Index Home About Blog
Subject: The Palpable Superiority Of Horizontal Landing
From: Mitchell Burnside Clapp <clappm@plk.af.mil> 
Date: Oct 02 1995
Organization: Phillips Laboratory
Newsgroups: sci.space.policy

************************************************************************
                        HUGE MASSIVE DISCLAIMER:

This is just me talking here.  Do not construe this posting as any 
official opinion of the US Air Force, the Phillips Lab, the Reusable 
Launch Vehicle office, or even the author.  I have not been given my 
official opinion yet.  This is posted solely to stimulate debate and 
discussion, and not to get me in really serious trouble, which 
misconstruing my intent here might otherwise cause. 

           WE NOW RETURN TO YOUR REGULARLY SCHEDULED POSTING

************************************************************************
   There has been a lot of debate, here and elsewhere, about the merits 
of vertical landing.  I thought it would be helpful and informative to 
argue the case for horizontal landing, which in my opinion is the only 
acceptable option for landing aerospace vehicles, especially relatively 
near term SSTO vehicles.

   I want to set aside altogether any discussion of takeoff modes, 
vertical or horizontal.  I'm not trying to push Black Horse or 
airbreathing solutions, or anything like that.  For similar reasons, I 
don't think discussion of landing on airless planets is relevant either.

   The issue in any landing, powered or unpowered, propulsive or 
aerodynamic, is the management of energy.  In propulsive-landing 
vehicles this is accomplished by brute force.  Gliding vehicles, on the 
other hand, convert their vertical potential energy into horizontal 
kinetic energy by means of aerodynamic lift. Either technique will work, 
of course.  We have ample evidence from soaring for the usefulness of 
glide landing, as well as the standard and routinely practiced flameout 
landing patterns for single engine fighters, as well as the seventy-odd 
successful landings of the shuttle orbiter and the several hundred 
lifting body and X-plane landings.  The evidence for the usefulness and 
safety of power-only vertical landing is considerably less, but the 
experience of the DC-X program (with which I was and am thoroughly 
involved) and the LLRV program (one of the pilots of which I have met 
and talked to at some length) argue that power-only vertical landing is 
not inconceivable.  

HOW AIRPLANES LAND

   Conventional aircraft, naturally, use both power and lift to manage 
energy in the landing.  The best way to think about this is to observe 
that the powered vehicles have an engine which permits them to flatten 
their otherwise lackluster glide ratios.  

   A typical fighter like the F-16 has a glide ratio, or lift to drag 
ratio, if you prefer, of about seven at the sort of airspeed one likes 
to land at.  The engine is set a little bit above idle, so that the 
pilot can use throttle to control his or her sink rate.  The typical 
glide slope for an aircraft in a conventional landing pattern is about 
2.5 degrees.  The corresponding L/D for a 2.5 degree landing is the 
tangent of that number, or 23.  An aircraft with an L/D of greater than 
23, such as a sailplane, will have a glide slope flatter than 2.5 
degrees.  Most aircraft have less L/D than this, so they need power to 
maintain a standard glideslope.

   There is nothing magical about this 2.5 degree number.  It is equally 
possible to land an aircraft at steeper glide angles, as for example, in 
the fighter engine-off landing patterns mentioned earlier.  Once the L/D 
declines below about 4, extensive research indicates that the handling 
qualities decline from satisfactory to acceptable, according to the 
scale test pilots use to measure such things (see for example, Jackson, 
Rivers, and Bailey, "Effect of Lift to Drag Ratio in Pilot Rating of the 
HL-20 Landing Task", J. Spacecraft and Rockets, Sept-Oct 1993, Vol. 30, 
No. 5, p. 543-548).  

   When landing an aircraft with the power off, the pilot must keep the 
airspeed high until he or she is ready to bring the flight path level 
with the ground and land.  This is called flaring.  The aircraft begins 
to lose airspeed, but the pilot has the gear down and locked and is 
ready gently to set the aircraft on the runway.  In this type of landing 
the pilot has committed to the landing before flaring.  In a real 
flameout landing situation, the pilot decides whether to attempt the 
landing or jump out at 2000 feet above the ground.  This is before the 
flare, at a point called "Base Key."  

   Naturally, there are redundant extremely reliable hydraulic systems, 
with power available to drive them from any of several sources, such as 
Ram Air Turbines (A-7s for example) or Emergency power units (F-16s).  
These hydraulic systems permit the control surfaces of the aircraft to 
operate normally even if the engine is off.  They also can operate the 
landing gear and other utilities on the aircraft.

   The reason to point all this out is not to be pedantic (which I am 
being, slightly, I admit) but rather to define the context of what is 
meant by talking about landing aerospace vehicles horizontally.

HOW ROCKETSHIPS LAND

   By rocketships, here, I mean the DC-X-type vertical landers, for want 
of a better spectrum of examples.  Whether the reentry is pointy-end-
first, as in the Delta Clipper designs, or hot-end-first, as in other 
concepts, the key thing that absolutely must happen is an engine start.

   For the sake of argument, I will assume that the vehicle is a LO2/LH2 
burning bipropellant vehicle. The vehicle weighs ten percent of its 
gross lift off weight (GLOW) at landing. It has solved the problem of 
throttling an engine to 10 percent of its rated thrust level, perhaps by 
having dedicated landing engines or some other means.  Similarly, it is 
not concerned with flow separation in the nozzles, perhaps because of an 
advanced nozzle design.  

   Before the engine can be turned on, it requires chilldown.  The 
amount of propellant needed to do this is difficult to quantify, but a 
typical value is the equivalent of 4 seconds of full thrust for the 
engine.  The propellant valves are opened, the igniter is fired and does 
not fail, combustion begins, the turbomachines are brought up to full 
speed, and the engine begins to perform as designed.  None of the 
motions of the vehicle cause an interruption in propellant supply. The 
power takeoff for the engine gimbals works flawlessly. As the vehicle 
approaches the ground and enters the ground effect, the normal shock 
between the vehicle and the ground causes no instability or disturbance 
beyond which the propulsive control system cannot cope.  The thrust 
declines from a maximum value of three times the empty weight at startup 
(Digression: this value derives from the need to decelerate the vehicle 
rapidly from terminal velocity and also to account for engine out) to 
80% of the weight at landing. No wind gust tips the vehicle over during 
or after landing.

   If this sounds a little complicated, it is.  To start a rocket engine 
is not an easy matter, especially in field conditions.  People who have 
studied the SSME argue persuasively that that engine is so difficult to 
thread through the labyrinth of its start sequence that it is 
essentially impossible to start in flight.  Admittedly the SSME is far 
from the last word in operability or ruggedness.  LO2/LH2 engines have 
been started in flight before (RL-10, J-2), but never thirty seconds 
away from a catastrophic failure if they fail to light.

   Why thirty seconds?  Well, if we assume that the vehicle has a sea 
level Isp of 360 seconds, then time of landing engine operation is given 
by the product of the landing propellant mass fraction and the specific 
impulse, divided by the mean thrust to weight ratio for the landing 
maneuver.  I used a 10% propellant mass fraction and 20% excess thrust 
from the initial start to engine off.  (A wing, by the way, appears to 
weigh between 5 and seven percent of the landed mass).  This includes no 
allowance for chilldown, reserves, crosswinds, go-around, or anything 
else that requires propellant. If a failure occurs with the engine 
system, of course, the time to catastrophe is much less than 30 seconds, 
since there is nothing to decelerate the vehicle. 

   It is possible to design an engine that is reliable enough to bet 
your life on for a vertical landing.  The Lunar Lander Descent Engine 
was capable.  It had hypergolic propellants, was pressure fed, was 
designed with an extremely forgiving injector design, and had redundant 
fluid lines to the engine. That's in 1/6 g and with no atmospheric 
effects to complicate the landing.  Doing the same thing on earth would 
be significantly harder.

WIND EFFECTS

   Any pilot has to be respectful of the wind during landing.  Let us 
assume that a horizontal lander and a vertical lander are hit by the 
same crosswind gust.  They have similar dimensions, and so are subject 
to similar side forces.  The vertical lander counters the side forces by 
gimballing the engines.  Depending on the drag coefficient, this 
requires swinging the engines by as much as 10 degrees to counter a 35 
knot wind.  A horizontal lander is traveling through the air at about 
150 knots for landing, and requires a relatively small deflection of a 
control surface to maintain runway heading.  There is also the ability 
to back into the gust and land slightly wing-low.

  Wind shear affects both types of vehicles in similar ways.  The 
effects of a microburst are bad on either type of vehicle, but it is a 
mistake to assume that a horizontal lander suddenly stops flying as soon 
as a microburst hits it.  There is a loss of lift, of course, but since 
the landing pattern is not flown close to the stall angle of attack, the 
aircraft merely increases its sink rate.  This could cause a hard 
landing, but not a loss of controlled flight.  The commercial aviation 
mishaps attributable to wind shear are generally with aircraft operating 
at slow speeds, at very high lift coefficients, and hence closer to the 
edge of their envelopes than a landing SSTO vehicle.  Vertical landers 
have thrust to overcome the downdraft, but if the microburst strikes at 
the wrong moment a hard landing is also a possibility.

   Wind shear is highly correlated with observable meteorological 
phenomena, and frankly the sensible thing to do if there are 
thunderstorms within 25 miles of the landing site is to pick another 
landing site. 

GO-AROUND

   Mandatory unpowered landings are utterly commonplace in soaring, in 
military aviation, and in Shuttle operations.  They can be rehearsed and 
practiced until they are as safe and reliable as any other sort of 
landing.  However, a go-around capability is always nice to have, and 
can be achieved with a horizontal lander by carrying back propellant and 
lighting an engine.  The thrust required is less than that needed for 
the vertical lander by a factor of L/D at least, and the duration is 
likewise less than that for a vertical lander, since all that is needed 
is enough additional energy to return the aircraft to high key.  About 
six seconds at thrust equal to 30% of weight gets a horizontal lander 
enough energy for another landing attempt, which might be desired on as 
many as ten percent of sorties. The vertical lander, on the other hand, 
could possibly include a few seconds of additional flight to divert to 
another landing spot.  This is probably more than six seconds and 
certainly more than 30% of vehicle weight.

WHERE TO LAND

   I believe that the supposed advantages of vertical takeoff and 
landing in terms of landing sites are illusory.  Even the pilots of 
Harrier jet fighters cannot land anywhere they choose, but must be 
concerned with the temperature of their exhaust on the terrain beneath 
them, and that is with subsonic exhaust from a non-afterburning 
turbofan.  Infrastructure will be required to take off, especially since 
one needs to be a few bell-diameters off the ground to ignite your 
engines and also to land safely.  Likewise a concrete pad will be needed 
for landing, or something like it.  Horizontal landers are constrained 
to runways, but we have many many runways, and an SSTO vehicle will have 
landing distances much less than the Shuttle orbiter because of its 
lower wing loading and so forth.

THE BOTTOM LINE

   Vertical landing is far too risky.  If the engines don't start, 
you're dead in seconds, and engines of the quality needed to get to 
orbit in the first place are not easy to start.  The design compromises 
needed to achieve similar levels of safety and reliability to horizontal 
landers are crippling.  There is at best a thin case for vertical 
landing being desirable even if these other criticisms did not apply.


   Okay gang, feel free to flame me.


-- 
                1965        1975         1985         1995
Small Car     VW Beetle   VW Rabbit   Honda Civic   Dodge Neon
Fighter         F-4         F-15        F-117         F-22
Passenger Jet   707         747          767          777

Space launch   Delta       Delta        Delta         Delta
               Atlas       Atlas        Atlas         Atlas
               Titan       Titan        Titan         Titan

The opinions expressed here are solely my own, and do not reflect 
those of the USAF, DOD, or anyone else.


Subject: Re: The Palpable Superiority Of Horizontal Landing
From: henry@zoo.toronto.edu (Henry Spencer) 
Date: Oct 03 1995
Newsgroups: sci.space.policy

(My, this started out to be a short response, and turned out a wee bit
long...  I should know better.)

In article <44poai$4og@fg1.plk.af.mil> Mitchell Burnside Clapp <clappm@plk.af.mil> writes:
>...We have ample evidence from soaring for the usefulness of 
>glide landing, as well as the standard and routinely practiced flameout 
>landing patterns for single engine fighters, as well as the seventy-odd 
>successful landings of the shuttle orbiter and the several hundred 
>lifting body and X-plane landings.  The evidence for the usefulness and 
>safety of power-only vertical landing is considerably less, but the 
>experience of the DC-X program ... and the LLRV program...

I miss mention of the Harrier and the Yak-36 here.  Many Harrier landings
are made rolling, which arguably disqualifies them, but some are done
truly vertically because the clear area is not large enough for anything
else.  And the Yak-36 normally does (did?) both takeoffs and landings
vertically, and in fact for a while Western observers thought that it was
incapable of rolling takeoffs and landings (although I believe this turned
out to be incorrect).  On the Harrier, and presumably the Yak-36 as well,
a vertical landing is purely powered, with control by RCS jets, no
aerodynamics involved except incidentally.  I don't have numbers on tap,
but this surely amounts to a considerably larger body of operational
experience than DC-X and the LLRV/LLTVs. 

>   For the sake of argument, I will assume that the vehicle is a LO2/LH2 
>burning bipropellant vehicle...

Nit-picky point here:  you are assuming that pump-fed cryogenic main
engines imply pump-fed cryogenic landing engines, which is plausible, but
not a logical necessity if different engines are used.

Although it would be stretching a point, the Soyuz landing rockets might
be considered a relevant example here.  If your terminal velocity is low
enough and the reliability of the landing engines good enough, they can
be lit just before touchdown.

>The vehicle weighs ten percent of its 
>gross lift off weight (GLOW) at landing. It has solved the problem of 
>throttling an engine to 10 percent of its rated thrust level, perhaps by 
>having dedicated landing engines or some other means.  Similarly, it is 
>not concerned with flow separation in the nozzles, perhaps because of an 
>advanced nozzle design.

This strikes me as slanted in contradictory directions. :-) If you have
dedicated landing engines, flow separation is a non-issue with plain
ordinary vanilla bell nozzles, because the nozzles are optimized for
landing.  If you are worried about flow separation and need advanced
nozzles, it is because you don't use dedicated landing engines. 
Furthermore, any SSTO capable of a vertical takeoff has solved the
flow-separation problem somehow; this whole issue is a red herring. 

Apart from the possibility of engines with wide throttling range -- the
RL-10 was demonstrated down to 1% thirty years ago -- the orthodox
solution to the throttling problem is to use only some of the engines. 
This dovetails nicely with engine-out capability on launch.  It would be
fairer to cite this approach than to raise the bogeyman of dedicated
landing engines. 

>...The propellant valves are opened, the igniter is fired and does 
>not fail, combustion begins, the turbomachines are brought up to full 
>speed, and the engine begins to perform as designed.  None of the 
>motions of the vehicle cause an interruption in propellant supply. The 
>power takeoff for the engine gimbals works flawlessly...

Why did you not list the same issues for the APUs powering the hydraulics
in the horizontal lander?  Again, I would call this a slanted presentation;
if these problems can be solved for APUs, they can be solved for engines,
and if they're scary for engines, they're almost as scary for APUs.

Also, using a subset of the takeoff engines for landing permits a highly
redundant start sequence (at the price of some small extra propellant
consumption):  attempt to start *all* the engines, pick a suitable subset
of the ones that started, and shut down the rest.

>As the vehicle 
>approaches the ground and enters the ground effect, the normal shock 
>between the vehicle and the ground causes no instability or disturbance 
>beyond which the propulsive control system cannot cope...

This is, or should be, a design-and-test issue rather than a reliability
issue.  Get it right once and it should stay right.  This might make or
break an SSTO design, but it should not represent a hazard to individual
operational vehicles.  It doesn't belong in this part of the discussion. 

>...No wind gust tips the vehicle over during or after landing.

A problem which is shared by horizontal landers, a particularly notable
example being the U-2.  This should not be cited as if it were unique
to vertical landers.

>   If this sounds a little complicated, it is.  To start a rocket engine 
>is not an easy matter, especially in field conditions.  People who have 
>studied the SSME argue persuasively that that engine is so difficult to 
>thread through the labyrinth of its start sequence that it is 
>essentially impossible to start in flight.  Admittedly the SSME is far 
>from the last word in operability or ruggedness...

So why cite it as the main example?  It would be fairer to base the
discussion on the J-2 and RL-10 experience -- since those engines have
spent their entire operational careers (with the exception of the RL-10s
on DC-X) being started in flight -- and just mention that it's not as easy
for the SSME.  The sound of grinding axes is heard here. :-)

>LO2/LH2 engines have 
>been started in flight before (RL-10, J-2), but never thirty seconds 
>away from a catastrophic failure if they fail to light.

Actually, I believe certain cases of multiple J-2 ignition failure on the
Saturn V second stage involved near-certain catastrophic failure within a
small number of seconds.  Since we have engine-out capability for the VL
SSTO on landing -- you assumed it when citing landing thrust requirements
-- then the two situations are closely parallel.  Nothing new and scary
here. 

Also, all ignition failures on Centaur are catastrophic; the RSO pushes
the button well within 30s if one engine fails to light.

>   It is possible to design an engine that is reliable enough to bet 
>your life on for a vertical landing.  The Lunar Lander Descent Engine 
>was capable...

Actually, one can argue the relevance of this.  Except in the final
seconds of descent, the LM ascent engine was always available as backup.

I also note that almost all of the return-to-the-Moon proposals from NASA,
in the brief time a few years ago when the idea was taken seriously, used
the RL-10 as the descent engine.  Evidently it is now considered reliable
enough for the job.  (P&W actually thought it was good enough in the 60s;
that's why they demonstrated deep throttling.)

>It had hypergolic propellants, was pressure fed, was 
>designed with an extremely forgiving injector design, and had redundant 
>fluid lines to the engine. That's in 1/6 g and with no atmospheric 
>effects to complicate the landing...

Why are the G-load and atmospheric effects relevant to engine reliability?
These, again, would seem to be design issues rather than reliability issues:
things that would be important to a test program but which would not take
operational vehicles by surprise.

>   Any pilot has to be respectful of the wind during landing.  Let us 
>assume that a horizontal lander and a vertical lander are hit by the 
>same crosswind gust.  They have similar dimensions, and so are subject 
>to similar side forces...

However, the overall forces can be substantially different, which
complicates matters.  For example, dihedral effect introduces roll in the
horizontal lander (because the effective sweep angle of the windward wing
is reduced, and that of the leeward wing increased).  As I recall, this
often defines the limits on crosswind performance for winged vehicles.
The vertical lander does not need control authority to cover this, since
with no wings it doesn't have dihedral effect.

>The vertical lander counters the side forces by 
>gimballing the engines.  Depending on the drag coefficient, this 
>requires swinging the engines by as much as 10 degrees to counter a 35 
>knot wind.  A horizontal lander is traveling through the air at about 
>150 knots for landing, and requires a relatively small deflection of a 
>control surface to maintain runway heading...

Mitch, I'm not a test pilot... but my impression is that a 35kt crosswind
is in fact at the edge of, if not beyond, what most horizontal landers can
handle.  The one aircraft for which I actually have the number handy is
the YF-12A, where the design limit is in fact 35kt.

>   I believe that the supposed advantages of vertical takeoff and 
>landing in terms of landing sites are illusory.  Even the pilots of 
>Harrier jet fighters cannot land anywhere they choose, but must be 
>concerned with the temperature of their exhaust on the terrain beneath 
>them, and that is with subsonic exhaust from a non-afterburning turbofan...

Actually, Harrier exhaust is almost entirely benign -- the concern there
is for kicking up debris rather than temperature damage.  Harriers
operated successfully for years off the old Spanish carrier Dedalo, which
had a wooden flight deck.

I concede that more caution will be needed with rockets.  However, DC-X
has already supplied an example of an emergency rocket landing -- at
relatively high gross weight, without preliminary fuel dumping -- on
unprepared terrain with minimal damage. 

>...Infrastructure will be required to take off...

Certainly; in the event of an emergency landing on an unprepared site, it
will probably be preferable to move the vehicle by crane or helicopter (at
least, as far as the nearest place where suitable temporary infrastructure
can be set up easily) rather than trying to fly it out under its own
power.  However, there's still an advantage here for emergency landings. 

>...Horizontal landers are constrained 
>to runways, but we have many many runways, and an SSTO vehicle will have 
>landing distances much less than the Shuttle orbiter...

Some areas have many many runways.  Others don't.  Clearly the need for
landing facilities is not an issue (or not much of one, anyway) if we are
bringing the vehicle down as planned.  When it comes to emergencies, I
note that with some frequency, the outcome of aircraft accidents turns on
whether there was a runway in just the right place or not.
-- 
The problem is, every time something goes wrong,   |       Henry Spencer
the paperwork is found in order... -Walker on NASA |   henry@zoo.toronto.edu

Newsgroups: sci.space.policy
From: Henry Spencer <henry@zoo.toronto.edu>
Subject: Re: MSFC engine
Organization: SP Systems, Toronto
Date: Sun, 24 Dec 1995 16:56:19 GMT
Lines: 14

In article <4bhuda$1buq@usenetw1.news.prodigy.com> MMFF37A@prodigy.com (Michael Gallagher) writes:
>...[vertical landing]  Tha'ts not even remortely an "off the 
>shelf" technique.  (And also why horixontal landings are nicer --- no 
>feule required beyond the needs of your retros.)

The weight of fuel needed for vertical landing is less than the weight of
the wings, control surfaces, etc. needed for horizontal landing.  Horizontal
landing is "nice" only because it is familiar; in many ways it's an inferior
method, historically necessary because aircraft engines were so feeble.  The
thrust:weight ratio of rocket engines is high enough that the idea deserves
to be taken seriously.
-- 
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |   henry@zoo.toronto.edu

From: burnside@bix.com (burnside)
Newsgroups: sci.space.policy
Subject: Re: MSFC engine
Date: 26 Dec 1995 00:12:06 GMT

Henry Spencer (henry@zoo.toronto.edu) wrote:
: In article <4bhuda$1buq@usenetw1.news.prodigy.com> MMFF37A@prodigy.com (Michael Gallagher) writes:
: >...[vertical landing]  Tha'ts not even remortely an "off the 
: >shelf" technique.  (And also why horixontal landings are nicer --- no 
: >feule required beyond the needs of your retros.)

: The weight of fuel needed for vertical landing is less than the weight of
: the wings, control surfaces, etc. needed for horizontal landing.  Horizontal
: landing is "nice" only because it is familiar; in many ways it's an inferior
: method, historically necessary because aircraft engines were so feeble.  The
: thrust:weight ratio of rocket engines is high enough that the idea deserves
: to be taken seriously.
: -- 
: Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
: Pay no attention to that cliff ahead...            |   henry@zoo.toronto.edu

I dispute the assertion that the weight of the fuel needed for
vertical landing (assuming by 'fuel' you actually mean propellant)
is less than the horizontal landing hardware.  Most studies that
examine horizontal landers wind up with 5-6% of the landed mass
allotted to HL-specific systems.  I have never seen a vertical
lander with less than ten percent of its mass at landing set aside
for landing propellant. (Think of it as thirty seconds).

Mitchell Burnside Clapp

Newsgroups: sci.space.policy
From: Henry Spencer <henry@zoo.toronto.edu>
Subject: landings (was Re: MSFC engine)
Date: Fri, 29 Dec 1995 04:43:19 GMT

In article <4bnekm$o47@news1.delphi.com> burnside@bix.com (burnside) writes:
>...Most studies that
>examine horizontal landers wind up with 5-6% of the landed mass
>allotted to HL-specific systems.  I have never seen a vertical
>lander with less than ten percent of its mass at landing set aside
>for landing propellant. (Think of it as thirty seconds).

As I've said in the past, I don't dispute Mitch's calculations but I do
dispute some of the assumptions underlying them (and the HL studies he
refers to).

As a case in point, the VL studies are for powered landings.  The HL studies
are not.  To make the comparison realistic, the HL studies must include a
go-around capability, which most of them lack at 5-6%.  How much does this
add?  Well, off the top of my head, assume we need a delta-V of 100m/s for
a worst-case go-around; that's enough added energy to return to maybe 2000ft,
depending on details.  That's equivalent to ten seconds of hover, another
3-4% of landed mass.  Suddenly the difference looks a lot smaller...

Of course, you can say that we don't need go-around because we'll always
land right the first time.  In which case, I can say that we don't need
thirty seconds of fuel for VL, because the guidance problem is, if
anything, rather simpler for VL.  We need maybe ten seconds, speaking
very conservatively, to kill the terminal velocity.  Now we're talking
5-6% for HL and 3% for VL.

I would also observe that there are mass penalties in systems which don't
look HL-specific at first glance, such as primary structure (more varied
loads from more points and more directions) and thermal protection (larger
area and [assuming a high-L/D reentry] more prolonged heating).
-- 
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |   henry@zoo.toronto.edu

From: Mitchell Burnside Clapp <clappm@plk.af.mil>
Newsgroups: sci.space.policy
Subject: Re: landings (was Re: MSFC engine)
Date: Fri, 29 Dec 1995 12:05:55 -0800

Just because the landing is powered does not mean that a vertical lander 
has a go-around capability. 

We know that the thrust, T, for a vertical lander is related  to the 
required number of g's of deceleration (a), the weight of the vehicle 
(W), and the specific impulse (Isp), and the weight flow rate (dW/dt) by:

                       T = a W = Isp (dW/dt)                      (1).

We can calculate the landing propellant mass fraction by multiplying by 
the time (dt) and rearranging terms:

                     1 - (W-dW)/W =  a (dt)/Isp                   (2).

This part is just physics.  The engineering comes in when you have to put 
in some numbers.  I will use the sea level Isp of the SSME at 370.7 sec. 
The value of (a) is a little harder to substantiate, but McDonnell 
Douglas is decelerating at 2 g's after a 5 second chilldown period last I 
heard, then throttling back to 0.8 g's only at the very end. Let's skip 
the chilldown part and say that it all time-averages out to 1.5 g of 
deceleration. 

Under these assumptions, with no other margins for anything, a vertical 
lander can just squeak down with four percent of its landed mass consumed 
in ten seconds.  This is enough to kill a terminal velocity of 288 knots, 
assuming no control penalties, no chilldown, and deceleration to a 
perfect stop exactly on the pad, with no winds.

Chilldown consumes, at a bare minimum, four seconds of equivalent 
full-flow propellant.  Residuals for LO2/LH2 systems run about 1% of tank 
capacity in a well-designed system.  Control penalties (steeering, 
targeting, and the like, to enable the vehicle to land on its intended 
point, again with no margin or reserve) would, I assert without proof, 
consume an additional 30 percent of the landing fuel.  The result is an 
equivalent time of (10+4)* 1.3 * 1.01 seconds, giving 7.4 percent 
propellant at landing.  Most prudent VTVL designers add some margin to 
take care of things like winds and so on, which adds about another 5 
seconds of flight time. This brings you up to a ten percent propellant 
mass fraction at landing, and you still don't have anything like what a 
pilot would call a go-around capability.

Mitchell Burnside Clapp
X-33 Operations Officer
-- 
                1965        1975         1985         1995
Small Car     VW Beetle   VW Rabbit   Honda Civic   Dodge Neon
Fighter         F-4         F-15        F-117         F-22
Passenger Jet   707         747          767          777

Space launch   Delta       Delta        Delta         Delta
               Atlas       Atlas        Atlas         Atlas
               Titan       Titan        Titan         Titan

The opinions expressed here are solely my own, and do not reflect 
those of the USAF, DOD, or anyone else.

From: Mitchell Burnside Clapp <clappm@plk.af.mil>
Newsgroups: sci.space.policy
Subject: Re: landings (was Re: MSFC engine)
Date: Tue, 02 Jan 1996 11:17:04 -0800

Richard A. Schumacher wrote:
 
> In <30E44A23.2E7A@plk.af.mil> Mitchell Burnside Clapp <clappm@plk.af.mil> writes:
> 
> >Just because the landing is powered does not mean that a vertical lander
> >has a go-around capability.
> 
> You argue persuasively that a vertical lander will have some margin
> but not enough for go-around capability. How does this lead to the
> conclusion that a horizontal lander, with no more margin and less
> operational flexibility, is the better choice?

Under the limited terms of the argument presented (compare no-go-around
HL with no-go-around VL) I conclude that HL is the better choice
because a: it weighs less and b: it does not require an engine start
(the safety arguments against which have been made elsewhere). It is
also easier to add glide-stretching or an actual go-around to an HL
than it is to a VL, but that's a bit outside the envelope of the
argument I was making.

What is your case for claiming that a VL has better operational
flexibility? You can't just land it any old where. It is wanting in
crossrange, typically. You cannot identify an engine failure until it
is too late to do anything about it.  It's not the airless planet
argument, is it?  I'm interested.

> Changing the subject:
> 
> >X-33 Operations Officer
> 
> What does that job entail?

Operationally, emptying coffee cups and filling up wastebaskets, much
like everyone's job.  It would be much more inmpressive if there were
an X-33 flying.



Newsgroups: sci.space.policy
From: Henry Spencer <henry@zoo.toronto.edu>
Subject: Re: landings (was Re: MSFC engine)
Date: Wed, 3 Jan 1996 21:46:16 GMT

>[vertical landing]
>You cannot identify an engine failure until it is too late to do 
>anything about it...

"If they hand you a lemon, make lemonade."  If ignition failures (the most
likely form of engine failure) are considered a significant issue, attempt
to light more engines than you need, and then shut down or throttle back
the unnecessary ones.  There is a fuel cost for this, but no fundamental
technical problem.  A variety of related schemes are conceivable, notably
the popular engine-out-during-launch idea of throttling the remaining
engines up to emergency maximum when one fails.

(I would add that such startup sequences would naturally be automated,
just like the startup sequence of an individual engine, so that decisions
can be made and acted on in fractions of a second.)

Once running, engines usually keep running.

It is also possible -- again at some cost in fuel -- to just light the
engines early, when there's time to think about problems, and hold them at
low throttle until needed.  This does depend somewhat on throttling range;
it's much more practical with RL10s (which can go down to 1%) than with
SSMEs (which hit stability limits somewhere around 65%). 
-- 
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |   henry@zoo.toronto.edu



Newsgroups: sci.space.policy
From: Henry Spencer <henry@zoo.toronto.edu>
Subject: Re: landings (was Re: MSFC engine)
Date: Thu, 4 Jan 1996 16:44:21 GMT

In article <4cft34$rke@news.ro.com> nhendrix@ro.com (Doug Hendrix) writes:
>>"If they hand you a lemon, make lemonade."  If ignition failures (the most
>>likely form of engine failure) are considered a significant issue, attempt
>>to light more engines than you need, and then shut down or throttle back
>>the unnecessary ones...
>
>Carrying more engines than are needed for a decent T/W liftoff ratio
>is also a weight penalty...

Indeed correct, but a VTVL SSTO weighs perhaps ten times as much at takeoff
as it does at landing.  Enough engines to give decent takeoff T/W is far
more than needed for landing, and landing is the critical case for ignition
failures.  (If an engine doesn't light at takeoff, abort the takeoff.)
-- 
Look, look, see Windows 95.  Buy, lemmings, buy!   |       Henry Spencer
Pay no attention to that cliff ahead...            |   henry@zoo.toronto.edu

Index Home About Blog