Define the operating window before adding downforce
Generated from
content/lms/race-aerodynamics/04-control-the-aero-platform/05-define-a-safe-operating-window.md; edit the source file, not this page.
Source path: content/lms/race-aerodynamics/04-control-the-aero-platform/05-define-a-safe-operating-window.md
Course: Engineer downforce you can actually use
Module: Control the aero platform
Estimated duration: 45 minutes
The principle
You do not start an aero change by asking how much more downforce you can bolt onto the car. You start by defining the operating window in which the car can use the aero it already has. For this lesson, the operating window means the set of speeds, ride heights, pitch attitudes, yaw attitudes, traffic conditions, and driver phases where the car's aerodynamic balance stays predictable enough that you can drive it hard without guessing.
That word predictable matters more than the word maximum. A wing, splitter, floor, spoiler, diffuser, or body modification can create a bigger downforce number in one condition and still make the car worse if the useful range is narrow, if drag costs more lap time than the downforce returns, if the load arrives mostly at one axle, or if the flow separates when the car pitches, rolls, yaws, follows another car, or reaches a running ride height different from the setup sheet. The window is the boundary between a setup you can exploit and a setup that only looks quick in a single measurement.
The intermediate driver already understands that more grip is good. The next step is learning that aero grip has conditions attached. Aerodynamic forces are generated by the way air interacts with the bodywork and devices on the car. A wing produces downforce from both upper and lower surface effects, with the lower surface especially important in the source material. That force is only useful if the flow remains in a condition that produces repeatable load. If a wing or diffuser produces more load until the flow separates or stalls, the number you care about is not the largest load before the cliff. The number you care about is the usable range before the cliff, and how the car behaves as it approaches that edge.
Aero also cannot be separated from the tires. The tires are still the only contact between the car and the road. They transmit acceleration, braking, cornering, and your control inputs, and they are also where you receive much of the sensory information that tells you whether the car is still under you. Aerodynamic download adds to mechanical grip, but it does not replace the need for a balanced mechanical platform. Carroll Smith's tire chapter is especially useful here because it warns against treating suspension and mechanical grip as irrelevant just because a car has aero. The grip generated by aerodynamic download is additive to mechanical grip. The basis is still a linear car with good mechanical grip, and the average racing corner apex is often at a speed where mechanical grip remains dominant.
That is why defining the window comes before chasing the number. A car that is mechanically nervous, inconsistent at ride height, or unbalanced across the phases of a corner will not be fixed just because you add a device that makes more load in straight-ahead air. You may hide one problem and create another.
What you are defining
Your operating window is not a single setup sheet value. It is a practical statement you can test. It should tell you where the car is reliable, where it is marginal, and where you do not yet have evidence. A useful window statement sounds like this in plain language: above a certain speed on clean straights the aero balance is stable; during hard braking the front ride height stays in the range where the floor and splitter remain useful; in medium-speed yaw the rear wing does not create a sudden rear balance shift; in traffic the car needs offset air before turn-in; and below the low-speed apex range the car is mostly a mechanical-grip car, so wing angle is not the first knob.
That statement is more valuable than a proud downforce estimate. It gives you a boundary for decision making. It tells you whether to add wing, reduce drag, raise or lower ride height, change rake, adjust roll stiffness, protect the floor from a bad attitude, or leave the aero alone and work on mechanical balance.
This lesson sits after the lessons on running ride height, front and rear aero sensitivity, floor honesty, and separating the setup sheet from the car at speed. Those lessons give you the measurements. This lesson tells you how to combine those measurements into a driver-useful rule. You are not merely asking what the car was on the scales or in the garage. You are asking what the car becomes at speed, under braking, in cornering yaw, and near another car.
The mechanism behind the window
Aero load changes the vertical load on the tires. That is why it can make the car faster. More load can let a tire generate more cornering, braking, or drive force, but the tires are still the limiting interface. Performance prediction in the McBeath material is built by combining car data, track data, tire behavior, power, braking, gearing, and an aerodynamic model so that tire loadings and straight-line performance can be worked out under changing conditions. That is the right mental model for a driver too. Aero is not a decorative layer on top of the lap. It changes tire loading, drag, acceleration, braking, and balance, all at once.
The trap is that those effects do not all point in the same direction. A larger wing angle may increase downforce and increase drag. A lower front ride height may make the splitter or floor stronger until it becomes too close, too pitch-sensitive, or too easy to upset. A rearward balance shift may feel secure in a fast sweeper and lazy in a slower transition. A setup that is confidence-inspiring in clean air may lose grip or balance when it follows another car. A car that is impressive in straight-ahead wind-tunnel style thinking may be less good when it is yawed in a corner.
McBeath's discussion of sports prototypes at small yaw angles is the key warning. Straight-ahead testing is useful for basic mapping, drag optimization, and high-speed straight-line braking work where downforce distribution shifts during the phase. But cars spend much of a lap in corners, and cornering means attitude, yaw, roll, and changing ride height. Formula 1 teams test at yaw angles representative of cornering because the aerodynamic performance can differ between straight-ahead and cornering conditions. The intermediate lesson is simple: if your window is defined only in straight-line air, it is not a window. It is one slice.
Flow attachment is another boundary. A wing or body device can be arranged so that flow remains attached across more of its span for longer, allowing more downforce before large-scale separation and stall. That wording points at the real goal. You are looking for a shape and ride attitude that produces load progressively and keeps producing it through the phase you need, not a setup that creates one heroic number just before separation.
The same principle applies under the car. The corpus does not give enough detail in these chunks to teach a complete underbody design lesson, and that is handled by the sibling floor lesson anyway. What it does support is the conservative rule: underbody designs differ, professional teams use CFD to explore many configurations and wind tunnels to validate them, and amateur racers should use practical tools to understand what their own car is doing. For you as the driver, that means the floor's useful range is an evidence problem. You do not assume it is honest. You prove where it works.
The five boundaries of the window
The first boundary is speed. Downforce grows with airspeed, but drag also matters. The performance-simulation material emphasizes the downforce-to-drag compromise, especially for high-downforce categories and aerodynamically efficient cars such as Le Mans prototype sports cars. At the driver level, you ask whether the extra load helps in the speed band where you need help, and whether the drag penalty hurts the straights or acceleration zones enough to erase the benefit. A device that helps only at a speed you rarely reach, or hurts every straight to help one corner, may be outside the useful window.
The second boundary is ride height and pitch. The car's aero state in the paddock is not the car's aero state on the straight, over a crest, under braking, or at turn-in. Running ride height matters because the body and floor meet the air differently when the car is loaded, braked, rolled, or compressed. If you have no evidence of the car's running attitude, you do not yet know the window. You only know the garage condition.
The third boundary is balance distribution. You are not simply adding grip to the car as a whole. You are changing how much load each axle receives and when it receives it. A car that gains rear load with speed may feel planted in a fast bend and then push at turn-in. A car that gains front load sharply during braking may invite you to brake later until the release phase exposes a rear that has not kept up. The exact behavior has to be measured or felt, but the principle is universal within the corpus: the aero model influences tire loadings, and tire loadings influence performance.
The fourth boundary is yaw and cornering attitude. Clean straight-ahead mapping is not enough because the car is rarely in only that condition when lap time is made. Yaw can change how devices see the air. More lap time is generally spent in corners than in perfect straight-ahead attitude, so a window that excludes cornering attitude is incomplete. This is especially important when you are evaluating a car that feels stable on the straight but becomes vague or inconsistent at entry, mid-corner, or exit.
The fifth boundary is surrounding air. McBeath's case-study material is blunt that cars running close together interact aerodynamically. The source describes the following car mitigating losses by offsetting laterally, with the left offset being useful for the tested left-hand-drive configuration. You do not need to turn that into a universal passing rule. You do need to treat traffic as a separate aero condition. If your car is quick only in clean air and unreliable when following, your operating window must say so.
A practical definition process
Start with the car as it is, not the car you hope it is. Write down the aero configuration, the static ride heights, the visible floor and bodywork condition, the wing or spoiler settings, and the mechanical setup that affects platform control. The performance-prediction material names mass, dimensions, and roll stiffness as inputs used to calculate forces, accelerations, and tire loads. You may not have a full model, but you can still respect the idea: the car's mass, stiffness, dimensions, and tire behavior shape whether an aero change becomes useful grip or an unstable compromise.
Next, divide the lap into phases. Use straight-line acceleration, high-speed braking, turn-in, mid-corner, exit, and traffic. Do not mix them together too early. A change that helps one phase can hurt another, and you need to know which phase moved. If you only ask whether the lap time improved, you will miss why it improved and whether it is repeatable.
For the first run, establish the baseline window. You are not tuning yet. You are trying to identify where the current car is calm, where it asks for correction, and where it changes balance with speed or attitude. Your notes should be phase-specific. The useful question is not whether the car had understeer. The useful question is whether it had high-speed entry understeer after brake release, mid-corner front wash in yaw, exit traction loss after aero load decayed, or straight-line nervousness over ride-height movement.
When you have data logging, use it carefully. The corpus points to data logging as a practical way to improve both car and driver, but it also stresses installing and calibrating systems to give useful results and using strategies to extract useful information. That matters. A speed trace, steering trace, throttle trace, brake trace, ride-height channel, damper channel, or yaw-related channel is only useful if you know whether it is calibrated, repeatable, and relevant. Bad data does not become evidence because it appears on a screen.
When you do not have full data, use disciplined observation. McBeath supports practical visual understanding of airflow around wings, spoilers, diffusers, cooling intakes, and outlets, including methods that can be used on track during testing or competition when test time is limited. The important point is not the brand of visualization method. The important point is that you look for evidence of what the air is doing, especially near devices that decide whether the setup is inside or outside its window.
Then make only one meaningful aero change. The larger the change, the easier it is to feel, but the harder it may be to separate from side effects. The smaller the change, the more repeatable the conditions need to be. The source material is realistic about this. Simulation can test changes in equal conditions, while the track rarely gives equal conditions. Your job is not to pretend the track is a lab. Your job is to use common sense, repeat runs where possible, and avoid declaring a result from one messy lap.
After the change, compare phases rather than adjectives. Did the car gain speed into the braking zone but require earlier braking because drag or balance changed? Did it hold mid-corner speed with less steering correction? Did top speed suffer? Did the car become better in one clean-air section and worse behind another car? Did the first lap feel good and the later laps become inconsistent as tires, ride height, or driver confidence changed? These are window questions.
Finally, write the window in a form that can guide the next decision. A good window statement names conditions and boundaries. It might say that the current wing setting is acceptable in clean air from medium to high speed, but not yet confirmed in close traffic. It might say that the splitter change helped high-speed front support but made the car too pitch-sensitive under hard braking. It might say that the car needs more platform control before adding more rear wing, because the balance change with speed is already larger than the driver can exploit. The exact content depends on your car. The discipline is the same.
How to feel the window from the driver's seat
You feel a useful aero window as repeatability. The car does not have to be easy. A race car can be demanding and still be inside its window. The important thing is that the demand is consistent. If the car asks for a known steering rate at high speed, gives you similar brake-release behavior lap after lap, and lets you make small corrections that produce proportional responses, you are operating in a usable range.
You feel a narrow or failing window as suddenness. The car may feel strong and then abruptly vague. The front may bite on the initial brake and then wash when the platform changes. The rear may feel planted on entry and then step as yaw or ride height moves the air out of the condition that made the load. You may find yourself delaying inputs because you are waiting to discover what the car will be this lap. That hesitation is information. It usually means you are past the useful window, or you have not yet identified it clearly enough to drive to it.
Be careful with confidence. Extra rear downforce can make a car feel safer because it reduces high-speed oversteer. That can be useful. It can also mask drag cost, low-speed understeer, or a balance shift that costs rotation. Extra front support can make the car feel eager, but if it arrives mostly in braking and disappears at release, it can teach you to overdrive entry. The driver cue is not whether one lap felt exciting. The cue is whether the car lets you repeat the phase with less correction and better speed.
Telemetry signatures to look for
The simplest useful signature is repeatability in speed and inputs. If an aero change expands the window, you should be able to enter the same phase with similar brake release, steering, and throttle timing while carrying equal or better speed. If the lap time improves only because one straight gained from weather, draft, or traffic, you have not proven the aero. If the minimum speed increases but steering correction increases sharply, you may have gained load but narrowed the control margin. If top speed falls and corner speed barely moves, you may have bought too much drag for too little useful downforce.
Do not overread one channel. The McBeath simulation discussion makes clear that performance is integrated from many influences: car mass, roll stiffness, dimensions, power, torque, gearing, braking, tires, and aero. On a real car, that means a speed trace alone does not identify aero. A driver who braked later, a tire that came in, a gust, or a better exit from the previous corner can change the same trace. Your evidence gets stronger when several signs agree: the car feels more repeatable, the same phase improves, top speed cost is understood, and the change appears in more than one comparable run.
The correct mindset
The source material is full of practical caution. McBeath emphasizes that professionals use CFD and wind tunnels, but also that the amateur has practical tools. It also warns that it is difficult to generalise in competition-car aerodynamics, that what works on one car may not work on another apparently similar car, and that trial and error are essential parts of development. That is not permission to guess forever. It is a warning to define the window on your car, with your evidence, before chasing someone else's downforce number.
This lesson should make you slower to add parts and faster to ask better questions. If the car is inside its window but lacks load in one speed range, an aero increase may be rational. If the car is outside its window because the floor, wing, or platform becomes inconsistent with speed, more downforce may make the problem sharper. If the car is mechanically unbalanced below the speed where aero dominates, aero tuning may be the wrong first job. If the car is quick in clean air and weak in traffic, the next window test should include following conditions rather than another clean-air wing sweep.
The phrase before you chase more downforce is not anti-aero. It is pro-performance. Downforce is valuable when it arrives at the right axle, in the right phase, through a wide enough range, with an acceptable drag cost, and with a platform that lets the tires use it. Define that range first. Then the next aero change has a target, a test, and a reason.
Worked example: sports prototype at small yaw angle
Imagine you are testing a prototype-style car that looks excellent in straight-line running. The car is fast at the end of the straight, stable under the first hit of the brake pedal, and the driver reports that it feels planted when the wheel is nearly straight. If you stop there, you might conclude that the car can take more downforce or that its current aero balance is solved.
McBeath's sports-prototype yaw discussion gives you the reason to keep going. Straight-ahead testing is good for basic mapping, drag optimization, and the braking phase from high speed in a straight line, especially because you need to understand downforce distribution and how it shifts during that phase. But more lap time is spent in corners, and cornering means the car is not simply straight to the air. At small yaw angles, the aerodynamic performance can differ from the straight-ahead condition.
The window definition for this car has to include at least two tests. First, confirm straight-line braking behavior: does the car stay stable and repeatable as speed falls and downforce decays? Second, confirm cornering attitude: does the same balance remain usable when the car is yawed at entry and mid-corner? If the car is calm in straight-line braking but loses front support as it takes yaw, the safe window is not high-speed aero in general. It is high-speed straight-line aero only. If the rear becomes inconsistent at yaw, adding rear wing may not fix the core problem; it may only increase load in the condition that was already good.
A good driver note after this test would be specific. It would not say the prototype needs more aero. It would say the straight-line braking window is acceptable, the mid-corner yaw window is unconfirmed or narrow, and the next test should target the condition where the car is actually losing repeatability. That note protects you from tuning the wrong phase.
Worked example: club racer balancing drag and useful load
Now imagine a club racer adds wing angle because the car feels light in a fast corner. The first run feels better. The driver can commit earlier, and the car takes less steering correction. That is real information, but it is not the whole window.
The McBeath performance-simulation material treats downforce and drag together because both affect lap time. The tire model, power and torque curves, gearing, braking model, and aero model all feed into performance. In driver terms, the added wing may have helped the high-speed corner while costing speed on the straight that follows. If the corner is long and important, the trade may be worth it. If the straight is long and the corner-speed gain is small, the change may make the car feel better while making the lap slower.
The operating-window method prevents you from stopping at feel alone. You would compare the fast corner entry, mid-corner speed, exit quality, and top speed on the following straight. You would also ask whether the gain appears in repeatable laps or just in the first confident lap after the change. If the car gained meaningful mid-corner speed and the straight-line loss is small, the window may have expanded. If the car gained confidence but lost too much acceleration or top speed, the useful window may be narrower than the sensation suggests.
This example is also where mechanical grip matters. Smith's warning about mechanical grip being the basis of cornering power should keep you from blaming aero for every fast-corner problem. If the car is below the speed range where aero is dominant, if the platform is not linear, or if the tires are not being used well, wing angle can become a distraction. The operating window should separate the fast-aero phase from the slower mechanical phase so that you do not tune one with the other.
Worked example: following another car in disturbed air
A clean-air setup can be a poor race setup if it gives away too much predictability when you follow. McBeath's case-study material states that competition cars running close together interact aerodynamically, and it describes the following driver mitigating downforce and grip losses by offsetting laterally from the car ahead. For the tested configuration, offsetting left also helped the left-hand-drive driver see.
Do not turn that into a universal rule that every car should always follow left. The source itself limits the conclusion to cars of the same overall configuration as the ones tested. The lesson is broader and more useful: traffic is part of the aero window. If you define the car only in solo clean-air laps, you may have no idea what it will do in the first race stint, the restart, or the closing laps behind a slower car.
For an intermediate driver, the test is simple. In a controlled session where it is safe and appropriate, note how the car behaves at the same corner when you approach in clean air, directly behind another car, and offset laterally. You are not looking for a passing hero move. You are looking for steering correction, brake confidence, front wash, rear security, and whether the car's balance changes in a way you can anticipate. If clean air is good but close following requires a different margin, write that into the window. Then your race craft and setup decisions can reflect the real car.
Common mistakes
The peak-number trap is the most common mistake. You judge an aero change by the largest downforce number or the strongest seat-of-the-pants moment, instead of by the range where the car remains repeatable. Good looks like a phase-specific window: where the load helps, where drag costs, where balance shifts, and where flow or platform behavior becomes uncertain.
The straight-line-only trap is next. You test on a straight, like the stability, and assume the car will be good in the corner. The corpus supports straight-ahead mapping for basic work and braking analysis, but it also supports yaw testing because cornering conditions can change aerodynamic performance. Good looks like separating straight-line braking from yawed entry and mid-corner behavior.
The mechanical-blame reversal is when you use aero to chase a problem that is mostly mechanical. Smith's tire and suspension framing matters here. The tires are the contact patches, mechanical grip is still the basis, and aero download adds to it. Good looks like asking whether the problem occurs in a speed range where aero should be decisive, and whether the mechanical platform is linear enough to let the aero work.
The drag blindness error is when you feel more planted and ignore the speed you gave away. McBeath's performance analysis treats downforce and drag as a compromise because lap time depends on both. Good looks like comparing the phase that improved against the straight or acceleration zone that paid for it.
The uncalibrated-data error is trusting a logger channel just because it has a number. The data-logging chunk emphasizes installation, calibration, and strategies for extracting useful information. Good looks like using data only when it is calibrated enough to answer the question, and pairing it with driver feel and repeatability.
The traffic omission is defining the car only in clean air. The corpus describes close-car aerodynamic interaction as unavoidable in racing. Good looks like explicitly writing clean-air and following-air behavior into the operating window, even if the first version is only a driver observation.
The universalizing error is copying another car's aero answer. McBeath cautions that competition-car aerodynamics are difficult to generalise and that what works on one apparently similar car may not work on another. Good looks like treating other cars as ideas to test, not conclusions to import.
Drill: one-event aero window map
Run this drill over three sessions at your next test day or race practice, using whatever data and observation tools are realistic for your car. The goal is not to find the perfect setup. The goal is to leave with a written operating window that is better than your starting opinion.
Session 1 is the baseline map. Run the car in its current configuration. For five clean laps, write notes by phase: straight-line acceleration, high-speed braking, turn-in, mid-corner, exit, and traffic if it occurs naturally. The success criterion is a phase-specific baseline, not a lap-time judgment. If you cannot say where the car was stable and where it changed balance, the session is not complete.
Between sessions, choose one aero question. Do not choose three. A useful question might be whether a wing setting helps a fast corner enough to justify drag, whether the front support remains usable under braking release, or whether the car behaves differently in yaw than it did straight ahead. If you have airflow visualization or calibrated ride-height and speed data, aim it at that question.
Session 2 is the controlled change or controlled confirmation. If conditions allow, change one meaningful aero variable. If conditions are not stable enough for a change, repeat the baseline and improve the evidence. Run another five clean laps and compare the same phases. The success criterion is identifying which phase moved and whether the movement was repeatable.
Session 3 is the window check. Return toward baseline or hold the better setting and test the boundary that remains uncertain. If the car improved in clean air, look for a safe opportunity to observe following behavior. If it improved in straight-line braking, pay attention to turn-in and mid-corner yaw. If it felt stronger but lost straight speed, decide whether the corner gain paid for the drag. The success criterion is a written statement with three parts: confirmed operating range, uncertain range, and do-not-chase-yet range.
A finished drill note might read like this: current setup is repeatable in clean high-speed braking, front support improves with the change, top-speed cost is visible, mid-corner yaw behavior still needs confirmation, and no further downforce should be added until the yaw and drag boundaries are understood. That note is not fancy. It is useful. It tells you what to test next and what not to pretend you know.
When this principle breaks down
The principle does not mean you must finish a professional aero map before making any change. Club racers often have limited track time, incomplete data, and no wind tunnel. McBeath explicitly frames practical tools for racers who do not have professional resources, and the book's broader guidance is to use the tools carefully and with common sense. If a car is unsafe or obviously unstable, you may need to make a conservative change before your evidence is complete.
The principle also does not mean every window must be wide. Some competition cars are narrow-window tools by design. A high-downforce car may demand precise platform control and a driver who respects its speed range. The question is whether you know the range and can drive inside it. A narrow known window is different from a narrow unknown window.
Finally, the principle does not make aero development clean or final. The corpus is honest that trial and error are essential in motorsport aero and that what works on one car may not work on another. Your operating window is a living description. Each test should either expand it, sharpen it, or reveal that a tempting change is not yet supported. That is development. Chasing more downforce without the window is guessing.
Author Review
No quiz questions are attached to this lesson.
Sources
| # | Document | Chunk | Pages | Score | Collection |
|---|---|---|---|---|---|
| 1 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 893cce66-5e94-8af0-6d98-00acc7cbd324 | 383 | 1 | uio_books_raw_v1 |
| 2 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 90b5a640-d9b2-b0ef-2f6e-f9a0dadce5aa | 411 | 1 | uio_books_raw_v1 |
| 3 | Racing Chassis and Suspension Design Carroll Smith | 148524fa-62af-201e-6dff-3b729c84477a | 8 | 1 | uio_books_raw_v1 |
| 4 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 2abb3a1a-1abc-3549-8f79-9fce704061d6 | 334 | 1 | uio_books_raw_v1 |
| 5 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 6edca499-2988-7702-ccc8-3d17b516edff | 385 | 1 | uio_books_raw_v1 |
| 6 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 9f0edfc1-9e8c-3a96-a48d-b0d658513db3 | 385 | 1 | uio_books_raw_v1 |
| 7 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 17fd5a9b-5fdf-ead1-ff69-572014594b23 | 477 | 1 | uio_books_raw_v1 |
| 8 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 9e3001fd-e626-5565-9b11-bc3cab151d27 | 281 | 1 | uio_books_raw_v1 |
| 9 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 61068e74-0999-1e25-03bd-8c545f352d25 | 26 | 1 | uio_books_raw_v1 |
| 10 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 10acd525-ae45-7603-2847-9b1b9db65585 | 9 | 1 | uio_books_raw_v1 |
| 11 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 43f9ecd8-7336-a0ec-07a9-5149279141e4 | 43 | 1 | uio_books_raw_v1 |
| 12 | Competition Car Aerodynamics 3rd Edition McBeath Simon | 9a496275-f006-9cdc-8647-b7acc6459056 | 42 | 1 | uio_books_raw_v1 |
| 13 | Competition Car Aerodynamics 3rd Edition McBeath Simon | cd94958f-1042-ceff-8d99-06fa06ac633b | 504 | 1 | uio_books_raw_v1 |
| 14 | Competition Car Aerodynamics 3rd Edition McBeath Simon | d788f877-dfdc-2c41-96e0-e6a0de38e907 | 412 | 1 | uio_books_raw_v1 |