12-22-36-Inch Draft

If we're going down the four-way route at the outset in order to prove the concept, then we may as well optimise it aerodynamically. To do so based on 20mm alloy and tube-connectors, I've set the centre-section at 12" square and the frame at 36". At the same time propellers have been upped to 22" for while they produce less thrust than at 18", they do so more efficiently.

These proportions match both the airframe that flew in December of 2021 at a larger scale, beside the design registered as number 6145740 in June of the same year.

(Proven) Proof of Concept:

Realise at 04:00 a.m. or thereabouts that for a proof of concept the airframe needs only support four motors and propellers, so long as it offers the potential to up these eventually in number to eight... or even sixteen. It will after all be flying with a crash-test dummy in the first instance.

It has the benefit of practically halving the cost and while researching the options in the small hours I also find a supplier who can provide 20mm sections as well as half-metre square sheet for the centre-section.

My outlines on the lounge floor of the past two days are not wasted either, as they fix dimensions going forward to that half-metre-square centre-section, along with metre-long perimeter lengths.

One benefit of the approach is not least the fact that this outline flew successfully in December of last year, albeit in an underslung configuration:

https://www.youtube.com/watch?v=HAIYGQ-CgQ8

Re: Draft

© TELEDRONE LTD. 4th November 2022

I've re-drafted the outline so that the tubes can all be drawn from a stock 5.0 metre length. This meant that what I call the 'dog-bone' supporting a pair of motors was to become a 525mm length, and the cantilevers 675mm (excluding stop-ends).

This meant in turn that the combined length was 525 + 675 = 1200mm and a total of 4.8 metres four-off. To do this I reduced the clearance between props to a half-inch, and so in theory we could enlarge each length proportionally and it should hang together and improve the clearance.

Five-metre stock being over sixteen feet it is worth upping the length of dog-bone and cantilever combined to four, or 1220mm. To up the measurements using school math, we arrive at 537mm and 683mm respectively. A benefit of going larger is that we can always cut it back to what we see in the image if it all goes tits-up.

And whilst the centre-section here is a half-metre square, we can still use stock alloy sheet of this size to cap a slightly larger outline... the important thing being the overall width still fits inside the wheel-arches of the trailer.

Note to self therefore, 537mm the shorter and 683mm the longer (total 1220mm).

Metric... or Imperial?

© TELEDRONE LTD. 3rd November 2022

Mix-and-match really because propellers follow US practise and come in inches, while stock material here and elsewhere in the world is retailed in metric. A good example is that centre-section, which at 20" will suit a half-metre square shear-web.

I've come over a bit Jackson Pollock today and transferred sketches from an A5 pad to lining paper taped to the floor. I was interested to see that Tesla still model in clay at full-scale, a method dating back a century which auto-manufacturers still adhere to today for the final 'look and feel' of each model.

Similarly when ships like the Titanic were designed, each cross-section or 'station' would be chalked out at full-scale on the floor like this prior to their engineering.

The overall dimension here stands at 42" if you include the overhang of each motor, but discount the arc of its propeller. Folding two-blade propellers are slowly making their way up to the larger drones, so it is conceivable that an airframe like this might yet be stored in a four-foot box.

Alloy... or Carbon?

There are options in the UK to purchase 20mm pultruded carbon-fibre sections with a wall-thickness of 1.50mm to suit tube-connectors, but there are reasons not to do so:

Despite recent rises in commodities, aluminium remains almost ten times cheaper.

It can be cut to length from stock unlike preset fibre lengths brought in from Holland.

It can be re-purposed in future prototypes or else recycled.

I can drive up the road to collect and doughnuts usually do for cutting by the staff.

It comes in a range of dimensions that are easy to scale, whereas carbon does not.

You can re-work metal whereas composites are effectively single-use use plastics.

Enough said?

Re: Bar

Conventionally with the sort of hand-built commercial drones that sell for five-figure sums, the means of mounting the motor is to design a plate or outline that can be cut from carbon-fibre sheet using a laser-cutter... which is expensive.

I bolt them direct to inch-square tubular alloy as I have done generally for each of the prototypes that we have flown (including the last in December of 2021).

At the scale considered for our current effort, 20mm square alloy is a better choice, but then at the thickness it comes from the supplier (2mm) the tube-connectors that I use do not fit inside. Our world rotates round stock dimensions of one inch or 25mm.

Accordingly looking at that section superimposed on T-motor's technical drawings, we can see that the four bolt-holes for fixing U-7 motors range around a 30mm diameter circle... which means that the distance between each is rather less. In fact, at around an inch (as seen in the figure on the left) it does not at all suit direct attachment.

What I've done in the past is to rotate it through forty-five degrees and then fix the motor with two bolts directly through the spar, with the outliers pinned to angle-alloy fixed along its sides. In this case as you can see, it is a close call... whereas with U-11 motors which I used previously, the bolt-holes were arrayed around a 40mm circle.

Nonetheless it should be do-able. In fact we discovered that airframes flew perfectly safely with just two bolts ~ those within the confines of the spar ~ in use. 

Experts in the field had endless worries over this with torsion and resonance issues, none of which materialised. Whilst people with giant drones are generally patronised, on one hand I flew jet airliners that had been brought to grief by such issues, and on the other I'm aware that the Wright Brothers were also dismissed as cranks.

So as Andy Warhol once said, If people hate what you are doing then keep on doing it.

Power Play

The first choice to make when it comes to fitting battery-packs to eVTOLs is between lithium-ion and lithium-polymer. The former carry more weight, but deliver power over a longer period and are therefore the cells of choice for electric cars, aeroplanes and upcoming air taxis. The latter pack more punch and are preferred at the lower end of the scale when budgets dictate shorter flight times at lower weights altogether. They also come as flat-packs, instead of bumper packs of cylindrical cells the size of AAs.

The second choice is whether to 'bus' the batteries together, and supply the motors by a sort of spider-web array of wires from the centre-section, or to have packs address motors directly. Smaller drones are conventionally wired from a single battery in this way, but many personal air vehicles fit the batteries directly adjacent to motors, like for instance LIFT's 'Hexa' eVTOL.

An advantage of a single energy source is that it might feature one large battery-pack or a pair thereof wired in series to attain a higher voltage, instead of say eight smaller. 

Thirdly there is a choice between higher and lower voltages, with the number of cells in each pack determining the voltage in multiples of 3.7V. T-motor recommend running our motor of choice on six-cell packs (6S) like the one above, which supply around 22 of the rated 24V. Voltage determines the RPM of the motor and is thus tailored for the fixed pitch propeller to optimise thrust at expected operational values.

On balance I prefer to go down the distributed energy route, for even though a single battery failure would affect one of the eight motors, this will not be an issue because there are eight more independent motors and power sources available for recovery.

Finally we can calculate the flight endurance from the battery rating, by examining the T-motor data below. This says that the motors will be drawing around 22A at a typical setting, and a 22000mAh battery-pack like those I have used in the past would run the motor for one hour therefore. The smaller pack (2200 mAh) seen here would therefore be expected to run for a tenth of that time, or six minutes. LiPo batteries however are not run flat, however, so that the ones in your phone and laptop for example are wired to a battery-management circuit (BMC) that prevents this from happening.

Accordingly eight of the above packs (else the slightly larger at the same price) would suit the prototype, at a cost of £520. This brings the budget nearer to £3000 all in and shows why few people build large drones for fun, let alone people-carrying types. My plan this time however is to let you power up my prototype, the way Liberace used to tell people that he hoped they liked his jewellery... as they had paid for it.

(Note that the figures below are the results of tests at 25V instead of 22, although we can address this by upping the number of cells in the battery-pack with a 7S type, so that we've nearer 26V available to us.)