Thinking about making the switch to FAR.

[NathanKell]

Jet Isp is meaningless because intakeAir also counts towards Isp. in effect you get 16x the Isp it says you do--which means those 1800s Isp jets are in fact 28,800s (!) regarding the fuel that's actually your own, not the stuff you generate while flying.

For this reason, and because jet thrust doesn't decrease with altitude at all, it pays to get as high as you can as fast as you can, and that's why I don't even bother looking at jet delta V, it's basically meaningless. Make sure you have enough LF for however long it takes you to get up to speed, and a bit extra if you miss KSC on return, and that's it.

[capi3101]

Alright - new design:

Let me guess what y'all's advice is going to be before you make it -

1) Fill in the gaps between the external tankage and the fuselage with wing connectors, both fore and aft.

2) Move the CoM forward.

3) Move the CoL a shade closer to the CoM.

4) Fix the funkiness with the entire tailplane assembly. Maybe increase the sweep of the tailplane or loose it entirely.

5) Reduce the wingspan and length of the tail because reasons.

6) It doesn't look enough like a plane because reasons. Make it look more like a plane. The MiG-21 doesn't count.

This plane made it up to about 45k/1900 m/s in rocket mode before I lost control of it; plane broke up due to aerodynamic failure as I was attempting to recover from the resultant spin. I'm guessing that was the point at which the control surfaces could no longer control the inherent engine torque RCS Build Aid was telling me about.

Ultimate hope for this design is eight tonnes in the cargo bay or thereabouts; this flight I was empty so it had a little more oomph to it.

So - was I close? Yes, I was being facetious on #5 and #6; call that a bit of frustration creeping in...

[UnusualAttitude]

A few ideas:

-Keep the tailplane, but having it clip into the fuselage structure is probably a bad idea. Have you tried a T shaped tail (horizontal stabiliser at the top of the fin)? They're great for improving stability. Also, does that horizontal stabiliser have pitch control ? If not, you will be severely lacking pitch authority as your only control surfaces are close to your CoM. I'd say that's why you lost it...

-Your engine torque looks reasonable. That shouldn't be a problem with SAS.

-If you reached 45k then you're through the hard part. You should be able to just use hold prograde and push your apoapsis. Try it, it works.

-you're nearly there, never give up !

- - - Updated - - -[/CO

Edited April 5, 2015 by UnusualAttitude

[Wanderfound]

You should be able to get that torque figure down to <10kN just by shifting the lateral tanks and engines up or down a whisker. Expect it to take a bit of fiddling, though. Tweaking tailfin mass is also useful for fine-tuning thrust torque.

As for the rest...yeah, basically. CoM as far forwards as you can get it, make it look like a "normal" aircraft. Yes, the Mig-21 was a tailed delta, but it's a very uncommon planform. You should be able to get it flying okay with sufficient fiddling, but a more common wing design would make it easier.

You can make unusual aircraft work in FAR, but it's a lot harder than doing things more conventionally. Start by getting the basics locked down, then get fancy. As the cliche goes, learn to walk before you try to run.

I wouldn't be pushing the CoL any further forwards than it already is, however. Unlike in stock, you really don't need to have your CoL right on top of CoM, and there are often good reasons not to. So long as you have sufficient pitch authority to lift the nose when you want to, rear-biased CoL isn't a problem.

[Van Disaster]

CoL vs CoM is an continual argument I put CoL on top of CoM, but I set my tanks up so CoM stays still - there's really not a reason to seperate them, I think. If those are all-moving horizonal tail surfaces I'd probably set CoL up without them.

I would attach the horizontal tail parts to the base of the vertical fin so they're not clipping anything, otherwise... yeap looks ok. Mebbe use b9 control surfaces for whatever those bits on the main wing do. You'll probably get some huge CoL shift with the length of that delta at Mach1 so if that's a problem just make it shorter & wider, otherwise if the wing loading is ok ( so you're not pulling 20 deg AoA at 25km ) that looks fine. I wouldn't bother filling the gaps in personally.

[capi3101]

I think tonight I'll ask you guys about how to build a proper rudder...

Up until now I had been building a rudder/fin out of a moving wing, but y'all have been recommending building a T-tail or at least putting the tailplane up on the fin for a while now, and so I decided to give building a fin with a dedicated rudder a shot. This was my first try. Doesn't look right.

Changes to the Bigger Waste of Time were otherwise reasonably successful; FAR greenlit the redesign at 30k/Mach 4 and the torque was down to a little over 1 kNm. Damn thing failed to reach orbit because (of all damn things) the engines overheated and exploded when it switched to rocket mode (missed it because the FAR flight data was covering that part of the screen). I was able to maintain control long enough to get the plane over land (1.2 degree grade grassland) and get the horizontal speed down to 80 m/s and vertical speed to -5 m/s; still lost the plane on landing.

I usually equate engine overheat to having the engines trying to handle too much mass or there being too few heat-dissipating parts. Which one do y'all think it was?

Edited April 6, 2015 by capi3101

[Van Disaster]

Just put a short piece of something structural ( even a tank ) just ahead of the engines, it's a KSP-ism that we call a bug & devs apparently call a feature. IIRC it's because the centre of mass of the part in front is too far away. Presumably fixed with 1.0 heat changes.

Tail looks ok, make the tip of the fixed bit shorter & keep the same offset, that might look better. I think there's something up with your B9 pwings though, that pattern usually means the shader is borked.

[capi3101]

Right - so stick the Engine Nacelles back on in front of the engines and then re-balance the torque. Got it.

Running a fair number of mods on a box with only 4 GB of memory, 2.9GHz Dual Core and an on-board card whose equivalent was high-tech about eight years ago. I'd upgrade my box (I want to be able to play Star Citizen when it comes out) but there's this annoying thing called "life" that refuses to stop happening.

[Wanderfound]

Make the bottom of the rudder parallel to the top and it'll look fine. Make sure the rudder is set to yaw only. Set the outer wing surfaces to influence roll only; set the inner wing surfaces to be flaps only. Increase the flap deflection a bit. Probably crank up the pitch influence of the stabilator, too.

Next up is to tune the control surfaces; take it up at a few different speeds and test the control surfaces all the way to full stick. If it struggles to obey your inputs, increase max deflection on the relevant control surfaces. If it flips out of control, reduce max deflection. You'll often need a bit of excess low altitude authority in order to have enough in thin air, so test high as well as low.

While you're doing that, check your tanks to be sure they're draining evenly and in the order you want. RAPIERs are just prone to overheating, especially when grouped. The easiest solution is often to just cut throttle to 85% or so when they switch mode.

Edited April 6, 2015 by Wanderfound

[capi3101]

Cool - I'll try those out those tweaks on the BWOT when I get a chance.

I'd like to digress from my designs for today - yesterday, another forum member posted this FAR plane design:

I went ahead and cloned it and tried flying it around a bit last night. Only real change I made was adding a bit of dihedral to the wings.

Now, from what y'all have told me - and this is what I told them - that design has A) too many engines and insufficient pitch authority. My general advice was to cut the number of engines to two and add a set of canards up front (I suggested AV-R8s high up on the fuselage; high up so they wouldn't be in-line with the main wing). Was that good advice or not? I've read that canards are de-stabilizing surfaces...

My experience with the plane was that it had really lousy pitch authority, a high takeoff speed (150 m/s before it even wanted to start nosing up), a tendency to tear itself apart on takeoff due to excessive Q, and that nasty un-controllable pitch up there at 20k/Mach 3. FAR also predicted the design would have L_ß problems at 30k/Mach 4 - not enough fin/rudder, but I figured the design wouldn't make it past the M_w instability at 20k/Mach 3, so I didn't bother trying to fix it...

What's involved in designing a good tail-less delta, in general?

[Crzyrndm]

Advice sounds about right, although I would have added excessive CoM/CoL separation and small tail. Canards are destabilising I assume is referring to the forward shifting of CoL

What's involved in designing a good tail-less delta, in general?

Well, the main issue with a straight delta (single wing: no tail plane or canards) is normally how far back the CoM gets and excessive separation between Com and CoL. That in turn leads to low control authority in pitch and yaw. Following that, my focus with delta's is always shifting mass forward and increasing lift/wing area (extra lifting surface keeps AoA lower and improves control).

[Wanderfound]

Now, from what y'all have told me - and this is what I told them - that design has A) too many engines and insufficient pitch authority.

It has more engines than the minimum required; whether that is "too much" depends entirely on the intentions of the builder. The difference between a just-flying economy cruiser and an overpowered sportster can be a fifteen minute saving in time to orbit. It's also quite fun to be able to climb vertically if desired.

That airframe could provide sufficient pitch authority, but you'd need to crank up the max deflection on the elevons close to maximum. Control authority is basically max deflection x control surface area x distance from CoM. The rearward engine and wing placement of a delta tends to give them rear-biased CoM, reducing the last factor in that equation.

My general advice was to cut the number of engines to two and add a set of canards up front (I suggested AV-R8s high up on the fuselage; high up so they wouldn't be in-line with the main wing). Was that good advice or not? I've read that canards are de-stabilizing surfaces...

My experience with the plane was that it had really lousy pitch authority, a high takeoff speed (150 m/s before it even wanted to start nosing up), a tendency to tear itself apart on takeoff due to excessive Q, and that nasty un-controllable pitch up there at 20k/Mach 3. FAR also predicted the design would have L_ß problems at 30k/Mach 4 - not enough fin/rudder, but I figured the design wouldn't make it past the M_w instability at 20k/Mach 3, so I didn't bother trying to fix it...

What's involved in designing a good tail-less delta, in general?

Adding canards would ease the pitch authority issue, but it could also induce involuntary pitch-up instability problems at altitude. The take off problems are just a matter of having the gear too far back; shift them forwards to just behind CoM, and raise them on hardpoints if necessary to avoid tailstrike.

The main problem with tail-less deltas is related to that rearwards CoM. The horizontal and vertical stabilisers just don't have enough leverage. So, get the CoM as far forwards as you can, and make sure that the vertical stabiliser/s are decently sized and that there's a fair bit of wingspan right up the back. Compound sweep wings (i.e. reducing sweep and increasing span towards the rear; stick some strakes on the front of your wings) often work nicely on deltas.

Re: canard instability. Think about what happens when the nose of a plane starts to pitch up out of control. Wing area at the rear of the plane will tend to pull it back to level flight; wing area at the front will pull it further over. Hence canards promoting instability. But this is also why they're good for manouevrability; you get heaps of pitch authority for just a relatively small surface. They're good when used appropriately.

Edited April 7, 2015 by Wanderfound

[capi3101]

Yeah - I was thinking about suggesting moving the mono tank up between the cockpit and the drone core, but I didn't know if that would've moved the docking port into an impractical position or not (I never got a good look at the underside of the plane so I don't know if the original designer put an extra set of structural intakes down there or not; if he hadn't, I wouldn't see any issue with setting the port up to open up from the ventral side instead of the dorsal). Small fin crossed my mind when I ran the 30k/M4 and got the red/unstable L_ß figure.

Say I was working on that same design; how much more wing does it need? I'm assuming more chord, not necessarily more span. Maybe some strakes up in front of the intakes?

EDIT: Whoops - Wanderfound posted while I was typing...yeah, I wasn't sure suggesting canards to the guy was a good idea, especially after I'd cloned the craft and saw that it already had pitch-up tendencies.

Edited April 7, 2015 by capi3101

[Wanderfound]

It could probably get away with the wing area it has now, especially if they've had their mass shaved down a bit. One of the joys of overpowered planes is that the excessive TWR allows you to push through draggy high AoA flight while you get up to speed.

But if it was me, I probably would add the canards, and maybe some strakes on the central fuselage in front of the intakes. Deal with the pitch-up problem by lengthening the fuselage to shift CoM forwards. Apart from anything else, he could use some dedicated LF tankage for jet fuel.

Swapping the central two RAPIERs for a single central turbojet would also provide some room for more elevators and wing (as in the above pic).

Edited April 7, 2015 by Wanderfound

[Van Disaster]

That design was basically fixed with canards - I've never run into random pitch-up issues from canard designs that weren't caused by something else, and if you're using FAR you can program in some alpha protection using the AoA control setting. This was the last solution I posted in that thread, it's completely free of vices although fuel balance is a bit terrible ( it's not like I spent any time on it ).

Deltas need the chord length to be a significant percent of the fuselage length so you get some longditudinal stability - and that will also give you enough pitch moment by shifting the rear of the wing back away from CoM, although you'll never really have great pitch authority with a delta. You get a lot of CoL shift and takeoff needs careful undercarriage positioning, otherwise nothing really any different. Here's a straight delta version which also doesn't have enough fuel, but was also put together 15 mins ago

Honestly though with two Rapiers that thing is more of a missile than an aircraft.

[UnusualAttitude]

This is the most stable delta-wing I've managed so far...

Long enough fuselage, heavy stuff up front to pull CoM away from the tail, reducing the need for your typical massive FAR peacock tail-fin

The gull wing is great for stability and looks cool; feels a bit weird when pulling gees lower down though.

[capi3101]

Just wanted to let y'all know that I have been paying attention...

Kinda wish I hadn't reverted the flight - I could use the landing practice. It's become reflexive...

[Hodo]

Good job, that is a mighty nice looking craft there.

As for landing practice.

I did touch and goes for a couple of hours in my craft to get the hang of their landing characteristics. What I would do is launch the craft with about 1/4 of the fuel they would normally launch with, this way I could simulate a landing while light. I would also do a few with full fuel, this would give me an idea of what the craft is like when heavy.

[Wanderfound]

Nicely done.

A turbojet on the tail would give it a lot more atmospheric performance, as well as saving quite a bit of fuel during the early phase of the rocket ascent. Alternately, a central LV-909 would give a high-efficiency orbital manoeuvring option. You've got at least one more SAS unit than you need, and why so much monoprop? The docking port alone has more than enough for docking use, and a couple of spherical tanks stuck in the cargo bay would do if you needed extra for some reason.

As Hodo suggests, for landing practice you want to take off with a mostly empty ship. Do one named quicksave just after take off, and another after you've got the ship turned around and lined up for approach. Repeatedly restart the first one until you've got the hang of lining up an approach, then repeat the second one until you're comfortable landing.

Kerbal Flight Indicators makes landing much, much easier.

Edited April 8, 2015 by Wanderfound

[Van Disaster]

Alternatively NavUtilities.

I take off & dump fuel so I get to test either way; that needs TAC fuel balancer or the smartparts fuel dumper though.

UA, if it's got canards, it's not *really* a delta in the way we were discussing

[eddiew]

Excellent Good choice on B9 procedural wings too Do you really need that big monoprop tank though? Usually I find the docking port's own supply is plenty to pull of 2-3 docking manoeuvres. Might be better to get LFO in there for a bit more range.

Note for landings; radial parachutes don't cause drag even under FAR. Generally I put on 1 per 2.5 tons of dry mass, around the dCoM. Don't always use them, but it's nice to know that in an emergency you can come down in the rough on chutes

[capi3101]

Had a couple of you ask about the monoprop tank - this design started out as a tweak of SalehRam's design from the day before yesterday, and it had an inline monoprop tank. I didn't see any harm in leaving it in since some of the advice with that particular design was "make the fuselage longer and draw the CoM forward". In my experience, monoprop makes for reasonably good ballast if nothing else. I did move it well forward from where it was originally; in retrospect I wonder if it would've been better to put it in between the cockpit and the cargo bay.

Design actually incorporated a third engine originally - I took it off and replaced it with the drag chute once I started flying it. TWR was over 1 with the extra engine and the design was getting some pretty high Q at low altitude. Forgot to drain that back fuel tank after I removed the engine though; from my final fuel stats I definitely didn't need it. I suppose the design has a good fuel reserve as a result though.

I was pretty sure I'd overSASed it. On the other hand, RCS Build Aid was telling me I had 10 kN of torque from the main engines, so I knew I needed at least one. Plus I figured the extra mass would draw the CoM forward a bit. I think that if I were to double up the PB-NUK and move it over to the sides of the cargo bay there'd be enough space for a small payload. Another reason to move that monoprop tank in between the cockpit and the cargo bay...get the cargo bay closer to the CoM. Of course, that would make the SAS less effective, wouldn't it?

Got both NavUtilities and TAC Fuel Balancer installed - can't really see that from my UI on account of low graphics settings.

Will have to think about those chutes...

[Van Disaster]

Use RealChutes if you're going to use anything - you can add proper drag chutes, which can help considerably given spaceplanes can have some fairly high landing speeds.

[Wanderfound]

The trick with high-power ships is to climb steeply enough that you're out of the soup before you get too fast, and be sensible with the throttle if cruising at low altitude.

Drag chutes can work, but they're really not necessary except in very rare circumstances. S-turns and airbrakes do just fine (as does getting down low with engines off early enough for drag to do the job). And when you do want extra after-landing brakes (Duna etc), a couple of tiny radial engines aimed forwards do nicely.

I'd change the Mk2 monoprop for a Mk2 LF tank, change the nacelles to LF/O tanks, move the drone core, docking port and passenger bay to the front and put the cargo bay over CoM (so that cargo doesn't unbalance the ship). If not using a third engine, consider replacing the drag chute with a shielded docking port and changing the inline port to more tank or cargo space. If you must have drag chutes, radials work as well as inline.

I'd also add some structural intakes at the bases of the tailfins, slid back so as to make them look more supported. One SAS unit is plenty, and rotate the PB-NUK so as to not obstruct the bay. The placement of the RCS doesn't appear to give any lateral thrust; pairs midway angled 45° above and below the wing on the lateral tanks, in line with CoM, would work.

To sort out engine torque, shift the lateral tanks microscopically up and down until it reduces.

Edited April 8, 2015 by Wanderfound

[Van Disaster]

Oddly enough I end up using drag chutes on my smaller spaceplanes - the large enough ones seem quite happy with wheelbrakes, the little ones tend to attempt all sorts of tarmac acrobatics.