MEMs rockets and jets - a fabulous opportunity

Micro-electro-mechanical systems use techniques developed over that
past 50 years making ever tineir integrated circuits, to make ever
tinier mechanical devices. This is not nano-technology, its thousands
of times bigger. Yet, MEMs devices are thousands of times smaller
than traditional macroscopic devices.

This could be very important for rocket engines. Why? Because of
favorable scaling laws.

That's because rocket thrust is a function of nozzle area while rocket
weight is a function of part volumes. So, when you reduce the size of
a rocket by half say, the area is 1/4 and the volume is 1/8 th, so,
thrust versus weight is doubled.

So, imagine taking something like the RL-10 rocket engine which has a
nozzle diameter of about 1 meter - and reducing it to 0.25 mm - that's
about the size of a pixel cavity on a HDTV plasma screen (smaller than
can be resolved by the human eye) 0.25 mm is 1/4000th the size of a
1,000mm diameter RL-10. Its area is 1/16millionths and its volume is
1/64billionth - so, the thrust to weight of these engines if 4,000x
greater than a full-scale RL-10. So, if a full scale engine has an 80
to 1 thrust to weight ratio, the sub-scale engine has 320,000 to 1
thrust to weight ratio!!!

What can you do with tiny thrusts?

Built large numbers of engines and operate them in large arrays!

In fact, just like you can paint pictures in colorful gases by sending
signals through an array of pixels on an HDTV plasma screen, you can
paint thrust vectors across a surface made up of millions upon
millions of tiny rocket engines. This forms what I call a 'propulsive
skin'.

In fact, just as you can have three primary colors form any color in
the rainbow, and create an image plane that is made up of three
different sets of independent pixels - so too can you can form any
thrust vector relative to a surface by a triad orthogonal engines at
each point on a surface. Similarly the same signal processing and
control mechanisms that make HDTV plasma screens possible, are
directly adapted to a propulsive skin that cna produce any thrust
vector on its surface. Furthermore, unlike an imaging device that
must be flat, a propulsive skin can be shaped to achieve specific
efficiencies for a given mission or flight cycle.

There are other advantages and possibilities with this sort of engine
array

First is supreme safety. While the explosion of a 15,000 kgf thrust
engine that is 1 meter in diameter can make a bad day, the explosion
of a 1 gram thrust engine 0.25 mm in diameter is hardly noticeable.
While operating 100 million engines that cannot be serviced by any
means available today in a propulsive skin may mean a virtual
certainty that some engines will fail on each mission, those failures
need not be life threatening or even bring the vehicle to harm.

Second supreme reliability. Again, 100 million engines operated on
each flight cycle may see 100 engines fail in one way or another every
time the system is used. Howeve,r such failures are harly noticable,
and cause a performance reduction of less than 1 part per million.
10,000 flight cycles - which is typical of most aircraft today over
their service lives - mean that 99.99% thrust is available at the end
of a vehicle's life after that many flight cycles. The odds that a
sufficient number of engines fail to cause any noticible degradation
in the mission is virtually impossible.

Third - ultrasonic noise. The word rocket derives from the German
word 'racket' which means noise. Rocket engines have a lot in common
with steam whisltes, where jets of gas enter the still air at high
speeds producing violent sound waves. The principal components of
this sound wave of an engine is determined in part by its size. This
means that 0.25 mm engine's primary noise is well above the level of
human hearing. Its smaller than a dog whistle, and equally
unhearable.

Fourth - controllable thrust. Just as gray scale on a HDTV plasma
screen may be implemented by quickly switching the pixels on and off -
so too may engine average thrust levels be similarly controlled.
While such operation is not recommended in larger engines, scaling
laws favor such changes, which if cleverly exploited, can produce
radically improved reliable and less costly designs.

Fifth - controllable sound - Switching engines on and off in the
audible range produces controllable sound sources acrosss the
propulsive surface. A sonic hologram that can literally blow you away
(or alternatively lift the holographic surface) creates a number of
interesting design opportunities.

Sixth - very low mass - low cost. Imagine lifting a payload of 1 kg
can carrying it ballistically 3,000 km. this means the vehicle must
achieve a delta vee of about 3 km/sec. Assuming an engine performance
of 420 sec Isp . This implies a propellant fraction of 52%. With a
structural fraction of 3% - this leaves 45% for payload. Thus the
entire vehicle masses 2.23 kg of which 1.16 kg is propellant and
0.07 kg is mostly low cost packaging. Designing the system for 5 kg
top thrust - which is more than adequate for this mission - and a
320,000 to 1 thrust to weight ratio - implies a propulsive skin
overlaying the packaging massing only 16 milligrams !! Aerospace
hardware ranges from $5 million to $25 million per ton. That's $5 to
$25 per gram.- that's $0.08 to $0.40 for the propulsive skin in this
case. Since a kilogram of most products costs less than this figure,
and since 95% of the cost of a product stems from the logistics of
delivering that product to market, propulsive skin has a monumental
potential to change the way products are handled on Earth.

* * * EXAMPLE * * *

Imagine workers in an apple orchard harvesting apples. Instead of
dragging behind them a large gunnysack of freshly picked apples, they
have a bag of propulsive skin propelled containers. Apples are loaded
into containers, and the containers fly off to a nearby fueling
station in the field and are programmed with a destination received by
internet or cell phone or pda ordering - and fly off immediately to
their customer anywhere in a 28 million sq km area.

* * * ANALYSIS * * *

A kilo of freshly picked apples costs $1.50 the container costs
$0.08. At 1/5th cent per kWh (low cost solar panels see http://www.usoal.com)
the 1.16 kg of propellant costs $0.02 made from 1.74 liters of water.
Packaging is another $0.05. The order processing software that
programs the propulsive skin costs another $0.02 - a total of $0.20 to
deliver $1.50 worth of apples in seconds. Empties only require 5% of
the propellant to return them and propellant costs less than the
package in this analysis, so it likely that packages would be
recycled. It also gets rid of a disposal problem. Customers may
receive a $0.02 credit if their packages come back in good shape.
Alternatively, they may be charged an additoinal $0.20 if they are
not.


* * * EXAMPLE * * *

A restaurant fulfills takeaway orders. Their average food mass per
customer is less than 1 kg. Their average check size is $9.00 - they
deliver orders to anyone within a 3,000 km radius for $1.20 surcharge
and they get the 'packaging' back - with delivery in minutes.

* * * EXAMPLE * * *

A disposable (100 uses) flying cylinder carries a passenger up to
3,000 km for less than $20 in less than 10 minutes. The passenger
steps into the vehicle as they might an old-style telephone booth.
The vehicle is programmed via the customer's pda or cell phone. No
controls are present inside the vehicle - which has a low density foam
interior and a transparent ablative aluminum coated PET skin. The
vehicle takes off vertically, and as it ascends, it reorients itself
in flight to put the passenger in a reclining position, during ascent
and descent at high-gees. As gee forces lower, and soft landing is
approached, the vehicle resumes its vertical 'upright' stance and the
customer then exits the vehicle. The vehicle goes to a nearby
refueling station and thence to another caller nearby. After 100 uses
(3 flight hours) the vehicle flies to a disposal center to be
recycled.

NOTE:

MEMs based pumps heat exchangers compressors and fans and other items
used in life support and fuel handling and cryogenic refrigaration
also benefit being reduced to tiny sizes as well enhancing their
performance.


..

On May 7, 10:45*am, Williamknowsbest wrote:
Micro-electro-mechanical systems use techniques developed over that
past 50 years making ever tineir integrated circuits, to make ever
tinier mechanical devices. *This is not nano-technology, its thousands
of times bigger. *Yet, MEMs devices are thousands of times smaller
than traditional macroscopic devices.

snip
.

NOTE:

MEMs based pumps heat exchangers compressors and fans and other items
used in life support and fuel handling and cryogenic refrigaration
also benefit being reduced to tiny sizes as well enhancing their
performance.

.

Here's an intresting site that has pictures of mcro-rockets

http://www.me.berkeley.edu/mrcl/rockets.html

Here's a good review article (sorry about the ad)

http://www.technologyreview.com/Infotech/12415/?a=f

Here's a major aerospace company's efforts

http://www.st.northropgrumman.com/ca...ropulsion.html

Here is an interesting development in small rotary engines

http://www.me.berkeley.edu/mrcl/

Of course, small rotary engines can also be small rotary pumps. or
small cryogenic refrigerators

http://www.grc.nasa.gov/WWW/RT/RT200...5440moran.html

http://www.cstl.nist.gov/projects/fy...5radebaugh.pdf

Heat transfer operates by diffusive laws described by Fourier a while
back. These laws tend to favor large surface areas communicating heat
through thin volumes. Well, these sorts of relations scale well for
tiny arrays of refrigerators. So, arrays of cryocoolers are ideal for
managing the thermal environment of cryogenic fluids.

One application is the management of the thermal environment of liquid
cryogenic fuels. One of the reasons hypergolics were used in Apollo
was their storeability and of course their ease of restart.
Hypergolics are about 1/2 as efficient as liquid oxygen and liquid
hydrogen. Reliability isn't a real problem with large engine arrays
as already described. Storeability is the only remaining problem,for
cryogens. However,with MEMs based cryocooler arrays built into the
tanks themselves, and powered by the ullage gases in hydrogen oxygen
fuel cells, provide a means to efficiently store cryogens reliably for
long periods of time.

The difference between cryogenic propellants and hypergolic
propellants is obvious. It takes a delta vee land on the Lunar surface
from lunar orbit and return

LANDING STAGE
Propellant = N2O4 Aerozine 50
Vf = 3.05 km/sec
Mp=8,165 kg
Mt=14,596 kg
Vf=2.50 km/sec

ASCENT STAGE
Propellant = N2O4/Aerozine 50
Vf=3.05 km/sec
Mp=2,363 kg
Mt=4,547 kg
Vf = 2.23 km/sec


Using Cryogens

LANDING STAGE
Propellant = LOX/LH2
Vf = 4.5 km/sec
Mp=6,221 kg
Mt=14,596 kg
Vf=2.50 km/sec

ASCENT STAGE
Propellant = LOX/LH2
Vf=4.5 km/sec
Mp=2,536 kg
Mt=6,490 kg
Vf = 2.23 km/sec

The empty ascent stage in the hypergolic case is 2,194 kg. The empty
ascent stage in the cryogenic case is 3,954 - nearly twice the payload
to the moon and back.

With imrpvoed structural fractions possible today its quite feasable
to consider a single stage lander/ascent vehicle that carries its own
light weight thermal protection and is fully reusable!

Another advantage is that any savings on landing fuel (note there was
a surplus) is converted to power and water in fuel cells.

There are other uses of arrays of tiny cryogenic refrigerators.

Cooling air to remove liquid oxygen for use in a rocket array - thus
obtaining a combined cycle rocket & jet array.

LOX has a boiling point of 90.2K and a freezing point of 50.5K.
Liquid nitrogen has a boiling point of 77K and a freezing point of
63K.

Room temperature on the Kelvin scale is 300K.

A cryocooler operating at 3.4 bar compressing air to less than 1/3 its
volume and letting it cool to room temperature, then reducing the
pressure of the air to atmospheric by letting the roomtemperature gas
expand to its original volume, will liquify the oxygen. This can be
done rather quickly if done with a large array of cryocoolers. At 90K
the nitrogen is still gaseous, while the oxygen is liquid. The gas
portion,which is 70% of the total,is exhausted in a way that it cools
the incoming gas,so that energy is not wasted. The LOX is fed to an
engine array that burns some sort of fuel. Either liquid hydrogen, or
jet fuel. Hydrogen and oxygen used in an array of tiny fuel cells
(which also has scalaing advantages since current goes up with
electrode area) - power the refrigerators.

A jet is possible by noting that many cryocoolers uses compression
waves to cool a gas and also noting that MEMs rocket engines operating
intermittently produces compression waves - a multi-layer system that
transmits vibrations may directly link the compressor portion to the
thrust portion of the 'jet'.

That is, separating the gases nitrogen and oxygen from air,and
liquifying the oxygen, compresses the oxygen dramatically. It is
likely that arrays of jets such as these could be 3x more efficient
than traditional jets since these compress only the oxygen. NOx would
not be an issue either. Finally,by controlling the jet of hot nitrogen
- one could create an auxiliary thruster to the oxygen/fuel driven
engine array - which would be very similar to a high bypass engine in
many respects - lower noise, higher thrust higher effective specific
fuel efficiency.

Since only enough oxygen need be removed from the air to burn the fuel
efficiently, the auxiliary jets would also include a measure of
oxygen.

I am experimenting with what I call mini-rockets, not micro-rockets.
They are made out of metal foils (primarily titanium) that are shaped
with multiple punches chemical etching folded and welded together and
stamped and etched some more!

http://en.wikipedia.org/wiki/Progressive_stamping_die
http://www.composidie.com/about/video-vault.html
http://en.wikipedia.org/wiki/Chemical_etching

1800 engines per minute each 5 mm in diameter producing 375 grams of
thrust and massing 24 milligrams each. 2.6 mega-engines per day from
one tool setup made from a titanium foil.

Engine array plates or 'lift rods' of 2,500 engines, forming a surface
125 cm by 5 cm -10x250 engines- produces a maximum thrust of 937.5 kgf
massing only 60 grams.

1,000 plates every 24 hours from a single press once in production.

Compressors that use air for oxidizer, and use jet fuel for propellant
produce an effective specific impulse of 4,000 seconds using a high
bypass engine design. High bypass is needed for cooling and noise
reduction at this scale.

With 250 kg of fuel, 20 minute flight times are possible at full
thrust. One hour flight times carrying 100 kilos payload and 250
kilos of fuel - 38.7 gallons - using engine thrust for lift only in
vertical flight.

http://en.wikipedia.org/wiki/Bell_Rocket_Belt
http://en.wikipedia.org/wiki/Jet_pack

Lift rods mounted inside airfoils with slots near the leading edge for
air intake and slots near the trailing edge for exhaust have reduced
noise level improved lift to drag in forward flight allowing nearly 10
hours of high speed flight.

The system consists of a wing with a 3 meter span, the central 1.25
meter is equipped with a 'lift stick' buried in the root of the wing.
A tear drop canopy and body 2.2 meters long and 0.7 meters wide and
0.7 meters deep admit a prone pilot. The wing rotates and folds back
along the aircraft body to enter and exit VTOL flight as engine thrust
increases as the wings fold back along the length of the fuselage.
The lightweight construction masses less than 50 kg. Each wing tip
contains up to 20 gallons of fuel.

On Wed, 7 May 2008 07:45:51 -0700 (PDT), Williamknowsbest
wrote:

Micro-electro-mechanical systems use techniques developed over that
past 50 years making ever tineir integrated circuits, to make ever
tinier mechanical devices. This is not nano-technology, its thousands
of times bigger. Yet, MEMs devices are thousands of times smaller
than traditional macroscopic devices.

This could be very important for rocket engines. Why? Because of
favorable scaling laws.

That's because rocket thrust is a function of nozzle area while rocket
weight is a function of part volumes. So, when you reduce the size of
a rocket by half say, the area is 1/4 and the volume is 1/8 th, so,
thrust versus weight is doubled.

So, imagine taking something like the RL-10 rocket engine which has a
nozzle diameter of about 1 meter - and reducing it to 0.25 mm - that's
about the size of a pixel cavity on a HDTV plasma screen (smaller than
can be resolved by the human eye) 0.25 mm is 1/4000th the size of a
1,000mm diameter RL-10. Its area is 1/16millionths and its volume is
1/64billionth - so, the thrust to weight of these engines if 4,000x
greater than a full-scale RL-10. So, if a full scale engine has an 80
to 1 thrust to weight ratio, the sub-scale engine has 320,000 to 1
thrust to weight ratio!!!

Except that the micro-RL-10's fuel pump will instantaneously self-destruct
due to bearing friction, the rest of the engine will fail due to thermal
shock a short time thereafter, the residence time of the propellants will
be inadequate for complete mixing and combustion, and most of what thrust
you do get will be lost to viscous and free-molecular flow effects in the
micro-nozzle.

NASA/JPL's Juergen Mueller, who has put a whole lot more thought into this
than you have, estimates the thrust:weight ratio of an actual micro-rocket
engineered to deal with all the various adverse scaling effects, at "only"
2000:1. The Air Force Research Laboratory's Dr. Andrew Ketsdever, one of
the more optimistic proponents of microthrusters, notes that the specific
impulse is likely to be reduced by 45-55% from the large-engine value[1].


There are some applications where this sort of inefficiency is tolerable,
but generally, microthrusters are for microspacecraft. If you're going
to try and make bignum microthrusters work together to produce lots of
thrust, you'll almost certainly find that a single larger rocket would
work better. And if it seems like a stupid idea to use a large rocket
engine, that's probably because you're thinking of a problem where
rockets aren't the solution regardless of the scale. Rockets are for
missiles and spacecraft, and not much else. Waving a magic wand labeled
"Micro!", or even "Nano!!!", doesn't much change that.


[1] _Micropropulsion for Small Spacecraft_, Micci & Ketsdever, editors,
AIAA Progress in Aeronautics and Astronautics series vol. 187, 2000.
Chapters 3 and 4 for the Mueller and Ketsdever work, respectively.


--
*John Schilling * "Anything worth doing, *
*Member:AIAA,NRA,ACLU,SAS,LP * is worth doing for money" *
*Chief Scientist & General Partner * -13th Rule of Acquisition *
*White Elephant Research, LLC * "There is no substitute *
* for success" *
*661-718-0955 or 661-275-6795 * -58th Rule of Acquisition *

On May 10, 11:50*am, John Schilling wrote:
On Wed, 7 May 2008 07:45:51 -0700 (PDT), Williamknowsbest





wrote:
Micro-electro-mechanical systems use techniques developed over that
past 50 years making ever tineir integrated circuits, to make ever
tinier mechanical devices. *This is not nano-technology, its thousands
of times bigger. *Yet, MEMs devices are thousands of times smaller
than traditional macroscopic devices.
This could be very important for rocket engines. *Why? * Because of
favorable scaling laws.
That's because rocket thrust is a function of nozzle area while rocket
weight is a function of part volumes. *So, when you reduce the size of
a rocket by half say, the area is 1/4 and the volume is 1/8 th, so,
thrust versus weight is doubled.
So, imagine taking something like the RL-10 rocket engine which has a
nozzle diameter of about 1 meter - and reducing it to 0.25 mm - that's
about the size of a pixel cavity on a HDTV plasma screen (smaller than
can be resolved by the human eye) * *0.25 mm is 1/4000th the size of a
1,000mm diameter RL-10. *Its area is 1/16millionths and its volume is
1/64billionth - so, the thrust to weight of these engines if 4,000x
greater than a full-scale RL-10. *So, if a full scale engine has an 80
to 1 thrust to weight ratio, the sub-scale engine has 320,000 to 1
thrust to weight ratio!!!

Except that the micro-RL-10's fuel pump will instantaneously self-destruct
due to bearing friction, the rest of the engine will fail due to thermal
shock a short time thereafter, the residence time of the propellants will
be inadequate for complete mixing and combustion, and most of what thrust
you do get will be lost to viscous and free-molecular flow effects in the
micro-nozzle.

NASA/JPL's Juergen Mueller, who has put a whole lot more thought into this
than you have, estimates the thrust:weight ratio of an actual micro-rocket
engineered to deal with all the various adverse scaling effects, at "only"
2000:1. *The Air Force Research Laboratory's Dr. Andrew Ketsdever, one of
the more optimistic proponents of microthrusters, notes that the specific
impulse is likely to be reduced by 45-55% from the large-engine value[1].

There are some applications where this sort of inefficiency is tolerable,
but generally, microthrusters are for microspacecraft. *If you're going
to try and make bignum microthrusters work together to produce lots of
thrust, you'll almost certainly find that a single larger rocket would
work better. *And if it seems like a stupid idea to use a large rocket
engine, that's probably because you're thinking of a problem where
rockets aren't the solution regardless of the scale. *Rockets are for
missiles and spacecraft, and not much else. *Waving a magic wand labeled
"Micro!", or even "Nano!!!", doesn't much change that.

[1] _Micropropulsion for Small Spacecraft_, Micci & Ketsdever, editors,
* * AIAA Progress in Aeronautics and Astronautics series vol. 187, 2000.

Google will let me post, but when I try to pose a specific response to
your points, Google will not let me post. Is the URLs?

I've attempted to respond to a critic's response to my MEMs articles
several times. While Google will gladly post test statements, I don't
seem to be capable of posting my lengthy reply with references and
whatnot. Now when I started by Word program to transfer my clipboard
there, I lost what I had copied to the clipboard. So, I must retype
everything... hopefully it will take this time.

I've attempted to respond to a critic's response to my MEMs articles
several times. While Google will gladly post test statements, I don't
seem to be capable of posting my lengthy reply with references and
whatnot. Now when I started by Word program to transfer my clipboard
there, I lost what I had copied to the clipboard. So, I must retype
everything... hopefully it will take this time.

JS said that the rotor would fail, that's not the case;

High Rotor Speeds

I've attempted to respond to a critic's response to my MEMs articles
several times. While Google will gladly post test statements, I don't
seem to be capable of posting my lengthy reply with references and
whatnot. Now when I started by Word program to transfer my clipboard
there, I lost what I had copied to the clipboard. So, I must retype
everything... hopefully it will take this time.

JS said that the rotor would fail, that's not the case;

High Rotor Speeds
http://www.
greencarcongress
..com/2006/09/
memsbased_turbi_1.html

Okay - so it posted up to this point, so I've got to believe it has
something to do with the URL? and maybe JS name?

So, I've broken up the URL - just reassemble it to read it. Rotor
speeds are really high and they don't melt down as JS asserts.

JS said that the combustion wouldn't be efficient, again not the case

Microcombustion
http://
ronney.usc.edu/
publications/ISC31
SwissRollModel.pdf

JS said that thrust would be inadequate, not the case

MEMs rockets
http://www.me.berkeley.edu/
mrcl/rockets.html

JS said that viscous effects would degrade performance - while it is
an important design factor, it can be mitigated in careful design.
For example, increasing chamber pressure, or careful control of
surface chemistry

Heat Transfer and Viscous Effects
http://www.uvm.edu/
~vacc/pdf/AIAA%202007
-3987.pdf

Viscous Effects
http://www.engin.brown.edu/
Faculty/breuer/KSB%20
Papers/ASME98_Nozzles.pdf

Scaling of Viscous Effects
http://raphael.mit.edu/
Technical_Reports/
Bayt_Thesis.pdf

Surface chemistry and viscous effects
http://ame-www.usc.edu/
afrlmicrofluidics/Oral%20
Breuer.pdf

JS misquoted Dr. K stating that he said there would be a 50% specific
impulse drop in micro-thrusters. Dr. K actually said that careful
attention must be paid to details of microthruster design or one might
suffer such large losses.

JS said that rockets are only useful for rockets and spacecraft. He
did not say why. Perhaps it is their high cost. Perhaps it is their
inherent risk. Perhaps it is their complexity. i showed that a
320,000 to 1 thrust to weight ratio would produce propulsive devices
of immense capability at trivial costs which clearly have huge
potential effects. I also did an analysis of a spacecraft application
- to the lunar module. So, I am talking spacecraft as well as other
potential applications.

It should be up to the market to decide what use this technology has
once its further along. While 320,000 to 1 and 100% large engine
performance is agressive, its a reasonable target. Yet even at 2,000
to 1 and 50% engine performance might prove useful.

Consider a 1 kg lifting surface made of a microengine array that has a
2,000 to 1 thrust to weight. The surface itself masses 500
milligrams. At $5,300 per kg - typical of aerospace equipment - this
costs $2.65 - with a specific impulse of say 200 seconds - a
propellant fraction of 20% achieves 446 m/sec. Enough to toss 1 kg
across Manhattan. It costs $20 to $50 to have a package delivered
same day in Manhattan. One can imagine office supply stores selling
half pound packages with propulsive surfaces at office supply stores
10 for $100 - and them being used to send packages cross-town in
seconds.

So, even accepting all of JS' conclusions - of modest thrust to weight
and modest performance, we still have the potential of interesting
commercial uses.

This doesn't say that such uses are certain. This does say that its
premature to prejudge things based on faulty reasoning.