Announcement

Collapse

PMDG Forum Rules

1) SIGN YOUR POSTS. Since 1997, we have asked users to sign their real name, first and last, to all posts in the PMDG forum. We do this in order to keep conversations personal and familiar. You took the time to be here, we want to get to know you. This is one of the few rigid rules that we enforce regularly. We do so because we feel that forums in which users must engage one another personally are generally warmer, more collegial and friendly. Posts that are unsigned will be quietly removed without comment by the moderators, so to make your life easy- we recommend enabling your forum signature so that you never need to remember. Do this by clicking the username pull-down at the top right, then selecting "User Settings." You will find the signature editor on the ACCOUNT tab, about half way down the page. Look for "Edit Post Signature." Be sure to click the "Show Signatures" box.

2) BE NICE. We are all simmers here and no matter our differences of opinion, we share a common love of aviation, computing and simulation. Treat everyone else in the forum with respect even when you disagree. If someone frustrates you, walk away from the conversation or ask for a moderator to get involved. Speaking of Moderators, they prefer not to be treated as "The Thought Police" but if any behavior infringes on the enjoyment of another user or is otherwise considered to be unacceptable in the moderator's judgment, it will be addressed in keeping with our view of ensuring that this forum remains a healthy environment for all simmers.

3) BE LAWFUL: Any behavior that infringes upon the law, such as discussion or solicitation of piracy, threats, intimidation or abuse will be handled unsympathetically by the moderators. Threats and intimidation may, at the moderator's discretion, be provided to law enforcement for handling.

4) BE FACTUAL: When you post, always be factual. Moderators will remove posts that are determined not to be factually accurate.

5) RESPECT COPYRIGHTS: Posting of copyrighted material such as flight manuals owned by Boeing or various airlines is not allowed in this forum. If you have questions related to copyrighted material, please contact a forum moderator for clarification.

6) RESPECT PMDG: We love to hear what you like about our products. We also like to hear what you think can be improved, or what isn't working. Please do tell us and we will always treat your feedback with value. Just be sure to treat the team respectfully, as they do put a significant amount of effort into building and maintaining these great simulation products for you.

7) RESPECT PMDG DEVELOPERS: All of the developers will spend some time here. Given the ratio of developers-to-users, it simply isn't possible for us to answer every post and private message individually. Please know that we do try to read everything, but developer workload is simply too high to manage personal contact with tens-of-thousands of users simultaneously. In most cases, members of the development team will stick to conversations in the forum and will not answer private messages.

8) RESPECT OTHER DEVELOPERS: PMDG has always advocated for a strong development community and we have many friends within this community. Every developer offers something unique that helps to make the simming community larger and more vibrant. We insist that you treat our friends respectfully.

9) RESPECT MODERATORS: Moderators have a tough job, and none of them enjoy having to stomp out negativity. If a moderator has to weigh in to keep a thread peaceful, please respect that effort and refrain from giving the moderator any grief.

10) If you require official support for any of our products please open a support ticket through the support portal, https://support.precisionmanuals.com

11) This forum is designed primarily as a vehicle for the PMDG development team to interact with our customers, and for customers to interact with one another in a manner that is positive, supportive and assists in the general advancement of understanding the simulation and helping to make this and future simulations better. Any other use of this forum is not permitted, including but not limited to discussion of pricing policies, business practices, forum moderating policies, advertising of non-PMDG products, promotion of events, services or products that are not approved in advance by PMDG or any other topic deemed unacceptable by any forum administrator

12) HAVE FUN: This is the whole point of it all.
See more
See less

Engine Spool Dynamics

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Engine Spool Dynamics

    Perhaps some of the real NG pilots on the forum can comment on this issue, but I have noticed that the engine behavior of the NGXu is still not correct from my understanding of jet engine behavior, namely the spool dynamics. I'm not an NG pilot but have watched countless clips of engine spool ups and I have other addons that simulate this behavior correctly (only 2 exist). When the throttles are advanced on jet engines such as the cmf56 the engines should be slow to respond in the initial N1% (from 20%-35%) and then respond rapidly to the commanded N1 from there. The behavior is incorrect on the ngxu, the engines respond linearly from 20%-40% and are too quick to spool, they should be very sluggish to respond at first, like an exponential curve. Additionally, firewalling the throttles from idle to max results in a very quick engine spool up, with no lag (around 3 seconds to max power). I have submitted a ticket regarding this because I believe this is incorrect, and I don't think the position of the throttle levers should have as much impact on the overall spool timings of the engines.

    -Mike Iacovetta
    -Mike Iacovetta

    #2
    Agreed.

    It's also noticeable that in the NGX/Xu the engines take a while to settle at the commanded power setting, even above 40% N1 - they will slowly creep up to the power setting instead of reaching it quickly and abruptly stopping at the target N1. This makes it seem like the engines are sluggish in their entire range.

    The difference can easily be spotted when comparing to real world videos and other addons, and it seems to be present even in the other PMDG products (so the cause is likely in their underlying codebase shared across all their products). The behavior seems to be that of an overdampened PID controller somewhere.
    Santiago Vegega

    Comment


      #3
      Originally posted by Santi871 View Post
      Agreed.

      It's also noticeable that in the NGX/Xu the engines take a while to settle at the commanded power setting, even above 40% N1 - they will slowly creep up to the power setting instead of reaching it quickly and abruptly stopping at the target N1. This makes it seem like the engines are sluggish in their entire range.

      The difference can easily be spotted when comparing to real world videos and other addons, and it seems to be present even in the other PMDG products (so the cause is likely in their underlying codebase shared across all their products). The behavior seems to be that of an overdampened PID controller somewhere.
      Yeah, I've noticed this with the 777 and 747 as well

      -Mike Iacovetta
      -Mike Iacovetta

      Comment


        #4
        If there is some sort of verifiable data, send it to support, please.
        Kyle Rodgers
        PMDG Developer Emeritus

        Comment


          #5
          Originally posted by krodgers View Post
          If there is some sort of verifiable data, send it to support, please.
          Will send to support but don't think there is verifiable data of this sort - if anything you would know. Telling the engines guy to compare to real world videos and look for "overdampened PID" in the simulation should suffice for them to know what to look at.

          E: Also would be nice if some 737 pilots can chime in.
          Last edited by Santi871; 20Nov2019, 03:14.
          Santiago Vegega

          Comment


            #6
            Santiago,

            "Telling the engines guy to compare to real world videos and look for "overdampened PID" in the simulation should suffice for them to know what to look at."

            This made me laugh. If we do as you say, we will be reducing the complexity, depth and thoroughness of the methods we use for research and verification by a wide, wide margin.

            If I were the prickly type, I would find it to be an offensive slight against the intelligence of the very very bright individuals who build our products. So instead, I'm just going to laugh at the suggestion you offered and presume it was just a stab at being helpful, while using impressive words in hopes that it would sound smart in front of the audience.

            Robert S. Randazzo
            PMDG Simulations
            http://www.pmdg.com


            Comment


            • Santi871
              Santi871 commented
              Editing a comment
              It should suffice for you to know exactly what to investigate and find what needs fixing, if it needs fixing. Why would I go digging for hours for documentation which you already probably have, or generate a 20-page long back and forth thread about what is right or wrong when I can simply point you guys at a suspected bug and let you guys do your homework?

            #7
            Never timed it but she definitely is slower to spool from ground idle to 40% than from 40% to take off thrust.
            ATPL - Class I Instructor - Seaplane - B737 - BE1900D

            Comment


              #8


              Here's a video of the real 737-800 spooling up from idle to around 44% N1, I have overlayed the NGXu engine instrumentation showing the exact same thrust lever position indicated by the n1 command sectors to highlight the discrepancy between the actual and simulated aircraft. Look closely, you can see that there is no delay on the NGXu, the engines begin responding immediately, but then they dont get any faster, they actually slow down when reaching the commanded N1. The real aircraft behaves opposite to this, it is slow to respond initially, gains momentum and spools until around 38% N1, then quickly reaches the commanded N1. After 45%, the NGXu is pretty accurate, as the engines in both the simulation and the real aircraft respond quickly. The area that needs work is the initial spool.

              Now watch the NGXu if I bypass the step to 45%, the engines just respond immediately and reach takeoff power in around 6 seconds.

              Unfortunately I don't have a real world video to compare the second case with, as you would not just firewall the throttles like this in the real airplane. However, I am sure that the real aircraft would have several seconds of delay before spooling.

              -Mike Iacovetta
              -Mike Iacovetta

              Comment


              • Santi871
                Santi871 commented
                Editing a comment
                First video is a perfect comparison, nice! It's interesting to see there can even be some overshoot in the real deal.

              #9
              Looks like that’s settled then, Robert?
              Marc Bonaldo
              Armchair Pilot

              Comment


                #10


                Here's more. @9:28 Notice how the engines "unlock" after roughly 33% N1 and how quickly they reach 42%. They actually overshoot to 43-44% while the EGT spikes to over 600, then they go back down. Not the case in the PMDG which ever so slowly creeps up to the 44% target.
                Santiago Vegega

                Comment


                • Windshear
                  Windshear commented
                  Editing a comment
                  I see your point, but they are very fast after 40% could it be environmental conditions?

                  I have another thing that puzzles me, why does he have 38000 ft in the MCP at take off? I've seen others do it too, any explanation?

                • Santi871
                  Santi871 commented
                  Editing a comment
                  They are very fast after 40% because they are, which the PMDG bird doesn't capture and hopefully will get corrected. To sum it up, I'd say the PMDG is too quick below 40% and too slow above 40%.

                #11
                The spool behaviour really needs some tweaking, I submitted tickets about it right on release of the NGXu and I was promived that the team will look into it.

                For those simmers who want some actual reference, the FSL320's CFM engines are really pretty pretty good. I never saw any engine getting closer yet.
                Greetings,
                Emanuel Hagen

                Comment


                • Emi
                  Emi commented
                  Editing a comment
                  Surely it isn't perfect, but it's getting closest there is in flight simulation.
                  Your findings of the N1 overshooting when spooling up to 40% is something that happens regularly by the way when adjusting thrust to 40% for takeoff.
                  I never noticed it at other thrust settings, but then again I don't particularly look there this close during flights either.

                • Santi871
                  Santi871 commented
                  Editing a comment
                  Yeah, I'm no expert, but it sure looks like the engine is really pushing itself in the low N1 range to produce a sufficient response time. I'm also sure all of these things will vary to a degree depending on the engine's wear.

                • Emi
                  Emi commented
                  Editing a comment
                  They don't really vary Santi, modern EEC controlled engines are by far not as vulnerable as older ones were. At least from a pilots perpective, I'm sure there will be some engineers out there who want to lynch me for this comment.

                #12


                I submitted a ticket as well, I was told that the team is aware of the issue and will try to fix it if they can, but they are limited by the way the engines are coded. I would think the custom coding would enable them to get it right though, but i'm no developer. Just for reference here is a clip of a certain simulated airbus showing the correct spool behavior of an electronically controlled CFM56 engine, first to 45% N1, then from idle to full power. You can see the thrust lever commanded position with reference to the blue dot. Hopefully PMDG can figure out how to simulate this properly, i would love to see this behavior in the 737.

                -Mike Iacovetta
                -Mike Iacovetta

                Comment


                  #13
                  Was there any follow up to this? I've been flying the 737-800 a lot recently and noticed the same behavior, I wasn't sure if there was a difference in a 7 to an 800 but the spool is incredibly fast and not realistic. I know PMDG wants verifiable data but I've yet to fly any turbine airplane that has data on engine spool times. I've only seen mx forms on what is considered an 'out of limit's engine spool response time due to a fault.
                  Herb Barrett-King

                  Comment


                    #14
                    Certification requirements require the engine to be able to go from flight idle to rated take off thrust in 5 seconds or less.

                    I found a chart from an engineer describing EEC spool up logic. 15 seconds to go from ground idle to take off thrust if advancing straight from idle. 5 seconds to go from a stabilized 40% to take off. ~5 seconds ground idle to 40%.

                    5 seconds after touchdown she goes into ground idle. Go around would suck after this point. The procedure for touch and goes is to stand up the thrust levers to 50%, stabilize then go, because of it. Having done touch and goes, I will attest for the increased sluggishness. You feel like there is no power for a looong time and your eating up runway quickly.
                    Last edited by MeatServo; 21Jan2020, 14:31. Reason: Removed chart as its proprietary info than I'm not sure I have permission to share
                    ATPL - Class I Instructor - Seaplane - B737 - BE1900D

                    Comment


                      #15
                      Originally posted by MeatServo View Post
                      Certification requirements require the engine to be able to go from flight idle to rated take off thrust in 5 seconds or less.

                      I found a chart from an engineer describing EEC spool up logic. 15 seconds to go from ground idle to take off thrust if advancing straight from idle. 5 seconds to go from a stabilized 40% to take off. ~5 seconds ground idle to 40%.

                      5 seconds after touchdown she goes into ground idle. Go around would suck after this point. The procedure for touch and goes is to stand up the thrust levers to 50%, stabilize then go, because of it. Having done touch and goes, I will attest for the increased sluggishness. You feel like there is no power for a looong time and your eating up runway quickly.
                      I hope that data can be given to them so the NGXU can be more realistic.
                      Herb Barrett-King

                      Comment


                      • MeatServo
                        MeatServo commented
                        Editing a comment
                        Already done. From a stable 40% N1 to take off thrust is pretty accurate, 4-5 seconds. To set a stable 40% N1 its reasonably accurate too in my opinion, maybe should be a sliver longer. However, 20% to take off thrust takes the same amount of time, 4-5 seconds. So in conclusion, If you're following proper take off technique of stable engines 40% before setting take off thrust, it's reasonably accurate. So you'd really only see it during touch and goes, botched landings after 5 seconds when she goes into ground idle, and hitting toga early.

                      #16
                      First and foremost, PMDG product line is fantastic. I have being a customer from day one and now with the NGXU and the 747-V3 but there is a Trent I see with developers. They are very hard to take criticism, I have seeing that with FSLABS and PMDG. Robert replay in a little concerning and reminds me to an old boss I used to have, sadly now I'm his boss. I'm not a pilot but, I'm surrounded by them some of them evening flying A380 and such. When it comes to PMDG 737 NGXU spool up time needs some twists, indeed it does. It's very surprising so say that the FSLABS feels more in tune and realistic. In the passed the PMDG 737 was my favorite airliner but now, mmm I believe the FSLABS A320 family it's a little ahead.

                      Comment


                        #17
                        I don‘t know if this can have an impact in n the NGXu or 747 (or other competitors‘ airplanes) but I have noticed something else:

                        I‘ve digged out my old Realair Turbine Duke and the Aerosoft Twin Otter after a really long time. Actually it was only the Turbine Duke first. It has always had a great engine model regarding beta range and the condition levers, high idle, low idle, all that stuff. But in previous versions of P3Dv4 the engines were quite responsive as turbines of that size normally are. They do need some time to spool but it‘s still pretty quick. Now in the latest version of P3D they are increeeeeeedibly slow. It‘s almost impossible to properly set take off thrust or later reduce power quickly enough to avoid overspeed. I had to rocket out of the airport with 25-30 degrees nose up to reduce the speed as the engine were sooooo slow to respond. The 737‘s development has likely been started under 4.4 (?) or at least its engine model. maybe this has had an effect on the 737 too?

                        I then reinstalled the Twin Otter to check its not that bad engine simulation. It is extremely slow now too. EXTREMELY slow, really. Even the FADEC simulation of the Vertx DA62 has become slower while at the same time its ground stickiness has improved a lot without any updates from the developer.

                        I‘m pretty sure something has changed with one of the latest versions or hot fixes regarding the engine responsiveness. Maybe PMDG could start there? I do agree that for such small engines like CFM56 they FEEL a little too sluggish.


                        (BTW: to tell PMDG that they should look at their competitors, their engine model was correct, is pretty rude. I wouldn‘t react friendly on this either at all.)
                        Last edited by Ephedrin; 22Jan2020, 03:47.
                        i7-6700k, GTX 1080TI, 32GB DDR4 RAM @1600MHz, 4k

                        Comment


                          #18
                          Originally posted by Pilot53 View Post
                          Perhaps some of the real NG pilots on the forum can comment on this issue, but I have noticed that the engine behavior of the NGXu is still not correct from my understanding of jet engine behavior, namely the spool dynamics. I'm not an NG pilot but have watched countless clips of engine spool ups and I have other addons that simulate this behavior correctly (only 2 exist). When the throttles are advanced on jet engines such as the cmf56 the engines should be slow to respond in the initial N1% (from 20%-35%) and then respond rapidly to the commanded N1 from there. The behavior is incorrect on the ngxu, the engines respond linearly from 20%-40% and are too quick to spool, they should be very sluggish to respond at first, like an exponential curve. Additionally, firewalling the throttles from idle to max results in a very quick engine spool up, with no lag (around 3 seconds to max power). I have submitted a ticket regarding this because I believe this is incorrect, and I don't think the position of the throttle levers should have as much impact on the overall spool timings of the engines.

                          -Mike Iacovetta
                          Depends how fast you do it no? I mean if you watch videos of the engines, you don't know what the pilot is doing and how fast... Or slow...
                          Yours Truly
                          Boaz Fraizer

                          Comment


                          • MeatServo
                            MeatServo commented
                            Editing a comment
                            It does and doesn't. Thrust lever position is merely your commanded thrust. Thrust is not as direct as say turbo props or piston engines.

                            Simply put, the EEC and HMU controls the rate the engine accelerates to prevent surges, stalls etc. You can jam the thrust full forward from minimum in under a second without damaging the engines, surging, over temping or stalling them. It will control the rate of acceleration to protect the engine. That's actually a certification requirement. The other is from flight idle or 15% of rated take off thrust to 95% of take off thrust in 5 seconds or less.
                            Last edited by MeatServo; 22Jan2020, 17:22.

                          #19
                          Originally posted by Comando1221 View Post
                          They are very hard to take criticism,
                          Yeah it's almost not worth being on this forum anymore because properly type rated pilots offer input and are all wrong. oh well....
                          Herb Barrett-King

                          Comment


                          • MeatServo
                            MeatServo commented
                            Editing a comment
                            They just want verifiable data, and not subjective feelings and experience. For example, I feel less situationally aware flying the sim then a real airplane. Things feel like they happen faster so I think something is off. When I go and actually look into it the timing is the same. Temporal Distortion.

                            They have been really good when I bring things up in explaining the limitations they are bound by in P3D. Sometimes a compromise needs to be made. Its not airplane breaking by any means.

                          #20
                          Originally posted by Comando1221 View Post
                          First and foremost, PMDG product line is fantastic. I have being a customer from day one and now with the NGXU and the 747-V3 but there is a Trent I see with developers. They are very hard to take criticism, I have seeing that with FSLABS and PMDG. Robert replay in a little concerning and reminds me to an old boss I used to have, sadly now I'm his boss. I'm not a pilot but, I'm surrounded by them some of them evening flying A380 and such. When it comes to PMDG 737 NGXU spool up time needs some twists, indeed it does. It's very surprising so say that the FSLABS feels more in tune and realistic. In the passed the PMDG 737 was my favorite airliner but now, mmm I believe the FSLABS A320 family it's a little ahead.
                          Nameless please add a signature to your profile in order to ensure that your posts are signed, real name first and last.. You were advised/agreed to this requirement when initially establishing an account. Thank you!

                          https://forum.pmdg.com/forum/main-fo...pdated-25feb19
                          Chris Makris (Olympic260)
                          PMDG Technical Support
                          http://www.pmdg.com

                          Comment

                          Working...
                          X