Multiple Engines???

(George William Herbert) wrote:
Henry Spencer wrote:
[...]
The practical problem with biamese and triamese is that almost anything
you do to simplify the boosters starts you off down the slippery slope of
building two different vehicles. It's very hard to stop that.

Just leaving systems out looks easy, but often it means a lot of extra
engineering to assess what *happens* when you leave those systems out,
and what drives development cost is not materials but engineering effort.
Later on, when weight is excessive or there's a bit of a performance
shortfall, well, we're already building two different configurations, so
we'll just make them a little *more* different...

Biamese or triamese is a win only if the boosters are the *same* as the
orbiter. Same TPS; if it doesn't get as hot, that's nice. Same OMS;
okay, we can leave its tanks empty on the boosters. Same systems, all of
them. Maybe we fill the boosters' cargo bays with tanks, but if so, any
permanent fittings we need to add go in the orbiters too. It takes very
strong engineering leadership to make this work.

I would pushback from that some. But the sentiment is clearly
thinking smart.

If you're trying to do highly commonal biamese/triamese, I would
lay down some groundrules:
1) Same airframe structure. That means same parts, same holes,
same brackets, the whole nine yards.
2) Same TPS, per Henry.
3) Same systems for power, control, guidance, etc.
4) Same mechanisms (control surfaces, doors, gear, etc).
5) Same wiring harness.
6) Same main engines powerhead; a different nozzle is acceptable
between the orbital and booster models, but one should be able to
put a box over the nozzle end and tape cover any other identifying
markings and stump a tech on whether it's the long or short nozzle
model engine.

{...]
Basically... there should be two sets of things. The Airframe,
which is structures and systems which are common, and those should
be *common*... the fitout for changing one model into the other
model should be not significantly more than normal minor overhaul.
And then modular equipment sets that change between the two
missions (or more, if you have low/hi/orbital rather than
booster/orbital). All the interfaces need to be common,
and you need to enforce on the design team (and ideally on
the operations team) that airframes are not going to be
shoehorned into either role. Establishing that in the
operations and maintenance schedule model would be great.



I would think that one of the few ways to enforce the rule is to make
any unit be the orbital unit at some point. For instance, units ABC
are used in the first launch, A going to orbit, B to high staging, and
C to low staging. B then becomes the orbital unit for the 2nd launch
with CD staging, C the orbital unit with the 4th launch, with DA
staging....

Having modular assemblies would help a great deal, too, I would think.

/dps

(George William Herbert) wrote:
Henry Spencer wrote:
[...]
The practical problem with biamese and triamese is that almost anything
you do to simplify the boosters starts you off down the slippery slope of
building two different vehicles. It's very hard to stop that.

Just leaving systems out looks easy, but often it means a lot of extra
engineering to assess what *happens* when you leave those systems out,
and what drives development cost is not materials but engineering effort.
Later on, when weight is excessive or there's a bit of a performance
shortfall, well, we're already building two different configurations, so
we'll just make them a little *more* different...

Biamese or triamese is a win only if the boosters are the *same* as the
orbiter. Same TPS; if it doesn't get as hot, that's nice. Same OMS;
okay, we can leave its tanks empty on the boosters. Same systems, all of
them. Maybe we fill the boosters' cargo bays with tanks, but if so, any
permanent fittings we need to add go in the orbiters too. It takes very
strong engineering leadership to make this work.

I would pushback from that some. But the sentiment is clearly
thinking smart.

If you're trying to do highly commonal biamese/triamese, I would
lay down some groundrules:
1) Same airframe structure. That means same parts, same holes,
same brackets, the whole nine yards.
2) Same TPS, per Henry.
3) Same systems for power, control, guidance, etc.
4) Same mechanisms (control surfaces, doors, gear, etc).
5) Same wiring harness.
6) Same main engines powerhead; a different nozzle is acceptable
between the orbital and booster models, but one should be able to
put a box over the nozzle end and tape cover any other identifying
markings and stump a tech on whether it's the long or short nozzle
model engine.

{...]
Basically... there should be two sets of things. The Airframe,
which is structures and systems which are common, and those should
be *common*... the fitout for changing one model into the other
model should be not significantly more than normal minor overhaul.
And then modular equipment sets that change between the two
missions (or more, if you have low/hi/orbital rather than
booster/orbital). All the interfaces need to be common,
and you need to enforce on the design team (and ideally on
the operations team) that airframes are not going to be
shoehorned into either role. Establishing that in the
operations and maintenance schedule model would be great.



I would think that one of the few ways to enforce the rule is to make
any unit be the orbital unit at some point. For instance, units ABC
are used in the first launch, A going to orbit, B to high staging, and
C to low staging. B then becomes the orbital unit for the 2nd launch
with CD staging, C the orbital unit with the 4th launch, with DA
staging....

Having modular assemblies would help a great deal, too, I would think.

/dps

The practical problem with biamese and triamese is that almost anything
you do to simplify the boosters starts ... building two different vehicles.

Yes indeed. I know this well.
I would be happy if, after operational testing, 20% commonality
remained. But design and fly first, optimise later.

The practical problem with biamese and triamese is that almost anything
you do to simplify the boosters starts ... building two different vehicles.

Yes indeed. I know this well.
I would be happy if, after operational testing, 20% commonality
remained. But design and fly first, optimise later.

Even at mach 8 separations, is the first stage all that far down range,
more than 100-200 miles? It'll have a lot of energy from altitude and ...

My trusty spreadsheet says 2,247 m/sec, 74 km up and 93 km downrange.

Even at mach 8 separations, is the first stage all that far down range,
more than 100-200 miles? It'll have a lot of energy from altitude and ...

My trusty spreadsheet says 2,247 m/sec, 74 km up and 93 km downrange.

(Henry Spencer) wrote in message ...
In article ,
Charles Talleyrand wrote:
In a world where engines explode upon failure, having exactly one
engine per stage is best.

Fortunately, this is not such a world. Explosions of fully-developed
liquid-fuel engines are very rare. (Solids are another matter.) They
may refuse to work but they seldom do anything messy.

It has been suggested to me in private email that the correct answer to
this delima is to have one engine
FAMILY, but with multiple engine sizes per family.

As has already been noted, this doesn't work, because many aspects of
rocket engines don't scale straightforwardly.

The only good example I can recall of liquid fueled rocket engines
failing in catastrophic and dramatic fashion in a developed rocketry
program (i.e. after the initial development of orbital rocketry) would
be the N-1, and there are some fairly good reasons to take that as a
special case.

(Henry Spencer) wrote in message ...
In article ,
Charles Talleyrand wrote:
In a world where engines explode upon failure, having exactly one
engine per stage is best.

Fortunately, this is not such a world. Explosions of fully-developed
liquid-fuel engines are very rare. (Solids are another matter.) They
may refuse to work but they seldom do anything messy.

It has been suggested to me in private email that the correct answer to
this delima is to have one engine
FAMILY, but with multiple engine sizes per family.

As has already been noted, this doesn't work, because many aspects of
rocket engines don't scale straightforwardly.

The only good example I can recall of liquid fueled rocket engines
failing in catastrophic and dramatic fashion in a developed rocketry
program (i.e. after the initial development of orbital rocketry) would
be the N-1, and there are some fairly good reasons to take that as a
special case.

(Henry Spencer) wrote in message ...
In article ,
Mike Miller wrote:
I suspect the TSTO can glide back. Even at mach 8 separations, is the
first stage all that far down range, more than 100-200 miles?

For an orthodox trajectory, a Mach 8 separation will bring it down more
like 300mi from the launch site. That tends to require powered return.
And that's not a particularly high separation speed.

It'll
have a lot of energy from altitude and speed that can be spent turning
around and gliding back.

Not unless it has truly excellent aerodynamic performance, which is
another nasty can of worms. It's not all that high up by the time it
can get turned around.

That always seems like the sticky wicket for RLV TSTOs. There are
ways to go about solving the problem but none of them are simple. One
idea I kinda like is a hopping sub-orbital first stage. Build a
nearly orbital single stage RLV (but enough less than orbital to
realize substantial savings design and operations wise), set up
several launch/landing sites around the world and then hop the
sub-orbital vehicle from one to another for each launch. For example,
a ring of 2, 3 or more spaceports along the equator. Since you are
(or should be) running a high flight rate launch service with these
vehicles there isn't much downside to not having them all piled up at
one centralized launch complex. The other idea I like is a simple
drop tank on a quasi-SSTO. Though with orbital rocketry there's very
little that's ever simple.

(Henry Spencer) wrote in message ...
In article ,
Mike Miller wrote:
I suspect the TSTO can glide back. Even at mach 8 separations, is the
first stage all that far down range, more than 100-200 miles?

For an orthodox trajectory, a Mach 8 separation will bring it down more
like 300mi from the launch site. That tends to require powered return.
And that's not a particularly high separation speed.

It'll
have a lot of energy from altitude and speed that can be spent turning
around and gliding back.

Not unless it has truly excellent aerodynamic performance, which is
another nasty can of worms. It's not all that high up by the time it
can get turned around.

That always seems like the sticky wicket for RLV TSTOs. There are
ways to go about solving the problem but none of them are simple. One
idea I kinda like is a hopping sub-orbital first stage. Build a
nearly orbital single stage RLV (but enough less than orbital to
realize substantial savings design and operations wise), set up
several launch/landing sites around the world and then hop the
sub-orbital vehicle from one to another for each launch. For example,
a ring of 2, 3 or more spaceports along the equator. Since you are
(or should be) running a high flight rate launch service with these
vehicles there isn't much downside to not having them all piled up at
one centralized launch complex. The other idea I like is a simple
drop tank on a quasi-SSTO. Though with orbital rocketry there's very
little that's ever simple.