ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Posted (9 months 1 hour ago) and read 11574 times:
As there is quite a bit of data released on the GE9X, which will now be the exclusive engine on the 777X, I made a little analysis of the engine. Here the key data according to my fiddling with Gasturb.
How close is it to reality? I would say the biggest errors are probably in the fan parts and in the spool speeds as the fan pressure maps I had for Gasturb are standard ones and the fan of the GE9X is quite new technology with better throughput, setting new standards for a CFRP fan (ie thinner blades, more like a TI fan). This means I should have put in higher spool speeds. It does not affect any cycle values however, they should thus be fine. So overall it is probable reasonably close except for the RPMs of the spools.
First a pictures from the GE presentation from ISTAT earlier this month:
Based on these and other data I came to this first analysis:
As can be seen the engine has a 61:1 pressure ratio at the design point (GE info). This is Top of Climb for the 777-9X where the GE9X delivers some 20+ klbf to get the 777-9X to initial cruise FL330. These 61 then translates to a PR of about 52 for cruise and 54 for take off static conditions. Bypass ratio hovers around 12. TSFC at best cruise FL (370-380) and thrust (15klbf at average cruise weight of 284t) is about 0.49 lbm/lbf/hr, a very good value.
The funny thing with these high PR engines is that optimal core efficiency happens at lower turbine temps then for the engines that have PR of around 50 at ToC, optimum TSFC is here at around 1800K, 3200R so their is less of a problem with new Turbine materials and cooling etc. The work can therefore focus on using bettter materials and cooling technology to reduce cooling flows, I have cut the cooling flow to the Nozzle Guide Vanes with 50% based on the parts of it being made with CMC.
Here the working line for the engine:
As can be seen nice low TSFC all the way up to max continuous at these FL, from the table we see that climb consumption is also low, below 0.6 once past FL80 or so.
ContnlEliteCMH From United States of America, joined Mar 2005, 1442 posts, RR: 46 Reply 1, posted (8 months 4 weeks 1 day 23 hours ago) and read 11484 times:
The excellent chart you've provided shows that bypass ratio is variable. My understanding is that the geometry of the engine is unchanging. May I assume that the numbers for bypass ratio on the chart represent an estimate of air volumes at certain conditions, and not the geometry of the engine?
Christianity. Islam. Hinduism. Anthropogenic Global Warming. All are matters of faith!
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 2, posted (8 months 4 weeks 1 day 18 hours ago) and read 11310 times:
Quoting ContnlEliteCMH (Reply 1): The excellent chart you've provided shows that bypass ratio is variable
Bypass ratio of all bypass engines varies as the engines are throttled and moves from thicker to thinner air. The engine manufacturer gives us one value and we have come to believe that is it, it is the value for the engine and it stays the same, nothing could be falser, it varies all the time. Even the value they give us is given a bit arbitrarily, they can give us the value for Top of Climb ie the point where the engine is the most stressed aerodynamically or the value for max Takeoff thrust at standing still at sea level (Static SLL) where the engine is working he hardesf from a mechanical point of vue. The lighter load on the engine the higher bypass ratio and vice versa. Look at the engine throttle diagram in the graph, the bypass goes from 11.5 at max thrust to 25 at min thrust at M0.85 and 33000 feet.
If you look at the manufacturer values for e.g. the T1000 in the RR product sheet from their web site you will see that they give different values for different versions of the engine;
Thrust 53000-74000 lbf
Bypass ratio 10-11
Inlet mass flow 2400-2670 lbm/sec
In this broshyre they don't claim any pressure ratio but they have given the value of 50:1 for the T1000.
These are all the same physical engine (they only produce one version with different thrust ratings plugs, i.e. fixed derate plugs put in the engines FADEC) and the values comes from spinning it to different max RPMs. Now all the values are correct but it is a serious meal of apples and oranges as they don't tell you which value is measured where and which belong togetheer. RR is not unique in this respect, they all play this games with the journalists (and us) and we eat happily without noticing that they are not telling us a real story, they serve us numbers and we eat them, the bigger number the better engine when it comes to bypass.
So lets disect what all these values tells us. First the ones that belong together:
Most derated engine (and herefore lowest price and longest on wing life because it spins the slowest and has the lowert temps) 53000 lbf, bypass 11 and mass flow 2400 lbm/sec, hence then 74000/10/2670 for the top version. All these values are probably TO sea level static but no guarantee. The pressure ratio of 50 has nothing to do with any of these values, it is the highest value they achieve for the 74000 lbf engine at Top of Climb, the T1000 pressure ratio at TO is around 40-43, 40 for the 54klbf and 43 for the top engine. To RR defence, they are almost the only one stating Top of Climb (sometimes) when they give such a value, GE has told us pressure ratio 61 for the GE9X but not anything more like where and at what flight situation or is it installed in the nacelle or not. It must be top of climb, TO SLL would be extreem and such jumps are not done in this world, thus one who understand engines can classify it to a high probability and go on from there.
So no values in a Turbfan is fixed, it varies with thrust (the higher thrust the lower BPR, the higher PR and mass flow), altitude and also temperature. The engine manufacturers knows this of course and the fact that they tell the journalist a real saga and they don't understand that they are served an apples and oranges meal, they happily put this into a table and tell you this is how the engines stack up, B_S. The engine manufacturers don't really want to tell anyone anything of great value to protect their competitive position and to have the marketing guys to be able to serve nice big numbers.
The same goes for when they give you their fuel efficiency, which they very very seldom reveal other then a fussy X% better then some old engine. Now the big question is at what flight regime and from what version of the old engine, is it installed in the nacelle (you loose 2%) or on test bench and at what altitude and thrust setting? As you can see above the TSFC varies between 0.3 at sea level and Takeoff power up to almost 0.6 when climbing and then settles down to 0.49 at the coldest/deneses flight level of around FL380, this is all uninstalled. Then everything can vary with the variation from ISA temp but that is things at a higher level and beyond what we need to deal with here.
So no thing is very fixed in the engine world, that is why you need to understand what the values mean, put it in a model and then you an start to compare engines. The best comparison available where all manufacturers are forced to give their values and tell you what they mean is the ICAO emission databank http://easa.europa.eu/environment/edb/aircraft-engine-emissions.php but even here it is apples and oranges as the manufacturers don't have to measure at sea level, GE gives the values for 800 feet for instance, RR and PW sea level. Here one can at least read the individual datasheets where it is stipulated how the values are measured.
So a caveat is really needed like "not only your MPG can vary, just about everything can "
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 3, posted (8 months 4 weeks 1 day 17 hours ago) and read 11257 times:
Quoting ContnlEliteCMH (Reply 1): The excellent chart you've provided shows that bypass ratio is variable. My understanding is that the geometry of the engine is unchanging. May I assume that the numbers for bypass ratio on the chart represent an estimate of air volumes at certain conditions, and not the geometry of the engine?
Reading your question after having fired of this long answer I am probably also guilty of a bit apples and oranges , I think your question come from an understanding the bypass ratio is a physical value of the engine, it is not. It is defined as the ratio of the mass flow of the bypass divided with the mass flow through the core. As the core work hard this ratio goes down (proportionally more flow through the core to produce the hp to drive the fan) and when the core throttles back and have leisure time the ratio goes up .
ContnlEliteCMH From United States of America, joined Mar 2005, 1442 posts, RR: 46 Reply 4, posted (8 months 4 weeks 1 day 9 hours ago) and read 11069 times:
I found both answers excellent. I was inquiring as to whether or not the number varied because of a change in the ratio of mass flow rates, and you affirmed that this was the case. I'm ashamed to say that despite a degree in mechanical engienering and more than a passing interest in high-bypass turbofan engines, this now-obvious reality had never occurred to me.
Christianity. Islam. Hinduism. Anthropogenic Global Warming. All are matters of faith!
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 5, posted (8 months 4 weeks 1 day 3 hours ago) and read 10944 times:
Quoting ContnlEliteCMH (Reply 4): I'm ashamed to say that despite a degree in mechanical engienering and more than a passing interest in high-bypass turbofan engines, this now-obvious reality had never occurred to me.
Welcome to the club, I was there not long ago. Then I decided this should change, one book and one book level software license later I am starting to get the hang of it (all described in the TF42 thread here in Tech Ops).
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 7, posted (8 months 3 weeks 6 days 10 hours ago) and read 10524 times:
Quoting SCAT15F (Reply 6): So why is the BPR for the GE9X only 10? The RB3025 is 12, and the P&W proposal was at least 12 as well. Both also have bigger fans at 132.5" or more. GE9X is still 128"
The GE9X is also announced as 12, you will find the BPR in the 6th line in the table. As I said it varies according to how hard the core is working. The engine manufacturers can pick any part of the envelope to say this is our engines BPR, normally they pick the static seal level case at full Take Off power (5 minutes limited), in my analysis I have this at 11.8, it could also be 12, it does not change anything in the analysis as the variation in engine characteristics at BPR 12 for eg TO is rather slow, so 11.5 or 12 is essentially the same engine. A change from 4.5 to 5 does count more, past 8-10 things are flattening out.
I don't understand your point that it should not be a great engine because GE has kept the fan dia at 128'', fan dia is not a quality measure. If GE can make a BPR 12 engine at 128'' and RR at 132.5'' it shows that GE has a smaller core which together with a 128'' fan gives BPR 12, this is actually then better then what RR can do, both delivering 100klbf net thrust at TO. The core quality measure is the TSFC at cruise and secondly at climb. It seems GE was very close on the cruise TSFC (if not on the RR value) and probably also close on the climb TSFC (where 3 spools have an advantage) with a lower risk project. IMO GE is ahead in development of this evolutionary engine, RRs RB3025 was fully clean sheet.
jetlife2 From United States of America, joined Jul 2006, 217 posts, RR: 25 Reply 8, posted (8 months 3 weeks 6 days 3 hours ago) and read 10399 times:
Great work. I can assure you though that the GE9X is indeed an entirely clean sheet design. If you think about the operating cycle compared to the GE90-115B you will see that it is not an evolution. In fact I would say that there are only 3 things in common between the two engines, and they are "G", "E" and "9".
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 9, posted (8 months 3 weeks 6 days 2 hours ago) and read 10383 times:
Quoting jetlife2 (Reply 8): n fact I would say that there are only 3 things in common between the two engines, and they are "G", "E" and "9"
Hi Gareth, nice to have you back . I will do an analysis of the GE90-115B as well in this thread so we can compare. I understand that no part is the same but from a working principle it seems like a further evolution of the GE90. The most exiting part seems the further evolution of the fan where you come down to the blade thickness of the Ti fans that RR and PW has, thus increasing the throughput and still keeping the weight gains. It might be that the tip/hub ratio is increased as well and the blades more advanced profiles/curvature allowing you to go one tick higher in tip speeds (thus also helping the booster with it's blade speeds as well).
Anyway we will see the detailed cycle differences once I finish the -115B, this has a ToC PR of around 50 so there is quite some difference .
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 11, posted (8 months 3 weeks 3 days 6 hours ago) and read 10024 times:
Thanks for the link .
So a rainy day later I have the first shot of a GE90-115B analysis, it was a bit trickier then thought mainly due to it's application, the B 777-300ER. I will handle this in a new thread in order to make it easier for folks to find the different engines, it is here:
To finish off the GE9X I want to fill in something I missed, the requirements from the 777-9X (one does not construct an engine without the spec what it shall fulfill on the airframe). So here my take on the data of the 777-9X as it relates to engine sizing (I do it this time as text instead of the normal table, it might be easier to follow then what all these figures means in my engine requirements tables
Aerodynamic design point:
The engine shall be able to climb the 777-9X to at least FL330 where it will start it's first cruise segment at some 95% of MTOW (~330t). The height is set by the wings desire to operate within the acceptable Cl window of some 0.55-0.50, climb higher at this weight and it comes to close to transonic buffeting, settle lower and the wings area causes to much drag (due to the thicker air). This means minimum 17200lbf to overcome the airframe drag and another 2400lbf to keep the climb at 300ft/min minimum. This sets the level for the aerodynamics flows and compression/temps needed in the engines. In this case the engine has power to spare for this requirement (this is often the case for twins which has two other thrust gates see below), its 22klbf means the climb will finish with some 600ft/min.
The engines components are optimized to make the cruise part as efficient as possible, to the point where efficiency for Take Off and to some extent climb are sacrificed. At the cruise thrust of 18-12klbf (set by the drag of the 777-9X at FL 350 all the way to 430 as it burns of fuel and the speed, M 0.85) the compressors and turbines have their sweet spots in terms of efficiency and the temps for the stators and turbines are comfortable at some 1600K Turbine Inlet Temp (TIT or T41). At such cruise thrusts settings one can start to think about throttling back on the cooling flows like the RR TXWB does, this augments the cruise efficiency and thus improves TSFC further.
This is the most stressing moment for the engine, it operates at it's max temps and the highest mechanical loads as the sea level air density makes it produce 115klbf of thrust at close to the same RPMs as for the last part of the climb, this is almost 5 times more thrust and thus load on the engine. As the outside temp is now 20° C or more instead of -60° at ToC and the engine power is set by the difference between this temp and the TIT, the TIT is now at its all time high at 1900+K. This high temp is thus only allowed for 5 minutes until the frame has passed V2 at 35ft. Should an engine fail after V1 the single engine thrust must get the frame to a climb angle of 2.4% at this height, it means some 75klbf for the -9X. The GE9X delivers 78klbf so this is OK as well and there is probably some margin in the engine should B run into weight problems with the 777X (and my drag model for V2 is not that accurate, +-5% I would guess). This is probably what has determined the overall size of the engine rather then Top of Climb.
Single engine cruise
Should an engine fail the -9X must be able to cruise at a sufficient flight level for it's ETOPS time to clear any mountain ridges etc, the max continuous thrust at FL100 nd 200 is around 30klbf as can be seen from the climb tables, normally the single engine thrust is good for around FL150 (I do not analyze this drag vs FL in my airframe model at the moment).
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 13, posted (8 months 2 weeks 4 days 8 hours ago) and read 9554 times:
Quoting SCAT15F (Reply 12): Looks like GE has increased the GE9X fan diameter to 131.5 inches and BPR to 10.3 per Aspire Aviation...
Yes, it was originally posted by Flightglobal after a talk with the GE9X program manager, GE have increased the thrust to over 100klbf on the demand of Boeing and therefore increased the fan diameter. To drive it they need more shaft hp and have therefore increased the core mass flow a bit, it has lowered the BPR.
I also got some general insights into how to handle margins for deterioration and flat rating from jetlife2, I will adjust the model and try to give these things a first shot as well, it is learning by doing .
tortugamon From United States of America, joined Apr 2013, 2118 posts, RR: 9 Reply 14, posted (7 months 2 weeks 4 days 3 hours ago) and read 8465 times:
Quoting ferpe (Reply 13): I will adjust the model and try to give these things a first shot as well,
Looking forward to your updated analysis ferpe; its almost the end of the weekend (your time) . In another thread you mention that the increase is probably to increase range for EK. That sounds very feasible and probably likely. Is there anything to make you think that the increased fan size might lead to a slightly larger aircraft? Could they be working on stretching the model closer to 80M to the delight of many?
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 15, posted (7 months 2 weeks 3 days 3 hours ago) and read 8241 times:
Quoting tortugamon (Reply 14): Looking forward to your updated analysis ferpe; its almost the end of the weekend (your time)
It is always nice to know your work is hotly awaited . I was putting the finishing touches on the updated analysis yesterday when my new laptop with the analysis software on it decided it didn't want to find any hard-drive anymore (probably a serious virus after trying to find engine maintenance manuals on these "you want XXXX, we got it!" websites ) so here comes the analysis without the finer adjustments:
First some fundamentals (these are better described in the TF42 thread (http://www.airliners.net/aviation-forums/tech_ops/read.main/329330/):
1. Thrust is coming from the engine accelerating the mass of the airflow passing the engine faster then the airplane speed:
thrust = airflow mass * airflow overspeed
2. The funny thing is that the closer this overspeed is to the planes speed the more efficient it is, but to get any reasonable thrust out of it you then need to accelerate a lot of air = high BPR.
3. The core needs to drive a big fan to push this big mass of air out the back with this low overspeed. The power (hp) is directly related to the core airflow and how hot this is before it enters the turbines (T41). High temp = lots of shaft HP generated by the turbines at a given core airflow, at a given max T41 you need to increase the core airflow if you wnat to drive a higher fan airflow.
So GE had produced an engine concept for 100klbf which has a BPR of 12 i.e. the core needs 8% of the airflow to produce the shaft hp to drive the fan. Jointly the core and fan channel flow produce 100klbf with some 3430lb of air pushed out the back at 940ft/s (multiply the two and you get 100klbf. The overspeed is also called specific thrust and is a good measure on what type of engine one has. Engine for SR71 = very high specific thrust otherwise there is no overspeed at M3 , engine for C-series which flies at M0.78 needs less overspeed and benefits from lots of air = high BPR and low specific thrust which brings good propulsive efficiency).
So when Boeing asks for more thrust GE can either increase the overspeed (higher fan pressure ratio, ie spin the fan faster) or increase the airflow and try keep the fan PR low (good for low noise level which is directly dependent on low specific thrust). GE increased the fan diameter with 2,5'' to 131,5'' which is 4% more fan area then before. The airflow in my analysis at TO has therefore also increased with some 4% from 3430 to 3560 lb, see the tables (original engine first, stronger one below, click on the tables to see the figures better and ease the comparison as you get them in separate tabs):
For elegance sake I rated the new version at 105klbf (I have no info that this will indeed be the rating). A number of things can be seen from the tables:
I have introduced flatrating of TO, V2 and ToC thrust i.e. the thrust needed comes from the airframe model and the Turbine entry temp TET (T41) and EGT (T45) is then dictated by this thrust and the +15°C ambient temp for TO and V2 and +10°C for the climb thrust at ToC. I could not add these point to my original analysis as my computer went sic . You see that this cost us some 78K on T41 and 66K on T45 (EGT). If we assume that the design has a redline T41 at 2000K and T45 at 1100°C we can see that the original design left no room for in service deterioration.
Jetlife2 helped me with a rule of thumb for deterioration: start with 20°C and then add 5° for every 1000 cycles you want the engine to stay on wing. So with the typical 5000 cycles for a longhauler engine we need an additional 45 °C for the EGT on top of the 66°C we need for flat rating. In all we need the engine to come out of the factory with an EGT of 1055°C max at TO +15°C which is what we have with the revised analysis engine. The original run to hot, ie had a to small core, GE probably had slightly larger core (or they have higher redlline temps then I assume). The core size can be seen at the airflow, we had 267 lb/s originally and now 320 lb/s, a 20% increase. I would guess GEs increase was around 15% or so.
So we now have a larger core in relation to total flow, this means we loose about 2% in cruise TSFC at average weight ( 0.49 goes to 0.50 ). As mentioned before my original engine was a bit tight on core airflow so all in all GE has probably lost about 1% uninstalled, on an installed aiframe level it can be slightly higher as the engine dia has increased which increases engine and nacelle weight and nacelle wetted area (this is always the drawback of a larger fan dia).
With the above I have tried as best I can to describe how engine OEMs have to reason when the airframe OEM ask for more thrust. One has a certain technology level for e.g. the Turbines (I assumed 2000k T41 and 1100°C T45) and a certain max PR and component efficiency (these are coupled, see the TF42 thread). If you then need more TO thrust you have to increase airflows and a higher relative core airflow cost you TSFC.
This is just as far as I have understood it and I am still at a corse level with my understanding but as the saying goes "it is at least better then handwaving" .
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 16, posted (7 months 2 weeks 2 days 12 hours ago) and read 8062 times:
Friend of order might question why the larger airflow does not show up at the design point, Top of Climb (ToC). It is because the old table was done with T41 as a limiter and then the thrust ended up at 22klbf, for the new analysis I have another control tactic, I put the trust at the ToC as needed by the frame (20klbf) and look at the resulting T41 and T45, same tactic for TO and V2.
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 18, posted (6 months 2 weeks 2 days 3 hours ago) and read 6366 times:
Quoting cobra27 (Reply 17): How relaible is this Data, considering the engine hasn't been built yet
GE has revealed quite a bit of data on the engine, thrust, fan diameter, pressure ratio, bypass ratio and TSFC at cruise (10% lower then GE90-115B). They have also given technology levels on fan, high pressure compressor, high pressure, combustor, turbine and low pressure turbine. So the data is probably pretty close to reality, say +-5%.
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 19, posted (6 months 2 weeks 1 day 18 hours ago) and read 6292 times:
In aftermath I realize another part of your question, how can I generate this endless stream of data points? I use a professional Gas turbine analysis software, Gasturb 12. It is made by the former head of Gasturbine performance analysis at MTU, Joachim Kurzke:
I had a hard-disk crash on the computer where I had the (not very expensive for a hobbyist) licensed software so things have been a bit quiet lately but now it is running again and I am working on a comparision of the GE9X vs GE90-115B. There is also a free version on the website, it does not do all the losses but does the principal stuff, try it, you have a lot of data you can put in it from these threads (to get you going).
ferpe From France, joined Nov 2010, 2677 posts, RR: 58 Reply 21, posted (6 months 1 week 6 days 3 hours ago) and read 5931 times:
Quoting cobra27 (Reply 20): Last year I was introduced to IPSE Pro which is similar program, but used for Industrial powerplant calculations
Had a look at it, it seem more modular but therefore more demanding of it's user. Gasturb is really simple to use with good manuals and also good tutorials for the licensed version. The only thing one need to acquire over the usual parameters are the efficiencies of the fan, compressors and turbines (90% is a good starting point). There are for each type of engine a demo engine, one can leave many parameters unaltered from these like combustor efficiency, mechanical losses etc.
IMO Gasturb is a joy to use, simple enough to get going immediately but capable enough that I haven't explored more then 1/3 of the analysis capabilities after 6 months of use. And then one can show off with the 3D diagrams , like this one which shows the TSFC variation vs Altitude and TET/T41 for the GE9X:
This diagram uses a fixed engine model (happens to be the original BPR 12 version of the GE9X) and then runs it over these different altitudes and burner temps with TSFC as figure of merit, all nicely plotted in a 3D mountain like diagram. Observe that the iteration is done with burner temp (T4) but I asked the plot to be done with T41 (Stator outlet temp SOT = Turbine entry temp TET) as plot parameter.
Once again we see that the optimal fuel burn area start over FL 360 and that we need to have a T41 over 3060 R or 1700K to get it. I wrote somewhere that the optimal burner temp for low fuel consumption declines with increased pressure ratio, this is not correct, it mixes cause and indirect effect. Optimal burner temp declines with increasing component efficiency, but then high pressure ratios requires high component efficiency to make sense, so indirect high pressure ratios means the optimum burner temps decline .
Fully possible , don't think GE has cut any metal yet, they are testing the components in principal rig test. Then they can scale the engine for final spec and start detailed design. After that it will be harder to rubber stretch the engine .