Some time ago, the car magazines used to routinely include, in their road test results, estimates of road horsepower vs speed and/or road horsepower broken down into "rolling resistance" and "aerodynamic drag."

In this connection, the term "road horsepower" was used to mean the amount of power necessary to propel the car down the road at a constant speed (with no wind or incline).

Often, these were determined from "coast-down" tests.

Has anyone seen such data for the TCH?

Has anyone doe a coast-down test on their TCH? (It's not difficult to do. All you need is the right road conditions, a speedometer, a stopwatch and a buddy.)

Has anyone seen published road horsepower data from any source for the TCH or any late model Camry? (or even an ES350? )

 

Smilin' Jack — I haven't seen any published data like you're looking for on the TCH. But I have done some modelling of the TCH, using Toyota's published figures and formulas gleaned from available texts. For example, here are my formulas and parameter values (in metric units) for the TCH. Here Pss is the steady-state power (in kW) required to overcome air resistance and tire rolling resistance at speed v (in km/h) on a level road; and Lper100km is the corresponding FC (in L/100 km):

# rho=density of air (kg/m^3); A=car's effective frontal cross-sectional area (m^2); Cd=car's coefficient of drag; epsilon=ICE's thermodynamic efficiency including transmission losses (fraction) — epsilon1 is at low speed (65 km/h) and epsilon2 is at high speed (100 km/h); v=car's speed (km/h); w=wind speed (km/h); alpha=wind direction (rad CW from car's direction of travel); R0=tire rolling resistance coefficient; m=car's total mass (kg); g=acceleration of gravity (m/s^2)

rho:=1.225; A:=2.30; Cd:=0.27; epsilon1:=0.185; epsilon2:=0.33; R0:=0.008; m:=1919; g:=9.8; #2007 Toyota Camry Hybrid

Pss = (rho*Cd*A/2*(v/3.6)^3 + R0*m*g*(v/3.6))/1000

Lper100km = 45/4*Pss(v)/epsilon/v

The graph of Pss versus v is attached for your information.

Stan

 

... and a partridge in a pear tree! LOL!

 

Cool!

Not that I understand it all, but . . .

Cool!

 

Thanks very much to Stan for this highly useful info.

This pretty much covers my question for "otherwise"

I'd still be very interested in the coast-down results if anyone has the info.

Jack

 

It would be interesting to see such a graph with the upper and lower bounds of R0 based on different tire pressures. Have a guess as to how much R0 would vary from an under-inflated tire to an over-inflated tire?

 

Since tire rolling resistance loss is directly proportional to R0, the rolling resistance coefficient, it's easy to estimate how it is affected by changes to R0. Reliable values for R0 don't seem to be published, however. I used a guesstimate of 0.008 for the "low rolling resistance" Michelin Energy tires, based on some numbers I found. Maybe 0.009 would have been a better number to use?

It may be useful to break out the two components of Pss separately:
Pss = Pair + Prr
where Pair is the power loss due to air resistance, and Prr is that due to tire rolling resistance. Since Pair is proportional to v^3, whereas Prr is proportional to v, air resistance loss dominates the total loss Pss above ~100 km/h. Nevertheless, tire rolling resistance loss is dominant at in-town speeds below ~70 km/h, and is a very significant contributor to the total power loss at all speeds. [You may be wondering why Pair is proportional to v^3 and not v^2. The air resistance force is usually taken to be proportional to v^2, but power = force x velocity, and this introduces the third v factor.] I am attaching a new graph that breaks down Pss into its two parts Pair and Prr.

Stan

 

Stan,

Thanks for the additional info.

A few questions:

To do the fuel consumption estimate at various speeds, would you interpolate for epsilon ? (Obviousy that won't do for an extrapolation beyond 100 kph because past peak torque rpm the efficiency decreases with higher engine speed and presumably at some point with higher road speed.)

I suppose we could go a step further and back-calculate engine efficiency as a function of engine speed and power output from a set of BSFC curves if we had those data. I've seen references in this forum for the bsfc data for the Prius. Ever see any for the TCH?

With such efficiency vs. RPM and power output, we could combine further with the Pss vs. road speed formula and your equation relating the component RPM's and road speed to calculate estimates of fuel consumption vs. steady road speed either for the normal mode or for the Heretical mode. You ever consider doing that?

Jack

 

Andy,

It certainly would be interesting, and that's exactly the sort of question that we could answer experimentally (and quite precisely) with the coast-down test.

Jack

 

Smilin' Jack — Those two numbers for epsilon were just two numbers that I came up with when seeing how well I could match the predictions of the Maple worksheet (from which I excerpted the formulas above) with some fuel-consumption data that I gathered with my TCH. My main purpose at the time was to see if I could explain the consistent difference in fuel consumption that I found when driving the 125-km route from Waterloo to Toronto, Ontario, versus what I found when driving the same route in the reverse direction. After allowing for the elevation change, I satisfied myself that the difference was roughly consistent with the effect of the prevailing moderate westerly winds that occur in this region. The higher value epsilon2=0.33 was what I had read somewhere (I forget offhand where) for the Atkinson-cycle ICE's thermodynamic efficiency — high by ICE standards. The lower value epsilon1=0.185 gave a better match with my measured fuel-consumption data taken at 65 km/h. One could probably interpolate for intermediate values of epsilon, but I don't really think that my values have any real solid validity to them. No, I've not seen any BSFC curves for the TCH's ICE.

I have thought of trying some further experiments to refine the epsilon values, but haven't done any yet. Low-speed measurements (where air resistance is negligible) could refine the R0 value. High-speed measurements could then be used to refine the value of epsilon (around that ICE operating point). I encourage you (and others) to do some careful experiments.

Stan