Originally Posted by grael

Good to see the science and observations matching!

Yes, indeed the data collected from the telemetry sensors confirms what I always expected.
Now it is possible to make a much better estimation of the pitch.
In most 3D CAD programs you can calculate the volume of the blades.

Concerning a mono-foil, seen this?

Who needs a stabilizer? (3 min 54 sec)

And talking about a fully submerged blades hydrofoil actively controlled:
https://www.modelbouwforum.nl/thread...rtboot.279272/
Google translate is your friend

Originally Posted by unusual_rc

@Steven_M, noticed this movie of yours on youtube:

https://www.youtube.com/watch?v=_XM2nKv9fks

Very interesting, since you noticed what I am too experiencing.
This IMHO is the difference between large and RC scale pumps using the same medium (water):

1. Blades take in space in the water column, which reduces the amount of water to be pumped per revolution,
2. Overlapping blades are not needed for small pumps due to fact water does not scale.

Yes it is very interesting and I still do not know where this experiment will take me.
Regarding these points:
I think 1) is negligible. The "blocked volume" of a single impeller blade is relatively much smaller than the change in performance.
2) Yes, small pumps are a completely different story than big pumps, it seems. But you also need to take priming into account.

Originally Posted by unusual_rc

What I did after last run, was calculating the true pitch of my impeller based on RPM and power usage.
My 40mm pitch (.9) is in reality 35mm (.8) I found.
If I use this value in my calculation, everything is spot-on!

Sounds great but how do you mean that? Did you do static thrust test of your pump? Or did you do some calculation from your telemetry when the boat is moving? The problem is we cannot measure the thrust when boat is moving.

Originally Posted by unusual_rc

A friendly model boater tested the 3 bladed impeller in the KMB28 pump for me with telemetry logging, and found out he had less performance with it. Pitch of the blades was equal, just added a 3rd blade. (the Bronze AliXpress version).

Great to hear that my tests are not alone showing this!

Originally Posted by unusual_rc

So I am eager to test a 2 blade non overlapping impeller soon.
Why 2 blades instead of 1?
Overlapping is not necessary, and 2 blades do not give unbalance.

In the early days of model airplanes those 2 stroke diesel IC engines features a single blade propeller, with a counter balance weight on the other side.
Those were stationary engines, so the dynamic balance could be estimated.

I do not think you can with a proportional throttle, although the enclosed pictures shows something different?

Nevertheless 3D printing a single bladed prop, which makes a full 360° revolution is feasible for sure!

So first, regarding the balance. It is less imporant than I thought and it works even unbalanced. I am working on balancing it with grub screws on the other side.

And regarding the 360° revolution single blade impeller, yes I tested it as well (not in the video). It will not work as a single stage because it almost cant prime. But when it finally primes, the efficiency is great. After more testing I was able to achieve the same efficiency with two stage configuration so there is no more reason to work on single stage 360° single blade version for me.

Originally Posted by unusual_rc

And can you tell more about your stator blade testing? More blades better or worse?

I shared very detailed info about many more things I tested in a discord group created by the kwikjet X creators, you are free to join it, the link is in their fb group.
Basically 7 is the best in my case. Anything more is worse, anything less is also worse. So the 11 vane stator produces more laminar flow but is less efficient.

Originally Posted by Steven_M

Yes it is very interesting and I still do not know where this experiment will take me.
Regarding these points:
I think 1) is negligible. The "blocked volume" of a single impeller blade is relatively much smaller than the change in performance.
2) Yes, small pumps are a completely different story than big pumps, it seems. But you also need to take priming into account.

Concerning 1st point, If I compare the volume in CAD it is quite something in the water column the total axial length the blades are long.
For my latest 40mm pitch impeller it is 13% of the total volume for 1rpm.
So if I remove 13% from the 40mm pitch, I get 35mm of pitch.
And if I fill in just that in the excel sheet, including correct measured rpm, it returns exactly the used electrical power in Watt!

Agree that priming is a very important factor. However my 2 bladed Graupner jet primes fine, as do most KMB-28 units do in general.
I have not calculate it (it is possible for sure to do so), however I do not believe the under-pressure in air created by an impeller makes a whole lot of difference with overlapping blades. I have to say my FJD45mm jet primes really quickly. However the model is heavy (the stern sits deep) and the inlet channel is quite large.
Personally, having experience with hydrofoil surfing, aeration is the biggest issue: the ability for air to find a route through the water into the inlet channel. With hydrofoiling, once your submerged wing comes too close to the surface, due to under pressure in the water air can form channels through it, out of nothing. Amazing to see on camera! I think something similar has to do with repriming.
Worth the investigation though

Originally Posted by Steven_M

Sounds great but how do you mean that? Did you do static thrust test of your pump? Or did you do some calculation from your telemetry when the boat is moving? The problem is we cannot measure the thrust when boat is moving.

I am still wanting to buy a cheap digital suitcase scale. Found one which goes up to 40KG.
My latest calculation, which compensates losses shows 12,5L/second with 1600W of power.
1 Litre = 1 Kilogram.

There is a difference in static thrust and thrust at top-speed. The inlet channel is loaded much easier, less drag.
That is why I did not bother, but it is on my list.

Again, this shows that shoehorning a large drive in a small boat makes not much sense to reach top speed.
It is brute force, for a large heavy model.
As a general rule of thumb the weight of the boat should be half of the maximum thrust, to have a very well performing model.
For example, with 800W of power a KMB-28 can do up to 3,5Kg of thrust. Their jetsprint model weights around 1,6-1,8Kg.
Topspeed 35km/h
My findings of the last 25 year rc jet-boating.

Originally Posted by Steven_M

So first, regarding the balance. It is less imporant than I thought and it works even unbalanced. I am working on balancing it with grub screws on the other side.

As a rc model airplane flyer unbalance is always a bad thing.
Good thing about a jet is that the axis is at least firmly supported on two positions.
So any unbalance has less effect, especially with two stage counter positioned single blades

Originally Posted by Steven_M

And regarding the 360° revolution single blade impeller, yes I tested it as well (not in the video). It will not work as a single stage because it almost cant prime. But when it finally primes, the efficiency is great. After more testing I was able to achieve the same efficiency with two stage configuration so there is no more reason to work on single stage 360° single blade version for me.

Repriming issues? Interesting.
What do you mean with the last sentence, that you could match efficiency with the two stage?
With two 180° single bladed impellers?

Originally Posted by Steven_M

I shared very detailed info about many more things I tested in a discord group created by the kwikjet X creators, you are free to join it, the link is in their fb group.
Basically 7 is the best in my case. Anything more is worse, anything less is also worse. So the 11 vane stator produces more laminar flow but is less efficient.

Interesting find!
Unfortunately, I dropped all social media accounts because I do not agree with their business model

Originally Posted by unusual_rc

Concerning 1st point, If I compare the volume in CAD it is quite something in the water column the total axial length the blades are long.
For my latest 40mm pitch impeller it is 13% of the total volume for 1rpm.
So if I remove 13% from the 40mm pitch, I get 35mm of pitch.
And if I fill in just that in the excel sheet, including correct measured rpm, it returns exactly the used electrical power in Watt!

You are right, this is important. I did not think of it this way. So 13% is a lot and will be more if the blades are thicker. I should also incorporate this into my sheet.
But still, IMO this is only part of the reason why less blades can be more efficient.
With that electrical power used, do I understand correctly that your expected input power when you insert these values exactly matches your actual measured input power? I am not sure how much this means. Go with me (or correct me):
1) The sheet calculates the pump as being 100% efficient. It expects the input power same as the output power. But it is not, in reality it is very safely said under 60%.
2) The sheet only works for static thrust testing. In contrast, you are measuring the input power when the boat is driving, right? But when the boat is driving, the sheet does not work because the water going into the intake already has velocity and kinetic energy. The power calculation in the sheet only works when input water has zero kinetic energy.
So when the actual pump works so much differently than in the calculation and so many things change, why should it be "correct" that the expected input power exactly match the output power?

Originally Posted by unusual_rc

Agree that priming is a very important factor. However my 2 bladed Graupner jet primes fine, as do most KMB-28 units do in general.
I have not calculate it (it is possible for sure to do so), however I do not believe the under-pressure in air created by an impeller makes a whole lot of difference with overlapping blades. I have to say my FJD45mm jet primes really quickly. However the model is heavy (the stern sits deep) and the inlet channel is quite large.
Personally, having experience with hydrofoil surfing, aeration is the biggest issue: the ability for air to find a route through the water into the inlet channel. With hydrofoiling, once your submerged wing comes too close to the surface, due to under pressure in the water air can form channels through it, out of nothing. Amazing to see on camera! I think something similar has to do with repriming.
Worth the investigation though

Interesting thing with the hydrofoil, if you share a video I will take a look

Originally Posted by unusual_rc

I am still wanting to buy a cheap digital suitcase scale. Found one which goes up to 40KG.
My latest calculation, which compensates losses shows 12,5L/second with 1600W of power.
1 Litre = 1 Kilogram.

There is a difference in static thrust and thrust at top-speed. The inlet channel is loaded much easier, less drag.
That is why I did not bother, but it is on my list.

I understand that you mean the 12,5kg/s mass flow only when boat speed is achieved?
If this test was done static, for FJD 45mm with 33mm exit nozzle, 12,5kg/s mass flow would mean that it is producing over 18kg of thrust - IMO unrealistic value for pump of this size.
I am curious, how does your amp draw change when you hold the boat static and go full throttle? Does it go higher than when the boat is driving?

Originally Posted by unusual_rc

Again, this shows that shoehorning a large drive in a small boat makes not much sense to reach top speed.
It is brute force, for a large heavy model.
As a general rule of thumb the weight of the boat should be half of the maximum thrust, to have a very well performing model.
For example, with 800W of power a KMB-28 can do up to 3,5Kg of thrust. Their jetsprint model weights around 1,6-1,8Kg.
Topspeed 35km/h
My findings of the last 25 year rc jet-boating.

As a rc model airplane flyer unbalance is always a bad thing.
Good thing about a jet is that the axis is at least firmly supported on two positions.
So any unbalance has less effect, especially with two stage counter positioned single blades

I finally agree on this
I will try to solve the unbalance indeed.

Originally Posted by unusual_rc

Repriming issues? Interesting.
What do you mean with the last sentence, that you could match efficiency with the two stage?
With two 180° single bladed impellers?

With the single blade 360degree single stage impeller, I achieved 36% pump efficiency (estimating 80% brushless motor efficiency) when static thrust testing. Then after more testing I got up to 40% static efficiency with two stage units (two single blade 180° spaced impellers) with uncomparably better priming as the result. I also tested different blade counts on first and second impeller and the results are interesting, I am still working on it.

Originally Posted by unusual_rc

Concerning a mono-foil, seen this?
Yes, I think it's Horst Sergio's friend.
Here's Horst's thread about unifoiling:
https://kiteforum.com/viewtopic.php?...35&hilit=horst

And talking about a fully submerged blades hydrofoil actively controlled:
https://www.modelbouwforum.nl/thread...rtboot.279272/
Google translate is your friend

Thanks!
Google translate wouldn't translate for me but it was still interesting. The Page 3 Mini40 rc foiling yacht was interesting too.

Fellas, this is a very interesting subject matter indeed! I am thrilled that there are people like yourselves who are anxious and willing to take the time to find the answers and realities! I find I no longer have what it takes to endure all the necessary testing and measuring, but am enthralled with interest in your quest for the truth Please keep up the good work!!

Single blade impellers can even be beautiful.

Originally Posted by Kayaker

Single blade impellers can even be beautiful.

Kayaker! Many years ago I had a propeller like that! It was from the swarf from cutting gun barrels. A beautiful long corkscrew mono blade.

I believe the lily shaped impeller above runs in a tube at the bottom of a fish tank.
http://www.viva-technology.org/New/IJRI/2021/24.pdf

Originally Posted by unusual_rc

Blades take in space in the water column, which reduces the amount of water to be pumped per revolution,


So I am eager to test a 2 blade non overlapping impeller soon.
Why 2 blades instead of 1?
Overlapping is not necessary, and 2 blades do not give unbalance.

And can you tell more about your stator blade testing? More blades better or worse?

I’ve been testing several stators and impellers to make improvements to a 35mm jet drive and here is the short version:

Different stator vane shapes and lengths all give similar thrust and top speed. The biggest difference is in steering with a little difference in re-priming and boat handling. Grael’s 11 vane stator lost prime sometimes (not often) when it sucked air but his 7 vane stator rarely looses prime. I’m going to test his 5 vane stator in a speed boat I’m redesigning, hopefully this winter. A straighter water column makes the boat run straighter and turn more predictably. A straighter water column also makes a boat hop more making trim tabs necessary.

I’ve been able to make an impeller that gives very good boat handling and an impeller that has good top speed but can’t get both in the same impeller. I’m working on a boat hull that can handle more speed before I get too carried away with impeller design.

Thanks Robin, looking forward to your plans and findings.

Only experience I had was that the 4 vaned large AoA stator blade version of the FJD 45 had the same top speed as the modest angled one. However amp draw was much higher.

Greal made some nice designs.
I will put more time into the stator housing design this winter, besides a new boat.

Not sure if I step over to a 35mm drive, or maybe a 40mm.
I have to say I am quite happy with the 45mm version. It is large, but still runs on 4 to 6S (above 6S the Esc become very expensive so it appears) well.
Need a bigger boat

@unusual_rc
Can I ask you one thing? Sorry if you already wrote this somewhere, but how does your motor RPM and amp draw behave when you hold the boat on full throttle static, compared to when you drive it at speed normally? How big is the change?

Originally Posted by Steven_M

@unusual_rc
Can I ask you one thing? Sorry if you already wrote this somewhere, but how does your motor RPM and amp draw behave when you hold the boat on full throttle static, compared to when you drive it at speed normally? How big is the change?

With nylon 12 impeller Ø45 Pitch of 40mm, FJD original stator with Ø33mm exit nozzle static: 140A with 14krpm
With nylon 12 impeller Ø45 Pitch of 40mm, FJD original stator with Ø33mm exit nozzle dynamic: 140A with 14.75krpm

Checked the Voltage under load at these peaks, and those are comparable.

However these are the plastic impellers, which did show a lot of loss of efficiency compared with the metal version I am currently testing.
That difference is more dramatic than the difference in RPM.

Nevertheless it seems RPM is a tat higher when dynamic. Speed was between 30 and 35kmh at that moment.
By comparison: metal impeller with better flow stator-housing (smaller 4 blade stator vane angle, thinner blades) does 110A at 16krpm and 40kmh +
A huge difference!

And do not forget the shape of the inlet channel is very important when testing static and dynamic.
I know some jetsprint teams did some testing in a flow tank. Even with flow angles diagonal on the axis direction.

Time for an up for this topic!

In the last year I learned a lot!
Again the inevitable rule of all "assumption is the mother of all f...-ups" was my enemy, in despite in my professional work I never make assumptions.
Why do I do it in this hobby then?
Example, my assumption of brushless motors were running at their highest rpm possible when efficient.
I was wrong! Part of my assumption was based on the Esc I used, which later proofed to be malfunctioning from the day I used it.

Steven_M has been working with the Excel calculations, and so did I.
He brought up two points:

1. maximum possible thrust of a jet is only present during a static test
2. What is the total needed power given by the sheet exactly telling us compared with real-world testing?

Point 1: Max thrust is only present during a static test
True! As soon as the boat start to accelerate the 3rd law of Newton becomes present.
Forward motion of the object is in the opposite direction of the water stream propelled from the exit nozzle. At a certain moment the thrust is equal to the forward motion plus drag, which levels out the equation to zero.

Still at that moment the same amount of water is being pumped, and probably even more (assumption alert!) due to the inlet water velocity creating less drag.
So I still think the calculation is valid for top speed.

This is hard to derive from the measured data and boat speed.
Nevertheless I have adapted Steve's static thrust test, which I will perform more in the near future.
Problem with a drive in a boat is it starts to become an issue keeping the boat on the water in the high rpm range.
I might need to build a rig too!


2. Total power (in Watt) from the formula's inside the excel sheet
The efficiency of brushless motors for RC use is rarely given by the manufacturers.
For an inrunner the efficiency is lower than that of a outrunner.
Still 80% is a usable an realistic figure.
Also the Esc uses energy to make the motor turn. Never found any usable figures for that.

Which means 1000W electrical power makes 800W mechanical power on the motor axis.
200W lost as heat due to resistance and a small bit as mechanical losses.

The sheet calculates the maximum thrust given without any resistance, drag and so on. Utopian situation.

So a random jet-drive starts already with the 20% loss of the motor.

What happens with the rest of the 80%?

Weird thing is when I give in in real life measured rpm and the measured thrust by the static test (11kg in my situation until the boat left the water. That was not full throttle) I come really close by the thrust in the sheet.
That said, I took another assumption (alert!):
The volume of the blades inside the water column of a single revolution of the impeller is not taken into account by these formula's.
In my case that is 13% of the volume of a single revolution of the impeller, so if I take that amount my real-life pitch of 40mm is more 35mm.
If I fill in that figure for pitch, the figures are spot on!

Does that mean my drive is crazy efficient?
No, impossible of course.
But what is it then?


Did some CFD simulations where I replaced the impeller of my drive for the one I currently use, a progressive pitched version with the same pitch, and a 3 blade short (non-overlapping) impeller bladed version with the same pitch.
Result:
My current impeller gives the highest exit velocity.

The progressive pitched impeller gives the second highest velocity, only 0.1 m/s of a difference.
However the power needed to turn the impeller is less (10%)!

The short bladed impeller gave more than 2.0m/s less exit velocity, however 45% less power needed than my current design.

Assumption alert again: are my CFD simulations correct?

Originally Posted by grael

Unusual_RC, If you are going to play with the FJD stators and impellors again I suggest you have a look at this:

https://www.youtube.com/watch?v=bCwtcDNB15E

It's a truly mind expanding look at foil design taking the founder of common aerodynamic theory's later refinements to his earlier elliptical wing theory and calculations that people have been using almost exclusively for the last century.

To put it in a nutshell, Ludwig Prandtl superseded his earlier and simpler(but now widely taught) theory after some years with new calculations mirroring bird's wings superior efficiencies, not invalidating his previous calculations, but with a new and superior grasp on how to flow between lift and no lift at the wing tips. I think in his head he was content in the knowledge that he had achieved something amazing but by then people were preaching the religion of his first formula and his new pages lay ignored for almost a century. Apparently he had some brief contact with the Horton brothers who understood him and went on to produce some exceedingly efficient delta planes. Recently some new people are investigating and reaping the rewards. Note in his video the discussion of fan blades for the space station, lessons learned from Prandtl's second papers are eminently transferrable to increased jet pump efficiency too. Do persevere, I started watching the above video for the second time last night and in the second viewing I'm absorbing important details that escaped me the first time.
Graham.

Took a while before I had the chance to look at the video, and learn more about Ludwig Prandtl's wing load theory!
Indeed very interesting because I love to fly both real as in RC (more affordable )


The biggest difference between water and air is compressible versus in-compressible fluid.
Nevertheless on the webiste of the NASA they state this theory also applies to turbo machinery (the summarizing name for all propellers, impellers and compressors) including water propulsion!

The link to water-jet propulsion is clear to me!
I am not starting straight with this theory, but for sure I will dive into this deeper concerning impeller design in the near future!
The 3D printed cooling fan design for the ISS is amazing! 13% more efficient and much more quiet.
A correct implementation of this theory in a impeller could create in excess of 10% efficiency, which makes it more than worth to try!

I could be wrong, to what I understand Prandtl and the Horten brothers never met in despite they lived in Germany before WWII.
Their theories come very close.
The Horten brothers are the fathers of among other the B2 bomber of the US air-force.
After WWII the US confiscated a fly-worthy plywood flying wing powered by two Jumo jet-engines. which is still in procession of the the NASA till this day.

Found a somewhat shorter and less academic explanation of the Prandtl wing load theory by Albion H. Bowers from the NASA.

NASA's Albion H. Bowers - " Why Birds Don't Have Vertical Tails" - AMA EXPO 2014 (38 min 48 sec)

This does not mention the cooling fan design based on this theory though.
I cannot find any info online about this fan unfortunately.
As Grael already pointed out this can be very interesting for impeller design.
Especially since this could potentially move the cavitation tip speed much further away!