Schlechterer Motorwirkungsgrad nach Tuning? | Posts 16+

bafische wrote:

Hehe - Ulf, you still don't understand me! Please detach yourself from those old injection amounts in the dynamometer. Please compare the actual power curves - in real-world conditions!!! Here you can see how the ratio of increased fuel quantity to increased power looks like and whether this is constant across different engine speeds.

Oh... the problem is that, besides the official figures for torque and power at 1900 and 4000 rpm, I don't have any other "official" data.
But maybe your mystery bag will reveal something else?

Quote:

If you are operating in typical Western European outdoor temperatures and your low-level cut-off (LLK) with a Duosensor does not exceed 60-70°C at the nominal power point for VL, then your tuning should not thermally overload the motor (assuming you have set up the same conditions as the BPX).

Okay, with the cooling system running at its maximum capacity, I'm now getting 50 kilowatts of "net" power at 4000 rpm and a top speed of 160 km/h.
So, with an outside temperature of 20°C, the outlet temperature is approximately 70°C, and at 30°C outside, it's around 80°C...

Quote:

I find the consumption comparisons at 85 and 110 kW in the DOC much more interesting. Here, despite all the hardware-based measures to improve efficiency in the 100 kW system, one can clearly see the disadvantages of the extremely high peak power in the specific consumption at higher speeds.

"You're right. In that case, the ARL is slightly more fuel-efficient below 2000 rpm, although the low-temperature catalytic converter probably doesn't play a significant role there yet, because the exhaust gas temperature (EGT) only starts to increase rapidly around 2500 rpm (in my experience)."


EDIT:
You really only need to compare the quantity-performance ratio when looking at the key specifications; that usually gives you a good general idea of what's going on.

Maximum torque at 1900 rpm.
-> At ASZ, the torque was 310 Nm with a force of 56.0 mg, resulting in a specific torque of 5.54 Nm/mg.
-> With the BPX system, a torque of 330 Nm was achieved with 60.0 mg, resulting in a value of 5.50 Nm/mg.

Pmax
-> With the ASZ engine, you get 130 horsepower at 4000 rpm and 229 Newton-meters of torque with a compression ratio of 47.0:1, resulting in 4.87 Nm per unit of compression.
-> With the BPX engine, you get 160 horsepower at 3750 rpm and 300 Newton-meters of torque, resulting in a specific output of 5.08 Nm per milligram (59.0 mg).

According to this, the BPX has an efficiency that is approximately 1% worse than the ASZ at maximum torque, but approximately 4% better at Pmax .

"That contradicts the efficiency relationship between AJM and ARL at higher speeds, according to your diagrams... I'll check the performance curves again to see if there's an explanation (that I may have previously overlooked)."

ulf wrote:

You really only need to compare the quantity-performance ratio when looking at the key specifications; that usually gives you a good general idea of what's going on. Maximum torque at 1900 rpm. -> At ASZ, the torque was 310 Nm with a force of 56.0 mg, resulting in a specific torque of 5.54 Nm/mg. -> With the BPX system, a torque of 330 Nm was achieved with 60.0 mg, resulting in a value of 5.50 Nm/mg. Pmax -> With the ASZ engine, you get 130 horsepower at 4000 rpm and 229 Newton-meters of torque with a compression ratio of 47.0:1, resulting in 4.87 Nm per unit of compression. -> With the BPX engine, you get 160 horsepower at 3750 rpm and 300 Newton-meters of torque with a boost pressure of 59.0 bar, resulting in a specific torque output of 5.08 Nm per milligram of boost pressure. According to this, the BPX has an efficiency that is approximately 1% worse than the ASZ at maximum torque, but approximately 4% better at Pmax . This contradicts the efficiency relationship between AJM and ARL in the higher speed range, according to your diagrams... I will check the performance curves again to see if there is an explanation (that I have previously overlooked).

Hi Ulf, if you're only comparing individual points, make sure they have the same rotational speeds relative to each other.
I also wanted to point out how the power increase is distributed across the RPM range. I want to go back to the initial question: why doesn't your engine with the "BPX" application have 21% more power throughout the entire RPM range, but only "at" the optimal point at 4000 RPM? The power increase at, for example, 2500 RPM is only about 10%. That's why the dynamometer will never display absolutely accurate values.
It is highly unreliable to determine differences in efficiency based on the relationship between published performance data and actual production volumes, as there are too many uncertainties regarding the physical output quantities, among other things.
It's still interesting though. I'll search for and find the performance curves of both engines for all speeds, and you can then compare them with the quantity and publish it here.
In my opinion, the BPX should have a slightly lower efficiency compared to the ASZ, despite its potentially better LLK (liquid cooling) and possibly adjusted ATL (air temperature limit).

Is it already known whether the BPX or ARL ATL has a different running gear geometry? I mean, the "G" index ATL supposedly only has friction-optimized bearings – does anyone know more about that?

bafische wrote:

The performance gain at, for example, 2500 RPM is only about 10%. Therefore, the dynamometer will never display absolutely accurate values.

Hi Bavarian,

I can also compare it using a different method:

Based on the standard values, I can adjust the torque curve proportionally to the respective increase in quantity, i.e., assuming a constant efficiency.

If I adjust the dynamic resistance (DZR) for the series performance by reducing the weight in such a way that, based on the theoretical data, my measured time is calculated.
(so that the "too short" real-time performance is achieved not through the usual factory-installed enhancements, but through virtual overclocking).
"Then, if I enter the tuned torque values calculated based on a specific quantity adjustment, it calculates a time that I cannot actually achieve in reality."

To adjust the calculated tuning time to match the measured value, I need to enter a correction factor of 4% power reduction (or torque reduction) in the engine control unit (ECU).

This means that my engine, with the tuning compared to its stock configuration, loses approximately 4% in efficiency between 2000 and 4000 rpm.

Or am I making a fundamental conceptual error here?

Quote:

Is it already known whether the BPX or ARL ATL has a different running gear geometry? I mean, the "G" index ATL would only have a friction-optimized bearing system - does anyone know more?

Regarding the ARL loader, I recall something about it being "flow-optimized." Unfortunately, I no longer know the exact source; it was likely just an article in some popular magazine (not a technical journal like MTZ, etc.).

Regarding the LLT (Language Technology):
You write that temperatures of 60-70°C should not be exceeded as much as possible according to the manufacturer's instructions (LLK).
My Polo came from the factory with a Naturally Aspirated Supercharger (NSW) in the low-temperature coolant (LTC) air intake, which allowed it to achieve a net LLT (likely referring to lambda value or similar measurement) of over 65K @ 4200 rpm (without any tuning).
Even at an ambient temperature of approximately 15°C, the boost pressure was reduced from around 3500 rpm onwards, deviating from the normal performance curve.
The constructors attribute the following characteristics to the ASZ in the Polo:

Simply eliminating the turbocharger reduces the net LLT to approximately 58 kPa at 4200 rpm (without tuning).

By modifying the cooling air intake, I now have approximately 52 kW at 4200 rpm - including a boost pressure increase of 0.1 bar (which would certainly make any naturally aspirated engine in a Polo with turbocharger technology very desirable ).

I would like to further reduce the LD (load deviation) above 4000 rpm because there is still approximately 8 mg of diesel and 140 mg of air mass in the exhaust "unnecessarily" (a mass ratio of at least 1:15.9 is required in the exhaust).
Or would you rather interpret the cloudiness as being caused by excess air and leave the LDR (liquid-to-dry ratio) unchanged?

Or am I making a fundamental conceptual error here?


Well, just the basics, I guess... old song!

DZR is not a reliable measuring instrument; I've heard and read from multiple sources that this device can calculate scatter values of up to 20 horsepower, which don't actually exist. It only works when calibrated on a dynamometer; everything else is just guesswork.

Julian wrote:

DZR is not a reliable measuring instrument; I have now heard/read from several sources that this device calculates scatter values of up to 20 horsepower, which do not actually exist. Tuning should only be done on a dynamometer; the rest is just nonsense.

... because test benches are always so accurate .
http://www.hs-elektronik.com/leistungspruefstand.html

... because test benches are always so accurate .
http://www.hs-elektronik.com/leistungspruefstand.html

Come on, Ulf... Test benches are definitely light-years more accurate than your ASZ-DZR guessing machine, provided they are used with the same environmental variables and in a professional and correct manner.

ulf wrote:

My Polo originally came with a New Standard Wiring (NSW) system in the low-temperature coolant (LTC) air intake, which allowed it to achieve a net LTC torque of over 65,000 Nm at 4200 rpm (without any tuning). Even at an ambient temperature of approximately 15°C, the boost pressure was reduced from around 3500 rpm onwards, deviating from the normal performance curve. So, the constructors attribute these characteristics to the automatic transmission system in the Polo

.

That's outrageous. It's hard to believe that something like this made it into the series. It's possible that, due to the good performance-to-weight ratio and the lowest possible cost for adapting the ASZ engine to the Polo, no changes were deemed necessary.

By the way, have you ever thought about adding a water spray cooling system to the liquid coolant (LLC)? I'm talking about an open system. Subaru even uses this on their turbocharged models as standard equipment.

Quote:

I would like to further reduce the LD above 4000 rpm, as there is still approximately 8 mg of diesel and 140 mg of air mass in the exhaust "unnecessarily" (in the exhaust, a mass ratio of at least 1:15.9 is required). Or would you rather attribute the cloudiness to more air and leave the LD unchanged?

Be careful not to rely too much on the HFM (Humidity Flow Meter) when tuning, or ideally, never at all. In principle, you are right; a Lambda value of 1.15 at idle is acceptable given the conditions – I'll explain more about that shortly.
Trust your own judgment; try driving alongside a vehicle with a trailer on the highway, or even better, observe it yourself and look at the smoke trail. If the smoke becomes easily visible at high speeds and under consistent conditions, you're likely doing it right.
TIP: When tuning, attach a heat-resistant hose from the exhaust pipe to the rear window – this way you can see for yourself how much blurring you're willing to tolerate. (I should probably get that patented – but it looks really ridiculous).

Then just add a temperature sensor before the turbine, and you're almost perfectly equipped. (Why doesn't anyone do that where you are?)

"By the way, I saw some Golf IV 1.9l TDI cars on the highway today with a sticker that said something like "Z-Tuning" on the back. When one of them accelerated slightly, it was like the sun went out and there was an eternal night for 30 seconds. Since I was right behind it, I couldn't see anything for a few seconds. That was Lambda 0.5 in diesel!"
Is that one of you?

Okay, going back to the point, if you're not experiencing any issues with lambda/fuel trim, it's always recommended to reduce the boost pressure at high mass airflow rates (high RPMs, especially after reaching Pmax). This lowers the temperature before the compressor, and therefore all intake air temperatures. It also slightly reduces peak pressure, but most importantly, it lowers the turbocharger speed, which is the most critical parameter for your tuning setup.
These are all just my personal opinions and are not universally valid, again .

Julian wrote:

Test benches, with the same environmental variables and proper and correct operation, are definitely much more accurate than your ASZ-DZR device...

Please pay attention to the small details that are highlighted in bold...

Regarding absolute results, the DZR certainly requires more effort in determining the correct inputs (especially wheel circumference and weight).

However, in before-and-after comparative measurements (without claiming to provide accurate absolute performance data), taken within a few minutes on the same track under calm wind conditions and with sufficiently precise timing, the DZR does not need to hide behind test benches – provided that one accepts that it only provides an average overall statement for the range of 2000 - 4000 rpm.

ulf wrote:

http://www.hs-elektronik.com/leistungspruefstand.html

Excellent and informative report; I highly recommend it to anyone who enjoys tuning in their free time.

Thank you for the link.

bafische wrote:

By the way - Have you ever thought about adding an optional water mist cooling system for the engine oil cooler?

It's a maybe... I hope to eventually get an offer for a plug-and-play LED upgrade that can replace the standard LED without requiring major modifications.

Quote:

In principle, you are right; a Lambda of 1.15 at the VL is acceptable considering the turbidity.

As far as I know, Lambda 1.15 would be a mass ratio of 14.5 x 1.15 = 16.7?

Then, the throttling setpoint should probably be set to this minimum value, and then the LD (load demand) should be reduced so that at approximately 25°C, around Pmax, the limiter only shows minimal or no air reserves (above the throttling setpoint). . . however, I am again relying on the LMM .

Quote:

If the smoke becomes easily visible at high speeds and under constant conditions, you are correct.

I'll try that first, but for clarification:
Do you mean becoming visible in broad daylight, or in the darkness within the beam of a following car's headlights, which highlights every speck of soot as if under a magnifying glass?

Quote:

TIP: When tuning, attach a heat-resistant hose from the exhaust pipe to the rear window – then you can see for yourself how much blurring you're willing to tolerate. (I should probably patent this – but it looks really ridiculous).

Good idea... Do you think a piece of flexible aluminum ventilation ducting could work for that?

Quote:

Then quickly add a temperature sensor in front of the turbine, and you're almost perfectly equipped. (Why doesn't anyone do that at your place?)

"Quickly" cut a thread into the exhaust manifold and install a 1000K sensor along with (an expensive?) measuring device?
That definitely exceeds my current capabilities...

Quote:

But let's go back to that point: if you don't have any issues with lambda/fuel trim, it is always recommended to reduce the LD (boost pressure) at high mass flow rates (high RPMs, especially after Pmax). It lowers the temperature before the compressor, and therefore all intercooler temperatures; the peak pressure decreases (slightly), and most importantly, the ATL (air throttle lift) speed, which is the most critical parameter in your tuning.

The temperature before the compressor, for me, is the fresh air drawn in by the charger - how can that be heated up by the subsequent compression process?
Or do you mean the temperature before the turbine?

I consider the ATL boost pressure control to be less critical (in my case) because I replaced my stock intake manifold with the BPX version. This increases the pressure in the air filter box by approximately 40 mbar (if I recall correctly), and should allow for an increase of almost 100 mbar above the standard pressure of 2.3 bar (absolute), without requiring the turbocharger to spin faster than stock.
My turbocharger would only need to be spun up higher to handle the increased mass airflow resulting from the improved intercooling – although a net gain of 50 Kelvin at maximum power is probably just an average value...?

Quote:

These are all just my personal opinions and are not universally valid again

. Nevertheless, I believe that you are not wrong if you trust your own opinion .

Quote:

Lambda 1.15 would be, as far as I know, a mass ratio of 14.5 x 1.15 = 16.7

.
Sure, that's a rough estimate for the PD values.

Quote:

Then, the turbidity compensation factor should probably be set to this minimum value, and then the LD (likely referring to a control parameter) should be reduced so that at approximately 25°C, around the maximum power point (Pmax), the limiter only shows minimal or no air reserve (above the turbidity compensation factor)... However, I am again relying on the mass flow meter (LMM) in this process.

No! Don't do that, otherwise you'll end up within the smoke containment area with every slight downward movement (e.g., during load changes).
Nur so eine Idee: Die Drehmo ist bei dir ja fix, die Leistung steht ja bei dir. Das max. Lambda was du an der VL zulassen willst, kannst du ja mit der Rüsselmethode optisch bestimmen durch ausprobieren oder dich auf das orig. Rauch-KF und HFM verlassen. Du weisst ausserdem schon bestimmt schon wie hoch deine max. Unterschwinder im LD und damit in der Luftmasse beim Einregeln an der VL bei allen möglichen und möglichen VL-Beschleunigungen sind. Jetzt senke den LD solange ab, bis im worst-case also bei max. LD-Unterschwinger das Rauch-KF gerade so berührt wird. DaIt's perfect.

Quote:

I'll try that first, but for understanding: Do you mean "becoming visible" in the bright daylight, or in the darkness within the headlight beam of the pursuer, who presents every speck of soot as if under a magnifying glass?

Under bright sunshine and the highest outdoor temperatures...

Quote:

Good idea... Do you think a piece of flexible aluminum ducting could work for that?

I prefer heat-resistant rubber tubing; it won't scratch anything, and you can easily attach it to the windshield wiper with zip ties – it takes just seconds.

Quote:

The temperature before the compressor, for me, is the fresh air drawn in by the charger - how can that be heated up by the subsequent compression process ? Or do you mean temperature before the turbine?

I misspelt it; I meant "compressor." But thanks for thinking along those lines.

Quote:

I consider the ATL boost pressure less critical (in my case) because I replaced my stock intake manifold with the one from BPX. This increases the pressure in the air filter box within the maximum pressure range by approximately 40 mbar (if I recall correctly), and should therefore allow for an increase of almost 100 mbar above the standard pressure of 2.3 bar (absolute), without the turbocharger having to spin faster than it does in stock configuration. My turbocharger would only need to be spun up higher for the increased mass flow due to the improved intercooling, although 50K net might just be an average value at maximum power...?

Okay, I'd be cautious about that. The key factor for n-ATL (negative air pressure) is the pressure ratio, correct? However, it also means you need to consider the total pressure loss across the entire intake and pressurized section, including the LLK (liquid cooling system) and hoses. And from what I can see, the ASZ LL (low-temperature coolant) line doesn't look as good as the BPX in terms of performance – but I don't have any specific numbers. You're right about the serial LD (low-drag) components for the ASZ, but I thought you were running with the ~2.5 bar from the BPX?

bafische wrote:

Now, lower the LD until, in the worst-case scenario, where the maximum LD underswing occurs, the smoke KF just touches it. Then it will be perfect.

Okay, but I actually don't need any lower-frequency reserves there, because I haven't seen any significant lower-frequency oscillations on the LDA yet.

Quote:

I prefer heat-resistant rubber tubing; it won't scratch anything, and you can easily attach it to the windshield wiper with zip ties; it's done in seconds.

Hmm... Where can I get something like that? What's it called? What temperatures should it be able to withstand approximately? And what caliber are you using? Is it for tightly fitting onto the end pipes (I have the classic TDI dual exhaust, even though it's almost completely hidden by the bottom edge of the bumper), or is it meant to be inserted, or...?
I'm sorry, but I don't really understand it...

Quote:

You are right about the serial LD from ASZ, but I thought you were using the ~2.5 bar pressure from the BPX.

The BPX has 2.55 bar, the ASZ series has 2.35 bar, and I use a maximum of 2.45 bar.
So, the aforementioned increase of 0.1 bar, which within the Pmax range should ideally be "speed-independent" when using the BPX nozzle, based on the pressure ratio, would actually result from a compressor inlet pressure of 1.0 bar (with the BPX nozzle) instead of 0.96 bar (with the standard nozzle).
2.35 x 1.0 / 0.96 = 2.448 bar

Assuming further pressure losses occur up to the compressor inlet, and that the 40 mbar difference caused by the nozzle persists all the way to the turbocharger,
For example, 0.90 and 0.86 bar.
then, with a constant nATL, a significantly stronger pressure increase occurs at the output:
2.35 x 0.9 / 0.86 = 2.459 bar (which is at least 11 mbar more ).

If we add a speculative 0.1 bar to the compressor outlet pressure for the low-pressure cooling system (so, adding to the standard value of 2.45 bar), then the Cupra nozzle provides even more benefit.
2.45 x 0.9 / 0.86 = 2.563 bar before the low-level control valve (LLK) = 113 mbar gain at constant relative to atmospheric pressure (nATL).

... however, always in addition to the required speed for the increased mass flow .

Can you now understand my relatively minor concerns about the charger's speed?

ulf wrote:

Okay, but I actually don't need any underswing reserves there, because I haven't seen any (clear) underswings on the LDA yet.

And if you ever drive into the Bergisches Land region, or if it gets very hot? Add about 3mg of tolerance to the smoking threshold. That usually works well (if the HFM is correct).

Quote:

Hmm... where can one get something like that, under what name or brand? What temperatures should it be able to withstand approximately, and which caliber are you using? Is it for tightly pushing onto the end pipes (I have the classic TDI double pipe, even though it's almost completely hidden by the bottom edge of the bumper), or is it for inserting, or...? Sorry, but I don't really have a good grasp of that.

You can just slide it onto the pipe and then use a hose clamp to secure it. You can estimate the diameter by looking at the exhaust pipe. Try searching online; I found mine by chance. They should be able to withstand temperatures of around 200 degrees Celsius.

Quote:

Assuming further pressure losses up to the compressor inlet and that the 40 mbar difference caused by the nozzle persists until the turbocharger, For example, 0.90 and 0.86 bar. then, with a constant nATL, a significantly stronger pressure increase occurs at the output: 2.35 x 0.9 / 0.86 = 2.459 bar (which is at least 11 mbar more ). If we add a speculative 0.1 bar pressure reduction due to the low-pressure cooling system (meaning the standard pressure is 2.45 bar), then the Cupra nozzle provides even more performance. 2.45 x 0.9 / 0.86 = 2.563 bar before the low-level control valve (LLK) = 113 mbar gain at constant relative to atmospheric pressure (nATL). ... however, always in addition to the required speed for the increased mass flow . Can you now understand my relatively minor concerns regarding the charger's speed?

Partially, the calculation isn't quite correct. Unfortunately, n-ATL (isothermal efficiency), pressure ratio, and volumetric flow rate (efficiency!) are not linearly related. To calculate something accurately, you need the compressor performance curve. But it will probably be in the right direction.
How and where did you measure the 40 mbar lower vacuum pressure?

That's interesting.

Let's say hypothetically:
The vehicle could have larger nozzles with slightly higher injection pressure, resulting in the same atomization, and the injection duration would not deviate from the standard.
The turbocharger delivered linearly more pressure relative to its torque output. This means the combustion would be just as good and efficient as with a stock setup (perhaps even better, because everything happens at a higher molecular density).

Then:
-If only the engine had a higher efficiency! Since it hasn't been physically modified, and assuming internal frictional forces and losses due to bearings and auxiliary components remain constant, this means that proportionally more power is available for use on the road.

To what extent can this be included in the measurement? Or, if at all, were we talking about a value of 0.03% of the overall efficiency?

bafische wrote:

ulf wrote: Okay, but I actually don't need any underswing reserves there, because I haven't seen any (clear) underswings on the LDA yet. And if you ever drive a little bit into the Bergisches Land region or if it gets very hot?

Nothing has come to mind yet... however, my VTG rod is set up a bit longer.
As a result, the overshoots are smaller than usual, and the undershoots that react to them disappear (mostly?).
I'm willing to accept the slightly slower pressure build-up...

Quote:

Allow for a 3mg tolerance relative to the smoking KF value. This always works (if the HFM is correct).

So that at full load, the opacity (with a Lambda of 1.15) is approximately 3mg above the engine speed?
The lambda value is likely correct; it reports a value of approximately 1200 mg at 2500 rpm.

EDIT:
I just reviewed the serial blurring kernel again.
For example, with 750 mg of air, up to 46.2 mg of diesel can be released -> A lambda value of 1.11 is possible from the factory!
(At 925 mg of air: 1.16)
I'm going to experiment with what happens when using the lowest possible tuning lambda of around 1.09, specifically focusing on the point where knocking occurs. This will allow me to maintain a difference of approximately 3 mg between the knock threshold and the torque limit at maximum power (Pmax).
With a ratio of 1.15, I would need 984 mg of air for 59 mg of diesel, and then the engine would likely start misfiring due to excessive heat before reaching its optimal operating speed.

Quote:

Try searching online; I found mine by chance.

They should be able to withstand temperatures around 200 degrees Celsius.

Quote:

Okay, let's see what I can find...

Unfortunately, n-ATL, pressure ratio, and volumetric flow rate (efficiency!) are not linearly related to each other. To calculate something, you need the compressor's characteristic curve.

Quote:

Yes, I would also like to see that (from KP 39).

How and where did you measure the 40 mbar lower vacuum pressure?}
Regarding the external pressure connection for the engine pneumatic system (disconnected) on the air filter housing, using a VDO-LDA with vacuum and a resolution of 0.1 bar.
Series pressure sensor (slightly expanded range): approximately -0.04 bar.
Cupra nozzle: no vacuum detected.

It's just a reading from a simple iron sensor, but it has consistently repeated itself across multiple measurements.

Hi Ulf,

Things are starting to look good. If you've got the LDRs under control, that's even better. Your values seem reasonable.
Hopefully you can find a suitable hose. In about a week, I might be able to tell you more about where to get one (I've heard something about it) if you're not having any luck. Just reach out to me again. You'll probably hear from me soon anyway.

I don't have the specific type of compressor you mentioned from your charger, but I'll see if I can find something similar in the literature. That should give you a starting point.

The 40 mbar reading might be close. It's almost impossible to measure the suction pressure before the air filter accurately enough anyway, even professionals mostly have to estimate it, despite using expensive measuring equipment.