The story so far:
We have 2 impeller designs for the turbopump to test.
Lloyd Droppers came up with the original design and Peter Fairbrother
came up with a Barsky design.
I can machine the Barsky design.
Russel Blink said he would machine the Lloyd design. (which I have not
produced the CAD drawings for. There was some work to be done to
eliminate thrust in that design also)
Eric Gustafson and Anthony Cesaroni: Both rendered the impeller design
in 3D.
Peter Hokanson: Contributed some script files that can be used with
OpenFoam for CFD analysis.
Troy Prideaux and Andrew Burns: Had been working on a Barske design and
offered up their work.
I'll have to go back later and figure out everyone that offered details
or information that got me this far.
David Weinshenker just added some comments that made me simplify the
pump I'm building right now.
The pump I'm working on right now:
Mass Flow = 0.89 lb/s
DP = 870 psi
N = 50 kRPM
density 0.771
Impeller:
Diameter 38mm
6 blades. Blade width 2.2mm. Blade outer ends machined to 38mm diameter
(not rounded). Inner ends rounded off.
Center blade-free hole 12 mm diameter
Blade height 5.2mm at theoretical center, 3.4mm at edge.
Impeller back plane: 2.4mm thick. Cut away as much of the back plane as
practical, something like the attached drawing.
Volute:
Casing:
41.5 mm interior diameter, circular, concentric.
Diffuser:
Output hole should in theory be 2.8mm diameter - but try 3.0 - 3.2mm
actual. A rectangular hole is good too, about 3.2 tall and 2.2 wide.
Output taper 1 in 10 diameter, ie 1 in 3.16 area, at least 35 mm long.
Edges where taper meets casing interior should be sharp-ish.
Some design notes: The number of blades is set at six partly to make
machining easier, partly to strengthen the impeller, and partly because
in a smaller pump large numbers of blades/high solidity makes less
difference - TD's impellers are 12" diameter, TD's test conditions are
not really applicable to rocket pumps, and scaling laws don't work well
for small size pumps anyway - in fact it can be counterproductive for
small high-head pumps.
The main function of the "sparse" impeller back plane is to reduce end
thrust to low values, but it should also also increase head and
efficiency a bit.
I am still in two minds about the clearance between the 38mm dia blades
and the 41.5 mm casing... there are a lot of things to take into
consideration, but overall it's my best guess without actual
experimentation.
As for the blade inner ends, I don't really know whether they should be
rounded or not, or otherwise shaped, but I went with rounded.
One thing I did not explore was double-sided impellers.
-- Peter Fairbrother
I've changed that design to simplify it even more after reading the
original Barsky paper but it's essentially the same.
In order to run test we need a prime mover that can exceed the
requirements by some margin be reliable and rugged and as inexpensive as
possible.
Hence where we are now.
I acquired a TD03L-10T Turbocharger Centersection
Machined a pedestal mount (which contains the oil return line)
Machined a shroud (to put nozzles ect on the turbine)
Purchased an electric oil supply pump.
Made the oil tank.
To do:
Wind coil for rpm pickup
build spider to mount pickup coil
Drill nozzles into the shroud to provide for initial test.
Determine turbine flow requirements.
Get system up to speed and power to run pump test. (I have a 5.7L engine
I can convert to a compressor or I can use the exhaust from my current
5.7L truck.)
Once we have a proper test stand and safety precautions in place.
MACHINE THE PUMPS!
Once we get working pumps we design a turbine specific to rockets.
To make work easier and faster we will at some point want a balancer of
our own. For the time being because of the size of the pumps we intend
to test they can be balanced at a turbocharger shop.
I like rocket engines not rocket motors so I suppose the heart of our
rockets are the turbopump so that means we specialize in turbopumps.
I intend to do a lot of work on turbopumps over the next few years
obviously so this test equipment is the necessary start to an enduring
process.
-------- Original Message --------
Subject: [AR] Re: Turbopump prime mover
From: Ed LeBouthillier <codemonky@xxxxxxxxxxxxx>
Date: Thu, December 03, 2015 11:43 am
To: arocket@xxxxxxxxxxxxx, arocket@xxxxxxxxxxxxx
In the beginning there will be no load at all is all I'm saying.
Yeah, but there'll still be no horsepower. That's still fine. It's definitely
a beginning.
The turbo did not have impinging nozzles
The gap around the turbine is the nozzle (it contracts but does not
expand). The point is to produce choking to raise the velocity of
the gases to the speed of sound in that gas. The scroll directs the
mass tangentially to the turbine so that it impinges at a better
angle. If they don't have enough mass flow for choking, there is
still some acceleration of the gases. Ideally, though, there will
be choking.
I really don't know the math to do this obviously and I appreciate the
help.
What are you trying to do? I haven't been following this discussion lately.
What I seem to gather is that you are trying to produce a drive system able
to produce 30 HP at 50K RPM. Is that correct?
Personally, I think you're going to need something more like a jet combustion
chamber/gas generator to produce the pressure and mass flow rate you want.
But,
you can definitely start the thing spinning with compressed air.
Look at the jet engines that people have produced using turbo turbines
and you'll see the kind of gas generators that I think you'll need:
http://www.junkyardjet.com/
There are a LOT of variables and it is boggling and there is a lot of
information and maps for turbo compressors but the amount of data for
the turbine is far less, companies keep that data a little closer to the
vest.
Yes, I know the number of variables is boggling...I agree with you. There's
a lot to understand. I don't consider myself anywhere near an expert on the
subject, but I spent a few years studying the math of turbines and turbopumps.
At best, I'm a junkyard craftsman on the subject even though I've never built
one.
But some of the math that I reviewed the other day about mdot and Ve are the
basics. You can back out from the horsepower and RPM you want to the mdot and
Ve relationship you need. That'll dictate the kind of gas generation system
you
need (by gas generation system, I'm including compressed air tanks).
You can definitely work your way up from quarter-inch fittings to test
the lubrication system. The lubrication system is going to be crucial
to sustaining operation of the thing.
It is a learning process but we need a prime mover to run any
experiments on a real pump.
That's a good place to start.
Meanwhile if we bat around the details of the flow problem we can pretty
much solve that and we can all learn a bit more about turbines.
I'll try to put a few hours in the next week but I need to know the
particulars of the system you're planning on using. Here's my take
on what I need to know:
TURBINE FINAL TARGET PARAMETERS
-------------------------------
RPM: 50K RPM
Horsepower: 30 HP
Turbine Diameter: 2 x 0.834 inch = 1.668 inches
Gas Source: ? I think it'll have to be a combustion gas generator (or a car
engine)
Gas Pressure: 60 PSI sustained
Gas mdot: ? we'll have to work this out
Attached is a spreadsheet I use for turbine design. Hopefully
you'll find it interesting and educational. It won't apply
precisely to this application, but it'll be useful nonetheless.