Effect of EGR? Or is it?

[07heri]

This EGR stuff is making my head hurt.  I understand, in theory, how to adjust it.  What I don't fully grasp is how it's measured.

Lets say I'm looking at the RPM rows 1500, 1750, 2000, 2250 and 2500.

I see a rise in the 2000 and 2250 rows....5% TPS to 20% TPS.

I have read a few hours worth of threads on EGR.  Based on what I've read it's to smooth out the VE's.  Makes the math easier for the ECM is my best guess. 

What I can't get my head around is this.  Are we only "assuming" it's EGR related?  How do we know it's EGR influenced?  Doing a street tune we don't know what the AFR's are in that region.  All we can do is trust the VE's are real close, even with the hump.  Could the AFR be where we want it even though there's a hump in the 2000 and 2250 RPM rows?

[FLTRI]

...Could the AFR be where we want it even though there's a hump in the 2000 and 2250 RPM rows?

Sure. See it all the time and is normal in most situations...if VE values are correctly calibrated.
The hump is the change in fuel requirement to meet target AFR. You can use the EGR tables to lower the hump...however the increase in the smoothness/rideability is proportionate to how big/steep the hump is.
HTH,
Bob

[mayor]

Here's how I view the egr settings with my laymen thought process.  The airflow in the engine generally will not make a major abrupt change, and when you see the airflow (VE) making a major change there is likely due to another table affecting how the ecm is calculating that value.  The area that is affected by these EGR tables are somewhat predictable, they are only ever going to be at the lower load areas of the throttle position.  This means that when you see a major swing between light load and heavier load in the same area, that swing is likely caused by how the ecm is interpreting the supplied values.  The EGR tables are to be used in conjunction with data collected (o2/lambda sensor feedback) and that data determines whether or not a change is needed to smooth out the ve transitions in areas that have too much or too little egr table influence.  Anytime a value is changed on the EGR table, the ve values must be repopulated (reVtuned) because the values that were there before are no longer accurate.  TTS has included in the vtune software a feature that indicates where the transition between where the EGR tables are active and where they are not. This is very helpful in finding transition points.  I would caution though, this can be a moving target.  I have seen TPS areas on one run that are affected, only to be moved to the left or right the next run.  Adjusting the EGR values to smooth out the populated VE tables can be frustrating and time consuming, but I still use these tables when I see major abrupt changes.  As a DYI street tuner, I think making the ve transitions as smooth as possible does help limit the potential for lean/rich spots. 

[07heri]

lets say in the rpms row i listed i have

  90  90   90   90   90 
  90  90   90   90   90 
100 100  100 100  100
100 100  100 100  100
90   90   90    90   90

The VE's are correct.  The AFR is assumed to be correct but we don't know.  Just relying on the tuner did its job and VE's are good.

Next, I use EGR table adjustment and the 100's all come down closer to 90.  Did this just cause a leaner situation when the VE's came down from 100 to 90?  Won't the ECM now work harder to get those areas back to 14.6?

My confusion is this.  We don't know the AFR is good or bad in the hump area.  If it's good with the hump isn't that an easier math process for the computer?

Now we lower the hump.  VE's come down.  Map looks smooth.  But isn't the AFR different due to the lower VE's.  Didn't we just cause more work for the ECM to bring the AFR back to where it was good WITH the hump?

What I'm failing to grasp is why is a hump or a valley automatically assumed to be due to EGR?  If like FLTRI is saying, it's normal.  Why mess with it?  If the VE's are correct and the AFR is correct and we have a hump why mess with it?  By getting rid of the hump, assuming the VE's and AFR is good with the hump, aren't we now forcing the ECM to make a correction (due to the lower VE's) that it didn't have to make when the hump was there?   

It's hard to put in words what I'm getting at...

[mayor]

if the ve's are correct, then the only way that that occurred is the afr would have to be correct.  The collected afr values is what determines the ve values and those ve values have variables  connected to them.  The EGR tables is one of those variables.  You could have two different ve table values that record the same measurable AFR outcome if the variable table values change.  Whether you have a hump in the ve tables or not, if the constant is the desired measured afr output the ve value is correct.  here's a simple way to look at it: 

let's represent AFR with a constant value of 10, represent the ve table value as a variable x, and the EGR table value is a fixed number. 

if 10 = 5X, then x= 2
if 10 = 4X, then x = 2.5
if 10 = 2.5X, then x =4

10 is the constant in this case, which is what the collected afr is since everything is based on the afr data collected.  The target is staying the same, but anytime you change the variable X (EGR table value) the value needed to reach the constant (VE table value) moves.  Does this change the constant, no.   

Don't get to wrapped up in the name EGR, these tables are not exactly indicative of how much EGR is actually occurring.  The just needed to have a name, and that's the name that TTS gave those tables. Is egr happening in these affected areas, yes.  The only thing you are doing when you are adjusting the EGR tables is limiting the amount of correction that is applied in the affected ve cell calculations to ensure that areas that are affected by these tables aren't being over corrected.

[Tsani]

The "hump" is not neccessariy due to EGR, but could also be an effect caused by the cam and other various components of your build. Therefore it can be normal. Now with EGR, that will cause issue with readings made by the O2's and generally comes into play in the lower to middle rpm areas and the 60 kPa and lower areas will be affected the greatest. Use the EGR analyzwe to help you in this reguard. I also use the 2D graphing to help me see whot is going on with EGR.

A case in point: I had a couple of cells that were consistanly maxing out in the VE % range in the rear cylinder. So I lowered the EGR table by 10% in the rpm area closest to the area giving me fits. Yes, I t did smooth the graph a bit but most importantly for me is it gave me more head room in those cells so I could be sure (as much as I could be) that I was getting the correct readings. By using the EGR analyzer I started adjusting the EGR Table in certain rpm areas, + or -, to get the data to go as flat on the graph line as possible. Result was a better running bike, less heat, detonation, and VE swings. The over all after effect was some areas of the VE table seemed to flatten because I believe it was adjusting for EGR effect and getting better data. But there comes a point were adjusting the EGR tables will gain you little to none. To just have to play with it and learn. With my VE table graph, I can clearly see the effect of the cams stated range.

And to really muck this all up is you need to figure how much of all this is due to O2 sensor placement. Eventually you reach a point where it is what it is and you either accept it or make modifications. This is where the Pros have an edge because they see it more often than we do.

I have also been using the DataMaster graphing feature to help me visualize all this crap as well. And I can tell that my front cylinder is definately different than my rear cylinder and I strongly believe I can see the effect of the O2 placement and the EGR. Looks to me like the Front has a worse placement than the rear and it makes sense to me when I physically look at the header and think about the flow in it. WWhen I get a chance, I will take some screen shots and throw em up. Might be a while tho. But I can say with out a doubt, that adjusting the EGR tables has been a great help to the point where it took only a few runs to stablize the mapping in the mid to upper rpm areas that could be done without a dyno. Readings there now do not change there by more than 3%+- for me. Also keep in mind that while street tuning it is also impossible to create the exaction same conditions time after time and you will see changes that reflect this. 

[07heri]

if the ve's are correct, then the only way that that occurred is the afr would have to be correct.  The collected afr values is what determines the ve values and those ve values have variables  connected to them.  The EGR tables is one of those variables.  You could have two different ve table values that record the same measurable AFR outcome if the variable table values change.  Whether you have a hump in the ve tables or not, if the constant is the desired measured afr output the ve value is correct.  here's a simple way to look at it: 

let's represent AFR with a constant value of 10, represent the ve table value as a variable x, and the EGR table value is a fixed number. 

if 10 = 5X, then x= 2
if 10 = 4X, then x = 2.5
if 10 = 2.5X, then x =4

10 is the constant in this case, which is what the collected afr is since everything is based on the afr data collected.  The target is staying the same, but anytime you change the variable X (EGR table value) the value needed to reach the constant (VE table value) moves.  Does this change the constant, no.   

Don't get to wrapped up in the name EGR, these tables are not exactly indicative of how much EGR is actually occurring.  The just needed to have a name, and that's the name that TTS gave those tables. Is egr happening in these affected areas, yes.  The only thing you are doing when you are adjusting the EGR tables is limiting the amount of correction that is applied in the affected ve cell calculations to ensure that areas that are affected by these tables aren't being over corrected.

OK, so VE's are assumed good.  Hump is evident.  If the Ve's are good the AFR should be good.  What kind of over correcting would be happening?