Sign in to follow this  
StoneC0ld

FS9 engine start, how to make them spool correctly!

Recommended Posts

if you are trying to replicate the real aircraft...Find out the correct idle N1 & N2. Set the idle N1 in table 1502 CN1 vs CN2...set the N2 idle in table 1503 CN2 vs throttle.find out the N1 & N2 at standard day 16*C for engine start up.(You need to know the moment the pilot should introduce fuel.)Let's say you get 4% N1 and 22% N2 for start:Open table 1505: Corrected fuel flow vs. CN2find the farthest right line and verify your x= is at least 101 and you y= is more than x.then,at x=0 set the y= to 22.01. the y= at X=0 should NEVER BE LESS THAN N2 idle % at fuel on!!! That is where the MS team screwed up! Then make a guess and draw a gradual slope upwards from left to right starting at x=0 y=22.01 up to x=100 y=>100.The slope should start out flat with x= lower than y=. x should catch p to y near the middle and then y should catch up to x and pass x by the x=100 line.Last thing, set your N1 at N2 start speed to N1 start speed, in this example, 4% N1 at 22% N2.do a test start in the sim.Fine tune the slope to get the spool timing. If the aircraft starts without any jumpin in N1 or N2 but spools too quick, try adjust the fuel flow scalar in table 1501: spool up rate or in the aircraft CFG under fuel_flow_gain.In my aircraft I have video from the cockpit during start up so I was able to play the video and run the FS Start simultaneously. The engines spool within +/- 1 seconds and +/- 2 % N1 from fuel on to idle. No jump or spike...smooth as a baby's bottom.My spool up rate in table 1501 is set to 0.0015 My fuel_flow_gain in the aircraft CFG is set to 0.00075

Share this post


Link to post
Share on other sites
Help AVSIM continue to serve you!
Please donate today!

Hi,I really don't want to contradict you but,>if you are trying to replicate the real aircraft...>Which one? aircraft behave different on startup accoring to the type/brand/model of their engines, amongst other things.>>In my aircraft I have video from the cockpit during start up>so I was able to play the video and run the FS Start>simultaneously. The engines spool within +/- 1 seconds and>+/- 2 % N1 from fuel on to idle. No jump or spike...smooth as>a baby's bottom.>My spool up rate in table 1501 is set to 0.0015 >My fuel_flow_gain in the aircraft CFG is set to 0.00075>If your aircraft is a turbofan/turboprop, I think you're missing one critical point: engines need bleed air aid to start, and spool up time has a direct dependency on bleed pressure. FS doesn't simulate this, so in the end if you want to stick to the REAL procedure, you'll need to fake the gauge indicators for N1/N2/N3/EPR through a set of polinomials and other similar stuff. Setting the correct relationship between engine spools, fflow and spool rate as you stated is really important, but not complete IMHO. I know this very well, as I needed several months to develope a precise simulation of my RB211 engines (including all kinds of abnormal situations: hot starts, hung starts, compressor stalls, high altitude inflight starts, etc)Tom

Share this post


Link to post
Share on other sites

Tom, you missed the point.The point is most FS9 FDEs notoriously spools up erradically and unrealistically. Mine don't. That's the point. For those out there who have problems getting them to spool correctly on start up (turbo prop and turbo fan), above is the solution. The engines spool up so realistically when started properly that I don't even use a starter wave sound...I just let the interpolated engine sounds do all the work.As for your point of bleed air and engine pressure, that was beyond the scope of this post...but since you bring it up, yes FS does model bleed air pressure, but it defaults to 20 psi for start combustion and it assumes you always have an input source. It is never off or low at the defined "start" RPM but using Lvars it is easily manipulatable. For instance, in my aircraft, 32 psi are required for a clean start. I set up a variable that checks to make sure there is another air source available that is producing the default 20 psi, either the apu or the opposing engine, and scale the air gauge to display 32 psi. I could likely set up a ground started too but it is never used procedurally for my aircraft, unless the APU is broken. Lucky for me the FS bleed air is tied to CN2 which adjusts itself for temp and ambient pressure.During start up, if the system does not sense required pressure it will deny or delay combustion, all through standard A:vars. Makes it all simpler and more reliable.Hot Start:If fuel is added before air pressure and air flow reach certain levels, a "fake" temperature is added to the standard TGT. This extra temperature will take 2-3 times as long to cool. If the engine starts while in this condition, the temperature goes with it, and as in the real aircraft, a start attempt with a residual temperature of 150*C will most likely lead to an exceedance and engine shutdown.Hung Start:I have a bug in my code that I am glad to have. Didn't try to make this but, on my last flight I unintentionally hung the start. Why? It was hot and the barometric pressure was low, and I (in a complacent, FS always works fine and I am in a rush kinda way) failed to check that I had 32 psi. I had 30! it got up to 28% N2 then fell then tried to re-light then fell and I had to shut it down because the TGT was rising.Compressor Stall:I do not have enough knowledge of compressor stalls to model this so if you could explain how it happens I would appreciate it. The only of this kind I have modelled is attempting to ground start with a tail wind, and N1 is below 4%. Will lead to a compressor stall and can lead to an engine fire.High Altitude Starts:I haven't tested this one yet! Thanks for reminding me!!!!! Gonna try it now.Using these same ideas, I even set up cross bleed starts. This particualr aircraft requires over 50% N1 for cross starts and by measuring the FS9 bleed output of one engine at 49% N1 I set the minimum psi for a successful cross bleed start. Everything else works the same without gauge trickery or polynominals. Because the Bleed air is tied to CN2 ambient pressure and temperature are already factored.Now if I could figure out how to measure water ingestion!XML variables in use:(A:Turb Eng1 Bleed Air,psi)(A:Turb Eng1 Corrected N2, percent)(A:General Eng1 Mixture Lever Position, percent)(A:General Eng1 Exhaust Gas Temperature, celcius)(A:General Eng1 Fuel Valve, bool)(A:General Eng1 Combustion,bool)(A:Master Ignition Switch,bool)

Share this post


Link to post
Share on other sites

>Tom, you missed the point.>>The point is most FS9 FDEs notoriously spools up erradically>and unrealistically. Mine don't. That's the point.I didn't miss your point; only said that what you stated was true but incomplete to be as real as it gets; and not because of your fault but rather because of FS's own limitations.>>As for your point of bleed air and engine pressure, that was>beyond the scope of this post...but since you bring it up, yes>FS does model bleed air pressureFS models ENGINE bleed air, once the engine is running...but air source needed to start the first engine can only come from APU or Ground starter, and neither of them are modeled in FS9, just APU is in FSX but only limited to sound/fuel/power so far I've tested.>but it defaults to 20 psi for start combustion Where do you measure this value?? What I know is FS waits for a certain N2 value to start the combustion once fuel is added (i think is around 20 % CN2); if you add fuel immediately after open the starter, it won't fire up until that value, and then N2 rpm will increase abruptly, the same as EGT..very unrealistic. In the real a/c, RPM won't accelerate much more than normal and EGT will rise in a steady line.>>For instance, in my aircraft, 32 psi are required for a clean>start. Yes, that is a common value. However, a turbine engine will probably start with lower values, but with a noticeable delay to obtain proper N2 rotation. >During start up, if the system does not sense required>pressure it will deny or delay combustion, all through>standard A:vars. Makes it all simpler and more reliable.>And how do you sense bleed pressure, using an Lvar? then you're faking a bit, like me :-). Otherwise, if you use Engine AVars, that's unrealistic.>Hot Start:>>If fuel is added before air pressure and air flow reach>certain levels, a "fake" temperature is added to the standard>TGT. This extra temperature will take 2-3 times as long to>cool. If the engine starts while in this condition, the>temperature goes with it, and as in the real aircraft, a start>attempt with a residual temperature of 150*C will most likely>lead to an exceedance and engine shutdown.If you aborted a start with an EGT/TGT near limits, there is no problem to redo the start process immediately (in fact it is done in real cases); crancking the starter is the best way to lower EGT temperature fast, then when reaching 110-150 C you can put the fuel on and go for a new try. THis is not simulated in FS.>Hung Start:>>I have a bug in my code that I am glad to have. Didn't try to>make this but, on my last flight I unintentionally hung the>start. Why? It was hot and the barometric pressure was low,>and I (in a complacent, FS always works fine and I am in a>rush kinda way) failed to check that I had 32 psi. I had 30!>it got up to 28% N2 then fell then tried to re-light then fell>and I had to shut it down because the TGT was rising.>With 30 psi instead of 32 the engine should start in ALL cases. Is a fact in real aircraft. Probably with a slower acceleration and a bit than normal rise in EGT, but that's all. Hung start would happen whenever engine combustion occurs, engine starts to accelelate towards idle speed and you (or a failure) cut the starter on beforehand, removing bleed air source. The engine will never reach idle speed and will stabilize at low N2 RPM, with higher than normal EGT. When you add power with the throttle, the engine will not speed up and EGT will tend to rise beyond limits. Time to shut try a restart. Not simulated in FS as well.>>Compressor Stall:>>I do not have enough knowledge of compressor stalls to model>this so if you could explain how it happens I would appreciate>it. The only of this kind I have modelled is attempting to>ground start with a tail wind, and N1 is below 4%. Will lead>to a compressor stall and can lead to an engine fire.Strong tailwinds on startup is a common source for engine surge/stall (mostly on older/non FADEC engines) It's simply the N2 spool's compression cycle becomes unstable and or disrupted, and the compressed air escapes forward instead. This produces a loud bang, sometimes a short fireball, and the N2/EPR instrument readings rise and fall abruptly. Could lead to engine damage.Something interesting to "fake": put your plane at service ceiling (40,000+feet),do an abrupt pitch up, and figure out what would probably happen with the engine instruments...>>High Altitude Starts:>>I haven't tested this one yet! Thanks for reminding me!!!!! >Gonna try it now.You need to build up speed to ensure a minimum N2 RPM (due to ram drag,usually 15% will be enough) then put fuel on and wait near two minutes for a slow acceleration and stabilization of the engine. AND be prepared to exceed the EGT limits, but who cares in these cases? So, as you see, real engine behavior is a little more complicated than the basic (but efficient) FS handling. It's up to every one how far to go and match it by code.Tom

Share this post


Link to post
Share on other sites

>FS models ENGINE bleed air, once the engine is running...but air >source needed to start the first engine can only come from APU or >Ground starter, and neither of them are modeled in FS9So how is it an unstarted engine produces 20 PSI?if that were the case we wouldn't need an air source. It produces this before start. Whether it is an input or output, we can use it either way.>Where do you measure this value?? What I know is FS waits for a >certain N2 value to start the combustion once fuel is added (i think >is around 20 % CN2); if you add fuel immediately after open the >starter, it won't fire up until that value, and then N2 rpm will >increase abruptly, the same as EGT..very unrealistic. In the real >a/c, RPM won't accelerate much more than normal and EGT will rise in >a steady line.That is a negative. The start N2 is set as I outlined above. about 19 PSI is calibrated to it. regardless of whether the engine starts at 30 N2 or 10 N2 it will have ~20 PSI of bleed air. Your theory on the way FS starts the engine is actually backwards. In my tests I controlled the fuel valve manually. I have tested the engines with valves and fuel levers on from the beginning. FS requires:16 PSI of fuel pressure18.8 PSI of bleed air115 pounds of fuel (set by thrust) and that is all regardless of N2 or fuel valve position.If you add fuel it doesn't simulate combustion until those factors are met. The reason for the horrible starts is most FDEs have the fuel and thrust set wrong. At the moment of ignition the Fuel flow in a real a/c starts at zero correct? then counts up to idle...well look at the chart:Most FDEs start at x=0 and y=0x is your fuel flow and y is your N2x needs to be zero at the point the engines will be started.In my A/c it is 25% N2 so my x=0 is y=25.01, or zero fuel flow at 25 N2. if you cut in the fuel early will continue to spool due to the starter but it will not start until reaching 25.If you start your FDE with x=0 and y=0 then at x=0.24 and y=24 your are injecting 24% of the total fuel flow into the engine at ignition. In reality you would hear a loud KABOOM followed by sirens...in FS it burns all that fuel in a huge spike.So when I hit the fuel cut on...using DEFAULT A:VARs... N2 hangs for approximately 3 seconds then gently spools from 25% to 28% in about 10 seconds then steadily gets faster and faster thru 50% where it slows down. At 60% it is almost a 0.1 percent per second creep up to 63% N2 which is idle. The great thing is this uses no tricks or faking and it also brings the rise to imbalanced thrust as the engines are no longer perfectly synced. I was even able to "start" the engines in one of my tests where it took almost five minutes for the engine to spool from 25% N2 to 30% N2. >And how do you sense bleed pressure, using an Lvar?I read it off the engine during the start sequence...the magic number 18-20 PSIMy APU runs as a 3rd engine in FS so it PRODUCES bleed air that I have to feed during starts.>If you aborted a start with an EGT/TGT near limits, there is no >problem to redo the start process immediately (in fact it is done in >real cases); crancking the starter is the best way to lower EGT >temperature fast, then when reaching 110-150 C you can put the fuel >on and go for a new try. THis is not simulated in FS.Although you are correct you explain it wrong. You cant re-start immediately if you have to dry crank to cool...lol. In my aircraft the delay between false starts is one minute dry motoring to flush fuel and aircool the engine. That's what I mean when I saw I had to restart and wait for cool. Anything below 150*C is fairly safe.>Hung start would happen >whenever engine combustion occurs, engine starts to accelelate >towards idle speed and you (or a failure) cut the starter on >beforehand, removing bleed air source. The engine will never reach >idle speed and will stabilize at low N2 RPM, with higher than normal >EGT. When you add power with the throttle, the engine will not speed >up and EGT will tend to rise beyond limits. Time to shut try a >restart. Not simulated in FS as well.You prove my point. FS tried to start the engine. EGT was rising, the engine was spooling up and hung out around 28% N2. N2 fell, rose again, and I SHUT IT DOWN. Had I waited for the APU to stabilize as called for in the procedure before pulling the start knob I would have had a clean start. What FS9 FDE have you seen that will do that? and without trickery!I have had starts that hung at 25% and started after 15 second delay, and starts that have gotten to weird numbers and fallen and risen. Of course it wont actually get to idle and be stuck there. As for airstarts, it just goes of the air pressure... at FL200 starts are occuring at 24.3% N2, 6.9% N1, and... 19 PSI...

Share this post


Link to post
Share on other sites

""So how is it an unstarted engine produces 20 PSI?""Because it is not realistic! Simply put.""That is a negative. The start N2 is set as I outlined above. about 19 PSI is calibrated to it. regardless of whether the engine starts at 30 N2 or 10 N2 it will have ~20 PSI of bleed air"" FS requires:16 PSI of fuel pressure18.8 PSI of bleed air115 pounds of fuel (set by thrust) and that is all regardless of N2 or fuel valve position.""Do I need to repeat that a real engine will be able to start with a variable range of bleed air supply,fuel flow and fuel pressure, depending on present conditions? Those fixed values enable FS to simulate startup in a generic set of engines; that is good for the sim but not real if you are modeling a specific engine, which is my case.""Most FDEs start at x=0 and y=0x is your fuel flow and y is your N2x needs to be zero at the point the engines will be started.In my A/c it is 25% N2 so my x=0 is y=25.01, or zero fuel flow at 25 N2. if you cut in the fuel early will continue to spool due to the starter but it will not start until reaching 25.If you start your FDE with x=0 and y=0 then at x=0.24 and y=24 your are injecting 24% of the total fuel flow into the engine at ignition. In reality you would hear a loud KABOOM followed by sirens...in FS it burns all that fuel in a huge spike.""Must I assume that you are putting fuel on immediately after opening the starter?? How bad :-( Fuel switch must be turn on ONLY when N2 spool rate has reached a certain number (usually marked with a magenta indicator on turbofans). Anyway, as long as N2 core rotates fuel pump is working, then if turn on fuel switch, fuel will enter and soak the combustion chamber at low RPM, also generating a dense white vapor trail from the engine's exaust, and most probably end in a hot start once the fuel ignites in combustion.I do not mean to offend you, but I think you should review the way a real engine works on startup and the parameters that are taken into account. Some of these are really difficult to understand in a whole.Tom

Share this post


Link to post
Share on other sites

you must have a language barrier because you appear to not comprehend anything i say.The fuel on from the beggining was a test to see at what point FS9 simulates the start.The variables I pointed out are constants. If you don't understand how that makes them important regardless of how unrealistic they are I can't help you.They can be manipulated however you want without complex coding and all using A vars.As far as how an engine works, there is a difference between needing to understand an engine and purposely breaking procedure in an effort to understand the sim.

Share this post


Link to post
Share on other sites

Sir,If you have ATP experience, then you must be familiar with CRM.Act that way and don't humiliate people.Jan"Beatus ille qui procul negotiis..."

Share this post


Link to post
Share on other sites

>The variables I pointed out are constants. Not to pick nits, but something is either a "variable" OR a "constant."It cannot ever be both simultaneously... ;)

Share this post


Link to post
Share on other sites

>you must have a language barrier because you appear to not>comprehend anything i say.>>The fuel on from the beggining was a test to see at what point>FS9 simulates the start.Maybe I can't comprehend you because you confuse your concepts too much. My fault anyway, and I apologize for that.>>The variables I pointed out are constants. If you don't>understand how that makes them important regardless of how>unrealistic they are I can't help you.But it was you who started the thread saying "if you are trying to replicate the real aircraft"... and was me who was trying to help you understand that those concepts pointed by you were important to the sim but not realistic in terms of actual engine behavior. That was all about, and in the end you state the same in the above paragraph, so yes, it becomes difficult to comprehend you, sorry again.>They can be manipulated however you want without complex>coding and all using A vars.Of course, and I suspect that lot of people in this forum know how to configure their engine tables to obtain what they want (I being one of them). Anyway, your findings do add another perspective; thank you for sharing that with us.>As far as how an engine works, there is a difference between>needing to understand an engine and purposely breaking>procedure in an effort to understand the sim.I don't see the point in doing an effort to understand how the sim works, without undrstanding the basic concepts of the real procedures just to help the sim behave as real as it gets.Tom

Share this post


Link to post
Share on other sites

Jan, I am not hear to humiliate. I am simply pointing out that his responses suggest he doesn't comrehend what I am saying.. My responses started with "Tom, you missed the point" and I still feel that way.If you guys think I am being negative, my apologies, but I am not at all. Bill,What I mean is the values the variables produce during engine starts are constants "at the point of ignition." That in itself is useful to know...to those who want to really know what FS is doing.Tom,We're all here to learn. All I am trying to say is every aircraft has a different system for start, and there are 10 million ways to design that, but if you want your engines to spool as the real aircraft...meaning once ignited, the spool at realistic rates of rotation with the ambient air and weather affecting it and smooth spool up...this is how. I am not trying to show anyone how to code the parameters outside of that, which is where you keep focusing. That's why I say you miss the point.I think where you get confused is when I say the variable outputs unrealistic values that can be used to get the engine to perform realistically. I just want the end product to perform realistically... which is what you want so we are on the same side.And trust me Tom you are a friend, not an enemy. You posted useful information in a civilized way for all to read, which is why I am here.

Share this post


Link to post
Share on other sites

>I don't see the point in doing an effort to understand how the>sim works, without undrstanding the basic concepts of the real>procedures just to help the sim behave as real as it gets.>Because if you understand 100% of the real thing and 0% of the sim you do more work than you need to and I don't have the time to work harder than I have to!It goes both ways. If you understand both you can do more.

Share this post


Link to post
Share on other sites

>Tom,>>We're all here to learn.Cannot agree more on this.>All I am trying to say is every>aircraft has a different system for start, and there are 10>million ways to design that, but if you want your engines to>spool as the real aircraft...meaning once ignited, the spool>at realistic rates of rotation with the ambient air and>weather affecting it and smooth spool up...this is how. Well, I have to disagree on this, not with the technical aspects of your sentences but with the mode you state them. I would prefer to read something like "this is a good way to improve startup" rather than "this is HOW a real aircrat starts". Because your method, no matter how great could be, is just one of many that other talented people use here; I for one don't use your approach and my engines start smooth as well, no matter how deep I fake the gauges in a further step. >I am not trying to show anyone how to code the parameters outside>of that, which is where you keep focusing. That's why I say>you miss the point.True, but as you were trying to show FS generic parameters as realistics when they are clearly not (some of them not even close), I tried to clarify that point. If that wasn't your intention, then probably you're right and I missed your point in the essentials. >>I think where you get confused is when I say the variable>outputs unrealistic values that can be used to get the engine>to perform realistically. I just want the end product to>perform realistically... which is what you want so we are on>the same side. Never thought that we were on different ones; maybe we have a different level of acceptation for some FS technical aspects.>And trust me Tom you are a friend, not an enemy. You posted>useful information in a civilized way for all to read, which>is why I am here.>Never could have post a line in this thread if I would consider you an enemy. I don't ever talk with enemies (don't know of having ones) but for sure can dialogue in a respectful manner with people that don't share some or all of my points of view. I see you are doing pretty much the same now, then let's keep our future exchanges this way :-)Tom

Share this post


Link to post
Share on other sites

>>It goes both ways. If you understand both you can do more.>I agree on this sentence for sure.Tom

Share this post


Link to post
Share on other sites

Would it be possible to post an example screenshot or two of the curve, and corresponding table of x/y values?I'm trying to plug this into the Dreamfleet 727, but I'm now stuck with hung starts (or whatever it's called when it just stops spooling up) at about 22-23% N2.... I'm also extremely confused with the x on the 1505 graph in pretty much all of the aircraft I checked being 0.6 or less... Including the Level-D 767 and PMDG737... I changed the scale to have the highest x=101 and scale the intermediate x values accordingly, and have had the failed starts ever since.....Any further advice, or clarification?Thanks. :)

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this