Exhaust back pressure anomaly

[kovir]

Hello
Toyota Land Cruiser 3.0 d4d Diesel engine
Problem - engine vibrations
What could be the cause of backpressure spike?
Firing order 1-3-4-2

[Steve Smith]

Hello and sorry for the late reply:

Exhaust gas pulsation diagnosis is challenging for sure given the number of variables that can affect the shape and timing of pulses captured at the tail pipe

These variables include:
• Exhaust design and routing
• Exhaust form (i.e., V-Engine with 4 into 2 into 1 then perhaps back to 2 exits)
• Active exhaust systems
• EGR activation upstream and downstream
• After treatment systems
• VVT-I and VVT-iL
• Choice of tail pipe to measure (when multiple tailpipes are utiilsed; one maybe for cosmetic purposes only)
• Measurement location (At the rear of the vehicle, i.e., tailpipe)
• Active intake systems
• Turbocharger influence on gas flow
• Mechanical integrity of the engine
• Measurement variables (disturbance to sensor from tailpipe movement)

With all that said, exhaust gas pulsation diagnosis remains a valid technique assuming you know the characteristics of the engine/exhaust system and use the acquired gas pulse combined with other channels to base your conclusion.

Ideally a known good [signature] capture with information such as temperature would be awesome
Typical non-intrusive signals that will assist with pulse identification include ignition/injection events (for “Sync” and TDC estimation) crankshaft signals for speed and angle of rotation and perhaps MAF/MAP for intake pulsations. Remember pulses found within the exhaust will also have an impact on intake pulses and vice-versa during overlap. Now add into the mix a mechanical/combustion problem and pulsation identification becomes ever more challenging!

Note: Exhaust gas pulsation evaluation is most valuable during crank, non-start scenarios as we remove the complexities of combustion, EGR and VVT from the equation. (Our engine is simply an air pump in this condition)

Below I have opted to capture the exhaust gas pulsations from a BMW 2.0 Litre 4-Cylinder N47 Diesel engine that is warm (not hot) during cranking (non-start) with a WPS500x inserted into Cylinder 1 and a First Look sensor installed into one of the tailpipes

Image 1

Before we move on with the analysis above, there is one key fundamental to remember about exhaust gas pulsations! Prior to the exhaust valve open event (EVO) towards the end of the expansion stroke, there will be a negative pressure above the piston that is introduced into the exhaust. This will result in a momentary “suck back” of exhaust gasses within the exhaust system

During analysis of the majority of signals, repetition and uniformity is the name of the game and as we can see above, the “suck back” is not uniform!

We can detect EVO from each cylinder with reasonable accuracy and the time rulers confirm the correlation of the EVO event captured by the WPS (channel C) and the actual detection of the “suck back” at the tailpipe (approx. 11.5 ms later) Be aware of such delays which vary given the infinitely variable dynamics of exhaust gas flow between vehicles

So why the lack of uniformity? The repetition suggests we have “relatively” good timing between events (suggesting our engine is either “in time” or all cylinders are equally “out of time”) however, the uniformity is lacking. (See “suck back” for cylinder 1)

The fault that has been deliberately placed on this engine is an increase in cylinder volume thanks to the addition of our WPS500 compression hose and dummy glow plug (which could add approx. 10 ml) See here viewtopic.php?p=97815#p97815 for more information about compensating for additional cylinder volume when measuring peak compression with a WPS500

The effect of increased cylinder volume is lower peak compression (usually caused by a cylinder leak) and a reduced expansion pocket which therefore results in a negative pressure that is closer to atmospheric, hence the reduced “suck back” in the tailpipe during the cylinder 1 EVO event

Outside of exhaust pulse analysis for one minute, look at the crank math channel and note the increased crankshaft speed during compression and decrease during expansion for cylinder 1 (due to the increased cylinder volume). This non-uniformity in speed has an effect on the following compression and expansion strokes of cylinder 3 which has the job of “building” crankshaft speed back up to the uniform speed we see for cylinder 4 and 2.

This momentary “catching up” of crankshaft speed has a direct effect on the “suck back” for cylinder 3 as we can clearly see it is better that cylinder 1 but not as good as cylinder 4 and 2

The message here is be aware that an under-performing cylinder will have an effect on the neighboring cylinder (in the firing order) with respect to intake pull (inside the intake manifold) and exhaust “suck back”

Something else you may have spotted in the crank math channel is the area I have highlighted with red circles that appear to coincide with the EVO events of each cylinder and suggest a momentary increase in crankshaft speed! Note how cylinders 3, 4 & 2 look well defined yet cylinder 1 appears to lack uniformity by comparison?

My initial thought here is the EVO event has the effect of momentarily unloading the crankshaft due to the fact we have released the negative pressure above the piston into the exhaust system, resulting in an instantaneous increase in crankshaft speed.

To quote Brandon Steckler who uses the syringe analogy; imagine pulling back on the plunger of a syringe whilst holding your finger over the exit. As the plunger approaches the end of its travel it will become very difficult to pull due to the negative pressure within the body of the syringe.

Now imagine removing your finger from the exit of the syringe whilst still pulling back on the plunger; the effort required will decrease and so plunger speed will increase (assuming you are applying the same effort) The plunger in our scenario is the crankshaft and the effort is applied by the starter motor during cranking

So, moving on, lets take a look at exhaust gas pulsations on this engine during cranking with all compressions equal (Cylinder 1 volume has now been restored to normal)

Image 2

Note the uniformity and repetition of both the crankshaft speed and exhaust gas pulsations. Here we don’t have to worry about the influence of combustion/EGR etc. confirming we have even contribution to exhaust gas pulsations with “typical” EVO timing events

Below we now have the engine with compressions equal and all cylinders firing as normal at idle speed (No faults)

Image 3

Note above how we have repetition but the exhaust pulses that were present during cranking have become less defined and complex in nature given the introduction of combustion dynamics. The capture therefore serves as a signature (known-good) for this vehicle idling at approx. 23°C

In addition to the repetition above, I have highlighted what appears to be repeat disturbances (multiple oscillations) during the exhaust stroke of cylinder 4. At this stage I don’t have an explanation other than the arrangement of the exhaust manifold design which incorporates the EGR port and turbocharger flange between cylinders 1 and 2 (see below); could this be buffeting inside the exhaust manifold? Once again, this is something we could discuss for weeks but one thing is for sure, the vehicle runs fine and the above capture forms our “signature”

Image 4

Below we introduce a misfire, by disconnecting injector number 2 (Compressions are normal)

Image 5

The left-hand image has our “signature” waveform, the right hand captures the disruption to the exhaust pulsations due to cylinder 2 misfire. I have highlighted the areas of interest which are now clear because of our signature comparison. Note how a misfire on a single cylinder disrupts the entire gas flow resulting in loss of repetition and uniformity.

Interestingly below, the EML illuminated and no doubt set a fault code which has the beneficial side effect of switching off EGR.

Image 6

I think we can agree above, with EGR inactive this is the best repetition and uniformity we have seen via the tailpipe with the engine running (good compression and no misfire)

Moving onto your capture below, we have a reduced “suck back” for the EVO event of cylinder 3 combined with excessive pressure during this exhaust stroke. (I have used the “Pressure Waveform Overlays” chart from Microsoft to assist with the identification of all cylinder strokes.)

Image 7

We certainly have repetition to our capture but I would expect improved uniformity rather than the peak highlighted above. (This is where your signature capture would be worth its weight in gold)

Could I ask, what is the customer complaint and what is the symptom exhibited by the engine?

Do you have fault codes, misfire data or excessive smoke?

The exhaust gas pulsations alone are one piece of the puzzle not the answer (to quote Brandon again) Could you repeat this test when cranking only?

I will make an assumption the vibration is due to a misfire, the following tests in addition to exhaust gas pulsations will help:

• Relative compression
• Intake manifold pulsations
• Crankcase pulsations
• Cam and crank correlation
• Graphing of the crankshaft sensor signal (Crank math’s)

I hope the above information is of some help and has not deterred you from using exhaust gas pulsation monitoring during diagnosis. If anything, I hope it highlights the challenges involved and that measuring such pulses during cranking yields a better return for the time spent.

If we could measure exhaust gas pulses as a matter of routine maintenance and add them to the Reference Waveform Library that would be most appreciated. (These captures are as valuable as Cam and Crank correlation waveforms)

Take care…….Steve