Paper on Turbopump Alternative

Since the subject of turbopump alternatives comes up fairly regularly
the following recent (and interesting) paper might be of interest:

Knight, Andrew, "Designing and Testing a Lighter, Simpler, Less-
Expensive Liquid Propellant Pump", Journal of Propulsion and Power,
Vol. 20, No. 1, Jan-Feb, 2004.

The pump he describes seems remarkably similar, in appearance at
least, to the Comprex supercharger.

The author, Andrew Knight, identifies himself as the proprietor of
Andrew Knight Space Propulsion. A Google search didn't turn up any
additional information.

If you can't find a copy of the paper you can look up his patent #
6,499,288 online.

Does anyone have any additional information or opinions about Mr.
Knight and his pump?

Jim Davis

Jim Davis wrote in message .1.4...
Since the subject of turbopump alternatives comes up fairly regularly
the following recent (and interesting) paper might be of interest:

Knight, Andrew, "Designing and Testing a Lighter, Simpler, Less-
Expensive Liquid Propellant Pump", Journal of Propulsion and Power,
Vol. 20, No. 1, Jan-Feb, 2004.

The pump he describes seems remarkably similar, in appearance at
least, to the Comprex supercharger.

I have not read the paper. My university does not receive that
journal. After reading the patent i'm not going to bother with
inter-loans etc.

I seems like a bad idea to me IMO. It is just a version of the
piston-less pump really. He has just replaced the valves with a
rotating slide thing..... ok so i'm not good at explaining. IMO slide
seals are not easier than valves.. well valves have sliding seals too.
I see no advantage in the design over piston-less pumps.

While on the subject. Positive displacement pumps don't have the NPSH
problems that dynamic pumps have. Pistons have advantages over
piston-less. But what about other designs. Like a screw pump with a
screw expander? That way you get the gas to do some work while
expanding, rather than "workless" free expansion...

Greg

In article ,
Greg wrote:
While on the subject. Positive displacement pumps don't have the NPSH
problems that dynamic pumps have.

The problem isn't entirely absent, although in a more subtle way. The
delivery stroke of the pump is all push, no pull, so cavitation isn't an
issue there. But the refill stroke is all pull, so cavitation *is*
possible then, depending on flow resistance etc., and it could cause
various kinds of trouble.

...But what about other designs. Like a screw pump with a
screw expander? That way you get the gas to do some work while
expanding, rather than "workless" free expansion...

Please elaborate; I don't believe I've heard of screw pumps, and I can't
picture what you're envisioning.
--
MOST launched 30 June; science observations running | Henry Spencer
since Oct; first surprises seen; papers pending. |

(Greg) wrote in message m...
Jim Davis wrote in message .1.4...
Since the subject of turbopump alternatives comes up fairly regularly
the following recent (and interesting) paper might be of interest:

Knight, Andrew, "Designing and Testing a Lighter, Simpler, Less-
Expensive Liquid Propellant Pump", Journal of Propulsion and Power,
Vol. 20, No. 1, Jan-Feb, 2004.

The pump he describes seems remarkably similar, in appearance at
least, to the Comprex supercharger.

I have not read the paper. My university does not receive that
journal. After reading the patent i'm not going to bother with
inter-loans etc.

I seems like a bad idea to me IMO. It is just a version of the
piston-less pump really. He has just replaced the valves with a
rotating slide thing..... ok so i'm not good at explaining. IMO slide
seals are not easier than valves.. well valves have sliding seals too.
I see no advantage in the design over piston-less pumps.

The slide thingies are solenoid valves. This design
allows the valves the ability to withstand high
pressures and high flow rates without injury, since
there is no mechanical connection from outside the
valves to inside.

Greg wrote:

Jim Davis wrote in message news:
...
Since the subject of turbopump alternatives comes up fairly regularly
the following recent (and interesting) paper might be of interest:

Knight, Andrew, "Designing and Testing a Lighter, Simpler, Less-
Expensive Liquid Propellant Pump", Journal of Propulsion and Power,
Vol. 20, No. 1, Jan-Feb, 2004.

The pump he describes seems remarkably similar, in appearance at
least, to the Comprex supercharger.

I have not read the paper. My university does not receive that
journal. After reading the patent i'm not going to bother with
inter-loans etc.

I seems like a bad idea to me IMO. It is just a version of the
piston-less pump really. He has just replaced the valves with a
rotating slide thing..... ok so i'm not good at explaining. IMO slide
seals are not easier than valves.. well valves have sliding seals too.
I see no advantage in the design over piston-less pumps.

While on the subject. Positive displacement pumps don't have the NPSH
problems that dynamic pumps have. Pistons have advantages over
piston-less. But what about other designs. Like a screw pump with a
screw expander? That way you get the gas to do some work while
expanding, rather than "workless" free expansion...

Not sure what you mean by workless free expansion, but...
We did look a bit at rotary positive displacement pumps (and rotary
positive-displacement motors) for Mockingbird. They tended to be
heavier than piston pumps, and there's less integration of the drive and
pump sections, but they weren't out of the question; we just didn't have
the time and resources to look at them in detail. One problem with
rotary expanders is that they tend to be hard to seal; they need tight
tolerances which are difficult to maintain over a wide temperature
range.

Jordin Kare

My last post was not as clear as it could have been--sorry.

The context *I* was thinking of is for smaller (ie mockingbird)
engines and systems. For bigger systems I do not believe that
turbo-pumps will be beaten. High speed turbo-machinery can also be
made reliable especially if you leave out the LH2...

first Henry's post.....

The problem isn't entirely absent, although in a more subtle way.

Yes. I meant that NPSH is *less* of a problem than high speed
turbo-pumps where inducers/boost pumps are needed.

Please elaborate; I don't believe I've heard of screw pumps, and I
can't picture what you're envisioning.

Screw pumps, Screw expanders, gerotors(inner gear pumps) and gear
pumps, are all very similar. Screw compressors now make up the
majority of all commercial air compressors. A good intro with nice
pics for screw expanders/compressors:

www.staff.city.ac.uk/~ra601/purduco2.pdf

A company trying to improve upon the reciprocation engine:
www.starrotor.com/

For even pressure delivery from a piston pump, the driving gas must be
at constant pressure, it is then expelled to ambient pressure. But
this is not as efficient as getting the gas to do work by expanding.
We usually start with say 5000-10000psi and regulate that down to
300-1000psi, again without the gas doing useful work in the expansion
process.

A note about stored gas with a expander, power available could be
improved with some preheat. Without preheat there could be some
cooling problems (ie the gas gets to cold) but i haven't run the
numbers.

Clearly turbo-pumps solve this problem and provide very good power
density. But as has been noted before, piston pumps can be competitive
with small engines.

The question is: can using these rotary expanders/pumps provide a
advantage over piston pumps? I don't know, but i think that it is
worth considering and would be useful.

Jordain's post...

Not sure what you mean by workless free expansion, but...We did look
a bit at rotary positive displacement pumps (and rotary
positive-displacement motors) for Mockingbird.

Free expansion ie not doing useful work...Like the pressure drop in a
regulator. The clearance issue *is* an issue. There is definitely
plenty of development effort involved.

I have tried many times to find *detailed* info on mockingbird but
have failed. Is there any detailed info anywhere?

Greg

Jordin Kare wrote

Greg wrote:

I seems like a bad idea to me IMO. It is just a version of the
piston-less pump really. He has just replaced the valves with a
rotating slide thing..... ok so i'm not good at explaining. IMO slide
seals are not easier than valves.. well valves have sliding seals too.
I see no advantage in the design over piston-less pumps.

Nor me. Neither do I see anything patentable, but I won't argue that.

One thing that might be done would be to introduce a very light piston, to
thermally seperate the drive gas from the pumped fuel/propellant. A disk of
foamed plastic, or the high-tech equivalent.

While on the subject. Positive displacement pumps don't have the NPSH
problems that dynamic pumps have. Pistons have advantages over
piston-less. But what about other designs. Like a screw pump with a
screw expander? That way you get the gas to do some work while
expanding, rather than "workless" free expansion...

Not sure what you mean by workless free expansion, but...

The gas is stored at 6,000 psi and gets expanded to 600 psi before feeding
the cylinder - the expansion from 6,000 psi to 600 psi does no work.

At least that's what I'd guess he means. There are other possibilities, for
example the later expansion from 600 psi to atmospheric does no work either
... Energetically this is not a very efficient pump.

We did look a bit at rotary positive displacement pumps (and rotary
positive-displacement motors) for Mockingbird. They tended to be
heavier than piston pumps, and there's less integration of the drive and
pump sections, but they weren't out of the question; we just didn't have
the time and resources to look at them in detail. One problem with
rotary expanders is that they tend to be hard to seal; they need tight
tolerances which are difficult to maintain over a wide temperature
range.

Quimby expanders are good for tolerance requirements, as the pressure
difference is split between several seals.

If the shafts aren't parallel then more work can be extracted as well.
Another way is to vary the depth, groove/land ratio, or the pitch.*


--
Peter Fairbrother


* Reminds me a bit of triple-expansion steam engines - as a little kid I
once sat on the end of the pistonshaft/driveshaft coupling thingy and
travelled in 24-feet circles. Be totally illegal nowadays, from safety pov,
but it was great fun.

(On the "Waverly", if anyone wants to know. She was (still is) a sea-going
paddle steamer with magnificent triple expansion engines, she used to ply
the Scottish island routes, but her engine room is covered in plastic
windows now, the new engineer isn't as amenable to a friendly dram as the
old Chief was, and her Captain isn't a friend of my Grandpa anymore)

pps anyone know what happened to the erstwhile engineers's daughter? She was
a cutie. Ahh, memories.


*all independant variables, even with non-parallel shafts.

Henry Spencer wrote:

In article ,
Greg wrote:
While on the subject. Positive displacement pumps don't have the NPSH
problems that dynamic pumps have.

The problem isn't entirely absent, although in a more subtle way. The
delivery stroke of the pump is all push, no pull, so cavitation isn't an
issue there. But the refill stroke is all pull, so cavitation *is*
possible then, depending on flow resistance etc., and it could cause
various kinds of trouble.

...But what about other designs. Like a screw pump with a
screw expander? That way you get the gas to do some work while
expanding, rather than "workless" free expansion...

Please elaborate; I don't believe I've heard of screw pumps, and I can't
picture what you're envisioning.

They're usually known as "rotary screw compressors" and they're commonly
used for large, continuous duty air compressors. They basically work
like an Archimedes screw for gases: two helical shapes mesh together and
rotate inside an oval-cross-section housing. Each turn of the helix,
combined with the housing, forms a more or less gas-tight volume that
moves from the inlet end to the outlet end as the helices turn; the
actual compression occurs as each turn's volume is "squeezed out" at the
output end (although I think there are tapered versions that do the
compression continuously along the length).

The actual geometries are kind of interesting; typically one shape has
convex lobes (cross section like a child's drawing of a flower --
semicircles around the perimeter of a circular core) and the other has
concave ones -- but not the same number! one helix generally has 4
lobes and the other 6!)

There's a cutaway picture at http://air.irco.com/asg/rotary/index.asp

They're nice for big compressors for several reasons: they have a lot of
surface area for cooling the gas as it's compressed. They use
clearance seals, so they don't have seal wear problems, and they can be
directly driven from a motor shaft. They can be designed to give a
fairly steady flow, reducing the need for accumulators. And they have
the high pressure ratio capability of all positive-displacement pumps
(The same basic geometry is used in injection-molding machines at very
high pressures).

The bad news is they tend to be heavy because of the odd-shaped surfaces
that have to take pressure and maintain clearance seals -- the helices
themselves would need to be hollow for a flight-weight pump. They leak.
They're also lousy as motors, and they're not much if any easier to make
than turbopumps -- unlike piston pumps, which basically require a lathe
and not much more. The long life is (nearly) irrelevant for a rocket
pump.

Jordin

Greg wrote:


I have tried many times to find *detailed* info on mockingbird but
have failed. Is there any detailed info anywhere?

Greg

Alas, not on line. I keep meaning to scan in the one presentation that
was formally released (which actually had most of the interesting
information in it; we didn't get all that far into the design) and put
it up (or let some better-equipped site host it), but I never seem to
have enough time. Anyone in the US who'd like to do the job, let me
know in email.

Jordin Kare

From what I can tell, the gas is doing work in going from 6000 to 600
psi. The only significant losses are seen in the
turbulence/shear/conversion to heat and noise occurring in the
expansion valve. Envision this:

A layer of gas is at 600 PSI in the filled pump chamber on top of the
propellant. The propellant is at the required pressure, but is not
moving to the engine. More gas at initially 6000psi is admitted to
the chamber, and this difference in pressure forces propellant to move
into the engine. The force exerted on the top of the propellant is
moved by the newly admitted gas through the distance required to
equilibrate the pressure of the newly admitted gas to 600psi, when
motion and expansion stop.

This motion is of course continuous, not intermittent, and the mean
pressure of the gas during an expulsion phase is above 600psi, enough
to overcome resistance losses and maintain a high flow rate.

Where power can be recovered is when the expanded volume of slightly
over 600psi gas is released to ambient. This should in total be a
fairly continuous, cold/cool gas flow which could run a turbine to
pressurize hydraulics, generate electricity, and provide directional
control.

I am of the opinion there is no reason this could not be cheaply
scaled to a medium lift booster and higher stages, and that it would
be advantageous as size increases to use cryogenic nitrogen which is
not pressurized when cold to be the propellant pressurant, providing
propellants compatible with the short term of contact are used.

Thanks, TDP