It was a little over 18 months ago when I was trying to determine what affect a turbocharger blanket would have on spool up that I defined a measurement called the 2-11 time.  The initial discussion of this measurement centered on how I determined what the measurement would be of, but not why I thought this measurement was useful for comparing the responsiveness of different turbochargers.

Recent discussion about the measurement has given rise to some questions that provide a good opportunity to look more closely at the 2-11 time and consider if there are alternative measurements that may be better at expressing the turbocharger response characteristic it tries to represent.

One of the first questions was “why not measure vehicle acceleration? ”  Fair enough.  When I asked how the data should be presented it was suggested to use rpm vs time, as a line graph.

Here’s a sample of how this would look for the first couple of seconds of a few pulls:

This to me is not particularly clear, at least not with respect to which setup is more responsive.  Putting these pulls into the format I’ve chosen, the data for these two turbochargers looks like what is shown below:

Another advantage is the amount of data that can be represented while still being able to differentiate between the products.  The chart below has 255 pulls shown, if I were to try and put that many acceleration curves on a chart with lines it would be incomprehensible.

Some other comments made were as follows:

“Boost pressure is not spool.”

True.  But looking at a compressor map,

the pressure ratio, (turbocharger outlet pressure divided by the inlet pressure), does generally rise with the increasing rotation of the compressor wheel, and the manifold pressure is a result of the turbocharger outlet pressure, after some of the intake piping pressure losses are taken into account.

While rising boost pressure is not a perfect substitute for the rotational rate of the compressor wheel, (the line of boost pressure rising is not parallel with the left hand speed line), it is pretty close, and it is an adequate substitute for my purposes.

The 2-11 psi range is approximately 1.2 to 1.9 on the PR axis.

How about measuring the rotational rate of the compressor wheel, could that been done?  Yes.  Garrett sells a speed sensor kit that retails for around $450.  It involves machining the turbo’s compressor housing so the sensor has a line of sight to the compressor wheel.  Unfortunately it will only be good on a single turbocharger, and as my effort has involved multiple turbochargers this is not a cost effective option.  Of course if someone is willing to fund the effort to purchase and install these systems on multiple turbochargers and then pay for the altered turbo’s they should contact me to make arrangements.

Yet another concern related to the affect other variables may have on the outcome.  This is something that I make a reasonable effort to control, but have limitations to work with.   A comment made to me was in relation to the vehicle weight and is worth looking into further:

“the weight of the car will effect turbo performance the same as overall acceleration. Weight = load. Load = drive. Drive = boost. Every variable you have mentioned being afraid of skewing the acceleration test is also skewing your spool testing”

Rather than load the car up with extra weight I decided to simulate a weight change by performing a series of pulls going uphill, and a series going downhill.  Then using ECUxPlot I’ve graphed the calculated acceleration (g’s) versus engine speed for each pull, shown below.

The affect of the road angle on vehicle acceleration is clearly evident.

How about the affect on boost?

Or if you prefer, airflow:

I believe that these measurements are strongly influenced by the road angle variable.  But how about spool rate?  This is what ECUxPlot calculates:

The two conditions, uphill and downhill, only start to break out into distinctive trends around the 9-11 psi/sec (2600-2700 rpm).  On the boost only chart this occurs when the boost curve is between 9-11 psi for the two road conditions.  Spool rate is not affected as much by the road angle as the acceleration curves are.

Placing this series of pulls into the format I have chosen gives the following result:

Downhill appears to have a slight influence on the time, while uphill looks to follow the trend line for all of the data points.

The bottom line from this investigation is that I now feel the 2-11 time is less sensitive to vehicle weight than acceleration, dispelling concern raised about how spool and acceleration are equally affected by the vehicle weight.

Going forward it is worth considering the concerns raised about measurements such as this, but it is also important to weed out and ignore uninformed objections or those that are raised out of spite.

Today’s effort was directed at the engine that will be going into the Silver S4, the former Nogaro engine with 91k on it.  At some point in the recent past the timing belt was replaced, but on account of me not knowing when in the past it was done I decided to update the baseline.  The valve covers were also done recently, but I know those were within the last 3,000 miles so I’m not concerned about being dead on with the mileage that they were replaced at.

With the engine out and a number of components removed I decided this would be a good time to replace the coolant hardpipe o-rings.

I purchased the O-rings as part of a servicing kit that ECS Tuning offers, along with a few other parts I was in need of.

I can’t recall seeing pictures of where these O-ring are located in a picture of the engine, so I took a few pictures as I went along.

 

In preparation for installing the new thermostat and water pump I put a polishing wheel on my dremel and cleaned up the surfaces that the water pump gasket and thermostat o-ring will be seated against.

This post will be updated occasionally as more entries are submitted to the Audi B5 S4 Survivors poll.

As of 12/31/2016 there have been 506 entries submitted.

A concern that was raised about the results related to the data not being unbiased.  If the results are biased, trying to apply them to the entire population of B5 S4’s could make for incorrect conclusions.  To try to assess if there is some data bias I graphed the percentage of Tiptronic transmissions out of all the samples, as the number of entries increased.

The increasing percentage of Tiptronic transmissions does suggest that this data is ‘enthusiast’ biased, on account of B5 S4 enthusiasts being more inclined to buy manual transmissions.  The majority of early contributions to the survey were made by members of B5 S4 car forums and Facebook communities – enthusiasts.

With the continuance of time I have entered B5 S4’s that I have seen listed in for sale ads and on insurance auto auction sites.  These cars are more likely to be representative of the overall population, and as their numbers are increasing in the poll, the percentage of Tiptronic transmissions is increasing.

Another measurement that can be looked at in a similar fashion is the body style, something less likely to be influenced by an enthusiast, or non-enthusiast, owner.

In this case the percentage of Avant body style out of the the overall sample population has remained fairly consistent, around 15%.