F1 Engine Designer AMA

Proof where

What are your thoughts on formula-e?

at least the udder is aerodynamic

Pushrods or OHC

Will McLaren-Honda ever recover?

which F1 engines have you worked on

Why are you on /o/ instead of making your shitty honda powerplant work?

>>17025691

No specifics I'm afraid, Pretty much everything I work on is covered by NDA

>>17025693

Apart from the millenial-baiting social media gimmicks that they use I am actually a big fan, I can see it converging with F1 over the next 10 years

>>17025702

Depends on the application, I've worked mostly on DOHC as you won't find many pushrods in many series outside of Nascar.

>>17025706

Honestly? I think the rate of development of the other 3 manufacturers will keep them permanently 1-2 seasons behind in the development race. I could write a whole thread on why they aren't catching up but Honda's whole organisation is screwed by having their engineers in Japan and their test facilities in Britain so it takes forever to develop anything.

>>17025715

Like I said in the last post, their whole company just isn't set up in the same way as the rest of the F1 engine companies, they see it as an R&D excercise for their young homegrown Japanese engineers. If I were them I'd move their whole operation to Europe where the actually experienced F1 engineers are based and steal their knowledge.

>>17025686
Favorite engine layout?

Least favorite?

>>17025686
look at that downforce

this cow is a mountain

Good books on engine design/layout/tuning?

>>17025686

i used, and referenced, that image in my CFD assignment.

File: rudolf_at_mach_30.png (768KB, 1449x912px) Image search: [Google]

768KB, 1449x912px

>>17025856

Ineterestingly the current F1 IC Engine is a 90 degree V6 and despite all the hate this layout gets on here it's quite possibly the best engine in the world right now,

>600bhp/L
6,000 mile life (actually a lot for a race engine)
>50% thermal efficiency

My absolute worst engine I ever worked on designing was an I6 for a luxury automotive client, so go figure.

File: 1486236131524.jpg (41KB, 483x376px) Image search: [Google]

41KB, 483x376px

>>17025686
Corvette or GTR

and how do I become you

>>17025686

How are the inlet ports and combustion chambers in F1 engines finished? Any super secret coatings or novel surface treatments in use?

>>17025889

Are you looking for something technical or for the layperson? If you want something as a starting point Carroll Smith's books are all excellent. If you are looking for something in-depth then pic related is my bible

>>17025896

No shit, my college CFD assignment was on the aerodynamics of race horses, I'll see if I can dig out some of my images.

>>17025979
As technical as possible. I wanna really learn about the stuff

>>17025686
Long rod or short rod engine

File: Smith.jpg (22KB, 259x400px) Image search: [Google]

22KB, 259x400px

>>17025889

Sorry forgot pic.

>>17025930

I don't really get this meme about which is faster, a 13 year old in a Formula Ford could trounce either round any race circuit worth talking about.

Oh and to become me you'd probably need to stab me and steal my identity card.... Oh wait to become an F1 engineer? Study a STEM subject to Uni or at least technical apprentice level, then get to your local race track and get to know the pro's running there. Motorsport's a small world.

>>17025932

Personally we don't, but I can't speak for the other manufacturers. We did try coating the inlet but it was to reduce inlet temperature rather than increase flow, and it didn't actually help. With highly boosted turbo's the exhaust port is much more performance critical as that is where the heat is scavenged for the turbine. Same reason for the chamber.

>>17026047
What's the rough effenciency percentage between a canted pushrod head and a DOHC modern head assuming cam profile specs are the same. tl;dr about how much hp per cubic inch is different head style worth when equal

File: rod.jpg (5KB, 259x194px) Image search: [Google]

5KB, 259x194px

>>17026037

Conrod length is pretty irrelevant in engine design, I've never worked on an engine where the rod length has been the driving design factor, it usually just ends up being that length as a result of the deck height and bore/stroke dimensions. The V8 F1 cars ran some hilariously short rods though these have lengthened for the V6.

>>17026079

What kind of efficiency are you talking about? Volumetric efficiency? Overall thermal efficiency of the engine? Because neither of those are directly controlled by how you actuate the valves. If both engines have the same cam and same port you'll get the same volumetric efficiency and same power. The advantage of DOHC over pushrod on an NA engine is that it will generally allow for more favourable port which will improve airflow. How much better completely depends on how good the engineer who designed the intake system is.

>>17026159
Thanks. Meant more efficiency per cubic inch. I probably should've asked how much power combustion chamber shape and spark plug placement is worth

Why aren't OHV with 4 or more valves per cylinder more common?

File: CX0005.jpg (36KB, 300x275px) Image search: [Google]

36KB, 300x275px

>>17026208

I only know of Honda doing this for the CX500, pic related. I can only assume that the extra cost and complexity required in the fancy rocker arms means that for multi-cylinder engines DOHC is just a neater solution.

>>17026208
Geometry gets tricky with running the pushrods through the water jackets and cylinder heads to get a decent angle on a 4 valve head. Also the lobes on the camshaft would have to be incredibly thin as well as the lifters and you would probably just run into a lot of valvetrain deflection at the rpm range where a 4v setup becomes beneficial

>>17025686
How would I tune a turbocharged motorcyle engine?

I.e. how do I detect knock. I don't have access to spark plugs with pressure ports/pressure transducers.

Straight cut gears

File: 1492249433297.jpg (43KB, 640x620px) Image search: [Google]

43KB, 640x620px

>>17026047
I'm going as a CS student is that good do you get paid well? btw how hard is it to be or become a racecar driver.

>>17025686
Would a racecar with a proper CVT (read: no bullshit "shifting" pauses) ever be viable? I think Williams made a prototype for one in the 90s but apart from that I've never heard anything else.

I mean in theory, keeping the engine at the optimum RPM constantly should make it pretty fast,right?

>>17025777
what would you do to design a diesel engine to be built for high speed racing? of course it won't hit F1 tier but how would you minmax that shit for maximum vroom?

>>17025686
how does the intake charge behave above 6000 rpm ?
its just that for most engines this is where torque starts to drop off

I think the air has less time to get into the engine
even at higher pressure there would be less volume I suppose

Bring back rosaries and gas turbines..

File: T head V8 Gear driven cams -.png (27KB, 839x567px) Image search: [Google]

27KB, 839x567px

32 valve cross flow flat head v8 running with 30 psi boost ?
just with an aim to have some power and reliability by eliminating the need for head gaskets

main problem is bulk and low compression ratio
that and valve heads can still shear off

>>17025686
whats the difference between petrol spec oil say 15-40 weight and diesel spec oil of the same weight? Is it just additives or is it more then that?

File: Red_Bull_X2010_'10.jpg (1MB, 1621x981px) Image search: [Google]

1MB, 1621x981px

How feasible is the Red Bull X2010? I know it was said it was technically possible but it kinda sounds like bullshit.

>>17028585
Oh it's possible.

But the amount of G's it would induce on the human body would make it impractical for anyone to operate for longer than 5 minutes before they pass out.

It'd basically give you a concussion after two laps at Spa 'cause your brain would be smooshing back and forth against the sides of your skull. It has limits that are higher than the human body essentially.

>>17028589
what if it was piloted remotely via VR/Simulator?

>>17025686
What engine configuration would you guys develop if the new F1 owners said:
>1.0L
>Two turbo's
>Max 2.5 bar of boost
>18K RPM
V8? V10? I4?

>>17026208
Because, thanks to their truck heritage, OHV engines are generally designed with tall strokes, so they don't need to go to high RPM to produce power. High RPM is where multivalve really shines. GM tried a 4v OHV conversion for the LS7, wasn't worth the added complexity.

>>17026258
Also Cummins 24v.

>>17027981
Yes, assuming it doesn't break and doesn't have massive power losses, it'd outperform a conventional transmission. We're a long way from both though, because CVT development in racing is nonexistent.

>>17027500
Not OP, but you can buy standalone knock detectors. Also, just keep it on a safe boost level (below 0.5 bar) on stock internals, since those are really high compression on most sportsbikes. You use factory timing, add a couple degrees as safety margin, and then you make sure your wideband lambda sensor is on a really safe (12.5 or 13 to 1) level.

Or you pay someone to do it, who then has to repair your bike if he blows it up.

>>17025686
Why aren't turbo 6-cylinders more common?
Not really in motorsports, just in general.

OP is dead. Long live the surrogate engineer.

>>17027983
If you want to see high performance diesels, go look at Le Mans. The recipe there is mainly dictated by the rulebook (as is any racing engine, really), but it boils down to:
>Biggest displacement you can get
>Most boost you can throw at it
>Most cilinders you can squeeze into your chassis
>Try and get the most RPM out of it
>Add a hybrid system, including MGU-H
More displacement and boost are logical steps, as is the hybrid system. What's wierd about diesels is that they're RPM limited not by valvetrain (like gasoline engines are), but by their flame front. Roughly speaking, the flame front dictates the maximum piston speed, so you want a relatively short stroke to be able to realise that high RPM. This means you need smaller cilinders - and if you want high displacement and small cilinders, that means you need a lot of cilinders. At the peak of diesel Le Mans cars, Peugeot and Audi were both rocking 5.5 litre twinturbo V12 diesels. After that, they got smaller and smaller.

>>17028011
It doesn't even get supersonic or anything fancy in a commuter car, it's just that most bottom ends (pistons, crank, rods) aren't designed for high-RPM use. In most engines, torque drops off around that mark, because we want it to. We limit the camshaft, and other parts of the system to make sure it doesn't significantly overpower everything, which would break our warranty expectations - and that gets expensive quick.

By limiting the camshaft the air has less time to get into the engine and less lift, too. For high-RPM running, you design camshafts (and the rest of the entire system) to breathe only at that high mark, which usually means you have to severely compromise the low-RPM portion. That sacrifice is not something we can make on a production car.

>>17028663
>Not OP, but you can buy standalone knock detectors.

They are 100% useless (too much noise due to straight cut gears).

>Also, just keep it on a safe boost level (below 0.5 bar) on stock internals, since those are really high compression on most sportsbikes.

I plan to use toluene (like 1/2...2/3 + 98AKI gasoline), this should help with detonation

> You use factory timing, add a couple degrees as safety margin, and then you make sure your wideband lambda sensor is on a really safe (12.5 or 13 to 1) level.

I plan to retard ignition like 1 degree per 2-3 psi.

However ignition timing vs IAT s still an open question for me.

AFR of 13 at full boost at maximum VE could be a disaster.

>>17025686

Any chance of F1 engines getting closer to the 14000/15000/whatever it is rev limit this year?

Do you guys have any sort of special valve tricknology like that Koenigsegg Freevalve shit?

How far ahead do you believe Mercedes is right now in the engine department?

>>17028053
No. Also, Google InMotion Le Mans.

>>17028129
Great idea, but practically stupid. Even a regular OHV engine is going to flow significantly better than a flathead, and you'd be better of just trying to make that unbreakable instead.

Also, why 32v? More valves = more cilinder head volume, so remove some valves and your CR will go back up. With a regular 2v flathead, you can aim for about 9:1 CR, which is perfect for boost.

>>17028690
They're pretty common in modern performance cars though.
>Alfa QV
>GT-R
>ATS-V
>C43
>M3/M4/M2
And that's just a few off the top of my head.

>>17028690
carfag here

>Packaging (cant fit under cuck bonnets and safety regs
>More costly to develop
>Not easily usable in fwd/awd cars
Or do you mean inline 6's?

Turbo v6 have always been pretty popular

VR6 and I6 are rare tho

>>17028708
Can't you filter out the noise then? I mean, some wideband knock sensors send the entire readout to your ECU, which can then figure out if it's knock or gear whine - unless it's the same frequency of course, which I doubt.

If you really want high boost, you'll need low compression internals eventually - although I'd much rather have high compression/low boost instead. Owned a 650 Seca Turbo, shit was not fun when it kicked in.

Isn't ignition vs. IAT fixed in the stock mapping though? Maybe you could fix it with a few degrees permanently off, to add to that 1 Deg/3psi.

>>17028711
Not OP, but F1 does have pneumatic valvesprings.

File: concern.jpg (24KB, 267x274px) Image search: [Google]

24KB, 267x274px

>>17028724
>>pneumatic valvesprings

>>17028724
>Can't you filter out the noise then? I mean, some wideband knock sensors send the entire readout to your ECU, which can then figure out if it's knock or gear whine - unless it's the same frequency of course, which I doubt.

I bet there is a reason why knock sensors aren't used on most of the bikes and it's the gear noise. Sure thing you can try to filter it out but still you wont be able to detect light knock events (which is what I'm after).


>Isn't ignition vs. IAT fixed in the stock mapping though? Maybe you could fix it with a few degrees permanently off, to add to that 1 Deg/3psi.

That's true but stock ECU isn't likely to handle IAT of over 70-80degC

File: Ford-Falcon-XR6-Turbo-2.jpg (99KB, 1200x767px) Image search: [Google]

99KB, 1200x767px

>>17028690
Not OP, but I can think of a few reasons for an automaker to pick another engine layout or leave a six cylinder engine n/a:
• In a laterally mounted V6 (obviously FWD), performance really wouldn't as much of a priority as trim-level-whoring is. I'm not saying FWD V6 cars are shit (the Toyota Aurion has a good kick), but it wouldn't be on the minds of the designers of these cars.
• If it's a RWD performance oriented car, they already have to deal with the longer hood due to longitudinal mounting, and an I6 would make the hood even longer, regardless of aspiration.
• In either a V6 or I6 RWD, I guess that automakers don't want to be boycotted by "mums against hoons" or some shit, as it would damage their reputation if more than a select few of their lineup had turbo six cylinder engines.
• Companies aren't selling enough of the cars that are equipped with these engines to justify them (pic related)

Just my ideas, it's great to see a non-shit thread with decent conversation; although I don't expect it to last as OP has fucked off and the first americans will be waking up in a few hours, and the thread will devolve into post after post of "EU cuckery"

>>17028776
>handle IAT of over 70-80degC
I'm looking at the moment at fast EGT probes that are suitable to be installed pre-turbo (i.e. inconel sheath and stout construction so its tip wont fell of after few heat cycles). I hope that would do the job as EGT is expect to rise (however I dont have any data on that).

>>17028776

Audio engineer here.

Knock is wideband noise, you can't filter it out as it occupies all frequencies.

>>17029035
So what if you filter out the gearbox noise then? That's probably a precise frequency.

>>17025686
F1 guy please stay.
We could really do with some intelligent life around here

>>17029043

The issue isn't gearbox noise. Driving over gravel will set a knock sensor off, even stuff like large fuel injectors cycling can trigger knock.

It's why people map with det cans as it takes a human to intelligently work out what is and isn't knock.

>>17029035
>Knock is wideband noise
>>17029043
>So what if you filter out the gearbox noise then?

engine knock is not that wideband, as far I know it's determined by cylinder bore mostly.

anyways the key here is its amplitude and duration with regard to valvetrain and gearbox noise,

I believe the best what can be done is either converting spark plugs to add a port for pressure transducer or trying to monitor EGT.

>>17029094
>It's why people map with det cans as it takes a human to intelligently work out what is and isn't knock.

that's for sure should be one of the options to try.

I was going to make a homemade dynostand for that, I got 400kg+ flywheel machined for that lying in my garage, however the machinist that did the job didn't make crosshatch (as it'd end too $$$$$ for me).

Do you work on ABAQUS?
How do I make a tubular steel-concrete column behave as a composite section?

What CFD application would let me import 3dsmax model and calculate downforce/lift values on both axles?

>>17028129
Excuse my noobness but what does "crossflow" mean? I'm a Finnish automotive mechanic student.

>>17029195
>cross flow flat head

https://en.wikipedia.org/wiki/Crossflow_cylinder_head

>>17029105

I've not run an FFT on a recording of det, but at a guess I'd say it sits between 7-10khz. It'll be hard to isolate that from random stones or even loose rattly bits.

Amplitude would be much higher like you say though, guess you could just gate it. The risk then is that your amplitude threshold cuts out anything less than catastrophic det.

Why did you get beaten by time attack cars. The hammerhead more specifically.

>>17029217
Oh, right. Thanks.

>>17029225
>The risk then is that your amplitude threshold cuts out anything less than catastrophic det.

It doesn't work like that, light det can kill the engine if it runs like for "long" period of time, while few severe knock events during transients (esp. when the engine isn't loaded 100%) are pretty safe, and severe knock events while running at max boost/max load could hinder reliability and kill the engine.

I believe the FI bike tune should be done on way safer side than on car due to lack of feedback (like knock sensors on cars) to probably that's a really big issue and I'm overthinking it.

Anyways runing over 250WHP on a motorcycle should be enough to have fun and it's easy with conservative tune.

>>17029251
I saw few Finnish guys on youtube filimig their project cars and how they fabricate turbo headers, mate gearboxes, etc, I really enjoyed watching that stuff.

Looks like 50% of guys in Finland either drive a rally car or do time attack or drift or whatever

Great country

>>17029273
Yeah. We really like our motorsports.

Will having a bachelors in mechanical engineering give me a gateway into working for company's like Honda/Nissan?

>>17029283
Yes.

>>17029249
Because F1 is extremely limited.

>>17025686
Can you please some good books on engine and drive train management (EFI, traction control), engine design (head design, lubrication, bearing design/load/etc, coping with natural frequencies/vibration)?

>>17025686
>Bored Formula 1 engine designer
Better not be Honda, those fuckers better be working 25/8 to fix their shit

File: 4138ZZXH4QL.jpg (29KB, 323x475px) Image search: [Google]

29KB, 323x475px

>>17029720
OP already recommended The Design and Tuning of Competition Engines in >>17026047.

I'll add pic related to that list, it covers a lot of frequency tuning and a big of head design. EFI/ TCS is beyond my scope as a mechanical engineer, and bearing design and load bearing is a 4 year education in and of itself.

>>17029729
OP confirmed for Mercedes.

>>17029749
>OP already recommended The Design and Tuning of Competition Engines in >>17026047.
It's like 60y.o. one, didn't man kind made huge progress in engine design and tuning?

>>17029955
That's better materials, tools, methods, machines allowing better/more nuanced designs.

The physics is the same as 60 years ago

>>17029980
ok

>>17025686
I want to see the drag co-efficients of more farm animals

If I'm studying computer engineering, what are my prospects for getting hired in the automotive industry?

>>17025912
How do you improve an engine's thermal efficiency?

>>17029605
Isn't it kinda sad though?
You guys dump millions after millions on research just to get cucked hardcore at the end.
What is the usual opinion on this topic around there?

File: smug anime girl.jpg (346KB, 1104x1104px) Image search: [Google]

346KB, 1104x1104px

Do you work for Honda, friend?

How does it feel attaching your name to a failed program?

>>17030524
>You guys dump millions after millions on research just to get cucked hardcore at the end.
No, you get cucked by the rules beforehand, and then spend several millions to be the best within that design parameter.

If it weren't for these rules, humans couldn't even operate F1 cars anymore.

>>17030535
Again, I'm not OP, but if I were a Honda engineer I probably wouldn't have any spare time to be posting on a Silesian engine block weaving forum

Whats the most effective rotary engine you've ever designed?

What did you think of Bahrain?

>>17025686
would adding full VVT/L and increasing injection pressures significantly improve efficiency?

Do you "agree" with all current forms of cost cutting carried on by FIA ?

>>17025912
>50% thermal efficiency
COMBINED efficiency (whatever that means)

>>17026047
there are no race tracks in my country senpai

>>17025686
bump

>>17025686
why is Ferrari's power unit so based

File: F06C4553.jpg (111KB, 750x500px) Image search: [Google]

111KB, 750x500px

will

>muh v10

ever return? or will liberty media possible bring back the v8

What do you think of NASCAR?

>>17030545
Yet they can operate fighter planes just fine

>>17032809
>f1 going down the mountain
scary as

File: giulia_colorizer_02_big_361.png (170KB, 820x300px) Image search: [Google]

170KB, 820x300px

>>17025686
Is really the engine of the Giulia Quadrifoglio a V6 derived from the Ferrari turbo V8 unit?

What do you think of that V6?

>>17025686
Hey, I'm looking to learn more about the design and programming of fuel injection systems. Got any ideas?

>>17032911
This a joke? No hairpins in the sky nigger

>>17027960
>btw how hard is it to be or become a racecar driver.
Unless you have enough money to sponsor yourself or have been racing your entire life, very hard. Make enough money in your day job to go racing on weekends.

>>17033311
bump

How much better would F1 be if they raced Formula 3 cars instead?

You don't need to answer this one because I've got the answer and it's a lot.

>>17032932
only the design of the heads and fuel injection system
the rest of it is a Chrysler pentastar v6 iirc
this does mean you could Frankenstein a pretty mean dodge charger

>>17028701
>diesel performance
>add cylinders

stfu

>>17026208
They are, and have been since, oh, the '40s.

>>17028656
And the big Cummins (NH, NTC, J, etc)
And Detroit 71s (though that was 4 exhaust valves)
And Detroit 60s
And 3xxx series Caterpillars

File: Cat V8 cutaway.jpg (767KB, 3888x2188px) Image search: [Google]

767KB, 3888x2188px

>>17035649
Actually, scratch that, the DD60 is OHC. As is the Cat 3508.
But this here I think is a 3408. See the bridges between the valves? Two pushrods, four valves.

>>17032809
Why would anyone want V10's to be back, except for noise? This has been the best F1 season since 2010 ffs.

>>17035600
How do I know you're american?

>>17033566
>igblyin
https://www.youtube.com/watch?v=1EpJ3KoUNmI
I doubt a F1 driver will ever manage to pull 12Gs in a corner.

how little engine go big fast

>>17025912
>>17025686
Anyone have more of these pics?

>>17033311
Google and a real-time PIC like an arduino.

If the engine is running at a constant speed and load you don't need much more than a wideband lambda sensor, a PWM generator (like an astable 555) and a couple MOSFETs.
You can do TBI, batch-fire MPFI, or if you wanted to get fancy with sequential injection, either a 555 per cylinder or a single 555 and some flip-flops running as frequency dividers.

>>17032911
>what are pressurised suits

>>17028559
bump would like a answer on this yo dood

>>17028559
>>17035837
It is additives.
Diesel oil has more zinc and shit that's good for old cars.
Poke around on this site and you'll see: http://www.pqiamerica.com/

Would you trade blowies with me?

>>17035804
I'd better use some DSP EVK/EVB

Usually you get better environment, better debuggers, a lot of libs to interpolate/extrapolate/implement various control loops/etc.

>>17028589
hoomanity fuck yeah

Is turbocharged 2stroke diesel engine possible and why hasn't it been made yet?

>>17031210
>would adding full VVT/L and increasing injection pressures significantly improve efficiency?
VVT/VVL only serve to meet emissions/fuel economy goals while also making a nice peak hp level. In F1, you'd never, ever, run on the ''small cam'' side of VVL, although VVT might add a bit of turbo response. You're not going to see added power from either addition, just some throttle response from the VVT systems.
Higher injection pressures? If it's limited, then there's bound be something gained if they ran, say, 150 bar instead.

>>17035625
>this does mean you could Frankenstein a pretty mean dodge charger
Why do that when you can buy a Hellcat Charger?

Also, why hasn't Alfa made a Giulia SS, using the Hellcat engine yet?

>>17036066
2 stroke diesels actually require boost to run, so they're always supercharged or turbocharged, often a combination of both.

>>17036099
>Why do that when you can buy a Hellcat Charger?
the shorter v6 helps reduce understeer
and the ZF 8HP transmission could make decent use of the power

>>17036099
>VVT/VVL only serve to meet emissions/fuel economy goals while also making a nice peak hp level.

VVT broads power band, VVL helps with mixture formation (more boost w/o knocking). Also VVT is used for EGR action but that's not neccassary

racing engines operate in narrow rpm band and that makes VVT basically not needed.

>>17025686
What is wrong with mclaren honda

>>17036099
>Why do that when you can buy a Hellcat Charger?
>Also, why hasn't Alfa made a Giulia SS, using the Hellcat engine yet?
he's full of shit, it's not a Chrysler engine

>>17025889
> Good books on engine design/layout/tuning?

bump for more books

Post more aerodynamics of animals

>>17025686
Why not LS swap it?

>>17035786
except airplane Gs are usually up and down, not side to side.

>>17035671
2012 desu

Why is mazda the only one that does NA engines? They have pretty good fuel efficiency in MX5.

Is diesel dead?

Is the future downsizing or back to upsizing?

When new regulations come for emissions and fuel consumption testing, how fucked are engine makers?

>>17041539
>Why is mazda the only one that does NA engines? They have pretty good fuel efficiency in MX5.

false, many manufacturers do NA's

>Is diesel dead?

Nope.

>Is the future downsizing or back to upsizing?

I believe both, e.g. we will have broader spectrum of cars, both tiny and large

>When new regulations come for emissions and fuel consumption testing, how fucked are engine makers?

It's consumers who are fucked, not engine makers.