General Question Space Shuttle launch window

[StargazerBranden]

Okay, guys,
I found and read this thread:

http://orbiter-forum.com/showthread.php?t=16557

It pertains to the Delta-Glider and works quite well! But when it comes to the shuttle, i can kiss my remaining fuel goodbye. I can never get a solid launch. when calculating launch azimuth i dont always come up with 42 degrees. If i use either the value i come up with or just use the 42 degrees i get a Rinc above 1 degree and thats sianara to my fuel on the alignment burn with the Shuttle's limited fuel and thrust. So my question is, as it relates to the space shuttle, is there a different launch time besides the 300 second time to node that the link recommends to get my RInc at MECO to below 1 degree?

---------- Post added at 12:53 PM ---------- Previous post was at 12:25 PM ----------

I mean, surely there is some formula to calculate this?

 

[Thorsten]

The formula mentioned in the thread is an approximation for a non-rotating Earth - if you need to do more precisely, you have to take the velocity vector you get from the rotation into account as well.

Dependent on what Shuttle you use and whether you pilot manually, just zeroing Rinc during the ascent by yaw vectoring actually works quite well.

 

[StargazerBranden]

I use shuttle fleet. Built in autopilot and I use the direct insertion profile. No OMS burn. Autopilot has full control, NASA-style. Just transition right into orbit synchronization (usually just by bringing my perigee to 180km and apogee to 245.6 km).

The azimuth isn't the problem. The DGIV documentation has a formula that I use where you calculate target Inc (51.6 ISS), launch latitude (28.625 KSC according to Map MFD), rotational velocity of 407.8 m/s and target velocity of between 7750-7650 m/s which gives me a value of 42.80 degrees or 138.20 degrees for azimuth depending whether I'm launching "uphill" or "downhill" so to speak. So I don't think azimuth is the problem. The problem is time. There has to be a set launch time, where I can launch at that time every launch window and have a RInc of < 1 at MECO every time. The thread gives a ballpark of around 300 seconds until node passage. However if I launch at that time, on my azimuth, I have a RInc > 1. That's bad for the shuttle. I have to take my RInc value, exit orbiter, add or subtract that value to my azimuth in the autopilot profile, then relaunch and see where that gets me. Entirely trial and error. I want to be able to remove the trial and error. The answer has to be in the launch time. I just haven't been able to calculate a time since Tn =300s doesn't seem to be the answer.

 

[DaveS]

That's because 300 seconds equals 5 minutes. The ascent profile of a shuttle is approx. 8.5 (510 seconds) minutes long from T0 to MECO. Also a Direct Insertion involves one OMS burn, done at the first apogee. This burn is known as OMS-2. At no point is the ET carried into a stable orbit. The perigee is always in the atmosphere, around the 60 km altitude. This is to ensure that the ET re-enters the atmosphere and is destroyed.

 

[StargazerBranden]

DaveS,
Thank you for pointing out my error. I meant to say OMS-1 but typed it out as only OMS. As for launch time vs ascent profile I have tried liftoff at 510 seconds to node still with RInc > 1 degree. However, I have not fully tested this out. What I mean is that I have launched at Tn = 520s (Item 777 begins countdown at T minus 10) but have noticed that until Tn gets to a lower value, one second of Tn is not equal to one second of elapsed Sim time so perhaps that is why I was still getting a margin of error in my RInc. I will test this tonight and post the results. My plan is to let Tn count down to zero, go back 8 minutes and 50 seconds from that, find out what the Tn value is at T minus 10, and see if that number works for multiple scenarios with the calculated azimuth. I will post my results for those interested. Thank you both for the valuable information!!

---------- Post added at 09:48 PM ---------- Previous post was at 05:17 PM ----------

UPDATE:

After multiple launch tests I am still getting nowhere. I let the Time to Node click down to zero, added 550 seconds from that (allowing enough time to get GPC MFD set up), launch at 510 seconds remaining. I get a RInc of >2 this time. Try again using launch MFD. I get the exact same azimuth with my own math, exact same data as displayed on Align MFD and the exact same end RInc result. However, I happened to notice, somewhere along the way the Time to Node increases. Going to do another launch, see how much time remains until node passage at MECO and then add that to my figures and then set the extra time and relaunch. (I did mention that I was trying to get rid of the trial and error process and have a mathematically accurate, NASA style launch right?! Yet here I am and elbows deep in trial and error... :beathead: ) Haha. Anyway I will post my results for those interested!

 

[boogabooga]

Get familiar with Launch MFD. It has a plane correcting feature that you can use in real time during your launch. You can manually control yaw during the shuttle fleet auto ascent while following its direction.

 

[StargazerBranden]

UPDATE 2:

Got time to intersection on launch MFD down to zero at MECO and I am left with a RInc of 2.34 degrees. Bye bye OMS fuel! So... All that and still nowhere. I'm going to scour the internet for some official NASA documentation for calculating launch times. There HAS to be something. A formula. Something on how to do this with as much precision as possible. Or else my next launch is going to be in a straight-jacket instead of a spacesuit. (The scene from Armageddon where Rockhound is duct-taped to the X-71 chair after going banana sandwich on the asteroid kinda comes to mind here... :lol: )

---------- Post added at 10:19 PM ---------- Previous post was at 10:18 PM ----------

boogabooga, I will most definitely give that a try!

---------- Post added at 10:33 PM ---------- Previous post was at 10:19 PM ----------

:idea:

Here's a thought...
Would a Hohmann transfer calculation work for figuring an exact launch time? Or does that only work for other celestial bodies? Is the time exactly accurate on a Hohmann transfer or is it a ballpark estimate?

 

[Thorsten]

So I don't think azimuth is the problem. The problem is time.

Why would a time error give you a 2.34 degree inclination error?

Assume Earth doesn't rotate. Your target groundtrack passes right over the launch site (aka, you're in the same plane) and you launch along the direction of that groundtrack (that's your azimuth).

You're going to end up in the same orbital plane, regardless at what time you launch - time doesn't affect inclination in this scenario, all it affects is anomaly at epoch - to get into a different inclination, you need to launch into a different direction.

Now rotate Earth. If you've done your computations correctly, you're *still* going to the same inclination - though a time offset can now displace the longitude of the ascending node. If you're three minutes off in time, only by 0.75 degrees though.

Or, in other words - you have pre-computed an inclination and a launch azimuth. Yet you do not end up in the same inclination afterwards.

Even without any rendezvous targeting, there's only two solutions to this puzzle: Either you've done the calculation wrong/inaccurate, or your autopilot doesn't fly correctly. But a time offset can't lead to a difference between planned and actual inclination reached.

 

[GLS]

Sorry to butt in the discussion, but the launch azimuth is just the initial direction the launch vehicle has to follow. As the vehicle progresses during ascent, the azimuth should (in this case) decrease a bit, but the "inertial direction" should remain the same (except for whatever yaw steering it has to do).
So one has to launch, turn to the calculated azimuth, and then pretty much hold the sideways orientation in relation to the stars. This will get the vehicle in an orbital plane with the desired inclination. To get the vehicle orbital plane to match a target, the launch time has to be chosen carefully, and/or yaw corrections have to be done during ascent. Given that it takes some time for the launch vehicle to enter orbit, the launch time has to be before the target's orbital plane crosses the launch site.

 

[Thorsten]

Would a Hohmann transfer calculation work for figuring an exact launch time?

Not at all - a Hohmann approximates your burns as zero duration impulses, whereas your ascent is an eight minute acceleration to orbital speed.

So one has to launch, turn to the calculated azimuth, and then pretty much hold the sideways orientation in relation to the stars.

He said he lets the AP fly - so I assume the AP would do this (haven't tried Shuttle Fleet myself).

 

[StargazerBranden]

Thorsten, very informative! Thank you for that. It has to be an autopilot error. My azimuth math that I computed out by pen, paper, and calculator checks out against both the DGIV built in calculator and Space Calculator Version 2 so I think it is a limitation of the shuttle fleet autopilot as you say. As the rest have suggested I'm going to have to manually adjust the remaining sideslip during the ascent stage.

---------- Post added at 07:52 AM ---------- Previous post was at 07:08 AM ----------

boogabooga,
This manual correction with launch mfd , is it the heading error value I'm looking to zero out or a different figure? Still learning the MFD interface

 

[zerofay32]

The shuttle fleet autopilot works but is not perfect. Also it is completely a closed-loop type of guidance. But you can manually yaw the shuttle while it is on the launch autopilot with the 1 and 3 numberpad keys. Wait until the shuttle rolls heads up and use Align Plane MFD to yaw the correct direction. If the initial angle is less the 5 degrees, you'll have plenty of time to get it down before MECO. Get it under .5 and you're good to go.

 

[boogabooga]

I included a Launch MFD plane alignment tutorial in this:
[ame="http://www.orbithangar.com/searchid.php?ID=6872"]Atlas V / Enhanced Cygnus Launch Scenario[/ame]

 

[Thorsten]

Get it under .5 and you're good to go.

Heh - and there I was agonizing for two days about a 0.01 deg residual after inclination acquire which I couldn't seem to beat down in my launch AP code...

Turned out it was the non-spherical gravity field which makes point mass orbital dynamics too inaccurate :hmm:

I wonder what accuracy the real Shuttle has, but I've never seen any numbers.

 

[zerofay32]

Obviously, you get it as low as you can with the SSMEs, and maybe Shuttle Fleets OMS engines is not accurately powered. But when I was running a mock station building program several years ago I found that under .5 after MECO and a +/- 3 day to rendezvous (standard shuttle approach from under and behind) leaves plenty of fuel for maneuvers. Also combining plane change maneuvers with Orbital adjustments make the flight profile more accurate.

My feeling is the same procedure would not be the case will SSU. But I believe eventually SSU ascent guidance will be smarter than Shuttle Fleets'. Part of the cheat with Shuttle Fleet is all the RCS (Fwd and AFT) and OMS fuel is combined in one source. Which allow a bit more flexibility. Also who knows if the fuel consumption rates are accurate.

EDIT: I never tried leaving non-spherical gravity on so there is another departure from reality/cheat.

EDIT2: maybe it was .05. Getting old and can't remember.... :huh: In any case, yawing near the end of the accent will get one closer to the desired orbit that just letting the guidance do its thing by itself. Also be sure to save the scenario before launch. Sometimes the Shuttle Fleet autopilot works... and 1 in maybe 20 flights it doesn't get you into orbit. But maybe that has been fixed...

 

[Thorsten]

It depends a bit on payload, but I think I remember I tried that 10 seconds of two engine OMS burn is worth 0.01 degree of inclination change.

That'd make 0.5 degrees a 500 second burn - seems a lot. Makes the 0.05 more likely...

 

[clive bradbury]

After much experimentation, I have nailed the shuttle fleet launch and alignment, at least to my own satisfaction. Absolutely no need to use launch MFD (Good, but too complex for my simple brain to process) and no messing about with yawing during the ascent either.

A successful launch to a low relative inclination relies on two things only – a correct launch azimuth and launch time.

AZIMUTH

My procedure for shuttle fleet for a target inclination of 51.6 degrees (ISS/MIR) is as follows:

The first point to note is that the shuttle fleet autopilot is not completely accurate in azimuth. An azimuth calculator result for a launch from Canaveral to a target inclination of 51.6 comes out as 42.8141.

Do not put that figure in the launch profile in your scenario file – use 40.8 degrees instead. I also use the OMS line, and PROFILE = 0. If you leave the latter line out it will default to the zero profile anyway. That’s the azimuth sorted.

LAUNCH TIME

Launch time = 200 seconds time to descending node.


Open Align Planes MFD, select your target, and watch Tn relative to the descending node.


Open GPC MFD in the other window, OPS Mode 1 (ascent profile), Select Item 777 (the launch code), then hit enter when Tn = 210 seconds.


Get coffee and sit back down. When everyone else is carefully yawing – drink coffee.

That method will give you a relative inclination of around 0.06 degrees to your target (You will notice that the inclination does increase around the time of the roll to heads up, but it will decrease again afterwards with the above result). Very little fuel is required to correct an inclination that low. Job done.

 

[StargazerBranden]

Clive, does that work on the opposite node as well? Input 139.2 launch azimuth for a southerly launch. I would assume so and can always check it.

 

[clive bradbury]

Never tried it, but I see no reason why not. I have found that if you subtract a couple of degrees off the 'proper' azimuth that seems about right for SF.

 

[Poscik]

Some time ago I was wondering how the hell NASA did whole rendezvous with almost NO corrections. After many simulations I realized, that non-spherical gravity sources plays big role here. In general, you launch little bit off plane - as far as I remember, your LAN should be ~0.05° to the east from station, but your inclination should be identical. Now staying on lower orbit deviates your LAN to the west little bit faster than station's LAN. This process is clearly visible on AlignPlanes MFD. The lower you are, the faster LAN deviates to the west. In result, after two days of chasing, LANs difference should be around zero and you don't need almost any plane corrections.