At What Speed Can an F1 Car Stay Upside Down?
Do you often drive F1 cars upside down in tunnels?
If so, bad news: in 2026, youβll need +11.7 km/h for the carβs downforce (upforce?) to match its weight!
2026 cars are 20kg lighter (-2.5%) than 2025s, but have ~13% less downforce. The latter prevails: the speed needed to equate its weight increased from 199.7 km/h to 211.4 km/h β slower in high-speed corners, lower g-forces.
In reality, 211.4 km/h would only let you drive upside down very briefly: at that speed, downforce=weight βNo normal force and thus traction! Drag would slow the car down β downforce drops β you fall!
In order not to fall, you would need enough downforce to produce enough rear longitudinal force to overcome drag.
[CFD analysis via @Air_Shaper]
Do you often drive F1 cars upside down in tunnels?
If so, bad news: in 2026, youβll need +11.7 km/h for the carβs downforce (upforce?) to match its weight!
2026 cars are 20kg lighter (-2.5%) than 2025s, but have ~13% less downforce. The latter prevails: the speed needed to equate its weight increased from 199.7 km/h to 211.4 km/h β slower in high-speed corners, lower g-forces.
In reality, 211.4 km/h would only let you drive upside down very briefly: at that speed, downforce=weight βNo normal force and thus traction! Drag would slow the car down β downforce drops β you fall!
In order not to fall, you would need enough downforce to produce enough rear longitudinal force to overcome drag.
[CFD analysis via @Air_Shaper]
β€12π3π₯°2
π KEY SUZUKA NOTES
π₯Extreme cornering speeds and tyre stress:
- Ferrari's superior chassis/aero will help in quali, BUT...
- ...High wear will hurt them in the race.
- Crucial to ace per-corner ERS deployment/harvest to save energy (and tyres)... meaning that quali won't be flat out as usual :(
Very PU-demanding track, like Australia: Mercedes' ICE power advantage will matter more than in China.
Crucial indicator of each team's performance for the rest of the season... or at least before the technical directives and big updates.
πΈ @pirellisport
π @GaborWeber
π₯Extreme cornering speeds and tyre stress:
- Ferrari's superior chassis/aero will help in quali, BUT...
- ...High wear will hurt them in the race.
- Crucial to ace per-corner ERS deployment/harvest to save energy (and tyres)... meaning that quali won't be flat out as usual :(
Very PU-demanding track, like Australia: Mercedes' ICE power advantage will matter more than in China.
Crucial indicator of each team's performance for the rest of the season... or at least before the technical directives and big updates.
πΈ @pirellisport
π @GaborWeber
π11β€5π₯2
π JAPANESE GP - PRACTICE 2 | LONG RUNS
π’ Mercedes is looking untouchable: quick and consistent.
π΄ LEC was next best, yet over 0.6s/lap slower!
π PIA set the quickest lap, but long run didn't impress.
π£ Alpine and βͺοΈ Haas are looking good again!
π΅ VER again nowhere. HAD was as quick as him once slow laps are removed!
Pace per team:
1) Merc/ANT (M)
2) Ferrari/LEC (M) +0.662 s/lap;
3) McL/PIA (M) +1.144 s/lap;
4) Alpine/GAS (H) +1.301 s/lap;
5) Haas/OCO (M) +1.410 s/lap;
6) RBR/VER (H) +1.493 s/lap;
7) Audi/HUL (H) +1.550 s/lap;
8) RBs/LAW (M) +2.050 s/lap;
9) Williams/ALB (M) +2.468 s/lap;
10) Cadillac/BOT (M) +3.205 s/lap.
π’ Mercedes is looking untouchable: quick and consistent.
π΄ LEC was next best, yet over 0.6s/lap slower!
π PIA set the quickest lap, but long run didn't impress.
π£ Alpine and βͺοΈ Haas are looking good again!
π΅ VER again nowhere. HAD was as quick as him once slow laps are removed!
Pace per team:
1) Merc/ANT (M)
2) Ferrari/LEC (M) +0.662 s/lap;
3) McL/PIA (M) +1.144 s/lap;
4) Alpine/GAS (H) +1.301 s/lap;
5) Haas/OCO (M) +1.410 s/lap;
6) RBR/VER (H) +1.493 s/lap;
7) Audi/HUL (H) +1.550 s/lap;
8) RBs/LAW (M) +2.050 s/lap;
9) Williams/ALB (M) +2.468 s/lap;
10) Cadillac/BOT (M) +3.205 s/lap.
β€16π₯2π1
ππͺ« JAPANESE GP - QUALIFYING | CLIPPING ANALYSIS
(130R CORNER)
+ TOP SPEED & MINIMUM SPEED
π₯ Ferrari had the WORST top speed: just 320km/h. Ferrari-powered teams were slowest on average (worse than Aston-Honda even)!
π© Honda had a massive 60km/h clipping: just 267 km/h while still full-throttle! (worst ICEβ½οΈ?)
Mercedes-powered πͺ Alpine was fastest overall: 333 km/h top speed, like RBs, but better minimum speed.
- Best top speeds: RBPT PU (strong deployment).
- Best minimum speeds: Mercedes and Audi PU (strong ICE).
Clipping was so absurdly high during and after the 130R corner because the low-drag mode was not allowed during and after that corner.
Before the corner: ERS deployment + low drag mode.
During and after: ERS harvesting + high drag mode.
This wasn't the case in Australia or China.
[Made via @JMP_software]
(130R CORNER)
+ TOP SPEED & MINIMUM SPEED
π₯ Ferrari had the WORST top speed: just 320km/h. Ferrari-powered teams were slowest on average (worse than Aston-Honda even)!
π© Honda had a massive 60km/h clipping: just 267 km/h while still full-throttle! (worst ICEβ½οΈ?)
Mercedes-powered πͺ Alpine was fastest overall: 333 km/h top speed, like RBs, but better minimum speed.
- Best top speeds: RBPT PU (strong deployment).
- Best minimum speeds: Mercedes and Audi PU (strong ICE).
Clipping was so absurdly high during and after the 130R corner because the low-drag mode was not allowed during and after that corner.
Before the corner: ERS deployment + low drag mode.
During and after: ERS harvesting + high drag mode.
This wasn't the case in Australia or China.
[Made via @JMP_software]
β€11π₯1
π JAPANESE GP - RACE | RACE PACE
1) Merc
2) McL +0.24 s/lap;
3) Ferrari +0.32 s/lap;
4) Alpine +0.97 s/lap;
5) RBR +1.02 s/lap;
6) Haas +1.59 s/lap;
7) Audi +1.71 s/lap;
8) RBs +1.74 s/lap;
9) Williams +2.13 s/lap;
10) Cadillac +3.14 s/lap;
11) Aston +3.99 s/lap.
My take on the race:
- ANT could be the most likely WDC. 2 wins in 3 races, already matching RUS's best season! Must improve his starts, though.
- McL is back, having achieved peak PU performance (either through testing or Mercedes assistance).
- Ferrari, and LEC in particular, maximised their points once again. Performance will come.
- GAS beat VER thanks to pace and better straight-line performance.
- Regulations work better for races than for quali
- COL was going NINETY-TWO km/h slower than BEA: something must be done about these regulations, creating dangerous situations.
Rate the race from 1 to 10!
1) Merc
2) McL +0.24 s/lap;
3) Ferrari +0.32 s/lap;
4) Alpine +0.97 s/lap;
5) RBR +1.02 s/lap;
6) Haas +1.59 s/lap;
7) Audi +1.71 s/lap;
8) RBs +1.74 s/lap;
9) Williams +2.13 s/lap;
10) Cadillac +3.14 s/lap;
11) Aston +3.99 s/lap.
My take on the race:
- ANT could be the most likely WDC. 2 wins in 3 races, already matching RUS's best season! Must improve his starts, though.
- McL is back, having achieved peak PU performance (either through testing or Mercedes assistance).
- Ferrari, and LEC in particular, maximised their points once again. Performance will come.
- GAS beat VER thanks to pace and better straight-line performance.
- Regulations work better for races than for quali
- COL was going NINETY-TWO km/h slower than BEA: something must be done about these regulations, creating dangerous situations.
Rate the race from 1 to 10!
β€8π4
Formula Data Analysis
π JAPANESE GP - RACE | RACE PACE 1) Merc 2) McL +0.24 s/lap; 3) Ferrari +0.32 s/lap; 4) Alpine +0.97 s/lap; 5) RBR +1.02 s/lap; 6) Haas +1.59 s/lap; 7) Audi +1.71 s/lap; 8) RBs +1.74 s/lap; 9) Williams +2.13 s/lap; 10) Cadillac +3.14 s/lap; 11) Aston +3.99β¦
COL vs BEA images by @SamF1__
π13β€1
π JAPANESE GP - RACE | Hamilton's Power Issue
Hamilton was right: he had NO POWER to defend from LEC and RUS! πͺ«
LEC reached 317 km/h without slipstream or overtake mode, HAM just 307 (Lap 42).
And that's nothing: LEC reached 330 when in HAM's slipstream, but the latter didn't exceed 306 when the roles inverted!
(Lap 43, and RUS overtook him.)
Behind HAM, LEC gained 0.391s on the main straight.
Behind LEC, HAM LOST 0.250s (while RUS gained 0.529s).
So:
- For some reason (bug?) HAM didn't have the ERS power needed to defend. β
- RUS could overtake HAM more easily than LEC did... even though he was 0.144s further behind on the straight! π
Hamilton was right: he had NO POWER to defend from LEC and RUS! πͺ«
LEC reached 317 km/h without slipstream or overtake mode, HAM just 307 (Lap 42).
And that's nothing: LEC reached 330 when in HAM's slipstream, but the latter didn't exceed 306 when the roles inverted!
(Lap 43, and RUS overtook him.)
Behind HAM, LEC gained 0.391s on the main straight.
Behind LEC, HAM LOST 0.250s (while RUS gained 0.529s).
So:
- For some reason (bug?) HAM didn't have the ERS power needed to defend. β
- RUS could overtake HAM more easily than LEC did... even though he was 0.144s further behind on the straight! π
π14β€6π4
F1'S BIGGEST ISSUE: ERS BEHAVIOR AT FULL THROTTLE
At full throttle, ERS behaviour (deployment vs harvesting) dictates acceleration: the difference approaches 10 m/sΒ², like going from constant speed to stomping the brakes of a road car!
Max deployment: +350 kW
Max harvest at full throttle: β250 kW
Difference: 600 kW (~800 hp)
ΞAcceleration = ΞPower / ( Mass Γ Speed )β 9.2 m/sΒ²
(Confirmed by the data!)
The driver stays flat out, yet total power swings by 800 hp (1000 hp β 200 hp, or Bugatti Veyron vs Polo GTI).
And it gets worse: This #F1 data... is from HAM's best QUALIFYING lap!
My solution:
For 2026:
- Limit ERS deployment from 350kW to 250kW (53%/47%β62%/38% ICE/ERS power split). π
- Limit the max harvest at full throttle to -150kW (keep the 350kW harvest under braking).
For 2027:
- Increase fuel flow by 20%, reaching ~1000hp again with 66%/34% split.
For the future: cheaper naturally aspirated 3000cc V8 engines on biofuel + rider-managed KERS. π₯
Made via @JMP_software
What's YOUR SOLUTION? π
At full throttle, ERS behaviour (deployment vs harvesting) dictates acceleration: the difference approaches 10 m/sΒ², like going from constant speed to stomping the brakes of a road car!
Max deployment: +350 kW
Max harvest at full throttle: β250 kW
Difference: 600 kW (~800 hp)
ΞAcceleration = ΞPower / ( Mass Γ Speed )β 9.2 m/sΒ²
(Confirmed by the data!)
The driver stays flat out, yet total power swings by 800 hp (1000 hp β 200 hp, or Bugatti Veyron vs Polo GTI).
And it gets worse: This #F1 data... is from HAM's best QUALIFYING lap!
My solution:
For 2026:
- Limit ERS deployment from 350kW to 250kW (53%/47%β62%/38% ICE/ERS power split). π
- Limit the max harvest at full throttle to -150kW (keep the 350kW harvest under braking).
For 2027:
- Increase fuel flow by 20%, reaching ~1000hp again with 66%/34% split.
For the future: cheaper naturally aspirated 3000cc V8 engines on biofuel + rider-managed KERS. π₯
Made via @JMP_software
What's YOUR SOLUTION? π
β€20π3π3β2
I've estimated the ICE (Internal Combustion Engine) power for each manufacturer: big trouble for Red Bull, while Ferrari can be optimistic! π
Mercedes: ~576 hp (best)!
Teams with β€564 hp (>2% deficit) earn one extra upgrade token per year. RBPT sits at ~565 hp: they may miss the cut!
Their PU isn't the problem; the chassis is, which is why Racing Bulls has matched them!
Teams with β€552 hp (>4% deficit) earn TWO extra upgrades per year. Ferrari, Audi, and Honda should all qualify!
Ferrari's gap to Mercedes (~29hp) is huge, yet their best-in-class chassis allows them to fight!
Which team will have the best in-season development? π€
Mercedes: ~576 hp (best)!
Teams with β€564 hp (>2% deficit) earn one extra upgrade token per year. RBPT sits at ~565 hp: they may miss the cut!
Their PU isn't the problem; the chassis is, which is why Racing Bulls has matched them!
Teams with β€552 hp (>4% deficit) earn TWO extra upgrades per year. Ferrari, Audi, and Honda should all qualify!
Ferrari's gap to Mercedes (~29hp) is huge, yet their best-in-class chassis allows them to fight!
Which team will have the best in-season development? π€
π₯11π6β€5
The 1953 British Racing Motor #F1 engine was more powerful than the best Internal Combustion Engine on the 2026 grid! π€―
The fuel-flow meter limits power excessively: no wonder it cannot produce enough energy to make ERS deployment sustainable! πͺ«
A 10 to 20% fuel-flow increase, coupled with an equal reduction in ERS peak power, would:
π - Limit clipping substantially;
π - Improve laptimes (same peak power, higher average power) and top speeds;
π β - Avoid major engine redesigns.
A no-brainer, in my opinion...
The fuel-flow meter limits power excessively: no wonder it cannot produce enough energy to make ERS deployment sustainable! πͺ«
A 10 to 20% fuel-flow increase, coupled with an equal reduction in ERS peak power, would:
π - Limit clipping substantially;
π - Improve laptimes (same peak power, higher average power) and top speeds;
π β - Avoid major engine redesigns.
A no-brainer, in my opinion...
β€18π1π€―1
Formula Data Analysis
The 1953 British Racing Motor #F1 engine was more powerful than the best Internal Combustion Engine on the 2026 grid! π€― The fuel-flow meter limits power excessively: no wonder it cannot produce enough energy to make ERS deployment sustainable! πͺ« A 10 toβ¦
Behold the beauty of the mighty V16 1496cc of the BRM Type 15: 16 tiny cilinders of just 93.5cc each! π
β€18π₯°5
NEW F1 ERS RULES (From Miami), EXPLAINED! π
Clipping time drops significantly, with minimal impact on lap times and the pecking order. At peak harvest, an F1 car can produce less full-throttle power than a Fiat Panda, but it's for the better π (read to the end!)
QUALI
β’ Max recharge: 8 β 7 MJ
β’ Peak superclip power: 250 β 350 kW
β’ Tracks with alternative lower energy limits: 8 β 12
RACE
β’ Boost gain capped at +150 kW
β’ ERS power not on straights: 350 β 250 kW
β’ Peak superclip power: 250 β 350 kW
Effects:
β’ Less energy to deploy (7 MJ in quali). Stronger superclipping (250β350 kW) means it lasts even less.
β’ Boost mode nerfed, but still enough to overtake.
β’ Laptimes roughly unchanged: stronger superclip (β perf) offset by less deployment not on straights (β perf).
Saved the best for last: during superclipping, the PU produces ~400 kW (ICE) β 350 kW (ERS harvest) = ~50 kW (68 hp) at full throttle. That's 1 kW less than a Fiat Panda 1.2 (51 kW)! πΌ
Clipping time drops significantly, with minimal impact on lap times and the pecking order. At peak harvest, an F1 car can produce less full-throttle power than a Fiat Panda, but it's for the better π (read to the end!)
QUALI
β’ Max recharge: 8 β 7 MJ
β’ Peak superclip power: 250 β 350 kW
β’ Tracks with alternative lower energy limits: 8 β 12
RACE
β’ Boost gain capped at +150 kW
β’ ERS power not on straights: 350 β 250 kW
β’ Peak superclip power: 250 β 350 kW
Effects:
β’ Less energy to deploy (7 MJ in quali). Stronger superclipping (250β350 kW) means it lasts even less.
β’ Boost mode nerfed, but still enough to overtake.
β’ Laptimes roughly unchanged: stronger superclip (β perf) offset by less deployment not on straights (β perf).
Saved the best for last: during superclipping, the PU produces ~400 kW (ICE) β 350 kW (ERS harvest) = ~50 kW (68 hp) at full throttle. That's 1 kW less than a Fiat Panda 1.2 (51 kW)! πΌ
β€13π3
The most powerful naturally-aspirated #F1 engine ever was... Japanese, likely π―π΅
Peak power reached in 2005 (end of 3000cc, 'V10' era)!
Renault ~900hp
Cosworth 915hp
Mercedes ~930hp
Ferrari 920-940hp
Honda >965hp (Last power step in Suzuka)
Toyota >1000hp, detuned to ~960hp for reliability
The straight-line acceleration of these cars was insane, much higher than the current ones' (and did so 20 years ago)!
Compared to current cars:
- Similar peak power
- Higher end-of-straight power (No ERS, always full power)
- 200kg lighter (600kg, including driver!)
- 20cm narrower track, reducing drag
Crazy!
Peak power reached in 2005 (end of 3000cc, 'V10' era)!
Renault ~900hp
Cosworth 915hp
Mercedes ~930hp
Ferrari 920-940hp
Honda >965hp (Last power step in Suzuka)
Toyota >1000hp, detuned to ~960hp for reliability
The straight-line acceleration of these cars was insane, much higher than the current ones' (and did so 20 years ago)!
Compared to current cars:
- Similar peak power
- Higher end-of-straight power (No ERS, always full power)
- 200kg lighter (600kg, including driver!)
- 20cm narrower track, reducing drag
Crazy!
β€14π₯5π2π1
The most powerful F1 engine ever?
The tiny 1500cc 4-cylinder 1986 BMW M12: over 1400hp in qualifying!π₯
Power is estimated: no dyno could accurately measure over 1000hp!π‘
The Benetton reached 352km/h in Monza, with much worse drag (and grip) than current cars!
Mind-blowing! π³
The inline-4 layout of the BMW engine gave it a clear advantage over the V6 engines used by Ferrari and Renault:
1 fewer turbo, 2 fewer cylinders, 8 fewer valves
β‘οΈlower frictional losses
β‘οΈmore power and better reliability.
The tiny 1500cc 4-cylinder 1986 BMW M12: over 1400hp in qualifying!π₯
Power is estimated: no dyno could accurately measure over 1000hp!π‘
The Benetton reached 352km/h in Monza, with much worse drag (and grip) than current cars!
Mind-blowing! π³
The inline-4 layout of the BMW engine gave it a clear advantage over the V6 engines used by Ferrari and Renault:
1 fewer turbo, 2 fewer cylinders, 8 fewer valves
β‘οΈlower frictional losses
β‘οΈmore power and better reliability.
β€20π₯6
π MIAMI GP - SPRINT QUALIFYING| SESSION OVERVIEW
Key F1 SPRINT QUALI notes:
β’ π₯ LEC quickest in S1, β¬οΈ ANT in S2, β¬οΈ RUS in S3.
β’ π§ McL (NOR) got pole with no purple sectors!
β’ The slowest F1 car (π© Aston) was slower than the SLOWEST F2 (Shields)!
β’ π₯ LEC missed power on SQ3 (-6km/h vs SQ2, 0.3s lost vs others!)
β’ Start of 4-teams title fight? π
Aero Performance
Best Sector Gap:
Mercedes and RBR struggled in S1, Ferrari in S2 and S3 (but much less so).
NOR vs ANT Telemetry:
Notice McL's much stronger deployment before T1 and after T3, which, however, limited their power on the two main straights!
McL, Ferrari, and Mercedes set the best sectors, so they have the most potential!
Ferrari performed worse than they could have, while NOR took the most out of that opportunity!
Red Bull looked better than in any other session this year, probably: excellent top speed and straightline performance, yet no evident chassis-side issues!
All-around improved car, if this trend gets confirmed.
Key F1 SPRINT QUALI notes:
β’ π₯ LEC quickest in S1, β¬οΈ ANT in S2, β¬οΈ RUS in S3.
β’ π§ McL (NOR) got pole with no purple sectors!
β’ The slowest F1 car (π© Aston) was slower than the SLOWEST F2 (Shields)!
β’ π₯ LEC missed power on SQ3 (-6km/h vs SQ2, 0.3s lost vs others!)
β’ Start of 4-teams title fight? π
Aero Performance
Best Sector Gap:
Mercedes and RBR struggled in S1, Ferrari in S2 and S3 (but much less so).
NOR vs ANT Telemetry:
Notice McL's much stronger deployment before T1 and after T3, which, however, limited their power on the two main straights!
McL, Ferrari, and Mercedes set the best sectors, so they have the most potential!
Ferrari performed worse than they could have, while NOR took the most out of that opportunity!
Red Bull looked better than in any other session this year, probably: excellent top speed and straightline performance, yet no evident chassis-side issues!
All-around improved car, if this trend gets confirmed.
β€12π₯3
π MIAMI GP - SPRINT | SPRINT PACE
βThe break was so long we got back to last yearβs Miami Sprint!β
The updated π§ McLaren returned to race-winning pace, while π¦ RBR is in the mix again.
π© Mercedes' pace was only the THIRD quickest (and RUS was no quicker than VER).
π₯ Ferrari can be optimistic!
Pace:
1) McL 0.00
2) Ferrari +0.24
3) Merc +0.35
4) RBR +0.52
5) Alpine +1.39
6) Williams (M/M) +2.14
7) Audi +2.30
8) Haas +2.68
9) RBs +2.79
10) Aston +3.55
11) Cadillac (H) +3.72
What's YOUR prediction for the F1 race? π
βThe break was so long we got back to last yearβs Miami Sprint!β
The updated π§ McLaren returned to race-winning pace, while π¦ RBR is in the mix again.
π© Mercedes' pace was only the THIRD quickest (and RUS was no quicker than VER).
π₯ Ferrari can be optimistic!
Pace:
1) McL 0.00
2) Ferrari +0.24
3) Merc +0.35
4) RBR +0.52
5) Alpine +1.39
6) Williams (M/M) +2.14
7) Audi +2.30
8) Haas +2.68
9) RBs +2.79
10) Aston +3.55
11) Cadillac (H) +3.72
What's YOUR prediction for the F1 race? π
π₯5β€2
π MIAMI GP - RACE | RACE PACE
BEST PACE: ANT, NOR β
Race decided by ANT's 1-lap earlier pit and clean overtake, not by differences in pace.
FAST PACE: PIA, LEC β
LEC almost matched PIA before the issue (analysis coming tomorrow)!
SLOWER: RUS, HAM, VER β
HAM had extensive bodywork damage.
VER did 51 laps on a Hard set!!
RUS had no excuses, while ANT carried the team!
BEST PACE: ANT, NOR β
Race decided by ANT's 1-lap earlier pit and clean overtake, not by differences in pace.
FAST PACE: PIA, LEC β
LEC almost matched PIA before the issue (analysis coming tomorrow)!
SLOWER: RUS, HAM, VER β
HAM had extensive bodywork damage.
VER did 51 laps on a Hard set!!
RUS had no excuses, while ANT carried the team!
β€15