Daniels geared inflatable (slowly built)

Finally finished my build after more than two years. When I started there were next to no dedicated parts for e-foils available, therefore I built most of it myself. It was fun but took too much time.
I started with a geared setup like Pacificmeister and the Hiorth brothers. As it was difficult and expensive to get a 5:1 Neugarth gear, I got a Chinese clone from Alibaba.

POD

1st iteration with 3D printed motor mount and back part: (in red)

2nd iteration with lathed Parts:

![IMG_7552|375x500]

Got a milled part from China to cool the gear, also added a thrust bearing to prevent axial load to the gear.

(upload://uAU5t6I3RaBQumOGzNzleBajmpJ.jpeg)


Cut a slot in the stainless steel shaft with a dremel in a 3D printed sliding tool


Coupler with a slot, shaft with key inside. This is a quite massive coupler, which is one of the reasons my pod is so long. I also balanced the couppler with an angle grinder to minimize vibrations as it has a bit of mass. Read to many posts of broken couplers…


3d printed mast clamp with cable glands, thread is 3D printed an cut into the part. Part is coated with 2k polyurethane floor coating to waterproof. There are 2 rubber seals on both sides of the pod tube.
g)

To prevent damage to the O-rings from the fixing holes in the tube while sliding the tube over, I built a tool with screws that fill the hole in the tube during the installation.

Inflatable board

Got this board from ali:


it ist 160x65x12cm, the opening is aprox. 70x30cm

Electronics and battery box

It was not easy to build a box, first attempt failed as I missed the point that the hole in the board gets bigger when you inflate it.


Fail! Can’t be put into the opening as one piece…

Next attempt worked out:
A box for the batteries and electronics, connected to a top plate that builds the lid, 8" hatch to access the batteries. separate ESC box in a wet area above the mast.

Mould from coated chip board and 3D printed parts.

Fiberglass sandwich construction with 3 Layers of 280g fiberglass, 5mm sandwich core from foam and 3d printed parts, 3 layers on top.
M4 inserts to fix stuff inside. Inserts were hold in place with 3d printed rails. Threads were filled with “coltogum” seal and drilled out after curing (where needed).

Vacuum bagging, at the top is a foamed PVC frame to get a flat and smooth surface (for proper sealing with the top plate).


End result is a relatively light and stable box.

After some iterations of filling with epoxy filler and sanding, coated with pigmented 2k polyurethane floor coating:

Cable glands screwed into 3D printed ASA-X insert with threads, which were glassed into the core of the sandwich.

Top and bottom plates
The idea was to build a CFK sandwhich for the top plate and a 2 part bottom plate consisting of a 4mm aluminium plate for the rear bottom part and a CFK board for the front. Top plate: 2 Layers of 400g CFK on each side, 5mm foam core, 3d printed edges and inserts, full 5mm CFK inserts for the spacers between top and bottom plates. 5mm acrylic glass window laminated into the sandwich. Wheight of the top plate is around 1.3kg which is ok but I think it could be built lighter and still be solid enough (the way I built the bottom plate with one layer of lighter CFK).


Acrylic glas window to see LED status and allow 2.4GHz signal to pass

Botom plate: lighter contruction with 5mm foam core, 1 layer CFK 400g/m2, 1 layer 160g/m2 and a light 80g/m2 fiberglass top layer. light and sturdy.

Electronics
Arduino nrf24 receiver, battery current hal sensor, GPS receiver (grey box) Dallas one wire Temp sensors for battery and ESC temp, water sensor with beeper and arduino logger There is also a magnetic lanyard safety magnet switch with a hal sensor, if the magnet plug is pulled out, throttle is set to zero. In the black box is a 100A contactor and an anti spark circuit with arduino and relays. After connecting the XT90 battery cables, power can be switched on/off
with an RF remote like a car key. This also servers as an emergency switch, if the magnetic lanyard shouldn’t work.


Lid lever to allow proper hatch sealing

Water test:

Flier ESC 240A 20S, passively cooled via alu housing and bottom plate, added some CorrosionX as @Flightjunkie suggested
power connectors are 3D printed, pwm throttle signal and temp sensor cables are “waterprof” 3 pin LED connectors, they are available with up to 5 pins on ali (" Male to Female Connector Waterproof Cable for LED")
They are quite compact and with some silicon grease added they seem to be waterproof, just make sure they have o-rings on them, some mine came without.


Update1:
added some details

15 Likes

Nice build Daniel! I have a similar gearbox but cooling is really needed for these… The heat softened my 3dprints after 30min of ride time. I’m switching to aluminium mounts right now for the same reason

Congratulations @sat_be! Looks absolutely clean and professional.

Very nice built
What kv is your motor ?

Thanks, Motor ist a SSS56104 with 420KV, so on the lower side but the prop is an “old” FR with high pitch. Max RPM without load is around 4000 (48V*420/5) with 12S LiPo dependig on the Battery level.

Good to know. Hopefully the extra aluminium cooling part also helps to keep it cool (as it adds some weight).

Very nice construction. You are talking about a machining in china, can you share their website if you are satisfied with the work done? What is the total weight of the board ?

Thanks

I know that why I asked and I have it on my geared setup 35km/h around 3100rpm so good to go

Alexandre, with “it” you mean he “old” FR prop? that was 7’’ pitch right?

Glad to hear,I had my first session yesterday and it feels powerful, I did not dare to fully pull the throttle, max was around 70%.

1 Like

Ordered from them, they produced the cooling part for the gear in aluminium, I just sent them a drawing and step file from fusion. It was very accurate. They also sent it with normal post on my request to reduce shipping cost.The 2 parts were 107$ + 15 for shipping.

I haven’t weighted my setup as I don’t have scales at home but I guess it’s around 25kg with 2 20000mAh Lipos, I’ll update as soon as I know.

1 Like

I measured at the time , I don’t remember but I think at least 9’’

1 Like

Got some scales as I was curious about the total weight myself…
It is heavier than I whish, 22.2kg without batteries, so 26.2kg with the small 16000mAh Lipos and 27.4kg with 20000mAh Lipos. With 12s12p, it would be close to 30kg. The Pod and the mast/fuselage/wing are quite heavy, there is some potential. The contactor also adds 500g. Alu plate for passive cooling weights 1.3kg.
What do other setups weight? I know some are even over 30kgs.

Did two test rides, still learning but today I managed to fly while lying on the board, lots of fun. It seems I have an issue with my duct, it is probably too soft because the propeller scraped some layers of it off. It happened twice, I replaced the duct and happened again today. I didn’t hit anything, the duct must somehow deform and then touch the prop. It is 3D printed in ASA-X. Prop is cnc’d aluminum, trailing edges are quite thin/sharp. Distance between blades an duct is around 1mm.
Any idea why this happened?

Not sure how thick your duct is, but it could deform during a changing flow while manoeuvring (up/down left/right). 1mm is quite a tight tolerance. Out of the water it is free spinning?
If you have the duct, spin prop, and by hand can very easily warp it so it touches then it is not rigid enough.
I also have a pretty thin duct and placed a layer of fibreglass on the inside and outside. Made it very rigid. My gap is about 2 to 2,5mm.

Yes it’s free spinning, duct is about 6mm at the thickest spot. I can push it so it touches the prop but it takes quite som force. The question is wether the round hull around the prop deforms or the struts are too weak and bend. I think if I do as you suggested and cover just the tube inside and outside with glass or carbon it should be stiff enough. I guess its try and error.
Or print it in different material like petg or some carbon reinforced material.

I also put a layer of carbon on the outside of the duct and put 2 layers of fiberglass on the fins to stiffen them. This reinforced the duct well. I haven’t been able to test it in water yet.

To strengthen the ribs of the duct I have printed them hollow, with a through hole of 3.3mm. In this hole I’ve epoxied an carbon rod of 3mm, huge difference in strength. Next duct I will do the same and make the ribs thinner.

Thanks for your suggestions, I’ll try to cover the outside with carbon and put 3mm carbon rods into the struts. Wasn’t that easy to get those holes into the struts in fusion (still easier than drilling them out).
It’s printing so I hope for a good result and that I don’t break the struts when removing the support.

Nice, good luck with the mods after printing