Tuft testing a C-Max

[fbov]

One of the topics I find interesting on this forum is car mods, specifically those that improve fuel consumption through areodynamic improvements. In researching here and eslewhere, the general rule is that there's not much you can do up front besides block air infiltration, but you can get big drag reduction in the rear. I wanted to see what airflow looled like over the back end, and what sort of changes might be beneficial. 

 

I had also noticed that the hatch lid didn't meet well with the roofline.

 

I laid out ~4" tufts of yarn and string in a 5" grid pattern over the back half of the car.

 

The idea is to drive at speed and look at the patterns the tufts take on the car. I took video, but need the right SW to pull snippets. Here are a couple stills.

 

General observations

Laminar flow is everywhere

- roof all the way back to the lip; that little step does nothing

- windows sweeping up toward the roof, right to the hatch glass

- door panels above rear wheel well, but ending under lights

 

Turbulent flow dominates the tailgate (lots of blurry tufts)

- lip under the glass down to the top of the bumper

- flanks under rear lights sweeping forward to wheel wells

 

The big surprise is the rear window - near calm

- strings lay flat and don't move much

- pattern is downward with slight outward flare

- pattern extends to L/R edges of the glass, right next to high-speed laminar flow

 

I see...

... evidence of clean separation from rear lights, up across upper hatch edge - a calm rear window.

... evidence of turbulent separation along the flanks, below the lights - a turbulent hatch panel below the window.

 

Take a look and let me know what you think. I'll note that I'm looking at video as well for my descriptions.

 

Have fun,

Frank

[ptjones]

Nice job Frank, I don't think it's right that your hatch isn't flush. GasPod time to test and also the front end. :)

 

Paul

[mtb9153]

One of the topics I find interesting on this forum is car mods, specifically those that improve fuel consumption through areodynamic improvements. In researching here and eslewhere, the general rule is that there's not much you can do up front besides block air infiltration, but you can get big drag reduction in the rear. I wanted to see what airflow looled like over the back end, and what sort of changes might be beneficial. 

 

I had also noticed that the hatch lid didn't meet well with the roofline.

Roofline.JPG

 

I laid out ~4" tufts of yarn and string in a 5" grid pattern over the back half of the car.

TT layout done.JPG

 

The idea is to drive at speed and look at the patterns the tufts take on the car. I took video, but need the right SW to pull snippets. Here are a couple stills.

Capture.JPG

Capture1.JPG

 

General observations

Laminar flow is everywhere

- roof all the way back to the lip; that little step does nothing

- windows sweeping up toward the roof, right to the hatch glass

- door panels above rear wheel well, but ending under lights

 

Turbulent flow dominates the tailgate (lots of blurry tufts)

- lip under the glass down to the top of the bumper

- flanks under rear lights sweeping forward to wheel wells

 

The big surprise is the rear window - near calm

- strings lay flat and don't move much

- pattern is downward with slight outward flare

- pattern extends to L/R edges of the glass, right next to high-speed laminar flow

 

I see...

... evidence of clean separation from rear lights, up across upper hatch edge - a calm rear window.

... evidence of turbulent separation along the flanks, below the lights - a turbulent hatch panel below the window.

 

Take a look and let me know what you think. I'll note that I'm looking at video as well for my descriptions.

 

Have fun,

Frank

Frank, I do videography as my second business.  I have considered doing what you have done and probably will do it for sure now.  If you are willing and can get me your video footage I can capture it into my editing SW Final Cut Pro and create a short video account of your testing.  I will supply you with a copy of what I create.  LMK if you would like me to assist?  I can extract clips of your footage and arrange them anyway you want in the final DVD.

Edited November 4, 2013 by mtb9153

[mtb9153]

Nice job Frank, I don't think it's right that your hatch isn't flush. GasPod time to test and also the front end. :)

 

Paul

It would be interesting to know if Suzanne has any laminar flow testing results of her Pods from the front and if this would be a worth while test since my pods are magnetic as are most if not all of our testers?

[fbov]

PM sent. Thanks for offering!

 

Once tufted, it would have been quite easy to add aero test factors to the car like magnetic GasPods, save for my wife's impatience. Just keep in mind that you're limited to sheet metal... but you get to try different variants and see if there's an optimum for turbulence reduction.

 

And here's the $64,000 question... (drum roll, please)... where would you put them for the greatest benefit? What are the group's thoughts?

 

HAve fun,

Frank

[wab]

PM sent. Thanks for offering!

 

Once tufted, it would have been quite easy to add aero test factors to the car like magnetic GasPods, save for my wife's impatience. Just keep in mind that you're limited to sheet metal... but you get to try different variants and see if there's an optimum for turbulence reduction.

 

And here's the $64,000 question... (drum roll, please)... where would you put them for the greatest benefit? What are the group's thoughts?

 

HAve fun,

Frank

 

SORRY i can't resist.

but first

I'm seriously thinking about getting some Gas Pod's.

I want to combine them with a Shark Fin with just 6 Pods.

 

Now to answer your question.

"where would you put them for the greatest benefit?"

Back in the box. :lol: 

[C-MaxSea]

Nice work again, Frank.  The pics are too small to see anything. Looking forward to more pics/video.

 

I don't have a clue, but based on your notes, looks like we may be seeing pods on the lower flanks? (not quite as sexy as on top)   Or perhaps under the rear bumper?  Or both in a 3-5-3 configuration ?  I'll take 11, please.

 

You are having way too much fun,

 

Nick

 

(Puzzled that few seemed to apprec./comment on your 13B07 results) 

 

(that is funny wab, but I don't dare 'like')

Edited November 4, 2013 by C-MaxSea

[fbov]

You can see the full size pictrue file by "right-click, save target as." Still, it's not the full HD frame and you lose a lot without motion... working on it. 

 

wab, I agree to an extent, humor aside. The point of testing is to see if there if a problem can be identified. without a problem, best leave them in the box. so...

 

Was an aerodynamic problem identified?

 

What would you suggest for remediation?

 

... these are not rhetorical questions, but rather the point of the exercise!

 

HAve fun,

Frank

[fbov]

I posted my results on the Ecomodder forum, and got a very concise assessment of the car's aerodynamic characteristics and my options in that regard.

 

*In elevation,Ford has done the K-form body, respecting boundary layer requirements for attached flow.
*In plan-view, they're borrowing from Paul Jaray's 1921 'Kombination-form' greenhouse.
*Both are truncated at a 'practical' length.
*You do have nice clean laminar flow up to where Ford exercised some aesthetics with the truncation. A trip down a dusty road would reveal your actual separation line.
*The wake is exactly as one would expect it should be.
*For lower drag you'd just elongate the body, picking up where Ford left off, extrapolating the curvatures along imaginary 'ideal' pathways.
*The effective-fineness ratio determines,more than any other factor,your potential for low drag.

 

If you understand all this, you're a better man than I. However, I learned a lot looking up what I didn't know. Ford did their homework aerodynamically, at least.

 

Have fun,

Frank

Edited November 9, 2013 by fbov

[fotomoto]

 Ford did their homework aerodynamically, at least.

 

 

Previously, I posted this discovery on my first oil change:  there's a NACA duct on the oil pan cover.   That told me all I need to know.  :rtfm:

[fbov]

Several folks have asked, so I am making the tuft test raw video available. While full resolution, the image is upside down, and not centered when it's interesting. If anyone with video edit capability is willing to help...

 

http://youtu.be/TJ-82y81_Lw

doesn't get interesting until 1:40 into the 2:00 file

 

http://youtu.be/aMdQqjd2KPM

interesting at 0:30

 

Have fun,

Frank

[fbov]

And here's an annotated still, showing what to look for...

 

Viewing the videos, you have to watch for fast moving tufts. In areas of laminar flow, the tufts don't move, so they're easy to see like parallel toothpics. When flow separates from the surface, the tufts go wild. You have to watch the video to get an idea how fast they move in turbulent air.

 

I've circled an area that's particularly telling - the rear window boundary. The tufts show a strong laminar flow along the side, around the RAS sticker. Turn the corner onto the back glass and the flow is laminar, albeit at much lower speed, but perpendicular. That indicates a very clean break with no wrap-around turbulence reaching the glass.

 

The same cannot be said for air along the flanks. Above the wheel well, flow remains attached until the tail light, when the bodywork wraps slowly around the corner. This slow curvature is perfect for the front of the car, but killer here, as it results in flow propagation around to the back of the car, unlike the clean break up by the rear glass. At wheel well level, the turbulence coming off the rear wheel never reattaches before the bodywork curves away and pulls this air in with it.

 

My aerodynamic conclusion is that the greenhouse above the beltline is well designed with good control of airflow over the glass surfaces on all sides. At the beltline, body contours begin to have a negative effect, causing turbulent flow separation at the rear, worsening as wheel well turbulence makes its presence felt lower on the body.

 

What do you see?

Frank

 

[fotomoto]

The same cannot be said for air along the flanks. Above the wheel well, flow remains attached until the tail light, when the bodywork wraps slowly around the corner. This slow curvature is perfect for the front of the car, but killer here, as it results in flow propagation around to the back of the car, unlike the clean break up by the rear glass. At wheel well level, the turbulence coming off the rear wheel never reattaches before the bodywork curves away and pulls this air in with it.

 

My aerodynamic conclusion is that the greenhouse above the beltline is well designed with good control of airflow over the glass surfaces on all sides. At the beltline, body contours begin to have a negative effect, causing turbulent flow separation at the rear, worsening as wheel well turbulence makes its presence felt lower on the body.

 

So are you considering moving the POD's to the sides in a vertical line behind the rear wheel well starting up at the belt line  and continue down to the bottom?

[fbov]

First off, I didn't test GasPods, but rather AirTabs.

 

My intent was exactly as you describe; install them in a vertical line from behind the taillight down behind the wheel well. As the link details, bodywork curvature proved insurmountable; I need to modify the AirTabs so I can mount them on the car's contours. Instead I tried the roofline, with negative results as discussed.

 

I would be interested in anyone's results who has put GasPods on the sides... or who has other ideas for addressing the flow separation, beyond new bodywork. Anyone could slap on a Prius-tail!

 

HAve fun,

Frank

[Jus-A-CMax]

Funny, I considered Air Tabs as part of the research but ended up being persuaded with the magnetic gas pods. I presume your airtabs was set around 1 to 1 1/2" from the edge of the rear shelf, similr to the way most of us set our gas pods?

Edited November 11, 2013 by Jus-A-CMax

[fbov]

I mounted them even with the front edge of the hatch lid, including the curvature as the tuft test showed that airflow isn't parallel the the direction of motion, it's slightly skewed there. If you recall that I reported the hatch was recessed downward a little? I was aiming for the tab's mouth to be flush with the roof when mounted on the hatch.

 

I was going to take pics but didn't; another sign of subliminal bias???

 

Have fun,

Frank

[Recumpence]

Frank,

 

I have been looking at a couple things to try on my car and I wanted to ask you a couple questions;

 

#1 I was planning on making a subtle Kamm-Back out of Lexan that would run around 8 inches rearward of the stock "Spoiler" and flanking the rear glass. It would be heat bent at the corners and follow the contour of the car. However, if we have good separation there already, the only benefit would be slightly reduced area for the air to detach from. Is this correct? What I mean is, there would not be any measurable turbulence reduction, but, rather, slightly reduced rear area for the air to detach from, thus minimizing the improvement there?

 

#2 I have long known the rear curves behind the rear wheel wells are far too gradual to promote clean separation and I have considered adding subtle side vanes that would attach just behind the wheel well from the bumper edge up to the bottom of the tail light giving the air a clean edge to separate from. This feature is clearly visible on the Volt and late model Prius, for example.

 

What are your thoughts on these two ideas? The side vanes would be very easy to make and install with simple foam double sided tape. They would also be very easy to remove.

 

Matt

Edited November 15, 2013 by Recumpence

[fbov]

You have reached many of the same conclusions I have. I just can't come up anything that's not ugly.

 

When you mentioned your plans for a Kamm-back, I wondered if you'd reviewed the videos, as they show much more of the dynamics than stills can. What struck me most was the outboard edges of the rear glass, where the airflow on the glass is substantially perpendicular to the high velocity flow coming off the side. It's a classic example of how to truncate the back end for a clean break, unlike the quarter panels below it. Things like this are one reason I did the tuft test; a Kamm-back was on my radar, too, but I think Ford beat us to it.

 

As to those wonderfully generic and aerodynamically horrible quarter panel curves you see on so many recent designs... the challenge is being subtle; Ecomodder's full of not-so-subtle examples.

 

Subtlty is what got me interested in vortex generators as a way of using air to modify the air flow. The challenges of putting flat devices on curved sheet metal led me to try them on the rear roofline, with negative results confirming for me the discussion above. I'm now modifying the devices to fit the bodywork contours, and will re-test if I'm successful. At least on the quarter panels, there's not much you can do to make things worse...

 

In fact, that's the second challenge to a VG approach; VGs work best when fed laminar air flow, so placement matters. They aren't as effective when placed in turbulent airflow, like that seen from wheel well level down on the rear quarter. The air never reattaches after being separated over the wheel well. The obvious solution is wheel skirts, or at very least, mini-skirts. Looked into variations of the Gurney Lip, placed parallel the air flow to segregate turbulent air from laminar regions. Most wheel turbulence comes off the top (it's going 2x car speed), so even a small skirt should extend laminar flow farther down and back. You'd need a full skirt to get smooth below bumper level.

 

So, I'm looking at VGs on the upper half of the quarter panel, starting in front of the tail light and going down to wheel well level. It's only half the available real estate, but it's the only laminar region. We'll see if I even get them on...

 

Have fun,

Frank

 

PS example of Volt rear quarters from Ecomodder's car modification FAQ. Click the video link (to an Ecomodder thread) and scroll down to aerohead's comments. He was the source for post #9. Insightful comments.

Edited November 15, 2013 by fbov

[Recumpence]

OK, here is what I would like to do (aside from the upper grille and other grille revisions) is to go with the modest Kamm-Back as I said before. This will reduce the size of the wake. I also am planning to make some simple vanes on the sides of the car below the tail lights above the bumper. These will be simple 1/16 inch thick black fiberglass or G10 panels just double side taped to the car extending rearward. The point of these extensions is to give the air a sharp end to detach from the body.

 

The only problem I see with the side vanes is the leading edge will not be flush with the body. The panels are 1/16" and the tape is 1/16". So, there will be an 1/8 inch leading edge there. I will do my best to bevel it and I can use some black tape as a transition (obviously just for testing purposes).

 

If you feel like doing more tuft testing, this would be a super easy thing to test.

 

Oh, also, I had already decided on making partial wheel well covers as well. Your testing has confirmed the value in doing that. Plus, rear wheel well covers would smooth the airflow going toward the side vanes I will be making.

 

What are your thoughts?

 

Matt

[fbov]

I've noted before that there's a step at the hatch; the door is lower than the roofline by 1-2mm (Others have confirmed the step, so I suspect it's intentional.), just enough thickness to attach a Kamm-back to the hatch and have it flush with the roof across the top. Bring it down so the bend follows the roofline taper and it's flush at the sides, still only attached to the hatch, and you'll have something I've already seen in my mind's eye.

 

I still can't see a nice version of the side panels. A box-like structure that's flush to the panels where it meets them?

 

I've done something here using the AirTab vortex generators I'd tried in stock form on the rear hatch lid last weekend.

 

As you can see from the last picture, I've taken a band saw to the standard AirTab, removing all but the "wishbone" shape. I then taped sandpaper to the body contours and sanded each piece into a custom contour that fit flush to the bodywork in that location. A little weatherproof silicone adhesive attaches them fairly well (I'm hoping), but takes 24 hours to cure. They're curing now.

 

I'm pleased as punch we hit 60F today, and 65F tomorrow, but warm days are windy, so I didn't do any baselining testing before, and any data tomorrow is suspect. Maybe I'll get a calm 49F evening like last weekend... But I've done a week of highway driving, so we'll see what happens in next week. Hopefully temps will be comparable... but beyond that, I'm not optimistic about December; nice weather usually averages out.

 

As to placement, I'm trying to bracket the little lip on the tail light lens (subtle but visible in pictures). One above, angled into airflow, three below, spanning the distance to the wheel well. You'll note that each side's pattern is different; there's a step at the seam from metal to plastic panel. The left side avoids the seam, while the right side avoids the gas flap - guess which one I did first? They're far enough back that none stick out past the fender profile, but not so far back that the flow has separated. That's why there's 4, not 8 per side. (I've got black for the bottom spot)

 

If I'm right, there will be a vortex coming off the lens. The VGs should add in theirs above and below, hopefully merging with the big vortex coming off the rooflline and not interfering with the good back window performance. To the extent that the vortices form an extended "air boat-tail" and reduce pressure in the wake, I should see an improvement in fuel consumption. If not, I trust the silicone adhesive will yield to dental floss, as the stock adhesive did the first time.

 

This time, however, Winter's on the doorstep, and we've got a lot of long trips to be taken on snow tires unless Winter's delayed. I've almost got pizza-pan wheel covers done, but I still expect the tire itself will still add turbulence, and there's no way I expect wheel covers to overcome snow tire rolling resistance. I figured I had one last chance to make an improvement (last week) and find this weekend a bonus, so things will stay for a while. I have to wait until Spring eventually!  

 

But I can tuft test the back window... the divergence in flow moving out from center is a sign of the roofline vortex, and I can see that while driving!

 

HAve fun,

Frank, who's having fun.

[frbill]

Impressive Frank, I look forward to hearing your results, Good Job! :victory:

[fbov]

Well the high-speed results are in... and there is no joy to be found on the expressway side of my C-Max testing.

 

The chart shows 3 data sets:

- Baseline is roughly Labor Day through Columbus Day

- TP_UG is a combination of the PCM Update plus tire pressures increased from 34-51 psi plus upper grill block

- VG is AirTab vortex generators mounted on the rear quarter panels.

 

The one caveat is the latter two data sets are taken over the last couple weeks - November - so there's very little overlap in temperature ranges. This data is for my expressway commute, taken in morning/evening pairs most days, mostly with ECO-cruise (didn't today). I did cull some data points from the baseline data as there could still have been some learning going on. That results in a non-zero correlation and a realistic curve fit, but...

 

Without the trend lines, this looks like one population of data with a strong temperature sensitivity. In contrast, my lower-speed commute showed real improvements from the PCM update (5%) and increased tire pressures (1-2%). One might argue there was a reduciton in temperature sensitivity, but one does so based on very little data.

 

Note that this data supports no conclusions regarding the efficacy of vortex generators. The PCM update has been widely praised for a positive result. Reduced rolling resistance from increased tire pressures can't go wrong. Given those changes show no affect, one cannot reach any conclusions regarding the slightly improved mileage seen in this data.

 

And I'm starting a new test tomorrow - snow tires. Temps are dropping, the lake effect snow machine will start up tomorrow night and we're going to Vermont for the Holiday!

 

This is one test that I don't expect to show an improvement...

 

Have fun,

Frank

[ScubaDadMiami]

Are you saying that you have had insufficient time to determine the answer, and that you will resume testing after the weather changes for the warmer?

[fbov]

No, but yes.

 

Yes, I plan to continue through the snow season into Spring. No guarantee I'll retain this configuration, but no reason not to...

 

No, the data are sufficient to reach some conclusions. The inability to detect a difference is a very common result in experiments. The conclusion is that low-speed testing for fuel consumption is a lot more sensitive to changes in C-Max configuration; PCM and tire pressures showed positive benefit driving 35-45 mph speed limit roads while no difference was seen driging expressways, 55-65mph.

 

HAve fun,

Frank

[Recumpence]

Thanks for the information, Frank. I am always pleased with whatever data can be gleaned by ABA testing.

 

Matt