S800 alternative props

[jes1111]

Please, Ferdinand - you're sounding defensive (I could be wrong, hard to tell sometimes with written exchanges). I'm not attacking or suggesting you are wrong about anything. I'm genuinely interested that your results seem to contradict the "conventional wisdom", which itself is based on the general science of propeller theory. What I find most interesting is that not only does your combo fly well, it also seems to be very efficient (assuming that's the combo used for your duration and distance attempts).

I'd really, really appreciate it if you could supply the figures necessary to generate an eCalc model for your setup (or you may already have that). I've noticed, as have others, that it is very difficult to get eCalc models up to the kind of flight duration you have achieved. This may be because eCalc's mathematical model breaks down in certain ranges, or because we're not trying the right combinations (guided as we are by this "conventional wisdom"). Or it could be a combination of these two. Either way, I'd like to find out why your combo works so effectively. If it proves eCalc's maths to be wrong then the author may be persuaded to fix it. And understanding this properly will advance the field of small multirotors for all.

Qualifiers:
- As Denny points out, maximum lift efficiency often comes at the expense of stability, and vice-versa. This represents a further compromise that needs to be understood.
- I'm definitely prepared to believe that Graupner puts more "science" into their blade designs than many other model prop manufacturers - that they know some stuff that we don't

 

[R_Lefebvre]

I was just checking for aluminium reducers 8mm - 6mm and almost fainted......the best price I could find (in a hurry) was one of our guys on the forum: http://www.multirotorforums.com/sho...m-Prop-Spacers-8mm-6mm-Reducers-Now-Available
...prices ranging up to almost $ 4,00 per piece with various vendors.

The most expensive place to buy raw material would be Amazon....and there you get a meter of seamless 8mm x 1mm tube for GBP 3,14.....http://www.amazon.co.uk/Round-Tube-8mm-1mm-Aluminium/dp/B005DR2GUM

Deeeeeeeeeeeeeenny.....any idea ???

Chris

Why not just buy them from Patrick? I have a lathe in my basement, but if I needed spacers, and had time to wait for shipping, I'd just buy them. $1.89 each, compared to setting up and making those... not worth my time. The only time I actually machine something is when I can't find, or can't wait for something purchased.

Sure, he's making a nice little profit at $1.89 if he's making these a bunch at a time. But if you just need 8, I'd probably spend an hour setting up and doing it. My time is worth more than $16/hr.

 

[jes1111]

It may be overkill, but if one were looking for high-precision collars to use as bore adapters, then Misumi offer (at a price) configurable machined collars in steel or stainless steel (with a choice of precision tolerances) and polyacetal/Delrin (with less precision). No doubt, though, that if Denny gets them manufactured they should be cheaper

Edit: and the ones at Aerial Media Pros are definitely cheaper

 

[R_Lefebvre]

Please, Ferdinand - you're sounding defensive (I could be wrong, hard to tell sometimes with written exchanges). I'm not attacking or suggesting you are wrong about anything. I'm genuinely interested that your results seem to contradict the "conventional wisdom", which itself is based on the general science of propeller theory. What I find most interesting is that not only does your combo fly well, it also seems to be very efficient (assuming that's the combo used for your duration and distance attempts).

I'd really, really appreciate it if you could supply the figures necessary to generate an eCalc model for your setup (or you may already have that). I've noticed, as have others, that it is very difficult to get eCalc models up to the kind of flight duration you have achieved. This may be because eCalc's mathematical model breaks down in certain ranges, or because we're not trying the right combinations (guided as we are by this "conventional wisdom"). Or it could be a combination of these two. Either way, I'd like to find out why your combo works so effectively. If it proves eCalc's maths to be wrong then the author may be persuaded to fix it. And understanding this properly will advance the field of small multirotors for all.

Qualifiers:
- As Denny points out, maximum lift efficiency often comes at the expense of stability, and vice-versa. This represents a further compromise that needs to be understood.
- I'm definitely prepared to believe that Graupner puts more "science" into their blade designs than many other model prop manufacturers - that they know some stuff that we don't

Interesting discussion here.

One thing that must be made clear: It's not true that there is "zero airflow/speed" in a hovering multicopter. There is an induced airflow. The air is definitely moving. Now, it may be the case that the prop could stall if you "punch it" quickly, ie: going from hover to full throttle, the prop will accelerate faster than the induced airflow, which might cause it to stall.

In any case, it's hard to argue with the real-world results Ferdinand is showing. I don't think anybody is trying to claim they are faked, or impossible... we're just trying to get our head around it because his methods do go against a lot of "conventional wisdoms".

But a big part of the problem is that, we are currently screwed for good propeller designs. (pun intended?) The simple fact is, we are using airplane propellers for an application they were never intended. Furthermore, they aren't even very good airplane propellers! The issue is that they tend to be designed as "scale propellers" modelled on full-size designs. But this doesn't work. Aerodynamics don't scale 1:1. You must consider the Reynolds number, and in this case, these scale propellers don't work well. The lift/drag ratios are lower than they could be if we had purpose-designed propellers.

I also play around with eCalc a lot, and if you believe it is perfect, you'll find it's physically impossible to build a copter which can fly for more than 30 minutes. I've "built" flying batteries. Big Lipos, with zero frame weight, and low kV motors with large props. I can't get it to go over 30 minutes. But, we know this is not the case, people have gone over an hour. So, eCalc is a useful tool for feeling things out, but it's not an exact science.

Something else I thought of while reading this thread, is in the case Ferdinand's distance record, he's obviously not just hovering, he is actually moving with a decent airspeed, so using higher pitched props makes total sense.

 

[cs8]

DennyR, your PM is full again =)

I would be interested in one set of your adapter, if there will be an option. If you can please contact me when they will be available. :nevreness:

 

[jes1111]

Something else I thought of while reading this thread, is in the case Ferdinand's distance record, he's obviously not just hovering, he is actually moving with a decent airspeed, so using higher pitched props makes total sense.

Just my thought on this particular statement...

When flying at speed the blades are not facing the airflow, i.e. the craft is not pitched forwards at 90º. So the incoming airflow is at an angle to the plane of the propeller. Therefore the consideration of the AoA in that condition has to be different (though I have no idea of how it might affect things). Logically a higher pitched blade may account (partially) for the flight distance achievement but then it would be even more strange if the same setup also achieved the demonstrated static hovering endurance.

 

[DennyR]

Actually the flow of air can move at the speed of sound and it will flow into the area of least pressure in the first instance. For example the venturi into a jet engine when taking off will draw air from behind in the opposite direction that it is traveling, that is why it has that large radius. (Sub sonic) Having done extensive research into race engine mass flow design the inlet entry is not affected as much by what is above it but more by what is around the sides of it. What determines the forward velocity of a multirotor is the thrust angle relative to the Cof G most of the air that will pass through the disc area will come from the sides

 

[R_Lefebvre]

Just my thought on this particular statement...

When flying at speed the blades are not facing the airflow, i.e. the craft is not pitched forwards at 90º. So the incoming airflow is at an angle to the plane of the propeller. Therefore the consideration of the AoA in that condition has to be different (though I have no idea of how it might affect things). Logically a higher pitched blade may account (partially) for the flight distance achievement but then it would be even more strange if the same setup also achieved the demonstrated static hovering endurance.

Ah, good point.

Trying to get my head around it... Nope, can't. Mind blown.

Hey, question for you guys: One of the last remaining unpleasant flight qualities of Arducopter is that when in fast forward flight and returning to level, or when holding position in a strong wind and then returning level, there is a very strong tendency that as the aircraft returns to level, it seems the forward airflow "catches" under the forward blades and makes the aircraft pitch back quite strongly, more than control system can accomodate. So for example, if flying forward, and then returning the pitch/nick stick to neutral, instead of "sliding to a stop", the quad pitches backwards hard, and actually comes to a stop quickly. It's almost like automatic braking. It's not so bad when flying around, as it is like automatic braking, but when trying to hold position in a wind, it's a nightmare.

Do other systems do this, or have they found a way around it? It's hard to find a good answer on this as not many people actually fly their multis aggressively like this.

I'm just wondering, can a smarter control algorithm get around this, or is the physical effect to strong?

 

[ChrisViperM]

As for the spacers: The easiest thing would be to buy 2 or 3 meter of 8x1 alu tube, find a machining shop close to you and pay a few cents per piece to be cut....I was just amazed how much money you can ask for nothing (almost...)

Following the discussion here (altough only understanding a fraction of the facts...) leads me to one conclusion: You guys are heroes...for the simple fact that there is some disagreement about a few facts, BUT nobody lost his temper or lost control over his ego....imagine this thread over at RC Groups....that would be a virtual nuclear war :tennis:

I played around with eCalc (which I never became friends with), I read lots of threads about props, ESC's and motors, but I think there is only one way to find out: buy a bunch of different props and find out for yourself, what suits you most....and that might be different from one person to another. The problem is, that a lot of people don't know what they are looking for...this forum is full with: I am looking for the best combination of XYZ....., what is the best gimbal ....., what is the best Lipo,...etc

I for my part are looking for smoothness, I don't need an agressive machine, so my prop choice will be different than somebody elses....

PS: for Ferdinand....have a look here: http://www.rc-heli.de/board/showthread.php?t=207880 (unfortunately in German)
One of the "best" discussions in connection with rotorblades....Either you laugh your head off....or you stop reading after the 1st page.

Chris

 

[jes1111]

I'm just wondering, can a smarter control algorithm get around this, or is the physical effect to strong?

If you want to try a smarter algorithm, stick a CC3D on there

 

[fritz99]

This is really an interesting discussion.

I have one phenomenon for which I do not know the explanation and no result :

I have an octocopter with gimbal and camera motor 1000kv prop : graupner 11x5 LIPO 5S 8000mah ca. 5 kilos
The flighttime is 6 minutes with 7000mah.

I have a S800 with Z15 NEX-5 LIPO 5S 8000mah about 5.2 kilos props and motors original.
The flighttime is 12 minutes with 7000mah.

Both with WOOKONG M

I do not understand the big difference in flighttime.
Can anybody tell me which props I should use on my octo so that the flight time gets longer ?
Thank you for your help.

Fritz

 

[R_Lefebvre]

If you want to try a smarter algorithm, stick a CC3D on there

Maybe I would if I could buy one.

Chris, I think you might be surprised what that would cost. They're going to charge at least $200/hour, and it will take them at least 10 minutes I'd think. You're not just hacking them off, you're also chamfering 4 places. At least... you should be.

 

[DennyR]

Ah, good point.

Trying to get my head around it... Nope, can't. Mind blown.

Hey, question for you guys: One of the last remaining unpleasant flight qualities of Arducopter is that when in fast forward flight and returning to level, or when holding position in a strong wind and then returning level, there is a very strong tendency that as the aircraft returns to level, it seems the forward airflow "catches" under the forward blades and makes the aircraft pitch back quite strongly, more than control system can accomodate. So for example, if flying forward, and then returning the pitch/nick stick to neutral, instead of "sliding to a stop", the quad pitches backwards hard, and actually comes to a stop quickly. It's almost like automatic braking. It's not so bad when flying around, as it is like automatic braking, but when trying to hold position in a wind, it's a nightmare.

Do other systems do this, or have they found a way around it? It's hard to find a good answer on this as not many people actually fly their multis aggressively like this.

I'm just wondering, can a smarter control algorithm get around this, or is the physical effect to strong?

That scenario in flight is pretty much the same with all MR's and some helis. it is a function of the stabilization algorithm, as you create less attitude hold it will fly like a typical model in manual mode. The higher the gains then the more it will try to self level when you don't want it to. If you are flying in GPS mode for example then you will need to keep forward stick pressure the whole time or it will try to stop and hold position.

 

[FerdinandK]

@jes1111, @R_Lefebvre
I used the same copter (with the 14x8 Gr Props) to fly 65min, so it makes sense to use the prop also when hovering too, sorry if I post this video again:

https://vimeo.com/38908434

I have no idea where that "conventional wisdom" comes from, but better using props that have been worked on for years, like the aeronaut, graupner, apc (I not 100% sure, but at least the first two have the same "father"). If you go for efficient (big) props on multicopter you should go for 18x10, 23x12, not for 18x4, or 18x6, 20x6, ...

@fritz99, that is easy it is the prop, the efficiency drops as the load increases. If you load a 11x5 with 200gr you are in the range of 10gr/Watt, if you load it by 800gr you are arround 6gr/Watt (of course the motor is also important).

best regards

Ferdinand

If only I would get a 18x10 CW and CCW ...

 

[fritz99]

@jes1111, @R_Lefebvre
I used the same copter (with the 14x8 Gr Props) to fly 65min, so it makes sense to use the prop also when hovering too, sorry if I post this video again:

https://vimeo.com/38908434

I have no idea where that "conventional wisdom" comes from, but better using props that have been worked on for years, like the aeronaut, graupner, apc (I not 100% sure, but at least the first two have the same "father"). If you go for efficient (big) props on multicopter you should go for 18x10, 23x12, not for 18x4, or 18x6, 20x6, ...

@fritz99, that is easy it is the prop, the efficiency drops as the load increases. If you load a 11x5 with 200gr you are in the range of 10gr/Watt, if you load it by 800gr you are arround 6gr/Watt (of course the motor is also important).

best regards

Ferdinand

If only I would get a 18x10 CW and CCW ...

Hallo Ferdinand !

So which props should I use for the octo ?

18x10. ?

They are too long for my octo.
The S800 has 15x4 props.

Fritz

 

[jes1111]

Here's the issue as eCalc sees it:

Ferdinand's endurance setup (hoping all my numbers are right):
View attachment 6613
eCalc gives up and says "No way, man!"

Reduce the propeller pitch to 7" and eCalc coughs up a solution, but still insists that everything is going to melt and burn:
View attachment 6612

Yet, by the evidence, this setup works and the motors and ESCs are still intact after 65 minutes flight. So we can presume that eCalc's mathematical model has departed from reality (whereas for more "normal" setups it has proved to be reasonable accurate).

(I've assumed Graupner's Propeller Constant to be the same as an APC-E, which may not be the case, but adjusting it makes little difference to eCalc's insistence that this shouldn't work.)

 

[jes1111]

Note, too, that eCalc is calculating the motor efficiency at hover as around 75% - which would normally indicate an "inefficient" setup.

 

[R_Lefebvre]

Wow, so 75% is bad... my setup is supposed to be 65%.

View attachment 6616

The issue is that, I think, it's way over-motored for what I'm actually planning on using it for. But that was necessary for a few reasons. I'm building it to be able to lift a 1kg camera/gimbal. (not factored into these figures yet). With that load, the efficiency is still only 71%. If I put the payload and change the 2P battery setup for 4P (so 4S 20,000mAh) then I get 14 minutes flying, 77% efficiency, 42% throttle.

One of my design criteria was to *not* be hovering at 60-70% throttle, even though that would be more efficient. I wanted the reserve capacity needed to keep flying with a motor failure.

Still, I would have hoped I could get better results than this.

 

[DennyR]

Ecalc has never come close to confirming what my own designs have used in the past.(after optimization) their apples are just not the same as mine I guess. Suck it and see works for me....

 

[DennyR]

Getting back to where we were. The new adapters with a 1 mm longer shaft will be finished tomorrow for those who what to get jiggy with the big props. Everyone who has ordered thus far, all have gone out. allow 4-7 days. :tennis: Phew, that was hard work.