Measured aero pressures on Insight
 

(Better measurements available now later in the thread)

Incidentally, for those following other threads, you can clearly see how absurd the notion is that The Template wouldn't develop lots of lift - unless of course a great deal of attention is paid to the underside of the car.

https://i.postimg.cc/fRSkrJgZ/Insight-and-template.png

(Better measurements now available later in the thread)

(I'll put up a new video with the revised measurements and measuring technique)

(Better measurements now available later in the thread)

I am curious about the reading of 500 pascals at the front of the car as this is greater than the stagnation pressure at 80 km/h.

velocity (m/s) = velocity (km/h) / 3.6 = 80 / 3.6 = 22.222 m/s

air density at 10 C is 1.246 kg/m3 for std pressure

Stagnation Pressure (Pa) = 0.5 * Density * Velocity ^ 2 = 0.5 * 1.246 * 22.222 ^ 2
" " = 307.6 Pascals

This compares with your measurement of 500 Pascals.

What actually causes the measured pressure to be greater than stagnation pressure?

Glad someone is looking closely at the data! Yes, the measured stagnation pressure is higher than it should be. That's why I always measure stagnation pressure when I am doing car aero pressure measurements. I then usually reference the other pressures against this value. That is, I normally convert stagnation pressure to 100 and then apply the same multiplier to all the other pressures, so the other pressures are relative rather than absolute. (For example, it's then interesting to compare the Insight pressure distribution with the XE Jaguar. - edit: I've now done this in the next post.)

Why is the measured stagnation pressure higher than it should be? I don't know. It could be the gauge, it could be the sensor, or it could be variation in measurement due to atmospheric conditions eg wind. In inches of water, the two stagnation pressure measurements were +1.7 (423 Pa) in one direction and +2.2 (550 Pa) in the other direction. That's a ratio of 1.3. Interestingly, the wake pressures also varied over the same range (+0.15 vs 0.2). In contrast, the other pressures were generally more consistent for the two directions.

For stagnation pressure you could use the calculated value of 310 Pa over the actual measured value (not rounded) of 487 Pa (ie 310 / 487) as a multiplier for the other pressures (ie 0.64) if you wished - of course it wouldn't change the relativity of the readings.

There's also another issue. I was using a 0-3 inches of water gauge and I picked a speed (80 km/h) that gave plenty of deflection (ie 2.2 inches of water) for the max measured pressure. However, that meant that I was working down the very bottom of the scale for measuring the smaller pressures. I could then have gone to one of my more sensitive gauges for these small pressures, but I didn't because I thought the chances of comparative error would be much higher (ie through using two different gauges). But if I were - say - developing a rear spoiler and wanted to measure the pressure change on the rear window, I would use a much more sensitive gauge (eg 120 Pa - 0.5 inches water - full scale deflection) for these measurements.

(now not relevant)

Thanks for the explanation.

Would it be feasible to use an electronic differential manometer such as this +/- 1 psi one. It has a resolution of 0.01" H2O and a full scale accuracy of +/- 1.0 percent.

https://www.az-instrument.com.tw/en/...-82012-AZ.html

Yes it would - do you have a price?

I try to avoid using instruments that are too expensive for normal modifiers. The low cost digital manometer I have has a highest resolution of 0.1 inch of water - no good.

Edit - this looks good https://www.amazon.com.au/Perfect-Pr...22441494&psc=1

Second edit - just bought it.

I actually own two +/- 2psi manometers myself and hence the reason for the question. They have 0.01" H2O resolution.

The funny thing is I went looking for a +/- 1 psi hoping for better accuracy but with the +/- 1 psi having a 1.0 % fsd accuracy and the +/- 2 psi having a 0.3 % fsd accuracy, then the latter would appear to have higher accuracy.

I paid a very similar price to yourself for mine.

Have you tried using them for measuring aero pressures? For 'simple stuff' like before/after the radiator, just place the ends of the hoses in the approriate locations. For 'complex stuff' (like panel pressures), you'll need to make a sensing puck.

At this moment in time I use mine for doing engine port flow measurements using a venturi and hence the 2psi works well.

For car aero measurements - particularly at the front of the car - I was wondering if there would be any merit in using a pitot tube protruding in front of the car to get a static reading to be used as one input to the differential manometer. This would be a substitute for the pressure box. How far the pitot would need to protrude in front of the car to get a clean static pressure is the question.

That's a very interesting idea - one I'd certainly not thought of.

The two issues I can see though is that I think the static port would need to be a long way ahead of the car (eg 1+ metres) and such a support could change flow patterns on the car.

But, depending on how quiet your roads are, I think both are not major deficiencies. I'll give it a go today and report back.

Initial measurement - it's looking very good indeed.

I decided to mount the pitot tube above the car - that's legal here whereas in front of the car wouldn't be. It's 2.2 metres above the ground.

Using the pitot tube as the static reference, the stagnation pressure on front numberplate measured 1.2 inches of water (300 Pa). That compares with the calculated stagnation pressure (see cr45's post above) of 308 Pa.

The reading also does not drift as it does when using the sealed vessel as the reference (it's impossible to keep temps exactly constant, so the volume expands/contracts a bit).

So well done cr45 for the great idea.

Now I need to do all the measurements - and revise the data I showed previously!

So I think that using the pitot tube as a static reference is a much better approach than using the sealed vessel as the reference. Thanks again to cr45 for the excellent suggestion.

I have redone all of the Honda's measurements. Note that the basic pattern remains the same as with the previous measurements - for example, no negative measurements became positive or anything like that! However, I think with the pitot tube static reference, the absolute figures are much more accurate.

I also did an additional measurement in the middle of the roof because I suspected lots of lift might be being generated here because of the roof curvature (and I was right).

Test:

• Pitot tube static reference located 2.2 metres above ground
• Magnehelic 0-3 inch of water gauge
• Custom measuring puck (easy to make yourself)
• 11 degrees C, fine and clear
• Light winds
• 80 km/h test speed
• Average of two measurements in different directions
• Pascals rounded to 5

Note: the Insight runs full front and rear undertrays, with the rear undertray incorporating a 10 degree diffuser.

https://i.postimg.cc/nV1xY2wf/80-kmh...ef-at-2-2m.jpg

https://i.postimg.cc/HW6Kkhhn/arrows...ref-80-kmh.jpg

https://i.postimg.cc/rm2FDbdv/IMG-0478.jpg

With the pitot reference giving such stable readings, I was able to do a test on the Insight that with the previous sealed reservoir approach, was difficult/impossible. I used a 0-120 Pa Magnehelic gauge (ie 0-0.5 inches of water FSD - so very sensitive).

I placed the measuring puck on the bumper corner and recorded a two-way average of -30 Pa at 80 km/h.

https://i.postimg.cc/wxQDDtKb/IMG-0482.jpg

I then added a large separation edge (always go large when testing - if it works, you can always go smaller.) Note that the measuring puck was not moved.

The pressure increased to -17 Pa (similar to the pressure previously recorded in the middle of the wake).

https://i.postimg.cc/fWC0J6XG/IMG-0479.jpg

So the separation edge measurably and clearly reduced the vertical suction peak down the side of the bumper.

Could you measure in the middle again? (with the separation edges on the bumper) The pressure should have increased there too.

And I'm curious if an improvement could be measured with your throttle stop testing.

I have some lip-style rear spoiler molding coming to make the proper separation edges (like this)

https://i.ebayimg.com/images/g/SBMAA...aZ/s-l1600.jpg

When it arrives - two to four weeks I imagine - then I'll do a bunch of measurements. My guess is that the separation edges will impact only the pressures near the edges, not generally within the wake.... but we will see!

This looks very promising indeed and stagnation pressure is now where it would be expected to be.

I decided to have a look at the sensitivity of pressure to temperature change for a pressure box.

Using Gay-Lussac's law for a closed container then P2/T2 = P1/T1

Hence P2 = (T2/T1) * P1

Lets assume that the atmos pressure is 101325 Pa and the temperature is 10 C.

Also lets assume that the temperature of the air rises by 0.1 C by the time you take your aero measurements.

The new value for the pressure in the box will be

P2 = ((10.1 +273.15) / (10.0 + 273.15)) * 101325 = 101361

That is a difference in pressure of 36 Pa which is very significant considering the magnitude of the pressure differentials you are trying to measure.

That much!

With the reference tank you can get it so that there's very little drift, but you need to first let the car's interior settle a lot in temperature, reset the tank pressure to atmospheric and then quickly do the test. Before then repeating the process for subsequent tests. (I also ended up insulating the tank.)

But I think the pitot tube ref pressure is better in every way - cheaper and easier for people to put together, too.

As I have shown above, the stability of the reading also gives me enough confidence to measure very small changes (eg changes in wake pressures).

I think your suggestion genuinely takes on-road aero pressure measurement to the next level. I am very excited by the possibilities that are now open.

All explained - using a pitot tube as the reference when measuring panel pressures

https://www.youtube.com/watch?v=WAg2...ature=youtu.be

Quick and easy trip strips! I love it. Thanks for the idea. I'll be looking forward to your results, if I can resist just buying them anyways and trying them out.

Rob Palin (ex Tesla aerodynamicist) has just given me a fascinating idea with regard to separation.

So I will be doing some more testing today.

But whether I can achieve what he suggests is another matter....

So that testing was interesting.

Rob's point was that if the boundary layer is thicker, the strength of trailing vortices at the point of separation are likely to be less strong, as the velocity gradient will be reduced.

So I wondered if thickening the boundary layer prior to the separation edge would result in a higher pressure (ie be improved) behind the separation edge.

Test:

• Honda Insight
• 12 degrees C
• light winds
• 80 km/h
• two way average
• 0-120 Pa Magnehelic
• Pitot tube 2.2m above ground as static reference
• Pressure sensor located just inside wake from separation edge
• One one side tested - other side standard

Separation edge alone: -14 Pa. (Note 1: a different separation edge from the one tested yesterday - smaller, reflecting what will actually be used. Note 2: the same pressure as recorded yesterday - new sep edge works as well as yesterday's, and very repeatable data)

I'd thought I would see if I could thicken the boundary layer with the disruption of a line of egg cartons.

https://i.postimg.cc/CxHGkcWf/IMG-0487.jpg

Vertical line of egg crate carton plus separation edge, 80cm forward of separation edge: -18Pa

Vertical line of egg crate carton plus separation edge, 50cm forward of separation edge: -17Pa

Vertical line of egg crate carton plus separation edge, 20cm forward of separation edge (as pictured): -33 pa

So the egg cartons didn't have an impact until they were close to the separation edge, whereupon they dramatically decreased pressure (the opposite of what we want).

I then thought I'd try some Airtabs. (You'd expect to them to make things worse if they inject momentum into the boundary layer.)

https://i.postimg.cc/G2NGdVs6/IMG-0489.jpg

Three Airtabs plus separation edge, 20cm forward of separation edge: -18 Pa

Three Airtabs in reversed orientation, 20cm forward of separation edge: -21 Pa

So the Airtabs basically made no difference to the measured pressure.

As a reminder, yesterday adding the separation edge increased the measured pressure from -30 to -17, so a dramatic improvement.

I tested a box cavity via pressure measurements in the wake.

https://i.postimg.cc/Dw7q0wvX/IMG-0495-lighter.jpg

Test:

• Honda Insight
• 14 degrees C
• light, slightly variable winds
• 80 km/h
• two way average
• 0-120 Pa Magnehelic
• Pitot tube 2.2m above ground as static reference
• Pressure sensor LHS, centre and RHS at bumper level

Without box cavity:

Centre: -18 Pa
LHS: -22 Pa
RHS: -20 Pa

Average: -20 Pa

With box cavity:

Centre: -16 Pa
LHS: -15 Pa
RHS: -18 pa

Average: -16 Pa

Now how significant is this? I am not sure. On the one hand you can say that the pressure was increased by 20 per cent - so that's excellent.

On the other hand, looking at the wake area (about 0.85 m^2), the actual force changes like this:

At 80 km/h, without box cavity, calculated wake drag = 1.7kg
At 80 km/h, with box cavity, calculated wake drag = 1.4kg

I doubt if a change of drag force of 300g at 80 km/h is significant - but perhaps it is?

But....

If the 20 per cent change also occurs on cars with big wakes, then it would become quite significant.

and...

This drag force will rise at the square of the speed, so at constant high speed, the drag force change (in absolute terms) of having the extensions will be much greater.

All very interesting.

(And yes, I am now much less skeptical of box cavities!)

Was wondering, have you thought of testing the separation edge with the air tabs alone (yes, realizing they also might not have a positive effect there, but still...)

Yes I did do this:

Three Airtabs plus separation edge, 20cm forward of separation edge: -18 Pa

Is that what you meant?

Oh, my bad! What I meant was testing the pressure without the modified edge... on the stock bumper.

No I probably wouldn't bother doing that. The separation edge works well, so Airtabs would have to improve things over having just the separation edge - and they didn't.

(I am first a car modifier, and so I want to do things that work on my own car.)

Remember all this testing equipment costs less than US$75 total - so anyone can do it to find out anything they want. (And on the car they wish to modify, which is best of all.)

Julian, you're a great resource, a bona fide testing guru!

Thanks for all the testing information you are contributing to this site!

One suggestion and one question:

The suggestion: You might want to try a simpler way to mount the pitot tube. I've been using a homemade pitot tube (a simple copper pipe without any side holes) for the past 20 years in the grill of my F250 truck, which sticks out maybe three or four inches in front of the grill. This unofficial, self-named pitot tube is connected to an aircraft airspeed indicator via a plastic tube that is run through the engine bay into the cabin where the gauge resides.

When I first installed the airspeed indicator, I left the static port open in the cabin, but as you noted, I too found this will not give accurate results due to the varying pressures in the cabin. So, I ran a plastic tube from the gauge's static port into the engine bay and got good results. I put a piece of open foam in the end of the tube to keep the mud dabbers out, and this may or may not mitigate any effects of air currents in the bay at speed.

I will note that I do have louvers in the hood to bleed off some of the intense engine bay heat produced by the turbocharged diesel engine. This helps preserve the battery and other heat-sensitive things that must live near the hot engine.

I know the pitot tube is working as intended via observing the same MPHs on the airspeed indicator as is demonstrated on the vehicle speedometer and also a GPS unit, but they are the same only in still air. If there is a headwind, I see appropriate increases in airspeed compared with the other two indicators of groundspeed. If there is a tailwind, the reverse is true. If there are sidewinds, the airspeed differential varies with the wind swirls. Also, when in traffic, all the moving vortex generators around me, aka cars, trucks and busses, make the airspeed indicator jump up and down like a neurotic cat on a hot stove.

All this to say that you might want to try a less elaborate way to mount the pitot tube. It doesn't bother my air speed indicator to simply place it in the grill.

Now, the question I have: The "pressure sensitive" colors on the hood of your car shown in the video seem to imply that the hood is experiencing high pressures not only at the base of the windshield, but also in the center of the hood. The reason I question this is I have installed hood louvers on several of my vehicles, and I have found the centers of my hoods see relative low pressure on the top as compared with higher pressures on the bottom (from the engine bay), which allows excessive heat to flow out and up through the hood at speed (and also when stopped, of course). An exception to this pressure differential would be at the base of the windshield, which does see high pressures due to the windshield vs hood confluence there.

So, are you actually seeing high pressures on the center of the hood as the colors in your video seem to show, or am I misinterpreting it?

I would be confident that in the fullness of time Julian will find that he can lower the pitot tube. However, as a starting point putting the pitot tube at a 2.2m height means that he can have 100% confidence in the measured static pressure.

The major difference between your use of a pitot tube and Julians use is that he is measuring static pressure and you are measuring total pressure. The air is almost certainly slowing down before it reaches your pitot resulting in a rise of pressure. This is not an issue if you are only measuring total pressure.

However it would cause a static reading to be too high and this discrepancy would increase with the vehicle speed. Pitot tubes have to be in free stream air in order to get accurate static pressure readings.

I'm sure Julian's requirements are more sensitive than mine.

That view is of the back of the Jaguar, not the front!

Typically, there is a low pressure across the leading edge of the hood (around the front curve) and a high pressure at the rear of the hood (base of windscreen, as you say). As you'd then expect, there is a gradient from low to high pressure across the hood, and at one zone the pressure on top is the same as ambient.

However, the influence of any yaw (cross wind component) can cause some changes in this pattern, as the airflow wraps around a 'side corner' (ie from the fender onto the hood).

To site hood vents, it's best to use a differential pressure gauge (eg Magnehelic) to measure the pressures under, and over, the hood simultaneously at the different possible vent locations. To vent the engine bay you want the biggest difference (highest under, minus lowest over). In the last car I did this with, a forward location on the hood was best.

Yes, as cr45 said, the requirements when measuring just airspeed are a bit different - and easier.

Incidentally, I liked your report on how measured airspeed varies in the different conditions eg in wakes of other vehicles.

I found this when I was measuring pressures on the top surface of an under-hood intercooler - the figures danced around a lot when I was behind other cars. Even, if memory serves me correctly, 20-40 metres behind other cars!

I was thinking about this. I think I'll try to find some really secure large suction caps that I can use to stick an upright to a window, with the pitot tube on top. That way the system can be easily swapped from car to car. (And then not measure stuff on that side of the car!)

If you have a second pitot then you could test it in an alternative position but keep the present pitot at 2.2m and input the two static outputs to the one manometer - ideally your most sensitive one.

If the new position is accurate, the meter will give a zero pressure reading at a range of road speeds.

Yes I do have two pitot tubes. Another really good idea!

I have a suction cap thingy coming and I'll do comparison readings at different heights when that arrives.

https://ae01.alicdn.com/kf/HTB1wupKa...317&hash=23377

I wonder how far from the front of the grill will one find still air that isn't influenced by the car.

As I noted, the pitot tube for my air speed indicator is just three or four inches ahead of the grill of my F250 truck, and in still air with no vehicles around, at 60mph ground speed, re the speedometer and the GPS, the air speed is also 60mph, re the air speed indicator with a smooth, steady needle.

So, I interpret this to mean the location of my pitot tube is OK, and that there is no forward air agitation going on.

Quote: "That view is of the back of the Jaguar, not the front!"

Ooops, it's my assbackwardness coming out again.

I guess I'm still surprised there would be high pressure on the lid of the trunk as well, rather than some lift.