Results of an Unloaded Power Turbine
2023 ж. 20 Жел.
38 521 Рет қаралды
Let's have a what if session about power turbines becoming unexpectedly disconnected from their load while running.
Here's the link to the video I mention which features a look at an LM1600 PT
• Power Turbines 4 Diffe...
I have a playlist all about power turbines that you may enjoy: " Power Turbines and what they do... "
I know I wrote to JayZ many years ago about this topic. 20+ years ago was in school. The class above me had "restored" a Boeing GT502 out of the swedish Stridsvang 103, yes it was partially Gas Turbine powered and partially V6 2 stroke diesel. This Gast turbine was a single spool with a singel stage tubine and an axial compressor, then it had a rear exiting single stage power turbine as Jay talk about at the end. It also featured a reduction gearbox so the RPM would better match up with the V6 diesel because they then had a combining gearbox. Anyways this Boeing GT502 was run by students at idle in a engine stand, probably intended for moving it mainly and working on it. This worked fine until on day a class ran it and the teacher gave it some "gas", at that very moment the power turbine exploded. I was just heading in for school and all the students that had a class cam running out of a smoke filled workshop and I can here the gas turbine of the Boeing GT502 spooling down being fuel starved as the heavy duty metal door was opened. I get in there and there is the Gas turbine with the outer casing of the power turbine blown away, the main turbine still spooling down, you can see its single stage turbine, the stator vanes for redirecting the gases to the power turbine are gone, the gearbox has oil pumps hanging off the sided etc because all the power accessories was broken off aa if they turned with there individual axles. The gearbox and power turbine was later taken apart by the student that rebuilt this engine and took the chance to look at the parts when they left (this was a class above me so not my project). The power turbine bearing and main gearbox bearing was cracked and the cracks went all they whay in to the gearbox casing in a rotating star pattern many inches long or like 100mm or so long cracks repeated 5-6times I would guess so it looked as if the axle seized and just torqued the gearbox until it shattered. The pumps on the gearbox for lubing etc where cracked at the flange that bolts to the gearbox so they must have basically been so over torqued they rotated with there own axles and fell out and was just hanging in hoses and bent pipes. There was probably some hydraulic pumps on it, the Stridsvagn 103S had a LOT of hydraulic systems, was like a feature of that thing. The power turbine itself, the axle was well fine, not broken off or anything, the Disc for the power turbine looked fine but all the blades left it. Idk if they left because it ate up the stator in-front of it (maybe the gearbox had failed at that point) or if they left because the root of the blades gave up before the disc but the disc was intact on the axle. And yea those small blades that are like 1inch or 25-40mm I would guess from memory went trough 4mm of steal in a lot of cases, we found them stuck in steel beams, what was thinner metal they just wen trough never to be found, probably orbiting the "moon" now because a lot of them left trough the roof literally. No one was killed luckily, but if you would have been in the way of the blades, yea.... The turbine casing was just like 1mm welded steel so it was like aluminum foil for your cooking in the kitchen, not a match at all for the blades when they exited, that part of the power turbine was basically evaporated and never found again. But in this failure scenario the disc survived but the blades left the disc instead for one reason or the other. The gearbox might have failed first and allowed the axle to move forward or sideways so the blades and disc hit the stator or casing (very thin metal, 1mm at the most), or if the blades broke at the root and took the casing and stator with it due to the overspeed of the power turbine. if you add this part to the YT url you can see someones Boeing 502 running watch?v=z-timBvjbYk&t=184s I would ot stand in line with the power turbine, nor would I run it without 8-10mm of steel around the power turbine because there is really no containment for a blade failure on those it seems. Anyways thats my story. And no I was not directly involved with this gas turbine project, I had my own project.
Actually, fan shaft breakage overspeed in a Rolls Royce is stopped by a MECHANICAL system. Not the FADEC. It uses a reference tube and pawl carrier with a cable that cuts off fuel to the OSU. It is on all BR engines including the newest Pearl 700. As well as commercial engines. If you are ever in Indianapolis, stop by the Rolls-Royce RCTC and I will show you around. Also, the LPT shaft is actually relatively thin, as it is hollow, and has to accommodate the reference tube. Your reference to Spey bog chains is very close to how the newer reference tube system works. The pawls will snatch the cable and stop fuel before 1/2 a revolution occurs. We still use this system as it works ALWAYS. Pearl 700 uses the latest technologies but still has mechanical overspeed protection.
HALF a revolution? That is insanely fast
Thank you! It's useful to get up-to-date information on current engines. I'm now 20 years into retirement. AgentJayZ mentioned my name in relation to a 'bog chain' (like an old WC high-level cistern) arrangement in a Spey engine. I do vaguely remember a Derby engine of that era having such an arrangement, and I think that one of the Trents (possibly the 500 in the A340?) had a similar system. The Olympus 593 in Concorde certainly did.
@@damonlandsdown The pawl heads actually drop at 35 degrees of shaft twist, and the cable is pulled at that time, another 1/2 inch of cable pull, and the shut off valve is pulled. The fan and LPT will obviously freewheel for a while, but fuel is cut super fast. The system is designed to stop overspeed immediately. Pearl 700 LPT runs 7500 rpm, has 4 LPT wheels, and an overspeed would be an astronomical amount of energy on burst.
On a Pratt 308A the cable is externally visible along with a pulley to redirect motion forward to the shut off. All the turbofans I've ever touched have the same safety shut off devices.
According to Stanley Hooker's book, they added a no load shutdown to the Proteus after the loss of a Britannia prototype.
In rocket launches they have a term “RUD” or Rapid Unscheduled Disassembly. I guess that would also cover the event described here 😊 Thank you for yet another great video.
Having contributed to this discussion about possible catastrophic failures in gas turbine engines, I am now at the gate waiting to board my flight from GVA back to BRS.
I like "brutally strong" (no unnecessary adjectives). Happy Holidaze, JZed ⚡
From what I witnessed of a CF6-6D fan shaft failure was the LPT rotor move aft until it contacted the stators and then proceeded to "grind" to a halt. Typically there is one ball bearing in the mainline stacking which acts as a thrust bearing for each powered shaft. If a shaft separated aft of the thrust bearing the lower pressure turbine would move aft into the stators and stop.
The Olympus power turbines with which I was familiar decades ago had a mechanical overspeed trip system. As I vaguely remember it, there was an asembly on the power turbine shaft consisting of a ring mounted on the ends of a pair of stiff cantilever arms. The CofG of the ring was offset and, if the turbine went into overspeed, the ring would move off centre, catching a lever, which would actuate a valve to shut off the fuel. That was to catch a potentially dangerous unloaded overspeed, of course. There was also a hydromechanical overspeed governor for reducing fuel flow if, for instance, the prop came out of the water in heavy seas.
great video, Actually the 767 was on takeoff when the failure occurred. One fragment which weighed about 50 pounds went over half a mile into a UPS building.
Love your talk about the LM1600. We run one with a generator 50 hz on thé PT. And use the exhaust gasses for a steam boiler.
From the recesses of my memory, I've just come up with another 'if all else fails' mechanism, which I vaguely recall being discussed at R-R Bristol during my time on Turbine Design. The discussion was around the form of the firtree serration, whether it should be 'coarse', which was supposedly typical of Derby designs, or 'fine', which was favoured at the time at Bristol. As I recall, the reason advanced for this was that, in the event of an overspeed, the dilation of the disc would allow the release of the turbine blades, thus preventing a disc burst. Having said this, as you've recently shown us, the failure of one or two blades can be uncontained, albeit with relatively low energy. However, the release of multiple blades would be highly likely to cause them to form a 'wedge', which could force its way through the turbine casing - it has happened. Nevertheless, it would potentially be less dangerous than a disc burst - but still with the risk of localised damage to the aircraft.
Thank you for a second splendid answer!
I overbuilt the driveline in my van for this reason. I don't think the N2 governor could catch it fast enough if it broke something. Keep up the good work Brother!
Hello again, AgentJayZ. There is one other, 'if all else fails', overspeed prevention system for LP/fan turbines. It's known as 'tangling'. In the event of an LP shaft failure, which detaches the turbine from the thrust bearing, allowing it to move rearwards, the blades are designed to clash with their NGVs. The intention is that this causes a huge amount of mechanical damage, substantially destroying the blades and their aerodynamic function, in adition to the massive friction of the destructive process.
That's a good idea. That would be why the thrust bearing is always on the cold end, ant the turbine is in a roller bearing... I never thought of that.
@AgentJayZ If you take a close look at your small GA of an early RB211 and compare the LP turbine with that of, say, a JT9D, does it look rather more robust? It probably wouldn't 'tangle', whereas the JT9D would.
The Willams FJ 44 and variants have a lever that goes through one of the rear housing struts that is right aft of the LP wheels. If the LP shaft breaks the aft movement presses on the arm and triggers a spring loaded Fuel cut off.
This video reminds me of one you made a few years back, about the "do to stand here" line painted vertically on the sides of the start cart.
Awesome video 😊😊🎉🎉
I operate a couple LMS100’s here in Colorado. Amazing machine, at our altitude on a cool T2 temp, we can make 94 MW gross.
Those are monsters! Are you allowed to take photos?
@@AgentJayZ yes I can. How do I send them to you?
email address on the channel page.
I bet that unloaded power turbine test attempt must have sounded terrifying. Now I have to wonder: what's the most terrifying sound you've heard while testing these engines? (aside from the random-xylophone tinkling of a dropped part)
if my math checks out 90000Nm on a 60cm diameter disk that weights 250kg result in an accelation of roughly 38000rpm per second
Something like missing an upshift under full power... if the car makes 35 thousand Hp...
So what you are saying is, when turbine parts fail due to over speed rotation, bad things happen in raptid succession.
I've never quite understood why they don't just connect the fan, compressors and turbines to a single drive shaft but now I kinda get it; the compressors need that high speed to pull in as much air as possible, but the fan running at the same speed, well, it would just shatter. I suppose having the multiple drive shafts/speeds is also where the difficulty comes from when designing a turbine engine, since you'd want the fan to be able to spin fast as safely possible (or within some range) while not disintegrating the high pressure stage from all that available air. I guess that's where they could just add extra steel to the turbines? Very interesting video!
Have a look at the playlist that I mention in the description. It will help you understand the need for separate systems/shafts.
I reckon you’ll be getting a warp core and the corresponding nacelles one of these days.
FYI, 66,000 ft-lb torque at 3600 rpm is about 45,000 shaft HP
That's right in the range of the various models of the LM2500.
So is this what happened with quantas flight32.?? An immediate failure of the IP turbine.? Love to know your thoughts on this particular flight.
Yes, mentioned in the vid briefly.
Here's an edited copy of my reply to another contributor to this discussion. In 2010, the No.3 engine (a Trent 900) of a Qantas Airbus A380 suffered a catastrophic failure of its IP turbine disc, with a classic overspeed burst into three 120deg segments. It's already been discussed numerous times on this channel. The main oil feed tube to the HP/IP turbine bearing housing failed, due to a manufacturing 'drop-off', resulting in a raging oil fire in the bore of the IP turbine disc. This overheated and softened the disc drive arm, which attached it to the IP turbine shaft. The drive arm sheared, and the disc accelerated to overspeed in a couple of hundred milliseconds. For your information, compressor and turbine discs are typically designed to have an overspeed margin of 140percent, or so. Naturally, R-R went into 'war room' mode and discovered that a number of other engines were potentially at risk, and these were subject to a 'campaign' action. I visited the site in Singapore where much of the work was carried out and I was told that engines were being turned around in as little as three days. I would hasten to add that the basic design of the oil tube had accumulated millions of safe flying hours, dating back to the days of the RB211. I'm also aware that R-R made changes to the software of the FADEC, which were intended to 'catch' an overspeed, by shutting off the fuel. How effective this would be in the event of another such event, which is extremely unlikely, I would need some convincing.
My day job of nearly 20 years getting commercial turbofans certified is coming to an end on 31 December, so I'll get my turbine fix here! (But I will no longer be able to caress GE9X fan blades when I walk into the office ...)
Would a shop-worn blade from a CF6-6 do? Got something cool to trade?
I am a non destrucive.test inspector for a major.airline. We inspect those shafts using a radioactive isotope inspection procedure on wing. We have to rope.of the are 300 feet.when doing this inspection.
Happy New Agent Jay Z year!Quick question hope not heard before. Could you shed some light as to why the flying engines don't burn gasious fuel & if liquid fuel is the only choice for flying jets,what are the pros & cons & limitations?or other uses of the liquid fuel serves to an aircraft before entering the fuel nozzles to produce that heat?Thanx!
My only partially ignorant observations for a turbojet (not an industrial free-turbine machine): 1) a fractured shaft might have immediate and 'not good' things to do to the bearings supporting that and the other coaxially located shafts--potentially reducing (!) runaway speed, 2) I'd think modern turbofan engines would have RPM monitoring on all each of the spools, for multiple reasons--wouldn't the FADEC immediately cut fuel (some or all) commensurate with any spool's overspeed?, 3) The fan provides some boost to core airflow--a sudden drop in inlet pressure (absent automatic adjustment mechanisms) might well cause a stall, and would certainly reduce power/airflow, and 4) the airfoils of each turbine will normally be operating at an angle-of-attack to gas flow, as if 'making lift' (if they were conventional wings), but if their disk suddenly became free-wheeling, their upper speed would be limited to that of the free-stream (my term) gas flow (zero angle-of-attack)--not some unlimited RPM. I don't know what percentage increase there would be between ideal power delivery angle-of -attack and free-stream--can't be terribly large numerically. As you said though, at max power things are closest to maximum stress, so only automatic mechanisms and various factors of safety can save the day. Thoughts?
Good thinking. A small difference: when a power turbine is disconnected from a load, the airflow into it does not change at all. What changes is the load goes to zero. The resulting acceleration is very large. The turbine does have a speed where further acceleration does not happen, but the breakup speed occurs before that. Kablammo!
66,000 ft-lbs. Imagine standing on the end of a cheater bar that's one city block (say 330 ft) long, and you weigh 200 lbs 😂
I worked at a nuclear power plant. It was cool to see, I just had the misfortune of hiring on 4 months before the utility decided to shutter the plant. I do remember one of my first days there I went up to the turbine deck and went on the turbine walkway. They had guards around each coupling. The guards used expanded metal, so you could see the coupling and the shaft turning. I went to the coupling between the last LP turbine and the generator. The generator produced 1.25GW. I remember going back to my desk and banging out that at 1,800 RPM and that power, it was 9.4 million ft-lbs. Using your metric, it would be an 8.9 mile long lever to produce that torque with a 200lb person on the end. Heck, even the turning gear was crazy. It had a 60HP motor to spin the turbine at 1.2 RPM when the plant was idle. That thing had over 500,000 ft-lbs of grunt to spin the turbine.
Torque measured in foot-miles? Oh, I did enjoy that!
From a reference I ran across: “One of the earliest turbofans was a derivative of the General Electric J79 turbojet, known as the CJ805-23, which featured an integrated aft fan/low-pressure (LP) turbine unit located in the turbojet exhaust jetpipe. Hot gas from the turbojet turbine exhaust expanded through the LP turbine, the fan blades being a radial extension of the turbine blades. This aft-fan configuration was later exploited in the General Electric GE36 UDF (propfan) demonstrator of the early 80s. One of the problems with the aft fan configuration is hot gas leakage from the LP turbine to the fan.” So interestingly a reverse configured turbofan with the fan blades concentric to the “power turbine blades” that drive the fan blades. Interesting how the industry didn’t ultimately settle on that solution.
Familiar with the -23. If only I had joined S&S Turbines a couple of years earlier... They had a few of those engines, dismantled them for useful parts, and scrapped the rest, including those incredible double blades. They are now extremely rare. I really wish I had one. The aft fan never worked as well as the front fan turbofan. Not enough bypass ratio, and it does not boost the airflow into the core. GE still calls the fan of their conventional turbofans a "core booster". Also the leakage you mention would have cost performance and fuel economy. But it was the start of a long process that brought us to the miraculously efficient and powerful airliner engines we have today.
@@AgentJayZ The closest “story” that I have heard about having a high power / torque load disconnect “rapidly” was from a person who did electrical characterization tests on power plants. If I remember correctly, they were characterizing (probably) the electrical generator on a GW nuclear power plant and the generator became unsynchronized (total load drop) from the grid. As this was a steam turbine the driving steam was vented rapidly and the reactor was scramed. The description of the reaction of the test personnel was impressive, something about a lot of running due to extreme noise and vibration (I think from the steam venting). Supposedly the automatic shutdown was successful and mostly the only thing hurt was their pride.
Here in the UK, Metropolitan-Vickers designed and tested an aft fan turbofan version of their F.2 turbojet towards the end of WWII, years before GE came up with their design. And yes, there were the inevitable leakage problems between the 'hot' and 'cold' streams, combined with the absence of a 'booster' effect for the core engine. Even so, the engine was significantly more efficient than the basic turbojet, because of the improvement in propulsive efficiency.
@@grahamj9101 Thank you for the interesting reply. Best of wishes for happy holidays.
Having previously mentioned MetroVick's aft-fan turbofan engine, I've checked on its history. It was designated as the F.3, and used the F.2/2 turbojet as its gas generator. The design was, started in 1942 and the aft fan had two contra-rotating fans, with the fan blades mounted on the tips of their turbines. Thrust was increased from 2,400lb to 4,000lb, with a corresponding reduction in specific fuel consumption. Of course, this was paid for by an increase in the weight of the engine. Based on their experience of the F,.3, MetroVick went on to design and test an open rotor aft fan turbofan. This was over 40 years before P&W and Allison began experimenting with their open rotor propfans. I should also mention that Frank Whittle came up with an aft fan turbofan arrangement, mounted behind his W.2/700 engine. It was intended to power the Miles M.52 supersonic research aircraft and included a form of reheat (afterburning if you must). In its final form, the combined fan/reheat arrangement was designated as the No.4 Augmentor. It actually ran towards the end of WWII, and there may be a clip of it somewhere on KZhead.
Will the hardware and software save the electrical generator end if it loses load?
They are designed to control the rpm of a power turbine that loses its load. Whether the disconnect is at the drive to the generator, or between the generator and the electrical "grid", the fuel would be shut off.
@@AgentJayZ I would be interested in seeing how fast they can shutoff fuel and if they have redundancy. I've never been in an aeroderivative power plant, just steam and industrial gas turbines (GE Frame 7 specifically.) The single shaft industrial generators have a whole lot more inertia and can't lose all load due to the compressor. I worked at a steam plant for a short while. With the steam turbine, there are redundant valves to shutoff the steam, operated by two different systems. On a load loss, they can shutoff the steam flow in about 5ms. This is most likely from a short on the outgoing HV lines, as the load into a short on an AC generator is counter-intuitively nearly zero. In the plant I worked at, the normal throttle valves were shut by the normal governor system. The backup stop valves were all mechanical and hydraulic. If a spring loaded flyweight snapped out, it would hit a hydraulic dump valve and that went through a quick dump valve to more quick dump valves at each stop valve. These valves were massive, and when I saw them operate you really couldn't see them move. One moment they're open, and then in the blink of an eye they're shut.
A generator has vastly more rotating inertia compared to the power turbine rotor. The rotor are usually solid steel with slots cut in them, and then the slots are full of copper and insulation. They're as big in diameter and probably 10X longer than the power turbine.
Are there bearings between the concentric shafts, or are each shaft's bearings held to the engine's structure? At least one would would need to be attached to the structure, but which one?
I made a video on this subject a while ago....kzhead.info/sun/eLeHptF8saWFeoU/bejne.html
There seems to be significant variations on the placement location and configuration of the bearings in multi shaft turbofan engines depending upon which engine. Some engines may not use bearings between the concentric shafts. It would be an interesting video to compare the significantly different bearing configurations and there motivations. Thank you for motivating me to go down that rabbit hole.
We tested the LM2500 without the power turbine and installed it later if the water brake was not available.
Yes, we use a nozzle much like what an aircraft engine would use. Power is calculated using mass flow and jet pipe pressure rise.
Can an N1 compressor stall produce an "unloaded " effect/ damage?
Compressor stall means the compressor no longer is working, so the engine is getting too much fuel/not enough cooling air. EGT rises, which can cause damage. Pilots can help reduce a stall by lowering power setting... just what you want when climbing.
@@AgentJayZ I was thinking along the same lines, but never really studied the catastrophic effects of an unloaded turbine. Thanks, you're a great mechanics' teacher.
LOT Polish Airlines Flight 007 in 1980 and later LOT Polish Airlines Flight 5055 in 1987.
Parts of the turbine disks destroyed the control system of the aircraft in flight. in 1980 on the approach to landing, in 1987 during the increase in altitude. there was no rescue.
I need to apply for internship I’m student aerospace technology
It's amazing, the consequences of a power turbine over speeding, are so much worse than the consequences of a piston engine over revving.
They both break, but the turbine engine is usually heavier and moving faster.
The RR Trent 900 that blew it's IP turbine to pieces was on QF32. An engineering look at the failure is in this video. kzhead.info/sun/a7iKhKtoiXdqZZ8/bejne.html
All RB211 variants and all Rolls Royce Trent engines have a separate low pressure turbine that drives the fan.
Yes. In my opinion, the three spool engine is a heavier, more complicated design than the geared turbofan, but it is more mechanically reliable. What do you think?
@@AgentJayZ seeing as an RB211 holds the WORLD record for time on wing, 40,000 hours which equated to 38 round trips to THE MOON I'd say they are the most reliable engine made. And the efficiency was game changing back when it was new.
@@AgentJayZ A three-spool engine is more complicated than a geared turbofan? What about all those gears and bearings in the reduction gearbox? It's a different form of complication. with the objective of saving weight by reducing the mass and the number of stages in the LP/fan turbine, by allowing it to run faster than the fan.
@@vtxrecruiter An RB211-535 achieved 40,000 hours 'on the wing' decades ago, while I was still working for R-R - and I took early retirement 20 years ago.
@grahamj9101 Yep, last I looked, that record still stands today. Rolls still lists that engine as the world record holder on our website. Funny how the competition can't beat it with all their new engines!
Double plus ungood?
The chocolate ration has been increased from 3 to 2 squares.
Canadian expressions for bad things is VERY different from Australian expressions. Amazing differences for two different former colonies.
Orwell was English though
Turbine blade speeds as I understand are in the speed range of 9mm ammo. A 10kg lump of hard metal travelling at 9mm bullet speeds should scare the pants off anyone with a functioning brain.
A bullet weighs, what? 140 grains? What's a grain? The power turbine rotors we talked about in this vid weigh over 500 lbs. Travelling at bullet speed, that's a load of heavy duty come-aparted-ness, I tell ya.
"What's a grain" indeed 😂. Yeah I'm missing a zero I see. Stunning amount of energy.
@@AgentJayZ the conversion is 7,000 grains per pound. 500 pounds is therefore 3,500,000 grains.
Just saw Blancolirio's video on the recent 747 engine fire. It'd be interesting to see your take on it. Also, he referenced AD 2012-19-08, where apparently these engines had LP shaft defects found when an engine failed.
I just commented on the video. I think those comments go to the patreon posting. I'll copy it to the YT post.
If you want to see the actual NTSB reports that were not linked to in that video see: POWERPLANTS GROUP CHAIRMAN’S FACTUAL REPORT, NTSB ID No.: DCA12IA114 & Materials Laboratory Factual Report - Failure Analysis-Master They are extremely informative!
Hello
Double ungood.
This is your footage isn't it? Did he ask your permission to use? kzhead.info/sun/lb2jgZ2esZukh3A/bejne.html
Yes it is. No they didn't. The way YT works, it's only a few seconds and they don't care if someone steals footage. I have tried many times, and YT always takes the side of the thief. Funny that. Welcome to the future, where all the machines are better, and all the people are worse. Oh... jeez, I said that on Xmas eve.
@@AgentJayZ I'm a firm believer that if somebody steals footage from another channel without permission, even a few seconds, they should be subject to penalties, up to and including channel termination for egregious repeat offenders.
Well, I asked the mighty algorithm about it, and have no response yet. Left a salty response on the offending vid... The crappy thing is, even if a bunch of my friends were to go and like my complaint, that would only give them more views.
@@AgentJayZ it would be funny if you could copyright claim it and get all the ad revenue from the video, but the thief would likely remove monetization to screw you or they already don't get enough views to make it worth the time. But it would send a message...
@@WarrenGarabrandtI think that's worth doing
Geee-Willikers! And I do mean the gen-yoo-wine 'willikers'. p.s.Did'ja ever get the willies?
It happened during the flight of Polish airliner in 1987: en.wikipedia.org/wiki/LOT_Polish_Airlines_Flight_5055