Cause of incorrect lambda reading ?

[TurboME]

I'm having trouble pinpointing the cause of a puzzling situation. I'm working with a GDI engine and using an injector with a sub optimal spray pattern which I believe is causing excessive cylinder wall wetting. Cold start is very bad. At WOT the engine's wideband lambda sensor is outputting a reading significantly richer than flow bench data for the injector suggests would be possible. Flow calculations based on lambda and air flow readings exceed the flow bench by ~12%.

Could the incorrect spray pattern cause the lambda reading to be incorrect in the rich direction? My initial thought is that a bad spray pattern would result in unburnt fuel which should result in a leaner than expected condition, but this is not the case.

 

[mattsooty]

It could be a problem with the lambda sensors O2 pump circuit. Do you have any means of measuring lambda brettschneider/spindt?

MS

 

[Lou Scannon]

My experience with UEGO sensors in natural gas engines is that unburned fuel causes the UEGO sensor to read falsely rich. I have not used them with gasoline but I was told on another forum that this does not occur with gasoline.

 

[TurboXS]

Are you working in lambda or AFR?

What is the fuel composition? Any ethanol or oxygenated fuel?

 

[TurboME]

Working with lambda values. Have tested both standard 93 octane gasoline and 93 oct with 10% ethanol, non oxygenated. Similar results with both.

I should also mention that another injector was tested that had an appropriate spray pattern for the application. It showed a higher flow rate on the flow bench, but resulted in leaner lambda values with the same fuel pressure and injector pulse times.

The engine is also turbocharged. One theory I have is that the excessive hydrocarbons in the exhaust created by the poor spray pattern are being concentrated towards the outside of the exhaust stream due to the centrifugal forces created by turbine. This could created a richer mixture at the tip of the lambda probe, however I have not found any published data to support this theory. Does this sound at all plausible?

 

[Bribyk]

What is the error in your calculations?

I'd think the turbine would mix the exhaust gases enough to be homogeneous.

 

[turbocohen]

Has the ego been exposed to anything leaking on it? How many crosscounts per sec are you seeing? What other fuels, stuff has it been exposed to? What type(s) of oil are you using? 4?