WOW! You guy's are really kicking in the goods here! Thanks! Henry
OK! I think it's time to do a reassess and review of the work done so
far. And where we are at with the project.
I'll get that out tonight.
-------- Original Message --------
Subject: [AR] Fwd: Electric centrifugal pumps
From: Henry Vanderbilt <hvanderbilt@xxxxxxxxxxxxxx>
Date: Wed, July 15, 2015 8:21 pm
To: arocket@xxxxxxxxxxxxx
This may be the Barske spreadsheet you mentioned...
-------- Forwarded Message --------
Subject: [AR] Electric centrifugal pumps
Date: Tue, 16 Nov 2010 22:00:13 +1100
From: Andrew Burns <burns.andrew@xxxxxxxxx>
To: Arocket List <arocket@xxxxxxxxxxxxxx>
Hey all, there's been a bit of discussion about electric motor powered
centrifugal pumps recently and I've been interested in designing and
making one for a little while. I think I remember reading somewhere that
Paul Breed had done some investigation into them, is there anywhere I
can read up on this?
I read up a book about centrifugal pumps and talked to some people I
know who are pretty experienced with them and I set out to design a
little test-bed pump within my manufacturing capabilities. I set the
mass flow rates and pressures to something small enough to be of use to
me in the short term (0.3kg/sec at 800 psi output, I'm not looking to
make an orbital rocket yet :p ). This sets the specific speed at 417
which is quite low. Due to the low specific speed I was told to have a
look at Barske impellers. A Barske impeller is a very simple design
which shows good performance at very low specific speeds when compared
to more conventional impellers. The real lure of the Barske design to me
is the simplicity of machining, the impeller is un-shrouded with
straight radial blades and the volute is a concentric circle 'bowl'
design with a tangentially mounted conical diffuser. As such one
wouldn't need a 5 axis mill or even a CNC mill to make the parts and the
literature I've read leads me to believe that efficiencies of 40 to 50%
should be attainable which is actually quite competitive at such low
specific speeds.
At my stated design conditions the fluid power output from the pump ends
up being around 1.5kW so for a pump efficiency of 40% and with some seal
losses you'll be up for a little under 4kW of motor power. Looking at
modern brushless motors and lithium polymer batteries a drive system for
such a pump wouldn't need to weigh much over 2kg, with enough power for
at least 20 seconds of pumping (limited by battery and motor power
output limits, not battery storage). Better yet the motor manufacturer
I'm basing my numbers on can make motors with shafts out both ends so in
a biprop system an impeller could be attached to each end of the motor,
and they're rated at a maximum speed of 60k RPM which is up where you
need to be with such small pumps.
The biggest sticking point I've encountered so far is the dynamic seals
for the pump. Without access to special ceramics or metals and precision
grinding/surface finishing I'm not sure how good a seal I could make and
although I'm really only interested in the seals lasting minutes, not
years the fluids being sealed against tend to be particularly nasty both
in terms of leaks and reactivity to hot/high friction environments. I've
considered using an inert gas to pressurise the stuffing box behind the
seal so that there would never be any fluid leakage out of the pump but
that's not a simple solution, especially if you want to keep the number
of seals down but also circulate fluid around the motor for cooling.
Does anybody on these boards have any experience or ideas when it comes
to home-brewing centrifugal pump dynamic seals?
I put together a quick spreadsheet for my designing, keep in mind that
there are a number of values in there pulled from design charts in books
and the equations are mostly also empirical and particular to the Barske
impeller design but I thought it might be useful to somebody.
Andrew
