More realistic turbo simulation

[Nick7]

If anyone has driver real turbo car, they'd notice turbo emu/simulation is .. well... not really right.

Building boost in neutral is impossible, unless there's specifically option in ECU for it, which moves ignition way back, so most of gas expansion actually happens after exhaust valves are open.
This can be found on some race cars to enable for faster start, but in more 'regular' cars is, well.. not an option.
Also, such thing can/will damage engine if used too much due to high exhaust temperatures.

Another thing is how boost is built when you lift off and press again on throttle, as it's not that much realistic either, just doesn't have realistic feel.

For better turbo emulation you should have a matrix similar to turbo map itself, which could ease way the spool/boost is calculated, and which could keep in account turbo rotational speed too.

Well, it doesn't need to be calculated as is in real world, but make it feel more real :)

[pasibrzuch]

For better turbo emulation you should have a matrix similar to turbo map itself, which could ease way the spool/boost is calculated, and which could keep in account turbo rotational speed too.
I guess Scawen will make it more realistic than matrix. In old versions of LFS turbo gauge was turbine rpm gauge :) so, if it was simulated back than we can be sure about changes going in good direction.

[Nick7]

Well, what I meant by 'martix' is simple matrix holding data simlar to turbo-maps themselves.

What creates boost is exhaust gases, so in neutral it's simply not possible to build max boost by flooring it.

On a same note - flooring gas in neutral should not consume same amount of fuel as floored pedal in some gear while driving :) (amount of fuel should be somewhat equal to power produced).

[5tag]

I have not driven many cars and neither of them had a turbo but it sounds realistic that you should not be able to build up a usable amount of boost in neutral because there is barely any load on the engine (only on the way up until the rev limiter).

[Ball Bearing Turbo]

It's entirely dependant on the engine & turbocharger in question. In some vehicles, you can hit full boost breifly just by revving the engine and not in others. In fact as a generality, a modern production car would be more likely to do that since they're generally built for low RPM torque, not screaming high RPMs racing around a track.

LFS does however need better turbocharger modelling no question.

[5tag]

It's entirely dependant on the engine & turbocharger in question. In some vehicles, you can hit full boost breifly just by revving the engine and not in others. In fact as a generality, a modern production car would be more likely to do that since they're generally built for low RPM torque, not screaming high RPMs racing around a track.

LFS does however need better turbocharger modelling no question.

XRT would have surely been developed before 1990, so if would feature a bigger turbocharger than those commonly used today.
RB4 and FXO would fit somewhere near before 2000 so I'm not really sure.

Also spooling up a turbo with a low compression bike engine in the MRT should not be that easy without load on the engine.

And of course XRR and FXR feature very big turbos.

[mitbrown]

mnm.. problem is that in most cars the rev limiter is an ignition cut-off, and LFS seems to have a "mechanical" limiter so the engine is at max load when full rev.

[Zebediah_S2]

XRT would have surely been developed before 1990, so if would feature a bigger turbocharger than those commonly used today.
RB4 and FXO would fit somewhere near before 2000 so I'm not really sure.

Also spooling up a turbo with a low compression bike engine in the MRT should not be that easy without load on the engine.

And of course XRR and FXR feature very big turbos.

what are you talking about?
have you ever driven a turbo car or any real car at all?
first...
turbos today are bigger than turbos of the 80's and 90's
in the 80's, turbos were just beginning to be used, and engines were not fully re-tuned to deal with the boost (ie, lowered compression, better pistons)
in the 90's and later, turbos were beginning to be used more widely in stock cars, and turbos got bigger because engines were being built with turbos in mind right from the start.

not only that, but cars today (in case you havent noticed) are more powerful than cars of the 80's. more engine means more turbo

the MRT uses a honda sportbike inline 4 IIRC.
any sportbike engine is a HIGH COMPRESSION engine.
a v-twin uses compression similar to your average car engine, ie, 8.5-9.5:1
an inline 4 bike engine shares more with an F1 engine than a typical car engine.
compression ratios are around 11.5:1 and as high as 13.8:1 in some cases.
most bikes are turbo-ed by only changing the pistons, which lowers the compression slightly to accept the added intake pressure.

that being said, compression has NOTHING to do with the tubo spool characteristics.
the spool characteristics are ENTIRELY based on the turbine size, weight, and efficiency in relation to the compressor size, weight, and efficiency.
the exhaust gasses are what spin the turbine which spins the compressor.
the more gasses, the more pressure is made.
the gasses passed thru the engine are directly related to the engine size and RPM



using a matrix that references points on a turbo map, in combination with throttle position, engine speed, and turbine speed, would be the best way to accurately simulate the turbo IMO

[tristancliffe]

What are you talking about?

Turbos today are generally smaller than in the 80s and 90s, as the turbos are more efficient and better optimised for each application.
In the 80s, turbos were well proven and had been used for donkeys years, and engines were scratch built to cope with them.
In the 90s turbos got a bit smaller as they were optimised more and more engines were designed with turbos in mind.

Not only that but cars of today are more powerful, but this has nothing to do with the size of the turbo. More engine most certainly doesn't mean more turbo.

The MRT uses a bike engine, but with the compression ratio reduced to around 8.5-9.0:1 to make use of the turbo. Even as standard they aren't that high (certainly not F1 high) as they need to be reliable and last a while on 'normal' fuel - you don't have to rebuild a 1000cc bike engine every 4,000 miles and use 105 octane fuel...
A bike engine shares more with a typical car engine than it does with an F1 engine.

Please don't make too much shit up.

[Forbin]

the MRT uses a honda sportbike inline 4 IIRC.
Yamaha YZF-600R, actually.


any sportbike engine is a HIGH COMPRESSION engine.
a v-twin uses compression similar to your average car engine, ie, 8.5-9.5:1
Suzuki SV650 v-twin -- 11.5:1
Suzuki SV1000 v-twin -- 11.3:1
Ducati 1198 v-twin -- 12.7:1
Honda RC51 v-twin -- 10.8:1

I happen to own an SV650 with a stock motor. It runs on 87 AKI ("octane") or 92 RON.


an inline 4 bike engine shares more with an F1 engine than a typical car engine.
compression ratios are around 11.5:1 and as high as 13.8:1 in some cases.
most bikes are turbo-ed by only changing the pistons, which lowers the compression slightly to accept the added intake pressure.

I think the real MRT has low compression pistons, which is probably why 5tag referred to it as a "low compression bike engine."


that being said, compression has NOTHING to do with the tubo spool characteristics.
Nothing at all? Not even on the temperature differential on either side of the turbine? Doesn't that affect spool characteristics?


the spool characteristics are ENTIRELY based on the turbine size, weight, and efficiency in relation to the compressor size, weight, and efficiency.
the exhaust gasses are what spin the turbine which spins the compressor.
the more gasses, the more pressure is made.
the gasses passed thru the engine are directly related to the engine size and RPM
Just engine size and RPM? So if you close the throttle but the engine is spinning at 9000 RPM, it'll spool a turbo?


using a matrix that references points on a turbo map, in combination with throttle position, engine speed, and turbine speed, would be the best way to accurately simulate the turbo IMO
What makes this method the best? Can you give some examples of other methods and explain how they are inferior?

[5tag]

I think the real MRT has low compression pistons, which is probably why 5tag referred to it as a "low compression bike engine."


No, sir. I actually thought sports bike engines have low compression ratio to be able to rev high better.

[KiRmelius]

+1 to this suggestion. I don't know much about cars, but when i first time driven a car with turbo, i was surprised how much it differs from turbo cars in LFS. Turbo lag is much bigger IRL. Building up boost is not as easy as in LFS. Even though it was a diesel, difference is still very easy to spot.

[_--NZ--_[HUN]]

+1 to this suggestion. I don't know much about cars, but when i first time driven a car with turbo, i was surprised how much it differs from turbo cars in LFS. Turbo lag is much bigger IRL. Building up boost is not as easy as in LFS. Even though it was a diesel, difference is still very easy to spot.

You've driven one car with turbo and you think you know how all turbo cars behave?
I have two Starions, one with the first generation TC06 turbo which is bigger, and one with the newer and smaller turbo. The older car builds up boost later but can hold it better in the higher RPM range. The smaller turbo comes on sooner but it's not as effective at higher RPM and feels more responsive.
These two cars are almost identical, but they behave quite differently just because they don't have the same turbocharger.
Of course I'm not saying LFS' turbo modelling is totally accurate but most of the people here(including me) don't have enough knowledge and experience to comment on this because it's a very complicated subject.

[PhilS13]

Turbo lag is much bigger IRL.

By IRL you mean "in the car you were driving with his specific engine with specific ignition timing map , specific compressor & turbine sizes & wear, specific manifold & exhaust flow, specific intake piping etc etc etc."

Right?

[Kamrock]

...Turbo lag is much bigger IRL. ... Even though it was a diesel, ...

This also depends on the car you were driving, not just the fuel... When I drove the Honda FR-V Diesel I hardly noticed the turbo lag, however, in the Focus it was horrendous, and I always found the car red-lined before you could sensibly use the power then you had to build up again after the change.

Although I suspect that the much higher red-line probably helped the Honda a lot (and knowing that Honda did a pretty good job building their new diesels almost from scratch, whereas I don't think the engine in the Focus hadn't been changed in years) and enabled the turbo to be much lighter.

All in all seeing how there's no sudden surge (as some turbos give) the low lag I feel isn't out of place.

[Nick7]

Actually it's even 'big lag' not low lag in most situations.
If you have high rpm's, spool will come much faster. On low rpm's you can't build boost.

Also, with big turbo's you cannot build max boost in at least first 1 gears, due how fast engine revs and lack of exhaust gasses to spin turbo that fast that soon.

[mdmx]

mnm.. problem is that in most cars the rev limiter is an ignition cut-off, and LFS seems to have a "mechanical" limiter so the engine is at max load when full rev.

I think this is the main reason for why the turbo seems so unrealistic..

It almost seems like there is no actual 'engine load' parameter involved when calculating turbine speed, but throttle position is used instead. Which means 100% throttle == 100% load, which is far from being realistic.

It seems the same applies to engine sound, you can clearly hear the intake sound when rev'ng to limiter full throttle on neutral. Sound starts to change immediately when you are releasing the throttle.

So i wonder there isn't that parameter anywhere in simulation. If that's true, then adding realistic 'engine load' parameter would greatly affect turbo behavior not only at neutral and when revving to limiter, but also on low gears and rpm's before/after torque peak, making it harder to build boost.

The more air in cylinder, the more fuel injected, the more exhaust gas produced, the more pressure build. So the pressure also builds more pressure, cos it's air in cylinder. That's why it's harder to start build from zero psi, and easier to maintain when you got max psi.

[Ball Bearing Turbo]

No, that would be a stupid way to do it and arbitrary. I kind of understand why people errantly say it, but it's a falacy that an engine's ability to produce boost is "based on load" :doh:

For the turbocharger simulation to improve the overall engine simulation simply needs to improve, and that was something that Scawen has indeed mentioned in the past.

[mdmx]

No, that would be a stupid way to do it and arbitrary. I kind of understand why people errantly say it, but it's a falacy that an engine's ability to produce boost is "based on load" :doh:

For the turbocharger simulation to improve the overall engine simulation simply needs to improve, and that was something that Scawen has indeed mentioned in the past.

Maybe i used wrong term, if i think it now the load term may also contain the amount of resistive forces, like air flow, friction, gravity. Of course these doesn't affect. In practice of course they seem to affect but not directly, if you are driving to very tight uphill with 1st gear it takes longer to accelerate to rev limiter, so there is simply more time to spin the turbine and produce pressure. Physics engine gravity takes care of that, definetly not part of engine simulation.

In fact what i ment by 'load' was intake pressure. The thing that causes the engine to sound like it's under 'load'. Intake pressure is often used by ECU to determine amount of fuel injected, in cars which engine has been tuned. Stock cars use airflow meter for that. Highly tuned turbo cars often use throttle position.

The whole point here is, when you hit the rev limiter ecu stops injecting/igniting. But intake pressure also doesn't change on rev limit, so i was wrong all way to the bottom. :) So the engine simulation algorithms needs to be fixed in a more authentic way.

[Ball Bearing Turbo]

The whole point here is, when you hit the rev limiter ecu stops injecting/igniting. But intake pressure also doesn't change on rev limit, so i was wrong all way to the bottom. :) So the engine simulation algorithms needs to be fixed in a more authentic way.

Indeed at the rev limiter the pressure differential across the turbine should drop substantially. But the way things are now is just a very simplistic implementation of the entire engine, including the turbocharger. Hopefully that will change one day.

[Tim_J_23]

One thing I hate, I don't know if it is like this on real cars but slap me with a tuna if it is.

If you revv to full RPM but not holding full throttle you get no boost.
If you press that extra 2mm on the throttle the boost will fly up but still, same RPM.
It feels wrong ...

Slap me with a tuna!

[Forbin]

At least in LFS, the blow-off valve (wastegate?) is closely related to throttle position. It's open at closed throttle, and gradually closes as throttle increases. Meanwhile, even though the BOV is open and there's no boost, exhaust gasses are still flowing through the turbine, spooling it up. When you give it that little extra bit of throttle, the BOV closes a little more, just enough to make the manifold pressure positive.

[Ball Bearing Turbo]

Well, blow off would be on the intake side of the engine. It should only open when pressure between the impeller and the throttle valve is too high to prevent stall & shock on the impeller.

Wastegate should only open when intake manifold pressure is as high as desired, and allows exhaust to flow past the turbine side and prevent excessive spool (overpressure).

What Tim said happens in LFS for the reasons explained in the previous few posts. The engines continue to build boost at the regular rate when pinned on the limiter. When he revs the engine to just under the limiter, the turbocharger is spooling up X amount up to some RPM, then when he pins the engine on the limiter, the turbo spools up the rest of the way very quickly because it was already spooled X amount and the engine builds pressure quickly.

Probably would not happen in real life although depending on the engine you would probably get some intake pressure even when bouncing off the limiter. How much would depend entirely on the engine in question.

[Nick7]

Blow off is connected to intake manifold.
It opens up on negative pressure (vacuum) in intake manifold, which is achived with closing throttle butterfly/low throttle.

Wastegate is generally also connected to intake manifold and will open up when there's excess of pressure in intake manifold and by this way regulates intake pressure.

This is just simplified version.

As for turbo spooling. It's spooled by exhaust gasses. More load, more exhaust gasses -> more boost -> more power/exhaust gasses.

Since this part does take some time you generally can only generate low boost in 1st/2nd gear, and very often none while in neutral (or very low - depending on engine/turbo size).

For making it easier calculation, using engine load 'can' be used as starting point to go and calculate how much boost turbo will generate. Ofcourse, factors as turbo spool time, etc cannot be ommited.

2 Tim: It depends. If you're in gear (like 4th/5th or so) - yes, boost will fly up, granted there's still no ignition cut.

[wheel4hummer]

Wastegate is generally also connected to intake manifold and will open up when there's excess of pressure in intake manifold and by this way regulates intake pressure.

You left out that the wastegate releases exhaust gasses, not air from the intake. :thumb: I'm sure you meant to say that, but the way you said it made it seem like you were saying that the wastegate released pressure from the intake manifold, which it doesn't.

[Nick7]

You left out that the wastegate releases exhaust gasses, not air from the intake. :thumb: I'm sure you meant to say that, but the way you said it made it seem like you were saying that the wastegate released pressure from the intake manifold, which it doesn't.
I said it was 'simplified version' :)

But ofc, wastegate opens flapper on exhaust part, so part of exhaust gasses goes around the impeller to reduce pressure turbo makes.

FYI: Was quite a lot into turbo's 5+ years ago, when I custom installed turbo on my Honda Civic + all changes (fuel, mapping) it needed. Hell, even understood how to read compressor maps for turbo's/calculate based on them :)

[mdmx]

Blow off is connected to intake manifold.
It opens up on negative pressure (vacuum) in intake manifold, which is achived with closing throttle butterfly/low throttle.


No, it's more of an opposite. Dump valve (blow off valve?) is in intake manifold, yes. But it opens up when pressure gets too high, not on negative pressure. And negative pressure in intake (on charged engines) happens when you flat the throttle and there is yet no pressure.. It happens when cylinder sucks more air than there is in intake. Not when you lift off.

Big turbine in big booster is quite heavy, which means it has quite a lot of inertia. When it rotates at it's full speed, it doesn't stop immediately when you lift off the throttle, but keeps spinning.

As long as it spins, it keeps building pressure (pushing air to intake). But when throttle is lifted off, the flap is closed and air has nowhere to go, it can't go to engine. So pressure in intake manifold very rapidly raises too high. It may even break the turbo, in case of very high boost. Boosters goes to state called surge, an unstable state between compressor and turbine, don't know the details of this.

And even with smaller boosters, it slows down the turbine dramatically meaning it takes a long time to build up pressure again after shift or quick lift off. With dump valve the pressure is released to atmosphere, which means the turbo can spin freely, without counterpressure, which means it slows down much slower, which means the pressure goes up much faster when you flat the throttle again.

Wastegate affects to pressure in a very different and much slower way, it simply adjust the pressure to wanted level by controlling the flow of exhaust gasses, if they bypass the booster or go thru it, so you don't blow up your engine/compressor with overpressure.

[senn]

LFS has the most realistic Turbo simulation i've seen in a racing sim so far. GTR Evo/Netkar/rFactor don't even simulate boost.

Also iirc, turbo cars, when free revved, don't make a lot of boost, until you rev them flat out....need someone with a boosted motor + gauge to confirm that, was a while ago i was speaking to someone about it, something to do with the engine not being loaded? not 100% sure..


if you want to see bad turbo "simulation" watch a video/play NFS Shift *shudder*
pretty sure it sends the boost gauge up with the engine off *facepalm*

[Nick7]

mdmx, you confused things a bit.

The pressures you mention are NOT in intake manifold but inbetween turbo compressor and flapper.
Blowoff valve/dump valve is connected via vacuum line to intake body - and to make clear, this is what open it. Blowof itself sits inbetween turbo and intake flapper.
When you release throttle, you close flapper -> at that point in intake manifold pressure goes to negative (vacuum) while turbo is still pushing air due to inertia. This air hits flapper and build still bust in intake pipe to manifold.
This is where blowoff comes to play. Due to vacuum in intake manifold it opens up BOV (remember, it's connected to intake manifold via vacuum line). While opening up, it releases boost either into air or back infront of turbo (VTA or recirtulating BOV) - air released is one inbetween turbo and intake flapper.
Reason for this is due to no gasses to spool turbo and pressure in intake turbo starts drastical deceleration which can damage bearings.

[mdmx]

mdmx, you confused things a bit.


Yes you are and were right. There is negative pressure in the intake manifold, after the flapper, and too much pressure before it in tubes. :x Confused it..

I have always thought that dump valve is opened by the overpressure before flapper. But it's much better to use manifold pressure after the flap, i think it reacts much faster that way, less space to create negative pressure in intake manifold than there is in pressure tubes before the flapper? And you don't have to readjust it every time you change the pressure.

[senn]

flapper = Throttle body/butterfly i assume

When people say "flapper" i imagine the old style AFM's that have a huge flap inside them for measuring air flow.....lol

[Nick7]

Well, BOV can also release air based on over-pressure, but that's like last-resort option, as it's not precise, and idea is to release any positive pressure out when flapper is closed.
It depends on spring stiffness in it.

2senn: yes, flapper = throttle body = butterfly

[senn]

BOV's are used to prevent compressor surge and subsequent damage to the turbo, AFAIK.

Closed BOV's, more boost than the engine can handle

http://www.youtube.com/watch?v=nCj6Spwl1CU

And BOV not set/working correctly vs working/set correctly

http://www.youtube.com/watch?v=Ffj01k64AZI

[Nick7]

BOV's are used to prevent compressor surge and subsequent damage to the turbo, AFAIK.
Exactly.
Closed BOV's, more boost than the engine can handle

http://www.youtube.com/watch?v=nCj6Spwl1CU
This was turbo surge, due to wrongfully dimensioned turbo for engine.
This is why you have compressor maps for turbo's to size them appropriately, and why there are different A/R ratios, etc..

And BOV not set/working correctly vs working/set correctly

http://www.youtube.com/watch?v=Ffj01k64AZI
Generally this is due to too stiff spring :)

[mdmx]

You can often heard that lift-off 'surge' sound, "dihdihdihdih", in very powerful track cars. +700hp audis and beasts like that.

Is it because the boosters and pressures are so big you can't just remove it completely, or is it intentional, a compromise where the pressure isn't released on small lift-offs? Or just bad design?

Like there, not very loud compared to dump valve sound but still exists:
http://www.youtube.com/watch?v=yvUXqa4Zo3Y&feature=PlayList&p=1CD8FC41832F8032&playnext=1&playnext_from=PL&index=29

And then again, +700HP road audi which doesn't have surge at all, this made me wondered if it's intentional in race cars?
http://www.youtube.com/watch?v=4uyYsLbRrY4

[atledreier]

I get the surge when I don't boost more than 0,4-0,5Bar. Anything over will open the BOV properly. If I loosen the spring in the BOV it won't hold my full boost (1,7Bar).

Edit: Here's a little clip. Notice how at lower boost I get surge, higher boost it opens properly. This was on snow though, so there is not many times I get full load and boost, since the revs go up so fast.
http://www.youtube.com/watch?v=1cwTerPgIIA

[mdmx]

I get the surge when I don't boost more than 0,4-0,5Bar. Anything over will open the BOV properly. If I loosen the spring in the BOV it won't hold my full boost (1,7Bar).


So does it mean that your BOV is not connected to intake manifold via vacuum line, and opened by negative pressure, but overpressure in intake tubes instead?

The more boost there is, the more far away it is from being negative pressure. Which means that more loosened BOV controller spring (if connected to manifold, opened by negative pressure) would only 'leak' on full throttle when there is yet not enough pressure and engine is vacuuming negative pressure to pressure tubes/hoses as well?

Can this even happen in boosted engines, negative pressure before flapper? Dunno, happens all the time on 'free breathing' (what's the english term?) engines. :)

[atledreier]

My BOV is indeed connected to the intake via vacuum, but the vacuumline alone is not enough to overcome the spring and pull the BOV open. If it were, the spring would be too weak to hold the BOV closed at full boost. So in a lower boost config like the car is stock (0.65bar) there is no problem having the spring weak enough for it to open at even a fraction of positive pressure and still keep closed at full boost. My setting is a compromise.

[Nick7]

Not entirely true... surge on BOV is situation you mention: when spring is too stiff and vacuum alone is not enough to open fully BOV.

Also, keep in mind vacuum alone >can< be enough to open BOV - it's pure physics.
Inside BOV is a piston which is pushed by spring so it's in closed position.
Area that's on pressurized side is for example 1 cm^2 (as example).
However on other side there's vacuum line and on other side of piston it's bigger surface area - example 4 cm^2.
This means if piston would move at 0.5bar negative pressure it is same as if it were pushed with 2 bar pressure on pressurized side (different surface size).

This way you can have stiff spring to hold high boost, yet have vacuum alone enough to open BOV.

What this also means - when there's pressure in intake it will also help keep BOV closed on boost, so it is not only the spring that is holding it closed.

This however is very dependent on BOV itself.

[erfrag]

if you have a diverter valve (aka recirculating BOV--or depending on what BOV you have) you can change its orientation around so the piston or diaphragm face away from the charge pipes and towards the intake side. this will effect pressure response, and can alleviate some forms of compressor surging.

http://i52.photobucket.com/albums/g27/frag85/Car/nromalvsreverse.jpg

that sounds like some really nasty surge you have happening. partial boost surge is caused from the turbo pumping out more air than the engine can suck in causing the air in the charge pipes to oscillate between the intake valves and compressor housing. over time this will destroy the turbo, and cause more wear+tear on the engine itself, since your intake is charged the BOV is being forced closed so the excess air cannot escape. my car did this when i installed less-restrictive turbo inlet pipe and using a few shims and a softer spring solved my issue.

your best bet would be to find an enthusiast forum for your car.


As for LFS, i hate how you don't loose boost when shifting, and the way the turbo spools on the TBO's. if you did more than a few shifts of pedal to the floor your clutch would be toast and your gear box would be taking a lot of damage. if we are going to have TBo's that can build max boost by just holding the pedal down, and not have to let off to shift we should at least have the benefit of shift lights so we don't have to do data logging to find our optimal shift points.

I guess i'm just used to my car, but in LFS the turbos seem to spool very slow for the boost and the power they generate. it seems like the turbo must be really lousy and/or is terribly mismatched for the application (too large of a turbo with the boost tuned down.)