How to do pics of tuft testing
 

One thing I really notice is that very few pics of tuft testing are posted here. As a zero cost way of seeing where there is attached and separated airflow - and also airflow direction - tuft testing is brilliant.

I like to use still image cameras for this (ie as opposed to a video camera), taking a sequence of photos and then picking the best single one that is indicative of the tuft pattern. Then I usually crop it a lot.

Photographing from the side of the road when the car passes you at ~60 km/h (35 mph) is usually fine. (Probably, you can do it in the street in which you live!)

To do this type of photography, I think that you need a camera that has:

• a fast shutter response, ie it takes the pic the instant you press the button (most dig cameras these days)
  
• the ability to set the required high shutter speed (most good consumer cameras and all 'prosumer' cameras)
  
• the ability to do a sequence of shots rapidly, one after the other (varies lot between cameras - look for burst mode of, say, 8-10 shots)
  
• a telephoto lens focal length that retains good resolution (again, most good consumer cameras and all 'prosumer' cameras, but be wary of digital zooms)

So, a sequence of pics as the car drives by me:

https://i.postimg.cc/fbdMK0d6/sample.png

And then after selecting and cropping:

https://i.postimg.cc/tgFHPGQb/000122.jpg

Final shot:

• Nikon D300
• Focal length 195mm (35mm equivalent)
• F9 aperture at 1/500th of a second shutter speed
• +0.3 stop exposure compensation

(This pic is showing the airflow pattern without Edgarwits front external air curtains)

And here is another, literally taken in the middle of suburbia (the driver was braking for the culdesac at the end of the street - note neighbour's post box!):

https://i.postimg.cc/yYyqVC43/Prius-8vg.jpg

An industry standard for testing is the 100ft circle (usually for G-forces).

With a camera on tripod and operator at the center you could record crosswind conditions efficiently.

I have done extensive amounts of skid pan testing, including at car manufacturer's proving grounds, on test facilities, and on my own circles (chalked on roundabouts for HPVs).

I do not think this is a good guide to crosswinds, as the air yaw angle is constantly changing - and the 'crosswind' component is relatively small.

On my 'home' skid pad:

https://i.postimg.cc/SsZ5v3R2/DSC-0019.jpg

(My first recumbent trike: it met the skidpan criterion but it was waaay too heavy.)

And my Lexus on the skid pan

http://us1.webpublications.com.au/st...3/1317_7lo.jpg

I did some tuft testing with my ugly old Honda Civic Wagon. Lousy photos, didn't get any help!

Hood, windshield and roof looked good.
https://ecomodder.com/forum/attachme...1&d=1600770451
https://ecomodder.com/forum/attachme...1&d=1600770598
https://ecomodder.com/forum/attachme...1&d=1600770598


Behind the front wheel:
https://ecomodder.com/forum/attachme...1&d=1600770598


It looked like air was going under the side skirt:
https://ecomodder.com/forum/attachme...1&d=1600770598


Side mirror:
https://ecomodder.com/forum/attachme...1&d=1600770598


Side mirror removed:
https://ecomodder.com/forum/attachme...1&d=1600770598
I was so surprised by the air moving upwards behind the A pillar that I added more tufts!

I have the same problem. I think I've convinced my Mother to drive my car and then I can try to take pictures from the side of the road or follow in her car and try to take pictures or atleast observe them and take notes.

Yes, very seldom any problem in these areas on any car of the last 30 years

Looks excellent with smooth wheel cover

Often the case - airflow is travelling towards the low pressure area under the car.

Flow on the door window is better with the mirror present! Mirror present: main door glass flow separation only behind mirror support. Mirror not present: main door glass separation more prevalent.

Flow pattern directly behind A pillar is common - I assume it indicates the developing of a trailing vortex off each front pillar. Removal of mirror appears to have increased the area of this flow substantially. It would be interesting to do throttle stop drag testing with and without the mirror.

How is this?

https://lh3.googleusercontent.com/pw...-no?authuser=0
https://lh3.googleusercontent.com/pw...-no?authuser=0

Need standard / modified shots.

I don't have a direct picture of the stock setup. I will get one tomorrow when it is daylight.

https://ecomodder.com/forum/member-m...53874367-n.jpg
https://ecomodder.com/forum/member-m...825-142223.jpg
http://momentcar.com/images/1995-escort-2.jpg

Here is a video I shot of it. I was trying to center the shot in the first 2 shots. The 2nd one didn't really work out. If you slow down the last few seconds it shows the modified version pretty well.
https://www.youtube.com/watch?v=RavN...hannel=M_a_t_t

I had also tested the right front window. That car never had a mirror there. I can't explain why the two sides were so different! Multiple photos show the same patterns. I taped over the mirror hole on the left side. I wish I had known about throttle stop testing when that car was still roadworthy. Maybe I'll do some testing with my Insight if I ever get it running correctly.

https://ecomodder.com/forum/attachme...1&d=1600828169

https://ecomodder.com/forum/member-m...5-dsc00892.jpg
https://ecomodder.com/forum/member-m...6-dsc00893.jpg

I also remade the modified GIF. Much easier to look at.

https://lh3.googleusercontent.com/pw...-no?authuser=0

Now for my armchair analysis, let me know how agreeable it is. It appears to be working, atleast better than the stock version. The part on the bumper is angled too sharply and/or because of the spinning wheel and is getting early separation as a result. The lower half shows this because it is flat against the body at the top. The upper half by the tail light seems to be working the best, but possible interference from the rounded trunk edge and wing. Tufts 2 and 3 (from the top) are what I should try to get all the tufts to do.

camera boom option
 

I've shot video of tuft tests on numerous cars by mounting a small camera on a boom that's attached aft of the area being filmed (obviously to minimize aero interference from the camera rig).

3 examples:


1) 2000 Honda Insight


https://ecomodder.com/forum/attachme...1&d=1600885085


https://www.youtube.com/watch?v=TmEeiGZwmS8



From thread: MetroMPG's Honda Insight boat tail extension (cardboard) tuft video; ABA test +9.7%

The cord tied between the two booms adds tension to prevent the camera from bouncing.

2) Similar set-up on my 2nd gen 2004 Toyota Prius:

https://ecomodder.com/forum/attachme...1&d=1600885582

https://www.youtube.com/watch?v=03vLhpPKAgE

From thread (starting around post #100): The Premier's Limo: 2004 Toyota Prius "winter beater" ecomodding thread

3) And the current car (1990 Miata MX5) - I didn't get a shot of the camera boom itself, but in this case it was a rigid DIY tripod of aluminum tubing, liberally duct taped to the rear & side of the car:


https://ecomodder.com/forum/attachme...1&d=1600886078


From thread: "MPGiata" modding thread: aiming for 50 MPG with a 1990 Miata. Update: success, 55 MPG


The other two tripod legs are out of the frame.

---


Cameras with wider angle lenses are better because you can use a shorter boom (Prius boom is probably 1/2 the length of the Insight's).


Obviously this approach is only practical/safe if the road you test on is free of obstructions/pedestrians/cyclists!

You got me thinking. This is what I came up with:

https://ecomodder.com/forum/member-m...8-dsc00900.jpg

I meant: I need to see pics of the car being tuft tested in standard and modified forms. I learnt long ago not to assume what the standard tuft patterns would look like!

Yes they look good! I've always been worried by the camera supports influencing flow (even upstream) but it doesn't look like that is happening, does it? Unfortunately I'd be reluctant to run that set-up on public roads (even one glance from a policeman and I'd be in trouble).

I don't understand. I posted a gif and a link to both sides being tested. Unless you mean it needs to be the same side compared?

I am not trying to test the wheel skirts. Just the separation edges.

https://lh3.googleusercontent.com/pw...-no?authuser=0

https://lh3.googleusercontent.com/pw...-no?authuser=0

Sorry, I assumed that I was meant to be commenting on the flow pattern with all the white panels attached? (ie versus no white panels?)

If all I am meant to be commenting on are the separation edges, then yes, they seem to be working fine. (But I have found with separation edges that measuring pressures is a better way of assessing their effectiveness.)

I attempted to make a puck for testing, first attempt didn't turn out so well. I suppose it is time to try again.

Not sure I've mentioned this here:

https://i.postimg.cc/1t5sCw3z/pressu...ch-cropped.png

They are commercially available.

Yes, that's always in the back of my mind! Definitely a potential drawback of this approach.


Lucky for me I live near the edge of town, about 30 seconds from some quiet rural roads.

How important do you think the plate/puck part thickness and shape is? I used 4.75mm (3/16") flat stock. I think I was successful this time.

That's a good question - and I don't know the answer. The professional Scanivalve one is 1.27mm thick, and mine is 4.4mm thick. I think yours will be fine. Round the upper edges.

In fact, I have often thought this would be a very valid criticism of my pressure measuring techniques with the puck: How thick is the boundary layer where you are measuring, and does the puck disturb that?

And the answer is: I just don't know.

Of course, flush holes in the car's body surface would be best, but that's a bit hard in your daily driver.

When developing the measuring puck, I went through a whole bunch of techniques, with Dick Barnard commenting, and this was the best I could come up with. (I'd not seen the Scanivalve one at that time.)

And, when my on-road measurements matched Jaguar's CFD, I became pretty confident.

And Cr45's suggestion (here) of using the pitot static port as the pressure reference (instead of a sealed vessel) was revolutionary. (I have, via PM, asked him for his real name so I can publicly thank him in my books and Youtube videos, but he is shy.) His suggestion took my pressure measurements to another level - indeed, a radical improvement, and a major reason for my doing the update book.

So I am now pretty confident that these aero pressure measurements reflect reality, but with the caveat that the further you are towards the front of the car (and so the thinner the boundary layer), the less valid may be the pressures.

Thanks for finding this; I just emailed them to see if I can buy some.

As a first approximation, the Scanivalve has a 45° chamfer.
Permalink #?

Yes, my puck has rounded top edges.

I can't remember when Cr45 made his/her suggestion - months ago.

You could look here: https://ecomodder.com/forum/search.p...nduser&u=15140

This was the post: https://ecomodder.com/forum/showthre...tml#post625892

A really great idea - so obvious in retrospect (but I'd not thought of it!).

Here in Australia I had to go through the Scanivalve local agent. That all went well, and I placed my order (and paid for it) only to then be told there was a minimum order amount (ie $ amount) - and I'd have to double my order. I declined, so I never got any of them.

Scanivalve sales got back to me this morning; in the US we can order direct from the company. The patches are $9.40 each, and there is a $75 minimum order. So, I'll have extras if anyone wants to try one.

I tried measuring the pressures on the separation edge I had been trialing, but my set up wasn't that great. So I will re do it, but I figured I might as well post where I put the puck for testing to see if anyone thinks I should try in a different location before I repeat the same test. I didn't do any testing with the bumper portion of it, just the lights. Is it important that I test on the same side if I do a there and back check?

I also noticed that the modified side is clearly dirtier than the stock side.

https://ecomodder.com/forum/member-m...k-modified.png
https://ecomodder.com/forum/member-m...puck-stock.png

I will have to look around and see if I can rig up a pitot reference. Right now I have the propane tank seen in the trunk as a reference.

That looks fine but you will probably be needing to use a pitot tube reference rather than a pressure tank (not sure which you are doing). Just leave it in that position for your two-way average.

I just edited to add I was using a tank, but will look for something to rig up the pitot tube.

Tank is fine for most testing but edge-of-wake testing is hard to do accurately without a pitot static port reference. Don't forget the pitot needs to be centre, front of car, at least 800mm above hood.

What is that number based on? Asking because in one of your other posts you used 2.2 meters off the ground and front corner (at least in picture it was).

Permalink:
https://ecomodder.com/forum/showthre...tml#post625899

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

Is it situational based on where you the probe you are measuring with is or?

That big pole's height and location were my starting point, but I found a central, front pole of 800mm gave the same results. And a bit less obvious!