Jack, I think your experiment would be a useful data point for comparing minor changes to your 3.8 wing. eg. Angle of attack, height, cord, etc. Your wing is high and reasonably isolated from the body, that seperation will reduce pressure changes on the body itself.

But a comparison to a duck or other tail would really fall down with this approach. I'll go back to my early post; the tail affects pressure on the entire back end of the car (maybe the whole car!). Clearly a duck creates a different pressure on the rear window than does a 3.8 flying wing. That difference won't be measurable at the latch.

True. The question is: what % of the total effect is that. My bet is that it is fairly small -- it is caused by "secondary" wind -- from eddies generated by the tail that flow back against the general flow direction past the car.

Any kayaker knows that such eddies are not to be trifled with... but they are surely smaller than the main current pressure.

Once we get Jack hooked into doing a simple test, we will be able to string him along into doing more and more complex things -- like smoke injection to trace out the eddies... (another great use for Ti).

I suggest putting one strain gauge on one of your rear springs and data logging it on a laptop if you are trying to do this on a budget.

Once that is done all you need to do is get up to speed and record the average deflection.

The relative numbers would then give you the force numbers you are looking for.

Thats as cheap and easy as this experiment is going to get.

Some more input for brain storming

Downforce will have an impact on the ride height resp.the attitude of the car (the higher the downforce the more the rear springs will be compressed and the front springs released). One could measure the actual spring compression or the distance between body and suspension. With torsion bars you could attach a potentiometer and measure the torsion bar angle. With coilovers you could attach a linear potentiometer to the coils. You will need a basic electrical or electronic setup which can be as simple as a battery with a voltmeter and a passenger who takes readings at defined speeds. The readings can be calibrated by measuring the suspension travel when loading the car with different weights (unless you know your spring rates and want to do the math).

A related idea would be to use an electronic protractor (e.g. mounted to the dashboard) to measure the change in attitude of the car.

In any case you will need a nice and smooth stretch of road to eliminate the impact of road bumps and repeated runs (Willow Springs or a drag strip will do, but a runway would be perfect, just watch out for those jets). One problem especially with cars with high spring rates will be, that suspension travel is relatively small. For example with a rear spring rate of 300 lbs/inch (30mm torsion bars) a downforce of 150 lbs should result in a compression of ¼ inch. Not too much to measure but probably enough to get some quick answers. It would be easier to use a stock car as for example 24 mm torsion bars will more than double the suspension travel.

Actually I think a height measure will be challenging due to the small magnitude of the change.

Jack wants to compare changes in downforce. A significant change might be 50 lbs, distributed over 2 wheels - 25 lbs per wheel. Couple this with Jack's 600 lb springs and we have a very small ride height change that would easily be lost in the noise.

Jack,
I have much experience in designing and applying bolt-on strain gages. I have a ton of R&D, design, testing, and field experience in analyzing structures and applying strain gages to measure the changes in stress/strain, due to loading/environmental changes and relating those changes to real world, useable data, such as changes in loading, weight of contents of vessels, etc. I hold a patent or two related to this area of expertize. I think that your idea has merit. Applying strain gages to determine actual and relative downloads should be something we can do.

In engineering (I'm not one), if it's possible, I think simple is best.

The simplest is just to run some consistent laps and see what combination of front/rear spoilers provide the best time. Factor in adjusting attack angles and there goes a whole afternoons worth of testing.

With instrumentation, I think Rob Channell and Guenter are on the right track: Linear potentiometer. In fact, use any type of potentiomter that allows you to measure vertical travel. Connect it between a control arm and the chassis, then "zero" the gauge with your Fluke meter or fancy-smancy data recorder.

Conduct several steady state, track-speed runs while plotting resistance against vehicle speed, return to the garage, get some grid girls that weigh 100 lbs. each (or equivalent) and have each of them sit on the bumper (splayed out across the hood is good for effect but you'll soon forget what you're measuring) until the resistance reading is the same as the road test. Plot the equivalent weight-resistance at 10mph intervals and you should get a pretty good idea of the wing's affect on suspension loading. Graduated weight sand bags are probably more accurate but not as fun. With GGs, you'll have to provide entertainment, food and wine to avoid chick boredom. Can I assist?

Cannibolize some ride height sensors (variable pots) from a wrecked vehicle with variable ride height control and adapt those onto your car. Pic-A-Parts boneyard is a good source for plentiful and cheap.

You'll probably end up realizing there's many combinations of spoiler type and adjustment (angle, height, etc.) along with the usual tire pressure, spring rate, antiroll bar settings, camber and ride height changes mixed in with front-to-rear balancing to make you dizzy. But those are the details few bother with. Those that bother usually benefit from having done it.

Glad I can help,
Sherwood

600 lb springs !!?? - now I know why it felt so different around Willow Springs in Jack's passenger seat.

Just one more idea:

Replace the center decklid lock with a small hydraulic cylinder so that the engine lid is only supported by the cylinder and is not touching the outer rubber stops. Now measure the hydraulic pressure at speed and you can calculate the vertical component of the downforce. If you just want to compare different spoilers you can zero out the effect of the decklid hinges.

Add a hydraulic pump and you can adjust the angle of your rear spoiler in full flight or enter the next import car show and win "best actuator award".

Jack, you want to spend $30, right?
The above seems a simple (brilliant idea)
way of spending even less and obtaining significant preliminary results. Not rocket science, but science. But I think a way has to be found to ensure that the aerodynamic profile of the car (decklid, wing) is not affected by the physical size of the measuring instrument i.e. the rubber bag changing the angle of the decklid.

Umm - maybe you should be careful about replacing the decklid lock with a strain gauge. According to Ludvigssen (p. 646) the ducktail cut the rear LIFT from 320 to 93 pounds at 152 MPH. The ducktail doesn't give downforce, it reduces lift.

I think this hits it on the head. In the conditions I'd be testing in, I don't think measuring ride height would work. At least, here's my thinking: 1) No road is perfectly smooth. 2) On top of that, even slight variations in speed are going to rock the car back (squatting) and forward (whatever the opposite of squatting is). 3) On top of that, any steering correction at all is going to load up one side of the car or another. And all of that will be coming as I try to measure differences in downforce from the wing to an accuracy (I hope) of 5 or 10 pounds. Distributed between two 600 pound springs, in a moving, even-slightly rocking car, on a less-than-perfectly-smooth road... I just don't think it would produce useable data.

That's why my thinking is to isolate the decklid's effect earlier in the chain. Obviously, if I could position a scale at each corner of the decklid, disconnecting the hinges and the latch completely, I could get a fairly clear idea of what the decklid and wing are doing. That doesn't seem to be do-able on my $30 budget, though. But I think I can get consistent results using the latch point for measurements. I agree that it isn't an apples-to-apples comparison when you're talking about different decklids, like ducktail versus RS3.8. The specifics of 'where' on the decklid the downard force is happening would get exaggerated based on its relative position to the edge with the hinges. But for wing adjustments, the latch point seems a viable way to test, to me.

I agree with Sherwood that the simplest way to test is to take it to the track and test. I've already done that. I've been playing with wing angle for about four years, now. But this gives me a chance to get some numbers independent of all the variables I face at the track. It's entirely possible that some of the 'differences' I feel from different wing angles are all taking place inside my head.

I won't be able to actually do this for a couple of weeks, since I'm behind on work (that's why I'm posting here, I guess. ;) ). My current under-$30 thinking imagines using some aluminum pieces and an analog bathroom scale. I can point one of my lipstick cameras at the display on the scale and the other at the speedometer. Voila. Data to analyze.

But that's my simpleton's approach. I really appreciate the input from those of you who know what you're talking about. Two important differences between me and an engineer are that my mind just doesn't work the same way as you guys, and you've all got a lot more specific education working for you. Island's bicycle wheel mount angle is something I would have never thought of. (I still think it would need unrealistically-perfect conditions for it to work, for the specific tests I want to do. But still, I never would have thought of it.)

I like the squeezing the play-doh idea, too. But I think that the duration of the drive would have to be controlled for it to be reliable. A longer drive with less downforce might squish as much as a short drive with a lot of downforce.

Thank you, sir. I want to put this (selectively-pulled) quote on a mug for while I work. ;)

"A longer drive with less downforce might squish as much as a short drive with a lot of downforce"

You can tell he's a writer... :D

I sure would be careful and not try it at 152 mph first. However from my understanding the ducktail is certainly pressing downwards (while the rest of the car want's to lift).

The ductail actually effects flow and pressures forward of and after the rear lid so its not just something that can be measured at the lid.

Its the reason why Jack's idea of trying to measure the force at the lid is limited if he wants to know the tails total effect.

If you read my post you will see that I suggested putting a strain measurement gauge on the spring which is not the same thing as a potentiometer to measure displacement of the rear of the car.

Even with stiff springs you should easily be able to get a 5 lb resolution.

Just glue one of the following to your rear springs and you are on your way to measuring what you need to know.

http://www.omega.com/Pressure/images/XDUCER_GAGES_m.jpg

http://www.omega.com/pptst/XDUCER_GAGES.html

I think it will give misleading results because different tails distribute the downforce differently. Measuring a single point could cause very misleading results if a tail with little downforce, but concentrated at the center were compared to a larger wing that distributed the downforce more evenly.

I've got to think that the wider the tail (IROC, RS America for example), not only is the downforce more effective, but it is spread across a wider area of the rear of the car. A decklid latch might capture a much smaller % of the total downforce with this type of tail versus a smaller, more concentrated tail (ducktail).

Making some numbers up just to make the example clear:

IROC downforce in wind tunnel = 100 lbs
Percent captured by decklid latch = 25% (low because the force has a wider distribution across the rear)
Decklid Measurement = 25 lbs

Ducktail downforce in wind tunnel = 50 lbs
percent measure by decklid latch = 50% (Higher, more concentrated downforce at the single point measured)
Decklid Measurement = 25 lbs

Unless the force from each tail is distributed the same way relative to each other, it could be very misleading.

Here's the article I referred to previously, it was Bobby Allison, not Steve McQ as I thought. They get into some detail about how they measured downforce.

http://members.rennlist.com/emcon5/cd_2_74.pdf

To follow on you lipstick camera idea Jack, why not rig up a simple gauge bar that would hang down past you rear wheel flares. You could mount your camera to view the gauge in relationship to your rear wheels (lips, centers?). If you use the same speeds on the same straight with your different tails, you'd get your relative data. Not precise, but under $30 I'd guess...

Bathroom scale jammed under ductail with lipstick camera duct-taped to inner fender.....30 bucks

A strain guage on springs on all four corners sending real-time data to a recorder or PC.......priceless

I'll bet the F1 dudes use the latter.

Too bad computer simulations aren't there yet for this application. And I mean really accurate ones.

You can't use a scale/strain gage under JUST the latch to measure downforce. The tail/lid look like an "L". The front hinge is at the lower right, the striker at the lower left, and the downforce/drag is produced at the top. Look at the moment produced by simply increasing drag. It will increase the force on the striker plate. SO, you wont be able to tell the difference between downforce and drag.

I have to vote w/ the guys for trying a suspension measurement device. I too think gluing a strain gage onto a spring would be easy and accurate.

SMD

Measuring changes at the spring might be quite difficult unless you were on a perfectly smooth road.

Jack,

Why don't you try and dis-assemble the digital scale and extend the readout into the cockpit....