Air Data

GREAT Ideal...

like this F1 testing array of air speed sensors

quote:

Originally posted by comp2:
... I wanted to quantify it.
Gary

“when you can measure what
you are speaking about, …
you know something about it;
but when you cannot measure it, …
your knowledge is of a meagre
and unsatisfactory kind…” Lord Kelvin 1883

quote:

Originally posted by JFB #05177:
GREAT Ideal...

like this F1 testing array of air speed senso

Certainly looks high tech and would definitely work. Problem is who is going to provide the Pantera for the test?

I built a mount for the pito tubes. They are mounted 1 inch apart and should roll up in the window:







I worked out the programing on an 18f4550 PIC. It is over kill but I had a circuit I could set it up on. I have it all data-logging now but Ive set up a circuit on a more appropriate 18f2520. I ordered the circuit boards. Should get them back next week some times. Then I'll adapt the programing for that processor.

BTW the 18f4550 has the about the processor power of a TRS-80 from the 80's; for those of you who remember those.

besides the 5 dP sensors, will you include another type of speed reference our just use the outermost as your referance?

will the 5 dp sensors values be scaled before storage.

do have any pre ideal what the boundary thickness should be? The equations assuming a flat plate starting at the A post are not that hard (but I no longer comprehend to solve)

Well, not sure I am answering your questions as asked but let me try.

The data logging is quite basic. Each airspeed reference will give me a value. The value will be arbitrary unless I can associate it with a value. If we assume the outer sensor is in clean air, then we can assume it is representative of actual air speed.

I may run a sensor even further out to confirm speed is outside of disturbed air.

As far as calibration, one of the data markers is GPS. The GPS speed will correspond to the outer most value giving a point of reference.

What I expect is a guess. What I expect is the thought of a boundary layer is convoluted and not an accurate representation. A wing for instance, has a boundary layer where the air never separates under normal conditions.

I have little doubt the air is corrupt at the A-pillar which prevent a text book boundary layer. What we have is a bit misunderstood unless we put it in a wind tunnel. I believe the air is turbulent which could still be a turbulent low pressure at the inlet behind the door where the air inlet is.

That does not explicitly mean there is more or less air in the inlet, only un determined.

Having said that, I don't believe a piece of string can give you relevant information. I don't think a pressure reading can be overall meaningful.

What I think is meaningful is forward airspeed which is * * * Pressure with direction * * *.

What I hope is 2 fold. One is a baseline for what the airspeed is
1. out of disturbed air,
2. at the duct.

Having a large discrepancy, I have ideas on how to bring the air back in. Without having a way of measuring it, make any changes impossible to quantify.

Beyond that, I don't have a crystal ball.

quote:

Originally posted by JFB #05177:
besides the 5 dP sensors, will you include another type of speed reference our just use the outermost as your referance?

will the 5 dp sensors values be scaled before storage.

do have any pre ideal what the boundary thickness should be? The equations assuming a flat plate starting at the A post are not that hard (but I no longer comprehend to solve)

Photo in first post. Bottom right of photo is GPS.

Threw the mount on the truck window. Worked great. Still have to finish a small bracket in rear to keep it pressed forward.

On the Pantera this will put the pitot tubes right in front of the side vents.

Couldn't you get one more sensor (or move one on the array) into the cockpit, and use it as your reference (static) sensor?

Maybe the middle one initially (but I like your grid layout). Seems like you need another channel!

Rocky

Keep in mind this is airspeed which is pressure with direction.

Static will be determined before you move. In fact, it when the unit is turned on I will put a towel over the sensors to gain a zero air movement reference.

What ever the values come back to will be tied to the GPS speed. Assuming the outer most sensor is in direct airflow. The data received is an 8bit binary value. 8 bits returns a range of 0-255. This starts in the middle which is about 133. From what I can tell the range is 133-255.


If at 30mph the outer sensor reads say 175 . . then at 60 the outer sensor says 195 then these will represent actual reference values.

So, at 60 if the outer sensor says 195 and the inner sensor says 175 then we would know the direct airflow at the inner sensor is 30.

I think once the initial data is taken, I'll move the sensors back where they are spaced about the same but one is inside the duct. and the others just behind the lip of the duct. This would help understand how much the air speeds up over the duct and how much air is actually going in.

My duct for now is going into the engine bay so the air in the bay will influence the flow too.

those are Prandtl tubes (Pitot-Static), thus true air speed as they have their own static pressure reference.

given the actual sensed will be a delta pressure (total minus static) and velocity is NOT linear, will the 133-255 bits be dP or V .

I never expected it to be linear.

Static pressure is on the outside of the airplane, not inside. This assumes an airspeed indicator has nothing to calibrate to.

On these sensors, the second tube on the sensors is the so called static reference.

This is a data logger so I should get a corresponding GPS reference for a wide range of speeds. To get values for the speeds I want I just have to accelerate slowly so they can all be logged. The logging process is a automatic so all I have to do is flip a switch and drive.

No amount of computations will substitute for actual data reference between actual GPS and the value recorded by the airspeed sensors.

quote:

Originally posted by JFB #05177:
will the 133-255 bits be dP or V .

The value will be pressure but converted to speed albeit the logging of corresponding values.

quote:

Originally posted by JFB #05177:
those are Prandtl tubes (Pitot-Static, thus true air speed as they have their own static pressure reference.

By the way, it's actually indicated airspeed not true. It's only true airspeed at sea level.

Me too. Here's one for ya. I took this photo years ago. Check the difference between Ground speed and indicated airspeed:

544 knots btw is 625 MPH

So if I read that right, you're at 41,000 ft. altitude, 204 kts IAS, 544 kts ground speed, and 0.7 mach. Does your mach number compensate for altitude? What was the wind, if you remember?

PS didn't mean to get off topic - neat little set up on the air flow sensors.

quote:

Originally posted by UFO-LOW:
So if I read that right, you're at 41,000 ft. altitude, 204 kts IAS, 544 kts ground speed, and 0.7 mach. Does your mach number compensate for altitude? What was the wind, if you remember?

That is correct

I don't specifically remember tailwind but it was over a 100. Long time ago.

https://www.nxp.com/docs/en/data-sheet/MPXV7002.pdf

It's kind of irrelevant. Data loging will essentially create that map for me and I won't have to guess if it's accurate.

quote:

Originally posted by comp2:
I think once the initial data is taken, I'll move the sensors back where they are spaced about the same but one is inside the duct. and the others just behind the lip of the duct. This would help understand how much the air speeds up over the duct and how much air is actually going in.

My duct for now is going into the engine bay so the air in the bay will influence the flow too.

My testing was going into a sealed air box from one of the side scoops. It would have been nice to had air speed, but we used pressure, which was positive up to 120 mph. I would have taken the car up to 140 mph, but there was some traffic on the road which prevented me from doing that. The speed is accurate because I have an electronic speedometer and it matches the speed when checked against a GPS.

You may find that your air speed will be less than the air speed on the side of the car when you have the side scoop exposed to all of the air from the engine bay.

I built a circuit dedicated for the project. I went from an over sized 18f4550 PIC to an 18f2520. It's a cheaper processor more appropriate for the project.

The next board the GPS will be on the board instead of an added board.

Next I'm going to 3d print a half case with LCD mount. It will have switch mounts so I can start and stop the recording process.

The SD card buffers are a challenge to solder. Their suppose to be oven flowed but hand soldering works. The resistors on the SD card board are even a lot smaller.

The circuit if functioning quite well now. Turn on and GPS initializes. When the data starts to flow the SD card is initialized. A txt file is created an opened.

Bottom right button starts and stops the data logging. Left button stops the data logging, closes the file and opens a new file for data logging.

The top button stops the data logging, closes the file and stops the program.

The circuit is a bit of a mess. I had to add another board to interface the components so I am redesigning the board to clean it up. I have some other applications for this circuit so it's worth the expense of new boards. It's about $135 to get 4 boards made so the more things I can set them up for the better.

Nice work!

I know there is a lot of interest your data.

Rocky

I've still got work to do but I wanted to see what the data would look like. The weather broke a a little today so I threw it in the truck. To be fair, I still need to do a few things (besides the new circuit). I need to mount it on the front of the truck so I can calibrate all sensors so I can make the sensor range the same.

The truck is going to be nothing like the Pantera. The truck has a rain deflector at the window and elephant ear mirrors so I expected a bit of a turbulent airflow.



Notice the high tech plastic bucket.





Here is the data with some corrections:






That's a little confusing so I reversed it and zeroed it in the spreadsheet before re-charting it. This makes a little more sense:



The actual value is irrelevant. What the air is doing and how it compares is relevant. Sensor 1 is 1" from the window. Each is spaced 1". Sensor 5 is 5" from window. Speed is 0-75mph.

What I can surmise is the air is quite turbulent along the side of the truck. I would not call this a laminar flow.

To make a little more sense of the jumbled data I looked at #1 and #5:



Here's the rest for comparison:

http://www.rc-tech.net/cars2/p...ransam/air/Data3.png

http://www.rc-tech.net/cars2/p...ransam/air/Data4.png

Sensor #4 and #5 almost mimic each other and #5 almost seems slightly less erratic....slightly:

Critical things for a good test will include a drive on a non windy day and with no cars around to influence the air.

If the air was mores stable, there is something else one could surmise:



If sensor #1 was smoother and averaged around "20" (Vertical number) at 70mph then you can compare that value corresponded to about 40mph using sensor #1.

cool...
would you mind sharing the raw data (speadsheet?)

I threw it in the directory the photos are in.

I built an updated circuit. It's a lot more clean:

OK...for the Layman...what the Hell does all the squiggly lines mean? lol

Ill exlain more later but its airspeed graphed to gps speed.

Nice work on the development of the instrumentation, and the electronics.

Very impressive.

I can barely keep my '74 vintage vehicle electronics running - you are building the kind of instrumentation used by aerospace corporations as a hobby!

Keep us posted!

Rocky

Waiting on spring weather.

While waiting for better weather I adapted the same circuit for another project. This circuit is a visual cruise control. The buttons set the speed. The lights light up or go out to indicate how close you are to your target speed. White lights on the left indicate speed up. Red lights on the right indicate indicate slow down.

The idea is, instead of constantly looking at the speedometer, you have a light indicator in the field of view; perhaps on a visor.

Just something I am playing with. One thing that suprised me was how much the cruse control in my truck floated speed. I thought it was pretty good at holding a steady speed. On a highway I think it does but on a road with minor dips it was fluctuating +/- 2mph.

Perhaps set it to reflect as heads up display on windshield. Cool idea. You are extremely talented. Wish I was that smart!

What I am doing is not as hard as it seems. If you like that sort of thing this is a great place to start. It's open source and easy to workwith:

https://learn.sparkfun.com/tut...s/what-is-an-arduino