Make aero changes answer with numbers
Generated from
content/lms/telemetry-systems-engineering/04-math-channels-and-derived-signals/03-aero-and-performance-metrics.md; edit the source file, not this page.
Source path: content/lms/telemetry-systems-engineering/04-math-channels-and-derived-signals/03-aero-and-performance-metrics.md
Course: Design and validate the telemetry system that feeds every decision
Module: Build the math channels that turn raw data into insight
Estimated duration: 55 minutes
Purpose
An aero change is not proven because the car felt better, looked cleaner, or produced one impressive lap. It is proven when the numbers answer a specific engineering question: did this configuration make the car faster overall, where did it gain, where did it pay for that gain, and can you repeat the result after the baseline has had a chance to drift?
That is the job of this lesson. You are not learning how to design a wing, diffuser, splitter, or cooling outlet. You are learning how to make the car answer an aero question with a disciplined test and a small set of derived telemetry signals. The bonded material supports a very practical view: even a trace of speed or rpm versus time can expose corner speeds, straight speeds, elapsed time, split times, and braking deceleration. With more serious logging you can go deeper, but the skill does not begin with expensive equipment. It begins with asking one clean question and refusing to let unrelated variables answer it for you.
The core rule
For every aero test, separate the gain from the cost before you decide.
Downforce and drag do opposite things to the lap. Downforce increases the tire's capability to develop cornering force. Drag reduces the engine power available to accelerate the car. A setup can therefore make a high-speed corner better and a straight worse at the same time. If your math channels only show lap time, you may know the final answer but not the reason. If your channels only show peak straight speed, you may reject a setup that was faster over the lap. If your channels only show minimum corner speed, you may accept a setup that gave away more time on the following straight than it earned in the corner.
A good aero analysis does not ask whether more aero is good. It asks a narrower question: for this car, on this track section, in these conditions, did this configuration produce enough useful cornering performance to justify the drag or other tradeoff? The answer has to be numerical because the trade is numerical. Corner speed, straight speed, sector time, lap time, and repeatability are the language of the answer.
Why a speed trace can be enough
The simplest useful aero evidence can come from speed versus time, or from rpm converted to speed by using the speed per 1,000 rpm in each gear. From that trace you can calculate or read the basic signals that matter: elapsed time, split time, corner entry speed, apex or minimum speed, corner exit speed, straight-line speed, and braking deceleration from the rate of change of speed in a braking zone.
That is not a complete aerodynamic force measurement. It is still powerful because aero changes show up as performance changes. If you increase downforce, the useful effect you are looking for is higher speed in the parts of the lap where aerodynamic load matters, especially higher-speed corners. If the same change adds drag, you may see a loss in straight-line speed or acceleration. The lap-time question is the net of those two effects. The sector-time question tells you where the trade happened.
This is why a basic logger can beat a stopwatch for aero testing. A stopwatch tells you the final elapsed time. A speed trace tells you how the lap was made. It gives you a permanent record of each lap or run and lets you go back later to inspect whether the gain came from the car, the driver, the track, or a confusing mix of all three.
The engineering question comes first
Do not begin by building channels. Begin by writing the question you need the channels to answer. A useful question is narrow enough that the data can say yes, no, or inconclusive.
Weak question: is this aero package better?
Useful question: does the higher rear-wing setting improve high-speed corner speed enough to offset any straight-line speed loss over the measured lap?
Useful question: does the lower-drag configuration improve the long straight without giving away enough high-speed corner speed to hurt total sector time?
Useful question: after returning to the baseline, did the original baseline still perform like it did earlier, or did weather, track condition, or tire deterioration move the reference?
Each question implies the channels you need. You need a net time channel because the decision is about elapsed time. You need sector channels because aero gains and drag losses happen in different parts of the lap. You need corner-speed channels because downforce earns its keep in cornering. You need straight-speed or straight-acceleration channels because drag taxes acceleration. You need a baseline-repeat check because changing conditions can make the answer look cleaner than it is.
The minimum useful channel set
For an intermediate Tracky telemetry workflow, build the smallest set of derived channels that can answer the question without burying you in signals.
First, create lap elapsed time and sector elapsed time. The source material specifically calls out elapsed time and split times as information that can be extracted from speed or rpm traces. Use these channels as the scorecard. If a configuration is faster in one corner but slower over the lap, the lap and sector channels keep you honest.
Second, create corner entry, minimum, and exit speed channels for the higher-speed corners you are using as aero evidence. The bonded material points toward higher-speed corner entry, apex, and exit speeds, with the rough neighborhood of 60 mph or 100 km/h and above depending on downforce level. The exact threshold is not a magic rule. The point is that you should not use a slow, traction-limited corner as your primary aero witness when the source material is telling you to look where downforce actually has a meaningful performance signature.
Third, create straight speed channels. These may be terminal speed at the end of the straight, speed at a fixed point on the straight, or time through a defined straight sector. The question is not simply whether the car was fastest at the end of the straight. The question is whether the drag cost changed the time budget enough to matter. Straight speed is evidence. Straight sector time is stronger evidence.
Fourth, create a braking deceleration channel if the speed trace and braking zone are clean enough. The bonded material identifies braking deceleration rates as something available from the rate of change of speed in a braking zone. Use this as a supporting channel, not as your only aero verdict. A change that alters high-speed behavior can influence how the car arrives at the brake zone, and a speed trace can show whether the braking segment changed along with the corner and straight segments.
Fifth, create a baseline drift check. This can be as simple as comparing the first baseline run against a later return-to-baseline run over the same lap and sectors. The bonded material explicitly warns that weather, track condition, and tire deterioration can move the baseline during a session. If the baseline moved, the configuration result has lower confidence until you account for that movement.
Sixth, keep a driver-feedback field beside the math. The source material does not treat driver feedback as a replacement for data. It says configuration changes can be supplemented by driver feedback on aerodynamic handling balance. Your driver note should be short and tied to the same sections as the channels: stable in the fast entry, nervous at exit, reluctant in the long corner, or balanced but slower on the straight. Do not let the note overrule the data. Use it to explain where to look.
The test method matters as much as the channel
A perfect math channel cannot rescue a bad aero test. The bonded corpus is blunt about discipline. Carroll Smith's documented comparison of two wings used five laps for each configuration, changed only the wing configuration, averaged the lap times, and discarded abnormally high or low times. The text also stresses returning to the baseline periodically, especially when weather or track conditions change, because some variables can always be relied on to alter the baseline, including tire deterioration.
That gives you a practical track-day method.
Start with a mechanically sensible baseline. The source material describes aero configuration work on a racecar with an optimized mechanical setup. That matters because you are trying to isolate an aero effect. If the car is still being sorted mechanically, the aero channels may be answering a suspension, tire, brake, or driver question instead.
Change one aero variable at a time. If you change wing angle, splitter height, tire pressure, brake bias, and driver approach in the same session, the lap may still get faster, but the aero change did not answer with numbers. The whole point is to compare different data sets so the effect of a setup change can be revealed. Multiple changes make the comparison ambiguous.
Run enough laps to average. The bonded example uses five laps per configuration. Five is not magic, but it is a useful minimum discipline because one lap can be driver adaptation, traffic, a missed shift, a different tire state, or a track condition change. Averaging lets the configuration answer as a run rather than as a single lucky or unlucky lap.
Discard obvious abnormal laps before you compare averages. Do not use this as a way to cherry-pick. Use it to remove laps that clearly do not represent the run: traffic, major driver mistake, a missed braking point, or a lap that is abnormally high or low relative to the rest of the run. The corpus calls this crude but effective. The important part is that you define the discard honestly and apply the same standard to each configuration.
Return to baseline. This is the part many amateur tests skip. If your first baseline run happened on fresher tires or a cooler track, and your new aero run happened later, you may credit the configuration for something the conditions caused. A later baseline run gives you a reality check. If the baseline is slower later by the same amount as the new configuration, your apparent gain may not be a gain at all.
Build the lap into evidence zones
Before you drive the test, divide the lap into zones. The zones should match the physics of the question.
Use high-speed corners as downforce zones. A high-speed entry, apex, or exit is where the added tire cornering capability from downforce is most likely to show up in speed and sector time. Do not ask a tight low-speed corner to prove an aero change unless the car and track make that corner genuinely aero-sensitive. The corpus specifically names higher-speed corner entry, apex, and exit speeds as useful traditional test parameters.
Use long straights as drag zones. Drag reduces the car's ability to accelerate, so a straight is where the penalty is easier to see. You may look at terminal speed, speed at a fixed point, or time through the straight. If the more aggressive aero setting gains in the fast corner but loses on the straight, that is not a contradiction. That is the expected trade you are measuring.
Use whole-lap and sector time as decision zones. The lap decides whether the package is faster overall. The sectors explain why. A sector that combines a fast corner and following straight can be especially useful because it captures the gain and the cost together. A pure corner zone and pure straight zone help you diagnose. A combined sector helps you decide.
Use braking zones as supporting zones. The source material allows braking deceleration to be derived from speed change. This can help identify whether a setup changed the way the car approaches or survives the braking event, but be careful. Braking traces are sensitive to driver execution. Treat them as supporting evidence unless the test was controlled tightly.
How to read the signatures
A useful aero gain has a pattern. The high-speed corner speed or corner sector improves. The driver note may report a better aero handling balance. The straight speed may fall if drag increased. The whole lap or the relevant combined sector must still improve before you call it a net gain.
A useful low-drag gain has the opposite pattern. The straight improves. The high-speed corner may lose speed if you removed useful downforce. The decision depends on whether the straight gain is larger than the corner loss. Do not accept a low-drag change just because the car reaches a higher peak speed. Drag matters because it changes acceleration and time. Peak speed is only one symptom.
An inconclusive test also has a pattern. The baseline does not repeat. The best laps all come from one run but the averages are close. Corner gain and straight loss are both small relative to run scatter. Driver comments change more than the speed traces. The car feels better but the sector times do not move. In that case the correct answer is not to invent certainty. The correct answer is that the current data does not prove the change.
A driver-learning result has another pattern. Lap time improves across the day, but the return-to-baseline run also improves. The new configuration may not be the cause. Comparative analysis can reveal setup changes and driver performance, but it only does so if you structure the comparison. If the driver was still learning the corner, the aero package may be riding on top of a driver improvement.
Calibration and installation are part of the lesson
The bonded corpus includes a reminder that data systems need to be installed and calibrated to give useful results. For this lesson, calibration does not mean a laboratory-grade aerodynamic force balance. It means your basic speed or rpm-derived speed channel is trustworthy enough to compare runs.
If speed comes from rpm, confirm the gear and speed-per-1,000-rpm relationship before you use the channel. If the car uses different gears through a section on different laps, note that because the conversion and interpretation can change. If speed comes from a logger, make sure the trace is clean enough to identify the same zones each lap. If lap or sector boundaries move around, your derived time deltas will answer a timing-marker question instead of an aero question.
Also calibrate the human procedure. Use the same out-lap rhythm where practical. Keep the driver's objective the same. Record the aero setting immediately. Record tire state and obvious track condition changes. Do not rely on memory after the session. A small logger plus disciplined notes is better than an impressive channel list with no record of what changed.
What the final answer should look like
At the end of the test, the output should be a short engineering answer, not a pile of traces.
State the configuration compared. State how many laps were used. State which laps were excluded and why. State whether the baseline repeated. Then report the lap-time average, the relevant sector-time averages, the high-speed corner speed changes, and the straight speed or straight-sector changes. Finish with a decision: keep the change, reject it, or retest because the baseline drift or scatter is too large.
A strong conclusion sounds like this in substance, without needing to be dramatic: the higher wing setting improved the fast-corner sector and the combined lap average even though it reduced straight speed, and the return baseline confirmed the track had not moved enough to explain the gain. A weak conclusion sounds like this: the car felt better and one lap was faster. Do not ship the weak version.
Where simulations and visual tools fit
Performance simulation, CFD, wind-tunnel data, and flow visualization are not competitors to track data. They answer related questions. The bonded material describes performance simulation as a way to evaluate varying aerodynamic configurations and predict a probable best setup before reaching the circuit, especially when practice and qualifying time are limited. It also describes CFD and full-scale wind-tunnel data as tools for understanding aerodynamic effects, and flow visualization as a way to see what the air is doing around wings, spoilers, diffusers, cooling intakes, and outlets.
Use those tools to aim the test. A simulation may suggest which configuration is worth running. Flow visualization may explain why a wing, diffuser, or cooling outlet behaves the way it does. Wind-tunnel or CFD work may help you build the candidate setup. But the Tracky math-channel task is still the same: once the car is on track, make the configuration answer through speed, sector, lap, and baseline-repeat numbers.
The discipline is proportional to the budget
The sources make a useful point for amateur racers: the budget can vary, but the need for careful use and common sense does not. Simple tools can provide useful information. Complex tools can provide more information. Neither protects you from a sloppy comparison.
If you have only speed over time, build the core channels and run a disciplined comparison. If you have a more serious data acquisition system, add more resolution, but do not lose the basic question. If you have simulation output, use it to choose what to test, but still make the track data confirm the net effect. If you have flow visualization, use it to explain and improve the package, but still ask whether the lap got better.
The habit you are building is metric-driven thinking. You extract the maximum information from the hardware at hand. You compare runs to previously collected data. You draw the right conclusion quickly from a large data set by designing the metric before opening the data. That is how a driver or engineer with ordinary logging can make aero decisions that are better than guesses.
Cross-references inside this module
The sibling lessons on designing math channels are the foundation for this lesson. This lesson is one concrete application: every derived channel exists to answer a setup question. The sibling lessons on slip angle and chassis state are adjacent but separate. Slip-angle and chassis-state channels can help explain how the car is behaving, but this lesson's primary decision channels are speed, elapsed time, sector time, straight performance, braking deceleration, baseline repeatability, and driver balance notes. Keep the scopes separate so the aero test does not turn into a general handling analysis.
The takeaway
Make the aero change argue its case in numbers. Start with the physics trade: downforce can improve cornering capability, drag can reduce acceleration. Choose the zones where those effects should appear. Run a disciplined baseline and configuration comparison. Average enough laps to avoid worshipping one lap. Discard abnormal laps honestly. Return to baseline when conditions may have moved. Then decide from net time, sector time, corner speed, straight speed, and repeatability.
If the data says the change is faster, keep it. If the data says it moved the gain to one part of the lap and gave it back elsewhere, reject it or refine it. If the baseline moved too much, retest. The win is not just finding the faster aero setting. The win is knowing why it was faster and being able to prove it the next time the car goes back on track.
Worked example: two wing configurations over repeated runs
You want to compare two rear-wing configurations. The supported method is simple and demanding: run one configuration for five laps, change only the wing configuration, run the other configuration for five laps, average the useful lap times, discard abnormally high or low times, and return to baseline periodically if the session conditions may have moved.
Before the first run, define the evidence zones. Pick the high-speed corner or corners where downforce should matter. Pick the longest straight or the straight most likely to show a drag cost. Define the combined sector that includes the corner exit and following straight if that is where the trade will be paid back. Build the channels before looking for the answer: average lap time, sector time, corner entry speed, minimum or apex speed, corner exit speed, terminal or fixed-point straight speed, braking deceleration if the braking zone is clean, and a baseline repeat comparison.
Run baseline wing A. The driver should not be asked to hunt for a heroic lap. The target is a representative run. Save notes on handling balance in the fast corner and on straight-line feel, but keep those notes secondary. Change only to wing B. Do not change tire pressures, alignment, ride height, driving objective, or anything else unless the test scope explicitly includes that change. Run the same count again.
Now compare the averages. If wing B raises the speed in the high-speed corner and improves the corner sector, that is the expected downforce-side evidence. If wing B lowers straight speed or increases straight-sector time, that is the drag-side evidence. The decision comes from lap average or from the combined sector that best represents the lap's trade. If the high-speed corner gain is larger than the straight loss, B has made a numerical case. If the straight loss is larger, the car may feel better but the stopwatch and sector math reject the change.
Finally, return to baseline A if there is enough running time or if conditions are changing. If the return baseline is different from the first baseline, mark the result with lower confidence. The source material warns that weather, track condition, and tire deterioration can move the reference during the session. A disciplined aero test does not hide that problem. It exposes it.
Worked example: the straight-line drag question
Now suppose the question is narrower: did the lower-drag configuration improve straight-line performance enough to matter? The bonded material says indirect measurements of configuration effects on sector or lap times and speeds are often all you need, and it also notes that drag is the easier of the two aerodynamic forces to begin measuring directly, requiring a long, straight, flat, smooth piece of road for that kind of work.
For a track-based math-channel answer, stay with the indirect performance evidence unless you have a validated direct-force method. Define a straight sector and make sure it starts after the car is settled on the straight and ends at a consistent point before the braking event. Use speed at a fixed point, terminal speed, and straight-sector elapsed time. Keep the linked high-speed corner sector in view because a lower-drag configuration may also remove useful downforce.
Run the same disciplined comparison: baseline, one change, repeated laps, abnormal laps removed honestly, and a later baseline check if the day is moving. If the lower-drag setup improves terminal speed but the straight-sector time barely changes, do not overstate the result. If it improves the straight-sector time but gives away more in the high-speed corner before the straight, the lap may reject it. If it improves straight-sector time and the linked corner sector remains stable, then the change has a stronger numerical case.
The key lesson is that straight-line speed is not the same as performance. Drag reduces acceleration capability, so straight speed is a symptom, but elapsed time is the score. A higher number at the end of the straight is useful only if the sector and lap show that it bought time rather than just moving the shape of the speed trace.
Common mistakes
Mistake one is proving the change with one lap. One fast lap is not a configuration result. It may be driver execution, traffic, a cleaner line, a different tire state, or track evolution. Good looks like repeated laps, averaged runs, and honest removal of abnormal laps.
Mistake two is changing more than one thing. If you alter the aero setting and the mechanical setup at the same time, the data cannot identify the cause. Good looks like one aero variable changed while the rest of the car and the driver objective remain stable.
Mistake three is using the wrong part of the lap as the witness. A low-speed corner may be important to lap time, but the bonded material directs aero assessment toward higher-speed corner entry, apex, and exit speeds when looking for downforce effects. Good looks like choosing aero-sensitive corners and drag-sensitive straights before the session starts.
Mistake four is ignoring drag because the car feels better in the corner. More downforce can make the car easier to trust, but drag can reduce acceleration. Good looks like reading the corner gain and the straight cost together, then using sector and lap time to decide.
Mistake five is ignoring the baseline after the first run. A baseline at the beginning of a session is not guaranteed to remain the baseline later. Weather, track condition, and tire deterioration can move the reference. Good looks like a return-to-baseline run or at least a baseline-drift note that prevents false certainty.
Mistake six is treating driver feedback as either useless or decisive. The corpus supports driver feedback on aerodynamic handling balance as a supplement to logged performance. Good looks like a short handling-balance note tied to a specific sector, then checked against speed and time.
Mistake seven is overbuilding the channel set before the basic answer exists. The sources emphasize extracting useful information from the hardware at hand and using tools carefully with common sense. Good looks like lap time, sector time, corner speed, straight speed, braking deceleration where useful, and baseline repeatability before exotic analysis.
Mistake eight is keeping only the final verdict and losing the trace. The value of logging is that the record can be inspected again. Good looks like saving the runs, settings, excluded laps, and notes so the decision can be audited later.
Drill: the five-lap aero answer
At your next appropriate test opportunity, run the five-lap aero answer drill. The purpose is not to find the perfect setup in one day. The purpose is to practice making one aero change answer with a clean numerical structure.
Choose one aero variable and one question. Example: does the higher wing setting improve the high-speed sector enough to offset straight-line speed loss? Before driving, define one high-speed corner zone, one straight zone, and one combined sector. Build or prepare the channels for lap time, sector time, corner entry speed, corner minimum or apex speed, corner exit speed, straight fixed-point or terminal speed, and baseline repeatability.
Run five laps in baseline configuration A. Exclude only laps with obvious abnormal causes, and write the reason. Change only the chosen aero variable. Run five laps in configuration B. If the schedule and conditions allow, return to A for five more laps to check whether the baseline has moved.
The success criterion is a one-page answer, not a faster lap. The answer must state the question, the configurations, the lap count, excluded laps, baseline-repeat result, the high-speed corner change, the straight change, the lap or sector time change, and the decision: keep, reject, or retest. If you cannot write that answer from the data, the drill did its job by showing that the test design or channel set was incomplete.
When this principle breaks down
The principle breaks down when the corpus-supported assumptions are not met. If the car does not have a stable mechanical setup, the aero comparison may be polluted by mechanical behavior. If the driver is still changing the way the lap is driven, comparative analysis may reveal driver performance more than setup effect. If the track, weather, or tires move the baseline and you never return to baseline, you may assign the change to the wrong cause.
It also breaks down when the selected segments are not sensitive to the aero question. If you test a downforce change mostly through slow corners and short straights, the signals may be too small or too mixed to interpret. If you test a drag change without a meaningful straight or without stable speed data, the result may remain indirect and weak.
Finally, it breaks down when you ask the telemetry to provide aerodynamic force detail that the available channels do not support. The bonded material says direct aerodynamic force measurement is possible and that drag is the easier starting point, but it does not provide a full force-measurement procedure in these chunks. In this lesson, the safe claim is performance-based: corner speed, straight speed, split time, lap time, braking deceleration, and repeatability can make many aero changes answer. They do not magically turn every track test into a wind tunnel.
Author Review
No quiz questions are attached to this lesson.
Sources
| # | Document | Chunk | Pages | Score | Collection |
|---|---|---|---|---|---|
| 1 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 2dcc6067-583b-6042-00b6-d306f5d46cd6 | 344 | 1 | uio_books_raw_v1 |
| 2 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 4adf8cb4-89c7-1b45-bd4d-9bb03634ecf3 | 345 | 1 | uio_books_raw_v1 |
| 3 | Competition Car Aerodynamics 3rd Edition McBeath Simon | c0cd0f54-6d9c-7f08-e9af-37c31b3421d3 | 345 | 1 | uio_books_raw_v1 |
| 4 | Analysis Techniques for Racecar Data Acquisition | 623bcaf5-8885-a44d-d415-2838517701c3 | 17 | 1 | uio_books_raw_v1 |
| 5 | Analysis Techniques for Racecar Data Acquisition | ad559d04-3651-61c2-d02b-5455aba0d7cc | 7 | 1 | uio_books_raw_v1 |
| 6 | Analysis Techniques for Racecar Data Acquisition | 5eeea298-6191-0fb2-1054-b10fe574a804 | 2 | 1 | uio_books_raw_v1 |
| 7 | Analysis Techniques for Racecar Data Acquisition | d0db9128-dc9a-aec3-14a8-5f101654753f | 3 | 1 | uio_books_raw_v1 |
| 8 | Competition Car Aerodynamics 3rd Edition McBeath Simon | d96fdf59-210e-7376-263d-0a249955f375 | 382 | 1 | uio_books_raw_v1 |
| 9 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 9f0edfc1-9e8c-3a96-a48d-b0d658513db3 | 385 | 1 | uio_books_raw_v1 |
| 10 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 576d96a1-00b7-66dd-f5b1-e33666cc457f | 334 | 1 | uio_books_raw_v1 |
| 11 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 4b5e1aa7-14cf-aacf-908a-c47094ea7ba5 | 504 | 1 | uio_books_raw_v1 |
| 12 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 61068e74-0999-1e25-03bd-8c545f352d25 | 26 | 1 | uio_books_raw_v1 |