CMax testing from Argonne National Laboratory

[jdbob]

This is back from October 2013 but I just ran across it and didn't see a link posted here. Covers both the Hybrid and the Energi 2013 models. Lots of techno-talk I don't understand but someone might find it useful.

 

Go to https://anl.app.box.com/s/n35jmex87dmvf8ulm27y and there is a download link at the top for the PDF version if you want.

 

[plus 3 golfer]

Garbage In  =  Garbage Out ;)

 

Dynamometer settings were based on wrong road load HP coefficients. Ford updated the coefficients in 2014.   As I posted in another thread, below is the history of the RLHP coefficients and HP vs speed curves for the C-Max compared to the Prius and Prius V.

 

[SPL Tech]

Garbage In  =  Garbage Out ;)

 

Dynamometer settings were based on wrong road load HP coefficients. Ford updated the coefficients in 2014.   As I posted in another thread, below is the history of the RLHP coefficients and HP vs speed curves for the C-Max compared to the Prius and Prius V.

 

 

What is rev 2013 and rev 2014? It took Ford three attempts just to determine the drag coefficient of the vehicle??

Edited December 26, 2014 by SPL Tech

[plus 3 golfer]

The original C-Max EPA rating of 47/47/47 was based on the Fusion Hybrid EPA tests.  The Fusion has a drag coefficient of 0.27 but the C-Max drag coefficient is 0.30.  So, the C-Max EPA mileage rating was overstated.  But, there was  a loophole in the EPA test procedures that allowed such to happen.  A manufacturer could use one set of EPA test results for one vehicle for other vehicles that basically have the same engine.  For example, the intent is to allow the manufacturer to run only one set of tests for the C-Max SE and SEL Hybrid and if the Lincoln / Mercury division had a C-Max like vehicle the test results could also be applied to the L/M vehicle.  That's the second line in the table and solid red curve.

 

Ford then actually tested the C-Max and reduced the EPA mileage rating in August 2013 but had yet discovered that the RLHP coefficients were in error.  That's the first line in the table and the solid dark blue line in the graph.

 

Then in 2014 Ford discovered that had used the wrong RLHP coefficients in several models including the C-Max.  Ford retested and again lowered the EPA FE numbers (last line in the table and light blue solid line in the graph).

 

I find it hard to believe that FORD (especially their engineers) didn't know what was going on with the EPA testing in 2012 and the fact the the RLHP numbers were incorrect.  Anyone familiar with the EPA test procedures and how RLHP is determined and used should have questioned Ford's C-Max tests in 2012 and 2013 when compared to other hybrids. How does a heavier, less aerodynamic vehicle have better EPA FE than the Prius V.  It's not magic, it's physics.    

[Smiling Jack]

The original C-Max EPA rating of 47/47/47 was based on the Fusion Hybrid EPA tests.  The Fusion has a drag coefficient of 0.27 but the C-Max drag coefficient is 0.30.  So, the C-Max EPA mileage rating was overstated.  But, there was  a loophole in the EPA test procedures that allowed such to happen.  A manufacturer could use one set of EPA test results for one vehicle for other vehicles that basically have the same engine.  For example, the intent is to allow the manufacturer to run only one set of tests for the C-Max SE and SEL Hybrid and if the Lincoln / Mercury division had a C-Max like vehicle the test results could also be applied to the L/M vehicle.  That's the second line in the table and solid red curve.

 

Ford then actually tested the C-Max and reduced the EPA mileage rating in August 2013 but had yet discovered that the RLHP coefficients were in error.  That's the first line in the table and the solid dark blue line in the graph.

 

Then in 2014 Ford discovered that had used the wrong RLHP coefficients in several models including the C-Max.  Ford retested and again lowered the EPA FE numbers (last line in the table and light blue solid line in the graph).

 

I find it hard to believe that FORD (especially their engineers) didn't know what was going on with the EPA testing in 2012 and the fact the the RLHP numbers were incorrect.  Anyone familiar with the EPA test procedures and how RLHP is determined and used should have questioned Ford's C-Max tests in 2012 and 2013 when compared to other hybrids. How does a heavier, less aerodynamic vehicle have better EPA FE than the Prius V.  It's not magic, it's physics.    

 

 

Many thanks to the Golfer for posting, re-posting and explaining all of this (with charts).

 

Something remains curious to me:

 

 The lines show the C-max PHEV (Energi) as having RLHP way greater than the original C-max HEV, essentially equal to the first revision, and noticably less than the second revision.  None of these seem to make sense.

 

Was the curve for the Energi never revised?  If so, how could it have so much greater RLHP than the Hybrid with the main difference being a few hundred pounds of weight ?  The weight penalty should not become more severe in the high speed region dominated by aerodynamics, and the aerodynamics of the two models should be essentially equal. No?

 

Does the Energi curve correspond to the first revision?  In that case it's hard to believe that the weight penalty would appear to be nearly nonexistent.

 

Does the Energi curve correspond to the second revision?  If so, how could the Energi possibly have lower RLHP than the Hybrid?

[plus 3 golfer]

Good questions. I wasn't focusing on the Energi. I'll have to look at what the Energi curve represents.

 

Ford originally tested the C-max Energi for the initial EPA numbers and used those FE numbers for the Fusion Energi. So, there was no retest of the C-Max Energi in 2013. In 2014 ford discovered the Error in the RLHP coefficients and retested many vehicles as shown in the link below and got the latest EPA numbers.

 

http://www.epa.gov/fueleconomy/updates.htm

 

I forgot to mention that one must remember that the Energi final drive ratio is numerically higher than the Hybrid. This coupled with the Energi weighing more than the C-Max will result in lower EPA FE for the Energi which can be seen in the link.

Edited December 27, 2014 by Plus 3 Golfer

[plus 3 golfer]

The latest Energi RLHP target coefficients are; 28.82,  0.345 and 0.02102  This data is not shown on the graph but would put the Energi curve slightly higher than the C-Max hybrid.  

 

The curve for the Energi shown in the graph is the original target coefficients that would be compared to the C-Max Hybrid HEV August Revision (solid dark blue line).  Remember the original C-Max HEV curve is the curve for the Fusion HEV which is the reason the original HEV and PHEV curves are so far apart.  

[Smiling Jack]

The latest Energi RLHP target coefficients are; 28.82,  0.345 and 0.02102  This data is not shown on the graph but would put the Energi curve slightly higher than the C-Max hybrid.  

 

The curve for the Energi shown in the graph is the original target coefficients that would be compared to the C-Max Hybrid HEV August Revision (solid dark blue line).  Remember the original C-Max HEV curve is the curve for the Fusion HEV which is the reason the original HEV and PHEV curves are so far apart.  

 

Golfer,

 

Thanks very much for posting these last two explanations.  That clears up a lot.

 

Thanks especially for the latest Energi RLHP coefficients.  Do those really correspond to the latest C-max Hybrid coefficients?  Why would the Energi's B coefficient be nearly 25% higher than the Hybrid's?

 

Also: What are the units to be used for V in the RLHP equation?

[plus 3 golfer]

Read this as it gives a fairly complete insight into dyno testing.  I don't think one can compare "fairly" one component of the RLHP coefficients among vehicles.  It's the total RLHP that's comparable.  One can guess as to why certain coefficients are higher or lower.  Units are mph for V.

 

Here's the target coefficients (which are the C-Max HEV Aug. 2013 coefficients) used for the C-Max HEV testing in the link I provided.  I believe the OP's link is a summary of the tests in the Argonne link in this post.

 

Target A [lb]

21.75

Target B [lb/mph]

0.365

Target C [lb/mph^2]

0.01859

[Smiling Jack]

 

Read this as it gives a fairly complete insight into dyno testing.  I don't think one can compare "fairly" one component of the RLHP coefficients among vehicles.  It's the total RLHP that's comparable.  One can guess as to why certain coefficients are higher or lower.  Units are mph for V.

 

Here's the target coefficients (which are the C-Max HEV Aug. 2013 coefficients) used for the C-Max HEV testing in the link I provided.  I believe the OP's link is a summary of the tests in the Argonne link in this post.

 

Target A [lb]

21.75

Target B [lb/mph]

0.365

Target C [lb/mph^2]

0.01859

 

 

Thanks.

 

OK, I get it.  The coefficient values are from a curve fit.  Still, there's physics behind the curve being whatever it is, and the physics has apparently influenced the choice of the form of the fitting equation; so I don't think that some physical interpretation of the coefficient values would be too far-fetched. 

[plus 3 golfer]

Thanks.

 

OK, I get it.  The coefficient values are from a curve fit.  Still, there's physics behind the curve being whatever it is, and the physics has apparently influenced the choice of the form of the fitting equation; so I don't think that some physical interpretation of the coefficient values would be too far-fetched. 

Correct.

 

I've done coast down tests on my C-Max and one can fit coast down curves (time vs speed) that match the data at very, very high r squared levels.  But, when trying to relate the curve to the Road Loads that affect the curve: rolling resistance (function of V), friction losses (function of V^2), and Aero losses (function of V^3) using known vehicle data (Cd, weight, frontal area) and external data (pressure, road surface, and so forth), the resulting coefficients may not totally represent a specific component of road load - for example, coefficient A may represent most of the rolling resistance and coefficient B a lesser portion of rolling resistance.

 

As stated in the link, the derived road load coefficient that are incorporated into the road load, but coupled, are:

•Aerodynamic forces

•Rolling resistance

•Friction loses

 

So, we just need to be careful in ascribing changes between vehicles to a specific coefficients due to the coupling. 

 

Here's a table showing the RLHP for the Energi and Hybrid at various speeds.  To me the data makes sense given the difference in weight and final drive ratios.  The difference in final drive ratios should increase friction losses in the Energi over the Hybrid transmission - MG1 and MG2 spin when coasting.  Coefficient C should represent virtually all aerodynamic drag and should be virtually the same between the C-Max and Energi and it is.  But what we don't know is how much of the changes in coefficients A and B relate to weight and how much relate to friction losses - what's the coupling between the two (if any) in the coast down data.

 

Edited December 28, 2014 by Plus 3 Golfer

[Smiling Jack]

 

Correct.

 

I've done coast down tests on my C-Max and one can fit coast down curves (time vs speed) that match the data at very, very high r squared levels.  But, when trying to relate the curve to the various components that affect Road Load it is less precise as one tries to deconstruct the curve into various components - rolling resistance (function of V), friction losses (function of V^2), and Aero losses (function of V^3) that match known vehicle data (Cd, weight, frontal area) and external data (pressure, road surface, and so forth).

 

As stated in the link, the derived road load coefficient that are incorporated into the road load, but coupled, are:

•Aerodynamic forces

•Rolling resistance

•Friction loses

 

So, we just need to be careful in ascribing changes between vehicles to a specific coefficients due to the coupling. 

 

Here's a table showing the RLHP for the Energi and Hybrid at various speeds.  To me the data makes sense given the difference in weight and final drive ratios.  The difference in final drive ratios should increase friction losses in the Energi over the Hybrid transmission - MG1 and MG2 spin when coasting.  Coefficient C should represent virtually all aerodynamic drag and should be virtually the same between the C-Max and Energi and it is.  But what we don't know is how much of the changes in coefficients A and B relate to weight and how much relate to friction losses - what's the coupling between the two (if any) in the coast down data.

 

 

 

Thanks again.

That's a lot of info to digest, but I do notice right away that the B term HP is not dominant at any of these speeds.

[SPL Tech]

It would be interesting to see some other cars on that graph like a full size sedan, a pickup truck, SUV, full-size truck, ect. It would be interesting to see how the drag compares to other cars out there besides the Prius.

[Smiling Jack]

 

......

.......

........, coefficient A may represent most of the rolling resistance and coefficient B a lesser portion of rolling resistance.........

........

 

 

In support of this possibility, I note that the Energi is 6% heavier, and its A coefficient is 6% larger.  At least in the comparison of these two cars, it would seem that the coefficient A is proportional to weight.

[plus 3 golfer]

In support of this possibility, I note that the Energi is 6% heavier, and its A coefficient is 6% larger.  At least in the comparison of these two cars, it would seem that the coefficient A is proportional to weight.

Did you notice the B coefficient for the Prius V is negative.  That's an example as to why I said one needs to be careful in ascribing a physical attribute of a vehicle to a specific coefficient.  Sometimes, it makes sense and sometimes it doesn't.

 

Another example is in my testing of grille covers for the C-Max.  I posted in another thread that I found about a 7% improvement in the coefficient of drag (Cd) with grille covers on vs off when analyzing coast down data with and without the covers.  So, if we assume that coefficient C is 100% related to aerodynamic drag, then we should expect a reduction of aero RLHP of 7% when using grille covers (C-Max Cd = 0.30 without grille covers and 0.279 with grille covers).  At 60 mph from the table above, 7% reduction would be about 0.847 HP (0.07*12.1) or a reduction of about 4.4% in total RLHP (0.847/19.13).  If we assume that this 4.4% is proportional to an increase in FE, we might get a FE improvement of 4.4% with grille covers at 60 mph.  If our FE without grille covers is 50 mpg, it might be 52.2 mpg using grille covers due to reduced aero drag. :)  Unfortunately, I can't say the same about gas pods. :sad:

[fotomoto]

As has been noted, there's a loophole where the manufacturer can use the same mpg results for a specific drivetrain when used in other, different model lines.  When shopping for a pickup truck a few years ago, I've also found this same loophole used by Toyota but within the same model.   The same EPA rating was listed for all the tacoma models using the base 4cyl/5spd drivetrain; whether it was the tiny two door, the bigger extra cab or the substantially heavier and longer 4 door model.  

[bigalpha]

I find it hard to believe that FORD (especially their engineers) didn't know what was going on

 

Especially since the CMax has been produced and sold around world since 2003!