Real-world testing
 

Enough of the CFD and armchair calculations, this summer I'm going to do some real-world testing. I could use some help planning what things to tests, and how to test them. I also need a second driver, provided you regularly track a Miata (why else would you be here), and know Watkins Glen like the back of your hand.

I belong to the Pineview Run country club which has a tight and technical 1.1 mile track, with about 50 mph average speed. I can go there pretty much whenever, and so I'll do a lot of initial low-speed testing there this Spring. I'm also 25 miles from Watkins Glen, which is an ideal high-speed venue to test aero. I booked a two-day open track (no run groups) event at WGI in early June. I may also do coast-down tests, as I live in a fairly rural area and nobody will care.

What do you want to know? Do you want to be involved, and in what capacity?

Might be useful to mention what you are looking to test.

Blackbird, obviously he is testing the real world. It is in his title for God's sake.

Poscat for a good laugh :rofl:

This is very open ended. Are you an expert driver giving your services of consistent driving? Do you have an aero package that you think will be competitive for Miatas? Do you have a data acquisition system and the necessary sensors to complete the testing? Are you looking for people to send you their aero package to test on your car? I would spark some discussion with more info on who you are, what you have, what you're trying to achieve, and take a stab at how you're going to achieve it.

Can you plz test miata with constants of 150-170 rwhp, air-dam, 2d wing, changing up WITH 3" splitter vs WITHOUT 3" splitter?

THEN test constants of 140rwhp miata air-dam, changing up BBFW spoiler vs 2d wing?

Lap-times | min speed per corner | max speed per corner | G-force per corner

Ok thanks, much appreciated.

Let me start over. In 1993 the Mazda RX7 had a Cd of .29; the Miata of the same year had a Cd of .38. What would a lap time or top speed look like if a Miata was shaped like an RX7?

So I got to thinking and looking at online Cd calculators and such, and it's a lot of fun, but in the real world, what does it mean to reduce drag by 20%? I do endurance racing, and a 20% decrease in Cd is around an 8% increase in fuel economy. That can be the difference between running an extra stint, and very worthwhile. But what does that drag reduction do for lap times? Or what if you take that drag reduction and put it into downforce, then what?

So let's take a Miata of average weight and power (2300 lbs and 120 hp lets say) and start putting different aero bits on it, see what happens. Fastback. Airdam. Spoiler. Side skirts. Flat bottom. Diffuser. Etc. What do each of these things do individually? What do they do together? What makes sense? If you're running a class like ST6, do yo want to run BTM aero and have another 6 hp, or use an airdam and fastback? Is a wing worth the full point? Or let's say in an endurance series where doing a 2-hour stint (or 2:40 in lemons) may be a big deal, what kind of things will reduce drag?

Some configurations that intrigue me:

1) Cheap stuff you can buy on Amazon and eBay - Does a hardtop spoiler do anything? What about those cheap plastic side skirts? Is a small spoiler worth it? Cheap wings, good or bad? We've all seen those cars pimped out with mostly Chinese parts, and you have to wonder if there's any function to it.

2) Spec Miata plus - What if you add a Supermiata style airdam and spoiler to an otherwise Spec Miata? OK, now add a splitter. And then a wing. Now remove the top. What happened?

3) Tops - I build boats for fun, and hardtops are way easier. I've built a fastback, shooting brake and 70s Supercar style top. How do they compare to a stock hard top? Or a Treasure Coast Chop top? Honestly, I could just test the tops and be done with this, but I have potentially 14 hours of track time to test things at the Glen, so why not do more?

4) Underbody - What happens when you add side skirts, a flat bottom, and a diffuser, one at a time, or in different combinations?

5) Open windows - What is the effect of running with open windows, and what can you do to improve airflow?

The list goes on. Not all configurations are worth testing, nor will I have the time or the inclination to test them all, but I have time to do more than just different shaped roofs, and I'm interested to know what are the most compelling things people would like to see tested. I have a couple good Miatas, and I wouldn't be against using someone else's but I need to fabricate brackets and such so that the different tops and front ends can all go on and off easily. So it's easier to do that in my garage on my cars. I have a 93 street car that is more or less a STS build, and a 94 that was a Spec Miata but has been modified well beyond that, with a fastback, wing, airdam, splitter, etc. Both have Megasquirts and could run different maps to simulate different HP, but I'd rather just stick to one each.

Telemetry-wise I have an AIM Solo, but if that's not enough, maybe someone will loan something better. The same goes for parts. I either have or can make most things, but if there's a part someone wants tested, and I can easily put it on, then why not? I'm in this for the info. I want to know. I'm sure there are things you want to know.

I'm looking for like-minded individuals who will spend the time and effort. I'm asking on MT because there's a high level of no-bullshit-allowed here. If you're questioning my ability to drive consistently, well that's why I'm looking for a second driver. I don't even need to drive at all, I'm the parts building and swapping guy. I want this to be a real investigation into Miata aero, and something that will serve as a reference well into the future.

Do you want to help?

So you dont really have anything for this test, just a real cool idea?

I don't have anything to test? I have two Miatas with six top options (open, hardtop, Treasure Coast chop top, fastback, shooting brake, RF-style), a cheap wing (hoping to get more wings, I'll make a spoiler), side skirts (with and without wheel spats, will make proper barge boards), four front end options (stock, r-package lip, Supermiata style airdam, and a 4" splitter extension for it), and a diffuser. I'd like to make a flat floor. I also have time, money, fabrication skills, 10 track days paid for at a short track, and two days paid for at Watkins Glen (14 hours open track, not a typical HPDE day).

I feel like I can get most of the important data from the AIM Solo, but I'm open to adding more sensors for data acquisition.

At some point I'll meet one of you who's interested in this test, and we can talk about what and how to test these things in a controlled manner.

There is going to be way too much noise in your lap times to be able to see much more than a massive change. I think you would get much more useful data using linear travel sensors on the suspension and testing on a smooth flat road to see how much your aero is pushing down on each end of the car. That combined with coast down runs to figure out drag. Pick the highest DF , lowest drag setup that fits your rule set and profit.

Well, Hopefully a little encouragement goes a long way. This is a really cool idea and I applaud the concept and your thoughts/efforts to try.

Personally I would try black and white changes. Coast down and lap times of big differnce items.

WIndows open or shut is a basic one im interested in, we all think we know but (to the best of my knowledge) no one has qualified this.

How much drag does a 3- 4" splitter create in coast down?

I'm not sure roof on or off is pertinent, almost all accept a roof is better than not and race with a roof anyway..?

Maybe roof versus fastback?



Also, 949 style air dam is known to work very well so im not sure there is any point in learning by how much or why....? although with and without splitter would be interesting.

If you believe the CFD work done by the Hancha group, an airdam reduces drag from .36 to .31, and a 3-4" splitter reduces drag further to .29. I don't really trust CFD tests, and that's why I'm doing this. It will be interesting to see if the splitter adds or reduces drag, and that should be apparent in coast-down tests.

There really isn't much reason to test an open top, but I'm interested to see if the HP Wizard drag calculator is accurate for that value.

For sure I'm testing an open top, hard top, Treasure Coast chop top, fastback, shooting brake, and what I call a 70s Supercar top, but is maybe more like a longer RF top. There's 18" of snow between me and the barn, but the tops are in there awaiting final fiberglassing and window cutouts. I'll post some concept drawings and construction photos.

The Treasure Coast chop top is a cheap way to run a lightweight hard top. Some people run a hard top without a rear window, and this top may simulate that situation well enough. This photo is from Treasure Coast Miata. I ran this top at Thompson, but without data, I couldn't tell if it was better or worse than a standard hard top.

https://cimg9.ibsrv.net/gimg/www.mia...1e3181d815.jpg

This is the fastback I built and raced with last year. Again, no comparative data from the same day. It felt faster. It looks faster! Is it? Who knows?
https://cimg3.ibsrv.net/gimg/www.mia...0a10cf7123.jpg

This is the next top I conceived of over winter break, I like Shooting Brakes.
https://cimg4.ibsrv.net/gimg/www.mia...dbcb3bdfb0.jpg

This is an early photo, it's pretty much complete now, but I couldn't figure out how I wanted to hinge the rear hatch. I got so tired of scratching my head, I started on another top instead.
https://cimg5.ibsrv.net/gimg/www.mia...95cacb2f8f.jpg

I call this a 70s Supercar top, because it reminds me of a 512, M1, Pantera, or other 70s car. Also kind of a ND RF, but longer.
https://cimg7.ibsrv.net/gimg/www.mia...3198307159.jpg

It was pretty easy to make using a single sheet of plywood with cuts so that I could bend it. The Shooting Brake is strip built out of individual strips, but slicing up a piece of plywood like this on a table saw is easier. The tops are surprisingly light, being more or less a surfboard style construction, with fiberglass skins over a lightweight core. The gaps are filled with a mixture of epoxy and microballoons. The neat thing about this top is I'll be able to add rear glass and change the height at the rear, so I can effectively test a fastback with different backlight angles.
https://cimg6.ibsrv.net/gimg/www.mia...3696a7bfb7.jpg

The next two tops I build will be for 24 hours of Lemons, one is a RX500 (flat top, almost wagon looking), and the other will have a central fin, like a Le Mans Prototype (calling it the LeMons Prototype, of course).

Love the first ply picture, very lotus europa style....

Great idea, thanks for attempting this.
For track testing without $$ in sensors/Senna, I'd be interested in only top speed at predetermined point before braking zone. Racechrono with 10Hz booster can easily document this accurately.
-longest straight/ consistent entry speed/pre-determined full acceleration point / end point pre determined and still @full accel.

-hardtop rear window delete. (windows down)
-hardtop vs fastback

I like where you are going. I'd like to see only windows down (what tracks allow windows up?!?). Give us the gold of the testing steps because I myself have considered a million different ways to remove as many variables as possible; but know in my heart I will never be consistent enough to judge laptimes this way. Love the hard top designs.

I've looked up linear travel sensors, and there are quite a few available. Has anyone here put one on their car? I suppose I need one each front and back, and some kind of gauge with a max level readout. Is there a plug-and-play solution for this?

Re: noise in my lap times, yes. That's why I'm looking for a Miata driver who knows WGI like the back of their hand. I could put my brother Ian or teammate Evan in the car and they'll be more consistent than me, but I'm looking for someone who is well acquainted with tuning downforce and can instantly feel setup changes and provide feedback. This is not a paying gig, but it is 3 days of free track time.

I'll have a RumbleStrip predictive lap timer in the car, and this will help keep a consistent speed at the exit of Turn 1, which determines the speed on the back straight (in a Miata, anyway), and can easily be over 110 mph (even in a Miata). The Aim Solo can get top speed on the bake straight, as well as G-forces in corners, lap times, etc. By overlaying data of multiple runs, large changes should be obvious, small changes probably not so much. I don't think I can test hardtop windows up at the race track, I'll probably be black flagged if I do that. But I'll coast-down test that one for sure. Hardtop window delete and hardtop vs fastback will be tested, as well as the other tops.

This would need to be logged, if you have any spare analog inputs in Megasquirt that would work. A max level readout won't work because a single bump in the road will skew the results. You'd want to drive at multiple steady state speeds, and average wheel travel during those speeds. Using suspension travel, spring rate, and motion ratio, you can calculate downforce. Then you can generate a downforce versus speed curve.

I found this sensor and gauge, which pair together, but you're saying it would need to log the data.
Shock Absorber Suspension Travel Sensor with Rod End Joints
Mini Digital Display Gauge for Shock Absorber Suspension Travel Sensor

I wonder if there's a DIY solution that is more like putting a zip-tie on a fork tube kind of thing? Although Turn 7 may be where the most compression occurs, and that wouldn't be measuring what I want to measure, which is the downforce due to speed alone. If I can get a downforce measurement on the back straight at WGI, that would suffice.

In response to the digital gauge, you'll find that the data is all over the place. It's not like you'll be able to look over as soon as you hit 100 mph and be able to say, "my suspension is compressed 28.5mm". Instead you'll be able to kind of read the gauge and say "I think it's bouncing between 26 and 30mm, so maybe it's like 28mm?". This is why it needs logged. Every pebble and bump on the road will skew the results, so being able to filter and average the data is necessary. Also it should be noted that for what you're trying to do, and the level of DAQ/analysis you have available, this testing would not be done on track, or at least at a typical track pace. It needs to be steady state, so hold at 60 mph for 5 seconds while recording data, hold at 70 mph, etc, etc. If you are driving for lap times, you are always on the gas or brake, both of which skew suspension travel measurements (unless you also have a way to cancel those forces out of the measurement).

In response to zip-tie on the shock shaft. This wouldn't work because you could never measure the travel you want. Imagine you have this zip-tie on the front shock, you're going 100 mph on an infinitely long glass bed and your suspension compresses 28mm due to downforce. If you ever go less than 100 mph after this, you've now decelerated and the body pitches forward, you've now moved the zip-tie more than 28mm. Then you're not on a glass bed and you hit a pebble, again the measurement is skewed.

The linear travel sensor at steady state will get you approximate downforce numbers. This is useful if you are trying to correlate CFD or input into a lap time simulator. If you have a good amount of test cases, you don't care about downforce numbers, you care about the result of the whole vehicle. Therefore, if you can get a consistent driver and record max speed at the end of the straight, average steady state cornering acceleration over the most consistent corner, and lap time, you'd get a real good idea of how the different setups are performing in relation to each other. Does the AIM solo allow you to plot a map with speed along the track map? That would be useful to overlay all the setups.

My Aim dash records G's for corners. presumably other race dashes do the same thing.

I wonder if averaging laps with one mod versus another to get a gain or decrease in the g reading is enough?

Aero is mostly aimed at achieving higher downforce and thus higher grip in corner, thus higher g. Isn't that closer to the end result and outcome required anyway?

yeah, windows up not feasible at every track. looking forward to the info.

I made my own ride height sensors with a $5 potentiometer and a 3D printer. They work well if you have something that can log 2 x 5v analog channels. The resolution is probably enough to gauge downforce alterations if you are running softish suspension. In theory you can figure out drag by how much the front end lifts vs the rear, but that really only works if you are adding downforce at each end and not the middle of the car.

I ended up not running them because I found my lap times were fairly consistent and I could evaluate changes easy enough. But then I haven't done much with aero yet.

https://cimg4.ibsrv.net/gimg/www.mia...98fd7b5e9d.jpghttps://cimg1.ibsrv.net/gimg/www.mia...a5be49bb80.jpg

I have some professional experience in vehicle (specifically tire) testing.
AIM solo is good enough for lap times.
Driver mod will be the hardest part for lap testing. You need a 5 lap standard deviation of < 0.25s (including tire evolution) to be meaningful.
A great way to test drag are coast down tests. Using the same stretch of pavement, coast from high speed to low speed in neutral (80mph -> 20mph) and measure the distance using the AIM Solo (we used a vbox). Shorter distance = higher drag. Need to have a weather station set up to make sure wind isn't too high (5m/s was max for us) but do it on a calm day and you'll be fine.

If you bring it down to VIR or Road Atlanta I could drive for you ;)

Is distance more important to measure or time? I found a neat coast-down calculator on Grassroots Motorsports, and their spreadsheet uses start speed, end speed, and time. But not distance. It measures drag in pounds, and I'd also like to be able to convert that to Cd, if someone knows how.

I'd love to take you up on VIR, it's so on my bucket list. But the thing about WGI is that 1) it's in my back yard, and 2) I have 14 hours of open track time (no run groups).

The driver I'm looking for can easily do 5 laps within .25 seconds. I'm not a great driver as evidenced by these overlays, but these three consecutive laps are all within .1 second. (This was from a tire test my brother and I did at Thunderhill). My brother kind of pokes fun at me for not being consistent enough, and so I'm looking for someone who can do five laps and make this chart look like one line.

https://yousuckatracing.files.wordpr...018/10/rs4.png

Warning! Short story ahead!


I am in the process of developing a setup for aero testing on track. Uses a pitot tube (dynamic pressure with yaw sensing), sensors for relative humidity, ambient temperature, barometric pressure, strain gauges bonded to camber plates, strain gauge temperature, strain gauge amplifiers, and analog to can bus converter. I am calibrating the microstrain output of strain gauge in a load cell then correcting on the car with offset coefficients to correlate for corner weights using some freshly calibrated corner scales. There are many ways to skin the cat but this is my approach for what might be considered steady state aero tests. I correlate corner loads to dynamic pressure instead of solely GPS based or wheel speed to eliminate some uncertainty in testing, (head wind, tail wind, crosswinds). For total drag force, I have been using the primitive approach of taking average longitudinal g force during deacceleration while coasting down correlated to dynamic pressure and the taking into account the mass of the vehicle. I then subtract low speed rolling resistance force found in low-speed coast down test with some components of wheels, brakes, and rotating component inertia are taken into account. I import the frontal image of the vehicle into Solidworks with a meter stick in the image which allows me to find the frontal cross-section area of the vehicle to be used with aero loads and drag to generate Coefficient of lift and drag values. The plan is to plug in test results from aero setup changes into Lapsim or chassis sim, or a similar lap simulation tool for various tracks and configurations and evaluate which setup is the best compromise for the track day on hand.

(Right) Pitot tube mounted using Go Pro Camera Suction Cup (hasn't fallen off yet, key word being "yet")
(Left) PCB board with environmental sensors and CAN-Bus Converter
(2nd image) Omega strain gauges

https://cimg5.ibsrv.net/gimg/www.mia...9cadd49ae.jpeg
https://cimg9.ibsrv.net/gimg/www.mia...729da95fd.jpeg

.

PM sent

We tested four different tops, two wings, no wing, splitter, vortex generators, and combinations of them. It's a lot of information that would be hard to sum up in a forum post, and so it's all on my website. https://occamsracers.com/

None of this would have been possible without jspeed.713, thanks buddy.

Great read. We run a Champ Car with a roof panel, but we've been debating building a fastback. This is what we needed to reignite that fire.

Wow. Very impressive work here.

Really enjoyed reading that. Thank you for all that effort!

It seems like a 9LR wing might be on my shopping list....you have me worried about drag with APR GTC200.

Something to consider is the Lift/Drag efficiency of the wing in question. Drag is not a bad thing, as long as you are generating downforce at a much higher rate it can often pay off. Assuming the drag and lift impacts on pace and fuel consumption are attractive or acceptable, then the primary concern might lie in the balance of the vehicle. Does the ruleset and car classification allow for the front of the vehicle to make sufficient downforce and establish a driveable car in regards to balance? I have seen on a few occasions people compensate for the imbalance experienced in higher speed aero influence conditions by using mechanical adjustments to sway bars, springs, suspension kinematics etc. Relying on heavy amounts of mechanical balance correction often made the car a handful to drive at lower speed corners.

Looks like I missed that in my reading. What was your concern with the APR GTC200 wing?

Apparently very draggy. My lap times haven't improved as I expected with extra power so I wonder if the higher drag is giving me front lift as mentioned in that article.

I got to thinking and it seems like I "may" be subconsciousley compensating for a bit of understeer. just a wild ass theory, could be far from the truth BUT worth playing with.....

So further to my comment about buying a 9LR wing, I realised a few things I can test first.

- Will try running it lower (without the lifters).
- Will try without Gurney flap.
- Run it flat or even negative to see what that does
- Go back to the original endplates.

If it is related to the wing, the new fast back is likely to make it worse so just trying to plan ahead....

I've always recommended the APR GTC-200 on S2000, owned a track S2000 with wing on Circuit of the America's and Harris Hill Raceway. I never got a back to back test at COTA, but Harris Hill I was able to pick up 5mph on a offcamber right hand turn (85-90mph). Same day removed APR wing and could not maintain grip on same turn. Being their both pretty similar platforms I always liked the concept of mitigating drag on center of wing due to aero affects of the chassis. To note, I never tested, but always was suspicious that risers did harm.

i really don't recommend running it negative... if, for whatever reason, you get turned around at speed bad things can happen