Sport Flying This forum is for those that like to fly just for the enjoyment of flying, and all the airplanes that help them. Discuss here the joys of flying and the airplanes that help you enjoy it.

Accelerometers

Reply
Old 01-30-2007, 11:07 PM
  #1
Sneasle
Thread Starter
 
Join Date: Jan 2006
Location: Talladega, AL
Posts: 1,164
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default Accelerometers

I had a thought yesterday at work and did some digging. I was thinking that it would be cool to install an accelerometer in your plane and then plot the data to see what kind of forces we actually put our planes through. Then I got to thinking more and came up with the idea that if you use a 3-axis accelerometer and you know your sample rate you could 3-d plot your flight path. If you took it even farther and put one in each side of the wing and in the tail, you would have an even more interesting plot, especially for the guys that fly 3d and pattern stuff.

So I got to looking and found a product that is available that is made by Eagle Tree. It looks pretty cool, but all I am really interested in is the accelerometer. Granted, it has other cool features and lets you plot your flight through GPS, but GPS is only so accurate. Also, the accelerometer that it has is an attachment to the main unit, which is not what I wanted. I was looking for something small (small enough for 1/2a planes) that could be self contained during the flight and store its own data to be offloaded later.

I looked through some IC sensor suppliers and found a couple nice options, but I was wonder what ya'll thought about this. Do you think it is practical to try and put something like this together? It would only really be for me and the guys locally who would be interested, maybe if someone here wanted to try it I would send it out.

I am going to try and keep it within a package at or less then the size of a standard servo with bare minimum weight.

Also, since I am not familiar with all aspects of our hobby, what range of g forces do I need to be looking at? How high do I need to be able to measure? 2-3g? 10g? 18g? higher? There are several sensors available but they are expensive enough that I can't just by them randomly and try them out to see if they survive.


Here is the Eagle Tree site for anyone interested. http://www.eagletreesystems.com/
Sneasle is offline  
Reply With Quote
Old 01-30-2007, 11:15 PM
  #2
jlingrel
 
Join Date: Jan 2005
Location: Rio Rancho, NM
Posts: 421
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Sounds like an interesting project, and it would be interesting to see your results. I have used accelerometers on simulators for years and I have never seen one with a recording device that is very small. I am sure there are things available. Maybe even using telemetry and recording the results on the ground. Guys are using video that is sent to the ground for viewing. Keep us posted

John
jlingrel is offline  
Reply With Quote
Old 01-30-2007, 11:46 PM
  #3
Sneasle
Thread Starter
 
Join Date: Jan 2006
Location: Talladega, AL
Posts: 1,164
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

I considered sending the data to the ground, but the transmitter/receiver equipment would complicate the package more then storage the data on the device would. There are many micro controller on the market that have a relatively large capacity on chip. There should be one available that should work. Right now I am looking at the TI MSP430 series because of the development board that is available. I have already bought one and have goofed around with it and it looks pretty cool.

For people that need/want fast data sampling or just want to store data for long flights (gliders, endurance planes) there is the possibility of using flash memory off chip. The SD interface is said to be very easy to work with, and under the assumption that I have the data channels available after interfacing with the 3-axis accelerometer it should just be a simple circuit and a matter of code to dump data to the flash chip. The great part about this is that the flash chip could be removed and if I dump he data in an intelligent manner, be easily read as a text file on the computer. This would also make using the data for plotting much easier.

Sneasle is offline  
Reply With Quote
Old 01-31-2007, 10:41 PM
  #4
jlingrel
 
Join Date: Jan 2005
Location: Rio Rancho, NM
Posts: 421
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Keep talking, you really have my interest. I have often toyed with the idea of sending airspeed and altitude info to the ground and I am sure your project would lend itself to that.

John
jlingrel is offline  
Reply With Quote
Old 02-01-2007, 12:14 AM
  #5
Sneasle
Thread Starter
 
Join Date: Jan 2006
Location: Talladega, AL
Posts: 1,164
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Oh, it is very possible to send the telemetry to the ground, the Eagle tree system does that. But, that system is expensive, and while it is cool, I am only really interested in smaller amounts of data. The eagle tree system is very very complicated. I am mainly looking to come up with somethign that I can put together in my free time and assemble with the tools I have available. If I need to I will have pcb's made, but only if I can make them cost effective (ie, less then $10/unit) otherwise it won't be possible for me to do this.

I am not looking to turn this into a business, just something to fiddle with in my spare time and if it works, pass a few on to friends/club members to see if they enjoy having the extra data.

These parts I am looking at are very very low power, on the order of micro amps. They even have power saving modes. This is what really got me thinking about this because if implemented correctly, you won't have to worry about extra weight or power draw. I am shooting for under an oz all up weight, and I think it is very possible when these IC weight as little as a gram. The heaviest part I might have to deal with is the addition of a pcb if needed. If i don't need a pcb, then the other weight factor would be a small watch battery to power it. I have thought about putting a regulator on it and powering it off of the receiver. Hell, it would weight less then a servo and pull less power. I think most guys could deal with that little bit in most airplanes, even in 1/2a planes.

The biggest issue I think I will have to deal with is data storage. I am looking at my chip options to try and find something that can give me 2-5 samples/second and still allow me to store 10min. or so worth of data. If I can only get 5 min., it won't be that bad, but I really don't want to have to add in external storage if I don't have to because that requires me to have more data lines to work with, more code to implement, and overall a more complicated system.

I need to look more into the SD interface accessories (SPI interface I think technically), but there might be other options that I can add for people who might be interested in them.
Sneasle is offline  
Reply With Quote
Old 02-01-2007, 02:04 AM
  #6
jlingrel
 
Join Date: Jan 2005
Location: Rio Rancho, NM
Posts: 421
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Do you have web addresses for the people you are looking at using. I wouldl be inerested in it too. And if you are going to have PCBs made, I would be interested in that too. I have been in electronics for about 40s years and could possibly contricbute something to your efforts. For taht matter, we are not that far apart, I only live in Huntsville.

Keep me informed

John
jlingrel is offline  
Reply With Quote
Old 02-01-2007, 04:54 AM
  #7
HighPlains
 
Join Date: Mar 2003
Location: Over da rainbow, KS
Posts: 5,079
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

You can calculate the range of g force in a loop or tight turn if you know the airspeed and radius of turn. For instance, a pylon turn with a racing airplane is in the 30 to 40 g range.
HighPlains is offline  
Reply With Quote
Old 02-01-2007, 10:38 AM
  #8
Sneasle
Thread Starter
 
Join Date: Jan 2006
Location: Talladega, AL
Posts: 1,164
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

30-40g is kinda high.

What is the usual radius of a pylon turn and what speeds are we talking about? about?

John,

I will gladly keep you informed. I am not sure what you mean by web addresses for the people I am using. I assume you mean for the hardware I am looking at using.

The MC is made by TI:
http://www.ti.com/corp/docs/landing/...Other+OT+ez430

I have not yet decided on an accelerometer. I have a few different ones coming in the mail from 2 different companies. I have two different chips coming from Freescale Semiconductor (mma6262q, mma7260qt) and one from ST Mirco (lis3l02as4).

The TI chips are nice and cheap, and if they are usable for this, should make assembly of that part a breeze since they can be soldered on directly to headers and easily replaced if needed. It might even be possible to use the Target board profile as a template for mounts for other MC should the need for more data storage arise.

I tried to find some TH accelerometers, but they only ones I found were SMT and I had a hard enough time finding these few that were offered as samples. I have no problem switching to a different chip if needed, I just didn't want to waste the money upfront on experimentation.

Going back to assembly, I don't mind soldering the small stuff. I am qualified to do it (IPC J-stnd), but for ease of use and debug I would rather make it as easy as possible to swap parts around. If I get stuck using SMT accelerometers (which looks to be the case) I will plan on making a board similar to what is produced by Bellin Dynamic Systems (www.beldynsys.com) with the proper footprint that is as small as possible with only the traces and holes needed. I would then try and cram as many as I could onto one of the 2.5x3 in. budget boards that ExpressPCB does and cut them apart myself. If I can keep the cost to under $10 a board, I should be fine.
Sneasle is offline  
Reply With Quote
Old 02-01-2007, 11:43 PM
  #9
HighPlains
 
Join Date: Mar 2003
Location: Over da rainbow, KS
Posts: 5,079
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Quote:
30-40g is kinda high.

What is the usual radius of a pylon turn and what speeds are we talking about? about?
170 to 190 depending on racing class. 75 foot radius for a loose turn and as little as 50 in a overly tight turn. Simple physics. Throw in gust loading on top of it and you can see why some wings need to be strong enough to stand on.
HighPlains is offline  
Reply With Quote
Old 02-02-2007, 11:28 AM
  #10
Sneasle
Thread Starter
 
Join Date: Jan 2006
Location: Talladega, AL
Posts: 1,164
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

thats awfully damn fast. I shall need to make a waiver then that pylon racers stay away .
Sneasle is offline  
Reply With Quote
Old 02-02-2007, 01:24 PM
  #11
Charlie P.
 
Charlie P.'s Avatar
 
Join Date: Feb 2003
Location: Port Crane, NY
Posts: 5,088
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

40G? I doubt it. And even a 180 mph capable R/C pylon racer slows his model for the turns. You couldn't do a 40G turn because the control surfaces would stall and the plane would side-slip because of the plane's inertia in the direction of travel. Something would rip off first. Dive bombers found that if they forgot the dive brakes they would pull up and the plane would keep going down but at a nose level attitude (wings stalled) until they pancaked in with their bomb.

Here's a FAA report on G forces during maneuvers.

http://www.faa.gov/library/reports/m...ia/AM72-28.pdf

a 70º pull-out "blackout" maneuver puts 6 G's on a pilot. A "tight level turn" (like a pylon) is 8g.
Charlie P. is offline  
Reply With Quote
Old 02-02-2007, 01:30 PM
  #12
HighPlains
 
Join Date: Mar 2003
Location: Over da rainbow, KS
Posts: 5,079
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Quote:
Gravity is weak
As is your argument.
HighPlains is offline  
Reply With Quote
Old 02-02-2007, 03:25 PM
  #13
Charlie P.
 
Charlie P.'s Avatar
 
Join Date: Feb 2003
Location: Port Crane, NY
Posts: 5,088
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

OK. I have stood under pylons as a pylon referee and lap counter in maybe 25 Quickie 500 races and I see they decelerate for the turns.

I know a "hot" R/C pylon plane can achieve 220 mph on the straight aways. Lets look specifically at the US FD5 team. A good race time is 66 seconds to complete the 2.5 mile circuit (10 laps). If that plane is doing the 180 mph in the straight-aways, but the whole circuit averages out to 136 mph (do the math), then he is slowing down appreciably in the turns.

http://www.supertrc.com/teamUSAf5d/index.htm

How am I doing so far.

Anyone ever recorded a "g" reading higher than 20 in an R/C model without making contact with the ground?

The sled test used for helmets and car restraints is a 50g impagt - equivalent to hitting a solid wall

http://www.gforce.com/products/srs1/srs1.html


http://www.nhtsa.gov/cars/rules/rulings/80g/80giv.html

Frontal impact of a car traveling 30 mph and hitting a non-moving wall is 49g.

1) For unbelted drivers

Depowering by 20-35 percent had very little impact on unbelted driver chest g's in both 30 mph and 35 mph crash testing. In Model I-96, chest g's went up by 2 (49 to 51) in 30 mph testing and went down by 2.9 (61.6 to 58.7) in 35 mph testing. Modelling efforts by Ford show that chest g's should go up by about 2-3 g's with depowering in both instances. NHTSA modelling shows chest g's going up at 40.2 and 64.4 kph, but down at 56.3 kph (see Table III-13). It is believed

that the energy absorbing steering column is the reason that chest g's do not increase much on the driver side with the driver unbelted. Model I-96's test results are very similar to an average of the fleet in NHTSA compliance testing for unbelted drivers of 48 chest g's.

The calculated increase in fatalities using the Model I-96 30 mph test results are:

49 to 51 g's, baseline to depowered tests = a 2 g increase



Only 9 G difference between supposedly making a turn and hitting a wall at 30 mph? I don't think so.

I'll wait breathlessly for your counter examples of that 40 g turn. (I'd love to know how the receiver batteries are prevented from blowing out the side of the fuselage).
Charlie P. is offline  
Reply With Quote
Old 02-02-2007, 03:43 PM
  #14
HighPlains
 
Join Date: Mar 2003
Location: Over da rainbow, KS
Posts: 5,079
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

I can see from this and your other posts that you have a very limited understanding of physics. It looks like you have a better understanding of what is commonly known as “Roadrunner Physics”.

The distance flown on 2.5 mile course is typically in the range of 2.7 to 3.0 miles, plus the time it takes to accelerate from a standing start. The 2.5 mile figure is a point to point distance for 10 laps.

At a speed of 170 mph, you have a velocity of 250 ft/sec

The formula for acceleration in the turn is V^2/R

V^2 is 62500 ft^2/sec^2

Divide by 50 (ft) and you have 1250 ft/sec^2

To convert to g, divide that by 32.1 ft/sec^2

The result is 38.9 g
HighPlains is offline  
Reply With Quote
Old 02-02-2007, 03:58 PM
  #15
Charlie P.
 
Charlie P.'s Avatar
 
Join Date: Feb 2003
Location: Port Crane, NY
Posts: 5,088
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Beep! Beep! Physics 162 & 163. I don't make my living at Physics but I use it every day.

Extraodinary claims require extraordinary proof.

Where is your source for the 50 ft turn at 170 mph? I don't disagree with the G force that would generate. That's why I maintain it is impossible. I bet it's closer to 250 ft at 100 mph. There are folks who use the throttle other than ltake-off and landing. It's a challange. Those Quickees take 100 ft at maybe 40 mph in the turn. I have had my car up to 145 mph but never would try to turn it sharp at that speed. And tires have a lot better traction with the road than a wing in air.


Quote:
"Pay no attention to the man behind the curtain"
Humbug
Charlie P. is offline  
Reply With Quote
Old 02-02-2007, 04:11 PM
  #16
HighPlains
 
Join Date: Mar 2003
Location: Over da rainbow, KS
Posts: 5,079
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Quote:
There are folks who use the throttle other than ltake-off and landing. It's a challange. Those Quickees take 100 ft at maybe 40 mph in the turn. I have had my car up to 145 mph but never would try to turn it sharp at that speed. And tires have a lot better traction with the road than a wing in air.

I never realized in my 25+ years of pylon racing in Quickie 500, Formula One, Quarter Midget, Q40, and Giant Scale that all my competitors were throttling down to turn. Please continue with your antidotal evidence.
HighPlains is offline  
Reply With Quote
Old 02-02-2007, 04:50 PM
  #17
Charlie P.
 
Charlie P.'s Avatar
 
Join Date: Feb 2003
Location: Port Crane, NY
Posts: 5,088
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Let's keep it to flying and not personal attacks.

Where is your evidence? So far your claims are also "anecdotal"? (Antidotes are for posions; like of the mind).

I suited three sites that reference g forces and the relative ranges. Four if you would like the link to the one that gives the 300L spec of 10g as the performance parameter.
Charlie P. is offline  
Reply With Quote
Old 02-02-2007, 05:50 PM
  #18
RCKen
RCU Forum Manager/Admin
 
RCKen's Avatar
 
Join Date: Jul 2002
Location: Lawton, OK
Posts: 26,829
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Ok, time to step in here. It's one thing to carry on a discussion, but it's another when personal insults start flying. Let's keep the discussion civil in here as this issue is dicussed, otherwise I'll have to lock this thread and end the discussion.

Ken
RCKen is offline  
Reply With Quote
Old 02-02-2007, 06:27 PM
  #19
HighPlains
 
Join Date: Mar 2003
Location: Over da rainbow, KS
Posts: 5,079
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

Quote:
anecdotal
My bad, that is the correct word.

First off, in post 9, I said that a 50 foot is overly tight, so it is a slightly smaller radius than should be used, though every pylon model can achieve it without breaking. Most racers do turn tighter at pylon 1 than at the 2-3 turn.

If you are flying the course in the States, the distance between pylon 2 and 3 is 100 feet. The smallest radius and shortest course length that allows one to smoothly fly 2-3 in a single turn is approximately a radius of 71 feet. This gives a g loading of about 27 g in calm wind. However to avoid traffic, it is best to be able to turn inside of a leading aircraft to avoid wake turbulence. To do this decreases the radius of the turn.

Now if you increase the speed to 190 mph, which is another class of racing, and do that same 71 foot radius turn, you end up with a loading of 34g. A tighter turn at pylon 1 of 60 ft radius and you have a loading of 40.3 g. So you can easily see that loading on a fast pylon model can be quite high.

When I was racing 5 lb. Formula One models, they went around 175 mph on the course. After a crash where the wings were only slightly damaged, I would subject them to the following test. I would support the wings on a pair of blocks that were 46” apart and would stand on the center section. I weighed 170 at the time, and never had a wing fail in the test. You might observe that a 5 lb. airplane at 40 g’s would require up to 200 lbs. of force. However, the weight of the wing is distributed, so it does not concentrate at the center section. Removing 20 oz of the wing, but adding 8 oz for the fuel removes ¾ lb. for a total of 4 ¼, so at 40 g, you end up close to the 170 lbs. of my test. At one contest, about 20 % of the airplanes folded their wings to due high winds that were quite gusty. The wings collapsed in the turns when the airplanes experience a gust that changed the instantaneous angle of attack of the plane.

I have an engineering degree. Engineers live and die with numbers and calculations. When the numbers are right, people don’t notice. When they are wrong, airplanes have structural failures, bridges fall to the ground, and buildings collapse.

Did you every hear of the Rolling Airframe Missle? It was an air to air missle designed to be a lifting body, so that it could pull 100 g's in it's final moments to impact the target. While maned vehicles are limited to far less g loading, due to their slower speeds they can be hard for a missle to hit when they jink correctly. The RAM was designed to match that tighter radius at a much higher speed, thus requiring the increadible g loads.
HighPlains is offline  
Reply With Quote
Old 02-02-2007, 08:01 PM
  #20
Sneasle
Thread Starter
 
Join Date: Jan 2006
Location: Talladega, AL
Posts: 1,164
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

God lord, I leave this thread for a few hours and Ken has to step in.

Please gentlemen, break it up.

Both of you have very valid points, but you are arguing from different sides of the fence.

Highplains,

I am a physics and EE major and know nothing of pylon racing. However, I do know the speeds that our models can achieve, and If you are describing the course correctly, these models could very easily be pulling 40g's. I don't dispute your claims, I converted to SI units and ran the number myself this morning when I got to work. If indeed they are taking a 50ft turn at 190 mph, they can hit 40g's easily.

However, your force calculation in the last post isn't quite right. F=ma says that your force is equal to mass*acceleration. Your mass is 5lbs, your acceleration is 40g, not 40, hence your equation is 5lb*40*32ft/s^2, which equals to 6400 poundals. If you wish to use pounds force as your force measurement you need to have your mass in slugs.

SI is your friend. English sucks.



Charlie,

You also have good points, and had I not run these numbers myself I would not have believed them. I do not fly pylon, and I do not fly speed planes, I haven't even been flying long at all so I don't know what kind of speeds they do, but the math does work out to 40g's at 190mph in a 50ft turn.

I do not dispute the sources you site, but from the little I looked though them they seem to be based on human limits. Humans are weak, and a well built plane is far stronger. Most modern military aircraft (ie, fighters) can take forces much higher then the physical limitations of the human body, but very few humans can live through a 10g turn and be coherent enough to keep the plane in the air.


Guys, thanks so much for the.. lively.. discussion. As far as I can tell, you are both right, so lets drop the argument and figure out if this accelerometer is worth building?

Thanks for the help Ken, much appreciated, but please don't lock the thread .
Sneasle is offline  
Reply With Quote
Old 02-02-2007, 08:23 PM
  #21
Charlie P.
 
Charlie P.'s Avatar
 
Join Date: Feb 2003
Location: Port Crane, NY
Posts: 5,088
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

I was mostly wrong. I posted a query on the Pylon Racing forum and came up with members who claim 75 ft turns pulling 30 g's, so I concede to HighPlains slightly higher numbers. Now I am ASTOUNDED that someone could feed two commands into a servo and have it respond in less than 1/4 second for both commands. 50 ft at 220 fps for the entry and the exit to the turn means the whole maneuver is completed in just 0.22 seconds. Shazam! Talk about flying ahead of your model.

Quote:
Did you every hear of the Rolling Airframe Missle? It was an air to air missle designed to be a lifting body, so that it could pull 100 g's in it's final moments to impact the target.
AIM-9M Sidewinder and AIM-120 AMRAAMs too, but they're using a better Gov. Spec. grade of balsa and MonoKote than available to most civilian modelers.
Charlie P. is offline  
Reply With Quote
Old 02-02-2007, 08:46 PM
  #22
HighPlains
 
Join Date: Mar 2003
Location: Over da rainbow, KS
Posts: 5,079
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

I forget what the sidewinder can pull in g's, but it is not very high.

As far as control input, if you consider that a very fast reaction is about .25 second, then a 170 mph airplane travels about 60-65 feet. We don't react, we anticipate. The airplanes only seem scary fast when flown by themselves. In a racing heat, you start looking at relative speed to another equally fast airplane or two. Then they seem quite slow. A good lap takes 6 seconds give or take.

You are correct that the airplanes slow down in the turns. This is due to induced drag to provide that 120+ or ++ pounds of lift. This is offset as much as possible by increasing the wing span as much as possible - limited by rules or wing structure. Induced drag is inversely proportional to the wing span squared. A modern FAI F3D pylon ship has a span around 65" currently. Fifteen years ago, they were around 50"
HighPlains is offline  
Reply With Quote
Old 02-04-2007, 11:32 PM
  #23
Sneasle
Thread Starter
 
Join Date: Jan 2006
Location: Talladega, AL
Posts: 1,164
Gallery
My Gallery
Models
My Models
Ratings
My Feedback
Default RE: Accelerometers

You pylon racers must kick but at video games .

Maybe I'll make a version with a 50g accelerometer if anyone wants to put one in a pylon racer. Shouldn't be too hard, I know they can take that, just the sensitivity (mv/g) would be a bit lower.
Sneasle is offline  
Reply With Quote
Reply


Thread Tools
Search this Thread
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -8. The time now is 05:20 AM.