Another thought on this topic.  My XLR got new tires along the way, and uses Michelin Pilot MXM4 tires, 235/50R-18.  The calculated diameter for this size would be 27.3″ per Tire Size Conversion, or 27.25 according to Virtual Dyno.  Tirerack lists the actual diameter for this size from this Manufacturer as 27.4″.  One would also assume that is on a new tire, and the diameter will shrink as the tire is driven and wears?  I have decided to use the Virtual Dyno theoretical calc of 27.25″ based on tire size of 235/50-18.

As a convention, I am going to put test equipment into Driver Weight, and remove gas used from car weight.  So Driver weight will include laptop, hptuner interface, etc.  Car weight for the XLR is 3647, which I assume is full of fuel.  So on the sample runs tomorrow I estimated the car was down 3 gallons at 6 lb/gal and deducted 18 lbs.

Virtual Dyno doesn’t adjust for altitude, so I am going to leave that out as well.  Plano is at 600+ feet of altitude, so we will just know the readings are lower than sea level.

Frontal Area & CD — in 2004 preproduction test C&D quoted a frontal area of 22.4 ft2 and a CD of 0.35. Cadillac marketing said the CD was 0.31, which is what I am using.  In 2006 XLR-V testing C&D quoted CD 0.33 x frontal area 44.6 ft2.  That is a large difference.

From a thread on Virtual Dyno on HpTuners, The quick way to calculate frontal area is (((w*h)(0.85))/12)/12 = Frontal Area.

For the 2007 base XLR, W=72.3in,  H=50.4in so for a first approximation (((72.3*50.4)*.85)/12/12)= 21.5 ft2.

Okay, so that settles a convention for Weight, Driver Weight, Drag Coefficient, Frontal Area, Tire Diameter, and I updated the gearing to match spec for 2007.

This chart only shows runs from yesterday.  The next task is to get more data with the same baseline to look for consistency.

UPDATE:

Studying the HP Tuner results, it appears yesterday’s run actually was lower hp due to higher engine coolant temps.  The ECT negative advance kicks in within that range.

This is the ‘before’ or premium comparison file showing 192F temps and 25.5 degrees of advance at 5995 rpm.

This is yesterday’s run showing 5995 RPM has only 23.5 degrees of advance because engine coolant temp has hit 207F.

This is a file from hptuners showing the ECT base advance adder going from 0 at 194F to -3 at 212F, so in this case at 207F adding most of (13/18ths)  -3 or -2 degrees of advance.

So the XLR made less power yesterday due to reduced advance from approx 299.1 hp to 290.6 hp due to those 2 degrees of advance.

This type of intricate timing and fueling in response to a variety of inputs makes back to back testing difficult, but I appreciate the insight that the tools like hptuners provide.

Using my initial and second (premium) run data captured from HPTuners, I did some Virtual Dyno Analysis (see my initial Virtual Dyno article here).  First I needed a new custom file for the XLR:

I had run the test data with the top down (worse aerodynamics), and I used a back-to-back wind tunnel test of a different car (Miata) to model the degradation of top down as .06 CD.  I will want to run virtual dyno tests with top up in the future?

I created a 2nd custom file to model the XLR with top down:

I edited the .csv saves from Hptuners to only include the acceleration range desired, and loaded both runs.

[click on image to zoom in, back to return]

I adjusted the SAE conditions for the pressure and temps during the test, 29.1 in/hg baro and 100F.  The car intake air was actually a bit higher.  I am not sure this adjustment is quite right — the reference site suggests a dyno adjustment of 4.5% for test conditions.  The Virtual Dyno not-SAE – to SAE goes from 251 to 259 or a 3.2% increase.  A 4.5% increase would be 262 whp, so not much difference really.

The results, 249 whp for the initial test, and 259 whp for the known-quality premium fuel test, seem consistent with expectations.  259 whp / 0.8 transmission losses = 324 hp.  Cadillac rated the stock XLR LH2 V8 engine at 320 hp.

One might conclude the difference between believed-regular unleaded and known premium unleaded in this test case was 10 whp (13 hp).  This is similar (9 hp) to what the calculated hp values showed.

The calculated engine hp vs virtual dyno:

Initial:  2nd gear 293 calc hp x 1.045 conditions adj = 306 hp
Virtual Dyno 249 whp / .8 trans loss = 311 hp

Premium: 2nd gear 302.3 calc hp x 1.045 conditions adj = 315 hp;
Virtual Dyno 259 whp / .8 trans loss = 324 hp

Or told in the other order:

Initial: 306 adj hp x .8 trans loss = 244 whp
Virtual Dyno: 249 whp

Premium: 315 adj hp x .8 trans loss = 252 whp
Virtual Dyno: 259 whp

Note in both tests 1st gear calculated hp was higher, and for comparison with Virtual Dyno analysis of the same runs I am using 2nd gear hp data.

Virtual Dyno depends on consistent, level test area, weights (with/without fuel for example), and conditions for comparable results.  It has however been tested over and over to be consistent with dynojet results when done with proper care.

My conclusion is that I need to do a few more captures of consistent test data for Virtual Dyno use, but the configuration files I have predict results consistent with my other indicators.