If you load a wing too much it stalls. If you laod a compressor blad too much it stalls.
Now imagine a whole ring of compressorblades, followed by a whole ring of stator vanes, followed by a whole ring of compressor blades etc.
You can't just overload one blade so you'll automaticlly overload the whole bunch. Now here comes the difference with a wing: A stalling wing just leaves a somewhat wider wake behind, the wake of a stalling row of compressor blades efectively closes the gap between the blades so the entire flow of air trough the entire compressor, all the stalled rows at the same time, are blocked. Imagine blocking a flow of air around 150 cubic meters of air per second(the approximate flow through the hot section of a CF6 at T/O thrust). If you are just testrunning the engine it causes the hangar doors to shake, If you are actually flying it causes the pilots to sweat.
Compressor stall occurs whenever the airflow in the compressor becomes unstable. Instability occurs when the effective angle of attack of one or more compressor blades is exceeded, just as an airplane wing will stall when its angle of attack is too high. In a compressor, the blades effective angle of attack will vary whenever the compressor's operating conditions, such as the inlet air pressure or rotational speed, are changed. Conditions which cause the onset of instabilities are referred to as unstable, or stall, operating conditions.
Because of its destructive nature, compressor stall is one of the most critical concerns in axial flow compressor design. Different types of stall can occur, ranging from mild to severe. In mild cases, stall can manifest itself by abnormal engine noise. In more severe cases, stall can produce loud banging noises and flame or smoke at the engine inlet and/or exhaust. Compressor stall causes excessive engine wear, degradation of performance, and a reduction in engine reliability and durability.
the rest of the article is at:-. .<a href="http://hpcc.grc.nasa.gov/catsslm/StallLine/chapter1.shtml" target="_blank">http://hpcc.grc.nasa.gov/catsslm/StallLine/chapter1.shtml</a>
Just to elaborate on the above comments.. .The airflow through the engine is part of the basic design.. .This airflow obviously takes some account for deviations. These deviations take care of engine wear and outside ambient conditions.. .When the engine is passed of the test cell there is a declared "stall margin" which is basically how much wear is allowed until the engine becomes "worn out". .There are several factors that attribute to compressor stall.. .The first is atmospherical conditions. Cross winds aggrevate the problem particularly with center engines on trijets.. .The second is general engine wear in the compressor section where the airflow becomes unstable.. .Most engines have variable inlet guide vanes and any mechanical failure will invoke compressor stall.. .The problem is that any events of compressor stall increases the wear and so lowers the stall margin.
Location: The home of Dudley Dooright-Where the lead dog is the only one that gets a change of scenery.
If you experience a compressor stall on a helicopter engine it could result in at the least the inspection of the drive line and dynamic system to the replacement of those elements and the inspection of the airframe structure.
> When the engine is passed of the test cell there is a declared "stall margin" which is basically how much wear is allowed until the engine becomes "worn out"<
Partial credit answer
Stall margin is actually the difference between the operating line and the stall line, both of which vary throughout the operating range of rotor speeds and airflow.
The stall margin can and does vary with wear (deteriorates) but also with altitude, temp and the accell or decell rate of the rotor (bodie stalls etc.)
Some recent examples of this have been the altitude stalls on the JT9D where the stall margin decreased due to combinations of altitude and wear (somewhat of a nuisance problem allowed by maintenance until the pilots squwaked it)
Also some recent problems with even young engines have been aggravated by temp changes and bleed loads rather than significant wear problems.
Stalls/surges cause temporary bowed rotors and aggravate wear and in some engines can even cause the blades and vanes to touch (not good) which significantly lowers the stall margin even further.
Location: The home of Dudley Dooright-Where the lead dog is the only one that gets a change of scenery.
On certain gas turbine engines they have an inducer which is normally a rotary compressor. This feeds the axial vane compressor, which compresses the air and pumps it into the combustor can(s). Under ideal conditions the inducer will provide the necessary airflow into the axial vane compressor to maintain stable combustion. Under certain conditions the inducer can’t provide the necessary airflow to the axial vane compressor resulting in combustion instability. This causes the axial vane compressor to slow down at which time the delivery and requirements are again matched causing the combustion to come on stronger than usual. This causes the free turbine to come up to speed so fast as to transmit a very high torque into the transmission. As I previously indicated this load can come on so strong as to damage the driveline and the dynamic system of the helicopter.
I do not dispute the explanations above but it seems that there are differing opinions relative to compressor stall and they are all most likely correct. One point I would like to make, a long time ago I lived next to the North runway at LAX and when the 747s first went into service they would make loud explosive noises as they built up speed. I was told that this noise was caused by combustion instability and to solve this problem they incorporated blow in doors on the inlet of the engine, which solved the problems. The explanation was that there was a mismatch in the needs of the engine and the capability of the compressor to deliver the proper airflow and pressure to the combustor cans that is, until the 747 built up speed during take off.
Loma,. .I did mention ambient conditions as affecting compressor stall. What was not mentioned was Surge bleed valaves and VIGV's!. .. .We have had center engines on B727 surging that when installed in the #1 or #3 positions worked fine.. . <img border="0" title="" alt="[Smile]" src="smile.gif" /> <img border="0" title="" alt="[Smile]" src="smile.gif" />
Small em,. .. .Gas turbine engine compressors are susceptible to two closely related, but distinctly different air flow problems. These are stall and surge. Because the two are so closely related, they are often confused with each other, or taken to be the same thing. If you were studying for your JAR ATPL exams, the distinction you would require is as follows:. .. .COMPRESSOR STALL is the partial breakdown of airflow caused by air approaching a blade or blades at an angle greater than the stalling angle. It may affect a single blade, or a small group of blades, but does not affect the entire compressor.. .. .COMPRESSOR SURGE is the total breakdown of airflow through the compressor. In extreme cases it can result in a reversal of airflow, such that exhaust gas is ejected from the front of the air inlet.. .. .The critical point in the above distinction is that stall is a partial breakdown of flow and surge is a total breakdown of flow. Unfortunately the above distinction often leads to the incorrect assumption that surge is simply a particularly bad case of stall. This assumption is potentially dangerous in that many surges do not originate in the compressor, but are in fact induced by the actions of the pilot. To understand why this is the case we need to take a closer look at stall and surge in isolation from each other.. .. .Virtually all modern turbojet and turbo fan engines employ axial flow compressors, so I will concentrate my discussion on how stall and surge affects these.. .. .AXIAL FLOW COMPRESSORS. .As Captain Squelch has stated, axial flow compressor blades are just like small wings. When air flows over the rotating blades at a suitable angle of attack, they generate lift. This lift tends to pull them out of the front of the engine, but this is not possible because they are fixed to discs attached to the compressor drive shaft. So instead of pulling the blades forward, the lift pushes the air backwards, accelerating it into the engine. But the spaces between the rotating blades are divergent, tending to decelerate the air. So some of the extra velocity is immediately converted into static pressure. The overall effect is that the rotors increase velocity, temperature and static pressure of the air. . .. .Each disc of rotating blades is followed a disc of stationary blades, which also form a ring of divergent ducts. As the air passing from the rotating blades moves through these divergent ducts, it slows down and its static pressure and temperature are again increased. The process is then repeated by further stages of rotors and stators. . .. .AXIAL FLOW COMPRESSOR STALL. .As stated above, the blades are just like small cambered wings. Axial flow compressor blades are generally of very high aspect ratio, and like high aspect ratio wings, they have a relatively low stalling angle. . .. .As angle of attack increases up to the stalling angle, the coefficient of lift increases. But at angles beyond the stalling angle, the airflow separates from the upper surfaces. This causes a rapid decrease in coefficient of lift, with the departing air forming a highly turbulent wake downstream of the blade. This turbulent flow then passes over the stator blades behind, and into the next row of rotors. . .. .The process from this point onwards can take a number of forms. In many cases the turbulence is simply flushed through the engine, such that only one or two blades stall. In this case there will be few if any outward indications that a stall has occurred. In the next level of severity, the initial stall might cascade rearwards, such that several rows of blades are affected, before the subsequent stages regain control of the air. In this case, the pilot might detect a sudden increase in internal temperature and increased vibration from the engine. The next level of severity, with the stall cascading all the way back through the compressor, will obviously increase the severity of the outward symptoms.. .. .The above processes do not necessarily mean that the airflow through the compressor has broken down completely. Some engine types can behave in quite bizarre manners. One of the US military turbofans introduced in the seventies or eighties for example, exhibited a locked-in rotating stall. In this case the stalling of a single blade cascaded not downstream, but onto the next rotor blade on that disc. This cased a small pocket of stall to rotate in the opposite direction to the engine, but at about half of the engine RPM. The engine would continue to run and produce some thrust, but in an extremely unhappy manner. Fortunately this particular phenomenon is not common to the majority of engine type. . .. .Stall occurs when a blade meets incoming air at an angle greater than its stalling angle. So any factor that increases angle of attack beyond the stalling angle is likely to cause compressor stall. To understand what these factor might be we need to look at the angle of attack of a rotating blade. . .. .To investigate this we can start with a blank sheet of paper. Draw a short (1 inch) line in the centre of the sheet, such that it is angled downwards from right to left at about 30 degrees. We will consider this to be the chord line of a compressor blade. The blade is rotating down the page and the airflow is coming in from left to right. Because the blade is rotating down the page, the airflow due to rotation is flowing up the page To represent this, draw a 2 inch arrow pointing vertically upwards to the trailing edge of the chord line. The remainder of the airflow is made up of axial flow into the engine. To represent this, draw a 1 inch horizontal arrow so that its head touches the tail of the rotational flow arrow. We can now draw the resultant of these flows, or the real airflow, by drawing a third arrow from the tail of the axial flow arrow to the head of the rotational flow arrow. The angle of attack is the angle between this third, resultant flow arrow and the chord line of the blade.. .. .To investigate increasing RPM, simply increase the length of the rotational flow arrow and move the axial flow arrow down to meet its tail, then draw a new resultant arrow. This will give a greater angle of attack. To investigate the effect of increasing the axial airflow velocity, simply increase the length of one of the axial flow arrows, then draw a new resultant. This will show that increasing axial flow decrease angle of attack, while decreasing axial flow increases it. . .. .By repeating this process using various combinations of rotational speed and axial flow speed we will find that angle of attack is increased by increasing RPM or by decreasing axial flow velocity. So any action which does either of these things in isolation, will increase the tendency to stall. We should however note that increasing RPM will usually increase axial airflow velocity, so stall will not necessarily be caused simply by increasing RPM. . .. .The important point here is that stall will occur when the ratio of RPM to axial airflow velocity reaches some critical value. So if for example we pull the aircraft into a very high angle of attack, or into a very high sideslip angle, the disruption of airflow into the intakes is likely to cause stall. Similarly if we accelerate the engine very rapidly, such that axial airflow cannot keep up with RPM, we will again cause stall. Common (pilot induced) causes of stall include:. .. .a. Excessive angles of attack or sideslip angles. This sometimes affects one engine position more than others.. .b. Operating with iced up intakes.. .c. Excessive engine acceleration or deceleration rates.. .d. Starting engines when parked in strong crosswinds or tailwinds (or with intake or exhaust blanks still fitted).. .e. Taxiing or flying too close behind other aircraft such that their exhausts or wakes flow into our intakes. . .. .But the stalling angle of a given blade is not entirely fixed. Structural damage caused by the ingestion of hard objects, erosion by sand or grit, icing, and surface contamination, can all reduce the stalling angle. So engines tend to stall more easily as they age, or if their compressors become iced or dirty. In the case or naval or SAR helicopters for example, this latter effect is particularly problematic. When hovering or operating close to the sea, the build up of salt crystals on the blades can quickly reduce stalling angle to dangerous levels. . .. .The degree to which the engine is able to control its airflow increases from front to rear. This is because the first row of rotor blades must accept whatever airflow they meet, whereas the subsequent rows receive air from the blades ahead of them. So the danger of stall in a serviceable engine is greatest at the front. The most common method of stall prevention is the use of variable angle inlet guide vanes or VIGVs. These are an additional row of non-rotating blades, immediately ahead of the first row of rotors. By changing the pitch angle of these vanes, the angle of the incoming airflow is varied to ensure that the angle of attack of the first row of rotors is always less than their stalling angle. This method is sometimes extended so that several rows of stator are of variable angle. I believe this has been applied to up to 10 stages in some engine.. .. .COMPRESSOR SURGE. .This phenomenon should perhaps be more correctly termed "compressor pressure surge". It occurs when the pressure downstream of the compressor, or in the downstream end of the compressor becomes too great. The airflow breaks down, the blades stall, and the pressure then surges forward, reversing the flow through the compressor. It is this that results in the ejection of exhaust gas from the front of the air intake. Like stall, the severity of compressor surge varies from minor to major. On an increasing scale of severity the effects can include minor vibration, increased internal temperatures, severe rumbling and loss of thrust, flame out, and (ultimately) immediate destruction of compressors and turbines.. .. .The most common pilot-induced cause of surge is excessive engine acceleration. If the throttle is opened too quickly, the pressure in the combustion chamber will immediately increase. If the compressor cannot accelerate quickly enough to match its outlet pressure to combustion chamber pressure, the airflow out of the back of the compressor will cease. This will increase the angle of attack of the blades at the rear of the compressor causing them to stall. Reduced flow through the stalled rear blades will reduce flow through the blades immediately upstream, causing them to stall. The stall will then cascade upstream, from rear to front. This stalling of progressively more of the compressor will reduce compressor outlet pressure still more, making the mismatch between combustion chamber pressure and compressor outlet pressure even greater. This allows the high pressure air in the combustion chamber to expand forward reversing the flow through the compressor.. .. .To understand another common cause of surge we need to look again at the compression process. The blades are little wings and like wings they are sensitive to airflow velocity. If the velocity is too low they will stall due to excessive angles of attack. If the velocity is too great they will produce shock waves, leading to shock-induced separation and shock stall. So the flow velocity through the compressor should ideally be kept constant. But the compression of the air as it flows through the compressor reduces its volume. If the cross-section of the air passage is constant, this reduced volume of air will slow down, or re-expand to fill the space, thereby undoing all of the work that has been done in compressing it. To prevent this, the cross-section of the air space gradually reduces from front to rear. . .. .This has the desired effect in maintaining constant velocity, but means that the engine is now committed to achieving a set rate of reduction in the volume of the air. This is no problem when running at optimum RPM, but at lower speeds, the compression process is less efficient. This means that the air is too big to pass through the reducing air space at the rear of the compressor. Left uncorrected this will cause the rear of the compressor to choke, with the airflow breaking down. This will stall the rear blades, reducing the compression efficiency still further. Once again the stalled rear stages will stall those in front of them so that a stall progresses from rear to font through the compressor. . .. .This problem is commonly prevented by the use of anti-surge bleed valves. These are controlled holes in the side of the compressor casing, usually about half or three-quarters of the way downstream. At low RPM they are open, allowing excess volumes of air to be dumped overboard. As RPM increases, the compressor efficiency improves, so the valves are gradually closed so that all of the airflow passes through the entire compressor.. .. .The subjects of compressor stall and surge are obviously very complex, and this post is already far too long. For readers with the required time (and interest) available, I would recommend "The Jet Engine" by Rolls Royce. ISBN 0 902121 04 9. (For the more cynical readers, "no I do not work for them, nor have I ever done so").
Keith . .. .Well done. Great stuff.. .. .A comment though if I may:. .. .15 years ago in the UK the Rolls people I dealt with talked only of Surge. But there were two types of surge, Self Clearing (your Stall) and Locked In (your Surge) and of course only one surge line to go with the running line.. .. .In the USA at that time everyone seemed to use the word Stall instead of the word Surge.. .. .I would be interested to know when the ATPL definitions you describe came into use.. .. .Regards. .. .John
John:. .. .I believe the two surges your are referring to are all surges. There are recoverable surges and irrecoverable surges. For the latter, you will need to shutdown the engine. Also, I don't think it's true that Americans use the word stall exclusively. I do notice the American engine company that has fewer surge problems with their products tend to use stall and surge interchangeably. Contrary to what Keith has said, my understanding is when a blade row stalls, it's (almost) always rotating stall, and there are more types of rotating stall than what Keith has explained. I would say a lot of people in the US would use rotating stall for stall (local phenomenon), and either surge or stall for surge (system phenomenon). And often times, a rotating stall is a precursor to a surge.. . . . <small>[ 03 March 2002, 15:48: Message edited by: casual observer ]</small>
T-shirt. .. .>We have had center engines on B727 surging that when installed in the #1 or #3 positions worked fine.<. .. .Yea, that center S-duct sure does "suck" <img border="0" title="" alt="[Razz]" src="tongue.gif" />
Casual. .. .Thanks for that. In commenting about the terms used in the US I did say it was a long time ago - 20 years actually - not 15 as I said, now I have looked in my logbook!. .. .The model I have in my head for pondering all these (many and complex) types of compressor aerodynamic events is a very simplified one, but it served me well when I was in the business and goes like this: . .. .When an engine is running, if any stage of the compressor suddenly drops its output pressure (whatever the cause) then there will be knock on effects for the stages behind and perhaps even in front. If those stages are themselves a bit near their aerodynamic limits then the disturbance is likely to spread very rapidly and in the end the combustion chamber pressure may not be prevented from relieving itself forwards. If on the other hand the first downstream stage is not particularly struggling there may be a chance to recover the situation without shutting the donk down. Most of the recoverable surges that I experienced (probably in the hundreds) could be sorted if you chopped the throttle quick enough after the first 'pop' or removed the cause of the trouble in the first place - often inlet flow distortion caused by excessive sideslip or incidence, especially when associated with too quick an accel or too much N over root theta due to a poorly adjusted PRL. . .. .Airframe type specific problems, caused by intake flow distortion, can themselves be very transient and as the engine starts to react can go away, making the pop an isolated and self recovering event even without reducing power. When establishing such manoeuvre related surge boundaries one could learn to yug to a high alpha, find the engine OK and yug then to a tad more and so on, until you just tickled a single pop. As opposed to going at it bull at a gate and have the first event become a locked in surge requiring shut down. . .. .Regards. .. .John Farley
John,. .. .I'm note sure (have no idea) when the distinction "stall is partial breakdown and surge is total breakdown" came into use with the JAA, but it was a frequently used question in the old CAA engines exams for many years. . .. .Looking through the few references I have to hand:. .. .a. The 1955 edition of the RAF's AP3456 (they called it the AP 129 then ) . .does not discuss stall and surge as separate phenomena. It discusses only the situation where "excessive pressure at the rear of the compressor causes flow breakdown, such that the blades stall".. .. .b. The 1969 edition of the same book, amended up to 1972 standard (yes, I do have a strange library), distinguishes between stall and surge. It includes the statements "The aerofoil sections of compressor blades can be made to stall if the relative airflow is varied beyond the critical limits" and, "In an axial flow compressor, surging indicates a complete instability of flow through the compressor". It also includes a surge/stall map in which the significant elements are marked "Running line", "Surge line" and "Unstable area".. .. .c. The 1952 (revised 1974) edition of the Gas Turbine Engine by Pratt & Whitney Aircraft Group, does not provide any detailed distinction between stall and surge, but includes the vector diagrams I described in my previous post and includes the statement "When enough blades in enough rows are stalled, the compressor surges". This book includes a large number of surge maps, none of which include the word stall. . .. .d. The 1986 edition of the Jet Engine by Rolls Royce, does not provide a detailed distinction between stall and surge, but includes the statements, "…..the blades may stall because the angle of incidence of the air relative to the blade is too high (positive incidence stall) or too low (negative incidence stall" and " If the engine demands a pressure rise from the compressor that is more than the blades can sustain, surge occurs. In this case there is an instantaneous breakdown of flow through the machine, and the high pressure air in the combustion chambers is expelled forward through the compressor".. .. .Knowing the way the old CAA examiners (who are also some of the new JAA examiners) work, they probably looked at the AP 3456 and the RR book and came up with the distinction, "stall is partial breakdown and surge is total breakdown". Looking back to my earliest studies of this subject in the mid sixties (I can just about remember it), the JAA definitions were used by the RN.. .. .Having said all of that, I think that in most aviation-related crewrooms around the world, the sound of an engine coughing or grumbling (however briefly) is often met with the comment " somebody is surging his engines". . .. .Casual Observer,. .. .My suggestion that the locked-in rotating stall is unusual was based largely on circumstantial evidence (or rather lack of it). When the US military was having this problem with its (then new) turbofan back in the seventies or early eighties, the various journals (Flight and AW&ST) discussed it as if were a new phenomenon. This, coupled with the lack of discussion of this subject since then, led me to conclude that it was comparatively rare. On reflection however, I can see that the reported events might simply have reflected the initial awakening of knowledge of a common phenomenon, rather than an isolated incidence of a new one. . .. .The turbulent flow generated from a stalled aerofoil flows mainly from its upper surface, so in a compressor it will affect the blade above its convex surface (in the disc) more than the one below it. So in addition to cascading rearwards, the stall might well migrate around the disc. But the camber on the rotors faces away from the direction of engine rotation, whilst that of the stators faces in the direction of rotation. So the circumferential stall migration in the rotors should be in the opposite direction to that in the stator. Also, because the stators are not rotating, the absolute value of the rotational velocity of their stall is likely to be greater than that of the rotors. So if a stall rotates, the predominate direction of this rotation should be in the direction of engine rotation. In the case of the engine referred to above, the stall rotated in the opposite direction to engine rotation. You are of course correct (if I understand you correctly) in implying that if a single blade, or small group of blades stalls, it will be a rotating stall. But it seams to me that this will turn with the engine. Much of this is of course supposition on my part, and I accept the fact that you might well be correct. . .. .I have more difficulty with Captain Squelch's argument that all of the blades in a disc must be stalled simultaneously. To see what I mean, let's consider the case of front end (first row of rotors) stalling due to turbulent airflow entering the intake. Turbulent airflow is by definition different at all points, so every blade (and probably every point on every blade) will experience a different angle of attack and flow velocity. The probability of even any two blades stalling at exactly the same instant is therefore remote. If individual blades stall at different times then there must be times at which some, but not all of the blades in a disc are stalled. . .. .It seams to me that the idea that all blades in a disc must be stalled simultaneously only works if we apply the arbitrary definition "a blade in any given disc is considered to be stalled only when all of the other blades in that disc are stalled". But if we apply this definition, the concept of a stall rotating in a disc become meaningless. If all of the blades are stalled, what exactly does it mean when we say the stall is rotating in the opposite direction to engine rotation? If on the other hand only part of the disc is stalled, then stall rotation means that blades move into and out of the stalled condition as the pocket of stall passes over them. . .. .I don't think we need to be too concerned (or surprised) about our different terminology for stall and surge; There are precious few areas in which American English and English English coincide in everyday matters. So we shouldn't expect too much in such technical areas as this! And anyway, faced with the type of symptoms we are discussing here, some airlines would simply say "That is a perfectly serviceable engine, so get your chickens and goats onboard. We are already 2 days late and I cannot shut down the engines because that one hasn't got a starter fitted". (when considering this scenario readers should pick an airline to suit their personal prejudices).. .. .On a more serious note, some of the earlier posts appeared to suggest that stall/surge is entirely due to engine wear. The main purpose of my post was to illustrate that many of the causes of stall/surge are within the control of pilots. But to understand what should be done to avoid and recover from these problems, it is necessary to know something about their causes and behaviour. Precise definitions of terms often matter more in the examinations than in the day-to-day business.. .. .Lomapaseo,. .. .Your experiences in the 727 illustrates the fact that some engine locations are more problematic than others. The centre engine in a trijet will clearly experience more difficult airflow conditions than the outer ones, particularly at high angles of attack. Similarly, tail mounted engine are likely to be more prone to stall/surge than underwing ones.. .. .It's funny how a simple question so often turns into a complicated one. There must be a term for this……..PPRUNERISM perhaps?
I dont know if small em is taking an exam or looking at the practical side.. .The above answers are obviously all very technically correct. The real problem is the application of all this theory.. .From a tech log point of view I have seen 'compressor stall' and 'compressor surge' all reported for very similar problems. The reports contain references to popping, banging, vibration, indication fluctions, etc.. .From a maintenance point of view they are one of two things, either a transient that cannot be repeated in which case it is an inspection for compressor damage or mechanical failure of bleed valves or VIGV's.. .I have heard 727 stalling/surging as they go by and yet the crew did not hear anything!. .Most manuals particularly big fans, warns against high power in crosswind situation and very often, on ground runs, just turning the aircraft a few degrees clears the problem. .Again the condition of the engine has a great effect on these situations and tired or dirty engines suffer more.
Well frankly I am stunned at the wealth of info in reply to my brief question. Thanks, guys!. .. . </font><blockquote><font size="1" face="Verdana, Arial, Helvetica">quote:</font><hr /><font size="2" face="Verdana, Arial, Helvetica"> I dont know if small em is taking an exam or looking at the practical side.</font><hr /></blockquote><font size="2" face="Verdana, Arial, Helvetica">It was partly just wanting to know How Things Work, and partly practical - but not even aviation related. I was musing over an engineering problem that involves a fan in a tube (no clues - patent pending!), and it struck me that it was similar to what I understood compressor stall to be. Turns out I was on more-or-less the right lines.. .. .em