MR60

The purpose of this posting will be to document the building of a video capable low vibration quad with FPV/OSD that flies for 1.5 hours.

- flight path range > 15 km (2km radius)

- max speed > 30 kph (capable of handling moderate winds)

- GoPro camera/video on gimbal

- optimal designed vibration dampened electronics platform

- fits in a suitcase

- quiet (will not disturb animal life)

It will document the following in installments:

- design validation

- frame+motor mount build

- propulsion system build

- vibration optimization method, test, and analysis

- vibration dampened EP build

- electronics build

- gimbal build

- battery optimization & build

- propulsion system optimization

- flight tuning

- flight & FPV test

- video test

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

Join diydrones

Email me when people reply –

Replies

  • MR60

    This installment is on using Carbon/Balsa panels where you have point loads.

    This is a microscopic photo of a carbon sandwich panel that has been cut. In this particular case:

    -        1 carbon woven cloth ply on each side (you can see the fibers in each tow and the weave)

    -        3.5 mm of balsa end core (so the core is perpendicular to the panel surface)

    3702253852?profile=original

    Carbon Sandwich Panels are:

    -        extremely strong or its weight in the following directions:

    • Shear (a force parallel to the surface of the panel)
    • Stiffness (3-point load force perpendicular to the panel surface)

    -        strong in:

    • Peel (trying to peel the skin off of the balsa core; peel is not strong with Nomex core)
    • Crush perpendicular to the sheet (the direction of the balsa wood core fibers)

    -        weak in:

    • Crush parallel to the sheet (if only 1 ply on each side)
    • Point loads (crush perpendicular over a small area; like a bolt head w/o a washer)

    These issues can be relieved by:

    • Increasing the number of carbon plies (skin thickness).
    • Increasing the density of the core.

    But, that can add weight everywhere when it’s only needed at specific locations. Airbus and Boeing have different approaches to solving this.

    When you fly, you are flying on (floor) panels made by sandwiching Nomex core between Composite skin (carbon or glass). So what happens in a 9g sudden stop (when mass suddenly exerts 9 times its normal force) … when the plane hits something unintended?

    If the seat you were in were just bolted to the sandwich panel, all of the passengers would end up in the cockpit as the bolts sheared through the panel (remember they are weak in that direction). Pilots really don’t care to be that cozy with 300 passengers and it would be sort of a mess to untangle, which the passengers don’t really enjoy and Boeing and Airbus like to avoid.

    To solve this issue:

    • Boeing, where I tested over 10,000 inserts in various materials, went to special aluminum inserts that are bonded to the floor panel. They each hold over 3,000 pound in shear.  They are easy to install (just drill a hole and bond the insert flange to the panel) and the inserts are quite light weight.
    • Airbus, continues to use Boeing’s old method of potting the floor panel. This involves:
      • routing out a large pocket where a fastener is to be installed.
      • moving the panel to a workstation where they fill the pocket with a compound.
      • moving the panel to an oven where they cure the compound.
      • moving the panel to a workstation where they sand the compound.
      • moving the panel back to the CNC where they then drill through the compound.

    The Old Boeing/Airbus approach costs a lot more, increases lead time for panels, takes more labor, and weighs more.

    And unfortunately, until we can buy special wee aluminum inserts for our wee panels, we are stuck with the old method. So, this is how it works.

    1) Design the part and identify where

        - Clamp-up loads from fasteners exceed the point load capability of the panel.

        - 9g crash loads would shear through the panel.

        - Since the part will have to be moved around, add location holes to the part and fixture.

    2) At each fastener location, route out a pocket that is more than 3x larger than the fastener OD.

        - Route through the top skin and balsa.

        - Do not touch the skin on the underside of the part.

     3) Fill the pocket with resin.

        - Something like West System 105 Epoxy Resin + 206 hardener.

        - Try to mix the resin to minimize bubbles.

        - Fill the pocket so that surface tension creates a slight bulge.

        - The bulge is needed because the resin will soak into the balsa and skin.

    4) Cure the resin (see microscopic photo).

        - Let stand per directions.

        - Or if you have an oven, accelerate the cure. For example, with 105+206

           i.     Cure at 98F until gelled

           ii.     Cure at 120F to accelerate cure time by 4x

    3702253738?profile=original

    5)Lightly sand flat with 220 grit sandpaper (see microscopic photo)

    3702253879?profile=original

    6) Put back under the CNC and drill the fastener hole (see microscopic photo)

    3702253824?profile=original

    The size of the pocket hole determines the bolt hole shear. Triple the pocket ID to 3x the fastener OD and fastener hole shear is more than 3x stronger. Why more than?

    The following microscopic photo shows normal 1-ply carbon skin with a balsa core. Note that you can actually see through the weave and see the balsa core below.

    3702253903?profile=original

    The following microscopic photo shows what happens to the resin matrix when poured into the pocket from above.

    3702253954?profile=original

    Notice how the resin (the white is glare off of the resin) fills the gaps in the weave and spreads out over more area than just the pocket. It’s typical for a 10 mm pocket to result in a 20 mm spread of resin through the skin. It also wicks into the balsa. The result is:

    -        The skin is stronger because it has a higher epoxy content to handle compression loads

    -        The fastener hole is gripping all around the pocket.

     

    In 250 and 500 size ships, I’d be surprised if you even need to do the above.

    But get into the 1000 and above, then most definitely. For example, if the ship weighs 10 kg, then during a fun landing that should be survivable, the ship will exert a load of 90 kg on various fasteners and joints.

    Getting back to the camera ship, I am making progress. But now I'm sidetracked on long-range FPV and radio control ... don't want to lose control of the ship when trees get in the way of radio waves. So, need to solve that first because that drives the footprint of the Rx for the radio and Tx for the FPV.

    • Developer

      Interesting. I have always been a fan of foam sandwich construction. I have generally found end grain balsa to be much too heavy though, because it soaks up a lot of resin especially if it is thin sandwich. I have got best results using PVC foam ( e.g Divinycell http://www.diabgroup.com/en-GB/Products-and-services/Core-Material/...), which is available in various densities and thicknesses.

      (For best results I used a vacuum bag arrangement with the vacuum pump salvaged from a refrigerator.)

      For adding fasteners, the best option is to replace the foam at the fastener with a harder material during construction of course. It is then also possible to add some extra layers of skin where the fastener is to be placed.

      For adding fasteners after the fact, I found the fastest way is to use a hole saw. If the finish isn't important the saw one skin, then go round and saw the other and re-skin both sides. This also gives the strongest bond with less chance of voids, else saw through one skin and route out the filling. The hole saw gives a nice neat edge.

      The same hole saw can also be used to create a mount for the fastener from ply or aluminium ( or even an inner aluminium outer ply ring)which will fit right in the hole. If you want an exact fit then add 3 small packers of thin ply to centre it in the hole. You can also use stacked penny washers instead or either side of ply etc.

      But there are many ways to skin a cat. Anything that reduces the amount of carbon plate used in quads has to be a good thing, much lighter and easier than cutting solid carbon plate but of course more labour intensive, which is why I guess you dont see it in use much.

      Divinycell H
      DIAB core material with excellent strength to weight. An all-purpose grade used in multiple industries.
      • MR60

        Impressive Andy. Any photos of your panel work? 

        When you bond carbon to the PVC foam, what is the weight per area? The balsa core carbon panel is about 1.61 kg/sq-meter. 

        What parts of the ship do you make with foam core panel? I've found because of it's low peel strength that it can't be used in some places like motor mounts. Thoughts?

        • Developer

          Unfortunately I dont have photos. I havent made a copter with this technique, though that wuld be pretty simple as you could make it  as one flat piece. The nearest I have is some photos of construction of a model catamaran

          http://www.zoomworks.org/oldsite/Kwiki/uber_frame.htm

          (Th one feature uses balsa stri plank but If you click on the link on the page Click here for pictures of my earlier multihulls  (N.b this is old. The link embedded here in this reply may not work since I did a lot of whacky javascript, so you may have to try the actual link on site) Then the Catamaran called "White Light" in there is actually built from PVC foam sandwich in female moulds. It is actually 60 kg/m3 foam of 3 mm thickness IIRC. The hulls and crossbeam were various layers of 50 g glass cloth layered dependent on where strength was needed and it was all done using  a vacuum bag.  

          As regards weight. It is dependent on whether a vacuum bag is used, in which case you can be sparing with resin since the skins are forced onto the foam at high pressure and air is evacuated( vacuum bagging is somewhat involved with various layers and preparation to make sure that air can escape but the details are somewhat long winded to go into here) 

          Assuming vacum bagging, you can calculate the weights for a 1 m2 sheet as follows. Let us assume 200 g/m2 carbon cloth. 

          I would allow the same weight of resin as cloth, so for 2 skins that would be 200 * 4 ==  0.8 Kg. (You  wet out the carbon cloth on some plastic and then apply it to the foam, rather than lay it up directly on the cores)

          The PVC foam comes in standard densities. Let us assume 65 kg/m3 foam at 3 mm thickness then the foam weight is 65 * 0.003 ~= 0.2 kg , so total for 1 m of sandwich == 1 Kg.

          Note here that this is only possible with a vacuum bag. If you lay it up be hand then you need lots more resin.

          Using a vacuum bag also greatly increases peel strength over hand layup. Obviously denser foam will have better peel strength I would guess

          • MR60

            A cat! Way too cool. Your boats are wicked looking!

            Just got back a few months ago racing across the Atlantic Ocean (ARC). Our primary competition was a French team sailing a trimaran. At the start in way too heavy winds, they went flying by us. It was unreal how fast they were going. During this two week race, we stayed just slightly above the great circle route and the French went way north. The weather severely changed for a few days putting the good wind south. We ended up beating them by just 12 seconds.  

            You've opened my mind to trying foam again. It's:

            - lighter than balsa core

            - heavier than Nomex core

            - not as good in peal as Nomex and neither are even close in the peal strength of balsa

            - but foam has a prettier edge than Nomex and might be worth the extra work

            Thanks!

    • Glup!! if I continue following your posts I'm going to finish building a Jumbo :O , thank's for this detailed explanations.

      Do you use 433mhz for long-range FPV? I love it, never lost signal.

      • I hate lost signal so all my equipment looks little exagerate, for video I use 1.2 with a diversity with a patch and a dipole tested 30 kms with a friend flying far but I never fly more than 3-4 kms :) ; the patch well oriented give me the best quality video. 1.2 video and 2.4 radio is the worse combination, I loose radio signal at 300mts so I changed to 433 on my big ships and solve it.

      • MR60

        Our FAA is going to shake things up in July, so am waiting to see what frequencies we can use. Good to hear that 433 Hz works well. Was planning to go low. Hopefully between there and 950Hz.

  • MR60

    how fast do you want to go?

    • Sorry, for some reason I did not get the notice of this/your reply, and just now saw it as I was documenting your post here. The problem situation is always the same:

      Craft weight (with equipment, minus batteries) - Battery weight - (possible Payload weight) - flight time.

      An hour or more of flight time is great, who could be unsatisfied with that, eh?

      So you say "How fast do you want to go?" ???

      Assuming the same build, and a flight time of hopefully no less than 1 hour, I want MAXIMUM speed possible. - If that is unrealistic, then hopefully 45 minutes minimum?

      - I felt I was already pushing the limits by doubling the speed, lol...

      IF you are including weight of a camera, that is negotiable. I have been considering either two small cameras good enough for FPV, one facing forward, and one facing straight down, or a crude gimbal that unlike most would only switch between the two positions, not necessarily anything else or in-between.

      - OMG !!! I forgot !!! Not having been able to read this fully, I do not know if you intended your design to be capable of FPV !!! (And possibly additional electronics for it)  (Slaps forehead).

      On a side note, the more I read from your posts, the more I am considering extensive use of carbon fiber, and its scaring me! lol. Although not the highest accepted wisdom, I tend to prefer Aluminum wherever possible, especially for construction, mounting and later adaption possibilities. But from learning a bit about you through your posts, I think that any argument I can come up with for aluminum you could defeat with carbon fiber! - Except for at least one... I am pretty stuck on using aluminum for motor mounts because it can act as a heat sink, especially if designed for such, better than carbon fiber. I otherwise have had the best success in designs by following the Murphy's Law approach. If, in a crash hard enough to compromise the material's integrity, carbon fiber can get messy, whereas with aluminum, I could potentially just bend it back in place.

      - Sorry if this reply is too long...

This reply was deleted.

Activity

DIY Robocars via Twitter
RT @chr1sa: Our next @DIYRobocars virtual AI car race is next Saturday. Compete from home using the @donkey_car simulator -- no physical ca…
yesterday
DIY Robocars via Twitter
RT @RoboticMasters: Students from @Sydney_Uni working hard on improvements and changes to @donkey_car simulator. @diyrobocars @adafruit…
yesterday
DIY Robocars via Twitter
Practice virtual race this Saturday; the real thing will be on Oct 3 https://www.meetup.com/DIYRobocars/
Wednesday
DIY Robocars via Twitter
Wednesday
Derrick Davies liked lisa TDrones's profile
Wednesday
DIY Robocars via Twitter
Sep 21
DIY Robocars via Twitter
RT @SahikaGenc: AWS DeepRacer & Hot Wheels Track https://youtu.be/4H0Ei07RdR4 via @YouTube
Sep 14
DIY Robocars via Twitter
Sep 8
DIY Robocars via Twitter
RT @davsca1: We are releasing the code of our Fisher Information Field, the first dedicated map for perception-aware planning that is >10x…
Sep 8
DIY Robocars via Twitter
RT @SmallpixelCar: How this works: 1)object detection to find cones in single camera image, 30 frames/sec on @NVIDIAEmbedded Xavier. 2)comp…
Sep 8
DIY Robocars via Twitter
RT @SmallpixelCar: Use two color cones to guide the robocar. No map needed, on onsite training needed. Just place the cones and it will fol…
Sep 7
DIY Robocars via Twitter
Sep 7
DIY Robocars via Twitter
RT @roboton_io: Great to see http://roboton.io running at 60fps on the cheapest #chromebook we could find! #edtech #robotics #educat…
Sep 3
DIY Robocars via Twitter
RT @openmvcam: Crazy in-depth article about using the OpenMV Cam for Astrophotography: https://github.com/frank26080115/OpemMV-Astrophotography-Gear https://t.co/BPoK9QDEwS
Sep 3
DIY Robocars via Twitter
RT @openmvcam: Hi folks, it's finally here! Our first draft of our Arduino Interface Library is out! It works over SoftwareSerial, Hardware…
Sep 3
DIY Robocars via Twitter
RT @chr1sa: Please let them have an open API. This would be perfect for @DIYRobocars races https://twitter.com/NintendoAmerica/status/1301513099707658246
Sep 3
More…