Power Systems Talk about motors, ESC speed controllers, gear drives, propellers, power system simulators and all power system related topics

Do propellers really "unload" in flight?

Old 09-26-2008, 03:33 PM
  #101  
Tram
Purveyor of EPP goodness.
 
Tram's Avatar
 
Join Date: Dec 2005
Posts: 249
Default

Let me know if you want me to test anything else on the same plane..

I'm not too much into the debate, but.. Have plane.. will fly..
Tram is offline  
Old 09-26-2008, 11:09 PM
  #102  
bz1mcr
AKA owner
 
bz1mcr's Avatar
 
Join Date: Aug 2005
Location: Troy, MI
Posts: 136
Default

Originally Posted by Larry3215 View Post
....

If we take Trams static run the power starts out at 318 watts and 10.74 volts.

Then 5 seconds into that static run the battery has dropped to 10.39 volts.

If the theory is correct, that 3.2% drop in voltage will result in a roughly 10% drop in power. That means the power should have dropped by 31.8 watts to 286 watts. The power actually dropped to 275 watts or about 13.5%

After looking at a number of static runs on the data files, the theory is a pretty close match to the recorded data. In practice, the power tends to drop off a little more than the thoery predicts. Probably due to motor heating.

I just took a quick look at Jeffs data file for the run where he did some level flight speed runs.

As I said his static run up folows the thoery very well with the power dropping off a bit more than the theory preditcs as the voltage drops.

On his first level speed run the total power drop after 5 seconds is aprox 27 watts from 290 watts down to 263 watts.

Of that 27 watt drop aprox 9 watts or about 1/3 is due to voltage sag from the packs. That doesnt take into account the motor heating.

So the power drop in flight on Trams speed plane is looks to be about 40/60 battery sag/motor heating vrs prop unoading.

The total power drop in level flight is about 9.3%

About 3%-4% is due to battery sag/motor heating and 6% is left over for "prop unloading".

I need to go through the file more carefully but thats the rough values for the first part of the data. I still need to look at the climbs and dives and later level flights to see if the data is consistant.
Larry,
I think the data and analysis is beginning to give us valuable insight.

A couple of questions:
1) Do you think the plane was at full speed after 5 sec.? Was the power fairly flat at that point or still going down indicating the plane is still accelerating?
2) How much of the difference in starting power (318 to 290) do you think was do to prop unloading? That 28 W drop is bigger than the the other drops.
bz1mcr is offline  
Old 09-27-2008, 04:00 AM
  #103  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Originally Posted by Tram View Post
Let me know if you want me to test anything else on the same plane..

I'm not too much into the debate, but.. Have plane.. will fly..
If you had an rpm sensor Id be working you to death As it is I think we have all the data you can provide - unless someone else come up with a different flight profile perhaps.

Originally Posted by bz1mcr View Post
Larry,
I think the data and analysis is beginning to give us valuable insight.

A couple of questions:
1) Do you think the plane was at full speed after 5 sec.? Was the power fairly flat at that point or still going down indicating the plane is still accelerating?
2) How much of the difference in starting power (318 to 290) do you think was do to prop unloading? That 28 W drop is bigger than the the other drops.
1)The plane was very close to full speed but not quite. Durring the 5 second section I chose speed went from about 68 to 92. Top speed was about 94. The power level did not stop dropping as you can see on the graph. Voltage continued to drop all along and Im now sure that motor heating is playing a small but significant roll in this as well.



2) Of that 27 watt drop Im thinking that about 16 watts was due purely to the prop unloading and the rest was due to motor heating and battery voltage drop.

Aprox 8 watts can be calculated as due to voltage drop. The remaining 3 watts Im thinking was motor heating or some other heat type loss.

I decided to go with the first 5 seconds of the speed run so I could roughly assume that the extra loss from motor heating would be aproximately the same % as during the 5 seconds of static run.

A bunch of "ass"umptions in there

A quick look at the static run shows that the power drop continues to increase as the voltage drop starts to level off.

Im assuming once again that the motor is heating up as the battery is heating up at the same time. The battery heating would lower its internal resistance but the motor heating would increase its resistance. It looks like motor heating is winning slightly and becoming more significant over longer power runs.


I really wish we had rpm data for the same time frames as that would help in the analysis.

Attached are some zoomed in sections of the graph data. You can see the static run and the first level speed run and well as one showing the level speed run and the following climb and dive runs.

I havent had time to go through the climb and dive runs yet.
Attached Thumbnails Click image for larger version

Name:	static1.jpg
Views:	147
Size:	388.7 KB
ID:	79358   Click image for larger version

Name:	level1.jpg
Views:	131
Size:	351.9 KB
ID:	79359   Click image for larger version

Name:	climbdive1.jpg
Views:	142
Size:	364.1 KB
ID:	79360  
Larry3215 is offline  
Old 09-27-2008, 04:57 AM
  #104  
bz1mcr
AKA owner
 
bz1mcr's Avatar
 
Join Date: Aug 2005
Location: Troy, MI
Posts: 136
Default

Larry,
you focused on the first 5 sec of each run, I see this data toward the end of each run:

At time = 30-35 sec ,last of static run:
V=10.3
W= 278
speed=0

At time = 235 sec ,last of level speed run:
V=10.3
W= 258
speed=90+
Note-- apparent unloading due to speed 278-258= 20 W for 0-90+ MPH
With no apparent Voltage sag!

I agree you really need motor RPM to sort out the data at other than level flight.
Don
bz1mcr is offline  
Old 09-27-2008, 05:01 AM
  #105  
Tram
Purveyor of EPP goodness.
 
Tram's Avatar
 
Join Date: Dec 2005
Posts: 249
Default

The only RPM sensor I have is the inline sensor.. Nothing fancy..
Tram is offline  
Old 09-27-2008, 05:27 AM
  #106  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Originally Posted by bz1mcr View Post
Larry,
you focused on the first 5 sec of each run, I see this data toward the end of each run:

At time = 30-35 sec ,last of static run:
V=10.3
W= 278
speed=0

At time = 235 sec ,last of level speed run:
V=10.3
W= 258
speed=90+
Note-- apparent unloading due to speed 278-258= 20 W for 0-90+ MPH
With no apparent Voltage sag!

I agree you really need motor RPM to sort out the data at other than level flight.
Don
You bring up a good point - comparing points on the curve where the voltages are identical. That was something I had planned but hadnt gotten around to doing.

I have an advantage in that I can be more precise collecting the data

If we look at the static data right about the 32 second mark the voltage drops to 10.39 volts and power is 278 watts.

In the first level run, at about 235 seconds is the point where the voltage first drops down to the same 10.39 volts. Power at that point is 260 watts.

The difference is 18 watts. Pretty close to what I said above.

Keep in mind that that was about 20 seconds into the speed run but only 15 seconds into the static run, so motor heating may have been different - or maybe not. In the speed run you would expect the motor be be getting more cooling that under static conditions. Im not sure how to interpret that.

yeah - we need rpm data. I tink it could help fine tune this.

Still - somewhere between 16 and 18 watts seems to be due to prop unloading out of a roughly 300 watt setup. About 5% - 6% total.

Much less than expected - at least by me.

Im going to try to zip the data file so any of you who have the software can play with the raw data if you want.
Attached Files
File Type: zip
vecjet136.zip (21.1 KB, 105 views)
Larry3215 is offline  
Old 09-27-2008, 05:28 AM
  #107  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Originally Posted by Tram View Post
The only RPM sensor I have is the inline sensor.. Nothing fancy..
The phase reading sensor would be best under these conditions I think.
Larry3215 is offline  
Old 09-27-2008, 05:44 AM
  #108  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Ok, now here is a bit of interesting data that may just confuse the issue.

If we look earlier in the static and speed runs the results are very different!

On the static run at the 24.25 second mark voltage is 10.54 and power is 296 watts.

On the speed run at 227 seconds voltage drops to 10.54 and the power is 290 watts at 62 mph.

Only a 6 watt drop at 60 mph!

Man I sure wish we had rpm!
Larry3215 is offline  
Old 09-27-2008, 06:24 AM
  #109  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

On second thought that data point actually fits in nicely. The plane was still accelerating at that point and not close to its terminal speed, so the prop handnt started to unload yet.

It occures to me that if Tram re-charged his packs after the static run we might even see a higher peak power early in the speed run than we saw in the static run.

This seems to follow the predicted performance of a hi pitch prop.
Larry3215 is offline  
Old 09-29-2008, 12:40 AM
  #110  
BEC
Bernard Cawley
 
Join Date: Jan 2006
Location: Auburn, Washington
Posts: 59
Default OK - here it is

OK Larry,

Here it is: throttle position, RPM, and power all logged. Static runup, part power takeoff, approximately level speed run, approximately level with pull to vertical, then a full power dive. The dive was short and far from straight down (more like 30 degrees down), but I think it's clear what's happening.

Grey is throttle (from about 40 to 200 is really zero to 100%), orange is power and red/pink is RPM. Data is from a flight of the Switchback Senior at the meadow at Weyerhauser about two hours ago.

EagleTree V3. I didn't plot motor or battery temp (or volts or current) but they were logged.

Oh - the power system is a Scorpion 3008-32, driven by a 3s Abosolute PowerXTreme Lite 2500 mAh battery through a Castle Thunderbird 36. Prop is APC 11x7E.

It was gusty from 6 to 12 mph with the odd gust to nearly 20 mph (measured) and the wind direction was variable, so that should account for some of the less-than-smoothness in the graph.
Attached Thumbnails Click image for larger version

Name:	annotated_graph.jpg
Views:	160
Size:	184.5 KB
ID:	79456  
BEC is offline  
Old 09-29-2008, 02:35 AM
  #111  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Thats great Bernard!

Can you attach the data file or e-mail it too me?

I didnt get to do a data run today because of the winds. It was way too gusty over on my side of the pond too. The lite foamy I want to do the first runs on would have been all over the pace.
Larry3215 is offline  
Old 09-29-2008, 08:17 AM
  #112  
BEC
Bernard Cawley
 
Join Date: Jan 2006
Location: Auburn, Washington
Posts: 59
Default

FDR file e-mailed.
BEC is offline  
Old 09-30-2008, 05:21 AM
  #113  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Got the file - thanks!

I'll run some numbers and post the results as soon as I can.
Larry3215 is offline  
Old 10-13-2008, 03:40 AM
  #114  
BEC
Bernard Cawley
 
Join Date: Jan 2006
Location: Auburn, Washington
Posts: 59
Default Here's another one

Same plane, but under almost dead calm conditions. Pretty similar, really. Oh - this battery isn't as good as the other one which is apparent in the short vertical.

I did get a much steeper dive this time.
Attached Thumbnails Click image for larger version

Name:	Flight11_unload_annotated.png
Views:	135
Size:	70.5 KB
ID:	80551  
BEC is offline  
Old 10-13-2008, 06:00 PM
  #115  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Thanks Bernard!

I havent had time to run the numbers on your other flight yet or to make any flights of my own. Life has been getting in the way of my fun!
Larry3215 is offline  
Old 10-13-2008, 08:12 PM
  #116  
Dr Kiwi
Super Contributor
 
Join Date: Aug 2005
Posts: 1,726
Default

Originally Posted by BEC View Post
Same plane, but under almost dead calm conditions. Pretty similar, really. Oh - this battery isn't as good as the other one which is apparent in the short vertical.

I did get a much steeper dive this time.
Bernard - here's my very naive analysis:

If we take the mid-points of (1) the "static run", (2) the "speed run", and (3) the "climb".... from your graph I read:

(1) 6700rpm from 210W - thus ~32rpm/W

(2) 6900rpm from 180W - thus ~38rpm/W

(3) 6750rpm from 217W - thus ~31rpm/W

I would take this to mean that during the high speed run the prop is unloading by 38-32/32 = 18.7% [or is it 38-32/38 = 15.7%].

And there is possibly a small degree of "loading-up in climb", evident from the "static versus climb" figures.
Dr Kiwi is offline  
Old 10-14-2008, 03:37 AM
  #117  
BEC
Bernard Cawley
 
Join Date: Jan 2006
Location: Auburn, Washington
Posts: 59
Default Same data - with speeds!

I was testing the new EagleTree airspeed sensor yesterday and ALSO had the GPS expander aboard. Here is the same graph (pretty much) with the speeds also shown. Note that in general the GPS and pitot/static-based speeds track well together. It was almost dead calm so as long as the pitot tube was pointed properly in the direction of flight I'd expect that....
Attached Thumbnails Click image for larger version

Name:	Flight11_unload_annotated_speeds.png
Views:	122
Size:	86.7 KB
ID:	80681  
BEC is offline  
Old 10-14-2008, 05:00 AM
  #118  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

Originally Posted by Dr Kiwi View Post
Bernard - here's my very naive analysis:

If we take the mid-points of (1) the "static run", (2) the "speed run", and (3) the "climb".... from your graph I read:

(1) 6700rpm from 210W - thus ~32rpm/W

(2) 6900rpm from 180W - thus ~38rpm/W

(3) 6750rpm from 217W - thus ~31rpm/W

I would take this to mean that during the high speed run the prop is unloading by 38-32/32 = 18.7% [or is it 38-32/38 = 15.7%].

And there is possibly a small degree of "loading-up in climb", evident from the "static versus climb" figures.
I think it would be interesting to see what the voltages were at those three points.

I tend to think that comparing points where the voltage was the same would give a better idea of the relative loading. The idea being that at a given voltage the motor should want to turn the prop at a similar rpm. Any differences should be due to a combination of heat losses and loading/unloading.

I was actually surprised at how large the heating losses seemed to be in the earlier graphs from Tram that I looked at. It was quite significant accounting for something like 20-30% of additional drop in power output during the static run over and above that due to voltage drop alone IIRC.

That kind of complicates calculating the pure prop unloading because I dont see a good way to account for the effects of extra cooling in flight other than to 'wag' it or ignore it.
Larry3215 is offline  
Old 10-14-2008, 06:07 AM
  #119  
BEC
Bernard Cawley
 
Join Date: Jan 2006
Location: Auburn, Washington
Posts: 59
Default

OK - here's the same period on the same flight with voltage, throttle, motor winding and battery temperature. Temp A is the winding temperature from a sensor blu-tak-ed to the windings in the back of the motor. Temp B is the battery - sensor held to the battery by the Velcro "safety belt" in the plane.

I'll let you match up the two....as you know the scales on the first would make the voltages almots impossible to see on the second.

Or, I can send you the .FDR file and let you pull it out. Warning - I was logging GPS parameters so in 10.8 minutes the recorder was FULL. (You can't pick and choose - it's all or nothing).

Gee, I should export to .kml and then put the flight path on Google Earth.....but not tonight.
Attached Thumbnails Click image for larger version

Name:	Flight11_unload_volts-temps.png
Views:	115
Size:	67.1 KB
ID:	80684  
BEC is offline  
Old 02-01-2014, 06:10 PM
  #120  
trebel-t91230l
New Member
 
Join Date: Jul 2007
Posts: 9
Default

Think of the prop as a rotating wing. with the brakes on and WOT {Wide open throttle) the blade angle and angle of attack are the same. As you release the brakes and roll forward the blade angle (on a fixed pitsh prop) remains the same but the angle of attavk lessens, therefore it does unload. This is the advantage of a constand speed prop. As the RPM tries to increase the blade angle increases (Moves toward high pitch) thus holding the RPM constant.
Hope this makes sense. Contact me if any questions
Dick
trebel-t91230l is offline  
Old 02-01-2014, 06:51 PM
  #121  
kyleservicetech
Super Contributor
 
kyleservicetech's Avatar
 
Join Date: Apr 2008
Location: Wisconsin, USA
Posts: 8,952
Default

Originally Posted by trebel-t91230l View Post
Think of the prop as a rotating wing. with the brakes on and WOT {Wide open throttle) the blade angle and angle of attack are the same. As you release the brakes and roll forward the blade angle (on a fixed pitsh prop) remains the same but the angle of attavk lessens, therefore it does unload. This is the advantage of a constand speed prop. As the RPM tries to increase the blade angle increases (Moves toward high pitch) thus holding the RPM constant.
Hope this makes sense. Contact me if any questions
Dick
Yup
Those props do unload during flight. No question about it. But, how much they unload depends on how fast the model is flying, and the blade pitch of the prop. When you make a full speed diving pass over the field, you can hear the propeller winding up. Which is unloading the motor.

ESC's such as the Castle Creations ICE series have data logging that shows this effect very clearly. Programs such as www.motocalc.com also have printouts and graphs that show the same thing.

This information can be quite useful for proper propeller selection. No sense to have a propeller whose prop pitch speed is 160 MPH, while your Piper Cub model is flying at 45 MPH. With this setup, there would be little or no prop unloading during flight.

The motor current on the models I fly typically unload on full throttle level flight to about 60-70% of full throttle current as measured on the ground.

Again, your results will vary.
kyleservicetech is offline  
Old 02-01-2014, 07:06 PM
  #122  
fhhuber
Super Contributor
 
fhhuber's Avatar
 
Join Date: Mar 2013
Posts: 4,786
Default

If you start out with the prop stalled (too high rpm for pitch and airspeed you stall the airfoil of the prop) then the prop can load up as the plane gains speed.

The goal in some cases used to be finding the prop that had the engine at full load at max level flight airspeed because that was maximum sustainable speed for the airplane.

So whether the prop unloads or not depends on the prop, rpm and the speed range you are looking at.

You can get a modern version of the old flexible white nylon props through Brodak. These are very interesting props. Under high load they "cone" forward and the pitch flattens. In essence an automatic variable pitch prop that gave some of the effect of a constant speed prop.
I used to watch the way the prop would change how much it coned as I performed aerobatics with control line models using the old Top Flite white nylon props we had to boil to keep them from becoming brittle.
fhhuber is offline  
Old 02-01-2014, 07:13 PM
  #123  
kyleservicetech
Super Contributor
 
kyleservicetech's Avatar
 
Join Date: Apr 2008
Location: Wisconsin, USA
Posts: 8,952
Default

Originally Posted by fhhuber View Post
If you start out with the prop stalled (too high rpm for pitch and airspeed you stall the airfoil of the prop) then the prop can load up as the plane gains speed.

The goal in some cases used to be finding the prop that had the engine at full load at max level flight airspeed because that was maximum sustainable speed for the airplane.

So whether the prop unloads or not depends on the prop, rpm and the speed range you are looking at.
.
Agreed.
I'm running my giant Big Stick model with a Hacker A60-16M motor, 19X12 APC-E wide blade prop, and a 12S2P A123 battery pack. You can actually hear the prop "Stalled out" at full throttle on the ground. And, as the model gains speed, hear that same prop bite in.

Plus, the Castle Creations ESC's data recorder shows the motor current increasing slightly from start of take off, to the model rolling forward. Then dropping off as the model gains flying speed.

Motocalc also plainly shows this same effect in the graphs and printouts on this same model setup.
kyleservicetech is offline  
Old 02-02-2014, 04:15 AM
  #124  
Larry3215
Look out for that tree!!!
Thread Starter
 
Larry3215's Avatar
 
Join Date: Mar 2006
Location: Gig Harbor, Wa USA
Posts: 7,061
Default

I never did finish the data analysis I was doing. Got side tracked some where along the line.

The conclusions I reached, from the data we all gathered 4 years ago, was that the power drop or power increase that someone could see in flight vrs static testing is highly variable.

You cannot simply say that all setups unload or that they all unload to similar degrees. There were very large variations in the degree of "unloading" shown in the data. From as little as 6% on up to possibly as much as 30%. Where ever I saw more than that degree of unloading critical data was missing. Unknowns such as throttle positions and maneuvering were missing and those two variables can make huge changes in the in-flight power levels all by themselves.

We didnt see much, if any, clear evidence of props loading UP in level flight, but there were some indications that it might have been happening. We didnt get to test enough of the right kind of setups - with adequate recorded data - to see that clearly. There were too many unknowns.

Its also very clear that whats going on with our electric models is much more complex than simple "prop unloading".

The difference in static power levels and in-flight power levels are highly dependent on several different factors: Motor heating, battery heating, power loading (how hard are you pushing the system - especially the batteries) as well as maneuvering (climbs/dives) all play significant roles. Thats not even considering different airframes, drag, etc.

One thing was consistent - there was at least some "unloading" going on in all the setups that were tested. Some not so much and some more so. For the most part, there didnt seem to be as much actual "prop unloading" as many people seem to think - but again, that was highly variable.

It was a very interesting investigation though and I want to thank everyone who contributed!!
Larry3215 is offline  
Old 02-02-2014, 01:11 PM
  #125  
rcers
Super Contributor
 
rcers's Avatar
 
Join Date: Jul 2005
Location: Trophy Club TX
Posts: 6,314
Default

Originally Posted by Larry3215 View Post
One thing was consistent - there was at least some "unloading" going on in all the setups that were tested. Some not so much and some more so. For the most part, there didnt seem to be as much actual "prop unloading" as many people seem to think - but again, that was highly variable.
That echo's my findings. Unloading in flight is variable, and the voltage drop is hard to differentiate from the "unloading" drop in current.

Square or over-square by far the most likely - on slippery ships.

I essentially account for little, if any, unloading and assume stalled static is about the right number to use for being safe with the equipment. If I was over on the ground I didn't use the setup, period. In other words I don't account for any unloading for safe margins with equipment.

Mike
rcers is offline  

Quick Reply: Do propellers really "unload" in flight?


Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information -

Copyright 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.

Page generated in 0.14241 seconds with 19 queries