Testing (Phase 5)

SeaDart Test 1 (SDT1)

From the very first meeting with the client I expressed a concern as to how heavy we need to make the model so it sits at the correct waterline. To have it bobbing on the water like a bath toy just isn't acceptable. We are going to great lengths to assure scale fidelity and the sit of any model either on the ground or in the water is very important to the overall look and ultimately is part of the identity of that aircraft. We know from our research the exact waterline required which sits 0.46 meters (18 inches) below the root leading edge and flush with the back of the wing.

by Alex Jones, on Flickr

We talked at length with Fighteraces about possible solutions and collectively came up with this basic plan.

With the basic principle that one liter of water is equal to one kilo in weight we can cut the hull along the scale water line to find the volume that we need to displace. This varies slightly with fuel state but not significantly. Having designed the hull in CAD while doing the ski retraction work we could easily cut across the water line and work out the volume. Assuming we make the Ski’s neutrally buoyant we need to displace 89 litres of water, hence the model needs to weigh 89kg (196lbs). We’d like the model to weigh 60kg (132lbs) as this gives an acceptable wing loading, any more and its starting to become a handful especially with our scale aerofoil sections. The reason for having the available thrust equal to weight is that scale models don't have the energy of their full size counterparts. With excessive power we can create the impression of energy using the thrust intelligently which allows us to fly the model in a scale manner.

We are now left with a model weighing 29kgs (63lbs) more than we would like. The solution with the obvious requirement for lots of tests is to build a 29 litre wet compartment into the hull. This would be shaped to distribute the water with regard to the CG. The tank would flood once the aircraft is sitting on the water and have several large openings concealed in the formers that the front and rear ski mechanisms will be mounted to. Its easy getting the water in but we now need to evacuate it during the takeoff.

Believe it or not the full size actually took on water, in one instance the ground crew forgot to remove the stoppers in the vents and the aircraft took off with over 1000kgs of water in the hull. The difference being is that the water wasn't wanted in the case of the full size but just accumulated from the open areas at the back near the jet nozzles.

by Alex Jones, on Flickr

Conveniently the SeaDart has a large water rudder at the back which doubles as a speed brake. The area available to evacuate the water from is large when the water rudders are cracked open so the hope is by the time the model has de-planed its skis the compartments water level should be in line with the current waterline of the model. We have lots of other ideas but this is by far the simplest so starting here seems the best choice, if this doesn't work we will move to the more complex ideas…

So this brings us on nicely to floating the pattern in our SeaDart Test 1 (SDT1)

The picture below shows the pattern exactly as seen sitting in the water at 42.35kgs. As you can see it is far too light and needs to displace considerably more water to get the correct waterline.

by Alex Jones, on Flickr

by Alex Jones, on Flickr
Last edited:
Laden down with some weight.

by Alex Jones, on Flickr

by Alex Jones, on Flickr

We added 37.5 kilos of sand making the weight as pictured here 79.85 kgs but we are still an inch off the scale water line. The total volume to be displaced of the pattern is 103 litres so 103kgs. If we subtract the patterns weight we get 60.65kgs that needs to be added to achieve the correct waterline. Subtract the additional sand weight of 37.5 kilos we are left with 23.15kgs which is what we should have added over and above the sand. This tallies with our estimations of the extra weight required to submerge the hull that extra inch.

If we remove the volume of the ski wells which are submerged we can shed 14 kilos so a final flying weight with the correct waterline would in fact be 89kgs. A flying model weighing 60kgs without a wet compartment would get close but we will try and get the correct buoyancy using the solutions outlined in the above post.

Hopefully we won't need the wet compartment but we have a plan if we do.

by Alex Jones, on Flickr
You can see here clearly that the pattern weighing 42.35kg sits far to high in the water making it look unrealistic. However the flying model will not be close to this weight when you bear in mind two full tanks of fuel and two P300's including their ancillaries weighs 22kgs alone. A dry weight of between 60-70 kilos is probably more realistic for the flying model.

by Alex Jones, on Flickr
Amazing project! Your considerations and calculations about the model's weight and it's position on the water are very interesting. Of course other calcs should be done for the model in flight. As we know giant-scale models can fly well even at very high wing-loading. But going over 250 gr./dmq. could be a real problem. Do you have estimated the stall speed at various wing-loading?
One more thing: your Sea Dart project is a VERY demanding project with many very complicated aspects. I believe that building a smaller/ cheaper test-model is mandatory for the success of the BIG THING. All the best!
Thanks for your input, it's great to see people get involved. We agree and small models by other people of the SeaDart have proved to be successful, obviously depending on weight we expect it to be between 220 and 260 gr/m2 varying with fuel load. We will be conducting hours of taxi testing which should give us a sound understanding and feel for the model at higher speeds. It is a real unknown but we are doing everything we can to minimise those unknowns. We would have liked to build a smaller test model but this was discounted early on by our client although we are confident it will fly but as you say its a very demanding project!
With our thoughts turing to the logistics of operating off water we decided to use a small model seaplane in this case an S6 by:


We asked FighterAces to put it together for us and then headed to Italy where we had access to a boat and a lake where we were almost the only people around. The model is small and unobtrusive, we managed lots of flights from the boat. Both with the boat moving under power and stationary without anchor. The objective was less about the model we used and more about the operation from water and how we will operate the SeaDart. As a team we need to put together procedures and understand what is need logistically.

We had a a successful few days testing and learnt plenty with regard to our operations which we will put in practice when we do the first lake taxi tests here in the UK.

by Alex Jones, on Flickr

by Alex Jones, on Flickr

by Alex Jones, on Flickr

The model S6 flew very well and I can recommend flying off water to everyone. A very enjoyable and rewarding experience.

by Alex Jones, on Flickr

by Alex Jones, on Flickr
With a better understanding of float planes we can move into the third test (SDT3). With filming for a short documentary taking place we decided a test stand for the engines was a good idea. This will give viewers a better understanding of the systems installed into the model. Just testing the engines on a generic test rig proves that they run but does little else in relation to our bespoke setup. We have designed the tanks and tried to keep everything where possible as normal as we can but we were keen to try our complete setup. To do this we took the aircraft structure and made this into a test stand ensuring everything is in exactly the same position as it will be in the model.
SeaDart Test 3 (SDT 3)

I think its worth talking about the engine setup and how we have designed the internal structure around the two P300RXG’s

I've built plenty of jets with twin setups but all with turbines in the 140 - 160N range. This will be the first with 300N turbines which shouldn't be that much different, however, the fuel draw on the larger engines is significant at full thrust. We don't want to find out that we have any positioning problems with the tanks, air-traps and thrust tubes once the model is built as this would be a huge problem to resolve on the completed model. With this in mind we decided to build a test stand using the aircraft internal structure so everything is in exactly the same place. This effectively tests all the components in situ but with full access. The other benefit is with a short documentary being filmed on the development of this model the test stand allows an easy way to explain things to those who don’t have any background in what we do.

The general arrangement per side is an 8 litre kevlar tank running to an FCT air-trap then onto the pump, filter, shutoff valve and finally the turbine. All the tubing is 6mm OD Festo which steps down to 4mm OD on the last run to the turbine. The turbines are housed in a water tight bypass ducting. This will allow any water coming into the intake or up the exhaust to flow freely through and not enter the fuselage. The ducting has recesses for O-ring seals designed into the tooling. The total weight of the dry tanks, tubing, turbines and ancillaries is 8.4kgs (18.5lbs), when we add 16 litres of fuel the full system weight is just over 21kgs (46.3lbs). Obviously this is significant but not unexpected.

We started the test stand by beefing up the internal structure of the model and adding two end plates which allowed it to stand. With a combined maximum thrust of 60kgs (132lbs) I wanted to keep the stand low to the ground and have attachment points for anchoring and tethering the stand. The stand is aluminium which is bonded together with Hysol.

The turbines need to be in place while we bonded the structure to ensure a perfect fit. The positioning has been designed into the structure so its just a matter of bolting them in place.

IMG_2217 2
by Alex Jones, on Flickr

by Alex Jones, on Flickr

IMG_2184 2
by Alex Jones, on Flickr

IMG_2208 2
by Alex Jones, on Flickr

IMG_2221 2
by Alex Jones, on Flickr
e’re using the same electronics the will be used in the model. Wanting multiple redundancy for obvious reasons we have decided to use a Jeti DC24. This has two 2.4 GHz receivers totalling four antennas and a backup 900MHz receiver with two antennas. All of this runs through a CB400, we have full turbine telemetry using the new digitech CTU’s along with, IAS, altitude, heading, accelerometer and a multitude of other flight parameters that will be recoded and analysed after flight. The model won’t be taken out of the water as a matter of routine on a normal flying day (after the testing) so an RC switch is used to turn the model on and off. Refuelling will be accessed through a panel where water ingresss isn’t a concern

by Alex Jones, on Flickr

IMG_2246 2
by Alex Jones, on Flickr

I’ll do short video to explain the setup to those interested, we have used 4 over-center switches on the transmitter to allow the turbines to have an ‘idle detent’ disabling pilot control. This allows us to start each turbine independently of the other and also help with asymmetric thrust in the water if required.


Staff member
We will certainly have to do a few runs to find the best spot to accelerate in, once the skis are planing it shouldn't be a problem as she will be skipping on top of the water.

After many emails and messages asking for video we have finally put together some footage of our first test taken back in September 2020. We hope everyone is keeping safe and well and when movement restrictions are relaxed and it's sensible to do so we will be ready for the next set of tests culminating in the first flights.