Yes, I think he was referring more to the centrifugal pump than the turbine
that drives it. From my very limited understanding, the smaller the turbine
diameter, the more likely you'll need to multi stage it as the efficiency
losses from specific speeds really hurt.
What's wrong with electric motors anyway?
Troy
-----Original Message-----
From: arocket-bounce@xxxxxxxxxxxxx [mailto:arocket-bounce@xxxxxxxxxxxxx]
On Behalf Of Peter Fairbrother
Sent: Tuesday, 11 September 2018 10:46 AM
To: arocket@xxxxxxxxxxxxx
Subject: [AR] Re: LOX/kero layout (was Re: Above 65000 ft for free)
On 10/09/18 18:57, rebel without a job wrote:
turbines?
At that size, between Reynolds number losses and tip losses, why stick with
Fun? Practice?
Tip path to outlet ratios become enormous, and ratio between tip clearancearea (with reverse flow) and driven area gets uglier and uglier and that size.
I don't know what a "tip path to outlet ratio" for a turbine is, but tip
clearance
isn't an issue here.
The turbine is shrouded over the outer edge; just inside that are the blades
which would look a bit like like C C C C C as seen from the side if they
weren't
covered by the shroud, then under that is the body of the disk. The blisk is
all
one piece, which is why it is so hard to manufacture.
No tip clearance, no reverse flow. Nor is there any significant flow in the
space (I won't call it clearance) between the shroud and the casing.
One possible design is a partial admission turbine with 21 blades and two
nozzles. The preburner output comes rushing out through the nozzles at 820
psi and 573 m/s directed at 20 degrees incidence to the plane of rotation. The
cross-sectional area of the nozzle output is a little larger than the cross-
sectional area of the inter-blade spaces.
But it is a pure impulse turbine, the cross-sectional area of the paths
between
the blades is constant, and there is no (intended) pressure drop over the
turbine.
It can help to think of the gas in the turbine space as incompressible.
The turbine then slows the gas stream to ideally zero velocity tangentially
and
573/(1-cos20) = 35 m/s axially (ignoring losses, slip etc), which then passes
into the main chamber.
The static pressure above and below the turbine disk is the same 820 psi,
roughly speaking. When the blade spaces are not being driven, they are
symmetrical top and bottom, and the pressure on top and bottom is the same,
so there is no flow through them.
Ideally the pressure drop between preburner and chamber occurs entirely in
the nozzles, accelerating the flow to 573 m/s and decreasing the pressure
from 11.6MPa to 6MPa. After the nozzles, and all through the turbine, the
pressure remains constant (ish) until the flow approaches the throat region.
Flow from the top of the disk to the bottom, excepting the intended fast flow
through the blade spaces, isn't the end of the world - in general, you don't
even need to try to prevent it.
The geometry is the same as eg the turbine in the Merlin engine, except afaik
that one is full admission.
Apart from bearings/seals, the only space which needs a tight clearance is the
gap between the static nozzles and the rotating blades - but with only 3
nozzles there isn't much of this. And it doesn't have to be very tight anyway,
roughly speaking just small in comparison to the cross-sectional dimension of
the nozzle output flow will do.
(yes, I can machine to a micron or two if I have to, but I try to avoid it,
especially for things like rotating blade clearances with changing temperature
profiles - no thanks, no, if at all possible no!!)
Positive displacement pumps such as gear and Roots pumps seem to exist atthis size and look easier to engineer.
I was thinking of Barske-style partial admission pumps. PD pumps, yes, nice
fit
to the specific speed, but how to drive them? A gear or Roots pump at
100,000 rpm?
If not a turbine, what?
Peter Fairbrother
--
400N rocket engine v2.0
LOX-rich staged combustion
Pumps:
LOX - 107 g/s @ 1.142 kg/l = 93.7 ml/s
kero - 45 g/s @ 0.78 kg/l = 57.7 ml/s
total = 152 g/s = 151.4 ml/s
O/F = 2.38
output pressure 12 MPa
delivered power LOX 1125W (93.7 ml/s x 12MPa)
delivered power RP1 693W (57.7 ml/s x 12MPa)
delivered power total 1818W
at 100,000 rpm-
LOX head= 1050m 93.7 ml/s specific speed = 166 (= 291 US units) kero
head= 1538m 57.7 ml/s specific speed = 97 (= 163 US units)
at 200,000 rpm-
LOX head= 1050m specific speed = 332 (= 582 US units) kero head=
1538m specific speed = 194 (= 326 US units)
Preburner:
11.6 MPa -> 6MPa exit
LOX/RP1 O/F 36:1
Ae/At 1.0046
Te 613C, 886K
Ve 573 m/s mach 1.07 26.1 kg/m^3
110 g/s flow rate
Turbine:
Gas stream kinetic power 18.06kW
Pumps delivered power 1818W
Turbopump overall efficiency required 10.0%
20 degree Turbine tip optimum velocity 275 m/s
Main Chamber:
Pc 6MPa Pe 1 atm
LOX/RP1 2.38
Ae/At 9.09
ISP: 295s 321s(vac)