It's capable of both. It can output live and then it also burns everything to a compact flash card. Live data would be useless for this application, I think. The differences will probably only be apparent in the averages that result from the thousands and thousands of samples.

That would be a good way to test the wing, but not a good way to test how the wing works on the car. For example, the airflow back where my wing is will be coming down at an approximately 10 degree angle, which is different from the flow over the roof. This 10 degree difference in what the angle of attack has always appeared to be and what it actually is is one of the things I'm interested in testing out. My guess is that I've always had my wings set about 10 degrees too steep.

I haven’t read most of these ideas so this may have been suggested....Why don't you rig up a mechanical bellcrank pushing / pulling a cable run through tubing (like a bike brake cable)? The bellcrank would amplify the reading for greater resolution. Easy to calibrate. You could run one from each wheel up to a panel, close to the speedometer, where the ends of the cable would move a flag or arrow in and out on a simple scale. Try to make it 2:1 to suspension travel.

I would think you could rig this up very cheep. Easy to calibrate and very accurate. You could capture readings of the panel and speedometer with a digital camera (on movie mode) for later review and analysis.

Unless you are set on an electronic idea as much for the challenge of making it work for pennies as the data you will get from it. Nothing wrong with that either.

i agree with tim. too subtle. find a real smooth road...replace your shocks with machined rods with integrated strain guages. the machined rods should be cheap...i don't know how much strain guages are.

Jack

I'm not so sure that this is the right way to do this.

You're trying to measure a change in pressure (using your dial gauge), which you will then convert into a force measurement.

For maximum accuracy, you should try to use nearly the full scale of the pressure gauge. This is because (on a hypothetical 50 psi guage) it incurs far less error to read a pressure change of 45psi (+/- 1 psi error) than it would be to measure 2psi (with the same errir) on the same gauge.

So, for a given expected force (100lbs ~ 45N), and for a given number of syringe supports to the wing (2), and for the above gauge (45psi ~ 0.310 MPa) you would chose syringes that have a cross sectional area of 0.75 cm^2 (using Pressure = Force / Area, and metric units because I can't understand imperial!)

i.e. smaller is probably better. As you've found, the volume change required to get a reading on one of these gauges is pretty small, and could be minimised by using braided hose to connect syringe to gauge, and possibly reinforcing the syringe?

Obviously you'd have to mess around with the numbers to suit your application, but I think the theory is pretty sound, and could give some really useful numbers about the DF that the wing can produce on its own.

Unless you can find a really really smooth section of road, I think this technique is going to give you the most comparable data, back to back.

Anyway, just my 2 cents worth. Good luck with the project.

Mark

*edit*

Should have made it clearer that I'm referring to a hydraulic system, using a virtually incompressable fluid. Air changes volume significantly under pressure, so you'd get too much movement at your syringe "transducer" end.

Even if your no forward movement static test worked the moving track surface would very likely scatter return bounce too much.

Quick, pulled it out of my ass in 5 min, needs fine tuning methods.
1. Make a pulley and cable system using bike brake type cable and parts. Well within the 3 tacos and a cold 6 pack cost zone This will allow you to do two things. One, to mount the sensors inboard where they are protected. Two, to "Amp" the movement as needed.
I would amp the movement rather than the signal as the signal is already in the mid range of your read/record system. To amp the movement you can use a disk pulley with the cable wound on the pulley. Mount an arm of any length you see fit on the pulley to "amp" the movement. Then aim the sensor at end of movement arm. Cable moves up and down 1 to 1 with suspension but the length of the sensor end pulley arm combo will set ratio. Disk should be sized so full scale movement will produce 90 deg rotation but wind one extra loop to cover extreme suspension movement and not snap cable. Easy to change as you see fit by moving arm in and out and easy to make. Use a small wound spring to take up slack as cable plays in and out. Mount front setups to the fender in the front on the shock tower. Mount cable to A arm at sway pick up. Rears, drill cable holes and mount setups in back above shock tower but in driver compartment. Run cable over 90 deg turning pulley and down along shocks in shock tube of frame. This will give max "unamped" movement to start.

2. Use same sensor end setup as above but wind the cable pickup points around the sway bars right in the center of front and rear bar and have one front and one rear sensor rig. This would remove single wheel travel inputs and only read overall ride hight data as the bar only rotates in the middle when the overall ride hight changes. My bet is that this would give the cleanest data overall and would be my first try for what you are looking for. Single wheel data would be interesting too but for other things.

Random notes. Do many runs over same track sector to get base line then make one change and repeat. Measure barometric pressure, ground speed and over car wind speed. Head wind will make for more down force at same track speed. Tail wind will reduce. 30 mph wind + or - can make a big dif.


Air pressure change measuring methods would be inaccurate as any temp change will skew your data. The act of moving the air in and out of a shock or piston will change the temp and pressure. Hydraulic methods would be less sensitive to temp but some would still be there. I would also expect to see a very choppy feed from a hydraulic system as there is no natural buffer for vibration from road noise unless you use soft tubing and the tubing would then add inaccuracy.

I do agree that measuring suspension height would be great, because you will see the net aero effect on the entire car - not just the wing.

But I think you will have difficulty getting meaningful data measuring suspension height changes. You've got stiff springs and they won't respond much to minor changes in downforce. If it were me, I'd like to measure changes of 25 lbs or less.

You've got 500lb? rear springs, x 2 = 1000 lbs spring rate. I think a 25 lb change might be lost in the noise.


Here is a way to measure both drag and downforce, pretty easily. It won't show total aero effect on the car, but it will show the full effect on the rear deck and wing combo.



http://forums.pelicanparts.com/uploa...1154581998.jpg


The key is that the wing uprights are attached to the ducktail with pivots. Easy to do. This allows you to put your pressure sensor (syringe) at the decklid latch yet allows you to isolate the effects of drag and downforce.

A second sensor, in this case in tension, is attached horizontally between the roofline and the wing. Thus you can measure drag too.

Key to this setup is the pivots, which allow drag and downforce to be isolated. The front pivots are of course the existing hinges. I would remove the hydraulic cylinder (decklid shock).

Note that to get good measures with the above, the pivot on the uprights needs to be directly beneath the center of pressure (COP) of the wing. The rule of thumb I've heard is the COP is 1/3 back from the leading edge. The wire must be horizontal, and strong enough to support the full effect of the drag.

Just got to find a way to mount that wire.

It could be two wires, from wingtip or upright to rear quarter window. You could still have just one horizontal syringe, and just multiply the measured drag 2X

With some further thought, it seems the drag of the ducktail will create some apparent downforce at the rear syringe. But to the extent that is a constant, you could still get meaningful data.

Is this a constant? I'm sure the wing height / angle has some effect on the ducktail drag. But I'd speculate (just intuition, nothing else) the change in ducktail drag is pretty small given the height of the wing and range of angle-of-attack you'll be playing with.

Jack/Chuck,
Thats not a good way to measure lift/drag. You already figured out the two can couple. And, The syringes would only make a measurement if the two MOVED (well, the tiny amount required to displace the tiny lever in the gauge). If the wing/lig/mounts are rigid enough there is nearly zero movement. If strain gauges were mounted to the uprights, the movement would be <<0.001" W/ the hydraulic setup, you would probably measure more slop in the hinges/mounts than actual deflection from aero forces.

How about building yourself a displacement transducer from a rotary pot? (they exist but are pricey) 1) Attach pot to car. 2) Wrap a fine string/wire around the pot shaft a few times and then to the suspension (so that susp travel 'unwinds' and turns the pot). 3) Wrap another string/wire around the pot the OPPOSITE direction and attach w/ a long spring to the car. The spring will keep the strings in tension (and the pot 'wound'). If the pot shaft is narrow enough you coul measure TINY defelections. A 1 turn pot w/ a 1" diam pulley/shaft is 3.14" of travel, a 1/8" shaft is 0.4" of travel... You would be able to power and hook directly up to your logger.

SMD

Jack,

I've got a ton of wind tunnel experience in a former life, including a bunch of work for Porsche Racing. I'm no longer in the industry (read: no access to the spendy tools and equipment), but am still a pretty big geek when it comes to testing my own projects. Hats off to you for your approaches and here is hoping I save you some time.

First off ride height is nice, but I doubt your going to see a clear correlation between height and downforce. It is there, but it is likely going to get buried in all sorts of noise. I've got some solutions if your after ride height for other reasons, but I'm cursed with fat fingers and rather not type.

As expected, the right way to do this is strain gauges or load cells. Realize that strain gauges are cheap (5 or six bucks each) but for absolute numbers you need to account for materials, side loading, temperature changes etc.. The easier way is to invest in "accurate enough" load cells which can be had for 40-50 bucks. Check ebay for "load cell" and I think you will be pleaseantly surprised. Best case a "S type" load cell, although other kinds will also work. I'll assume your data Acquisity system can deal with a load cell sans amplifier.

Reworking your uprights to incorporate a load cell should be pretty easy, giving you accurate downforce numbers without lateral load effects. Try to get the cell(s) up near the element for a short reaction arm. This gets structural effects of the support out of the picture. If you feeling fancy you can play with some cheap strain gauges at the wing/decklid mount to record drag, but this is slightly more involved.

If the up front costs are too expensive, realize that most of the digital bathroom scales at Kmart/Target have 2-4 strain gauges to yield a load cell of sorts. You'll have to play with getting a meaningful output for data acquisition, and put together a mounting arrangement, but all the parts are there. Only problem is your likely to have relative numbers between airfoil settings and not true calibrated downforce values (i.e. X lbs @ Y velocity). Probably not a big deal in the end.

Even cheaper, which I've used with good success is a hydraulic setup. Head to NAPA and snag some cheap drum brake wheel cylinders. If your really cheap, hit up your local shop or junkyard. Remove the internal spring (if they have one) fill with fluid, bleed and install a pressure sender in the outlet port. Mount as needed using some creativity and measure pressure. Decent pressure transducers can be had for cheap money. Recall that P=FA so measure the piston diameter. Unfortunately, force changes less than 30 or 40 lbs are lost due to seal friction etc. This is the 2 burrito approach your after and will likely be much more accurate than the syringe cable arrangement.

Sorry about getting long winded.

Vin

SMD, I'll respectfully disagree. The sensors need to measure force, not displacement. A bit of slop in the pivots would be irrelevant.

That said, the simple syringe is great for understanding the concept but isn't going to work in practice. It is simply not sturdy enough for the forces involved.

The syringe probably worked great for a model rocket engine producing around 5 lbs of thrust. I believe you need something that can measure full scale in the range of 200-300 lbs, for a single sensor in a given plane. That can be divided if multiple sensors work in parallel.

Vin's suggestion of a load cell is good one for the downforce sensor. You'd need something similar that works in tension for the drag sensor.

i think a cheap scale with a "telltale" on it. kinda like how the motorcycle guys put a ziptie on the shock, and see max movement. then just put the scale in the engine compartment with the various lids ajar slighty. see how hard the lid pushes down. something like a kitchen scale may work.

As for leaving the lid ajar, don't spin. Mine popped loose in a spin once. Not only did I endure lots of comments about my NASCAR style wind flap, I bent the decklid hinge and the decklid split at the seam on one side into an upper and lower section. Just be careful. Mayeb think about adding a safety wire so it can only pop up a certain amount.

How about taking two short pipes where one fit into the other. Then add a spring inside the outer pipe so that the inner pipe is pressed outwards. The principle being the same as for a decklid shock, but instead of gas you use a spring.

On the inner pipe you add an o-ring, and when the inner pipe is pressed into the outer pipe the o-ring will mark the max depression.

Then add four of these to the spoiler, like this:

http://forums.pelicanparts.com/uploa...1154627289.gif

The red lines marks the o-rings.

A rather crude method, but if you know the spring rates you can easily calculate the max force. This method obviously lets the spoiler move, but you can control the amount of movement by the choice of spring rate.

/Peter

Chuck, I see where your're coming from - we need to measure force. BUT *ALL* force measuring devices do so by displacement - however small. A strain gauge does it (you actually have to elongate the resistor), micromachined pressure sensors do it (by displacing a cantilevered beam and measuring the capacitance change), and simple air/hydraulic dial gauges do it (by displacing the bellows inside to move the spring scale).

I originally missed the second pivot in your picture. That does decouple any shift in the lid pivot from the drag 'sensor'. I assumed (wrongly) that it was rigedly mounted and thus any movement of the lid would put force on that sensor. I need to pay better attention to the pic.

It still seems far simpler to glue a couple of strain gauges to the uprights and the exact info we're after w/o any coupling (that can't be mathmatically eliminated) ;) THough I say that because Im an EE. Its also a 4 burrito solution...

SMD

SMD, yes the sensor needs to be a type that allows insignificant movement. Like a load cell, or strain gauge.

Peter, my concern with that spring-based setup is that movement great enough to be measured would allow significant change in angle of attack. Now replace your spring with strain gauges.....

- see what happens to the syringes then go back to this page for the next run(s)....


"I'll assume your data Acquisity system can deal with a load cell sans amplifier."
- see if your data logger lists accuracy, resolution, etc. and what env'l conditions are listed. Post that and folks can help you figure it all out.

If it doesn't list anything then that's a bad sign. But it can always be calibrated with some degree of labor.

Chuck's second pivot point is smart. It solves the problem I faced with the idea of a wire and only letting the decklid hinges pivot. I posted this in the 'Wingin' it' thread:

http://forums.pelicanparts.com/uploa...1154118406.jpg

It still seems to me that what Peter describes with his drawing is the simplest to implement, in a lot of ways. But the springs and a slipring will only show me a maximum reading, which could be the result of gusts or externally-generated turbulance.

If I elongate the mounting holes in the upright so that lateral movement is eliminated, but a small amount of vertical movement is possible, and then use four syringes to bear the actual weight of the wing, it seems like I could get live readings inside the car, easily corrrelated to speed.

It won't give me any data on drag. But drag is a pretty minor concern at this point.