Servers by jDrones

Building a Better Quad Frame


(Robert Giordano) #96

11 - Drone Frame Calculator

Its been a while since my last post but I have a LOT of stuff planned so I wanted to give everyone an update. Building frames from scratch, using a variety of materials, can be time consuming. You’ll buy some materials that seem like they could be lighter but when you cut the pieces and glue them all together, you may have only saved a few grams overall. Worse, your frame might even be heavier. Some people fill notebooks with measurements and weights of all the different parts, while others put everything into an Excel spreadsheet.

I don’t really like either approach, plus I wanted something to do the math for me…

I wanted to make a tool to help me calculate and keep track of all the weights of my frame parts.

What I thought would be a simple project, has turned into something much more complicated. Sound familiar, LOL? Anyway, I’m pretty far along with this new tool, and I’d like to share a preview of it…

Introducing the Drone Frame Calculator

I’m sure many of you have used eCalc and if you haven’t, its a great tool for calculating which motors and propellers to use for your drone. My tool serves a completely different purpose. The Drone Frame Calculator or “dCal” helps you compare different materials for building drone frames. It lets you store multiple lists of parts and weights online, so you can compare them against each other, or against your friends’ designs. And, it lets you calculate the weight of any size piece of material, if that material is listed in my database. I’m currently filling the database with hundreds of materials, from carbon fiber plate and tubing, to steel, aluminum, and nylon bolts, to plywood, balsa, foam, and carbon fiber composite sandwich materials.

An Example of dCal’s “Irregular Shape” Mode

Here’s one example of how you can use the Drone Frame Calculator:

Here’s the top plate from my H4 Alien 680mm quadcopter frame. It’s a fairly complicated shape. It would be difficult to estimate what that exact piece would weigh if it was cut from a different material. This is where dCal comes in handy.

As you can see, the original piece weighs 65 grams. Its made from 1.5mm carbon fiber plate with a matte finish on both sides. The material looks exactly like the 1.5mm carbon fiber plate I bought from GetFPV. The thickness measures exactly the same.

In dCal, I clicked the “NEW PART” button and selected “Irregular Shape” mode. I selected 1.5mm Carbon Fiber plate from the Materials drop-down menu. I entered 65g in the weight field, gave my part a name, and clicked “Save”.

Next, I made 2 copies of the part. On the first copy, I selected “1.0mm carbon fiber plate” from the materials drop-down menu and dCal showed me what the new weight would be, if I cut out the same shape from the new material.

The third example shows the weight if I cut the same piece out of 1/16" birch plywood. It would weigh less than half of the original!

More to Come…

I’m still building and testing this thing. It’s not quite ready for the public but its getting close and I wanted to share what I’ve done so far. This will be a FREE utility. I might run some Google ads to help with hosting costs but its running on my company’s server for now. I have many features planned, and I hope some will be influenced by this community, so feel free to pm me.

You can also follow my dCal blog :grinning:


(Robert Giordano) #97

Another Example

Here’s another example from my dCal app (Drone Frame Calculator). Earlier in this blog, we discussed steel bolts vs nylon and aluminum bolts. A common size bolt in many small to mid-size drone kits is M3 x 6mm. An H4 Alien quadcopter kit uses up to 96 of them, depending on the version and if you install all of the parts.

In dCal, I can just select different M3 fasteners from a drop-down list and enter the quantity. dCal instantly gives me the weight of my bolts:


(brandon macdougall) #98

There is one more option milling 3D frames subtractively. if you can cut 2d on a cnc you can cut 3D body’s breaking a arm requires you to mill another part for a few bucks and mill time once the patterns are developed. I don’t think its not in the realm of the DIY hobbyist. Some of the advantages are also low GPS interference low vibration good protection of FC and esc’s. just my 2 cents.


(Robert Giordano) #99

I agree. You can also get 3D parts printed locally now in many locations. I’ll be exploring both of these options in future posts.


(Robert Giordano) #100

More dCal Stuff to Help Build a Lighter Quad

As you might have noticed, I haven’t had a lot of time to work on this blog but I’ve been busy in the background, making tools to help accomplish the original goal - to design a better quadcopter frame, suited for DIY builders who have a limited budget and basic tools.

I’ve been adding materials and components to the database in dCal, my Drone Frame Calculator project. For example, I’ve added a fairly good selection of FPV cameras, by weight:

Visit this page here: https://design215.com/dcal/u/examples/camera-weights

I’ve also added a very simple flight time calculator when you know how many amps your drone uses to hover. The page has additional information for novice builders:

Visit this page here: https://design215.com/dcal/toolbox/flight-time-amps

I still have a lot of work to do on this project and I welcome any suggestions or requests. I’ve drawn up my next frame design and I’ll be using dCal to estimate its weight before I start building it. I’m going to be making a series of short videos demonstrating the build of this frame and hopefully it will accomplish the goal of this blog.


(U3350829) #101

I used 15x15x800mm carbon fiber square tube to build my frame.
It’s very durable,strong,and light. I build two quad 525mm and 650mm size as those pictures.


525mm quad’s weight is 1945g include 6S3P Li-ion battery, and over 60min flight time.

650mm quad’s weight is 2270g include 6S4P Li-ion battery, and over 50min flight time.
I think maybe it’s not the best frame,but it’s good for me.
For your reference.


(xbnr99) #102

Great blog, Robert! I really like the way youve based your findings on real-world tests and not relied on marketing hype or “the received wisdom”.


(Robert Giordano) #105

12 - Tubing Calculator

As part of my collection of tools designed to help everyone build better drone frames, I just launched another simple, but hopefully useful tool: https://design215.com/dcal/toolbox/tubing-calculator


UPDATE: These images reflect the new formula used for comparing torsional strength.

I’ve included some of the same information I first posted in this blog, back in April. After searching the internet and playing with many online calculators, I couldn’t find one that let me compare two types of tubing side by side, gave me the torsional constant for a thin-walled closed section AND estimated deflection, and worked well on my tablet. Also, I didn’t see any other calculator generating graphics that are perfectly to scale. I hope you like it.

I just finished this tool last night so if you find any bugs or see improvements I could make, please send me a PM. Thanks! I have a LOT more coming!!


(Chris Albertson) #106

It is actually easy to make a mold and lay down the carbon clothing resin, All using only simple hand tools.

If you are doing this at home you are not going to use an autoclave. C-clamps work and provide enough pressure for a small part like a drone body.

carbon/epoxy can be rather low-tech if you want it to be.

3D printing works too. You can print a stiff and light frame of you make the shell thick enough. Make it light be printing the in-fill at very low density, say 10% to 30%


(xbnr99) #107

Now I do :smiley: Also - how did you film the qav400 crash in episode 2? It looks as if the main camera is following the quadcopter. We’re three cameras on board?


(Robert Giordano) #108

Yes, 3 cameras. There’s a 7mm x 300mm carbon fiber tube extending backwards with a 808 Keychain camera attached to the end, hanging under it. Until I watched that camera I had no idea what was going on with my frame. I learned a lot from that experiment.


(Brad Kauffman) #109

Hi Rob,
I just found this thread and thought I would throw in a couple comments. I’m an aeronautical engineer and have been designing airplane structure for about 40 years now for a small airplane company in Seattle (Boeing). I have been messing with drones for a while and want to embark on a custom build for efficiency so I have found your posts to be excellent. The comments I would throw in are these:

  1. There has been discussion of torsional stiffness and ways to increase it using truss structures. In fact the most efficient way to increase torsional stiffness is by using a closed section. All current airplanes use a closed box section in the wing as the primary torsional stiffness. Truss work went out when stressed skin designs took over. If I understand your current build in plywood it looks like you have an open section in the center on one side and open on the other side on each end. If you cover the center section you can leave out all of the diagonal bracing and have a more torsionally stiff section. There is no need for any sort of core unless you want to make the skins so thin they lack stability.

  2. One topic I have not seen addressed (maybe it has been in other forums) is the loss of efficiency when the prop goes over the arm. The arm in essence is blocking a percentage of the prop’s thrust. The amount of blockage will be proportional to the arm’s cross sectional area and to the drag coefficient of the section. A square section has about twice the drag of a cylindrical section so this effect would be double for the same size square tube. If one was to fare in the cylinder to a symmetric airfoil shape it would drop the drag to one tenth of that of the round tube! So keeping the arms as skinny and as aerodynamic as possible should have a large impact on efficiency. Another spin off on this topic is that the prop- arm interaction will also cause vibration. There will be a pulse to the airframe as each prop blade goes over the arm. I want to build a test rig to test some of this with different arm shapes.


(Robert Giordano) #110

Brad,

Thank you so much for contributing to this discussion. This is exactly what I wanted. Your points are completely valid and I’ll address them:

  1. I completely agree. Maybe I’m not using the right terminology but much earlier in the thread, I mentioned a monocoque structure being the best and showed examples of commercial drones where the “shell” of the drone provides its stiffness. These are usually injection molded plastic frames.

    The overall theme of this thread is about building a frame from scratch without the need for special tools, exotic materials, and equipment the average person doesn’t have access to. So, I’ve been trying to design a frame that can be built by hand, with materials you can get at the local arts and craft store, using basic hand tools.

    Believe me, I’ve been trying to come up with a frame that uses a closed shell design, that would be easy to build, but they always come out heavier than my truss designs. I haven’t given up though because, a) I’d like to put all of the electronics inside the frame to protect them from the elements, and b) the entire ship would be more aerodynamic. The main reason I’ve been developing my Drone Frame Calculator is so I can calculate the weight of many different frame designs without having to actually buy, cut, and weigh all the pieces.

  2. Again, I agree but my question is, how much does that actually affect flight time and performance? I agree with keeping the arms as thin and aerodynamic as possible but the availability of materials is also a consideration. On a ship with a 20 minute flight time, do square arms cut the flight time by 2 minutes or by 20 seconds? On ships that use a box shell design, like the DJI Phantom series and the 3DR SOLO, the arms are much wider and block a considerably larger area of thrust than frames using carbon fiber tubing of any shape. With larger drones like the DJI Inspire, why do they use carbon fiber tubing with a larger diameter than would seem to make sense? I’m trying to figure than out as well. Finally, there is the DJI Mavic Pro, which uses square tubing for its arms and is advertised as “having the longest flight time of any commercial drone”. (which is only 31 minutes, lol)

    At the end you mentioned vibrations. I’ve found this gets complicated because it also has to do with the natural frequency of the components. To test some theories, I’ll be building 2 identical frames but one will use round tubing for the arms and the other will use square tubing. The square tubing is heavier and stiffer but won’t require motor mounts, while the round tubing is lighter and will have motor mounts. I will compare the vibrations and efficiency of both designs.

Finally, I’ve made some improvements to the previous design posted in this blog. The middle section will be enclosed and I’ll be testing it with and without the diagonal braces in the center, using my torsional test stand. I’ll be posting that design here soon. =)


(Brad Kauffman) #111

Hi robert,
For monocoque, if you cannot go thinner on the material a truss structure will win. Not that a true truss structure does not have any solid skins on the outside. Internal truss elements due very little for torsional stiffness, they need to make up the outer surface and enclose the box space. To make it simpler you could make a four sided box and then drill lightening holes in the unstressed area to reduce weight. I will see if I can get you a sketch of what I mean. If you want to compare torsional stiffness of various designs without having to build them and test them, I have some finite element software that can easily model box structures and apply a torque to it. I will be busy until the beginning of Oct though, after that I can provide some assistance.

My guess on why the produced designs have wide arms or poor drag characteristics is that this sport has been built primarily by controls guys, not aeronautical engineers. It has been all about the control algorithms and computing power up to now. I don’t think DJI had any aero guys, just look at the original Flamewheel arms. they are really poor from both a structural and an aerodynamic standpoint. The other point on the aero drag is about the use of the drone. The stuff I was talking about would come more into play for a slow or hovering drone like a camera platform where the prop wash is pretty much straight down. Once you add forward velocity, the airflow gets pretty confused and making an aerodynamic design would need either costly CFD analysis or wind tunnel testing.

Brad


(Brad Kauffman) #112

Here is a sketch of an efficient torque box made from flat panels cut into a sudo-truss arrangement. The torsional strength would be governed my the buckling stablility of the diagonals. The shallower the angle of the diagonals, the stiffer it will be. I didn’t know your dimensions so this box is just an example. If it buckles at too low a load rib posts could be added at each of the hubs where the members come together.


(Robert Giordano) #113

Thanks, that’s very cool. Many of the drone frame kits come with carbon fiber plates that are milled with cutouts in a very similar pattern. The problem is, these kits end up being extremely heavy because, instead of gluing the panels together like your drawing above, they include all kinds of aluminum spacers and steel bolts.

For a DIY drone frame, the following factors come into play:

  1. Availability and cost of materials. Many times, the “ideal” material isn’t available.
  2. Availability and cost of tools. Many people don’t have milling machines or 3D printers, although new online services are making these things more available at a lower cost.
  3. Knowledge of material strength, adhesives and bonding techniques, and the use of specialized tools is beyond many people who might want to build their own drone.

I’m trying to do something about number 3 while keeping 1 and 2 in mind. For example, Nomex is an amazing material but I’m not using it in any of the designs I post in this blog because its very expensive and isn’t nearly as accessible as say, 1/16" birch plywood I can buy down the street at Michael’s Arts and Crafts.

Have you seen the tools I’m building to help people (including myself) experiment with drone frames? I’d love it if you checked out the following links:

  1. https://design215.com/dcal/u/examples
    “Examples” is a user account on dCal for demostrating the app. The next two links are listed on this page but I wanted to draw specific attention to them.

  2. https://design215.com/dcal/u/examples/h4-alien-680-frame
    Here’s a commercial drone frame kit broken down piece by piece. The weights for the individual parts were not entered manually, but were calculated by the app using the entered quantity and dimensions.

  3. https://design215.com/dcal/u/examples/tubing-comparison
    This is a comparison of 4 similar sized pieces of tubing that could be used for a quadcopter. The description introduces my Tubing Calculator app and shows example screenshots. The Tubing Calculator uses HTML5 and SVG to render pixel accurate, scale models of any tubing sizes entered.


(Brad Kauffman) #114

I will take a look. My idea with the model posted is that you can make it out of 1/16" ply quite easily. For grins I 3D printed one today in 1/16" ABS. The one I printed is a 1"x3.5" box 6" long. It came in at 39g and is really stiff in torsion. I am thinking of doing a X8 frame made up of only 3D printed flat panels like this and graphite tubes for arms. When the glue dries I am going to test the one I made to failure to see the failure mode. One could easily design a frame like this that one could just print out on paper and use it as a template on 1/16" ply.

Many of the drone kits that come with truss like cutouts and spacers are terrible designs for weight. It is clear they have no engineering knowledge for weight efficient structure. The spacers have virtually no shear load path and introduce point loads in the panels where they are fastened. Usually they have the cutouts in the wrong place. Very heavy for the strength.


(cala2) #115

Brad, you printed flats parts and them glued? thank´s for share your ideas.


(Brad Kauffman) #116

That is correct. I was staying with the direction of this blog in making something from available materials and simple tools so I made the torque box as flat panels that could be cut out with a knife or saw. I was too lazy to cut them out of wood so I just printed an example. It came out stiff as hell. If you want stiff you have to have all four sides on the box.


(Brad Kauffman) #117

Hi Rob,
I took a look at the tool box you have so far and I have a couple comments.

  1. In the tube comparison you have J=2I which is correct for polar moment of inertia but it is incorrect for calculating torsional stress or deflection in a thin walled section. J for a thin walled closed section is given by: J=4A^2t/S where A is the cross sectional area at the mid line of the wall, t is the wall thickness, and S is the perimeter measured at the mid line of the wall. For the example of an internal dimension of 13mm and outside dimension of 16mm as in the example:

for square:
A=1414=196
t=1
S=14
4=56
J=4196^21/56=2744

for round:
A=14^2pi/4=153.94
t=1
S=14
pi=43.98
J=4153.94^21/43.98=2155.3

This gives a stiffness improvement of 2744/2155.3=1.273 or 27.3%, not 38.2%

It should also be noted that this applies to stiffness only. For strength calculation the sharp corners on a square section will change the failure mode and in composite may well be lower than the circular section.

  1. In the battery calculator it shows a linear relationship. The useful capacity of a Lipo battery drops as the current increases. You can increase the useful energy in a battery by discharging it slower. This is traditionally calculated with the Peukert formula. Here is a paper on the Peukert formula applied to Lipo batteries:
    http://www.mdpi.com/1996-1073/6/11/5625/pdf

The Peukert formula is C=T*I^k where C=capacity, T=discharge time, I= discharge current and k is the Peukert coefficient. From this paper it can vary from 1 to 1.3 for a Lipo. At 1.3 it is a huge effect.

Unfortunately suppliers don’t provide k (unless they are very good batteries). If you know the capacity of a battery pack at 2 different discharge rates you can calculate k. This is why you see a different size and weight for the same capacity battery pack rated at 10C versus 65C.