Quiz: Qualitative Flow Analysis
 

All right, all you armchair aerodynamicists of Ecomodder (alliterative and otherwise). Here's another quiz for you:

Below are three GIFs of tufts on the back window of a 2013 Prius liftback, taken on my cellphone taped to the trunk floor, looking up at the window. There is no antenna on the car (shark fin or wire), no washer nozzle, and no wiper. All three videos were taken at 55 mph indicated on the same section of road, northbound into a 15 mph NNE headwind.

The first GIF is your baseline. For the following two, I changed one thing--that is, there is exactly one change from the baseline configuration in each of those two videos. What are the changes?

Baseline:
https://media.giphy.com/media/yLevPn...Kvla/giphy.gif

Altered 1:
https://media.giphy.com/media/hG8hsy...J1Pc/giphy.gif

Altered 2:
https://media.giphy.com/media/iJ5bjh...zMgS/giphy.gif

Are you saying 2 and 3 had a piece put back on (as in return to OEM)? Or are we just guessing a random mod?

Also there is a poll option...

Love the challenge, and don't mind egg on my face if I am wrong.

For reference, but no antenna on the car, no washer nozzle, and no wiper:

https://images.hgmsites.net/med/2013...00413504_m.jpg


Baseline: Very good flow, isn't it? I'd be pretty happy with that.

Altered 1: I'd be reasonably confident you've added a fence/lip (eg a 1-inch high vertical plate or similar) that causes separation from the top of the hatch.

Altered 2: A harder one. My wife (who has watched aero testing tufts for about 25 years) says the wiper has been put back on. I say the antenna has been put back on.

Yep, just guessing a random mod based on tuft behavior. I will say that both mods are moderately-to-very common on Priuschat, but to my knowledge have never been measured by anyone using them (although Altered 1 is often claimed to "improve downforce" by people who fit it on their Prius).

I thought so too. I could see some of the movement of the outermost tufts that the camera didn't pick up so well, but I'll comment on what I think is happening after I pressure test (if that confirms what I suspect).

Bingo! Almost exactly right--I taped a 21mm high lip to the top edge of the window:

https://ecomodder.com/forum/member-v...917-183037.jpg

Yes, this one is more difficult. There's no disruption to the baseline flow pattern, but the tufts toward the trailing edge of the window are behaving slightly differently. (Actually, now that I look again the tufts at the leading edge have changed behavior slightly as well).

Yes, could be some trailing vortex development down the pillars.

Glad I got it right! And it may well reduce lift - but at a penalty of lots of drag.

Now that I know it's a common mod, I'd suggest (un-needed) vortex generators have been added at the top of the window.

That's what I suspect. The tufts on the edges of the window both a) fluttered at a higher frequency and b) were "tauter" than those in the middle of the window.

I may trial some fins along the hatch edges just to see what happens.

It may, but it is ruining the effectiveness of the stock spoiler so it's not certain at all.

Does the rear spoiler cause much of a change in angle to the rear hatch slope?

https://images.hgmsites.net/med/2013...00413858_m.jpg

I've never looked at a Prius with this in mind, so I am not sure. If the spoiler is just a continuation of the hatch angle (or makes it only very slightly shallower) then it will do nothing / very little to reduce lift.

Of course, if you want to get rid of nearly all lift, place a fence across the mid-point of the roof!

I'd be interested in your results.

Surely someone else has tried it and reported their results, but apart from my data, I've never seen anyone else reduce drag by added long fins down each side of a rear hatch that has attached flow (and I assume, therefore, trailing vortices).

What were you trying to do with the ones on your insight? Was the main goal just for something to mount the rear wing to?

The first version of fins was to provide better straight line stability by moving the lateral centre of pressure backwards. The fins also mounted the wing (that kept the fins upright).

https://i.postimg.cc/50nJW1h2/Figure-7-41b.jpg

The second version supports the fins by a steel frame behind the car and uses a spoiler instead of a wing. When throttle-stop testing this version I found lower drag with the fins, and the fins and spoiler, in place. My guess is that the fins are disrupting the trailing vortices off the pillars.

https://i.postimg.cc/XYMThprg/000599.jpg

https://media.giphy.com/media/iJ5bjh...zMgS/giphy.gif

Do we get the answer soon?

Right! I almost forgot:

https://ecomodder.com/forum/member-v...917-182858.jpg

It's the same lip spoiler, 21mm high, at the end of the factory spoiler.

There were a couple of employees waiting in their cars near where I parked to swap arrangements, for their shifts to start I suppose. I bet they wondered why the hell I kept coming back there.

Thanks - it's a good example of how flow changes can occur upstream of the modification.

You just need to get your Vape God friend to ride with you and tape a go pro to the outside of the car now to see the entire wind flow pattern. lol

I took some pressure readings yesterday for another picture of what's going on. These were centerline readings. First, no add-on spoiler:

https://ecomodder.com/forum/member-v...no-spoiler.jpg

Next, with the lip spoiler in place at the top of the window:

https://ecomodder.com/forum/member-v...of-spoiler.jpg

And finally, with the lip spoiler at the rearmost edge of the stock spoiler:

https://ecomodder.com/forum/member-v...ar-spoiler.jpg

The pressure readings decreased by 25-30 Pa over the rear window with the roof spoiler, but pressure increased on the roof ahead by 45 Pa.

Pressure on the window with the edge spoiler varied from baseline by + or - 5 Pa, but again the roof pressure was higher by 30 Pa.

I'm going to do some more testing of the rear lip spoiler today between teaching, with more locations on the window and roof.

pressure distribution
 

Looks like the 1st-condition, with no add-ons had the best performance.
The other two increased lift and drag.

Interesting. But what I'd like to see you do is to use a large spoiler - when pressure testing, I always go for something much larger than I intend using, see the results, and work backwards from there.

For example, we can see that the upper spoiler position increases pressures over the curve at the top of the hatch (from -170 in std from to -125) but puts the rest of the hatch in separated flow (lower pressures than standard) and also decreases wake pressure. So you'd say less lift but more drag.

The spoiler at the base of the hatch increases average pressures on the hatch, but because the spoiler is so small (and as we saw with the tufting) the effect is also quite small. A 25mm high lip is the smallest I'd ever test in the lower hatch position - I'd start with 75mm and then go smaller in 25mm increments.

So I'd not bother doing any more measurements with these configurations, but go larger in the modification so the effects are clearer. Some of these measurements may be in the range of noise - another reason I start with major changes first. Just use cardboard and tape to make the different height spoilers.

I don't agree with aerohead's assessment. The lip spoiler at the rear edge of the factory spoiler increased pressure by 30 Pa on the rear part of the roof! These pressures are the average of E-W runs; the data for no lip spoiler and rear lip spoiler are:

Roof:
No lip spoiler: -140 (E)/ -200 (W)/ -170 (avg)
Lip spoiler: -110 (E)/ -170 (W)/ -140 (avg)

Upper window:
No lip spoiler: -50 (E)/ -80 (W)/ -65 (avg)
Lip spoiler: -70 (E)/ -70 (W)/ -70 (avg)

Lower window:
No lip spoiler: 0 (E)/ -20 (W)/ -10 (avg)
Lip spoiler: 0 (E)/ -10 (W)/ -5 (avg)

Hatchback base:
No lip spoiler: -20 (E)/ -50 (W)/ -35 (avg)
Lip spoiler: -30 (E)/ -50 (W)/ -40 (avg)

I'm hesitant to ascribe significance to a difference in average pressure of 5 Pa. However, on the roof where the reading with the lip spoiler was consistently +30 Pa--a much larger difference--I think it's safe to say it is reducing lift. So is the spoiler placed at the top of the window, but at the expense of decreased pressure over the window behind it; the question is, how much further forward does the pressure increase over baseline?

I forgot to bring the spoiler with me so I didn't get to test it today, but I have some time tomorrow evening. I'm going to measure pressures further forward on the roof, and outboard of the centerline to see if I can get a better picture.

Have you thought about trying to put it midway through the window? That's not an educated guess, I'm just curious as to how it might change things.

assessment
 

* You lowered pressure acting at a distance on the body overhang. According to Hucho, that's a fail.
* You lowered base pressure and increased pressure drag. According to Hucho, that's a fail.
Just sayin'

By 5 Pa on average, which means realistically: inconclusive. 30 Pa on the roof: probably more trustworthy, especially since it was consistent. And that's an increase in pressure forward of the rear axle, acting on a surface that's close to horizontal.

forward
 

True, but in a position of no mechanical advantage, while simultaneously developing a deleterious torque aft.
You'll notice Indy cars and F1 moving wings as far forward, and as far aft of the axles, to get the force-at-a- distance leverage as Hucho advocates for in his chapter on high performance sports, and racing cars.:)

Nothing like complete inconsistency.

Yesterday Aerohead was complaining that rear spoilers create front lift; now he's complaining when a rear spoiler is placed so that it doesn't create front lift!

All right, more testing yesterday. This time, I took readings at 10 locations on the factory spoiler, rear glass and roof, up to just behind the B-pillar, at an inboard position (not centerline) and outboard.

One important difference between yesterday and Wednesday's testing: wind was out of the south both days, so there was crosswind in my earlier East-West testing. I used a different route yesterday, an exactly North-South county road (everything here is conveniently on a one-mile grid) and, with winds again out of the south, that means lower yaw, maybe even zero yaw. All testing was done at 80 km/h indicated with a pitot tube giving reference static pressure and readings averaged over 45 seconds or so each run.

No add-on spoiler:
https://ecomodder.com/forum/member-v...no-spoiler.jpg

Add-on lip spoiler:
https://ecomodder.com/forum/member-v...02-spoiler.jpg

Difference:
https://ecomodder.com/forum/member-v...difference.jpg
(all values given in Pa)

Observations:
-On the roof, pressure is lower toward the middle (faster flow speed).
-Over the window, this flips; pressure is lower toward the edge. This may indicate vortex formation there.
-Adding the lip spoiler produced positive pressure difference from ambient just in front of it, on the factory spoiler, and ambient pressure on part of the window.
-Adding the lip spoiler increased pressure a significant distance up the roof.

In the past I've not thought much of small spoilers such as this, believing that they wouldn't do much of anything. But based on these two tests so far, I think I was wrong!

Great testing - well done!

So, you may have thought of all of this already, but if not...

The increase in pressures on the rear hatch and roof from the little lip spoiler will decrease lift (the upwards component of the resolved force) and decrease drag (the horizontal component of the resolved force on the hatch). The decrease in drag will be very small, however, and probably more than offset by the larger wake.

That's why I ended up going with a spoiler than increased pressures on the hatch without an increase in wake area.

Yeah, and I'm not sure if I have a way of testing the increase in drag with any accuracy--since the Prius can't do throttle-stop testing. I might stick the spoiler on for my weekly hour-long drive (I teach one day a week this semester at a school an hour south of here) and see what happens; I've been getting just under 60 mpg depending on wind. Or, I might try the method from your book using a long hill; there's a section of interstate near here with a moderate grade for 1/2 mile. Or what I ended up doing with the air curtain ducts, measured fuel economy over a several-mile section of flat road with 3 or more runs in each direction in each configuration.

I tried reading engine load and horsepower on the Scangauge to see if rolling the windows down would show up, but there was no correlation--the problem, I think, is the variable torque from the electric motors. There's a MAF x-gauge I could use too but again, I think that would suffer from the same problem.

I might not get to do the Green Grand Prix this year because of Illinois' coronavirus positivity rate, which would have been at least a comparison of a long drive on a closed course with the two years I've done it already, the car in a different state each time.

If anyone has other ideas for testing drag changes...?

The downhill test might work if you can do it in neutral - not sure that is possible in the Prius. Otherwise, regen braking issues, etc.

Depending on your speed limits / interest of the law, I'd probably try to test fuel consumption over a short distance (eg 1-2km) on a flat, straight road at the highest possible constant speed eg 130 km/h+.

With either, do windows up/down to ensure your are getting real data.

complete inconsistency
 

* You know that Hucho advocated for an ' aerodynamic lever arm ' on page 231. And he provides all the context.
* I don't complain, I just post information germane to a complete discussion of a particular issue for which you've left holes big enough to drive a Titanic through.

pressure distribution
 

Where is the comparative underbody pressure profile.
What you're sharing has no context.

Well, bless your heart.

well
 

There's a reason I bring it up.
We're in subsonic flow and any change, anywhere, according to Hucho has implications to elsewhere around the vehicle.
And specifically as to spoilers, Schenkel's spoiler research ( also in Hucho's 2nd-Ed. ) revealed that a rear spoiler altered the pressure under the car as well as over the boot.
I'm not marginalizing your efforts, it's just that, without a full accounting, the total effect remains an unknown quantity.

Vman455 -

Is this the fancy Prius with the auto leveling headlights? I'm asking because I read it has a rear ride height sensor.

Forgive my exasperation, but I spent 4 hours, including the trip to the hardware store for materials to mount the pitot tube in freestream air above the car when my first apparatus failed, out of my very busy schedule (full-time job, half-time in school, and teaching part-time at another school) gathering this data only for you to chime in with your opinion that it is worthless, as you have criticized JulianEdgar on this forum before in an attempt to discredit any sort of on-road testing, even stooping to personal attacks as recently as this very afternoon. I don't know what axe you're trying to grind here; I don't think it is worthless, and if adding the spoiler lip has increased pressure under the car--well, I'll be measuring that soon and doing what I can to mitigate it.

Unfortunately, no--I added the LED headlights aftermarket, but not the auto-leveling. The OEM ride height sensor is $200+, a little rich for me, so I'll probably try to fab something up like JulianEdgar has done or use some other sort of (cheaper) distance sensor to measure ride height.

Keep doing your testing - I think your results are great, and we will all be able to learn a lot from them. In fact, you might want to think of keeping a Word doc with all your results in it - so much easier to refer back to than miscellaneous forum posts.

As for Aerohead, I have avoided the personal belittlement he is fond of, but it's obvious he has some issues. There's a lot of confusion in what he writes, and as you have previously found, many of the references that he cites don't actually support what he claims. However, I think the worst thing is that he refuses to learn from developments of the last 20-30 years and so keeps on rehashing the same old (mis)information.

Of course he isn't going to like your testing, because (as with mine) it will so quickly show that much of what he says is simply wrong!

Pressure and tuft testing tell you so much, and when you add lift/downforce testing, you'll be looking at information radically better than any guesswork / rules of thumb / 1930s models / templates, etc. (And, I'd also argue, a lot better than crappy CFD.)

Re using other height sensing techniques, I have tried ultrasonic height detection but couldn't get the resolution needed for lift/downforce testing at normal road speeds. So back to a simple P38 Range Rover analog pot-based sensor. I think my latest smoothing circuit is better than the one in my Veloce book, and seemed to work really well on the Impreza rear wing / spoiler tests.

So keep testing - it certainly takes time and effort, but it's also like taking off a blindfold and seeing the world for the first time.

I'll probably end up going this route; I looked up other height sensors and most of them can be had much more cheaply than the Prius arm, for some reason (maybe because they aren't very prolific?). Thanks for the advice on the ultrasonic sensor; too bad, because TI makes one that could be logged directly on my calculator.

Very much so. I've spent the last two years in calculus, physics, and engineering coursework, and it's the same sort of experience--especially physics labs pre-pandemic, where experiencing physical phenomena is worth 1000 times more than reading about them in books. One professor had us do a weekly write-up on important physicists of the past; one week I chose Emilie de Chatelet, a mistress of Voltaire who demonstrated (in the early 1700s, mind you) that kinetic energy is proportional to the square of velocity by dropping balls into sand and measuring how far they penetrated. I imagine a lot of people today would poo-poo that sort of experiment as low-tech and uncontrolled, yet it was instrumental to our current understanding of physics.

Good stuff. My father was a scientist and physicist and he instilled in me from a very young age that doing an experiment is often far preferable to reading theory. And in any area of car modification, that applies in spades.

I can't find a link, but Harry Miller was working (1931 4WD race car?) with German engineers who asked him about IIRC the diameter of the inlet to the supercharger.

He held up two fingers and the 'exasperated' German engineer used his caliper to measure the air. Turned out he was right on the money. That's what hands-on experience does.

I'd never known much about Harry Miller until I went to the Peterson Museum in Los Angeles. Then I was transfixed by this stunningly beautiful car - front-wheel drive, too!

https://i.postimg.cc/mD6dzrft/IMG-7469.jpg

https://i.postimg.cc/W12fHCnp/IMG-7467.jpg