Problems with Ducted fan design with standard PIDs

Hi Guys,

  So I have been working for a couple (few) months an this design it was a long held beleif of mine that ducted propellers would only have a detrimental effect on flight efficency. After discussing the dynamics 

3691177309?profile=original

of a ducted fan and its effects on efficency I became aware of the potential for such a design. One of the largest hurdles was to create a duct design that would be light, durable and hold up to the forces applied during flight. We wanted to have a minimal prop/duct clearance but avoid any contact. 

  A carbon fiber design was the only logical solution. For all four props we were able to maintain the same or less weight. Also each duct was designed to hold two motors each. The clearance is <= 1mm. 

We just had our maiden a couple days ago and I would say initial results were not bad. There is a lot of work to get the efficiency gains hoped for. 

3691177224?profile=original I am experiencing some problems that don't indicate a clear solution for me so I thought I would reach out for some insight from others. 

On takeoff the aircraft seems to be fighting with itself. There are no clear indications of high or low PIDs such as jittering or sluggishness. What I notice is a constant yaw correction from the controller. At some points during the flight I will be attempting to correct or maintain a heading and the aircraft is steadily rotating regardless of my input. It also seems to have a noticable problem with altitude control. 

I suspect there is some vibrations that could be aggravating the problems. I have looked at the accx, accy, accz values and noticed they did not look normal either. I noticed previous vibration readings in log files fell more in line with peaks and valleys. Here and more pronounced in previous tests the accx and accy are seperated quite a bit. 

3691177277?profile=original

I will get some video tomorrow and post it to give a better idea of what i am attempting to correct. I have attached a test flight log that might contain more info. Currently param is default and works well on a non ducted quad with the same core components and weight. I am also getting a lot of errors like alt disparity, acc error. The horizon view also seems to keep getting out of level after many acc calibrations that seem to fix the issue temporarily.

2015-02-03 14-31-58.log

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –

Replies

  • Moderator

    Wow that's a great project.  How were you able to mold the ducts from carbon fiber?  

    • Thanks Bill,

        We have been working on this design for several months and the processes of manufacturing the duct that was light weight, but able to hold its shape under the stresses encountered during flight with the propeller tolerance required to get higher efficiency. Short answer is a combination of positive and negative forms and a lot of experience with carbon fiber fabrication We have partnered with someone with two decades of experience fabricating  high performance carbon fiber components. We still have a way to go but this first iteration shows a lot of promise. Our next version should reduce the weight of the ducts significantly as well as improvements to the design of the body to improve stiffness and also reduce weight.

        I attached a video of the maiden.

    • I noticed in your video the aircraft seems to have a greater angle or pitch when in forward flight, I noticed the same thing in my testing.  I also notice the effects of the wind were greater then when flown without a shroud, mine was very difficult to fly in windy conditions .  Like driving a bus compared to a sports car, also seems to have to push harder in forward flight.

    • Relative to its predecessor (Mk 09 non ducted)  at the same angle the speed was much greater. Haven't tested in significant wind, but I suspect from initial tests wind drag is going to present a factor to be considered. How much is yet to be determined. 

    • Moderator

      Great video.  Seemed very stable for a maiden flight.  

      I have wanted to try to build my own ducting for a while now, but I have no experience fabricating carbon fiber parts.  I was going to try to cut the duct shape out of foam and then laminate it with carbon fiber tape that I found online.  

      I've used the tape for other things and found that it was very easy to laminate to foam using epoxy as well as a laminating resin called EZE-Kote.  

      Carbon Fiber Tape

      EZE-Kote

      I don't think mine would look anything like yours though!  Have you thought about selling the ducts ? 

    • I found there were a few issues that were resolved to some extent by replacing all the avionics (pixhawk, power module) after the first maiden. So this was actually the second flight. I think there were some issues with some of the hardware in the first flight. I had to fight like mad to keep it in the air. And accelerometer errors and compass errors were constant. During this flight it still was a little squirrely, but nothing that couldn't be tuned out I think. We did see an improvement in the current draw relative to a similar configuration (open prop) weighing about 1k less.  

        Using a foam mold with the layup ,likely, would not produce an acceptable duct based on my experience. The pressure reqired to push out the extra resin is substantial. For a light,stiff, duct that is dimension ally accurate, there are not many shortcuts to this type of component. It would probably serve to protect against prop strikes, but improving efficiency and  light weight may be difficult to achieve. 

      You may find experimenting with fiberglass cloth would be a more cost effective way to work out solutions. It uses similar methods for production.

      AeroTestra is focused on developing full solutions for high quality smallUAV for fleet commercial applications. The full aircraft will be available soon (1.5-2 months) though we will not be selling the ducts alone.

    • Moderator

      Please let me know when the full aircraft will be available.   What kind of improvement in thrust have you seen with the ducting ?   The way I understand it is that in most cases the thrust increase is often negated by the additional weight of the ducting.   The prop tips also need to be close to duct as well in order to take advantage of the tip vortices.  

      Your carbon fiber ones look like it would be light enough and still have the tight tolerance required.   I still might attempt to build one just to experiment.  Right now I use CF laminated to balsa/foam just as a cover. 

    • we won't attempt to quantify the gains in efficiency until the 2nd version of the ducts are complete. The first were a proof of concept, but the resin content was relatively high so it was clear a better production method be developed to reduce resin content. Our prop to duct clearance is <+1mm. This may be able to tighten up a bit, but that I consider that to be well within good tolerance. 3702876336?profile=original

       Overall the aircraft is a bit heavier than its predecessor. But we've got it on a diet and expect a lean machine with the 2nd iteration. I mentioned the current consumption was slightly lower than its lighter counterpart, with everything else being practically identical. We  are in the process of setting up a test bench to identify ,more precisely, the gains in efficiency and tweak motor /prop combinations. We are also looking forward to mounting the aft props to see what that gets us. The motor housing and mounting system were designed to accommodate a coaxial prop setup. We  will post something up front when the design is available.

        I also thought that looked like GaGa's aircraft. 

    • 3702904212?profile=original3702904227?profile=original3702904123?profile=original3702904253?profile=original3702904278?profile=original

      I have done a lot of experimenting "proof of concept" with coax counter rotating motor/props with really great results.  I have used some off the shelf motors and converted some pancake motors.  I think they would work great with a shroud with their lower profile.

      For better mechanical stability I installed an upper bearing between the props and their shafts, much less vibration and noise.

      The major difficulty encountered was mounting and maintaining the props, the flat CF props weren't available at the time but they will much easier to work with. 

    • Thanks very much for the detailed answer Ed!

      I assume mounting props using this set up where the props have a pretty much uninterrupted air flow between them is better than putting one motor and prop on top of the motor arm mount and the other (both motor and prop) underneath, at the bottom of the motor arm mount - be it contra-rotating?

      I'm asking since making the type of set up you did is complex and a lot of effort and if very similar results can be achieved through standard placement - I'd rather just try that.

      But if there is a big difference then of course it's worth spending the extra time and effort!

This reply was deleted.