Rolling Resistance: Explained & Reduced | Posts 16+

P = F x v, where F, the rolling resistance force, remains constant.


'Constant is logical... I would have assumed that higher speed would result in lower rolling resistance (not power consumption), but in reality (according to literature), it's the opposite. However, if you read my previous posts, it will make sense.'

The value you measured for 'rolling resistance' is almost ten times too high (according to literature). Your rolling test also measures the 'walking' effect of the tires, friction in the bearings, half-shafts, axle ratio, and the output side of the transmission is also rotating. And the actual rolling resistance is the smallest of these resistances.

So, forget about rolling resistance (I won't be upset if my 'great' calculation from the DZR is deleted ).

Best regards, and I reserve the right to be wrong.
Ernst.

Hello,

Just a thought... if the torque values of our tested TDIs are higher than the torque values calculated by the "pulling calculator" at engine speeds below the rated power speed, that would, in my opinion, explain the shorter times.
Because the rated power can still be the same, even though the acceleration is significantly faster at low speeds.
I suspect that the actual torque curves of our TDIs differ more significantly from each other in practice, as well as the underlying curve used in the drag calculation software.
The problem is that you can only measure the total transit time with a stopwatch, but not the individual delta-v values per unit of time.
I have started working on a suitable program for the diagnostic interface, but the project is currently on hold due to my broken laptop.
It would also be possible to use a PIC16C84 microcontroller on the transmission's speed sensor, allowing the measurement values to be read out later via a serial interface. In that case, a diagnostic adapter would not be necessary, and a laptop would not be absolutely required either.
I cannot yet estimate the effort required to replicate the communication with the engine control unit for speed retrieval. Based on the recorded data that flows through the serial interface and my comparison with the documentation available to me, I have not yet gained any clarity about what is happening.

Best regards, Rainer.

Good idea, Rainer!
For example, on websites like tuneline.at, you can find performance curves with statements like that. In one case, the AFN engine, although specified as having 235 Nm of torque, actually produced 245 Nm in the curve, and the torque only drops below 235 Nm at around 2800 RPM. However, the peak power is at the specified nominal value. It would be really amazing if one could measure the actual curve themselves.
Regards,
Ernst.

Ernst S. wrote:

Only the value for your measured "rolling resistance" is almost ten times higher (according to literature). Your rolling test also measures the walking work of the tires, friction in the bearings, half-shafts, axle ratio, and the output side of the transmission is at least also rotating. And the actual rolling resistance is the smallest of these resistances.

Hi Ernst,

Why should the rolling behavior of tires, specifically the tire tread , not be categorized under "rolling resistance"?

Due to the mere differences between my winter tires and a worn-out and a new set of summer tires, with a difference of over 20% in rolling braking force, I suspect (meaning "I don't know" *g*) that the rolling resistance is largely influenced by the tire compound.

Or, what "braking forces" do you understand by the term "rolling resistance"?

Rainer K. wrote:

If the torque values of our tested TDIs are higher than the torque values in the pull calculator at engine speeds below the rated power speed, this would, in my opinion, explain the shorter times. Because the rated power can still be the same, even though the acceleration is significantly faster at low speeds...

Of course, that could be the case.
The DZR (Dynamic Torque Response) increases the entire torque curve by the specified percentage with the multi-power input.

If a test run, for example, indicates a consistent increase of 10% in torque, but we are assuming a normal peak power output, then the "missing" Nm at the upper end of the measurement range must essentially be doubled in the lower measurement range.

In the case of the AFN engine, it wouldn't be 235 Nm + 10% = 258 Nm, but rather approximately 282 Nm.

And with our ALH Ibiza, which supposedly has 15% more power, it already produces 273 Nm - a figure that is usually quoted by more "reputable" power box suppliers for the tuned 110 hp engine... which makes me think this explanation is rather unlikely.

In the DZR, you can also create your own curves by entering the corresponding values into the yellow torque column.
Only the specific values to use are almost like a lottery .

Hi Ulf!

For the definition of rolling resistance, please refer to the information above (which I took from a vehicle dynamics textbook), as it does not actually include the walking work.

'However, you measured the rolling resistance correctly, and I completely messed up the calculation (I'm embarrassed, I forgot to use 9.81 to convert from kg to N). So, it's my fault... I'll correct it right away.'
The only thing I can imagine is that, according to the book, the rolling resistance might not be fully established yet at the low speeds in your acceleration test (after which it's constant). And regarding the acceleration, you're dealing with the mass, and the rotating mass of the wheels also has to be decelerated (yes, I know... it's probably not important). I just want to say that the rolling resistance is probably even a bit higher (0.15 * G) than what you measured, even though I claimed the opposite before.

This brings us back to the topic of the thread: Every time I try to correct the DZR to compensate for excess power... I end up finding even more excess power . My professor, like Ulf, also said that the losses of the entire powertrain are probably closer to 14% than to 12%.
Well, I'm starting to believe that, despite the inaccuracies of the DZR (German term, likely referring to a specific document or data), it still points to significant improvements.
I'm definitely too inexperienced to find any more errors in this code, maybe someone else will have better luck.
Regards,
Ernst.

Hmm, Ulf is right. Maybe I'll have time to analyze my performance data sometime.
The torque curve should also be calculable using the values from the airflow tests performed during the air filter test (available for download in the technical article). Who has the time and inclination?

Best regards, Rainer.

Hi everyone,

I once compared the DZR (drag coefficient) with a driving simulation diagram from VW for the 81 kW Golf 3.
According to the diagram (which shows "official" figures of approximately 10.8 seconds to 100 km/h, although without any noticeable gear changes in the curve - that poor virtual transmission! *chuckle*), it takes around 4.9 seconds in 3rd gear to accelerate from 72 km/h (shift point from 2nd to 3rd gear) to 100 km/h.

According to the acceleration calculator (12% friction losses), the car, with its empty weight and without any additional power, accelerates from 72 to 100 km/h in just 4.1 seconds. (Since 72 and 100 are not exactly within the calculation intervals, I have computationally interpolated at the ends.)

Converting the ratio of 4.9 to 4.1 seconds to the 2000 to 4000 rpm range would result in approximately 8.16 seconds instead of 6.83 seconds. This translates to roughly 14% less power compared to the VW's performance chart, or approximately 240 kg of additional weight on board, in addition to the driver. Alternatively, it could indicate a total frictional loss of 26% instead of 12% (the difference of 14% corresponds to the power loss – amazing, isn't it *gg*).

The commonly observed, reduced power values of 81kW TDIs in the G3 weight class, consistently showing 6.3 seconds (instead of 6.83) for the 2000-4000 rpm test, indicate a general power increase of approximately 8%. Adding this to the aforementioned 14% discrepancy between the dyno tests and the VW chart results in a total discrepancy of 22%!

To further complicate things, here's a comment from "Power-Paule" (long-time forum users might remember him) regarding efficiency measurements obtained on a dynamometer:
Transmission, axle drive = 0.91 - 0.96
Tires, wheels = 0.81 - 0.86
Overall efficiency = 0.74 - 0.83.

The rolling resistance of the tires I'm using consumes approximately 4 kW at 100 km/h (P = F x v), while the motor is delivering its maximum power. Therefore, of the 81 or 66 kW, 4 kW would be lost in the tires, resulting in an efficiency of approximately 94 to 95%, which corresponds to a 5 to 6% loss due to the tires.
Combined with the remaining DZR loss allowance of 12%, this results in a total loss of 17-18%, leading to an overall efficiency of 82-83%, which represents the most favorable scenario within the range mentioned by pp.
However, his explanation does not include the power consumption of the rotating masses in the vehicle, which Ernst has, as he mentioned before, supposedly increased by a few percent .
However, this should roughly balance out if we ignore the rotating masses and instead assume an average overall efficiency of around 78% for the pp system.
I certainly don't want to "beat around the bush" when it comes to removing those rotating masses from the DZR! On the contrary, I believe they belong there.

@Ernst: Perhaps your professor could take a look at the DZR? Maybe they would find the crucial detail in the story...

@Rainer, regarding "your" torque curve:
I'm not sure how much help this would be, considering how much your power supply is likely interfering, and that the mass airflow sensor control is probably causing the most enrichment at the top end.
I'm more interested in first clarifying the basic specifications, i.e., the performance as delivered from the factory.
So, if you were to record a test run without the enclosure, I think the analysis would be more worthwhile.
Regardless, I don't have time to analyze any curve right now.

Perhaps I can mention the additional performance figures in my VKM exam on Wednesday... specifically, what percentage they represent on average.

I will get much more information in September when I (hopefully) will be working on the dynamometer at my university. There, I will also be able to present the DZR (presumably a technical document) and have it checked.

Until then, I won't write anything more about that topic. Ulf has already done such a good job, and the chance that I might make things worse with a change is simply too high .

But the issue with the driving simulation program from VW speaks volumes. I also came across the concept of the moment of inertia because the DZR (digital tachograph) was showing me too little time. And now, VW's software shows even more . This should at least prove that the DZR still tends to show too little power, rather than too much.
Greetings.
Ernst.

Hello Ulf,

I certainly didn't want to "force" an analysis on you. My main concern was the possibility of generating a torque curve from the measurement data. Hopefully, I or someone else will find the time to do that soon.
Unfortunately, I won't be able to provide those values directly, as my modifications are too extensive.

Best regards, Rainer.

Rainer K. wrote:

I certainly didn't want to shove an 'analysis' at you.

That's not how I understood it either .

Quote:

My main goal was to explore the possibility of generating a torque curve from the measurement data. I hope that I or someone else will find the time to do it soon.

Calculated from the hip: Convert the measured time for an interval according to the DZR (German standard for engine performance) within the same RPM range (e.g., x seconds / 100 rpm). Based on the ratio of the measured time x and the DZR calculation used, convert the Nm (Newton-meter) value to the corresponding DZR row – and you have a rough estimate.

Quote:

Unfortunately, I won't be able to provide serial numbers so easily, as my modifications are too extensive.

If you just remove the box, shouldn't it be possible to almost restore the original condition? Or have I forgotten some essential things about your car?