Announcement

Collapse
No announcement yet.

exceed f18 edf

Collapse

Zenm Tech Pte Ltd

Collapse

Visit Zenmtech at rc.zenmtech.com

X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    exceed f18 edf

    Hi gents,
    Been actively flying my bix3 for a while now, so last week I managed to get my f18 up in the air for the first time . Modified it with cs 12 blade 70mm ducted fan and must say the sound is really awesome. Ratio is about 1:1.

    Unfortunately after about 30 seconds , it crashed. The hand launching wasn't a problem since it flew up well with enough power.
    However the turning wasn't as expected like my bix3. It seems like when I do a left, the plane didn't turn left, just bank left.
    That's the reason why it flew quickly out of sight , getting too small to see properly , and crashed when I somehow managed to get it turn hard to port 90 degree. Reason for crash is due to it being too small to be seen to control, but still within the field area.

    My question is, when turning , what is the critical technique ? What did I miss? Anyone can offer valuable advices will be appreciated.

    Thanks !

    #2
    Sorry to hear on the crash. F18 is a beautiful plane! Is it repairable?

    But hey, that seems like a big step from flying a bixler to an EDF warplane!

    Let's go down to basics. First did you get the CG right? Secondly, have you launch it into the wind?
    It seem like it did generate enough power to maintain the straight flight path but did you have enough thrust to make the turn? Jets need room for turns and at least maintain 3/4or full throttle making the turn or it will just stall and nose down. How's the thrust power like? If power is only moderate you don't want it to stall with more angle of attack than it can handle. The hard port to 90 degree probably lost it's thrust and power for maintaining lift. In general, edf jets takes time to build up thrust and speed so you probably need to more power and speed than you actually thought it would.

    Hope it helps..

    Comment


      #3
      Thanks a lot for your kind reply bro.

      I'm not sure if it's a big step from flying a bix3 to an EDF, since I'm not very keen on any other prop planes (on designs, behaviour etc). It's the trainer to train, and ultimately the EDF is the next destination.
      You may be glad to know this exceed F18 survived the crash, with the fuselage repaired. The foam is really of extremely good material, doubt it's normal EPO. Somehow it feels a bit rubbery, very robust, and easily bend back to it's original shape without needing boiling water technique.

      CG is correct, and launched into the wind with 4/5 throttle, about 15 degree upwards. I didn't throw hard, just sort of toss into the wind.
      Right after launch, the plane went shoot straight upwards. Thrust to weight ratio is 1:1 with proper tuning at home, so I guess that's how it can shoot up into the sky easily.

      I did however reduced the throttle when it went a little too far away of sight. That's where I can't tell whether it's flying towards or away from me. Panic button was pressed ("when in doubt, throttle out"), but I guess the same theory don't quite apply to EDF jets.

      You could be right that I need to maintain more power when turning, but this conflicted with the trainings I had with the trainer plane. Will bear this in mind when trying to turn this F-18 in future.

      The question now is, do I need to give it a slight pull up when turning?
      ie:
      left turn, when plane bank left to desired angle, move stick back to middle, and pull up.
      OR
      left turn, when plane bank left to desired angle, immediately pull up while holding that left position.

      Which one do you think is correct?




      Originally posted by ebpwu View Post
      Sorry to hear on the crash. F18 is a beautiful plane! Is it repairable?

      But hey, that seems like a big step from flying a bixler to an EDF warplane!

      Let's go down to basics. First did you get the CG right? Secondly, have you launch it into the wind?
      It seem like it did generate enough power to maintain the straight flight path but did you have enough thrust to make the turn? Jets need room for turns and at least maintain 3/4or full throttle making the turn or it will just stall and nose down. How's the thrust power like? If power is only moderate you don't want it to stall with more angle of attack than it can handle. The hard port to 90 degree probably lost it's thrust and power for maintaining lift. In general, edf jets takes time to build up thrust and speed so you probably need to more power and speed than you actually thought it would.

      Hope it helps..

      Comment


        #4
        Hi, looks like you'll to pull elevator on a turn. My concept-x edf has a similar behaviour. When rolled to vertical, it just keeps going (like knife edge), need to pull g like a fighter jet to bring the turn through. Have fun!

        Comment


          #5
          Thanks a lot bro for your advise, will try this then.
          Just hope my instinct won't kick in and throttle out when panic sets in!

          Btw I really wanna see the top of my plane when it's cruising. And no I'm not talking about inverted flight. Will start another thread to discuss and share this video once the project is stabilised.

          Cheers


          Originally posted by remauled View Post
          Hi, looks like you'll to pull elevator on a turn. My concept-x edf has a similar behaviour. When rolled to vertical, it just keeps going (like knife edge), need to pull g like a fighter jet to bring the turn through. Have fun!

          Comment


            #6
            Originally posted by xcfufu View Post
            Thanks a lot bro for your advise, will try this then.
            Just hope my instinct won't kick in and throttle out when panic sets in!

            Btw I really wanna see the top of my plane when it's cruising. And no I'm not talking about inverted flight. Will start another thread to discuss and share this video once the project is stabilised.

            Cheers
            That should be easy. Ask someone with a fast fpv quad or tricopter to chase your plane.

            Comment


              #7
              Originally posted by xcfufu View Post
              Thanks a lot for your kind reply bro.

              I'm not sure if it's a big step from flying a bix3 to an EDF, since I'm not very keen on any other prop planes (on designs, behaviour etc). It's the trainer to train, and ultimately the EDF is the next destination.
              You may be glad to know this exceed F18 survived the crash, with the fuselage repaired. The foam is really of extremely good material, doubt it's normal EPO. Somehow it feels a bit rubbery, very robust, and easily bend back to it's original shape without needing boiling water technique.

              CG is correct, and launched into the wind with 4/5 throttle, about 15 degree upwards. I didn't throw hard, just sort of toss into the wind.
              Right after launch, the plane went shoot straight upwards. Thrust to weight ratio is 1:1 with proper tuning at home, so I guess that's how it can shoot up into the sky easily.

              I did however reduced the throttle when it went a little too far away of sight. That's where I can't tell whether it's flying towards or away from me. Panic button was pressed ("when in doubt, throttle out"), but I guess the same theory don't quite apply to EDF jets.

              You could be right that I need to maintain more power when turning, but this conflicted with the trainings I had with the trainer plane. Will bear this in mind when trying to turn this F-18 in future.

              The question now is, do I need to give it a slight pull up when turning?
              ie:
              left turn, when plane bank left to desired angle, move stick back to middle, and pull up.
              OR
              left turn, when plane bank left to desired angle, immediately pull up while holding that left position.

              Which one do you think is correct?

              Hi xcfufu,


              I believe there is no right or wrong technique in making a turn. It is up to an individual's preference. Some people like to use aileron to turn, some prefer to use rudder to turn. A minority even use mixing to more the turn more efficient. For me, if I am flying an edf, I would like to input the stick with a little up elevator (nose up) and climb with a slight increase in throttle before making a turn. This provides me with a safety allowance in case it stalls so that I have time to recover.

              Comment


                #8
                Thanks ebpwu, your advise is good. Will try this technique. Not sure how my f18 will react to this method but it should work.

                Hi Paul,
                My house here has a long drain beside the field. Drain is probably 3m deep, runs along one side of the field towards a 'mangrove' type of area.
                Goal is to navigate the plane to fly as close to the bottom of the drain as possible inside towards the 'mangrove' area, so when i'm standing beside the drain, i can see the plane underneath me as bystander mode.
                The plane will fly pass with the 'whooshing' sound, towards the 'mangrove', fly underneath a small bridge, passing the bridge, before making a 90 degree full up, do a half cuban 8, and fly back to me within the drain again.
                Such may be achieved by Ardupilot, and have real time video feedback all actions to be via the cockpit camera.
                Once within my range (and by this time the plane should have drained out 80% of the battery power), i will regain control and try to land it in the field again.

                This routine estimated to take about 5 mins top.

                Comment


                  #9
                  Originally posted by xcfufu View Post
                  Thanks ebpwu, your advise is good. Will try this technique. Not sure how my f18 will react to this method but it should work.

                  Hi Paul,
                  My house here has a long drain beside the field. Drain is probably 3m deep, runs along one side of the field towards a 'mangrove' type of area.
                  Goal is to navigate the plane to fly as close to the bottom of the drain as possible inside towards the 'mangrove' area, so when i'm standing beside the drain, i can see the plane underneath me as bystander mode.
                  The plane will fly pass with the 'whooshing' sound, towards the 'mangrove', fly underneath a small bridge, passing the bridge, before making a 90 degree full up, do a half cuban 8, and fly back to me within the drain again.
                  Such may be achieved by Ardupilot, and have real time video feedback all actions to be via the cockpit camera.
                  Once within my range (and by this time the plane should have drained out 80% of the battery power), i will regain control and try to land it in the field again.

                  This routine estimated to take about 5 mins top.
                  You have a much higher chance of pulling that off flying in manual or at least fly-by-wire mode than using waypoints. Ardupilot isn't *that* precise and a slight delay would literally put your plane in the drain.

                  Comment


                    #10
                    The question now is, do I need to give it a slight pull up when turning?
                    ie:
                    left turn, when plane bank left to desired angle, move stick back to middle, and pull up.
                    OR
                    left turn, when plane bank left to desired angle, immediately pull up while holding that left position.

                    Which one do you think is correct?[/QUOTE]

                    Xcfuxu,
                    There are a lot of variables which will go into executing the turn (individual airplane characteristics, wind direction, gusts, and pilot preference as ebpwu mentions). In general, it will take a combination of aileron to initiate left roll along with some elevator or even some rudder inputs to make the plane change direction. Otherwise, only using aileron will make the plane turn on it's side but it will keep going in a straight line - unlike how a bixler or sky surfer would respond.

                    Out of your two options above, you would want to give the plane some left aileron to roll to the left, return the sticks to center and then pull back on the elevator to make the plane bank through the turn. Once it comes around the turn, you would return the elevator to center and then give some right aileron to bring the plane back to level flight where you would return the stick to center. In reality, the two actions along with optional rudder are done somewhat together to make the turn appear smooth and there are many constant corrections needed throughout the maneuver that you will get a feel for over time.

                    I would be careful jumping too quickly to programming waypoints to fly low through a drain area. It sounds totally cool but any wrong moves will end up in a very frustrating experience. I recommend picking a wide open flat field to practice some simple oval patterns and some basic successful landings. The F-18 doesn't have any dihedral or self correcting tendencies, so it can be a handful to fly.

                    Good luck, take your time learning it and most importantly have fun!

                    Comment


                      #11
                      Hi Paul ,
                      Thanks for your advise , will surely try on small open area on the field first to test the accuracy and reliability of ardupilot.

                      Hi Dave,
                      Thanks for the maneuver tips. Can't wait to give it a shot.
                      Totally agree with you to try on open flat field first. That's the whole intention. I don't plan to execute this using the f18 anyways. Bix3 will be a good test bed since the speed is relatively slow, and the platform allows me to put up lots of gadgets like camera , transmitter, additional lipo, and of course the ardupilot.

                      Just a quick question for any experienced ardupilot bros. I've done some searching on this topic but can't really find a concrete answer.
                      How do you control the speed of the plane using ardupilot way points ?I mean is it even possible?
                      Understand that mission way point allows you to set gps coordinates, with abs height. However to keep up that height , how did ardupilot knows what speed to turn the motor to achieve that? If say it half throttle, and the plane somehow stalls, will ardupilot actually detect this and increase throttle?
                      or ardupilot simply full throttle all the way without any configuration on the speed ?

                      Anyone?
                      cheers.

                      Comment


                        #12
                        Originally posted by xcfufu View Post
                        Hi Paul ,
                        Thanks for your advise , will surely try on small open area on the field first to test the accuracy and reliability of ardupilot.

                        Hi Dave,
                        Thanks for the maneuver tips. Can't wait to give it a shot.
                        Totally agree with you to try on open flat field first. That's the whole intention. I don't plan to execute this using the f18 anyways. Bix3 will be a good test bed since the speed is relatively slow, and the platform allows me to put up lots of gadgets like camera , transmitter, additional lipo, and of course the ardupilot.

                        Just a quick question for any experienced ardupilot bros. I've done some searching on this topic but can't really find a concrete answer.
                        How do you control the speed of the plane using ardupilot way points ?I mean is it even possible?
                        Understand that mission way point allows you to set gps coordinates, with abs height. However to keep up that height , how did ardupilot knows what speed to turn the motor to achieve that? If say it half throttle, and the plane somehow stalls, will ardupilot actually detect this and increase throttle?
                        or ardupilot simply full throttle all the way without any configuration on the speed ?

                        Anyone?
                        cheers.
                        You need to connect an airspeed sensor to APM. You can then design your waypoint to include the intended airspeed of your plane. Obviously, you need the speed to be more than your stall speed, plus a bit more for better control. The problem with what you will be doing is altitude. Inside the drain, the baro or even GPS might not be accurate and precise enough to keep your plane just above the ground. Even high in the air APM tends to fly the plane 1-2 meters +/- the target altitude. It also doesn't perfectly follow the waypoint path, just best effort.

                        Again, what you want (flying in the drain) is very doable, I wouldn't even consider it difficult, but the you have to fly the plane in either rate (gyro-stabilized) or fly-by-wire (accelerometer-stabilized) manner.

                        Comment


                          #13
                          Note that ardupilot has issues with near-stall flight control. The primary control mechanism makes linear assumptions, and if you are operating at low speeds near stall, which may be the case in manual flight in tight places, then expect the control loop to attempt to give control input that leads to highly nonlinear response and oscillation. This generally leads to a situation that is uncontrollable.

                          Additionally, state estimation for full autonomous operations in tight places is difficult. GPS does not work well in environments with dense multipathing, such as drains. You may wish to consider some alternative solution, such as LIDAR, but the processing time and bandwidth quickly becomes an issue (imagine dealing with 100-200Mbit/s of raw streaming data that needs to be turned into a surface, and planning the path through that)

                          For the least amount of pain, I strongly suggest that you first develop and characterize the vehicle response model through all regimes of intended operation, then run tests in simulation first. Also, it would be a good idea to simulate nonlinear conditions in a realtime manner, and to learn to deal and recover from them manually. Single surface stalls can be very exciting and exceed human reaction times.

                          Comment


                            #14
                            Paul and Tim,
                            My reaction to see your expertise advices was "wow".
                            Ultimately, what I'm getting from your tips was to try ardupilot of wide area, flying 3 mistakes high, moving on the simulate on wide area before the drain path, and ensure the plane has enough buoyancy during such flights to ensure the stalling characteristics are covered.

                            Thanks!

                            Comment


                              #15
                              Originally posted by xcfufu View Post
                              Paul and Tim,
                              My reaction to see your expertise advices was "wow".
                              Ultimately, what I'm getting from your tips was to try ardupilot of wide area, flying 3 mistakes high, moving on the simulate on wide area before the drain path, and ensure the plane has enough buoyancy during such flights to ensure the stalling characteristics are covered.

                              Thanks!
                              I would not go as far as to suggest that. If the end goal is to fly in highly confined areas, then it may be worth completely modelling the effects of the confined area on navigation, and validating this model.

                              If the objective is to handfly the thing, then it might be worth getting a simulator with an accurate dynamics model that you build from your actual model, and running that until you have sufficient confidence to execute the maneuver. Then perhaps consider flying the real thing in a simulated environment that results in minimal airframe damage. A cardboard/fabric tunnel is the typical solution to this.

                              Comment

                              Working...
                              X