Performance Variations: Causes and Solutions

Hello everyone,

Bertil's 100 hp PD engine, which apparently produces closer to 115 hp, reminded me once again of the question of the causes of power variations.

With VP-TDIs, this can be quite well explained by "overfueling" the pump settings (whether from the factory or through modifications), where simply more diesel enters the combustion chambers. The often associated jolting or rough running issues should be disregarded here.

"As far as I know, there are no comparable adjustment options or sources of error in PD engines. Moreover, a technology that enables D4 emissions in mass production should only exhibit very small deviations between the target and actual injection amounts."

According to this, the additional power in PD engines should largely come "from the engine block" itself, and therefore be directly related to efficiency (more horsepower from the same amount of fuel injection inevitably implies a higher efficiency, in my opinion).

My question: What parameters or mechanical manufacturing/adjustment tolerances could cause a sudden difference of around 10% in power output or efficiency, which, according to Bertil's performance test, seems to be distributed relatively evenly across the usable speed range?

For example, if the valve timing were shifted relative to the crankshaft, any potential increase in power would likely be limited to a narrower RPM range, while the rest would decrease accordingly, so that the overall result of a dynamometer test would hardly be "too good" by around 10%.

Or are all AXR models (or even all PD models) as powerful as Bertil's example?

Hi Ulf,

I might also phrase the question like this:

What are the differences between a properly broken-in engine and a poorly broken-in engine?

On the other hand, I think I've broken in my 90hp ALH engine properly. But even so, it doesn't run particularly well.

Perhaps someone has some advice...

Regards,

Eike.

Hi,

While I unfortunately didn't get to see what was discussed beforehand, when I see "115PD PS," I immediately think about factors like weight, the type and dimensions of the tires, fuel quality, outside temperature, additional power consumers, a possible slight downward slope of the test track, the condition of the engine, and driving style (e.g., slipping the clutch, shifting gears aggressively), etc.

I think the list of factors can be expanded indefinitely.

At least, I've read that you always get different values on various dynamometers. I've never been to one.

I can't imagine, no matter how hard I try, that the manufacturing tolerances for engine parts are so large that the performance of some engines of the same type could be 15 horsepower higher than that of others.

When I think about it, even with a gasoline engine, a "performance camshaft" can sometimes only result in a ridiculous 8 horsepower gain, and that's with intentional tuning. It's very hard to imagine that, with a TDI engine, the manufacturing tolerances alone could account for a 15 horsepower increase.

The human factor alone, as the driver, introduces so many variables that measurements taken in this way are too inaccurate.

Or am I making a mistake in my thinking?

Regards,
Thomas.

Thomas wrote:

... While I unfortunately didn't get to see what was discussed beforehand, when I see "115PD PS," I immediately think about factors like weight, the type and dimensions of the tires, fuel quality, outside temperature, additional power consumers, a possible slight downward slope of the test track, the condition of the engine, and driving style (e.g., slipping the clutch, shifting gears aggressively), etc. ...

Hi Thomas,

It's about my car and the thread: /viewtopic.php?t=1479

@Ulf

Regarding the question you asked, I'm also at a loss. That was also the reason for the thread I mentioned earlier.

eike wrote:

On the other hand, I think I have properly broken in my 90hp ALH engine.

But even so, it doesn't run particularly well. }

Hi.

On what criteria do you base your assessment of "not particularly good"?

Hello!

After attending a few lectures on automotive technology, I now have some new information. The power losses in the transmission should be approximately 2-5% for manual transmissions, and 6-9% for automatic transmissions.
The moment of inertia of the axles and wheels is more likely to be around 5 kgm².

And it would be possible to introduce a slip factor into the DZR (Dynamic Roadholding Coefficient). This is difficult due to the different tire types and axle loads, but if you simply subtract 2-4% from the vehicle speed column, the DZR value becomes more accurate again.

Overall, the DZR (Dynamic Rail Injection) shows a slightly better time, but the performance decreases by 4-5%. My 1998 Passat with the AHU engine is estimated by the DZR to have a performance difference of +/- 0% (according to the error database, the 1998 models are the weakest TDIs).
And Bertil's 100 hp engine, when tuned, will likely produce around 108 hp (a rough estimate)... That sounds reasonable, doesn't it?
Best regards, Ernst.

Ernst S. wrote:

And one could introduce a slip factor into the DZR (Dynamic Roadholding Coefficient). This is difficult due to the different tire types and axle loads, but if you simply subtract 2-4% in the vehicle speed column, the DZR value becomes a bit more accurate.

Hello Ernst,

I was already thinking, "Where is he?"

The issue with the slippage seems questionable to me, especially during the 2000-4000 RPM test in 3rd gear.
Slip of x% means, as far as I know, that the tire is rotating x% faster than the car is actually moving due to the transmitted force.
Then, at the end of the acceleration test, when you release the accelerator (or better yet, just disengage the clutch, allowing the tire to rotate freely = without slippage), the speedometer needle should drop from 100 to 96-98 km/h according to your values.

I have not been able to observe anything like that , even in the slightest way, although I have paid close attention to it several times.

To address a potential argument about the inertia of the needle in stepper motor tachometers :
I've often turned off the ignition during descents and then turned it back on. This causes the speedometer needle to drop to zero and then rise again. And it does so quickly, which I believe could very plausibly and noticeably show a "jump" of 2 to 4 km/h.

Sind die 2 - 4 % Schlupf nicht etwa auf maximal gängige Beschleunigungen im PKW-Bereich bezogen, d.h. auf den 1. How's it going?
It's roughly 3 times shorter than the 3rd gear in a TDI, which means only about 1/3 of the pulling power is available. Consequently, the slip (assuming it's proportional to the force) should be closer to 1%.

It's quite possible that someone might overlook something like that on the speedometer when disengaging the clutch at 100 km/h - although I claim that, with careful observation, one can clearly recognize even a change of 1 km/h (and distinguish it from the uniform speed loss due to air resistance).

Hi!

I'd like to get your tires: So, my cousin, who drives the same car as you, manages to make the tires spin when he's in first gear and then floors it. --> It's almost 100% wheelspin.

In 3rd gear, you have approximately 3000N of tractive force. With an axle load of 5000N (estimated for the dynamic load, as it lifts the front end during acceleration).
Therefore, a friction coefficient of 0.6 is required. Older tires can achieve this with approximately 5% slip. Newer tires require less.

'And the speedometer needle is too slow (too heavily damped): I noticed this during a test drive with a 130PS TDI. When I increased the engine speed from 2600 to 3000 with full throttle, I watched as the speed slowly increased from 85 to 97. By the time the engine speed reached 3000, the speedometer was still at around 94. Therefore, it's impossible to detect wheel slip with this speedometer.'

In the DZR (Dynamic Road Load Calculation), it's not a problem to account for slip. You can simply enter a factor in the vehicle speed column. This factor could also be made dependent on the force and weight to adapt it for all gears and vehicles. It won't be perfectly accurate, as it's also dependent on the road surface. But it's better than completely ignoring slip.
Best regards, Ernst.

On the other hand, I think I have properly broken in my 90hp ALH engine. But even so, it doesn't run particularly well.

Hi.

Based on which criteria do you determine that something is 'not particularly good'?}

Hello Ulf,

I base this on the fact that it simply performs worse than many of the similarly equipped vehicles owned by my colleagues.
My colleague's G4 runs without any problems up to 195 km/h. Mine only reaches that speed when it's performing well. Usually, it's limited to 188 km/h according to the speedometer.

Regards,
Eike.

The speedometers all behave differently. If your colleague's speedometer reads 195, it could be just as fast as if yours reads 188. Compare the RPM at your maximum speed with your colleague's. If his is truly faster, then he must be reaching a higher RPM.

Ernst S. wrote:

I would like to have your tires: So, my cousin, who drives the same car as you, manages to make the tires slip when he's in 1st gear and then gives it full throttle. --> approximately 100% slip

.


Hello Ernst,

Well, I generally avoid full-throttle acceleration in first gear because of the stress it puts on the materials.

Quote:

In 3rd gear, you have approximately 3000N of tractive force. With an axle load of 5000N (estimated for the dynamic load, as it lifts the front end during acceleration). Therefore, a friction coefficient of 0.6 is required. Older tires can achieve this with approximately 5% slip. Newer tires require less. And the speedometer needle is too slow (too heavily damped): I noticed this during a test drive with a 130PS TDI. The engine speed increased from 2600 to 3000 with full throttle... Then I watched as the speed slowly increased from 85 to 97. When the engine speed was already at 3000, the speedometer was still at approximately 94. Therefore, it is impossible to detect slippage with this.

These are clearly defined data.
However, even with a sluggish tachometer needle, after releasing the clutch at 100 km/h, something should still happen: either the needle should continue to rise to "catch up" with the actual speed, or the amount of slippage (plus the speed loss during deceleration) should decrease.
I find it extremely unlikely that both effects will cancel each other out to the point of being visually imperceptible.
I will be conducting some more targeted experiments in that regard soon.

Quote:

In the DZR (Dynamic Road Load Calculation), it is not a problem to account for slip. Simply enter a factor in the vehicle speed column. This factor could also be made dependent on the force and weight to adapt it for all gears and vehicles. It will not be possible to achieve perfect accuracy, as it is also dependent on the road surface. But it is better than completely ignoring slip.

Sure, that would be one of the easier exercises with the DZR.
To avoid making the process too complicated, I would use the vehicle weight and the traction force at the wheel at 3000 rpm as the basic parameters. From these, an automatic default slip factor could be calculated (and displayed in the internal calculations) to correct the starting speed, and that would be it.

But then I just remembered: Slippage creates additional friction, or "waste heat," which the engine has to provide as mechanical power in the first place.
Therefore, after subtracting the speed loss caused by wheel slip from the required power, the corresponding "energy loss" should be "credited" back to the motor.

Do you have an idea of the magnitude of the loss - or could both factors potentially cancel each other out in the end, so that no slip correction would be needed in the DZR ?

Hi Ulf!

At higher speeds, the rolling resistance increases slightly. It goes from approximately 0.013 times the weight force at 20 km/h to approximately 0.016 times the weight force at around 200 km/h. (This information is from a diagram).
I think there has to be some slip involved. Even at a constant speed of 200 km/h, you have 1.5-2% slip because the air resistance has to be balanced by the driving force.

Therefore, the reduced performance due to slippage is likely compensated for by the additional power generated through the 'slip work.'
According to my calculations so far: a 3% reduction in performance due to slippage and a 1% increase in performance due to higher rolling resistance during acceleration.

In the current DZR (Detailed Design Requirements), there are some items that would result in a 2% change... so, the slippage could also be included.

Regarding the speedometer inaccuracy: It seems to me that the slip is already taken into account in the speed display. Although, logically, this shouldn't be possible, because the speed is derived from the rotational movement.
Every speedometer I know lags behind the tachometer. To the point where, by the time the desired speed is reached, the slippage would already have to be compensated for.

Best regards, Ernst.

Ernst S. wrote:

Therefore, the reduced performance due to slippage is likely compensated for by the additional power generated through the "slip work." According to my previous calculations:

3% less power due to slippage and 1% more power due to higher rolling resistance during acceleration.

Quote:

Hi Ernst, Does that mean you want to recalculate/rethink/research the matter more thoroughly?

Every speedometer I know lags behind the tachometer. To the extent that, before reaching the speed, the slippage would already have to be compensated.

So, you mean that the digital tachograph reacts faster to changes in the display value than normal tachometers?
That would mean "too low" km/h values during acceleration, and conversely, "too high" km/h values during a braking process.

However, this shouldn't matter during a properly conducted "acceleration" test, because we defined it as a time range of 2000 - 4000 rpm, but not, for example, as a time range of 50 - 100 km/h }.

So, you mean that the digital tachograph reacts faster to changes in the displayed value than regular tachographs?
That would mean 'too low' km/h values during acceleration.

That's exactly what I'm getting at. I observed this in a Passat with a 130PS TDI 4Motion engine. And it was quite clear. The tachometer was spinning very quickly, the speedometer lagged behind during acceleration, and then increased the value even after I had already stopped accelerating.
Maybe there was also a fault somewhere... For example, the Climatronic display only increased after 1 second of pressing the button. Could it be that the control unit in that vehicle is already somehow overloaded?

I haven't been able to detect any speedometer inaccuracies in a 90hp Passat TDI. However, it might be because the car doesn't really accelerate much, considering its weight of 1480kg.

It's quite certain that there's some slip, approximately 2-6% (depending on weight, traction force, and tires), regardless of what the speedometer indicates. Tires simply need to slip to transmit power. Here's a slip curve illustrating this.
http://www.alfisti.net/german/ratgeber/technik/fahrverhalten.htm


However, this shouldn't matter during the 'properly executed' acceleration test, because we defined it as a time range of 2000 - 4000 rpm, but not, for example, as a time range of 50 - 100 km/h.

However, the time is determined by the difference in speed/acceleration.
And the speed range is slightly reduced due to the slippage. This also results in a slightly lower amount of air resistance... which translates to a total of 3% less power.

I'm not entirely sure about the extra work that results from the slippage.
Perhaps I can find out more next week.
I wouldn't mind if the DZR (Differential Drehzahlregelung - Differential Speed Control) completely compensated for the reduced power output caused by wheel slip.

Best regards, Ernst.

Ernst S. wrote:

In a 90 horsepower Passat TDI, I haven't been able to detect any speedometer inaccuracies. Perhaps this is because it doesn't produce much acceleration anyway, considering its weight of 1480 kg.

Hi Ernst,

The description seems a bit too harsh to me.
Then my car wouldn't produce any acceleration either. Perhaps older speedometers (with less electronics around them) actually show the speed more accurately.

Quote:

I'm not entirely sure about the extra work that arises due to the loophole. Perhaps I can find out more next week. I would have no objection if the DZR (Differential Locking System) completely compensated for the reduced power due to wheel slip.

Yes, please keep us updated.
See you later.

Hello Ulf,
Well, it's undeniable that a car's factory-set performance limits exist.
Take a look at the article!

http://www.amb-tuning.de/info/chiptuning.html


Sure, here is the translation of the text from German to English:

'Hello'