Pelican Parts Forums - Poor Man's Ride Height Sensors: some (more) data

Pelican Parts Forums (http://forums.pelicanparts.com/)

- Porsche 911 Technical Forum (http://forums.pelicanparts.com/porsche-911-technical-forum/)

- - Poor Man's Ride Height Sensors: some (more) data (http://forums.pelicanparts.com/porsche-911-technical-forum/305438-poor-mans-ride-height-sensors-some-more-data.html)

Here are some quick track day data.

I went out the first lap with the wing set in the middle position, the splitter attached, as well as the new 'dive planes' (which of course I'd misunderstood the purpose of when I made).

First lap: 1:32.XX (Wow, that's pretty good.)
Second lap: 1:30.XX (Okay Mr. Lap Timer, now I'm paying attention.)
Third lap: 1:29.8 (That's my best lap time ever at Willow!)
Fourth lap: 1:29.2 (Mother of mercy, that's even better still!)
Then, of course, some rich dude in a Cup Car spun and the course went yellow.

But for some reason I was a full second faster than I was a month ago. Same tires. Same track. The only difference was the addition of the 'dive plane' thingies.

Did they do something for my front ride height? Hard to say. At the time, I was so happy with the lap times that I decided to leave them on and save the really rigorous and detailed testing for a later straight-line test day. I still don't know what they're doing (if anything), but I was a full second faster.

Granted, temps were lower. Granted, Toyo RA1's continue to improve with age.

But still.

I continued driving, having too much fun for serious testing. I laid down a handful of 1:29.XX's. (The data logger says my 'best theoretical' lap is a 1:28.7.)

Later on, I took the wing and splitter and dive planes off, so I could at least get a data comparison between aero and no aero. Here's what it looks like when I set down a 1:29 with the aero against the best lap (1:32) I got with just the duck.

Remember, a higher number on the graph means the ride height was lower (springs more compressed).

http://forums.pelicanparts.com/uploa...1160621765.jpg

And to demonstrate that it's not just the ride height sensors jumping around, here are three 1:29.XX laps grouped with the fastest duck-only lap.

http://forums.pelicanparts.com/uploa...1160621867.jpg

And although I don't have time to go through the data carefully at this point (too exhausted, too hungry -- what a great day at the track), here are some raw numbers comparing the fastest with-aero lap against the fastest no-aero lap. It's not apples to apples, since the no-aero lap was in the afternoon. But even if you corrected a full second or a second and a half for the time of day, the aero is still helping out.

http://forums.pelicanparts.com/uploa...1160621969.jpg

Wow looks like the dive things are helping front down force. Not sure how after reading the other posts. Very interesting to see how your analysis turns out.

Also, in that Mitch Rossi book he mentions where he had his aluminum dive planes made (some place in LA). They are aluminum and look like they could easily be made removeable for track day installation. They actually look pretty good compared too...others :)

Jack,

My seat-of-the-pants meter really felt some improvements when I added the dive-plane winglets to the racecar for nationals. In fact when I moved them from around 30 degrees to 15 degrees I noticed no change in grip, but got 100 more RPM on the main straight.

If you think of it, perhaps some wool-tuft testing when you do the new runs can show us what they are doing!

One odd thing is that I was hitting the rev limiter with the new setup on the main straight. This means I was going 3-4 mph faster than previously. Maybe the dive planes do generate vortexes, and those vortexes (by reducing air transfer under the car) reduce drag?

It seems likely until you think about how randomly I shaped the dive planes. Still, there's always a chance I got it right on the first try. Or maybe dive-plane vortex generators aren't very sensitive to design.

Or intent. :confused:

wow, 3 sec from the wing, winglets and splitter. It would take about 75hp extra on the engine to improve lap times that much.

I looked at 20+ diferent variations and concluded that they aren't that design sensitive or that the designers hadn't a clue what they did.

Here's all the ride height data in one place. Some of the height changes can be explained by the torque the acceleration gives to the suspension. Some of it can be explained by elevation changes. Some of it is aero. If you feel your way through it (or if you're familiar with Willow), you'll see that ride height front and rear is always lower with the aero stuff in place than with the duck alone.

http://forums.pelicanparts.com/uploa...1160771370.jpg

And totally contrary to my initial findings, it looks like the front is getting a bigger push down with the aero in place than the back is. I might benefit from a more aggressive angle of attack (I was running the middle position) in the rear wing.

Jack, great data. Thanks for sharing.

The front spoiler being closer to the ground isn't creating downforce (in itself). It keeps air from going under the car. That allows the low pressure area behind the car to esentially 'flow' underneath and 'suck' the car to the ground. THAT creates downforce. Side skirts and rear venturies help the effect (by limiting air entry, and helping air extraction respectively). My understanding that w/o sideskirts, the lower front lip wont do much good.

The forward extending lip of the front spoiler *IS* creating downforce. There is a high pressure area infront of the bumper. Presses up on the underside of the bumper, presses down on the lip. pressure * sqin = # force. If you look at the IROC bumpers w/ an integrated lip, the lip only extends as far as the bumper above it. The downforce on the lip is only canceling the upforce on the underside of the bumper. Of course, stock bumpers have no lip, but they do experience the upforce. The farther the lip sticks out, the more downforce you generate (untill you go so far as to be ahead of the high pressure zone.) Why do you think noone make pointy nosed cars anymore (like the 924 and gen1 RX7)? ;)

Best,
SMD

I didn't go out to do any straight line tests this Saturday. But it's interesting to see the track data both confirming and sometimes contradicting the straight-line data. I think it brings out two shortcomings in my straight-line 100 mph tests. One is that they don't take into account acceleration, which includes the changing torque forces on the suspension as the car acccelerates. In the track data, you can see that there isn't a strictly linear relationship between speed and ride height. Part of this is elevation changes in the track, but some of it has to do with how ride height not only changes at different speeds, but also that the car behaves differently in a curve (turn 8) than it does on a straight path (front straight).

Here are two (big) charts that show front and rear ride height differences between a lap with the aero in place, and a lap with only the duck.

For reference, here's a track map.

http://forums.pelicanparts.com/uploa...1094433144.gif

(Remember, ride height is INVERSE to its position on this graph. When the line goes up, it means that that end of the car is going down.)

Front Height:

http://forums.pelicanparts.com/uploa...1160859911.jpg

Rear Height:

http://forums.pelicanparts.com/uploa...1160859950.jpg

The surprising part of this data is that the wing and splitter combo has a more dramatic effect on the front of the car than the rear. Contrary to my 100 mph straight line tests and my earlier track sensor data (I think it discredits my initial front sensor apparatus), it points to the fact that I might benefit even more (in terms of downforce) by running a more aggressive wing angle in the back.

Another interesting look at the data. Here's front and rear ride height on the same chart, from when I was running with the aero in place. I wanted to get a look at how the front and rear behaved under full throttle, both through turn 8, a sweeper, and also the front straight.

http://forums.pelicanparts.com/uploa...1160860239.jpg

If you look at the line for front ride height on the left side of the graph, you'll see that the front lifts up when the car begins to accelerate. Then, as the car finishes reacting to the initial throttle, the line for the front starts rising at the same rate as the speed line, which means the front is now dropping down -- presumably from aero.

The rear is doing a similar thing, sort of. It doesn't drop down so much initially (which you'd expect from accelerating), partly because the car is coming off of a hill in turn 6. But then it joins in with the front ride height line and the two are more or less parallel (both ends lowering) up until about 115 mph. Then the rear diverges. The front keeps dropping down (meaning the line keeps going up on the graph), while the rear end of the car starts to rise up. Some of this might be a cantilever effect coming from the front splitter doing its job. But I wonder, since turn 8 sweeps to the right, if the rear end is effected by air coming laterally under the car as the car turns. I say this because if you look further right to the acceleration on the main straight, the front and rear behave differently. Both lines rise up (meaning the car is getting pushed down), and both continue to go in the same direction right until the braking zone. There is a hesitation in the back end at 120 mph this time (that might be the shift from 4th to 5th gear), but the line keeps rising (rear end dropping down) on the straight, whereas it reversed on the sweeper (rear end lifting up).

Next time, I'll have to try different wing angles and also maybe attaching side skirts to see if that changes the way the rear end behaves through turns 7-8.

In an age of $30 accelerometers, it seems like it wouldn't be hard to rig an active wing that would engage whenever the car is cornering or braking. I'm sure the electronics in my data logger could support it. Heck, the data logger could run a programmed regimen of wing angles based on fixed GPS locations.

Jack,
There you go. Things are ticking. A couple of servo motors with feedback position and ........

Visit the local Pik-a-Part yard and scam some power seat position motors/sensors.

I'm feel fortunate to fantasize through your efforts.

Sherwood

I've been mulling over something similar using selenoids triggered by the brake circuit. Using a fixed wing angle for "normal" driving and then adding wing using the selenoid under braking... In a perfect world it would also measure accell/decell and work accordingly.

You’re having fun. Nice work!

So it crossed my mind to try and figure out what you theoretically should be reading on those sensors of yours, ie what will “fast” look like. My estimate (I’m not good enough to call it a calculation, but I think it’s better than a guess) is that you might want a ratio that looks something like .4” lower than static front, .6” lower than static rear. Here’s what I did, and why those numbers are no better than what you paid for them:

First, you’d think you want downforce roughly in proportion to the weight balance of the car. I estimated the following: 1060 lbs front weight, 1600 lbs rear, 23mm front torsion bars, 600 lb rear springs, giving a wheel rate of about 330 lbs per inch front and 575 rear (estimating here, because I don’t know the motion ratios for your wide-body, non-standard car).

Using these numbers, if you had 200 lbs of front downforce you’d want 300 lbs of rear. 200 lbs of front downforce (100 per wheel) would compress the suspension .3”, while 300 lbs rear would also round out to .3”.

*At this point in the calculation, it’s obvious that there is a conflict between the numbers I’m getting and what you’ve calculated. For example, your calculation suggests that the rear is sprung at 204 lbs/inch for both wheels (I took your 82 lbs and .26” at 20 degrees found earlier in this thread minus your 31 lbs and .01” at 12 degrees and subtracted them- 82-31= 51 lbs, .26”-.01”= .25”, 51 lbs/ .25”= 204 lbs/inch). That seems really soft for something with two 600 lbs/inch springs under it- probably means we both made math errors. It would be pretty easy to measure ride height with and without the 200 lb bag of sand over the axles to figure what the right numbers are.

If I use my numbers, your 20 degree wing position (.08” lower front, .26” lower rear) means 53 lbs front downforce, while the rear would be 299 lbs. Sound possible?
The .82” lower front, .4” lower back numbers sound less probable: 540 lbs front downforce (that’s a lot!), 460 rear.
*One thing to note here- I think it is probable that your front downforce really is changing as you change the rear wing angle- those front splitters are ride-height sensitive, and as you move them away from the ground (by pushing down with the rear wing) them make less downforce...

So if we buy into any of the above, if you could make the front and rear suspension compression the same (in absolute terms, from a static ride height) you’d have a balanced, fast car. This ignores two important factors that argue for more rear downforce, however:
First, race cars with aero are usually set up to understeer more as the speed builds. Seems a little predictable understeer and a high speed sweeper is a good way to keep the tail behind the nose, and that gives drivers confidence.
Second, and more significant, is related to nose-dive under high speed braking. Jump on the brakes and the nose goes down (obviously). This changes the “rake” of the car, which moves the center of pressure forwards (putting more downforce on the nose) and it moves the front splitter closer to the ground, also weighting the nose. The result- the front sticks and the rear doesn’t, resulting in scary stuff and even rear lockup in high speed braking. *BTW, some cars like the McLaren F1 and Merc SLR have spoilers that pop out only under braking to counter this effect.
Both of these points argue for limiting how much front downforce you make. Calculating the exact number is a little beyond me, but my calibrated guestimate says the car will be fast with a ratio that looks like .4” of front compression and .6” of rear. If I didn’t make 12 mistakes in the math.
In practice, all of this calculating is obviously pretty useless, but I find it interesting to check if what I *think* should be happening is what actually happens. I also worry all the other things could be effecting the data- if, for instance, your shocks are stiff enough in compression or rebound, they can “jack” the car up or down over bumps- might be a good idea to set them full soft in your straight line testing so you have an apples to apples comparison with static heights.

From a practical point of view, I don’t think you’re going to get better than an old racing rule I’ve read:

“Put as much downforce onto the front of the car as you can manage, then increase the rear wing angle until high-speed understeer becomes a problem.”

Good luck!

PeteVB, that's very helpful. I think you're correct about there being some errors in both my methods and my math. Your own estimates of "20 degree wing position (.08” lower front, .26” lower rear) means 53 lbs front downforce, while the rear would be 299 lbs." are probably very close to what's actually happening. It fits very well with theoretical calculations of what the wing should be doing.

The .82” lower front, .4” lower back numbers sound less probable: 540 lbs front downforce (that’s a lot!), 460 rear.

I agree. Those numbers look funny to me. I went back through the data and it looks like the rear ride height graphic is correct. However, my math was just plain wrong. Comparing the numbers in the graphic, the ride height difference is actually .28 inches at that point on the track.

http://forums.pelicanparts.com/uploa...1161553006.jpg

But the gap on the front ride height looks improbable, even if you don't re-do the math. The gap is too great at low speeds, where aero should be only playing a very small role. When the car's speed drops down to about 55 mph, near the middle of the graph, I would guess that the ride height difference is not as great as the graph suggests. I suspect the front chart needs a uniform correction, and I'll show what it looks like in this manipulated graphic:

Comparing that graphic to the one for front ride height, it jumps out as odd that the front ride height is so consistently lower with the wing/splitter -- in a way that seems to not be reduced by low speeds. Here's the graphic for the front ride height as it came in from the data logger, and then (immediately below it) a line that I've manually 'corrected' to what I think it probably should look like.

http://forums.pelicanparts.com/uploa...1161552185.jpg

Now, obviously it's not fair to just move lines like that and call it data. But I think I can explain the anomaly -- based on some new data I got doing some more testing yesterday.

Some of you may remember the diffuser I put together a while back. I did it without any input from actual aerodynamics studies. It was a quick and dirty look-alike, based on what I'd seen on other cars and what would fit under my engine. I've since learned that the angle at the exit is too steep, and that the shape of the flat-bottom piece on my car is not at the right angle for any kind of optimized aero performance. So I decided I'd use the old diffuser as a source for scrap aluminum for future projects, and work on a real underbody kit later. But before I cut it up, I thought I'd take it out and test it.

Before I get to those results, though, here's something that jumped out at me. My testing plan was to run my testing course in this order:

1) A 'control run' with the rear wing at 20 degrees and the 5-inch splitter in place.

2) A run with the diffuser under the car and also the front curtain attached to the front splitter.

3) A run with the diffuser under the car but with the front skirts removed.

4) Another 'control run' again with the rear wing at 20 degrees and the 5-inch splitter in place, nut no diffuser or front curtain.

To make it clearer, here's a picture of the front curtain on the splitter:

http://forums.pelicanparts.com/uploa...1160272005.jpg

(Although I also had my side dive planes attached for all the tests)

And here's a picture of the diffuser (from back when I was test fitting it):

http://forums.pelicanparts.com/uploa...1112209185.jpg

I'll get to the results in a minute, but here was the part I was surprised by when I looked at the data. Even though I did multiple runs in opposite directions to get a better data from each configuration, the front ride height was dramatically different between the initial 'control run' session and the final 'control run' session. The rear ride height stayed consistent, but the front appeared to be running 5/8 of an inch lower the second time I did the test, which simply wasn't possible.

But here's the catch -- between the first and last runs, I'd jacked the car up, which I believe is re-positioning my front ride height sensor when it gets fully extended and recompressed. I think the sensor is giving consistent and repeatable results in between jackings, but that it's not giving data that's consistent if you look at sessions that are separated by the car being raised up. It explains the gap in my track day test charts, and makes the data I got yesterday work, too.

So one of the things I need to do next is to improve the way the front ride height sensor is mounted so that it stops shifting position. I'll try to do that before my next track day.

So bad math and a shifting sensor mount are two of the bumps in the road to my aero data being reliable. But here's a new wrinkle that I didn't expect. The diffuser seems to have an effect on my car's downforce, in spite of the fact that it was a total shot-in-the-dark design. At least, the straight-line data have convinced me to test it next month on the track.

In a nutshell, here's the data from a series of tests (that did not have the car jacked up to interfere with data consistency):

With the diffuser in back and the skirt in front, the front end of the car was compressed 1/3 of an inch (.3214") lower than running with the splitter and wing alone, and the rear of the car was compressed 1/5 of an inch (.1921") lower than running with the splitter and wing alone.

And oddly, with the diffuser in back and no skirt keeping air out in front, the effect was more mild. With the diffuser but no front skirt, the front end of the car was compressed less than 1/10 of an inch (.0785") lower than running with the splitter and wing alone, and the rear of the car actually lifted less than 1/10 of an inch (.0856") higher than running with the splitter and wing alone. If my memory is correct, adding the front skirt alone in a previous test produced negligible changes in ride height. It looks like adding the diffuser alone has a similarly small effect. But both of them together just happened to produce a pretty big result.

That's my preliminary finding, at least. My math could be bad again, or my data could be too slender at this point (although I did a total of four runs with the diffuser and curtain attached, and six runs with the diffuser alone).

Next up is a track test, although the front curtain is going to probably have to be trimmed back to prevent contact when the car is diving under braking. Here's a photo from last week's track day with the splitter alone (no curtain). You can see that under braking, the nose is already pretty close to the asphalt.

http://forums.pelicanparts.com/uploa...1161555581.jpg

bump...whats new here??

Wow, Great work Jack. I'd been so busy with work I missed this thread completely. A couple of observations...

1) Back in the "Classic" Can Am days, both Lola and McLaren used wings mounted over the front deck or air-dam. Notice that no-one uses them anymore? I suspect this is because the wing is trying to create a low pressure area on it's lower surface, exactly where the front deck has a high pressure area, so neither is working to full effectiveness. I believe that Porsche was getting better downforce with their 917/10 and 917/30 designs which used a larger spoiler in front of a blunt nose on the car. Basically, the blunt nose created a high pressure area that pushed down on the large splitter surface.

2) Flip-up's on the outside corners may add downforce, but I don't think that they work out to be very effective, as in they increase drag significantly also. This is why they are generally an aerodynamic feature-of-last-resort, which designers would do without if they could.

3) I'd agree that trying to improve the front downforce will allow you to improve your aero-balance, rather then running the rear less effective.

4) Some ideas to try for increasing front downforce...
http://forums.pelicanparts.com/uploa...1167834607.jpg

A) Try to use the upper surfaces more efficiently, and eliminate lift (low pressure area) in those areas. A common "trick" is to vent air from a high-pressure area under the car to a low-pressure area above the car. In the case of the Audi, they've vented air from the area above the tire where the air "packs up" onto the upper surface of the fendor, which due to it's curvature acts like the top surface of a wing and generates lift. 935K3's used a similar arrangement. "Gills" like these are still common on LMP cars to this day, so I suspect that they're fairly effective. I would guess that directly above and behind the front headlights would be an area of lower pressure on your car.

B) Try to use the air that travels under the car more effectively. In the case of the DTM Audi, what you don't see is a large underwing behind the front spoiler. Picture something like the curved surface on an upside down wing. The trick is with this is to get it as close to the ground as you can without stalling it, since the downforce created increases significantly with proximity to the road surface up to the point when it stalls. Which is why one of the biggest aero-problems that designers used to have was "porposing" where the front end would get sucked down to the point of stalling, loose all downforce (ie. grip), and rise up before getting sucked down again. Needless to say, driver's were not terribly happy with the phenomenon. The solution was to run super stiff springs (see my Colin Chapman quote below).

C) If you are running air under the car through an under-wing to generate downforce, you need to vent it out, which usually done just behind the front wheels. You may not be able to run the huge cut-outs like the Audi does, but once again some "gills" in a lower pressure area on the side of the car may go a long way to reducing any pressure build-up under the front of the car.

Great work!