Engine control scenarios / modes

It's all about the code!
JRD McLAREN
contributor
contributor
Posts: 137
Joined: Mon Mar 04, 2019 10:19 pm
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Engine control scenarios / modes

Post by JRD McLAREN » Tue May 28, 2019 6:38 pm

I have vacum tank ...

and Vacum vs. TPS vs. RPM not correspond to fuel requirement ...

JRD McLAREN
contributor
contributor
Posts: 137
Joined: Mon Mar 04, 2019 10:19 pm
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Engine control scenarios / modes

Post by JRD McLAREN » Tue May 28, 2019 6:42 pm

Stock 'M' BMW has ITB, runner, plenum chamber, air hose MAF, air filter ..

This is the right solution / setup for "fuel economy" ITB engine ...
.. but not for all engines ..

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Tue May 28, 2019 7:02 pm

JRD McLAREN wrote:
Tue May 28, 2019 6:38 pm
I have vacum tank ...

ands Vacum vs. TPS vs. RPM not correspond to fuel requirement ...
That is a truly terrible setup that doesn't work well on any engine I've ever seen....it can't work work well if you stop to think about what's going on. On a 4 cyl you have 1 cyl on the intake stroke pulling vacuum and 3 that are connected but not pulling vacuum, also know as 3 vacuum leaks. You simply can't get a good MAP signal from a vacuum chamber on an ITB setup. It's a terrible setup.

An MAF is an option, but is a flow restriction.

The setup I have reads the MAP of each cylinder separately so there is nothing leaking vacuum, then uses a simple diode circuit to send the strongest signal to the ECU so you have a good MAP signal and speed density works correctly with ITBs.

Here's a link to the files Jared created for it.
https://github.com/jharvey/MultiMAP

puff
contributor
contributor
Posts: 2734
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: Engine control scenarios / modes

Post by puff » Tue May 28, 2019 9:10 pm

How much of an air flow restriction does the hot wire maf make?
Do I underestand it correctly, your solution suggests using 4 map sensors for a 4-cylinder engine

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Tue May 28, 2019 11:04 pm

puff wrote:
Tue May 28, 2019 9:10 pm
How much of an air flow restriction does the hot wire maf make?
usually between 1-5% for the sensor itself....but wait there's MORE! A plenum type intake needed is generally in the 5-10% restriction, the hose connecting everything another few %. It just all adds up. There is a reason you don't find MAF sensors on race cars unless the rules require it.
puff wrote:
Tue May 28, 2019 9:10 pm
Do I underestand it correctly, your solution suggests using 4 map sensors for a 4-cylinder engine
Yes, 1 per cylinder so in my case 12 MAP sensors. Only 1 signal to the ECU though, strongest signal goes. I suspect there is a quickly diminishing return over 4 but I haven't measured it yet, its on my list though. I've seen data for a 4 cylinder ITB setup with a small vacuum chamber vs 4 separate sensors on the same engine and the signal about doubles with the separate sensors and is much more linear, very similar to what you'd see with a normal plenum type intake. On my 12 cyl engine I have the strongest MAP signal I've ever seen on any engine even though its got huge ITBs and has cams that would not normal be expected to give much idle vacuum (240+ degrees at .050 lift, about 270 seat to seat in a 4valve with huge valves and ports), I guess because its picking up basically peak vacuum from each cylinder.

JRD McLAREN
contributor
contributor
Posts: 137
Joined: Mon Mar 04, 2019 10:19 pm
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Engine control scenarios / modes

Post by JRD McLAREN » Wed May 29, 2019 5:41 am

mk e wrote:
Tue May 28, 2019 11:04 pm
puff wrote:
Tue May 28, 2019 9:10 pm
How much of an air flow restriction does the hot wire maf make?
usually between 1-5% for the sensor itself....but wait there's MORE! A plenum type intake needed is generally in the 5-10% restriction, the hose connecting everything another few %. It just all adds up. There is a reason you don't find MAF sensors on race cars unless the rules require it.
Are you shure ..??
You have any dyno tests ..??

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Wed May 29, 2019 11:33 am

JRD McLAREN wrote:
Wed May 29, 2019 5:41 am

Are you shure ..??
You have any dyno tests ..??
A mix of dyno and flow data….but its not as simple as change this get that return because everything is connected. If the exhaust is a restriction then improving the intake will only have limited effect, or if the TB and/or the intake and/or cams, etc are an issue you many never see anything going up or down on the MAF size because something else is a bigger problem. When everything is right though you really can’t size the MAF large enough to not restrict flow and still have a useable signal at low power/idle and not have non-linear flow-signal effects or noise issues, that is the 1%low end restrictions that are possible, higher restrictions are more typical of oem setups which include noise mitigation, air filter, usually design elements to smooth flow and prevent reversion at the sensor.

Edit: I'll add this. There are a number of simulation softwares, I use Dynomation6 but there are others that are similar. Dynomation has a bunch of things it looks at but one is a general intake system max flow number, like the carb or TB size which are normally the main restriction, but really it means the total flow of everything connected to the intake port of the head. Playing with that number makes a really big difference in HP for most setups...which makes sense since an engine is basically an air pump and the more air it pumps the more power it makes. OEMs look at stuff in hp/$ or hp/lbs or hp/ exterior engine volume (a packing thing) so volumetric efficiency means very little to them, but most racing limits displacement and most hop ups are based on the stock engine so volumetric efficiency is everything which of coarse leads to very different choices than the OEMs would make. A typical 70s V8 has a VE around 85%, a well done race engine will be north of 120% and numbers as high as 145% have been shown to be possible.....but there is also a trade-off between high VE and just spinning a lower VE setup faster so 115-130% is were most higher end efforts I've see land.

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Wed May 29, 2019 1:15 pm

Edit 2: there is a reason things like sportbikes and F1 cars are all ITB and do not use an MAF....that reason is HP.

JRD McLAREN
contributor
contributor
Posts: 137
Joined: Mon Mar 04, 2019 10:19 pm
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Engine control scenarios / modes

Post by JRD McLAREN » Wed May 29, 2019 8:40 pm

So ...

Maximum efficiency for "high modified" Naturally Aspired (NA) engine is aprox 110-115%
(F1 engine for the example)

Is too difficult make NA engine with 100% VE , if you not have lots of money, friend in workshop
and requirement knowledge.

"Our" modified or "tuned" NA engines has VE aprox 92-96%,
if you can buy cams, piston, intake manifolds, exhaust (etc.)
..and put it all together ..

You can calculate VE with this simple math.
0.7854 x bore x bore x stroke
or use this calc - http://atgtraining.com/atg-volumetric-e ... alculator/
but it is still theoretical maximum only

MAF sensor, the right one, has minimum restriction for air flow.
(Air filter is bigger restriction than MAF)
(but we need them)

Engines on "racing cars" - something like Caterham - with ITB
have no catalytic converter, air filter (no ITG FOAM is not air filter)
have no MAP, MAF, CAM, IAT sensors..
Casue they don't need them .. (not need to be covered for any EURO X Y emissions limits)
They don't need "fuel economy" also

It is simple Alpha-N ... no corrections ...
4 injectors and one wasted spark module for 4 cyl. engine.

JRD McLAREN
contributor
contributor
Posts: 137
Joined: Mon Mar 04, 2019 10:19 pm
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Engine control scenarios / modes

Post by JRD McLAREN » Wed May 29, 2019 8:45 pm

Forgot for any "dyno simulator" software ...
or for any "air flow simulator" ...

..
And any XYZ HP or kW from dyno (like MAHA LPS)
is not "real numbers" ...

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Wed May 29, 2019 8:49 pm

JRD McLAREN wrote:
Wed May 29, 2019 8:40 pm
So ...

Maximum efficiency for "high modified" Naturally Aspired (NA) engine is aprox 110-115%
(F1 engine for the example)
That's total nonsense
Wiki says about 130 and dohc, but that's low.

I saw results for a 2v pushrod engine doing 138 on an epa certified dyno back in the mid 90s and an f1 engineer shared a 145 number with me, dyno queen, not a race spec engine but it was an exercise to see what was possible and 145 is what they achieved.

My 1990 zx6 stock other than jetting and exhaust did 112. My h-d roadrace did a 122 in the 1996 hp shootout at Daytona bike week.

110-115 engines are back markers ;)

https://en.wikipedia.org/wiki/Volumetric_efficiency
Volumetric efficiencies above 100% can be reached by using forced induction such as supercharging or turbocharging. With proper tuning, volumetric efficiencies above 100% can also be reached by naturally aspirated engines. The limit for naturally aspirated engines is about 130%;[2] these engines are typically of a DOHC layout with four valves per cylinder. This process is called inertial supercharging and uses the resonance of the intake manifold and the mass of the air to achieve pressures greater than atmospheric at the intake valve. With proper tuning (and dependent on the need for sound level control), VE's of up to 130% have been reported in various experimental studies

JRD McLAREN
contributor
contributor
Posts: 137
Joined: Mon Mar 04, 2019 10:19 pm
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Engine control scenarios / modes

Post by JRD McLAREN » Thu May 30, 2019 5:59 am

..yes wikipedia ...
Volumetric efficiencies above 100% can be reached by using forced induction such as supercharging or turbocharging. With proper tuning, volumetric efficiencies above 100% can also be reached by naturally aspirated engines. The limit for naturally aspirated engines is about 130%;[2] these engines are typically of a DOHC layout with four valves per cylinder. This process is called inertial supercharging and uses the resonance of the intake manifold and the mass of the air to achieve pressures greater than atmospheric at the intake valve. With proper tuning (and dependent on the need for sound level control), VE's of up to 130% have been reported in various experimental studies
with ITB .. has no intake manifold ... :roll:

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Thu May 30, 2019 11:35 am

JRD McLAREN wrote:
Thu May 30, 2019 5:59 am

with ITB .. has no intake manifold ... :roll:
Your ITBs are hanging by zip ties from the hood then?

The intake is everything from the intake valve to open air. The part you normally tune is the runner, so ITB to top of whatever stack/trumpet you have. The tuning is several things. There is the length which mostly sets the timing of the wave reflection pulses, the volume which sets the mass that is oscillating, the cross sectional area at'near the intake vale that sets the max velocity (E=1/2mV^2 so this is REALLY important, 0.5-0.6 mach is typical in race engines) and the taper which is about frictional losses and the total volume/mass in the runner. All intakes including ITBs have all this going on.

A multicylinder intake also has a plenum which isn't itself a problem but it is a challenge to minimize wave interference cylinder to cylinder and control the opening to minimize wall effect flow losses, but the biggest issue is normally there is a 90 degree turn the air needs to make, turns are very bad for flow. If you look at something like a ferrari V8 intake 355 and newer (355 was ITB, the newer are not but keep most of the features) they look very much like an ITB setup with an air box. They are 2 separate 4 cylinder banks (ferrari uses a flat or 180 deg crank so the V8 is truely a pair of 4 cylinder engines), this lets the plenum be tuned to oscillate and recapture the work used to stop/start flow each cycle as is done in the air box of 4 cyl motorcycles....so a plenum isn't inherently bad and can be a benefit but usually they do more harm than good.

Not mentioned in the wiki article is the most important part, the exhaust. The exhaust is where the magic really happens as its where the energy is. A well done header will drop the cylinder pressure to about 20-30kPa during the time period when both the intake and exhaust valves are open to set the intake charge in motion. There are 3 parts to intake, the crossover where the exhaust is during the work, the mid time when the piston is working, the end when the energy from the intake track is being recovered. That is how high VEs are achieved.

Attached is a graph of a well tuned engine at 9k rpm that shows intake and exhaust velocities and pressures through the engine cycle....there is a lot going on. This kind of shows why taping the runner and mixing them together to get a MAP signal is a bad idea. Most of the time the runner does't have any vacuum so its just a leak, but you're also bleeding off part the energy you want to use to fiull the cylinder. A normal plenum type intake takes the reading outside the runner and away from a lot this noise (unless its a tuned plenum then you have a similar issue).
Attachments
2013011-07 1000hp full graph set.JPG
2013011-07 1000hp full graph set.JPG (177.97 KiB) Viewed 366 times

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Thu May 30, 2019 12:28 pm

Another piece for the MAF path I didn't cover well is the noise but probably will make more sense after looking at the graph I just posted. Typically an intake is tuned to the 3rd reflection...sometimes the 2nd or 4th but 3rd is generally best....so there are 3 rpms where the intake is generating noise.

Then if you have high overlap cams, which are nonexistent in any modern OEM MAF setups but mission critical on race setups as I explained above, then you also have exhaust pulses in the intake track. Standard headers have primaries on the 3 pulse and collectors on the 5th, so 7 exhaust pulses (the peak rpm collector and primary match so 7 not 8). With tri-y type headers you have 2 primaries and 2 collectors effectively. My tri-y have main primaries on the 2nd pulse, secondary on the 3 do tuned to different rpm so 5 distinct pulses then the collectors are 5 and 6th, with last pulse matching each other and the main primary, which adds 9 more pulses for a total of 14 rpms there is exhaust noise in the intake track. Some have more.

Then the piping in the intake track itself will have its own frequencies and you might be getting more fro the exhaust piping is you aren't running a header choke or wave termination box of some kind. Getting a good clean MAF signal off a race engine is a challenge and anything you try to do to damp out the noise or improve the signal generally also damps out some HP. There is a LOT of engineering in OEM MAF setups that all gets buggered when you change cams, exhaust, porting, filters, ducting, etc....its not easy.

mk e
Posts: 85
Joined: Tue Dec 06, 2016 7:32 pm

Re: Engine control scenarios / modes

Post by mk e » Fri May 31, 2019 1:57 pm

:roll:
puff wrote:
Tue May 28, 2019 6:37 pm
hey! do we consider the rover v8 like this
Image
with a plenum chamber on top as an ITB?
Ferrari intakes are like that too. It's a design that flows well but MAP signal wise it's not like an ITB because the whole plenum area pulls vacuum and forces each runner to that level all the time....so you get a pretty good MAP signal.

With ITBs unless the throttle is basically closed each runner is moving between vacuum during the intake stroke and basically barometric pressure by the end of the exhaust stroke, you end up with a very weak signal that doesn't really corrilte well to engine conditions so its not very useful, maybe at idle but even then its sketchy.

.....which brings me to the question of why the OP is saying it's no good for fuel, because the signal is basically worthless, but then wants to use it to control spark? If I read it right? That's a pretty bad idea which is why most ecus don't offer an option like that.

Normally with ITBs alpha-n is used for ALL load sensing and a barometric correction between the load table the shows tps and the actual fuel/spark lookup or calculation. You can put the the baro correction right on the main tables and call the axis load....good so you know exactly what numbers are being used but bad in that you can only really tune WOT cells when baro is high....100% TPS with baro might only be 70% load it you live on a mountain for example so I don't really like that approach.

I read the issue had to do with 1800-2000 rpm cruise causing the mixture to bounce between 12-14 AFR.....that's just very poor tuning. Alpha-n load sensing makes determining the driver's intent ( cruise vs race), the TPS it a direct torque demand indicator, so for example up to 20 or 30% tps is cruise at 14.7 afr or whatever gives best milage, 70-100 is tuned for max power, 31-69 is a transition zone that should smoothly go from max efficiency to max power....so the AFR should not be bouncing. Also the spark table should be the exact same axis points as the fuel so the spark matches the fuel...there can be a few small exceptions but they are exactly that, exceptions and don't normally apply.

So the problem that started the threat is a tuning issue. To be fair, Alpha-n is quite a bit harder to tune than speed density because nothing is linear so most every single cell needs to be tuned not just interpolated and useful axis points are never ever evenly spaced, but it work well once its setup up correctly. The only issue I've ever seen with alpha-n is that the tps tends to drift with time so you end up messing with it always vs set it and forget it with a MAP based tune.

Post Reply