[Diy_efi] Re: Diy_efi Digest, Vol 6, Issue 23

[Bill Washington]

Mike,
    In my experience even when the diesel is pumping out lots of black 
smoke - ie unburned or part burned fuel - there is still lots of excess 
oxygen - the reason for the smoke is usually droplet sizes that are too 
large to burn in the available time - ie inadequate atomisation so the 
oxygen can't get to all the fuel during the power stroke - In this 
scenario, for a mechanical injection diesel, even if more fuel is 
injected than there is oxygen to burn there will still be excess oxygen 
because of the droplet size.
     The big adavantage with the latest generation of common rail 
diesels is the superior atomisation achieved by raising the fuel 
pressure in the distribution rail to extremely (even
dangerously) high  levels and changing from mechanical injection to 
electronic injection - the mechanical injection pumps had no hope of 
achieving the injection pressures the common rail designs use as normal 
- the pressure difference from mechanical injection to common rail is 
orders of magnitude.
    Therefore I believe that a normal O2 sensor (narrow or wideband) 
will not work on a diesel because there is always 'excess' oxygen. Note 
also that normally, diesels have wide open air induction - ie no 
throttle butterfly and engine speed and power are controlled solely by 
the volume of fuel injected, therefore the amount of air inducted in a 
normally aspirated diesel per stroke is essentially the same across the 
entire rev/power range (neglecting restrictions in the air filter and 
intake tract) - including these effects means that the volume of 
inducted air will reduce slightly as the revs increase.
    Adding a turbo changes this, of course
    That being said, I seem to recall reading somewhere about some of 
the common rail diesels using an oxygen sensor but no details of type or 
purpose though it stands to reason that it would/ could only serve any 
purpose in a maximum power condition in a modern common rail diesel when 
it may be possible to inject more very finely atomised fuel than there 
is air to burn it, then an oxy sensor could come into play to limit the 
volume of injected fuel.
Regards
Bill

>----------------------------
>
>Message: 7
>Date: Mon, 22 Aug 2005 22:10:37 +0800
>From: "Mike" <niche at iinet.net.au>
>Subject: Re: [Diy_efi] Wide band sensor, F1
>To: diy_efi at diy-efi.org
>Message-ID: <6.2.1.2.0.20050822220408.02951220 at 203.0.178.192>
>Content-Type: text/plain; charset="us-ascii"
>
>At 03:01 PM 8/22/05, you wrote:
>  
>
>>I would say "no", assuming you mean the usual WBO2 sensor used in petrol engines.  I have two reasons:
>>
>>1.    It's calibrated to provide a response which, while almost linear immediately about the stoichiometric mixture for petrol (c. 14.7: 1? memory fails) its response goes rapidly non-linear on either side of that figure.  Diesel engines, AFAIK, run in a regime where there is always excess air to burn the fuel which is expected to be metered by the injection pump.  So the sensor would be stuck "on the end stops" of its response curve.
>>    
>>
>
>Damn good point, I forgot all about this, power=fuel, not air flow based.
>
>I should have known as the first thing I did in Malaysia was rent a diesel
>with turbo and jammed the wastegate off - tut tut...
>
>  
>
>>2    Even if the injection pump is injecting excess fuel which cannot all be burnt, it tends to come out the exhaust as fine particles of carbon rather than as unburnt fuel, which is what the sensor is looking for.  I'm not convinced the sensor would notice, because of the different combustion process (chemically and physically) going on in a diesel.
>>    
>>
>
>mmmm Surely though if there is excess fuel any O2 sensor will indicate
>when (allowing for response time) and how far assuming it isnt jammed up
>against the 'rich' stop so quickly... ?
>
>So under those conditions would a wideband show lower than  say 11:1
>and not have its judgement clouded (pun) by soot particulates ;) ?
>
>ie. Even though its nonlinear wouldnt it be mappable, ie Calibrated to give
>a known output, with an error margin, at extreme ends of the scale, its been
>a long time since I looked at any O2 sensors admittedly...
>
>I wonder what they use for F1 (or do they) when running 4:1 or is it
>with that exception so they use such very rich mixtures just to cool
>the exhaust valves at selected parts of the race when acceleration
>isnt that important,
>
>
>
>Regards from
>
>
>Mike
>Perth, Western Australia
>VL Commodore Fuse Rail that wont warp or melt !
>Twin tyres for most sedans, trikes and motorcycle sidecars
>http://niche.iinet.net.au
>  
>
>  
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.diy-efi.org/pipermail/diy_efi/attachments/20050824/df2cedd0/attachment.html