Stratified Charge DI with EGR

[Bernd Felsche]

Mike writes:
>At 07:32 PM 11/3/2000 +0800, Bernd Felsche <bernie at perth.dialix.com.au> wrote:

>>Throttle-by-wire. The driver sets the amount of torque required by
>>the position of the pedal. The ECU adjusts throttle, EGR (up to
>>70%), injection and ignition accordingly.

>>The complication has its rewards; 15% less fuel consumption, partly
>>due to increased "volumetric efficiency" because the throttle is
>>often wide open... better find an updated abbreviation to replace
>>WOT!

>Aye - I'm a bit perplexed here, are we talking diesel if not but, talking
>petrol then I would have expected wide open throttle to consume more
>petrol due to maintaining  AFR ?

It's petrol. 

>On the occasions I have wide open throttle on my vehicle I'll either
>chew up huge amounts of fuel and/or leave rubber, but then my EGR is
>minimal...

Nor do you have stratified charge, or direct petrol injection one
would assume.

>In the event the engine is undersized/powered in comparison with an AFM
>or MAP only setup then I suppose there is value - but are you saying with
>this fly by wire that gradual pedal pressing under load by human can result
>in wide open throttle plate by fly-by-wire.

Up to a certain load, the engine operates in stratified-charge, lean
burn, with the throttle wide open. Yes; a slight depression - i.e.
low torque demand - results in WOT and a lean-burn cycle. 

Note that the semantics of the pedal have changed; the driver
determines the torque output, not the throttle opening. So if you
keep pressing further, the engine will switch to homogenous
operation and start throttling airflow - though you'd probably be
hard-pressed to pick the transition.

>70% EGR does sound awfully high though, any observations that this is
>used on turbo engines. And on the n/a ones is there much attention to
>EGR cooling ?

Sorry, I mis-read the section (it's quite a lot crammed into a few
pages) - it's a 70% reduction in NOx via EGR - No actual rate of EGR
is given other than being high.

>By the way whats does conventional "WOT" stand for ?

Wide-Open-Throttle; implying maximum torque demand.

>The stability and PID equations must be interesting ;-)

There are a few recent SAE papers; I don't have the funds to
undertake all the research in that direction. Interesting though it
may be. The first car to employ Bosch MED7 is the VW Lupo FSI.

>>The engine uses quantity-regulation at low load; the torque
>>developed depends on the amount of fuel injected. The amount of air
>>sucked in and ignition timing have little effect on torque.

>This sounds bizarre, especially as I would have expected AFR to be kept
>as close to stoich (or slightly) lean unless its a diesel...

Very much leaner than stoich - Bosch describes mixtures as lean as
2.5; any leaner than 1.5 and stratified charge operation leads to
soot formation; any leaner than 1.3 in homogenous mixture formation
and ignition can fail, leading to a prohibited band of AFR which the
engine management system must skip on the load transitions.

[In fact, the nature of the NOx storage cat (which stores the NOx as
nitrates) requires that the engine periodically switch from
lean-burn (LB) to homogenous so that sufficient CO is generated to
allow the reduction to N2 and CO2 to take place. This switching is
transparent to the driver as the ECU knows how to switch from LB and
back without any change in engine torque.]

Direct injection allows injecting the fuel immediately before the
spark (during compression) so that an ignitable mixture surrounds
the plug. The lower the load, the less the fuel injected, always
finishing just before the spark - I guess to ensure a dense mixture
cloud to light.

>Can you please clarify particularly this last paragraph of yours ?

Hmmm. I'd like to order a few of the tomes of SAE papers on DI and
EGR over the past few years first. Unfortunately, it adds up to many
hundreds of US dollars, which is probably better invested in me
exploding microcontroller chips as a hobby.

Meanwhile; I'd recommend that you try to get the Bosch reference I
wrote about earlier. It doesn't go down to component-specification,
but deals with the process of SI engine management well. They're not
going to give the game away by telling you everything; it's Bosch
and it's their game that they want to keep playing.

The english-language edition has a few more pages, so there might be
some more details about MED in there. Should be able to order the
book through a good bookshop.

-- 
Real Name: Bernd Felsche
    Email: nospam.bernie at perth.DIALix.com.au
     http://www.perth.dialix.com.au/~bernie - Private HP
----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org