Its all about measuring the force at each wheel, and I will repeat again that the easiest way to do so is by attaching a thin film strain gauge to the spring coils.

It would be cheaper, give better data resolution and be easier to install than the potentiometer solution which is just another way to get the same data via a different measurement device.

We agree on the force each wheel. Measuring the coil springs assumes some level of linearity and uniformity in the coil springs themselves....not necessarily an easy thing to measure once on the car. If you could measure the resistance of each coil spring to a load and could graph the strain gauges responses to changes in inertia, and thermal expansion, etc....you would indeed get a "maximum" average value. I would still want to know how the downforce specifically affected the size of the contact patch of each tire as the speed of the vehicle "increased or decreased".
Rick
'78 930

I'm not sure the tire pressure measurement would work. I think the tire pressure would stay more or less constant (disregarding changes due to heating). The extra force applied down on the tires would result in a spreading of the tire patch.

Some people have used this to do a rough corner weight measurement by measuring the tire pressures and measuring the area of each tire on the ground.

Area in square inches X pressure in pounds/square inch = pounds of weight on that tire.

Hence my idea of 1 strain gauge at the tail itself...

But I'm gonna switch my vote: Do the bike wheel -- in the time to read all these replies, Jack could have rigged the bike wheel & a small camera and be taking data....

After that he can do strain gages -- and then he can correlate the two. If there are big residuals, then he'll know something is wrong. OTOH, if they are well correlated, then he will know he is measuring the same thing (even if it's wrong)...

This is a great application for linear pots mounted on the suspension. One end on the lower control arm, the other mounted to the tub.

With this data, you can accurately measure wheel rate and get ride height changes, dynamically. Of course, you would have to log this data, filter it etc. See: http://www.celesco.com/press/clp.html

Then there is laser ride height data aquisition, which is absolute displacement:)


Just a thought..

A bicycle wheel (or even my spring-scale) would work with a camera recording results. It would only be as accurate as my ability to read the video (and the scale would be only sampling at the latch, a limitation).

I think an instrumentation approach (linear pots, strain gauge, etc) would work if I was able to run a signal into my DL-90 data logger, which would allow me to interpret the data with its software. Ultimately, it'd be more accurate.

Problem is, my DL-90 is in the process of getting fixed. So, next week I'll either address this with analog tools, or wait until the data logger is back up and running, which (fortunately) is now looking like it will happen sooner rather than later.

Interesting project!

I think Jack can determine some useful info. One conclusion I see is the determination of the "sweet spot" of the adjustable wing angle. At some point the steeper angle may not provide significanly more downforce, but drag will increase sharply.

To that end I think if a measuring device can be devised to measure the relative differences in downforce between wings and adjustment angles, and the drag can be measued he can plot the data points and he just might see some trends and some useful conclusions.

so.......for drag, why not use the DL90?

Make some baseline accelration runs. Find a LONG FLAT LEVEL road. Start the plot at speed in say, 4th, or better, 5th gear, and somewhere around 2000 rpm, floor it. Plot the time to rev numbers, lets say from whatever the rev point is that the engine pulls smoothly from, (3K)?, and stop the time reading at say 5K, and repeat several times. Maybe do it in both directions to eliminate wind abberations. Of course, borrow a little weather thingy from a bud to measure windspeed ..or buy one at a yacht supply store (return it later? sleazy, I know!) The result should show some consistant times. At the elevated seeds, drag will be the factor that will affect the acceleration times, so relative drag can be deduced. (The higher the speed, the more measurable the time difference)

Now, make a change and repeat the cycle. If the numbers are the same, drag wasn't meaningfully increased. Keep doing it for as many combinations that seem appropriate. I would start at the extremes just to make sure the results would be measurable and meaningful.

That should help you determine the drag relations.

One issue, of course, is if the bigger wings produce big downforce, it could change the angle of the car, (nose up) which would, of course, increase drag big time. The downside is that drag wouldn't soley be a function of the wing. And it could be adjusted out by either aero tweaks to the front, or ride height adjustments.

Data analysis later will help here.

I like the idea, (and I know it has limitations) of a scale to measure the wings/tails force on the chassis. (Of course, further determinations need to be made to determine the effect on the front, overall aero balance, etc)

Here are a couple of scales that are remote readout, always on (i think) and capable of reading 400 pounds.

http://www1.mscdirect.com/CGI/NNSRCM?PMK0NO=434411&PARTPG=NNLMK3

click on the blue model number for the pic and details.

Sorry, but I was unable to find anything cheaper that would be anywhere near accurate. I think that Tyson should be able to mount one of these and fab up some bracket to attach the decklid to the scale surface. Remove the rubber stops to get the decklid in the proper position and to tranfer the entire load through the scales. You may find that the deck lid flexes and needs to be positioned a hair higher. use clay or chewing gum to see if this is an issue. Again, do the extremes first, so that the others can be mounted the same way, or the already "semi seat of the pants" data will be further devalued and suspect. You may also find that the limits of the scale are insufficient. Perhaps a linkage to reduce the force can be fabbed in such a case. Simple lever arm math will help here.

I am sure that you will see significant differences, along with a lot of noise. Repeat runs will help the "filtering effect" as will common sense. Take an assistant or three! Make one a race car/fab guy with tools (begins with a T, ends in an N), and the other a not too anal engineer!

To make it easier, mount the display (lengthen the wire for this) right next to the tach and speedo, so that you can correlate downforce pounds and speed. Video tape it all! Take careful notes and keep al the DL90 data. Take gas in seperate cans to maintain consistant weight on the front of the car.

I know it's not a windtunnel, but I think you can get some good info (Not great!) about the relative differences between the wings, and more importantly, the point where steeper adjustment angles result in more drag than downforce.

(You can ponder the compromise of downforce vs lap times later!)

Jack,

I vote for potentiometers at each wheel. Something to measure wheel travel.

I suppose a simple set up would be to mount the pot with a linkage that rotates the pot relative to suspention travel. Sort of like the A-R bar links do. They can be calibrated against the force required for various wheel travel measurements.

You will have to smooth the data a bit because there will be some noise from road bumps, but it is doable. "Car Life" magazine did this in 1969 when they tested muscle car spoilers and Car & Driver did this in 1974 when they tested the Carrera duck tail.

Regards,

Jerry Kroeger

Glueing a thin film strain gauge to a spring coil is a lot easier than setting up a pot with levers and linkages and gives you exactly the same data. Mechanically actuated pot gauges are a pretty obsolete and innacurate way to measure small movements now that thin film strain gauges are cheap and abundant.

For those who are unfamiliar with these things here is a pic and a description of how they work. I worked for an instrumentation lab for a couple of years when I was still in college so you could say I have a little expertice in this area. Epoxying one of these postage stamp sized devices to a spring coil is a lot easier than all the other suggestionss that have been made on this thread.

http://www.omega.com/literature/tran...strainpic1.jpg

http://www.omega.com/literature/transactions/volume3/strain.html

Is there enough surface area on a coil spring to properly mount a strain gauge? How much per gauge?

Sherwood

Yes with the smaller ones. There are a lot to choose from but you can get a package of 10 for about $50.

http://www.omega.com/toc_asp/frameset.html?book=Pressure&file=STRAIN_GAGES

Explain more about the strain gauges -- and remember, I'm mildly retarded. I just epoxy one to each spring, and then run 12v through it and into my data logger?

Jack: As you remember...I come from the film industry...so my Idea would be to put the car on scales (standing still) and use wind machines to force air over the car.
We used this method to set up a Formula Ford some years ago...to check the angle of attack...and the proper place to put an oil cooler.
Using small strips of video tape stuck on the car, you can clearly see all the air passing over the surfaces.
With your car...the scales would tell you how much downforce and at which corner or front or back.
I am sure the tail would exert pressure on both the front and rear.
Bob

I just realised that this was my 911 th post...and I could not think of a better subject!

Do you know how much voltage your datalogger will resolve? several nV maybe?

Most strain gages are made of Constantin. It has a small change in resistance with applied strain.

You would probably have to get an instrument that is sensitive to small current changes. See that Omega site.

There are many place to read about strain gages. They will repeatably measure strains up to about 5%. You mount them on a beam sized for the range of forces that you expect. Thicker beams require more force for the same strain.

Beams in tension are the easiest to figure out. stress=F/A. strain= E/stress. E is a material property. Omega would have design manuals if you were really interested.

It would be a lot easier and cheaper to find an already calibrated force gage with a strain gage in it.

Good luck

The trouble with these glue-on strain gauges is thermal drift and calibration required. Although I have never instrumented a coil of a coil spring. Not sure how they would react.

Linear pots also require calibration and 'zeroing', at ride height, but should provide less drift as this is absolute displacement.
Chris

Even for a flat-bottomed road car, there is some ground effect to take into consideration that static testing may not reveal.

Although a little off topic, 'flow-vizzing' your car is a really cheap and easy experiment. Basically mix a colorant (white in your case) of some kind and gearbox oil. Place a small 1/4" dot in various places and run your car at speed. You'll get to easily visualize actual air flow and points of separation. See attached picture as an example..
http://forums.pelicanparts.com/uploa...1110852367.jpg

Then there is the other end of the spectrum:) Come on Jack, let build a 45% scale model..
http://forums.pelicanparts.com/uploa...1110852773.jpg

In a nutshell you will need to make a wheatstone bridge circuit (see link below) that will convert a 12 V tap from your cars power bus to a a voltage range that your data logger will be happy with. Hopefully your DL will have an input that will sample a 0-5 volt range with good resolution. Since the car's voltage will not be consistent it will also be necesary to splice a zener diode into the 12 v line powering the WBC to clamp the input voltage.

http://www.omega.com/literature/transactions/volume3/strain2.html

Once that is done you will need to peform some initial tests to determine the voltage change that occurs when you add several lbs to the rear of the car. From this you should be able to determine your lbs/v-diff factor you can analyze your data with.

The resting state resistance may change depending on the temperature of the spring gauges so once you have a data logger full of data you should reference everything to the resting state voltage at the end of your run.

Its all relatively easy, but since the strain gauges will be impeding small currents it will be best to use shielded wires between the WBC and the strain gauge, especially if you have unshielded spark plug wires.

You might try something alonge these lines:

http://www.a-dat.com/sensor/sensor_ride_height.php

It's a laser ride height sensor that outputs 0-5V. In the end, average ride height is probably a pretty accurate measure of downforce and this hootamagator seems pretty easy to set up.