Blog

In a previous post I put various wings and spoilers on my Veloster N and tested them at Pineview Run. What I found was that a ducktail spoiler or single wing was about .7 seconds faster than the OEM N spoiler/wing. And then I tried a dual element wing, which was another .6 seconds faster.

Pineview is a small, slow track, and there are only a couple corners where aero matters, and so 1.3 seconds is a significant difference. I opined that on a longer track with faster corner speeds, the lap time delta might double.

I got a chance to put my money where my mouth is at private open track day at NYST. The track closed for the season a couple weeks prior, but somehow Brad at S2K Takeover got his group a bonus round, and they invited me along. It was cold and damp at the start of the day, and this affected things slightly, but I got in a lot of consistent runs as well.

Rear aero options

For this event, I didn’t have much front aero on my Veloster N, just a hood vent and an aluminum undertray. Those are mostly for cooling, but they might provide around 60 lbs of front downforce at 100 mph. I have splitters and canards aplenty, and can get five times the downforce with those, but I wanted to test the exact same car I was using at Pineview last time.

Well, not exactly the same, because I didn’t bring the ducktail spoiler. At Pineview, the single wing was about the same performance, and so I left the ducktail at home. This would allow me to make faster setup changes, as I could keep the same roof extension on for every configuration.

If you are a regular reader, you know I’m a shit disturber; I like to challenge the status quo, and disprove common misconceptions using data. That’s another reason there’s no front aero on the car. Adding rear aero alone on a FWD car throws the aerodynamic balance out the window, and is the opposite of what most aerodynamic pundits suggest.

The rule of thumb is to match your aero balance to the chassis balance, so that you don’t change the balance of grip as the car changes speeds. And so if your car has a 50/50 weight distribution, then you add the same amount of downforce front and rear. Or in the case of this FWD hatchback, if you have 67% of the weight on the front, then you should add twice as much downforce on the front as on the rear. But I purposefully didn’t do that; I only added downforce to the ass end.

I tested seven rear aero options:

1. No aero – The base model Veloster didn’t come with a wing or spoiler, just a roofline extension. I built my wing mounts on top of this, and so the no-wing setup has two vertical wing supports. This might provide some stability in yaw, but they wouldn’t add downforce or affect the lap times much.
2. Single wing, 0 degrees – This is my DIY S1223 wing, which I recently modified with glassed-in endplates. This is also the lower wing for all of the dual element wing tests.
3. Dual wing 0/0 – This setting adds an upper wing element at zero degrees. This represents a dual wing at the lowest drag setting.
4. Dual wing 0/26, standard gap – I then set the upper wing to 26 degrees. This is a fairly conservative angle of attack for a dual element wing.
5. Dual wing 0/26, small gap – I kept the angle the same, but moved the upper wing down (closer to the main wing) and forward, so there was more overlap between the wings. This is the only wing setting where I used this gap and overlap.
6. Dual wing 0/40 – I put the gap and overlap into the original position, and then raked the upper wing to 40 degrees, which is typically the maximum angle you can use with a dual element. At this AOA I’m flirting with flow separation, but what can I say, I’m a flirt.
7. Dual wing 0/40 Gurney – I added a 1/4” Gurney flap. At high angles of attack, a Gurney flap can often help keep air attached. This setting should have the highest downforce, but also the most drag.

Let’s see how these compared in lap times, from slowest to fastest.

No wing 1:41.546

I tested the wingless car on the last run of the day. The track was in perfect condition by this time, now completely dry and warmed by the sun. I was also driving at my best, with enough laps under my belt that I wasn’t losing time anywhere for silly mistakes or a lack of reference points. And yet I set my slowest lap time.

The problem was mostly that I’d gotten used to rear aero, and without it, the car felt sketchy. On my best lap, the rear end got scary loose while braking on the front straight. And then I almost shit my pants braking into T5, as the car skidded in a way I hadn’t experienced before. I intentionally drive a tenth under the limit at this track, and I honestly thought I was about to put it in the grass.

That nervousness braking into downhill corners may be partially down to the Hyundai’s dynamic brake bias, which senses wheel speeds on each corner and then loads more and more front bias as the rears sense lock. This can put up to 93% of the bias on the front tires, making the rear end rather irrelevant. Without rear aero, that balance shifts earlier.

So let’s move on from braking zones to corners, and in the slower ones, the non-aero car handled the best and was the most fun to drive. Rather than forcing the car into rotation early in the corner, I could just lift a little and get a mid-corner pivot.

From start/finish to T5, the cars were very evenly matched, but from the downhill braking zone in T5 to T11, the car was 1.5 seconds faster with a single wing than without. Sans wing, I had less confidence in turns 8 through 11. From T12 to the end of the lap, the cars were again very evenly matched.

If you look at the speed trace, you’ll see that right around 52-54 mph is where the difference is; go faster than that, and you want a wing, go slower, and you don’t. Ergo, using rear aero alone on an autocross course could be useless, but on a race track, you definitely want it.

Double wing 0/0 1:40.370

Most people wouldn’t set their dual wing to zero degrees top and bottom, it doesn’t take advantage of the convergent gap, which allows more secondary wing angle. Where you see this setting is in Formula 1, when the upper wing goes into DRS mode.

And that’s exactly why I wanted to test this setting. Next year I plan to build a DRS upper wing, and I wanted to see what the speed advantage was on the front straight. So here’s how that shakes out.

The DRS 0/0 double wing had an average top speed of 107.4 mph, while the 0/26 setting had an average of 106.8 mph, and the 0/40 had an average top speed of 106.4 mph. So DRS was worth about 1 mph from the highest to lowest drag settings. This doesn’t take into account that the higher downforce wings would have a higher min speed through T18, so let’s call DRS a 2 mph advantage on this track.

Surprisingly, that’s not worth a lot in a lap time, as the reduction in drag isn’t noticeable until about 88 mph, which happens about half way up the front straight on the 3rd to 4th gear shift. From there until the braking zone is where DRS makes a difference. Still, DRS sounds like a fun project, and on a track like Watkins Glen or VIR, the drag reduction would be more worthwhile.

Single wing 1:40.081

The single wing is the bottom wing for all the dual wing setups. It’s my DIY S1223 with a 1/2” Gurney flap. In the wind tunnel, it was nearly the same downforce as a 55” 9 Lives Racing wing (185.4 lbs) and 55” PCI wing (182.5 lbs).

With the wing on, the car was much more planted in the braking zones and in the fast esses. The downside is that is the wing took away some of the fun in the low and medium speed corners. For the gain in lap time, and especially the braking stability, I’ll take that trade.

If I look at the theoretical lap times, there’s a full 2 seconds between the single wing and no wing. And this is without changing wing angle at all. If I had added more wing angle, the car would have turned an even faster lap time.

But let’s get back to the real data, and that was that the single wing at zero degrees was 1.5 seconds faster than no wing. This is just over double the delta I saw at Pineview, and so I guess I was right about that prediction. And with that, I’ll make another prediction, which is that this wing would be worth 3 seconds at Watkins Glen.

Double wing 0/40 Gurney flap 1:39.970

You may recall that the lower wing has a Gurney flap on it already, and for this run I put a 1/4″ Gurney flap on the upper wing as well. The upper wing measures 4.7″ chord, and so the wicker is about 5% of the chord, which is a typical size.

This configuration drew the short straw, because the track was cold and damp when I tested it. Top speed was down a hair from the other 0/40 test runs, but not as much as I thought it would be.

I’d like to re-test this one at Pineview Run, as the extra downforce might have some benefit in slower speed corners. But given that the 0/26 setting was faster, I don’t think there’s a reason to use double Gurney flaps and a 40 degree angle on this car, unless I add front aero.

Double wing 0/40 1:39.447 and 1:39.435

I did several laps with the Gurney flap on and then quickly pitted, ripped it off (it was duct taped on, just like I would in a wind tunnel), and went right back out on track again. Although the track was still damp, the conditions wouldn’t have changed much in those 60 seconds, and so I feel the comparative data here is really good between with and without the Gurney flap on the upper wing. The data shows that there’s about a half second between them, and no Gurney on the upper wing was faster .

On the second to last run of the day, just before testing the single wing, I retested the 0/40 dual wing, because I didn’t feel I gave it a fair shake on a partially damp track. The best lap was only .012 seconds faster, and so I think there wasn’t that much time deficit in the damp track after all.

Double wing 0/26 small gap 1:39.230

I wanted to see what would happen if I changed the size of the gap between the wings, and the amount the top wing overlaps the bottom wing. It would be better to run a sweep of offsets in a wind tunnel, but I figured what the heck, I’ll make one radical change and maybe I’ll notice a clear signal in the deltas. I did not.

The smaller gap had a slightly faster top speed on the front and back straight, but the differences were so minor I can’t attribute it to drag reduction. But the minimum corner speeds were all slightly less than with the larger gap, and maybe this is a lower drag, lower downforce arrangement after all. I can’t say for sure, and with lap times this close, kinda don’t need to retest this one.

Subjectively, I can’t say I felt huge difference between any of the 0/26 settings and 0/40 settings. They all made the car a little more stable, and required a little more effort to turn the car. But if you changed wing angle or Gurney on me without knowing, and sent me out on track, I wouldn’t know which I was using.

Double wing 0/26 1:39.104

The fastest laps were set with the top wing at 26 degrees with the standard upper wing gap. These laps are shown here in blue vs the 0/40 runs in red.

A – You can see the 26-degree angle has less drag and a higher top speed than the 40-degree.

B – The only place the 40-degree angle has a clear advantage is in the carousel before wheelie hill. More downforce means a higher minimum speed all the way around.

C – For whatever reason, the lower wing angle carried a higher speed through the esses.

Other than those areas, there wasn’t a lot to choose between them. Given that, there’s enough evidence to say that somewhere around 26 degrees is the best setting. Certainly the 40 degree angle was slower, as was zero degrees.

It’s funny, that when I originally built this wing I set the upper wing to 30 degrees and made no provision for adjustment. I was simply testing a proof of concept at that time, but damnit that might actually have been the best setting.

Conclusions

Here are all the fastest laps and theoretical laps for each configuration.

In the previous article, I forecasted that the single wing’s .7 second advantage and the dual wing’s 1.3 second advantage at Pineview might be worth double at a longer track, and it was almost exactly that. Like Pineview, NYST had only has a few corners where aero matters, and on a track with more fast sweepers, the advantage might double again

So I would also guess that the single wing’s 1.5 second advantage at NYST might be 3 seconds at Watkins Glen. However, I wouldn’t extrapolate this and say that the dual element wing would go 5 seconds faster at Watkins Glen, because the double wing is pretty draggy.

A single wing is more efficient, and a longer wing span might give similar downforce as the double wing at 0/26, while also having less drag. A single wing is also a lot easier to optimize, with just the angle of attack, height, and setback distance to monkey with. The dual wing has twice as many variables, and hitting the sweet spot may be down to luck. I’ve started building a larger MSHD, and I’ll test that next year, along with a new DRS wing.

Speaking of DRS, it seems like I’d want to have three settings:

1. Low speed – In slow corners, the car doesn’t benefit from rear aero, and it’s more fun to drive without it. I’d activate DRS at 50 mph and under.
2. Medium speed – For corners above 50 mph, I’d put the wing into full downforce mode.
3. DRS – On any significant straight, I’d go back to the low-drag setting.

Another discovery from this test is that I’ve found the tipping point where adding rear aero stops making the car go faster. This will help me figure out how much rear downforce I need when I put my splitter and canards back on. Those items are worth about 300 lbs of front downforce at 100 mph, so in order to achieve the same balance I found in this test, I’ll need to add at least that much more downforce in the back. Here’s why.

My car weighs 3200 lbs with me in it, and because I’ve removed some weight from the rear, the weight balance is about 67/33. So that’s 2144 lbs on the front tires and 1056 on the rear. The typical aero rule of thumb is that if I add 300 lbs of aerodynamic downforce, then I should split that 200 front and 100 rear. This retains the same chassis balance and mechanical grip that the engineers designed.

However, the fastest configuration I tested was with 60 lbs of downforce on the front (the hood vents and undertray) and probably 240 lbs on the rear. This adds up to 2204 lbs on the front and 1296 lbs on the rear, which works out to a 63/37 balance. Mathematically, I’ve moved the grip 4% rearwards at speed, using aerodynamic downforce. When I add my splitter and canards, I’ll want to be in the same range.

Now of course none of this make logical sense. When you put rear downforce on a FWD car, it takes away from the front tires, which is a problem because they are doing all of the steering and accelerating. But logic be damned, the reality is that I went 2.5 seconds faster with way too much rear aero bias.

If I did this same experiment on a RWD car, the results would likely be even more dramatic, because the rear wheels are responsible for acceleration and have more weight on them for braking. This supports the case for having heavily rear biased aero on a RWD car as well.

When you load a car with too much rear aero, it understeers in fast corners, but also makes for a safer and more stable car that requires fewer corrections when you lose control. The downside is the car will be boring to drive and harder to turn. But that’s what the brakes are for.

If you don’t care about lap times and want a car that is fun to drive, set your aero balance however you want. But if you care about lap times, ignore the rule of thumb and see what happens when you keep adding more and more rear downforce.

One day I’ll have enough information to write an in-depth article on aerodynamic balance, but until then, chew on this nugget: if you like the way your car handles, you don’t have enough rear downforce.

When I built my DIY S1223 wing, I chose a wing span of 53.15” and a chord of 11”. The chord was chosen just so I could say “this one goes to eleven,” but the reason I chose such a short wingspan was so that I could later make it into a dual-element wing. You see, you can buy a 135cm (53.15”) wing on Amazon or eBay for $60 or less, and I had that in mind from the beginning.

The other advantage of a short wing is not bumping into it. I have a decent sized “race barn”, but the garage door isn’t very wide and you’d be surprised how often a wing is just a pain in the ass. Or head.

The short wingspan and large chord combine for a low aspect ratio, and that’s not a recipe for performance. High aspect ratio wings are generally better because they suffer less from the detrimental effects of wing-tip vortices. Just the same, I had good reasons to build it this size, and so let’s get onto the dual wing aspect.

Dual wing end plates

To mount the upper wing I made new larger end plates and suspended the upper wing between them. In the past I always made end plates out of street signs, because I can get them for $1 per pound at my metal recycler. But for this project I decided to use marine plywood, which is considerably stronger at the same weight, and the additional thickness should allow me to reduce drag. Say what?

You might think a thicker end plate is more drag, but because I’ve shaped them as an airfoil, it should be less. That is, the leading edge is rounded and the trailing edge tapered to a sharp edge.

Rounding the leading edge it’s important, because it keeps air attached. For example, when the car is sliding through a turn, the car is in yaw, and airflow hits the end plate an angle. On an aluminum end plate, you might get flow separation on that sharp edge. However on a thicker end plate with a rounded nose, airflow should stay attached for longer.

I also tapered the trailing edge of the end plate on the inside. This, combined with the rounded leading edge, forms a curve like an airfoil. This should accelerate air under the wing, which drops the pressure further and creates more suction.

For initial testing, I set the upper wing at a fixed 30 degrees. I honestly didn’t know if this was going to work or not; I suspected that the extra downforce could make things worse, and I was sure there’d be more drag. So I wasn’t going to put effort into making the upper wing adjustable for gap, overlap, and angle, until I’d tested it.

Now if you’ve been reading my other articles, you know that it’s important to set the gap and overlap correctly, especially since the bottom wing has a Gurney flap on it. Putting a wicker on the bottom wing can increase downforce and L/D ratio, but the gap can be finicky. To avoid that I eyeballed a large gap and set a conservative angle of attack.

When all of the parts were completed, I weighed the lot: The main wing weighs 6.6 lbs including the bottom mounts, end plates, and Gurney flap; The upper aluminum wing is 4.6 lbs; The roof extension with wing mounts weighs 7.2 lbs. When I put it all together, that’s 18.4 lbs. Total cost was maybe $150.

For comparison, a 55” 9 Lives Racing wing with The Deuce kit, plus universal hatchback wing mounts would cost $1800 and weigh north of 30 lbs. And so I’ve saved a lot of money and weight. But I also did a shit ton of labor, and I can tell you right away it’s cheaper to buy one than to make one.

Note that the OEM spoiler/wing on the Veloster N weighs a bit over 11 lbs. When you consider my double wing is only 7 lbs more than that, it’s pretty damn light.

Testing

I tested the dual wing at Pineview Run, with the long track extension. This isn’t a fast track, and there are really only a couple turns where aero could potentially help out, and one straight where drag could come into play. I also tested other rear aero options, including the OEM wing, a single wing, and a ducktail spoiler.

You can read the results of that test if you have the password to the locked articles. If you don’t, you can get the password to it and all the other articles by buying me a coffee. Sorry about that, but this wing took a lot of effort to build, plus I had to get back to back data, analyze it, etc. All of it is a lot of time and expense, and I’m not going to give that away for free.

Version 2

After initial testing I decided to improve the wing further so that the upper wing can adjust for angle and gap height. Similarly to how I did it on the Wing Logic dual element, I made inner end plates that allow the upper wing to adjust X-Y-Z coordinates for gap height, overlap, and angle of attack.

I also decided to glue the wing directly to the end plates. I did this for two reasons: 1) To take any bit of flex out of the system. Since the upper wing is supported only at the bookends, I wanted those to be ultra strong. 2) To avoid intersection drag from where the end plate boundary layer interacts with the wing’s boundary layer. This can cause flow detachment on the inside corner, and so I rounded and then fiberglassed this joint.

As a result of the inner end plates and glassing, the wingspan got a little wider, and so now it’s 54” even. This is still a rather short wing, and I’m certain longer would be better.

So I’m already thinking about version 3, which will be a longer dual wing with manually activated DRS system. I’ll use a different upper element as well, probably with a 5.6” chord. But that’s a winter project, and I have a few more track events to close out this season and get some final testing data (gap, overlap, angle) for this dual wing. With any luck, you’ll read about that in a couple weeks. Onward!

A lot of people say that putting a spoiler or wing on a FWD car is stupid. Downforce shifts the weight balance rearward, and that takes away grip from the front, where you need it for acceleration and turning. There is some logic to this thinking.

Tire grip is roughly equivalent to the amount of weight on the tires, and when you add downforce on one end of the car, it changes the balance of grip. For example, let’s take a RWD car that weighs 3000 lbs and has a 50/50 weight balance. If I add 125 lbs of rear downforce at 80 mph, now there’s 1500 lbs on the front tires and 1625 lbs on the rear. At speed, that’s 108% more grip on the rear, and the balance of traction is now 48/52, biased towards the rear.

The result of that shift in traction should result in better longitudinal Gs, meaning better acceleration and braking. However, lateral Gs may suffer slightly, and the car would be more prone to understeering in the middle of the corner.

Now let’s take a FWD car that also weighs 3000 lbs and has a 66.7/33.3 weight balance. That means it has 2000 lbs on the front axle, and 1000 lbs on the rear. Adding the same 125 lbs of rear downforce makes 1125 lbs on the rear, which results in about 112% rear grip at speed. So you can see that adding rear downforce on a FWD car changes the balance of grip quite a lot.

Just like RWD, longitudinal grip should increase when braking in a FWD car, by utilizing the rear tires more effectively. However, the opposite is true when accelerating, and I’d expect to see lower long-G at speed as the weight shifts away from the drive wheels. It’s likely that the same scenario plays out in a corner, with too much weight shifted rearwards; since most FWD cars push already, adding understeer likely makes the situation worse.

Theoretically and mathematically, adding rear downforce on a FWD car, without offsetting that with about twice as much front downforce, sounds like a recipe for going slower. Right?

Fuck math

Let’s throw all of those calculations out the window and see what actually happens when you put rear downforce on a FWD car. For this practical application I’m using my Hyundai Veloster N, and driving it at a fairly low-speed race track, Pineview Run.

I tested four rear aero options:

• N wing – The N model came with a large chunk of plastic that is someone’s stylized idea of what a wing should look like. It doesn’t really have an airfoil shape, but let’s call it a wing since air goes underneath it, as well as over the top. In the wind tunnel it made 27.3 lbs of rear downforce and 3.5 lbs of front downforce at 100 mph.
• Ducktail – I tested a DIY ducktail spoiler in the wind tunnel, and it made over 140 lbs of downforce, but also a ton of drag. I felt it was too tall, and cut it lower. I don’t have wind tunnel data for the shorter version, but I’ll throw out a guess and say it makes 100 lbs of rear downforce and a couple pounds on the front.
• Single-element wing – This is a DIY 54″x11″ S1223 wing with a small Gurney flap. I’ve tested it in the A2 wind tunnel and it made 113 lbs of rear downforce, but unlike the spoilers, it lost a little front downforce through leverage. The wing is located further rearwards and doesn’t aggregate high pressure on the roof, so the rear downforce gains are offset by front downforce losses; this is normal for all wings.
• Dual-element wing – The above wing, but with a 4.7″ chord wing set at 30 degrees above the main wing. I didn’t test this in the wind tunnel, and I don’t have a good guess on downforce or drag, but it’s more of both.

N wing vs Ducktail

Take a look at the following speed trace, which shows the OEM wing vs the DIY ducktail spoiler. I started the session using the ducktail and after setting a decent lap, I quickly pitted, removed the ducktail, and replaced it with the N wing. This took about a minute, and I got right back out again and set a few more laps. The lap with the ducktail is blue, and the red lines are the N wing.

A – The first significant difference is a pair of left hand turns. You can see the blue line has a higher minimum speed in both of them, and results in a slightly faster top speed on the next straight.

B – These are two fast right handers, and again you can see there’s a significant difference in the minimum corner speed. On the following straight, the top speed ends up being the same, but on a longer straight, it’s likely the ducktail would slow down from drag. At this point in the lap, the ducktail has about a .7 second advantage, but over the rest of the lap where the corners are slower, there’s no difference.

Subjectively, I could tell the difference between the two rear aero devices. With the N wing, the car suddenly felt more lively, and the rear end moved around. This translated to more fun, but a slower lap time.

Ducktail vs single wing

Next I tried the ducktail vs the single wing, and they set nearly the same lap times. In the following speed trace, the ducktail is red, and the single wing is blue.

A – The long left-hand sweeper is pretty much the only place where the wing has an advantage over the ducktail, keeping a higher speed throughout.

B – The spoiler makes it easier to back up the corner, and I’m at full throttle earlier than when I use the wing.

C – Again, I back up the corner slightly better with the spoiler. It might be that the additional drag is helping me slow the car and get it turned earlier, but that’s just conjecture.

Subjectively, I couldn’t tell the difference between ducktail and wing. The wing was ultimately about .1 seconds faster, but I didn’t set enough laps to make that statistically accurate, and I’d call it a wash.

Ducktail vs double wing

Again I used the ducktail spoiler as the baseline aero, comparing the double wing to it. Since the single wing and ducktail were quite similar in performance, you could also say this is a single wing vs double wing test.

For this one I’m showing three laps from each, and I’ve also included lateral Gs, because there’s some additional interesting data points here. The ducktail is again red, and the double wing is in blue.

A – This is the long left where aero makes a difference, and you can see the blue lines in the upper graph are lower, indicating more lateral Gs. In the speed trace, you can see the difference is a couple mph.

B – What’s interesting here is that the lateral Gs at this spot are the same, but the vMin is higher for the blue line. Perhaps this means I have increased confidence in this corner, despite the same Gs?

C – The higher vMin in the previous corner allows me to consistently get to a higher speed.

D – The back straight is the only section where the ducktail is faster, and it’s entirely due to having less drag. You can clearly see the advantage in top speed.

E – Here the dual-element wing allows me to brake a little later, and that also results in a little time gained.

Subjectively, I can’t say I felt a huge difference between the spoiler and double wing, except that the double wing makes a small whistling sound at times. And on the back straight, I swore I could feel the drag from the double wing, but that could also be my imagination.

Conclusion

It’s pretty obvious from the lap times that adding rear aero makes this driver faster in this car. The Veloster has more rotation built in than the average FWD car (though maybe not as much as an Elantra N), and so it can benefit from additional rear grip at speed. It may also be that my particular driving style benefits from more rear grip. I didn’t do this experiment with another driver, so I don’t know if driving style is an important variable.

Caveats aside, the data shows that the double wing was about .6 seconds faster than the single wing or the ducktail spoiler. And both of those were about .7 seconds faster than the OEM wing. All said, I went 1.3 seconds faster using rear downforce alone, with no splitter or canards (although I do have hood vents, and I suppose that should count for something).

What’s surprising is that rear downforce made a quantifiable difference on a low speed track where aero only matters in a few corners. On a longer track with sustained high speed corners (Club Motorsports, Nelson Ledges, Palmer, etc.), I could see this being a three second advantage.

Another surprising thing I learned in this test is that I can make a minor change to the rear aero, and you can see it clearly in the 10hz GPS data. I know the track well and I’m a decent driver, but I make mistakes and I don’t have pro-level car control or consistency. So it doesn’t compute that there should be such an obvious signal in the data.

But I think what I’m seeing here is that I probably drive the car to the same level of instability. When the car is more settled, it allows me to release the brakes sooner and/or open the throttle earlier. And that flattens out mistakes I make from corner to corner.

So while I may not be able to drive the same line or make the same inputs exactly the same lap after lap, I end up driving consistently by maintaining a certain level of instability that I’m comfortable with. And that’s why you’re seeing strong signals in the data for what are relatively minor aero changes at low speed.

I would imagine that many other competent (but not professional) drivers do the same thing, driving to a certain level of comfortable instability. And because of that, I think using a 10hz GPS device to get long-G, lat-G, and a speed trace is acceptable data gathering for aerodynamic experiments.

I guess the final question is, why does adding rear aero alone reduce lap times? Mathematically, I’ve ruined the aerodynamic balance of the car by adding a lot of rear downforce and none in front. Engineers should be storming my house with pitchforks and torches, rooting out the madman who Frankensteined their masterpiece…. but for fuck’s sake, the car is actually faster. Either I need to do a bunch of math to calculate why that’s happening, or build an even bigger wing and fuggin send it. Yep.

Updated March 2025

I originally wrote this in 2024, but the 2025 classing rules have changed for both NASA and SCCA time trials classing. Changes below are in italics.

I’ve been thinking about taking my car to a time trial race, and as part of that, researching all of the time trials rules. But one thing that may throw a spanner into my plans is that I don’t want to turn any spanners myself. So if racing in time trials voids my warranty, I’m not racing.

But I have other options. If you’ve been following my blog, you know about Falconet, my Hayabusa-swapped Miata. It’s a serious race car that can do wheel to wheel or time trials. It will compete next year.

Anyway, as this is towards the end of the track season, I’ll ruminate on which series and classes my Veloster could compete in, if and when that happens.

NASA TT5

NASA time trials divides cars into classes by weight and horsepower. Power is averaged over a range of RPMs, so that peaky (low torque) motors have more parity with engines that have a broader spread of power. And there are modifiers for many things from aero to suspension to tire width. On paper, it’s all very fair, although some of the rules are questionable.

If I use the class calculator at 3200 lbs race weight, and specify that my car is FWD, has narrow tires, uses a splitter and wing, the car is allowed 256 average wheel horsepower. My car makes 244 peak horsepower at the wheels, and so you see it fits into the TT5 class with some room to spare (12 hp) for a cold-air intake and an exhaust. If I were to remove the splitter and wing, and use the base trim aero modifier (BTM), the car is allowed 275 average wheel horsepower. Since I’m not going to modify my engine, that puts my car 31 whp under the class maximum, and I’d have a hard time being competitive.

Now here’s the problem: TT5 has a maximum factory horsepower limit of 270 hp. So even though my car is 31 hp under the class limit, I can’t race in it! If I had the non-PP version of my Veloster N, at 250 crank hp, then I could tune the shit out of it, up to 275 at the wheels and be class legal. But my engine, which is 244 peak horsepower at the front wheels, is somehow illegal.

I’d like to see the logic behind this power cap in a lbs/hp series. Are they just trying to keep out the heavy cars? How does that even make sense in age when every new car is north of 3000 lbs?

Other cars that should be able to race in the TT5 class based on lbs/hp, but can’t because they have more than 270 hp, include the BMW 135i and 335i, Ford Mustang (EcoBoost), Dodge Challenger and Charger SXT, Hyundai Elantra N, Kia Stinger, Mitsubishi Lancer EVO, Nissan 350Z, Subaru WRX, and a slew of others. These aren’t popular track cars anyway, so who cares, right? And so instead of TT5 my Veloster N and the aforementioned cars all have to race in the next class up, TT4.

In 2025, NASA changed the maximum factory (crank) horsepower to 264 hp for AWD and RWD cars, and 275 for FWD. This makes the Veloster N legal, but the Elantra N isn’t, even though it’s heavier and ends up with the same power to weight ratio. While I don’t get the reasoning, I’d like to thank NASA for making my car legal for TT5, and also thinning out the competition. I may be racing with you sometime in the future!

If I’m going to nitpick the other rules, the other thing I don’t like is they don’t allow aero items from other classes. I mean… just use the point system you already developed, and instead of making canards, fender vents, diffusers, etc., illegal, make them cost points.

Grid Life

My Veloster can race in the Street class, which allows 4-cylinder turbos and 6-cylinder normally aspirated engines. Performance wise, there are some ringers in the class, but the Veloster isn’t that far off the pace. I wouldn’t be able to use any aero, or that would put me in Street Mod, which I could never be competitive in.

My Veloster also fits perfectly into GLTC, but that’s a wheel to wheel race class. However, COM has a GLTC-spec class in time trials, and I’ll get into that later.

In Canada, I could also race in Canadian Time Attack Challenge, which uses Grid Life’s rule set.

SCCA Time Trials Nationals

SCCA Time Trials Nationals (or is it Time Attack Challenge in 2025?) is moving their home track from National Corvette Museum to PittRace, and will hold the Nationals competition on Sept 18-21 2025. This is exciting for me, because PittRace is one of my favorite race tracks, and it’s a lot closer than NCM.

There are three classes I could race my Veloster in:

• Max 3: This class allows a lot of aero, suspension, and engine tuning. I could use all of the aero I’ve built for my Veloster and race in this class, but for reasons I’ll explain in a minute, the car would be considerably down on power.
• Tuner 3: This class has the suspension mods and engine tuning of the Max classes, but no aero mods. I won’t be racing in this class. (Side note, I feel like cars that are prepped to the Sport level should be able to use aero mods if they don’t take any other Tuner mods.)
• Sport 3: This class allows minimal modifications, like lowering springs and intake/exhaust plumbing, but that’s it. It would be easy to prep for this class, but the Veloster isn’t at the higher end of the performance spectrum.

But that was in 2024, and now in 2025, instead of Max 3, Tuner 3, and Sport 3, my car goes into Max 4, Tuner 2, and Sport 3.

Max 3 Max 4

I’m an aero guy, and so my first thought is to race in the Max classes because the Sport and Tuner classes don’t allow aero. The aero rules in Max are quite good and show a constant evolution, losing some of the silly restrictions (flat splitters) that the legacy rules had. Therefore, I could use all of the aero I’ve developed in the wind tunnel, without making more parts to a different rule set. Kudos to the TT rules makers for their willingness to adapt and change.

The Max classes use engine displacement to determine classing, and unmodified 4-cylinder turbos up to 2.46 liters go into Max 4. My 1998cc Veloster gives away displacement to engines like a 2.3 Ecoboost, resulting in a 75 hp deficit. So that puts me at a slight disadvantage to some turbo cars, but I shouldn’t complain, as 3-cylinder engines are only allowed 1.9 liters (?).

N/A engines are allowed up to 4.5 liters. In this category you have cars like the E92 BMW M3 and 718 Porsche Cayman GT4, both putting out 414 hp. That 140 hp deficit would be OK if my Veloster had a handling advantage, but these are pedigreed German sports cars I’m up against. Even worse, a 911 GT3 makes 500 hp from a naturally aspirated 4-liter flat six, and is in the same class as my Veloster. Grumble grumble.

All said, if I want to be competitive, the Veloster doesn’t look like a great choice in Max 2. But I could also build the car with aero and not care about how I place, and do this race just for fun. That’s compelling, but also exactly what HPDE is for. Why enter a competition if you’re not going to be competitive?

Tuner 2

I never really considered the tuner classes, but I did notice a lot of cars moved from Tuner 3 to Tuner 2 in 2025, including the Veloster N.

Sport 3

Given that a stock-engined Veloster is way outclassed in Max 3, let’s see how it fares in Sport 3.

Unlike the Max classes which were based on displacement, the SCCA uses a list to determine which cars go in the Sport category. The rules say that the list of cars in each class is not based on a balance of performance, but on a “balance of popularity.” But I feel like that explanation is selling themselves short. The SCCA time trial board has done a great job of re-classing cars annually, and there’s better parity every year.

For example, the Acura NSX was originally classed in S2, but has been reclassed to S3, and is now considered an equal sparring partner to my Veloster N. The NSX is lighter and more powerful, but the 20-30 years between the cars may be significant. Other cars that moved from S2 to S3 include the Chrysler Crossfire (310 hp), Dodge Challenger R/T Scat Pack (470 hp), Kia Stinger (368 hp) Lexus GS-F (467 hp), and Nissan 370Z Nismo (350 hp). While all of these cars are more powerful than my Veloster, they are also heavier. Maybe it’s a fair fight?

As of this writing, the following cars are in S3:

Acura Integra Type S 2024, NSX 1990-2005; Alfa Romeo Giulia Ti 2016+; Audi A3 2.0T All 2015-2018, S3 2013-2020, S3 Quattro Wagon 2006-2013, S4 2009-2016, S5 3.0T FSI 2008-2017, TT S 2009-2018; BMW 135i 2008-2013, M235i 2014-2016, 330i xDrive Sport Wagon 2017-2019, 335D 2004-2013, 335i All 2006-2013, 440i xDrive 2014-2020, 535i RWD 2004-2010, 650i Base 2005-2010, M3 2001-2006, M5 1998-2003, Z4 Coupe M 2006-2008, Z4 35i M 2009-2016, Z4 35is M 2009-2016; Cadillac ATS 3.6L V6 2013-2015, CTS V 2004-2007, CT4 V 2022; Chevrolet Camaro SS 1998-2002, Corvette Z07 1984-1991, Corvette Base 1984-1997, Corvette Non-Z06 1997-2005, SS 2014-2017; Chrysler Crossfire SRT6 2004-2008, Dodge Challenger R/T Scat Pack 2015-2022; Ford Focus RS 2016-2018, Mustang Ecoboost 2015-2023, Mustang GT 2011-2014, Taurus SHO 2010-2018; Genesis G70 3.3T 2019-2022; Honda Civic Type R 2016 +; Hyundai Elantra N 2022, Genesis R-Spec (All) 2013-2016, Veloster N 2019-2022; Kia K5 GT 2021-2023, Stinger GT 2018-2023, Stinger GT2 2018-2023; Lexus GS-F 2016-2020, GS350 FSport 2016-2020; Lotus Evora Base – NA 2010-2014; Mercedes C450 AMG 4MATIC 2015-2016, C55 AMG 2004-2007, CLK55, AMG 2003-2006, SL550 All 2007-2013, SLK55 AMG 2005-2011; Mitsubishi Evo All; Nissan 370Z Nismo 2009-2013; Polestar 2 Dual Motor, Performance Pack 2020-2024; Pontiac Firebird Firehawk/Formula/Trans Am 1998-2002, GTO 2004-2006; Porsche Boxster S 2005-2012, Cayman S 2005-2012, Cayman Base 2009-2012, Cayman Base 2014-2016, 911 Carrera 2, 2S, 4, 4S 1999-2008; Subaru WRX ALL 2009-2021, WRX STI All; Toyota Supra GR 2.0 2021-Current, Corolla GR 2023-Current; Volkswagen Golf R 2015-2021; Volvo V60 Polestar 2016-2018

I’m not super interested in what the SCCA thinks about a balance of popularity, what I really want to know is, how do these all compare in performance? Hello COM.

COM

Corvette Owners of Massachusetts (COM) has perhaps the fairest time trials rules in existence. They calculate a car’s performance using the weight of the car divided by its power and then factor in a Performance Adjustment (handling).

To calculate power, COM uses an average of horsepower and torque, which is smart because the torque curve is hugely important. COM then modifies the power-to-weight score using Performance Adjustment points that range from -25 to +28.5. This factor takes into account things like suspension, brakes, tire width, aero, etc. Those two numbers are added together, resulting in a Showroom Assessment score, and from this the cars are placed in classes ranging from T30 (the slowest) to T110 (the fastest.)

The Veloster N is listed at 71.7 points, at the bottom of class T70. I could compete fairly in COM taking modifiers to aero or tires and come right in at the class limit.

But the other thing COM is great for is comparing the relative performance of different cars. And so I did that for all of the cars on the SCCA Sport 3 list, and the average is 72.5 points. So on first blush, it looks like the Veloster is right in the middle of the performance range.

However, there are a lot cars that are ranked higher, including the aforementioned Nissan 370Z Nismo, several BMWs (E46 M3, 135is, M235i, and Z4M), and every generation of Mitsubishi Lancer EVO or Subaru WRX STi. Given that every one of these has a better power to weight ratio, and that they are AWD or RWD vs the Veloster’s FWD… I have to think I’d be on the back foot against them. But I also feel like I’m a decent driver and my car is close enough in performance that I’m up to the challenge.

But there are even faster cars in Sport 3, and at the pointy end of the class are several cars that COM puts in T80, meaning they have more than 80 points. Among these is the Veloster’s nemesis, the CTR.

When the Veloster N came out, it was regularly pitted in comparison tests against the Honda Civic Type R (80.2 COM points). The CTR beat the VN every time, but it was close. In the comparisons, many YouTubers would choose the Veloster for the fun to drive aspect, but the fastest lap times were all owned by the Civic.

Some of the Type R advantage is Honda power, with +31 hp and +35 torque out of the same displacement. But COM also rates the Performance factor at 8.3 vs 7.5 points. So not only does the CTR out power the VN, it out handles it as well.

Which I find surprising because I think the Veloster is the best handling FWD car I’ve driven. Admittedly, I haven’t driven a Civic Type R, but I think the Performance Adjustment discrepancy comes down to the size of the wheels and tires. The Veloster comes stock with 8″ wide wheels, and 8.5” is about as wide as you can fit. Ergo, most people are using a 235 width tire.

Conversely, the Civic Type R comes stock with 9.5″ wheels, and you can go wider still. So it’s quite possible to fit 265 tires on a CTR, and maybe even go all the way to the class maximum of 285? The Elantra N isn’t on the COM list as of this writing, but like the CTR, it has more power and wider tires than a Veloster.

But my competition doesn’t end there, the overdogs in the S3 class are the 987.2 Cayman (80.4 points), 992 Carrera (82.3), Ford Focus RS (82.5) and non-Z06 C5 Corvettes (83.1). These T80 cars are higher performance than anything else on the Sport 3 list, and might even stomp a CTR.

So even if my Veloster N is in the middle of the COM performance window for SCCA’s Sport 3 class, it’s likely that a majority of the cars will have a better power-to-weight ratio, and significantly wider tires. The question is, can a 275 hp FWD hatchback on 235 tires beat a 350 hp RWD on 285 tires or 350 hp AWD on 265 tires? To find out, let’s see some historical results.

Sport 3 lap times

SCCA TT lists the results for every race, and after dumping them all into a spreadsheet (and correcting the fact that they don’t put the fields in the same place each time), I got an idea of which cars are the fastest and most popular in Sport 3.

At the top of the list is the FK8 Honda Civic Type R, with seven wins and lots of podiums. The next most popular is the 2017+ Camaro SS with five wins. After that you have a pair of German sportscars, the Porsche Cayman and BMW E46 M3, with four wins apiece. The Ford Focus RS is the first of the AWD cars and nets three wins, and then you have a contrast of early Porsche 911 vs modern Ecoboost Mustang with a pair of wins each. Finally there’s a smattering of cars with just one win, such as a 2014 Cadillac CTS, 2012 Audi TT, 2022 GR86, 2006 350z, 2015 M235i, and finally a 2021 Veloster N. Which is especially poignant because it won the 2021 Sport 3 class at PittRace.

PittRace lap times

SCCA Time Trials has been to PittRace a couple times, and looking at the results, it appears that I should be aiming for a 2:05 lap if I want to be the fastest ever Veloster, or a high 2:02 if I want the fastest all-time S3 lap (yeah, right). I’m sure that people will up their game for Nationals, and so I don’t doubt that someone will make a new lap record, weather permitting, but can someone go under the 2:00 minute mark?

I also looked at the Tuner 3 lap times at PittRace, and most are a couple seconds faster, with one car going under 2 minutes. Meanwhile, two FK8 Civics did 1:58s in Max 3. So the delta between Sport 3 and Tuner 3 is a little over two seconds, which is about the same gap between Tuner 3 and Max 3. Comparing Sport 3 and Max 3, I would need over 4 seconds from aero to make my Veloster competitive in Max 3, and I have a hard time seeing that happening.

Racing and Sim racing

I’ve raced at PittRace a few times in 24 Hours of Lemons and Champcar, and I’m slated to race a B-Spec Honda Fit this year with the Garage Heroes in Training. The event is is a Lucky Dog endurance race, and so with no mechanicals or mishaps, I’ll get a decent amount of track time. Which is important, because by next year, I need to know PittRace like the back of my hand.

To that end, I’m also doing some sim racing. My brother Ian (yousuckatracing) set me up with a sim rig, and I do a few laps every morning. I’ve been playing Gran Turismo since GT came out, but with a controller, and so it’s taking me a while to get the feel of a wheel and pedals. I’m currently only racing Assetto Corsa, because it has the PittRace track. A lot of my friends seem to be on iRacing, but that costs money and doesn’t have PittRace, so….

I’m just getting started sim racing, so I’m not too proud of these times, but in the spirit of transparency and learning from others, this is where I’m at:

Unfortunately I don’t have a car that approximates at Veloster N, but Ian is good at modding things, so I’ll get him to make me one eventually.