[Project] G42-200 StarLiner

[halcyon]

Okay!
I did extensive testing tonight. I've prepared my findings along with, yes...pictures!
Firstly, this is what I tested it with:

i7 950 3.07ghz quad core 1333mhz FSB
GTX 465 1GB DDR5
8GB dual channel DDR3 1600mhz
Win 7 64 bit Home Premium
TrackIR 5
CH Fighterstick USB

I'm breaking my findings into two section; ascent, and reentry.

Takeoff/Ascent

Sitting on the runway at KSC, I got an average of 30 frames per second looking around the cockpit. This was with both engineer MFDs on.
With engineer MFDs off, I got an average of 32 FPS.
TrackIR performed flawlessly. No unexpected spikes looking around the VC.
For the takeoff and ascent stages of my test flight, I condensed the instructions in your readme and came up with basically this:

@ Ground
TURBINE CYCLE -> Start/Run
BURNERS/REHEAT -> Before Takeoff, On/After Takeoff, Off

@ 10K ALT
CANARDS -> In
VISER -> Raise

@ MACH 1.5
WING LEVER -> HPC

@ MACH 2.0/20K ALT
RAMX DOORS -> Open/Off
RAMX IGN -> On
RAMX THROTTLE -> 60%
RAMX IGN -> Off
MAIN ENGINE THROTTLE -> 0%
MAIN ENGINE STARTER SWITCH -> Cutoff
MODE SWITCH -> INT
RAMX THROTTLE -> 100%

@ MACH 4-5/35-50K ALT
RAMX -> HI/Auto

@ MACH 15/60K ALT
OXY FEED -> Auto
ROCKET CYCLE -> Start/Run
MAIN ENGINE THROTTLE -> 100%
RAMX THROTTLE -> 0%
RAMX IGN - Cutoff
RAMX DOORS - Close

@ 100K ALT
VISER - lower

Following that, out of two different ascent attempts I came up with an average of 17.65 fuel remaining in the main tanks.
Once in space, I had on average 47 FPS in the VC. With the engineer MFDs off, it increased to an average of 51 FPS.

Both ascent attempts I found that 10 degrees pitch up was the max pitch up I could maintain without losing sufficient altitude while still gaining speed at a nominal rate.
Once the pitch increased to 11+ my ACC ms/2 would start to decrease slowly.
Lowering the nose back to a pitch of 10-8 degrees would start increasing the ACC ms/2.

It was very easy to maintain a pitch of 8-10 degrees using only the default elevator trim controls (Insert and Delete keys). I got the feeling though that the aircraft was extremely light. Trimming up/down just 1 or 2 key presses caused almost 3 degrees pitch change at some stages of the ascent.
The G42 needs to feel heavier and respond as such when making control surface changes, especially at the speeds I was travelling.

Some of the switch locations are too far away from the pilot, such as the RAMX ignition and door switches.
I know that in real life an engineer would be handling some of the switch flipping and ease the workload on the pilot, but in Orbiter we don't have that luxury.
For me it was a little easier since I used TrackIR and just turned my head slightly to get to the required switches while still keeping a hand on the joystick (I'm right handed).
For most pilots though, it will require them to take their hand off the joystick, grab the mouse, pan the camera to the switch, make the change, pan back, grab the joystick again.
I realize they'll have to do this anyway even if all switches were right next to the pilot MFDs but at least they have less distance to travel.
This is assuming pilots don't use multiple keyboard shortcuts (which is what I'll eventually use if available. I'm a keyboard shortcut freak).
I also noticed that the G key raises/lowers gear but the switch doesn't visibly move.

Okay, on to the reentry which I'm sure a lot of you have been having trouble with.
Well, rejoice because I had great success on my first try!

Reentry

I decided to test an unpowered manual reentry using GPCMFD (part of the Shuttle Fleet addon) and GlideslopeMFD (I think it's just called Glideslope in the O Hanger).
This would put the G42-200 into a high 40 degree AoA while making S-turns to keep drag up without bouncing out of the atmosphere again.

I started with a PeR and ApR of roughly 350km, which is about the altitude of the ISS.
Over the Indian Ocean, just west of Australia I made my deorbit burn to a PeR of 57km.
At 150km I activated the GlideslopeMFD autopilot, which held my AoA at 40 even while making S-turns (it's the best AoA autopilot I've found).
Using the Ops 3 page of GPCMFD, which very closely models the actual Shuttle reentry velocity display, I waited about 2 minutes past EI (120km) until I saw my ship on the Ops 3 page and started my first S-turn to the right.

Now...I found that the G42 requires an incredibly sharp bank angle, 90+ degrees to get the VACC to go negative. I also quickly realized that a 40 degree AoA wasn't enough to give me the drag I needed to start descending down the Ops 3 vertical speed corridor.
I had to pull back on my stick all the way (I think AoA was about 70 something) before the drag starting showing at least 15 at which point I started down the VS corridor.

During these maneuvers only small corrections in the bank angle were necessary to change the VACC. In fact a lot of the time I overcorrected because my VACC was extremely sensitive to any bank angle changes.
This made me wonder if the G42 is getting too much lift.
On the space shuttle, the first S-turn bank is normally done around 60-70 degrees. The G42 needed more then 90 degrees of bank angle. There is a LOT of lift happening on reentry.

I followed the cues from GlideslopeMFD on what my altitude, VS, and VACC should have been. Ops 3 from GPCMFD took care of my required velocity.
Halfway between the middle of the Gulf of Mexico and the west coast of Florida I noticed that my DelAz was changing too rapidly.
Again I'm attributing this to the increased lift on the G42. At slower speeds when I made my S-turns I found that I needed less bank angle to achieve negative VACC (earlier it was 90+) but making the S-turns that steep was really changing the vector of the ship at speeds of about Mach 4-5.

I tried just pointing straight on my course towards the HAC turn and lowering the nose, but I had too much lift and my VACC quickly skyrocketed up into the positive 200+ range.
Getting closer to KSC I had to make an S-turn every 10 seconds or so and each time my VACC would shoot up which quickly put me out of my altitude profile according to GlideslopeMFD.
In the end, I was about 10km higher then I should have been when I got to the HAC turn. My velocity was right on the money though.
Here's a screenshot to show the HAC.



As you can see I was only about 2km higher then the planned altitude halfway through the HAC turn. Velocity was right in the middle of the vertical speed corridor.
When I came out of my HAC turn I was lined up with runway 33 at KSC, a little high though.
My speed was about 220 m/s. I had to do some more short s-turns on final to bleed some extra speed. I made straight for the runway very similarly to how the shuttle does it's final approach; a very steep dive before pre flare.
Lowered gear 200ft AFE and I crossed the runway 33 threshold doing about 170m/s velocity.
Flare was small because any more then 2-3 degrees pitch up made the aircraft climb away from the runway.
Touchdown was at about 150m/s velocity. Applied brakes and around 80-90m/s the aircraft actually lifted off again. I had to mash down on the elevator trim to keep the nosewheel back on the ground.
I stopped at the end of the runway on centerline.



So, my overall findings:

- too much lift. Aircraft needs to be heavier. Maybe it's a CoG issue or related?
- switches should be grouped closer to the pilot. The RAMX switches and the rocket/main engine switches should be together. Why are the RAMX switches all by themselves?
- if the oxygen feed is required for rocket operation during the last stage of ascent, shouldn't those switches be together? The oxy feed switches are currently next to the RAMX ignition/door switches
- nosewheel steering on the ground, as someone said is too sensitive. Needs to be turned down quite a bit
- OMS engines aren't powerful enough, as someone also said
- can't read any of the main panel/center pedestal switches if the sunlight isn't hitting it. The panel isn't emmisive like the overhead (also already mentioned)
- the default settings on MapMFD were weird. I missed having my terminator line/shading plus coastal lines and base boxes
- I also tried taking off without the afterburner. It was possible, just took longer. Don't know if it used the same amount of fuel compared to using afterburners with a shorter rotation time
- I used about 40% of the RCS thruster fuel on my reentry from holding constant AoA and doing the S-turns. On a mission to the ISS, I don't know how much would be left before reentry. Future missions might run out of RCS fuel

Okay, that's it for now, I'm heading to bed.
If I think of anything else I missed I'll add it tomorrow.
Overall Moach, this is one FANTASTIC vessel. The VC is a super winner, even in WIP-1. I get more enjoyment flying this then the default 2D cockpit of the XR-2.

This is only going to get better!

*edit*
I just realized that the steering autopilot was still active on GlideslopeMFD when I landed....this could have been the reason I couldn't flare too much.
But I still think the G42 has too much lift on reentry.

[Interceptor]

Loving this rare bird,but How do you lower the visor? BTW Moach This ship flys really nice.

---------- Post added at 01:58 AM ---------- Previous post was at 01:51 AM ----------

Nevermind, I figured out what the visor button was.DAH!

[cymrych]

My experience thus far is similar to Halcyon's, with some minor differences. My rig is comparable, with an AMD Phenom II 6 core clocking 3.2MHz, 8GB RAM, Radeon 6800 with 1GB dedicated, but without that nifty TrackIR tech (gotta check that out one of these days!) Somewhat perplexingly, my fps NEVER dropped below mid-50s, no matter what or how much I turned on.

Ascent:

I took a slightly shallower approach to orbit, with a velocity vector held via trim around 2-4 degrees above horizon for majority of non-rocket ascent. Throughout, the horizontal acc remained well within (NASA) tolerances, and may only have gotten higher than 1g at the very end of the scramjet routine and through the rocket orbit-insertion program. Believe I had roughly 16% in the tank once circularized at 250km.

As noted by Grover earlier, the rocket portion of the ascent seems rather overpowered; think it only took about a 7 second burst to raise PeA from -1.1M to my arbitrary 250k ApA. Rot RCS, however, I didn't feel was really underpowered ... at least for orbital attitude maneuvers.

Descent:

My to-date best entry went something like this. West of Australia, deorbit burn to a PeA of 70km. I rolled prograde, moved the tips to "reentry", inverted, and pitched to -30 degrees. Kept an eye on vertical velocity and acc in Surface mfd, and adjusted with subtle pitch controls to keep vert velocity around -200 m/s. I stayed inverted until my PeA dropped to around -500km when my orbital velocity was around Mach 25.5 or so. Rolled heads-up to positive 40 degrees ... and coasted, and coasted, and coasted.

Right about here I noticed everything halcyon related. The 42 bleeds energy incredibly slowly. So I rolled left around 95 degrees to keep vertical acc negative and pulled hard on the stick (AOA around 80ish, but hard to maintain. Probably 60-70 AOA on average through S-turns), and cruised right over Cuba and out over the Caribean for 700km or so. Eventually got my speed down to about Mach 7 around 50km altitude 1.3M from the runway, managed to get my vector to swing northwest, and began the run to the Cape. Still ended up doing a 100 km radius left turn over Miami to bleed off enough speed and loose enough altitude. When she finally slowed to under Mach 2, I neutraled the wings, and turned toward Kennedy. Passed subsonic at 15km altitude, 44 km from the tarmac. Fired up the turbines, but never touched the throttles, and coasted the whole way. Even had to use liberal down trim to keep the vector on the near end of the runway. Popped gear at 160m/s, about 3 km from runway. Long 0 vert acc flare at 25m, smoked the tires about 1/3 down runway. I used rearward linear thrusters to augment the brakes and simulate reversing engine thrust for wheelstop. (First landing took me on a fast taxi past the VAB. My legs were tired for our virtual pilots, I was slamming my toes on the non-existant brakes at my computer so hard!)

I don't think I'd go so far as to say she has too much lift, since this same lift makes for such a smooth ride to orbit. (No offense intended, halcyon. Afterall, you took a much more scientific approach to reentry, where as I just kind of winged it. And my virtual self even managed to forget his camera, lol. I'm looking at the problem more like an equation, where the left side (ascent) works, now we need to find the right side(descent) equivalency. I think a major enough mass increase or lift reduction to make the descent work would throw the ascent way out of kilter) I feel there likely must be some quality equilibrium point on the reentry side of the flight profile which will take advantage of this superior lift ability to help more rapidly decelerate the ship, which we haven't exactly found yet. I suspect we've been treating her too much like the shuttle in terms of reentry; think we either need to start deorbit burn considerably earlier than west of Australia and maintain a very long, shallow entry profile, or try a significantly steeper entry angle.

But both tests will have to wait until tomorrow. Well, later today, as I noticed just now that its past 2 in the morning. Oops! 2 hours of sleep is enough for a test pilot, right?

Moach, this bird is spectacular! And to think, this is just a 1st release WIP! Congrats again, and well done!

[Eccentrus]

Oh well much of the issues has been reported anyway, so I'll just add a thing, this one handles on keyboard much much better than your first G42, it doesn't oscillate like a horse now and the VC can only be described by one word: Awesome... :D all the complexity that is involved there reminds me of the cockpits of real aircraft

[Grover]

ok, ive now had a go at re-entry as well:

since it seemed popular, i used the glideslope MFD as a guide, but aimed to be a little slower, since i can always bump start the engines to extend my glide (ok, its not really a glide then)

it went perfectly, i never even made plasma, the AP dropped off around 1M from Canaveral (perhaps i went a little too slow), and i was starting up the RAMCASTERS (i was at mach 5) when orbiter... crashed.
it wasnt a CTD, coz it never went to the desktop, it just froze, ill check the log, but i doubt there'll be anything there.

so, re-entry is possible, just very RCS heavy (i was using 1Kg/second of systems fuel consistently throughout the re-entry) to hold AoA between 40 and 30.

i dont think theres really too much lift, once i got below 55k, i couldnt bank beyond 90*, and was at about 50* when i went back to level flight (or should have)

if anyone out there wants a guide (i used basesync for the de-orbit) i set ALT to 80, ANG to 0.7 and ANT to 50 (though i think 45 is better, to avoid falling short)

admittedly, i used the rockets, not the OMS for deorbit and plane alignment, becuase the OMS arent powerful enough, i think they should have a combined thrust of between 1Mn and 1.3Mn, to give acceleration around 5m/s^2

before i sign off for now, i noticed one more thing:
the craft doesnt sit still after taxiing to the runway, the AoA, Slip and speed indicators all fluctuate and the craft visibly bounces (though not very much)

i cant think what's causing it, perhaps having more than 3 touchdown points defined?

ill leave you guys to it for now

laters!

---------- Post added at 11:52 AM ---------- Previous post was at 11:46 AM ----------

one last thing:
for those that didnt like the MapMFD like how Moach set it, it was all configurable, heres the scenario i used, the G42 at Wideawake international ready to roll out of the loading hangar

BEGIN_DESC

END_DESC

BEGIN_ENVIRONMENT
  System Sol
  Date MJD 55685.7607013878
END_ENVIRONMENT

BEGIN_FOCUS
  Ship G42-200
END_FOCUS

BEGIN_CAMERA
  TARGET G42-200
  MODE Extern
  POS 1.46 -169.78 -13.32
  TRACKMODE TargetRelative
  FOV 40.00
  BEGIN_PRESET
    Track:Earth:50.00:RELATIVE 1.299 0.556 0.107
    Track:G42-200:40.00:RELATIVE 4.000 0.000 0.000
    Track:G42-200:40.00:RELATIVE 4.000 0.000 0.000
    Track:Earth:20.00:RELATIVE 1.553 -0.395 -0.680
    Track:Earth:20.00:RELATIVE 1.434 0.192 -0.403
    Track:G42-200:40.00:RELATIVE 4.000 0.000 0.000
    Track:Earth:20.00:RELATIVE 1.434 2.433 -0.488
    Track:G42-200:40.00:RELATIVE 4.000 0.000 0.000
  END_PRESET
END_CAMERA

BEGIN_HUD
  TYPE Surface
END_HUD

BEGIN_MFD Left
  TYPE Map
  REF Earth
  POS 0.00 0.00
END_MFD

BEGIN_MFD Right
  TYPE HSI
  NAV 0 0
  OBS 0.00 0.00
END_MFD

BEGIN_VC
END_VC

BEGIN_SHIPS
G42-200:StarlinerG422
  STATUS Landed Earth
  POS -14.4250000 -7.9576000
  HEADING 300.00
  RCSMODE 0
  AFCMODE 7
  PRPLEVEL 0:1.000000 1:1.000000 2:1.000000 3:1.000000
  NAVFREQ 0 0
  XPDR 0
  GEARS 2
  CNRDS 2
  VISOR 1
  INLET_MAIN 1
  INLET_RAMX 1
  WING_WVRD 1
  WING_FTHR 1
  MAIN_ENG_MD 0
  BURNER_TGGL 0
  RAMX_ENG_MD 0
  THR_AUTH_MD 0
  VC_SWS =======-=v-vvvvvvvvvvvvvv=--=----====vvvvvv^v===-=====
END
END_SHIPS

if you havent put ReFuel MFD into your orbiter installation, dont copy the last part concerning ReFuel MFD

[PhantomCruiser]

Wow, what a nice bird. I haven't had time to really test it out, but I did want to mention that this is one nice spacecraft. There is an amazing amount of detail in the VC. I love it!

[Moach]

oh wow! thanks everyone - your findings are of undescribable help

key noted facts (not ordered by any logic whatsoever)

• RCS overpowered
• OMS severely underpowered
• time too sensitive / ACS not enough
• too much lift (*see notes)
• RAMX and OXY FEED switch placement needs reviewing
• possibly too tiny SYS fuel tank
• vessel dry-weight possibly too low
• main panels not self-lit (i actuallly already knew that)
• the CoG-shift is very much necessary
• VC switches don't respond to keyboard commands (also already known)
• nosewheel steering is crazy - i'm aware (how on earth does one ajust that thing?)

notes:

i was aware that the '42 would be bound to feel excessively lightweight right now - but keep in mind, the cargo bay is EMPTY, a real mission would have a few extra tonnes to haul up...

as for the excess lift on reentry, did any of you remember to flip the wings to featherd mode (lever to RTY) before taking 'er down the slopes?
doing so eases back about 30% overall lift - and that's what it's there for


remember that unlike the DG, the '42 has real ground-effect lift modelled when close up to the surface, this would indeed cause it to "float off" if you overpitch when flaring... settings might be off, tho... perhaps excessive


last night i did some further checking... some clunkyness was caused by the EICAS page refresh, it got smoother when i removed the timeout-based update and set it to "PANEL_REDRAW_ALWAYS" so Orbiter commands the redraws on that from now on....


for some reason, disabling hardware pageflip gives me a 10FPS boost on my home rig... might be just me... that install is such a mess i don't even know what's what anymore



a thought for you folks with bad FPS - if you're on Vista (why??) or Win7 (much better) - check that activetype is disabled, as that bogs orbiter down when it tries to render text... not sure if it applies to WinXP, worth checking, just in case


anyways, thanks a million for your invaluable help - your flight testing info will go a long way back here

[Grover]

ok, ive done a full mission to the ISS (couldnt dock im afraid, due to there being no docking port), and i believe that it MIGHT be possible to do, but i had to use the no-oxidiser bug to complete the mission

we definatley need a bigger systems fuel tank, as well as higher thrust OMS engines, if we put cargo in there, we'll never make it to the ISS, let alone back down again. it may also be worth putting larger main tanks on it (though maybe not if we add droptanks and DARTS soon) to make sure its achievable

as for re-entry, i DID remember to lift the wings up, and i just flew a sucessful re-entry with the '200, its quite forgiving at the moment (possibly because you can't melt it), i only made contrails once, and that was when i started a panic-induced descent spiral over KSC after passing it at 40k ALT
despite the altitude, i was going quite slow, and getting down the the cape was really easy, i couldve done it without firing up the engines had i wanted to. though, in the next WIP, make sure there are speedbrakes, or increase the '200s aerodynamic drag is increased, because this thing just doesnt stop on its own unless you're below 1K ALT (in fact, you are probably best doing both, maybe then we can slow down enough on re-entry to keep glideslope MFD happy :D)

and i found the source of the movement bug: its the turbine engines. whenever they're turned on, they constantly supply a thrust (though not enough to overcome inertia at x1 time accel). Moach, for Orbiter's sake, you should remove the idle thrust whilst the whels are on the ground. i understand why you made that nice bit, but if you cant stop when the turbines are on, you wont be able to refuel.

ill leave it down to you now, i might try a little more, but i think bug-finding in WIP1 is mostly done now

see ya round

[Moach]

well noted...

ok, the turbines aren't properly modelled yet... flipping the cycle switch to "cut" -should- stop any thrust from being generated... but it might not be working too well just now


i did some tune-up

modifications:

• increased vessel dry mass a by couple of tons
• significantly reduced vtrim power (found a stray "zero" ten-folding the original value, killed it on sight)
• reduced thrust for dry turbine and rocket engines (only a bit)
• now the OMS are packing .5'gs worth of punch - they should be good for use now
• RCS power slightly reduced
• added capacty for systems fuel, from 8k to 10k now
• slightly reduced ground-effect multiplier (should land easier now)
• EICAS page refresh now controlled by Orbiter's PANEL_REDRAW_ALWAYS mode, instead of by 12ms dT timeout (runs smoother now)

with these revised settings i was able to make ~500k orbit with 5% oxy after circularizing...
i did takeoff with the fuel tanks maxed out, simulating a possible payload mass with the surplus fuel mass (full tanks are for plane-intercept legs, 80% is enough for straight-up ascent)


gotta try reentry now - wish me luck

[Grover]

if you're gonna re-enter, use Glideslope MFD and just control bank , i managed it first time round, im sure you will be able to as well

also, why is the landing speed (i found its below 100) much lower than rotation speed (about 130-150)? i knew it would be slightly lower, but not that much

---------- Post added at 02:09 PM ---------- Previous post was at 02:08 PM ----------

in fact, now youve reduced pitch trim alot, you may not be able to hold 40* AoA, even with full RCS, let us know how it goes

[Moach]

Quote:

Originally Posted by Grover

 if you're gonna re-enter, use Glideslope MFD and just control bank , i managed it first time round, im sure you will be able to as well

also, why is the landing speed (i found its below 100) much lower than rotation speed (about 130-150)? i knew it would be slightly lower, but not that much

---------- Post added at 02:09 PM ---------- Previous post was at 02:08 PM ----------

in fact, now youve reduced pitch trim alot, you may not be able to hold 40* AoA, even with full RCS, let us know how it goes

well, i took too long to deorbit and found myself in a much steeped reentry than i had hoped for... missed target by a couple timezones... but i didn't give in, i let it climb back and took 'er for a curved descent... all worked out in the end...

and the reason for the landing speed being so much lower than vR, is most likely because more than half of the ship's takeoff mass got burned up during ascent

unlike airliners, most of the G42's takeoff weight is fuel, so you can expect that the spped differences between t/o and landing will be a lot greater than what we're used to from FSX and stuff

[cymrych]

Quote:

and i was starting up the RAMCASTERS (i was at mach 5) when orbiter... crashed.
it wasnt a CTD, coz it never went to the desktop, it just froze, ill check the log, but i doubt there'll be anything there.

Ah yes, I forgot but I got that as well in one of my runs yesterday. No CTD, but a total freeze up of Orbiter.

Also noted the jittery HUD info on wheelstop, but that seems resolved. I should have remembered the motors were still spinning.

All in all, a fun, forgiving ship to fly!

[Hlynkacg]

I was running Orbiter on my Laptop. It has a 3ghz AMD Core Duo w/ Radeon 550HD Mobile Graphics card. With an in-game resolution of 1280/720 the frame-rate rarely dropped below 30 FPS.

My flight experiance last night was pretty much the same as Halcyon's and Grover's so I'll skip the Play-By-Play and just say that you've done some amazing work.

Now for my biggest gripe:
I normally fly with my cockpit FOV set between 45 and 50 degrees as I find that this is a good compromise between detail and periferal vision on most VCs.



Durring accent I often found myself having to "zoom in" to read my MFDs or fuel state, this is annoying and time consuming for those of us still flying with a Mouse/Keyboard.

I like how you Implemented the "Lean" feature on the engineer's POV camera and would suggest that rather having Ctr-Alt-Up move the pilot's camera closer to the HUD it should instead move the camera forward and focus on the MFDs.

Like so...


I feel that this would make fine-tuning Orbits and other MFD intensive activities (like docking) a lot easier on the pilot.

PS: I still want an overhead window for the pilot seat.

[Grover]

Quote:

Originally Posted by Hlynkacg

 (...)

PS: I still want an overhead window for the pilot seat.

seconded

[liber]

Nicely done with switch view clicking on co-pilot and revers...
Since it's alway low FPS I use to it...bad PC...