplane, no matter what size it is, your first objectives are building one that can get into orbit and get back again.

Soooo.. getting To orbit you only need enough oxidiser to get to beyond the atmosphere  to a point before your AP where the Nirvs can have enough time to circularise before you run out of sky.

A good test for this is an ore tank or two. (grab an ore tank as your first item and engineer's report will tell you how much your craft , the ore tank weighs. Add ore, add weight)

Getting back again means moving any and all remaining fuel to the front of the ship so that it is front heavy and so doesn't feel as needy about going down tail first. This is why a lot of people experience flipping on re-entry.

When you design a plane and you are testing it, Be a Beast! make it burn a bit more than you usually would when hitting the thick stuff.

When you are designing it remember the quirks of KSP. Open nodes cause more drag than they have any right to. Adding the weight of a nosecone to the back of an engine and then using the slidey too to slide it inside the engine so that you can't actually see it any more.(I hate clipping but reserve the right to be a hypocrite when circumstances are in my favour) It will bring the drag down to a more manageable range and some of those craft you were scratching your head over might actually perform a bit better.

Any road up.. you don't want to have to read the book that this is turning into... so crack on!

Chocks away Ginger!

 

Cheers. Like I said bad at planning, good at flying. Btw I'm blond and pale, so you are not far off.

[Daveroski]

1 minute ago, jsisidore said:

Cheers. Like I said bad at planning, good at flying. Btw I'm blond and pale, so you are not far off.

Ahhh that.. We were read to from a book at school about Biggles he was a pilot. Ginger was his mate.

 

[AeroGav]

On 29/11/2016 at 2:20 PM, jsisidore said:

Sorry wrong design, this is the old version. This won't escape Kerbin's gravity as it needs more nerv engines.

This is definitely it:

http://pastebin.com/VjHM7Zb4

Jeez Si,   I downloaded that thing 24 hours ago and have now gone through almost a full jar of coffee.

As regards to the handling issues you originally posted about :

• Vertical tail surfaces are still too small IMHO.  Would fly better with more area.  
• Those ailerons are enormous.   You probably only want to make small corrections to stay on heading 90 deg. East when flying to orbit, so i'd right click on them and reduce the authority to 25-30%.     At small deflection angles they make a little bit less lift and a LOT less drag, which in turn means less unwanted Yaw when you try to correct a roll.
• Centre of lift too far forward in the uploaded ship, i'm sure you know.

I also spotted signs that your spaceplane suffers from excess drag.   Eight nervs is way overkill to bring a CRG 50 to orbit, but you are carrying a buttload of oxidizer as well.     I see this approach a lot in people who are coming over from rocket design - moar boosters , then moar fuel, then moar boosters.     Sort out the lift and drag first, and the fuel and engine problem becomes way easier.

Easy to Fix Drag Issues

1. No 1.25m cones on the back of engines.   See the first 80 seconds of this video if you want an example of how to do this. https://www.youtube.com/watch?v=XXhIb5OlSRY&t=605s
2. It seems you got a "sandwich" of 3 FT800 tanks on each wing.    The upper and lower tanks each carry a triple engine mount, i was quite impressed with the way you did that tbh.   However the one in the middle has nothing on the back.  The fuel tank ends as a flat plate, even though you took the time to put a cone on the front.   Cones = good.  Notice the pattern?
3. Fuel ducts in cargo bay.  Mega drag ! Not needed since 1.2,  they create more drag than cockpits.  Get rid.
4. Shielded docking adapter nose cone.  Since 1.2 this has become a draggy part.   Basic nose cone or intake would be better unless you really must have docking.

Stage two drag reduction (more engineering effort)

1. Wings can be angled up by 5 degrees so the fuselage flies pointed straight into the wind.

2. RCS.   Those 4 way thruster blocks are very draggy and have poor heat tolerance.  Offsetting them inside the fuselage doesn't reduce the drag !   Single port thrusters are more powerful, have better heat tolerance and much less drag.   But the lowest drag solution of all is to skip RCS and just put some strong reaction wheels in the cargo bay.   If you really need the ability to translate up/down/sideways for easy docking, then mount all your RCS ports on the cargo bay itself.   They won't work when the cargo bay is closed (even if you mounted them on the outside) but they won't generate any drag either when the bay is closed (even if on the outside).

 

Aerogav looses the plot, disappears down a rabbit hole

 

Anyway your SSTO piqued my interest because I don't have a cargo plane with a ramp.   Also you have a 2:1 ratio of nukes to rapiers which is what you'd go for in an oxidizer free interplanetary design.

Now, those 1.25m tri adapters are draggy, i wanted to know the best way to mount three engines.  So , I built 3 rockets

The first had a mk1 pod, an ft800 tank, a 1.25m tri adapter and three "Dart" aerospike engines.    Launched straight up, it reached an AP of 125km.

The second had the smallest 2.5m fuel tank which has the same capacity as an ft800.  This required a 2.5m to 1.25m adapter so i could use the same pod.   And I still used three aerospikes , but the 2.5m tri adapter is known to have lower drag than the 1.25m version (it looks sleeker too).   This version reached 144km.

The last version had a mk1 pod, an ft 800 tank , and one aerospike on the bottom of the tank, as well as two more radially mounted with the aid of type B nose cones.   This one did best of all and reached 151km.

 

Long story short, i decided to try making a version of your plane with every drag reducing trick i know applied to it.   The rocket tests showed that ideally, i shouldn't use any tri adapters,  but mounting all those engines without them proved a headache.  

Fateful decision #1 - I got rid of the adapters by just mounting half as many engines and "hoping for the best".

Fateful decision #2 - that huge, enormous fuselage tank in front of the cargo bay adds massively to the weight and also causes a large shift in CG as it burns off.   I got rid of it, but that made the plane too short and stubby.   I had nowhere to put the strakes or canards.  So I added another cargo bay instead.

A few hours later - woaah, this think looks like it's doing 1000mph standing still ! 

Appearances can be deceptive, unfortunately.   20 minutes to climb to 10km,  where the air is finally thin enough that two Rapiers will push it past mach 1.

It was very stable and flys hands off with prograde only.  But as you can see I got rather bored and resorted to playing musical chairs to pass the time, moving the Kerbals around various locations in the cavernous fuselage.

It did in fact reach orbit, with 550 fuel remaining, but I was disappointed.  Not interplanetary, very slow climbing to orbit, and only a 2 ton rover on such a big plane is also a poor payload fraction.

Time for a group huddle.

 

Gross weight 64 tons

16 tons is engines

 22 tons is liquid fuel

Airplane stuff comes to about 9 tons (wings, gear, control surfaces etc)

3 tons of payload (2 ton rover plus 1 ton of monoprop)

Cargo bays, ramps, cockpit - 14 tons.

When you break it down like this, it's all rather obvious.  Fuel is not much more than a third of it's weight, so it's hardly surprising it doesn't go interplanetary.  That it even gets to orbit with such a low fuel fraction is testament to the design's economy.

The problem is the massive overhead posed by the mk3 cargo bays and ramp, which add 10 tons even when empty, plus the 4 ton mk3 cockpit.   On a 6 engine, 60 ton design this swallows most of your useful payload.    

@jsisidore ,  you had the right idea all along.  4 rapiers , 8 nervs and double up on the fuel.   Keep the cargo bay small, make the ship twice as big, and the  penalty of those cargo bays is more bearable.    It's all very well having a fine eye for detail to get good aerodynamic performance, but you can't let the basic mass ratios get out of shape while obsessing over drag.

But anyway, after putting hours in this version of the ship I just wanted some quick fixes to make it presentable.  Werner von Kerman had an idea..   

OK, back to work guys..

Panther boosters

Takes less than half the time to reach mach 1, and uses less fuel getting there.  They are on decouplers so ;

 

1218 LF remaining at an 82km orbit.   It now takes off, passes through the sound barrier and accelerates from mach 4.3 to orbit rather easily, so i reckon some extra fuel tanks would make it 100% interplanetary.  Unfortunately we can't use empty front part of the cargo bay as thay will mess up CG.  Underwing drop tanks?

Well, not today any rate.  I need a lie down. 

https://kerbalx.com/AeroGav/platypus-lifter-b

[jsisidore]

Hello AG. Quite a post. Kudos on reaching the orbit, I'm yet to break 10km and mach1. Too much dead weight can't argue with that, twr is horrible 0.5 need some 0.7 to at least have some hope.

Aerodynamics advices will be noted and implemented.

I need the adapter to refuel at Dune's mining facility which is already there waiting for me. 

Angling wings makes me nervous as I have a double wing individually connected to the fuselage.

RCS was a backup plan in case I will lose stability in Dune's atmosphere before touchdown, but now that I have biger SAS I might be alright without the RCS.

What I need is another two jets or less weight. Vaguely I can imagine I need more of everything to make the boot less noticeable.

[AeroGav]

1 hour ago, jsisidore said:

Hello AG. Quite a post. Kudos on reaching the orbit, I'm yet to break 10km and mach1. Too much dead weight can't argue with that, twr is horrible 0.5 need some 0.7 to at least have some hope.

Aerodynamics advices will be noted and implemented.

I need the adapter to refuel at Dune's mining facility which is already there waiting for me. 

 

RCS was a backup plan in case I will lose stability in Dune's atmosphere before touchdown, but now that I have biger SAS I might be alright without the RCS.

What I need is another two jets or less weight. Vaguely I can imagine I need more of everything to make the boot less noticeable.

 

Lift to drag ratio is king.

 

The better this is, the less rapiers you need to get past mach 1.   This is important because Rapiers are heavy, 2 tons each.  You'll be dragging them as payload from about 25km to Duna and back.   Above mach 1, the ram air effect boosts rapier so much, any plane that can get past mach 1 is limited in how fast it can go by heat.

Good lift:drag ratio also matters once you get above jet altitude.   If you have good lift and low drag, you do not need any oxidizer, one nuke per 15 tons ( a TWR below 0.5) is plenty.  

As you get closer to orbit, the effects of gravity are increasingly countered by the orbital free fall effect.  This means the amount of lift you need is also decreasing, which is just as well as the air is getting thinner and thinner as you cilmb higher.    Lift/Drag ratio is fairly constant from mach 1.5 to orbit, so less lift means less drag - in practice, this means as you're getting higher and lift is decreasing, so does drag because the air is so thin.

The upshot of all this is you don't need to bring loads of oxidizer or nukes, both of which are heavy, if you got good lift:drag ratio.

Landing on Duna

OK wow, you're stretching yourself in all areas here.

Duna has very uneven terrain and a very thin atmosphere.  Despite the low gravity, aeroplane landing speeds are two or three times higher than they would be on Kerbin.   This means you need plenty of wing area, very robust landing gear, and some kind of  lift thruster system. 

Take a look at my current best cargo plane 

https://kerbalx.com/AeroGav/Wyvern-Offroad

 

 

Empty weight is about 60 tons, it had very little fuel left in its own tanks at this point, but it was carrying a 36 ton orange tank.

Total landing weight 96 tons.

Lift rating is 50  So, you need 1 lift rating for every 2 tons of landing weight.

It has 12 vernier lift engines as well, inside service bays so as not to create any drag on the flight to orbit.  That works out at one Vernier per 8 tons of landing weight.

I wouldn't feel comfortable landing with less than this, as you can see from the video things get pretty exciting.

The main landing gear need to be very close to the centre of gravity.  If they are behind CG, as the wheels impact the ground the plane will tumble forwards.   If they are ahead of CG,  ground contact will make the nose shoot upward.   They Wyvern is a tail  dragger, with the mains ever so slightly in front of CG.    I figured it's safer to have the plane pitch up after coming into contact with the ground than to pitch down - if you're dropping hard onto the ground that means you don't have enough lift so pitching up will probably soften the next impact bounce.

It also has smaller gears one the nose and tail, to stop the nose or tail digging in.

 

Golden Ratios

One rapier per 30 tons takeoff weight on a good low drag design.   

Nearly 50% of the takeoff weight should be liquid fuel, if the cargo bay is empty.

If you are going oxidizer free, you need twice as many nukes as Rapiers.  If you are bringing oxidizer, then same number of nukes and Rapiers.  But keep the oxidizer to a minimum.  I'd say 220 OX per 30 tons takeoff weight is plenty.

At 120 tons , with some oxidizer, the Wyvern should only need 4 nukes and 4 rapiers.  It has 5 rapiers so it climbs quite eagerly.   And it has 6 nukes instead of 4, which means it needs very little oxidizer to reach orbit, less than the 220 per 30 ton i was recommending.

 

How do I check lift/Drag ratio?

From the game's main menu, go to settings, graphics, and make sure the "highlight fx" box is unchecked - or using the aero data displays can make the game crash.

From in game, press Alt-f12 to bring up the debug menu.  Go to the Physics Tab, the Aero.    If you enable the aero data gui checkbox, you get a big panel of stats.

Lift : drag ratio is on there.

Below 230 m/s, unrealistically good numbers are possible, but not relevant to where space planes operate.

If you don't have angled wings,  a mk3 design should get about 2.5 to 1 or so when supersonic above mach 1.4

If you angled the wings,  4.3 or even higher is possible.

Note that between 0.86 and 1.4 mach (250 m/s and 420m/s) is the high drag transonic region.   Your L/D numbers here will look bad but should improve a bit afterward, though they'll never match what you had at low speed.

Quote

Angling wings makes me nervous as I have a double wing individually connected to the fuselage.

I must admit, i'd call it an advanced technique but since what you are trying to do is hard, i think it might be necessary.

Your first wings attach to the fuselage.   Your nacelles attach to the first wings, then the second wings attach to the nacelles.

so

1. angle the first wings up 5 degrees, which affects everything.

2. angle the nacelles down 5 degrees, because you want the engines thrusting straight ahead and don't want the nacelles to create drag from being off prograde.  this undoes the angle on the second set of wings so..

3. angle the second set of wings up 5 degrees... done! 

You probably should also angle any horizontal tail planes or canards up by 5 degrees as well or the handling will get really strange.   That's it for the coarse adjustments.

 

Now you need to go into fine rotation mode and angle the canard up a little more, one degree at a time with test flights in between.  What you are trying to get is for the plane to fly with it's nose a couple of degrees above prograde with SAS off and you not touching the controls at all.    Prograde hold isn't actually perfect, it will not give you exactly 0 AoA but always either slightly above or below prograde depending on what the plane would be doing without SAS.

If you've adjusted the plane to fly 2 or 3 degrees AoA "hands off", the Aero Data gui will show you're actually at about half a degree of positive AoA  with prograde on , which is low drag.      If it flies 10 degrees nose up (for example) with sAS off, then Prograde will be more like 2 or 3 degrees nose up, at which point the fuselage is starting to make significant drag.   And you don't want the plane to fly at negative AoA because then not only is the fuselage making drag, it is also creating "negative lift".

 

 

PS.  I have not yet make a cargo Duna spaceplane with a ramp yet, partly because it's difficult, and partly because when you've got the Wyvern,  do you really need a rover?  Just pack you expedition into the cargo bay and use the airplane to explore Duna's biomes.

[AeroGav]

6 hours ago, jsisidore said:

Hello AG. Quite a post. Kudos on reaching the orbit, I'm yet to break 10km and mach1. Too much dead weight can't argue with that, twr is horrible 0.5 need some 0.7 to at least have some hope.

Aerodynamics advices will be noted and implemented.

I need the adapter to refuel at Dune's mining facility which is already there waiting for me. 

Angling wings makes me nervous as I have a double wing individually connected to the fuselage.

RCS was a backup plan in case I will lose stability in Dune's atmosphere before touchdown, but now that I have biger SAS I might be alright without the RCS.

What I need is another two jets or less weight. Vaguely I can imagine I need more of everything to make the boot less noticeable.

OK,   round two !

http://pastebin.com/V8c4KX6M

I made a new version of the aircraft, that's a lot closer to your original, in that it keeps the tri couplers and uses four rapiers and eight nukes.  It has only one CRG 50  but i did replace the normal size mk3 LF tank with the short version.   To balance out all that fuel ahead of CG,  are a big cluster of strakes acting as the tail surface,  and a sponson surrounding the aft half of the cargo bay (idea borrowed from the Wyvern) which also houses the lift jets.   As a result, CG don't shift when the fuel burns off.

First up,  I tried with wings that aren't angled.    It did get to orbit, and quicker than my "platypus", thanks to having twice the engines.  But despite taking off with nearly twice as much fuel, it has no more delta V in orbit.  Lift drag ratio was at best 1.5 to 1 at supersonic speeds. 

So, i went and angled the wings.

We're still only getting 3 to 1 whereas the Platypus version got 4 to 1.  But, thanks to having fuel tanks in place of empty cargo bays, it does perform slightly better.   Over 2550 LF remaining compared with 1211 on the Platypus.  It's heavier, thanks to twice the engines, but not twice heavier, so delta v should be better.  You can see how diminishing returns/the rocket equation punishes inefficient craft though.  Hopefully that's enough for a straight shot at Duna.

_{edit - i did a prograde burn and escaped from kerbin soi with 852 LF remaining in the tank. According to delta V maps, it takes 950 DV to leave Kerbin SOI from LKO, and only another 130 to get a Duna intercept.  Given that we can aerobrake and land dead stick, i say this shows a Duna mission is now possible.}

Why is our L/D worse?   Well, the tri adapters are making about 20% of our drag, and the nose cone another 10%  - features the platypus doesn't have.

Also, it's got marginally less wing and a lot more weight.  At the same AoA,  the wings are doing the same L/D ratio, but the fuselage parts will be going through thicker air due to it not gaining altitude as quickly.  Thus the extra fuselage drag pulls our overall L/D down.   Yeah, even with angled wings and fuselage pointing into the prograde, the body is still the main source of drag.

OTOH, it's a bit easier to keep the thing in the atmosphere longer during the speedrun.  It no longer tries to jump into space every time you exceed mach 4. Slightly better behaved airplane overall i'd say.

OH and before i congratulate myself too much on reaching orbit with 30% fuel, i took the Wyvern up for another spin just now..

 

Stuff like this makes me think the craft file is bugged.   in orbit with over 2/3 fuel remaining?

 

 

Edited November 30, 2016 by AeroGav

[jsisidore]

I'm glad you're enjoying yourself. As for me I just start to spin without any warning after 16km at 1000m/s and then of course explode. Need to rethink the whole aerodynamics shebang, maybe even bury mk3 cockpit altogether. With so many compromises might as well build boats instead.

Here's redesign: http://pastebin.com/di0FbXpJ

btw can you give me a link for that plugin which allows for precision spin?

[AeroGav]

13 minutes ago, jsisidore said:

I'm glad you're enjoying yourself. As for me I just start to spin without any warning after 16km at 1000m/s and then of course explode. Need to rethink the whole aerodynamics shebang, maybe even bury mk3 cockpit altogether. With so many compromises might as well build boats instead.

Here's redesign: http://pastebin.com/di0FbXpJ

btw can you give me a link for that plugin which allows for precision spin?

Also I use CorrectCoL which takes account of body lift when calculating your centre of lift. Left alone, it just makes the blue indicator in the SPH more accurate.  But you can bring up the GUI too for "static analysis", which tells you how much AoA you need to fly at a given altitude and airspeed ..      

 

Spoiler

 

It can help you work out if you got enough wing too.

The box on the right is where you can type the speed and altitude you want to simulate.  The vertical blue line indicates how much aoa is needed for level flight under those conditions.     The green line is stability loaded, the yellow is stability when empty.   You want both lines sloping downhill.    The X axis basically shows AoA and the Y axis shows whether the plane's is trying to nose up or down response to this.   So basically, when you're in negative AoA, you want the plane to have a natural nose up tendency to get back to an even keel, and a nose down tendency when AoA goes positive.    The point at which the line crosses the horizontal (x axis) is the AoA the plane will want to fly at if left on its own, with no control input and no sas.

 

I'll have a look at your latest version.   You are making progress, you've gone from not taking off to hypersonic, so it's getting there.

I must admit though,   the cargo bay and the ramp are a substantial drag penalty to take to Duna.  I do wonder if  a mk3 cockpit is a luxury we can't afford, though if you loose that 4 ton mass up front you're going to have to shift those engines waay forward to balance the ship.

[AeroGav]

34 minutes ago, jsisidore said:

I'm glad you're enjoying yourself. As for me I just start to spin without any warning after 16km at 1000m/s and then of course explode. Need to rethink the whole aerodynamics shebang, maybe even bury mk3 cockpit altogether. With so many compromises might as well build boats instead.

Here's redesign: http://pastebin.com/di0FbXpJ

btw can you give me a link for that plugin which allows for precision spin?

OK I can see what's wrong with your ship straight away.

You have four engine pre coolers for six rapiers.  The pre cooler, like the adjustable ramp intake, peaks about mach 3 , whereas the rapier has a bigger top end.   Even so this is not a problem provided you use at least one pre cooler per rapier.

I've found that the diverterless supersonic intake and those radial supersonic intakes are useless for rapiers.  They peak at about mach 2.5 then fade to almost nothing after mach 4,  no matter how many of them you spam.

Also i still think your vertical stabilizer is too small for such a big ship.

44 minutes ago, jsisidore said:

I'm glad you're enjoying yourself. As for me I just start to spin without any warning after 16km at 1000m/s and then of course explode. Need to rethink the whole aerodynamics shebang, maybe even bury mk3 cockpit altogether. With so many compromises might as well build boats instead.

Here's redesign: http://pastebin.com/di0FbXpJ

 

I've seen a much more serious problem though this this version.    The engines are too far aft.   That giant fuel tank right behind the cockpit balances this when full.  But once your fuel burns off the plane will be uncontrollable.  You need to shift those engines well forward, so they are close to your empty centre of mass.    Then probably get rid of that big fuel tank, or like I did, fit a smaller one up front, then something else that holds fuel towards the back.

Have you got RCS build aid?  Shows you a nice red ball in SPH which indicates where your empty CoM will be, makes this so much easier.

 

[AeroGav]

NEWSFLASH 

I added a few more intakes so you don't get asymmetric flameouts at altitude, and it goes to space.    It's not exactly efficient, the lift to drag ratios i am seing are about 1,  so you might as well remove the wings and launch it as a rocket.   But it does reach orbit  gratz! 

[jsisidore]

1 hour ago, AeroGav said:

NEWSFLASH 

I added a few more intakes so you don't get asymmetric flameouts at altitude, and it goes to space.    It's not exactly efficient, the lift to drag ratios i am seing are about 1,  so you might as well remove the wings and launch it as a rocket.   But it does reach orbit  gratz! 

Any fuel left? I would like to have a look at it. How do you calculate the ratio? Another plugin?

[foamyesque]

12 hours ago, jsisidore said:

Any fuel left? I would like to have a look at it. How do you calculate the ratio? Another plugin?

Alt-F12 to pull up the debug console, and then go to the Aero tab and turn on "Display Aero GUI". It'll give you a plethora of aerodynamic info.

[jsisidore]

1 hour ago, foamyesque said:

Alt-F12 to pull up the debug console, and then go to the Aero tab and turn on "Display Aero GUI". It'll give you a plethora of aerodynamic info.

Oh right! That's what AG told me, sorry.

Okay, before I start redesigning for less drag and more lift, a couple of questions.

My L/D ratio began at 2.5 then fluctuated between 1.1 and 1.3 - what does it mean? What ratio should I look for?

At around 27km and 1350m/s, just before I took this shot, the nose began to rise up progressively and I think you know what happened next... was that the fuel going back into the tail of the ship?

[foamyesque]

@jsisidore: I suspect (can you post the craft file?) that what's happening there has one of two causes.

 

1. The nuclear engines have burned off fuel towards the front of the craft, which moves the CoM backwards, making it unstable;

2. Your CoM has moved upward, and the combination of now-offset, non-gimballed thrust and thin air sapping the stabilization from aero surfaces causes it to start rotating.

 

Diagonising which is happening is tricky to do from a screenshot. I usually use KER for supplemental info on that.

[AeroGav]

1 hour ago, jsisidore said:

Oh right! That's what AG told me, sorry.

Okay, before I start redesigning for less drag and more lift, a couple of questions.

My L/D ratio began at 2.5 then fluctuated between 1.1 and 1.3 - what does it mean? What ratio should I look for?

At around 27km and 1350m/s, just before I took this shot, the nose began to rise up progressively and I think you know what happened next... was that the fuel going back into the tail of the ship?

Are you playing on 1.2 or later?   If that is the case, then the NERVs should drain all tanks evenly so that should not be an issue.

I did not switch the RAPIERs to  closed cycle mode when i flew your ship so possibly i held on to more fuel mass until out of the atmosphere.Your ship has engines quite far back, with no fuel it is tail heavy and unstable.   The fuel tanks are mostly at the front and make the plane stable, so long as they have plenty of fuel in them.

I can tell from the screenshot what is about to happen.

You have SAS set to stability hold, and the nose is above prograde.  If that thing was stable, it's nose would have a strong tendency to return to prograde so you'd see a large nose-up input (in the bottom left corner) being applied by SAS just to maintain your nose up attitude.  Instead it is working very hard to stop the nose rising further.  Once it runs out of elevator authority it's going to flip around and fly backwards.

As a band-aid,  try locking the forward most fuel tank (the one behind the cockpit) and only turn it back on when in space (or  the engines cut out)

Lift to drag ratio,   by adding more wing you might get it up to 2 or so in supersonic speeds.

By angling the wings you'll get more,  2.5 or 3?

Do both and 4 might be possible.  Highest i ever saw was about 4.3.

 

[jsisidore]

I see. Btw how do you activate that Editor Extensions plugin? I can't find it anywhere in game, does it have gui?

[jsisidore]

Would hiding the docking port in a protective shell work?

[jsisidore]

Hmm I tried to install the redux version still same results, I knew this would be extremely frustrating.

[AeroGav]

14 hours ago, jsisidore said:

I see. Btw how do you activate that Editor Extensions plugin? I can't find it anywhere in game, does it have gui?

If it installed correctly there should be an EEX icon in the bottom right corner of the SPH, like there is in this screenshot.  Mine installed automatically via CKAN so no worries there.

13 hours ago, jsisidore said:

Would hiding the docking port in a protective shell work?

I don't know because i haven't got much experience of procedural fairings, I find them hard to work with.

Docking vehicles on the ground isn't easy.   I tried it a couple of times, on Minmus's flats, and it wasn't easy.  One of the vehicles was now full of fuel and the other one empty,  the difference in weight meant it sat slightly lower on its springs. In space the magnetic attraction rotates the vehicles until the ports line up perfectly, but when they are on the ground stuff is not so free to move.

 Duna has lumpy terrain and more gravity, so getting the two to line up might be even harder.   You might have more success putting a claw on your base, but what i'd do is take advantage of the fact you got a docking port inside your cargo bay along with a nice flat surface (the bottom of the cargo bay).   I'd make my first vehicle payload out to Duna a tanker truck,  to shuttle fuel between your mining base and the SSTO. 

Alternatively you could install Kerbal Attachment System (which requires Kerbal Inventory System).  It allows you to put filler caps on your creations,  and allows Kerbals on EVA to link two filler caps with a fuel hose.   Duh ! Why isn't that stock ! 

KAS / KIS  can be used for an awful lot more than that, the problem is the size/complexity scares me a little. I'm happy using tiny mods like Editor Extensions and RCS build aid to create better stock vehicles, I tend to stay away from stuff that adds extra parts (filler caps and fuel hoses), since i want people to be able to use my craft without having to install mods on their game. 

Edited December 3, 2016 by AeroGav