Any ways to build a stable shuttle without adding canards?

[Firemetal]

If anyone wants, I will post pictures.

So whenever Ibuild a shuttle, I try to make it fly easily and not look nice. I add canards so I don't stall during re-entry. Now I have tried making shuttle replicas that are close to the real thing. But all of those just stall during or after reentry. I don't want to add canards to it because that looks terrible and isn't what the real one had. What I am saying is, are there any other ways to stop  reentry without adding canards? Thanks in advance.

Fire

Edited November 11, 2016 by Firemetal

[bewing]

Well, that's actually kinda tricky. The problem is landing speed and the wheel bounce bug. Are you willing to land on the polar ice caps? To not stall, you have to maintain your airspeed all the way to touchdown -- probably something like 200 m/s. You also need plenty of control surfaces and maybe a good amount of reaction wheel torque. 

[foamyesque]

Just gotta balance it right -- and, truth be told, you do want to be stalling, or very close to it, during your reentry's initial portion. Transitioning to a glide should happen somewhere around 1000m/s surface; prior to that you want a 45+ degree AoA for thermal control.

[AeroGav]

7 hours ago, Firemetal said:

If anyone wants, I will post pictures.

So whenever Ibuild a shuttle, I try to make it fly easily and not look nice. I add canards so I don't stall during re-entry. Now I have tried making shuttle replicas that are close to the real thing. But all of those just stall during or after reentry. I don't want to add canards to it because that looks terrible and isn't what the real one had. What I am saying is, are there any other ways to stop  reentry without adding canards? Thanks in advance.

Fire

Hi Fire,

I started a thread on something very similar, to which I posted two canard-free and flyable designs.

The first thing was to get CoM as close to middle of the ship as possible, even when empty of fuel and cargo.  This stops it becoming aerodynamically unstable on return from space. Since the only thing to counterbalance the weight of the engines up front is the cockpit,  it means using very light engines.   No vectors.   I've had success with an arrangement of 3 aerospikes for the main engines.   Couple of sparks or puffs for the OMS pods.

Since it's now going to be rather underpowered,  do your best to keep the aerodynamics clean.   Clip tailcones into the engines, remember to put a 2.5m tailcone on the central 2.5m attach node of the mk3 engine mount.  

The next thing was trim drag,  in other words trying to keep the nose up.  What you don't want is this situation -

Getting the CoM forward helps a bit since it gives the elevons a longer movement arm.  But your CoL is still way back making her very lawn dart-ish.  What helped was to angle the strakes upward slightly in the VAB,  which lessens how hard the elevons need to continually push downwards throughout flight.

Of course, while i managed to make a fairly authentic-looking orbiter,  the stack as a whole is not quite STS.

We have no external tank, it relies on 4 SRBs and internal fuel only 

I am not sure if Aerospikes have the power or gimbal range to push an external tank design to orbit.  I have never tried, but surely the devs wouldn't have given us the Vector if that was the case.

If you are happy with settling for an internal fuel design, then like me you'll probably start agonising over the 800 units of potential LF storage you got in its wings and strakes.

What I now believe is the most elegant approach is to use the Configurable containers mod, and tweak the amount of oxidizer in the fuselage tanks  up slightly, at the expense of LF.   Then when you fill the wings the vehicle as a whole will have just the right ratio of LF/O.   The beauty of this is that if someone downloads your ship in a stock game, they still benefit from those tweaked fuel capacities.   However, if you change the tank to hold something different to what it normally holds - for example, i built a version where i put monoprop in the wings for my OMS system - then in a stock game, what you get is wings with (full) monoprop AND also (empty) LF tanks, and it allows you to fill both, which is unacceptably glitchy for my taste,

 

EDIT 2 hours later I have this

Even by dropping down to terriers and sparks, the CG is still slightly aft of the middle of cargo bay.  

I eventually figured out how to put it in orbit. Goes from crazy power (4 kickbacks) to anaemic (terriers and sparks) and i don't fly many rockets, gravity turns aren't my thing. I would share it on KerbalX, but not sure i want to put my name to that pig.

Edited November 11, 2016 by AeroGav

[Firemetal]

8 hours ago, bewing said:

Well, that's actually kinda tricky. The problem is landing speed and the wheel bounce bug. Are you willing to land on the polar ice caps? To not stall, you have to maintain your airspeed all the way to touchdown -- probably something like 200 m/s. You also need plenty of control surfaces and maybe a good amount of reaction wheel torque. 

So... It is based on how you fly it rather than how you build it?

8 hours ago, foamyesque said:

Just gotta balance it right -- and, truth be told, you do want to be stalling, or very close to it, during your reentry's initial portion. Transitioning to a glide should happen somewhere around 1000m/s surface; prior to that you want a 45+ degree AoA for thermal control.

45 degrees doesn't work with an unbalanced shuttle. I put it at 45 degrees and it stalls. I put it at around 20 degrees and it stalls. It will stall no matter how. Also how do you balance it right? It seems I am just always just pushing the wings back more and more until I find the right place. I don't want to trial and error, I want to understand why it happens and be able to fix it quickly. 

 

@AeroGav. You're design is basically like putting canards in the front but instead you balanced the COM instead of the COL. I am trying to go with a design realistic to the STS. All the fuel in the back, External tank and a big cargo bay. I am looking for ways to make that design work rather than making an easy-to-fly, non-realistic shuttle.

I have made one shuttle like this that works but it is very easy to stall in. I also just kept moving the CoL back until I found the optimum place. I don't want to do that as I said.

But thanks fore the replies. Much appreciated.  

Fire

[AeroGav]

1 hour ago, Firemetal said:

 

@AeroGav. You're design is basically like putting canards in the front but instead you balanced the COM instead of the COL. I am trying to go with a design realistic to the STS. All the fuel in the back, External tank and a big cargo bay. I am looking for ways to make that design work rather than making an easy-to-fly, non-realistic shuttle.

I have made one shuttle like this that works but it is very easy to stall in. I also just kept moving the CoL back until I found the optimum place. I don't want to do that as I said.

But thanks fore the replies. Much appreciated.  

Fire

You may already know this but the real shuttle engines weighed 3.5 tons each, vs an orbiter empty weight of 68.5 tons

Vectors weigh 4 tons each,  vs an orbiter mass of about 23 tons.    See the problem? 

Also, you are aware that the SRBs were 83% of the liftoff thrust? So the all-powerful Vector isn't reflecting reality particularly well either.

Where is the CoG vs CoL in your design btw?   Is the CoL still behind CoM, making it stable , but hard to control and aerodynamically inefficient, or is it actually unstable, so that going more than a degree or two off prograde will set up a positive feedback loop that overwhelms the flight controls?  

BTW does  a mk3 shuttle really need to come in from LEO pitched up at 45 degrees to avoid overheating?   I've been playing an airplane only career mode and re-entering at 7 degrees only creates heat bars.    Admittedly, these are inline cockpits not right at the front of the ship and they actually have wings that work and make lift,  but i'm not sure telling someone to with a rear CG problem to pitch up more is a good idea 

[ForScience6686]

Pitching up early is the key,  create as big of a drag ratio as possible high up before the aero forces take over.   as you get into the thicker air you'll most likely pitch down anyways due to forces on the large wings located at the rear.   this essentially trades less drag for more lift.  a problem with most every ones shuttle design is the com  being at the rear,  this creates a large lever at the nose,  and very little control authority at the elevators.   which causes the tail to want to lead.

I've been working on some early career shuttles lately,  but even on my space planes, I avoid the rear mounted wings as much as possible.   I focus more on a centered and stable com,  and end up with more of a conventional plane design than shuttle.   This generally allows higher angle of attack for more aggressive aero  braking.   I usually deorbit  straight to ksc for best recovery and having a wider range of drag ratios helps to pinpoint the landing.   That allows for faster braking up high,  and better trajectory control  to make the runway.

My most recent crew shuttle, can deorbit from a 600km orbit direct to ksc.   The only way to manage that is to be essentially in a stall with belly to prograde until speed is manageable for flight  again.   By design this craft has a high drag ratio in order to pull off the reentry,  but again it's more of a traditional plane design,  which allows better pitch control to maintain  high aoa.

People must remember that the shuttle is a flying brick.   it only has the ability to glide in for landing and touches down at over 200 mph, much higher than I land most craft.

[Firemetal]

3 hours ago, ForScience6686 said:

Pitching up early is the key,  create as big of a drag ratio as possible high up before the aero forces take over.   as you get into the thicker air you'll most likely pitch down anyways due to forces on the large wings located at the rear.   this essentially trades less drag for more lift.  a problem with most every ones shuttle design is the com  being at the rear,  this creates a large lever at the nose,  and very little control authority at the elevators.   which causes the tail to want to lead.

I've been working on some early career shuttles lately,  but even on my space planes, I avoid the rear mounted wings as much as possible.   I focus more on a centered and stable com,  and end up with more of a conventional plane design than shuttle.   This generally allows higher angle of attack for more aggressive aero  braking.   I usually deorbit  straight to ksc for best recovery and having a wider range of drag ratios helps to pinpoint the landing.   That allows for faster braking up high,  and better trajectory control  to make the runway.

My most recent crew shuttle, can deorbit from a 600km orbit direct to ksc.   The only way to manage that is to be essentially in a stall with belly to prograde until speed is manageable for flight  again.   By design this craft has a high drag ratio in order to pull off the reentry,  but again it's more of a traditional plane design,  which allows better pitch control to maintain  high aoa.

People must remember that the shuttle is a flying brick.   it only has the ability to glide in for landing and touches down at over 200 mph, much higher than I land most craft.

Yes but I am going for a realistic shuttle design. Maybe that isn't a good way to go. Ej_sa likes to play without using canards on his planes. However I always find planes impossible to fly without them. How would one play without them? Thanks.

Fire

[foamyesque]

Well, the thing about trying to build a shuttle replica in KSP is that the Shuttle was designed to solve a much harder problem than exists in KSP; if you build a realistic orbiter, the proportional fuel tanks and boosters (which don't exist in stock exactly but can sort-of be reproduced with liquid boosters, which also give you gimballing like the Shuttle SRBs had) are enormously overkill and you'd make orbit before you jettisoned the boosters, let alone the ET :v

So you need to compromise somewhere, but an asymmetrical design layout like Shuttles are extraordinarily sensitive to changes in proportions (thrust balance, mass balance, aero loading, and their changes over time), so it's mostly not worth it; if you're doing it you're doing it for the challenge and to show off

 

@Firemetal: A KSP shuttle replica is going to have most of its weight at the rear of the orbiter. It's inevitable if you're using Vectors (which are the engine with the best look match and required gimbal range). This is also more or less true of the real shuttle (though in KSP the Mk3 cockpit is far too light for accuracy -- another gripe for another time); if you look at the wing layout of the real one, it's clearly built to have a CoP way the hell near the back (and to deal with lowQ), hence the oversized aerodynamic control surfaces on the trailing delta and the gigantic tail. What you want to do is drain your OMS tanks and remove the payload, see where your CoM is, and then build things so that your CoL is near, but behind, it. The actual shuttle's wing layout works pretty well for this; you can fiddle with the angle of the leading wing strakes to adjust things, if necessary. Turn off angle-snap when you do, because the adjustments shouldn't need to be large.

 

Another thing, not mentioned AFAICT, is that the Shuttle used RCS thrusters to provide attitude control; don't be afraid to do so too. They provide an awful lot of oomph in the early reentry to keep you in your desired attitude, and can be helpful in pushing you out of an intentional stall or near-stall and into a glide path, since your aerodynamic control surfaces may be stalled out and not provide an adequate change to the drag profile. The shuttle also had something called a "body flap" which is often ignored in KSP replicas; it's a large flap mounted under the engines, which provides heat shielding and can be deployed to reduce loading on the elevons.

TBQH my usual problem with shuttle flights (aside from the normal hassle of balancing the launch stack) is yaw. Even with the massive spaceplane tailfin, the CoM is often so far back that it struggles and results in a really nasty sideslip-roll interaction from adverse yaw.

 

Anyway! So I built one.

 

It's pretty close to the Shuttle, but not perfect.

Payload's approximately the same (~30t to low orbit), but this only carries 6 Kerbals to the Shuttle's 7 people. No Canadarm, because stock. Has the same split-tail tailfin (via clipping, which also helped fix my yaw issues :v) to act as an airbrake, OMS fuel is internal instead of mounted in a replaceable pod. Obviously, it uses a LFB since we don't have any SRBs that have either the proportions or the performance to match the real thing; instead, I used Mainsails. I really wish they had 2000kN and a wider gimbal range, but to get that I would need to build a 1.25m cluster and that just looks awful. The ET is somewhat smaller proportionately than the real Shuttle's because of the performance and flight profile differences between KSP and RL. Booster burn time is within a few seconds of the real thing (124s RL, 121.6s here). The Vector engines are thrust-limited; they provide about 25% of the liftoff thrust (vs. the Shuttle's 18%, consequence of the smaller ET).

 

Craft file: https://www.dropbox.com/s/56mup2zahma373s/Shuttle Stack.craft?dl=0

Action groups: 1 controls the body flap, 2 deploys the inner elevons, 3 deploys the split-tail airbrake, 7 deploys the ladder, 8 opens the cargo bay, and 0 deploys the payload.

Be careful flying it -- SAS does not like off-center thrust and the off-center gimballing makes manual steering sort of squirrelly.

Edited November 12, 2016 by foamyesque

[Firemetal]

1 hour ago, foamyesque said:

Well, the thing about trying to build a shuttle replica in KSP is that the Shuttle was designed to solve a much harder problem than exists in KSP; if you build a realistic orbiter, the proportional fuel tanks and boosters (which don't exist in stock exactly but can sort-of be reproduced with liquid boosters, which also give you gimballing like the Shuttle SRBs had) are enormously overkill and you'd make orbit before you jettisoned the boosters, let alone the ET :v

So you need to compromise somewhere, but an asymmetrical design layout like Shuttles are extraordinarily sensitive to changes in proportions (thrust balance, mass balance, aero loading, and their changes over time), so it's mostly not worth it; if you're doing it you're doing it for the challenge and to show off

 

@Firemetal: A KSP shuttle replica is going to have most of its weight at the rear of the orbiter. It's inevitable if you're using Vectors (which are the engine with the best look match and required gimbal range). This is also more or less true of the real shuttle (though in KSP the Mk3 cockpit is far too light for accuracy -- another gripe for another time); if you look at the wing layout of the real one, it's clearly built to have a CoP way the hell near the back (and to deal with lowQ), hence the oversized aerodynamic control surfaces on the trailing delta and the gigantic tail. What you want to do is drain your OMS tanks and remove the payload, see where your CoM is, and then build things so that your CoL is near, but behind, it. The actual shuttle's wing layout works pretty well for this; you can fiddle with the angle of the leading wing strakes to adjust things, if necessary. Turn off angle-snap when you do, because the adjustments shouldn't need to be large.

 

Another thing, not mentioned AFAICT, is that the Shuttle used RCS thrusters to provide attitude control; don't be afraid to do so too. They provide an awful lot of oomph in the early reentry to keep you in your desired attitude, and can be helpful in pushing you out of an intentional stall or near-stall and into a glide path, since your aerodynamic control surfaces may be stalled out and not provide an adequate change to the drag profile. The shuttle also had something called a "body flap" which is often ignored in KSP replicas; it's a large flap mounted under the engines, which provides heat shielding and can be deployed to reduce loading on the elevons.

TBQH my usual problem with shuttle flights (aside from the normal hassle of balancing the launch stack) is yaw. Even with the massive spaceplane tailfin, the CoM is often so far back that it struggles and results in a really nasty sideslip-roll interaction from adverse yaw.

 

Anyway! So I built one.

 

It's pretty close to the Shuttle, but not perfect.

Payload's approximately the same (~30t to low orbit), but this only carries 6 Kerbals to the Shuttle's 7 people. No Canadarm, because stock. Has the same split-tail tailfin (via clipping, which also helped fix my yaw issues :v) to act as an airbrake, OMS fuel is internal instead of mounted in a replaceable pod. Obviously, it uses a LFB since we don't have any SRBs that have either the proportions or the performance to match the real thing; instead, I used Mainsails. I really wish they had 2000kN and a wider gimbal range, but to get that I would need to build a 1.25m cluster and that just looks awful. The ET is somewhat smaller proportionately than the real Shuttle's because of the performance and flight profile differences between KSP and RL. Booster burn time is within a few seconds of the real thing (124s RL, 121.6s here). The Vector engines are thrust-limited; they provide about 25% of the liftoff thrust (vs. the Shuttle's 18%, consequence of the smaller ET).

 

Craft file: https://www.dropbox.com/s/56mup2zahma373s/Shuttle Stack.craft?dl=0

Action groups: 1 controls the body flap, 2 deploys the inner elevons, 3 deploys the split-tail airbrake, 7 deploys the ladder, 8 opens the cargo bay, and 0 deploys the payload.

Be careful flying it -- SAS does not like off-center thrust and the off-center gimballing makes manual steering sort of squirrelly.

Hey thanks. I just tried to build a simple shuttle that doesn't use canards but flies as if it has. It worked. I mean re-entering at 30 degrees is pretty darn stable. However I have figured it out in building it. 

I used monopropellant for the OMS so that I could use less mass in the back and make it easier to balance. Monoprop doesn't give a lot of DV, especially with vectors but it sure lightens the load.  In the end, it comes down to balance. I know you guys have been yelling this at me but you either have to add canards and balance the lift or balance the CoM by adding mass at the front or minimizing it at the back. It makes sense now. 

Thanks for your help guys. I really appreciate it. Will be a long time before I try to replicate it like that again.  

Fire

[foamyesque]

My biggest reason for using a monoprop tank, over a LFO one, was that the Mk3 monoprop tank is shorter -- which means the heavy vectors have less of a lever arm to work with. It's just long enough to mount a tailfin (when coupled with the engine mount) without looking incredibly stupid, too.

[Firemetal]

9 hours ago, foamyesque said:

My biggest reason for using a monoprop tank, over a LFO one, was that the Mk3 monoprop tank is shorter -- which means the heavy vectors have less of a lever arm to work with. It's just long enough to mount a tailfin (when coupled with the engine mount) without looking incredibly stupid, too.

Exactly. It is also a bit lighter than the rocket fuel tank which is why I thought using it for OMS would help my stability during reentry. But I used 6 1.25m mono prop tanks for fuel. This gave me a little more fuel to work with.

Fire

[ForScience6686]

14 hours ago, Firemetal said:

Yes but I am going for a realistic shuttle design. Maybe that isn't a good way to go. Ej_sa likes to play without using canards on his planes. However I always find planes impossible to fly without them. How would one play without them? Thanks.

Fire

I get by just fine without canards  myself.   it ask comes down to design and balance.  I also don't build for a specific look, I build for functionality.   this goes for planes and space worthy craft.   the biggest problem I see in other designs is improper placement of com.   This causes all sorts of engineering problems.   not saying it can't be done,  in fact my most prized ssto  is rear wing and uses canards.   but I spent a lot of time testing and tweaking to get mass stable and have the control I needed.

[foamyesque]

4 hours ago, Firemetal said:

Exactly. It is also a bit lighter than the rocket fuel tank which is why I thought using it for OMS would help my stability during reentry. But I used 6 1.25m mono prop tanks for fuel. This gave me a little more fuel to work with.

Fire

I have some additional tankagae in the front of the cargo bay -- helps reduce CoM shift and provides a little more deltaV for the OMS, allowing it to circularize more elliptical orbits and/or deorbit from higher up. Flown correctly you don't need much, but the margins are narrow -- I usually have under a second of burn left on the Vectors when I jettison the ET

[Firemetal]

2 hours ago, foamyesque said:

I have some additional tankagae in the front of the cargo bay -- helps reduce CoM shift and provides a little more deltaV for the OMS, allowing it to circularize more elliptical orbits and/or deorbit from higher up. Flown correctly you don't need much, but the margins are narrow -- I usually have under a second of burn left on the Vectors when I jettison the ET

It also helps if you don't move after re-entry until you are under 400-500 m/s. Stalling is quite easy when you are high up. However an unstable shuttle will stall anyways. But now I see the use of monoprop oms.

Thanks,

Fire

[foamyesque]

1 hour ago, Firemetal said:

It also helps if you don't move after re-entry until you are under 400-500 m/s. Stalling is quite easy when you are high up. However an unstable shuttle will stall anyways. But now I see the use of monoprop oms.

Thanks,

Fire

I intentionally stall out. I fly at 45-degree+ angle of attack, just like the real thing (and I do so on all my spaceplanes, it's just the best way to do it). The pile of control surfaces suffice to pull me back out of that, and the RCS acts as additional insurance. Then I cruise in on a 10-15 degree glide slope for range extension, starting from 1000-1200m/s surface velocity. The trick with a shuttle layout lies in not losing yaw stability when you roll. My free-form spaceplane layouts don't have that problem because I generally have a much further forward CoM than a KSP replica shuttle does, and use tail mounts as an extension, so my yaw stabilizer and rudder have much larger lever arms to work with.

[Firemetal]

55 minutes ago, foamyesque said:

I intentionally stall out. I fly at 45-degree+ angle of attack, just like the real thing (and I do so on all my spaceplanes, it's just the best way to do it). The pile of control surfaces suffice to pull me back out of that, and the RCS acts as additional insurance. Then I cruise in on a 10-15 degree glide slope for range extension, starting from 1000-1200m/s surface velocity. The trick with a shuttle layout lies in not losing yaw stability when you roll. My free-form spaceplane layouts don't have that problem because I generally have a much further forward CoM than a KSP replica shuttle does, and use tail mounts as an extension, so my yaw stabilizer and rudder have much larger lever arms to work with.

Yes. Sometimes when your spaceplanes get too hot, you have to. However I disagree when it comes to shuttles. The nose cones got buffed in 1.2 so they can survive 400 more kelvin than they could before. So right now, I don't think you have to stall when coming from LKO. The trick is having the COM further forward. I get that now.

Thanks,

Fire

[klesh]

Consider an internal ballast of fuel to move the COM around for re-entry; I carry an amount of monopropellant to shift my COM. 

[Firemetal]

15 hours ago, klesh said:

Consider an internal ballast of fuel to move the COM around for re-entry; I carry an amount of monopropellant to shift my COM. 

Good idea. Thanks.

Fire