几何尺寸与公差论坛

 找回密码
 注册
查看: 1889|回复: 0

urbine engine surge stall

[复制链接]
发表于 2009-9-6 09:36:20 | 显示全部楼层 |阅读模式
turbine engine surge stall
can someone please provide a basic explanation of a turbine engine surge and stall.  the specific application is for rotorcraft (turbo-shaft engines.  what are the specific differences between and surge and stall?  what causes them?  how do you prevent them?  does this happen only in the compressor area of the engine?  thanks for your help
check out our whitepaper library.
surge and stall are closely related phenomena and apply to compressors. in simple terms:
an axial compressor generates a flow and pressure increase down the engine. it achieves this through the flow of air over successive rows of rotor blades and stator vanes. both the blades and the vanes are similar to aircraft wing airfoils. if the flow separates over the airfoils badly the airfoil can be said to stall.
once stalled, the airfoil looses the ability to pump gas down the engine. there is then nothing to prevent the high pressure gas at the rear of the engine from flowing forwards to the lower pressure stages. this reverse flow is called a surge. it lasts for 10-50 milliseconds.
what causes stalls? strong cross-flows into the engine inlet during 'unusual' aircraft manouvres, bird injestion, maybe a few others.
the only surge protection system i have seen was a series of blow-out holes in the casing controled by a rotating ring - but that was on a very old engine.
hope this helps.
gwolf.
stalling (breakaway of airflow from the convex side of the compressor blade, just behind the leading edge) can also be caused by improper angle position of fixed blades, or more usually by a transient condition during run-up to full compressor load. the surge phenomenon, as gwolf explains, is a momentary flowback from high-pressure (aft end) to low-pressure (forward end) lasting only a few milliseconds, disappearing and returning some milliseconds later, creating a vibration which puts extreme load on blades and shaft. surge during transient conditions can be avoided by modifying inlet guide vane angle, compressor blow-off valves set to open to atmosphere above a certain pressure, or even (as north american sabre pilots did during the korean war) cutting off the fuel flow for a split second allowing the compressor to 'jump' from one speed/pressure characteristic curve to another.
hi, guys
some devices to prevent surge is to allow the air flow go downstream trought the compressor. this systems are known like bleed air valves or air bleed port. they light the pressure in order to gain air momentum of the air mass flow downstream.
fran ale
this wikipedia article explains something of surge and surge margins.  
scotty7,
this is way outta my league, but, isn't that why they come in at like 85-90% power (albeit in a high drag configuration)? i know they take time to spool up and the extra drag from flap/slat deployment also adds lift, but it seems like the engine is closer to max power this way. basically in it's "powerband" as would be with an internal combustion engine as in a car. been thru a few go-arounds in my life as a passenger-unnerving i find them!
scott
in a hundred years, it isn't going to matter anyway.
thats basically it.  when the gear goes down, an interlock takes the engine to high idle, which, depending on the engine, is around 10-30% full power.  just sounds loud after being sat at idle from 30,000 feet.  the regs say that the pilot must get power quickly in the event of a go-round or other panic.
it's not in a power band as such though.  gas turbines take time to run up from idle to max power.  you can't just whack the throttles wide open like on a car (the pilot thinks he can, the fuel control unit thinks otherwise), otherwise the engine surges.  clever control units let the pilot do (more or less) what he likes with the throttle.
schuyler...
a note regading this phenomena... stall/stagnation over-pressure loading is brutal on air-inlets. if i re  
surge
in the figure you can see the map curve for a given number of rpm for a compressor.
if you consider the pressure ratio of project the compressor can work in 2 point of the curve,
the one on  the right is stable (b)
the one on the left is unstable(a)
if your compressor is working in b and an inconvenience make the flow of mass increase the pressure ratio go down,
this means that the pressure of the air flow, going out the compressor, is lower than p2=const.
so the flow decelerates because there is gradient of pressure opposing
and the flow of mass decrease and consequently the compressor return to work in b.
in a similar way if the flow of mass decrease a positive gradient of pressure accelerates the air flow and the compressor return to work in b.
if we are in a every little variation of the flow of mass is amplified, so a compressor mustn't work on the left of the surge line.
if a compressor work on the left of the surge line the air is pumped instead of being aspirated.
and the compressor can crash.
usually the surge line interpolate the maximum of the map curves (but this is not a rule).
in axial compressor the field of stability is restricted if there are a lot of stages.
stall
when the laminar air flow become tubolent there is stall. this situation can happen over the casing or over the blading.
the stall is favoured by a positive gradient of pressure (delta_p/delta_x > 0) and it is delayed by negative gradient of pressure (delta_p/delta_x < 0),
for this reason it's really difficult to have stall in a turbine (delta_p/delta_x in a turbine is negative).
blading:
over a blade we can have stall by the suction side or by the pressure side, it depends on the angle between tangential velocity (u) and relative velocity (w).
if this angle differs too much from the angle of project there is stall. if there is stall it means that a part of the flow of mass can't pass through two consecutive blade.
so the exceeding flow of mass must pass through the nearest vanes(z-a & b-c in the figure); in this way the direction of the relative velocity of the nearest blades changes and after a short time the stall is passed in the next blade (from b to a).
we have a rotating stall.
by experimental tests we know that if the lieblein coefficient is < 0.4 there isn't stall on blading. (c is the blade chord, s is the step between two blades, wu is the component of w directed like u and wa is the axial component of w).
(in the figure number 1 indicates the entrance of rotor and number 2 the exit of rotor)
casing:
there isn't stall in the casing if the pressure coefficient in the rotor and the the pressure coefficient in the stator are < 0.5
the most critical point is the hub of stator blade.
it's usual to project with simmetric triangles of velocity (modulus of w2 = modulus of c1 and modulus of w1 = modulus of c2).
so the gap of pressure (delta_p) is the same for rotor and stator.
besides is usual to have the absolute velocity in entrance of the stage-disc equal in modulus of abslolute velocity at the exit. (modulus of c1 = modulus of c3).
(in the figure p3 is the pressure at the exit of stator, c2 is the absolute velocity - c2= w2 + u, it's a vectorial sum).
-----------------------------------------
excuse me for my english, but i'm italian.
i notice that it's difficult to see the figures, so i give you the links:
figure 1
hi,
a while ago i saw a picture of a general electric turbofan that had wavy fan blades. after reading these postings it seems to me that they are designed that way to prevent the phenomena explained here. does anybody know if this is the case, and if so could they provide a little more information on the way that they do this?
variable inlet guide vanes are also a means of helping to delay the onset.
gordon, can you be a bit more specific when you say "wavy?" are you refering to the swept/blended fan blade designs most of us have moved to?
the wavey fan blades are more aero dynamic although conventional ideas about aerodynamics dont explain why very well.  the idea can from a humpback s fin. not only is the leading edge not strat, but protrudes further outpart way down the fin before terminating. heres the best picture i found.
您需要登录后才可以回帖 登录 | 注册

本版积分规则

QQ|Archiver|小黑屋|几何尺寸与公差论坛

GMT+8, 2024-3-29 21:06 , Processed in 0.037266 second(s), 19 queries .

Powered by Discuz! X3.4 Licensed

© 2001-2023 Discuz! Team.

快速回复 返回顶部 返回列表