What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Help with measuring cowl pressure differential

lostpilot28

Well Known Member
I didn't give this much thought before I purchased a vacuum/pressure gauge. It's one of those single port devices that measures negative or positive pressure from 0 to 10 psi or 0 to 30 in/Hg. Exact model is here.

My plan was to place tubes at various locations under the cowl to read the pressure at that specific spot. I could run several lines, label them, then swap the gauge to a different hose in the cockpit while flying. But, I'm not sure if this instrument is going to work. Maybe it will....it just seems to need a pretty good amount of suction to move the dial, and I'm not convinced I'll be able to read that much variation under the cowl.

What do you guys think? I really don't want to build a manometer for inside the cockpit, so the dial idea was ideal. Any help is greatly appreciated.
 
Nor is it sensitive enough....1 inch Hg is 13.596 inches of water. Differential pressures under the cowl are often just a few inches of water. Heck, 100% dynamic pressure at 200 knots and sea level isn't quite 2" Hg and less than 1 psi.
 
Thanks guys...once I got it I sort of realized it wouldn't be sensitive enough. But, Bob, what do you mean about it being "absolute"? Couldn't I use a (more sensitive) gauge to measure the absolute pressure at various locations under the cowl, then compare the results?
 
Actually its gauge pressure, not absolute pressure. Gauge pressure means one side of the gauge is vented to ambient. And unless the gauge (if it were sensitive enough) would have to be in the area you are comparing for it to have a meaningful reading.

What you need is a differential pressure gauge that is sensitive enough. Something like a Dwyer Magnahelic.
 
Last edited:
I'm guessing something like this is the right thing???

http://www.transcat.com/Catalog/productdetail.aspx?itemnum=MAN02-A&utm_source=google&utm_medium=base

I think the difference is measuring relative (AKA: differential pressures) vs absolute pressures (not relative differentials). I don't know, but I'm just guessing that's the idea since this device measures differential.

I might snag one if it's the right tool. It sounds like something that could be worthwhile and save a whole lot of frustration.

Phil
 
Last edited:
Actually its gauge pressure, not absolute pressure. Gauge pressure means one side of the gauge is vented to ambient. And unless the gauge (if it were sensitive enough) would have to be in the area you are comparing for it to have a meaningful reading.

What you need is a differential pressure gauge that is sensitive enough. Something like a Dwyer Magnahelic.

Absolute pressure is zero-referenced against a perfect vacuum, so it is equal to gauge pressure plus atmospheric pressure.
Gauge pressure is zero-referenced against ambient air pressure, so it is equal to absolute pressure minus atmospheric pressure. Negative signs are usually omitted.
Differential pressure is the difference in pressure between two points.

So, why can't I use Absolute or Gauge pressure and read either at various locations under the cowl? That would provide the data I need. I really don't care about the values as much as I care about point A being higher/lower than point b, c, d, and e. If I'm just getting a number at each location, wouldn't I know how air is flowing in and out of my cowl?

Sorry for my ignorance...just trying to find an easy solution that I don't have to "re-rig" for several different test flights. And, I don't want to pay $200+ for an instrument (sorry Phil :eek: ).
 
Yeah, I hear ya.

I just wrote the check for a new IO-540. So from the new IO-540 perspective, $200 is a really cheap way to get some really helpful data.

I agree it's painful, but I don't want to sink any more money into engine repairs when I could just take care of the really expensive one I own today. So I don't mind paying an extra $200.

On a side note:
$200 is pretty easy to make up in fuel when comparing a trial and error approach with using instruments to collect exact data. It's probably close to a wash. :)
 
Last edited:
Again, ask your hvac guy to go flying with you or rent his. We use them all the time in this field for checking pres drops across coils/filters/rtn & sup ductwork, gas pres, proving switches. If you were closer I would go up with you. You really need a low and high pres port to be accurate.

Yes, you can reference cabin pressure. Take readings from two areas and you have your difference. Not as accurate as your cabin pres may change in between readings. You still need a gauge that reads "In H2O" and is accurate around 0-10".
 
Again, ask your hvac guy to go flying with you or rent his. We use them all the time in this field for checking pres drops across coils/filters/rtn & sup ductwork, gas pres, proving switches. If you were closer I would go up with you. You really need a low and high pres port to be accurate.

Yes, you can reference cabin pressure. Take readings from two areas and you have your difference. Not as accurate as your cabin pres may change in between readings. You still need a gauge that reads "In H2O" and is accurate around 0-10".

Great find, Chris! That's more in line with my budget!

So, quick question about measuring differential pressure. One of the primary things I'm trying to learn about is my oil temps. If I wanted to know how air is flowing through the oil cooler, would I place the ends of the tubes in front of, and directly behind the oil cooler? Or, would it be better to just measure the general area (upper left side of engine & somewhere just behind the left of the engine)? In reading about differences in local pressure areas, it seems like you have to be careful with probe placement.
 
Last edited:
If it were me I'd tap one side into my static line since presumably you've calculated your static errors. This will give you a consistant ambient reference when you factor in your error calcuations.

to measure cowl pressures you have two choices -

1. Local Ambient
2. Local Total or Stagnation Pressure

For local ambient place the other pressure line orthogonal to any flow at the location you are interested in.

For local total pressure place the probe facing into the flow just like your PITOT tube.

By having the reference tied into your static line you'll also be able to compare undercowl flow to your airspeed (using the Local Total Pressure technique), which would be particularly interesting at the inlet face, cowl exits, etc...
 
So, why can't I use Absolute or Gauge pressure and read either at various locations under the cowl? That would provide the data I need.

Because it will be more of a pain to measure drop across a cooler, for example. You'd have to do one flight with a hose in front of the cooler, then another flight with the hose behind the cooler. Then find the difference. Its easier to do differential pressure with two lines at the same time.

I'm thinking the accuracy of the digital manometer might not be up to task...its listed as +-.3% across the range of +-55" water, which the error could be significant since the range to measure pressure drop is just a few inches.
 
Magnahelic

As Wayne mentioned we HVAC guys use them all the time.
Since you are measuring pressure differential I would think the pressure difference would be very small. Probably .25" to .5". The total pressure could be as high as Dan stated but we don't care about total pressure only differential. I would think that the digital that was shown would not be sensitive enough. 2 psi is 55". This unit would not be sensitive enough to do what you want. Will you be running the hoses back into the cockpit? If so you could do this at almost no cost. Especially since you don't know how sensitive you need to buy.
The Dwyer magnahelic is the one to use but you buy them with different scales.
http://www.dwyerinst.com/Product/Pressure/DifferentialPressure/Gages/Series2000
This one goes to 1.0" which is probably about what you need.
A simple hose loop with water in the loop would also do the trick. That is all you are doing is measuring the water column pressure drop. If you have a hose loop with water in it and you blow light pressure air on one side and blow less on the other side the water will move up on one side of the loop. When you measure that rise or fall that tells you the inches of water column pressure. in "in.of water.
 
I'm thinking the accuracy of the digital manometer might not be up to task...its listed as +-.3% across the range of +-55" water, which the error could be significant since the range to measure pressure drop is just a few inches.

There is a "point" in front of that 3, Bob. ;)

I just bought one...I noticed that you can use it for differential and gauge pressure. My plan is to run several tubes to the cowl area and take all the readings on one flight. We'll see how it goes...

:eek:
 
I would think that the digital that was shown would not be sensitive enough. 2 psi is 55". This unit would not be sensitive enough to do what you want. Will you be running the hoses back into the cockpit? If so you could do this at almost no cost. Especially since you don't know how sensitive you need to buy.

Hi Mark, the range is 2 psi, not the sensitivity. The resolution of the meter is listed at .01 for the in/H2O function.
 
If you don't need too many decimal places, an airspeed indicator is a great differential pressure gauge:

knots.......psid...........in H20
0.............0.0000.......0.00
10...........0.0023.......0.06
20...........0.0094.......0.26
30...........0.0210.......0.58
40...........0.0374.......1.04
50...........0.0584.......1.62
60...........0.0842.......2.33
70...........0.1145.......3.17
80...........0.1496.......4.14
90...........0.1893.......5.24
100.........0.2338.......6.47
110.........0.2828.......7.83
120.........0.3366.......9.32
130.........0.3950.....10.93
140.........0.4582.....12.68
150.........0.5259.....14.56
160.........0.5984.....16.56
170.........0.6755.....18.70
180.........0.7574.....20.96
190.........0.8438.....23.36
200.........0.9350.....25.88
210.........1.0308.....28.53
220.........1.1314.....31.32
230.........1.2365.....34.23
240.........1.3464.....37.27
250.........1.4609.....40.44
 
Since you are measuring pressure differential I would think the pressure difference would be very small. Probably .25" to .5".

No, higher. Actual measurements from my notes, using a water manometer (yardstick and vinyl tubing):

Sunday 4-10-11 - Manometer on oil cooler entrance and exit. Manometer static was 18". Differential pressure = (reading -18) x 2

(reading 18.5) 1" 1450 RPM on ground

(reading 20.75) 5.5" 100 knots at 25/2500

(reading 24.125) 12.25" 168 IAS 188True 24/2400

196F oil temp 66F OAT
3.34V 142F exit air

After a few minutes 208F oil temp 67F OAT
3.37V 147F exit air
 
Diiferential pressure

I thought the point was to measure the pressure before a component and the pressure after giving a differential. Not just the inlet pressure against the static pressure.
In residential A/C we deal in about .5" of external static in a duct system and that is about 800 fpm. If I am right 188kts is 19,050 fpm. WOW.
I would still think the diiferential across a component would be less than 2.0".
Hey that would make the Chinese magnahelic perfect!
Good discusion.

My question is how does this help plenum, baffle or cowl design?
Do you want more or less?
 
Mark,

With higher velocity air like my -10 oil cooler gets through a 4" X 15" duct there will be a large pd.

In hvac systems we try to keep pres drops low for less noise, humidity removal and filter/total system efficiency. Total external static pres (esp)for residential 0.3-0.9" H2O. Higher for commercial systems.

In airplanes where we desire to keep cylinders and oil cool you want high pd. Something like 6"+ between top and bottom of cylinders. I will get some numbers after flying for all of you technical types like me.
 
If it were me I'd tap one side into my static line .....

Serious question; I know you can do it with three lines, but why not just connect the two measurement points directly to the meter? It requires a lot less meter range and one less tube.

For local ambient place the other pressure line orthogonal to any flow at the location you are interested in.

Hard to predict which direction is orthogonal in the plenum space. I'd connect to a piccolo tube, or shove an aquarium bubble rock in the end of the tube....anything to negate the effect of air motion. I have even wrapped the end of the tube in a wad of shop rag and masking tape.

NASA 3405 has a section of experiments with probe installations, and this SA article from our friend Chris shows a piccolo tube install:

http://www.rv-8a.net/106-111_BuildingBasics Cooling Drag.pdf

By having the reference tied into your static line you'll also be able to compare undercowl flow to your airspeed (using the Local Total Pressure technique), which would be particularly interesting at the inlet face, cowl exits, etc...

Can you explain the Local Total Pressure technique?

I'm just a student, so aero gurus please check me. I would reason that measuring local velocity inside the plenum (or in a duct leading from it) would require something like a coaxial pitot-static tube, ie a static source also inside the plenum, very local to the measurement.

Total pressure = static pressure + dynamic pressure. Think 11 = 10 + 1. Static and dynamic can trade back and forth.

A normal closed-end pitot-static system opposes total pressure with free stream static pressure to determine dynamic pressure. Think 11 - 10 = 1

Slowing the flow into the plenum trades some dynamic for increased static. The flow does not come to a complete halt so some dynamic remains (which we wish to measure). A pitot in that plenum sees total pressure just like the normal pitot. If you oppose it with free stream static, you'll again get 11 - 10 = 1, but the truth is more like 11 - 10 = 0.7 + 0.3.....0.7 representing the potion of the original dynamic transformed into increased static, and 0.3 being the remaining dynamic. When translated to airspeed the value will be too high. A coaxial static source would give you 11 - 10.7 = 0.3, the real value of the dynamic pressure.
 
Last edited:
static pressures around the airframe

DAN H is 100% correct in what he's saying. I tested the pressures around the door on my RV-10. (at 140 kts indicated) The pressure inside the cabin was -2.0" with respect to free stream static pressure.

the pressure differential outside the door vs cabin pressure was 7" water column. (36.4 lbs/sq.ft.)

Now it's obvious why the doors must be properly latched before t/o.

0-15" dwyer magnehelic is the gauge of choice-- I use them every day.

warren
 
There is a "point" in front of that 3, Bob. ;)

I just bought one...I noticed that you can use it for differential and gauge pressure. My plan is to run several tubes to the cowl area and take all the readings on one flight. We'll see how it goes...

:eek:

Since the accuracy is over 110" (fso = full scale) at the pressures you'll be reading the error will be significantly greater.
 
Serious question; I know you can do it with three lines, but why not just connect the two measurement points directly to the meter? It requires a lot less meter range and one less tube.



Hard to predict which direction is orthogonal in the plenum space. I'd connect to a piccolo tube, or shove an aquarium bubble rock in the end of the tube....anything to negate the effect of air motion. I have even wrapped the end of the tube in a wad of shop rag and masking tape.

NASA 3405 has a section of experiments with probe installations, and this SA article from our friend Chris shows a piccolo tube install:

http://www.rv-8a.net/106-111_BuildingBasics Cooling Drag.pdf



Can you explain the Local Total Pressure technique?

I'm just a student, so aero gurus please check me. I would reason that measuring local velocity inside the plenum (or in a duct leading from it) would require something like a coaxial pitot-static tube, ie a static source also inside the plenum, very local to the measurement.

Total pressure = static pressure + dynamic pressure. Think 11 = 10 + 1. Static and dynamic can trade back and forth.

A normal closed-end pitot-static system opposes total pressure with free stream static pressure to determine dynamic pressure. Think 11 - 10 = 1

Slowing the flow into the plenum trades some dynamic for increased static. The flow does not come to a complete halt so some dynamic remains (which we wish to measure). A pitot in that plenum sees total pressure just like the normal pitot. If you oppose it with free stream static, you'll again get 11 - 10 = 1, but the truth is more like 11 - 10 = 0.7 + 0.3.....0.7 representing the potion of the original dynamic transformed into increased static, and 0.3 being the remaining dynamic. When translated to airspeed the value will be too high. A coaxial static source would give you 11 - 10.7 = 0.3, the real value of the dynamic pressure.

Certainly many ways to skin this cat. I like having a common reference point which would be the airframe static source and the closest thing to ambient atmospheric pressure at the test condition. This allows one line to remain connected to the instrument and all sorts of others placed anyway you want inside the cowl. At the test condition simply swap the non-static line connections and read the differential pressures relative to ambient.

Agree determining orthogonality to flow is tricky and a piccolo tube would work - but probably overkill for most areas. I think the flow though an oil cooler duct would be fairly easy - not so much half way between the cyl and the cowl exit though.

You need a state equation Pv=nRT, a continuity equation or Mass flow in = Mass flow out (density in*Ain*Vin = density out*Aout*Vout) , the type of gas, and most likely a conservation of energy equation to account for heat transfer and flow losses to fully characterize the flow - if I had my Fluid Dynamics book by Zucker unpacked I'd look it up.

the required variables would be Ptotal, Ttotal, Area ratio, T, Pstatic, and a few ratios (gamma, sigma) and constants (gas constant) which can be looked up in tables.

a coaxial will certainly do the trick and in a sense I suggested it already by placing a probe orthogonal to the flow and one facing the flow. Subtracting those two dPs eliminates Pstatic.

If I measure 11 as Local Total Pressure directly
and 10.7 as local static under the plenum directly

Subtracting the two measurements gives me .3 its the same thing you are saying except I only have to change one tube connection from the instrument to measure readings throughout the cowl the other stays connected to the static line - anything making the pilots job easier while flying is good for me :)

Its technique only.





Ppdyanmic = Ptotal Plenum - Pplenum static

Pplenum static
 
....anything to negate the effect of air motion. I have even wrapped the end of the tube in a wad of shop rag and masking tape.

Hey Dan, no clue what a piccolo tube is...but, my MP line from the #4 cylinder has a plug in it with a very tiny orifice. Without it, my MP bounces around like crazy. Do you think doing the same thing on the end of a plastic tube would provide the same results?
 
Since the accuracy is over 110" (fso = full scale) at the pressures you'll be reading the error will be significantly greater.

Less than .4 inch/H2O...I can live with that. My guess is that if there is error, it's not going to swing wildly from 0 to the .4 with each reading. It'll most likely be "off" by somewhere between 0 and .4 and pretty much stay there. I take it you don't like cheap Chinese tools. ;)

It's instrument error...you have that with every instrument.
 
If I measure 11 as Local Total Pressure directly
and 10.7 as local static under the plenum directly Subtracting the two measurements gives me .3 its the same thing you are saying...

Understood. That would work, with a caveat...the static probe would need to be very close to the pitot. The pressure maps in 3405 show significant variation in pressure depending on probe position within the plenum. For example, I'm not convinced you can put a pitot in the cowl inlet and take static as an average with a piccolo in the mid plenum region.

.....except I only have to change one tube connection from the instrument to measure readings throughout the cowl the other stays connected to the static line - anything making the pilots job easier while flying is good for me :)

Yes, it would be easier if the test plan called for measurements from multiple locations during in one flight.

However, if the test flight was to measure a velocity only, two tubes from a coaxial pitot-static would hook directly to an airspeed indicator.....no tube swapping, no calculation, no position error, just a direct read of speed. And it would be a cleaner install, ie less stuff in the airflow to skew the readings.

Built this one for a round motor biplane some years back; tube within a tube.

ort37l.jpg


Hey Dan, no clue what a piccolo tube is...but, my MP line from the #4 cylinder has a plug in it with a very tiny orifice. Without it, my MP bounces around like crazy. Do you think doing the same thing on the end of a plastic tube would provide the same results?

No, but you're close. Take a 6" length of 0.1875 tubing, brass or aluminum. Crimp one end shut. Drill a row of tiny holes all the way through the tube, roll it 90 degrees, drill another row. Connect to a hose. That's a piccolo tube.
 
Understood. That would work, with a caveat...the static probe would need to be very close to the pitot. The pressure maps in 3405 show significant variation in pressure depending on probe position within the plenum. For example, I'm not convinced you can put a pitot in the cowl inlet and take static as an average with a piccolo in the mid plenum region.



Yes, it would be easier if the test plan called for measurements from multiple locations during in one flight.

However, if the test flight was to measure a velocity only, two tubes from a coaxial pitot-static would hook directly to an airspeed indicator.....no tube swapping, no calculation, no position error, just a direct read of speed. And it would be a cleaner install, ie less stuff in the airflow to skew the readings.

Built this one for a round motor biplane some years back; tube within a tube.

ort37l.jpg




No, but you're close. Take a 6" length of 0.1875 tubing, brass or aluminum. Crimp one end shut. Drill a row of tiny holes all the way through the tube, roll it 90 degrees, drill another row. Connect to a hose. That's a piccolo tube.


concur. Nice work Dan as usual.
 

Ha...now you have me thinking and digging in the hardware bins. You know, the place where workshop packrats store away odd bits because they might be useful someday?

I already installed three miniature bulkhead fittings, so I have three tygon lines available from the engine compartment. As you said.....

anything making the pilots job easier while flying is good for me

.....so how about three pneumatic toggle valves connected to a manifold?

v3lh75.jpg


I'll still try the direct read coaxial pitot-static for velocity, but the above should make pressure measurements very quick at each test speed.

I told you it was a serious question ;)
 
Less than .4 inch/H2O...I can live with that. My guess is that if there is error, it's not going to swing wildly from 0 to the .4 with each reading. It'll most likely be "off" by somewhere between 0 and .4 and pretty much stay there. I take it you don't like cheap Chinese tools. ;)

It's instrument error...you have that with every instrument.

Lets say you have a pressure gauge that measures 0-100 psi and the accuracy is +-1% FS, which is 2psi total error. Now lets say you are measuring a range of pressures from 0 to 5 psi. That equates to 40% accuracy in that narrow range, which is not good at all. Same thing applies here. If you are measuring 0-5" then you will have an accuracy of 13.2% and thats ignoring resolution or hysteresis errors based on the specs of the manometer if you believe them. If you got one that had a narrower range you would have far better results.
 
Lets say you have a pressure gauge that measures 0-100 psi and the accuracy is +-1% FS, which is 2psi total error. Now lets say you are measuring a range of pressures from 0 to 5 psi. That equates to 40% accuracy in that narrow range, which is not good at all.

You sure it's not 2 psi across a range of 100, thus 1/20 of 2 psi across 5?
 
You sure it's not 2 psi across a range of 100, thus 1/20 of 2 psi across 5?

No, it'd be 2 psi across a range of from 1-5. In any event, mil standards want you to avoid the top and bottom 20% of any gauge while measuring...stick with the middle 60. So I'd never measure 5 psi on a 100 psi reference gauge.
 
Lets say you have a pressure gauge that measures 0-100 psi and the accuracy is +-1% FS, which is 2psi total error. Now lets say you are measuring a range of pressures from 0 to 5 psi. That equates to 40% accuracy in that narrow range, which is not good at all. Same thing applies here. If you are measuring 0-5" then you will have an accuracy of 13.2% and thats ignoring resolution or hysteresis errors based on the specs of the manometer if you believe them. If you got one that had a narrower range you would have far better results.

Understood...but it's .3% across a range of 110 in/H2o. That's 3x better than what you're analogy is using. Maybe it's not ideal, but I'm OK with it. From what I've seen, the erros in digital instrumets are usually one way or the other (not randomly selected each time you measure). I don't really care what the values are...I want to know the difference between the values. I don't care if my upper plenum has +10 in/H2o or if it's +8. As long as it reads the same each time for the same test conditions.
 
No, it'd be 2 psi across a range of from 1-5.

I think you're saying the spec refers to offset error rather than gain error.

vqo7kl.jpg


Obviously I was thinking gain. However, I suspect offset is the correct view because a close look finds "FSO" in the spec:

Accuracy: +- 0.3% FSO @25C

Does FSO stand for "Full Scale Offset"?
 
I think you're saying the spec refers to offset error rather than gain error.

vqo7kl.jpg


Obviously I was thinking gain. However, I suspect offset is the correct view because a close look finds "FSO" in the spec:

Accuracy: +- 0.3% FSO @25C

Does FSO stand for "Full Scale Offset"?

My apologies, I'm not educated on instrument calibration. ;) Can you explain this in layman's terms? I can make a guess about the error being smaller the closer you are to "zero", but I'd probably be wrong.
 
I think FSO stands for Full Scale Output.

Thank you Chris. Leads me back to thinking it's a gain spec, but no matter.

My apologies, I'm not educated on instrument calibration. ;) Can you explain this in layman's terms?

I too am learning, but this part isn't hard.

Offset is a constant error. The readout is always some fixed amount too high or too low.

Gain is a multiplier. The error gets proportionally larger as you move up the scale. Gain error is listed as a percentage of the full range value ("FSO"), so the possible gain error for any given application is (percent of range being used multiplied by gain value).

Apparently the good instruments come with both an offset and a gain percentage stated in the specs. So you calculate possible gain for the part of the range you plan to use, then add the stated offset to know the total possible error.

In the case of the Chinese meter, we don't know if the stated 0.3% FSO is offset or gain. Doesn't matter. Check it with a tube full of water. It may be perfectly accurate. Or the 0.3% could be a total lie and error is far worse.

Me? I've been flying around with tubing zip-tied to a yardstick ;)
 
Last edited:
Dan, you're awesome. Thanks for that explanation. I may indeed test the instrument prior to flying with it just to see how far off it is. I still think that getting consistent results is better than actual numbers in this case because I just want to know the difference between two values. I hope that's not out of line - and if it is, then I'm sure someone will correct me. :D

Thanks again!
 
What are you going to do with the data once you get it? What is good...what is bad?:confused:

Dave (Swift driver) and practical engineer in the pressure and flow measurement business for 37 years.
 
Ha...now you have me thinking and digging in the hardware bins. You know, the place where workshop packrats store away odd bits because they might be useful someday?

I already installed three miniature bulkhead fittings, so I have three tygon lines available from the engine compartment. As you said.....



.....so how about three pneumatic toggle valves connected to a manifold?

v3lh75.jpg


I'll still try the direct read coaxial pitot-static for velocity, but the above should make pressure measurements very quick at each test speed.

I told you it was a serious question ;)

wow those switches are great!

I have a sneaky suspicion your "hardware bins" are far more organized and varied than mine are but I have one or two and always go there first when dreaming up a new project.

For me when these ideas come up I'm always thinking from a test pilot perspective rather than pure engineering - discussions with test engineers and test pilots can be lively but the pilots always win :) (well almost) I like to collect the most data I can on any given sortie to ensure test conditions remain as stabilized as possible rather than having to land, reconfigure and re-fly.

Is see benefits of measuring velocity with the coaxial as you describe as well. It would be a nice cross check to my method and vice-versa.

I've been researching self contained pressure/temp devices which record their measurements in non-volatile memory including time-stamps for download after the test. I'd like to place these throughout the cowl, fly, remove them and download their data. No wires, spaghetti hoses, or water in the cockpit :) Unfortunately I have not found anything under about $400 each and I don't get paid quite that well.

This is a great topic!

Next we need to discuss temperature measurements....
 
Gain or offset errors are what you get when you have some sort of sensor that amplifies a small voltage to a larger one appropriate to the range of the internal analog to digital converter that is translating the voltage levels to a digital number. Opamps have distortion numbers, gain errors, etc. I doubt highly this is a consideration with the sensor used in the manometer.

A water manometer has natural dampening due to surface tension and I believe it will be more stable and accurate than the ebay manometer. The digital manometer is worth a try though, let us know how it works...but I ain't gonna buy one :). I doubt that it has the digital filtering necessary to give you a stable reading...another consideration on top of the accuracy numbers. If it had a range of 0-15" H2O with 0.3% accuracy that would be far, far better.

I have a drawer full of cheap digital multimeters that have good specs but in practice I don't touch them since they are not nearly as accurate as some of the benchtop equipment I have....same idea here.
 
Last edited:
What are you going to do with the data once you get it? What is good...what is bad?:confused:

Dave (Swift driver) and practical engineer in the pressure and flow measurement business for 37 years.

Hi Dave...I have no idea! Actually, I have some idea, but I'm sure I'll be looking for guidance. What I know is from reading the other posts about cooling drag and efficient cooling of oil and CHTs. My engine runs a tad warmer than what I'd like it to, even after a few modifications. I've managed to bring my oil temps down to around 205 - 210 (F) in cruise on a hot summer day...but I'd like it to be under 200 in cruise.

My whole point to this exercise is to see whether the lower cowl pressure isn't different enough from the upper plenum pressure. If it's not, then I can start figuring out a way to increase exit air flow.
 
This is a great topic!

You bet.

Next we need to discuss temperature measurements....

Got the basics under control, thanks to Paul Lipps. Here's a quote:

For those who want to measure the cooling air at the bottom of the cylinder might I recommend the National LM135,235,335 familyof integrated citrcuits. They put out an analog voltage scaled to 10mV/K(C) over a range of -50C to 150C, and they only require an 11k resistor from 14V and return to power them. The output is easily read on a DVM and 20C would appear as 2.93V, corresponding to 293 deg. Kelvin.

An LM135, when packaged in heat shrink, is about the size and shape of cigarette butt. I rigged two of them on long leads and move them around the engine compartment as desired. A trusty old Fluke DVM serves as a meter, with a handy chart for converting voltage to temperature. The output is stable without filters.
 
LM's are used quite a bit in oil temp sensors for Lycomings for various engine monitors.

LM34 will read directly in deg. F. Thats what I use.

Paul also used Freescale MPX pressure sensors that he used to measure all kinds of crazy stuff. He actually was measuring standing waves in his induction tubes with an oscilloscope using them.

Interestingly enough I looked for one of their differential sensors with a pressure range suited for doing cowl measurements:

http://cache.freescale.com/files/sensors/doc/data_sheet/MPXV7002.pdf

Note that one of the most respected names in the pressure sensor business says their sensor is +-2.5% across the range of measurement.
 
Last edited:
You can use any digital voltmeter, but I have mine hooked up to the engine monitor I designed and built.

Darn, I was hoping you had discovered a cheap, scalable panel meter. Ok, the LM34 output is 10mV/degreeF, so 225F (for example) would appear as 2.25 VDC on a digital multimeter. That's obvious enough.

Basic hookup is simple:

24cwjfs.jpg


Any reason to think it needs filters, current limiting, whatever?
 
Actually I do have a cheap panel mount voltmeter that I used in my -6 for the ignition advance display that would work perfect with an LM34. I'll have a few in a drawer, will have to look tonight. It has some pins for setting the decimal point or take it out altogether.

edit: here it is: http://martelmeters.com/products.php?cat=4
 
Last edited:
What are you going to do with the data once you get it? What is good...what is bad?

Excellent questions. Let's turn the conversation to specific goals and practical methods. Aero pros, please join in, make corrections, etc.

I think there are three fundamental things you want to know.

First, how well are you converting dynamic pressure to static pressure? The static pressure in the upper plenum drives the system. More is better.

Second, what is the pressure drop between the upper and lower plenum? Pressure drop is proportional to mass flow. More drop means more cooling mass through the fins.

Third, how much heat are you transferring to the cooling mass? We want to transfer as much heat as possible to each pound of cooling mass. We want the exit air as hot as possible.

We have enough folks with manometers to establish a benchmark upper plenum pressure coefficient. Same for pressure drop and temperature rise. Known values become simple diagnostic tools.

For example, a common "cure" is the addition of louvers in the lower cowl. They lower engine temperature by increasing mass flow. However, if a pleasant afternoon spent rigging some tubing and a temperature probe showed low pressure drop and poor temperature rise, the actual problem is leakage. Improving the baffles and seals would have cured the cooling problem without additional mass flow, meaning no additional drag. (Please, no defense of louvers. It's merely a factual example.)

So, Sonny, you started the thread. What exactly do you want to determine or accomplish?
 
Back
Top