178 replies to this topic

[cheapracer]

Zirconia Deoxide...

Does anyone know of it's use or possible use in racecar application or other vehicles?

From Wiki;

Ceramic and engineering properties
Zirconium dioxide is one of the most studied ceramic materials. Pure ZrO2 has a monoclinic crystal structure at room temperature and transitions to tetragonal and cubic at increasing temperatures. The volume expansion caused by the cubic to tetragonal to monoclinic transformation induces very large stresses, and will cause pure ZrO2 to crack upon cooling from high temperatures. Several different oxides are added to zirconia to stabilize the tetragonal and/or cubic phases: magnesium oxide (MgO), yttrium oxide, (Y2O3), calcium oxide (CaO), and cerium(III) oxide (Ce2O3), amongst others.

Zirconia is very useful in its 'stabilized' state. In some cases, the tetragonal phase can be metastable. If sufficient quantities of the metastable tetragonal phase is present, then an applied stress, magnified by the stress concentration at a crack tip, can cause the tetragonal phase to convert to monoclinic, with the associated volume expansion. This phase transformation can then put the crack into compression, retarding its growth, and enhancing the fracture toughness. This mechanism is known as transformation toughening, and significantly extends the reliability and lifetime of products made with stabilized zirconia. A special case of zirconia is that of tetragonal zirconia polycrystaline or TZP, which is indicative of polycrystalline zirconia composed of only the metastable tetragonal phase.

The cubic phase of zirconia also has a very low thermal conductivity, which has led to its use as a thermal barrier coating or TBC in jet and diesel engines to allow operation at higher temperatures. Thermodynamically the higher the operation temperature of an engine, the greater the possible efficiency (see Carnot heat engine). As of 2004, a great deal of research is ongoing to improve the quality and durability of these coatings. It is used as a refractory material, in insulation, abrasives, enamels and ceramic glazes. Stabilized zirconia is used in oxygen sensors and fuel cell membranes because it has the ability to allow oxygen ions to move freely through the crystal structure at high temperatures. This high ionic conductivity (and a low electronic conductivity) makes it one of the most useful electroceramics.

The ZrO2 band gap is dependent on the phase (cubic, tetragonal, monoclinic, or amorphous) and preparation methods, with typical estimates from 5-7 eV.[1]

This material is also used in the manufacture of subframes for the construction of dental restorations such as crowns and bridges, which are then veneered with a conventional feldspathic porcelain.[2]

Zirconium dioxide can occur as a white powder which possesses both acidic and basic properties. On account of its infusibility and brilliant luminosity when incandescent, it was used as an ingredient of sticks for limelight.

Zirconia is also an important high-k dielectric material that is being investigated for potential applications as an insulator in transistors in future nanoelectronic devices.


[edit] Diamond substitute
Single crystals of the cubic phase of zirconia are commonly used as a substitute for diamond (diamond simulant) in jewelery. Like diamond, cubic zirconia has a cubic crystal structure and a high index of refraction. Discerning a good quality cubic zirconia gem from a diamond is difficult, and most jewellers will have a thermal conductivity tester to identify cubic zircona by its low thermal conductivity (diamond is a very good thermal conductor). This state of zirconia is commonly called "cubic zirconia," "CZ," or "zircon" by jewellers, but the last name is not chemically accurate. Zircon is actually the mineral name for naturally occurring zirconium silicate (ZrSiO4). Its transparent form is also used as a gemstone, and its opaque form as a refractory.


[edit] Patents
On August 7, 2006, Apple Inc. filed a patent for using zirconia oxide ceramics as casing for mobile devices. Some of the current iPods contain radios for built-in Nike+, Bluetooth, or wireless networking. Using zirconia-based ceramics rather than steel or aluminium for its radio transparency characteristics would enable such internal antennae to be within the device, without the need for changes in thickness or a window, like the iPod touch.[3]

[gruntguru]

Originally posted by cheapracer
Zirconia Deoxide...

Does anyone know of it's use or possible use in racecar application or other vehicles?[3]

Zirconium dioxide. No - apart from its current use in probably every F1 car as a lambda sensor providing AFR information to the on-board engine management system and the pit crew. "Lambda sensor" is a more accurate name than "oxygen sensor" as its response is more closely related to AFR than exhaust oxygen content.

[McGuire]

Originally posted by gruntguru

"Lambda sensor" is a more accurate name than "oxygen sensor" as its response is more closely related to AFR than exhaust oxygen content.

No, an O2 sensor works by directly measuring the oxygen content in the exhaust. From this info the AFR can be inferred by the ECU -- with a lean mixture not all the air finds combination, thus a surplus of oxygen, while with a rich mixture an abundance of CO and thus a relative shortfall of oxygen. However, since the sensor does not measure AFR directly, only infers it from O2 content, the system can be fooled. For example, if one spark plug fouls the AFR will be rich but the sensor will read it as lean due to the high O2 content.

How it works: One side of the zirconia element (thimble or plate) is exposed to exhaust gas and the other side to atmospheric air, and on either side is a porous platinum plating to serve as anode and cathode. Zirconium dioxide is an electroceramic with this interesting property: at temperatures >250 C, it conducts oxygen ions through its lattice. It's a kind of fuel cell. So when there is more oxygen on one side than the other, a charge is produced and a voltage potential appears across the electrodes -- which magnitude is log of the ratio of the difference between the two sides. (Nernst effect.) In practice, from ~200 to ~900 millivolts working a 450mv bias voltage on the ECU signal line. Since the O2 content on the reference side is known (21 percent more or less) the voltage produced by the element indicates the O2 content on the sampling (exhaust) side, rich or lean.

Note that the zirconia must be fairly warm in order to work. This is what heated O2 sensors are for -- the engine can go closed loop sooner and stay in under a wider range of conditions. The heating element may be PWMed by the ECU.

Wideband aka UEGO or WEGO sensors work the same way but with this added feature: along with the heated zirconia element there is also a diffuser cell or space with a control loop that monitors the current required to maintain the cell at a level equivalent to stoichiometry. (Analagous to a hot wire/film MAF sensor, sort of.) So instead of a discrete lean/rich, a wideband sensor can detect exhaust oxygen levels equivalent to ~10:1 on and output a signal accordingly. But with either type a zirconia sensor does the same thing: measures the oxygen content of the exhaust gas, allowing the ECU to calculate and adjust the AFR.

You can see how leaded gasoline is death for oxygen sensors, as the element is quickly plated with lead oxides and rendered inoperative. Zirconia is simply the common name for zirconium dioxide (usually with a bit of yttrium to stablize the crystal structure) rather like alumina (spark plug porcelain) is the common name for aluminum oxide. Alumina and zirconia look nearly identical in their raw forms.

[gruntguru]

Originally posted by McGuire
No, an O2 sensor works by directly measuring the oxygen content in the exhaust. From this info the AFR can be inferred by the ECU -- with a lean mixture not all the air finds combination, thus a surplus of oxygen, while with a rich mixture an abundance of CO and thus a relative shortfall of oxygen. However, since the sensor does not measure AFR directly, only infers it from O2 content, the system can be fooled. For example, if one spark plug fouls the AFR will be rich but the sensor will read it as lean due to the high O2 content.

I agree that gross changes in oxygen content (eg misfire) will have the effect described. However for reasonably combusted exhaust gas (no misfire), the lambda sensor is also responding to unburned fuel products in the exhaust (CO, HC). When the AFR is richer than stoichiometric, it is the predominance of these unburned products on the exhaust electrode that depletes the oxygen level (remember this is a platinum electrode so combustion reactions will be encouraged) and continues to deplete any oxygen ions that diffuse through the titania from the atmospheric electrode. This sets up a strong concentration gradient which drives the negatively charged oxygen ions across the element and generates a high voltage output from the sensor. This is actually a type of fuel cell which is generating a voltage by combustion of CO and HC from the exhaust with Oxygen from the atmosphere. The output from this "fuel cell" is understandably like a "light switch" with very little output everywhere on the lean side of stoichiometric and strong output everywhere on the rich side of stoichiometric. If we had zero oxygen in the exhaust and also zero combustibles (HC, CO) the sensor output would only be about 500 mV - a lot less than the 900 mV we get with excess CO and/or HC present.

So - while it is indeed the Nernst effect that drives the lambda sensor, the output of the sensor can look quite different to the log of the ratio of atmospheric oxygen to exhaust oxygen (try plotting it). This is because the local oxygen concentration at the exhaust electrode can be quite different to the exhaust oxygen concentration due to catalysis at the electrode.

Put simply - because of the catalytic effect of the platinum exhaust electrode, the oxygen concentration at this electrode will fall into one of three categories-
Nett oxygen (more oxygen than unburned fuel in the engine exhaust ie leaner than stoich'. Sensor output low.)
Nett fuel (opposite to previous. Sensor output high.)
Stoichiometric. (Unburned oxygen and fuel in exact proportion so neither predominates at the exhaust electrode and the sensor is in its "light switch" or transition state)

For all of the above reasons it is better to think of the lambda sensor as an AFR switch that changes state at stoichiometry rather than an oxygen sensor indicating exhaust oxygen concentration.

[McGuire]

Originally posted by gruntguru

I agree that gross changes in oxygen content (eg misfire) will have the effect described. However for reasonably combusted exhaust gas (no misfire), the lambda sensor is also responding to unburned fuel products in the exhaust (CO, HC).

No, the sensor is only capable of responding to O2. It cannot detect HC or CO at all. We can make inferences about HC or CO from the oxygen content, and combined with some other reasonable presumptions, program the ECU to adjust the air-fuel ratio accordingly, but the O2 sensor can only detect oxygen. The rest of the components in the exhaust could be peanut butter and jelly vapor.

The platinum electrodes in an O2 sensor have no gas-catalytization function by design or in practice. They are electrodes, nothing more.

In practice, it's easy to plot an O2 sensor in operation with an oscilloscope. A good narrow-band sensor should have a range of ~180 mv to ~920 mv or so, and will transit from lean to rich in about 75 to 140 milliseconds and from rich to lean in around 50 to 100 milliseconds, with 100 milliseconds as a reasonable benchmark either way. The loop response is not terribly fast, a few Hz.

"Lambda sensor" is a pretty decent name for a conventional oxygen sensor, as it operates in a rather narrow range on either side of stoichiometry. It's a Bosch trade name thing, very well then, and lambda is a totally valid and useful (if narrow) way to look at AFR. But as I see it, it is very misleading to say that the sensor measures the actual air/fuel ratio in any sense. That is not its physical function. It can only measure O2 content and allow AFR to be inferred from there.

[gruntguru]

Originally posted by McGuire


No, the sensor is only capable of responding to O2. It cannot detect HC or CO at all. We can make inferences about HC or CO from the oxygen content, and combined with some other reasonable presumptions, program the ECU to adjust the air-fuel ratio accordingly, but the O2 sensor can only detect oxygen. The rest of the components in the exhaust could be peanut butter and jelly vapor.

The platinum electrodes in an O2 sensor have no gas-catalytization function by design or in practice. They are electrodes, nothing more.

In practice, it's easy to plot an O2 sensor in operation with an oscilloscope. A good narrow-band sensor should have a range of ~180 mv to ~920 mv or so, and will transit from lean to rich in about 75 to 140 milliseconds and from rich to lean in around 50 to 100 milliseconds, with 100 milliseconds as a reasonable benchmark either way. The loop response is not terribly fast, a few Hz.

"Lambda sensor" is a pretty decent name for a conventional oxygen sensor, as it operates in a rather narrow range on either side of stoichiometry. It's a Bosch trade name thing, very well then, and lambda is a totally valid and useful (if narrow) way to look at AFR. But as I see it, it is very misleading to say that the sensor measures the actual air/fuel ratio in any sense. That is not its physical function. It can only measure O2 content and allow AFR to be inferred from there.

I will restrict my comments to narrow band lambda sensors for now.
1. If you fit 2 lambda sensors - 1 immediately before and 1 immediately after an oxidising catalyst and run the engine at - say stoichiometric AFR, you will find both sensors will have the same reading - about 0.5 volt. If you analyse the gases at the same locations you will find a small oxygen concentration just before the cat and a lower concentration after the cat.

2. If you place a piece of platinum in an exhaust stream and bring it up to the temperatures at which lambda sensors normally operate, you will experience catalytic oxidation reactions on the surface of the platinum - regardles of what it was designed for.

3. The lambda sensor is not very useful for measuring exhaust gas oxygen content and this is why I feel it is less misleading to refer to it as a lambda sensor. Its output level depends on a number of variables including oxgen content, sensor temperature and CO and HC content (if you don't believe the last, try bleeding a little LPG just prior to a tailpipe mounted heated lambda sensor). The one characteristic that IS useful is the step in output that occurs at a lambda value of 1 hence the name "lambda sensor"

[McGuire]

1 and 2 are irrelevant as I see it and as for 3, as I said earlier I have no problem with the term Lambda sensor. However, no matter what you call the thing it does not measure AFR.

[zac510]

Sounds like it would be easier to get this put in your wife's ring then trade the diamond for some wind tunnel time.

[gruntguru]

Originally posted by McGuire
However, no matter what you call the thing it does not measure AFR.

I agree - it doesn't measure AFR or Lambda very well either. What it does extremely well is indicate which side of stoichiometric the AFR is running.

[gruntguru]

Originally posted by McGuire
1 and 2 are irrelevant as I see it

The relevance of 1. ("If you fit 2 lambda sensors - 1 immediately before and 1 immediately after an oxidising catalyst and run the engine at - say stoichiometric AFR, you will find both sensors will have the same reading - about 0.5 volt. If you analyse the gases at the same locations you will find a small oxygen concentration just before the cat and a lower concentration after the cat.") is that we have changed the oxygen content in the exhaust using a cat, but the lambda sensor is still giving the same reading (same AFR - same lambda sensor reading)

[cheapracer]

Ahh ain't love grand. Will you 2 go and have sex with each other in someone elses thread.

For those who are interested I just helped the sale of a CIP to Syria and can get things made at around 1/3 to 1/2 the price at or above German quality.

For those who don't know what a Cold Isostatic Press is ..

Isostatic presses are used for compressing powdered materials into shaped pre-forms or general products. There are two main types of isostatic presses; cold isostatic presses (CIP) that function at room temperature and hot isostatic presses (HIP) that function at elevated temperatures. Applications for these processes include consolidation of powdered metals, ceramics, carbides, composites, pharmaceuticals, carbon / graphite, ferrites, explosives, chemicals, foods, nuclear fuel or other materials into compact shapes. Cold pressed metals or ceramic powder compacts may require additional processing, such as sintering, to provide a finished part. Cold isostatic pressing is a compaction process by which powders are turned into include refractory nozzles, blocks, and crucibles; cemented carbides, isotropic graphite, ceramic insulators, tubes for special chemical applications, ferrites, metal filters, preforms, and plastic tubes and rods. The powder material is placed within a mold, and then placed into the CIP processing chamber. A liquid medium, often an oil-water mixture, is pumped in and pressurized on all sides to create a uniform product. The pressure with a CIP chamber may reach as high as 100,000 psi. CIP applications Advantages of cold isostatic presses include the creation of product with uniform density, which leads to a reduction of internal stresses, eliminating cracks, strains and laminations. They also press products that have a higher “green strength” which allows for close tolerances, complex shapes, and better machinability. In addition, the CIP process is relatively inexpensive. Separate styles of cold isostatic presses are available for both industrial and laboratory applications. Specific CIP processes include wet bag processing (free molding), dry bag processing (fixed molding), and warm isostatic pressing (WIP).

[Andy Donovan]

Originally posted by cheapracer
Ahh ain't love grand. Will you 2 go and have sex with each other in someone elses thread.

For those who are interested I just helped the sale of a CIP to Syria and can get things made at around 1/3 to 1/2 the price at or above German quality.

For those who don't know what a Cold Isostatic Press is ..

Isostatic presses are used for compressing powdered materials into shaped pre-forms or general products. There are two main types of isostatic presses; cold isostatic presses (CIP) that function at room temperature and hot isostatic presses (HIP) that function at elevated temperatures. Applications for these processes include consolidation of powdered metals, ceramics, carbides, composites, pharmaceuticals, carbon / graphite, ferrites, explosives, chemicals, foods, nuclear fuel or other materials into compact shapes. Cold pressed metals or ceramic powder compacts may require additional processing, such as sintering, to provide a finished part. Cold isostatic pressing is a compaction process by which powders are turned into include refractory nozzles, blocks, and crucibles; cemented carbides, isotropic graphite, ceramic insulators, tubes for special chemical applications, ferrites, metal filters, preforms, and plastic tubes and rods. The powder material is placed within a mold, and then placed into the CIP processing chamber. A liquid medium, often an oil-water mixture, is pumped in and pressurized on all sides to create a uniform product. The pressure with a CIP chamber may reach as high as 100,000 psi. CIP applications Advantages of cold isostatic presses include the creation of product with uniform density, which leads to a reduction of internal stresses, eliminating cracks, strains and laminations. They also press products that have a higher “green strength” which allows for close tolerances, complex shapes, and better machinability. In addition, the CIP process is relatively inexpensive. Separate styles of cold isostatic presses are available for both industrial and laboratory applications. Specific CIP processes include wet bag processing (free molding), dry bag processing (fixed molding), and warm isostatic pressing (WIP).

Very useful things, CIPs. I'm doing an electroceramics* PhD and we have a CIP to help consolidate ceramics before firing. After uniaxial pressing in a hydraulic press, the CIP gets rid of all the internal flaws and makes a much nicer final product. The problem is when your bag (in our case a rubber glove) breaks and you end up with ceramic powder in all the valves, pump etc. Most CIPs last about 15 years before they end up clogged up and can't hold pressure anymore.

* including yttria-stabalised ZrO2, although I'm just making it for someone else so I couldn't comment on the discussion above.

[cheapracer]

Hey Andy,

Yup, uniaxial pressure is what it's all about, very useful for making missile parts by the way so it's a machine thats well documented when each one is made and for some reason the US are a bit unhappy if you try and sell any to Iran or Nth Korea, funny about that ;)

One of my good Mates here just around the corner has been making CIP's for the last 7 years and I help him on occasion with visiting foreigners.

He makes them from 200mm through to 1.5 meters diameter and has a patent in that he uses wound wire strap for the main casing and the bracing tower instead of solid steel. Just as strong but lighter and much cheaper to produce. I think he is willing to make up to 300mpa depending on size etc. and your wallet.

His customer's moulds are usually better quality than a rubber glove :-)

If you need info email me anytime :-)

[Andy Donovan]

Originally posted by cheapracer
He makes them from 200mm through to 1.5 meters diameter and has a patent in that he uses wound wire strap for the main casing and the bracing tower instead of solid steel. Just as strong but lighter and much cheaper to produce. I think he is willing to make up to 300mpa depending on size etc. and your wallet.

Jeez, that 1.5m model must be a beast. I dread to think how much that would weigh if it was done in solid steel, ours is about 200mm and even that's a serious lump of steel.

His customer's moulds are usually better quality than a rubber glove :-)

Ah yes, I guess it's not exactly what the manufacturer intended!

[penske1414]

Originally posted by McGuire



No, an O2 sensor works by directly measuring the oxygen content in the exhaust. From this info the AFR can be inferred by the ECU -- with a lean mixture not all the air finds combination, thus a surplus of oxygen, while with a rich mixture an abundance of CO and thus a relative shortfall of oxygen. However, since the sensor does not measure AFR directly, only infers it from O2 content, the system can be fooled. For example, if one spark plug fouls the AFR will be rich but the sensor will read it as lean due to the high O2 content.

How it works: One side of the zirconia element (thimble or plate) is exposed to exhaust gas and the other side to atmospheric air, and on either side is a porous platinum plating to serve as anode and cathode. Zirconium dioxide is an electroceramic with this interesting property: at temperatures >250 C, it conducts oxygen ions through its lattice. It's a kind of fuel cell. So when there is more oxygen on one side than the other, a charge is produced and a voltage potential appears across the electrodes -- which magnitude is log of the ratio of the difference between the two sides. (Nernst effect.) In practice, from ~200 to ~900 millivolts working a 450mv bias voltage on the ECU signal line. Since the O2 content on the reference side is known (21 percent more or less) the voltage produced by the element indicates the O2 content on the sampling (exhaust) side, rich or lean.

Note that the zirconia must be fairly warm in order to work. This is what heated O2 sensors are for -- the engine can go closed loop sooner and stay in under a wider range of conditions. The heating element may be PWMed by the ECU.

Wideband aka UEGO or WEGO sensors work the same way but with this added feature: along with the heated zirconia element there is also a diffuser cell or space with a control loop that monitors the current required to maintain the cell at a level equivalent to stoichiometry. (Analagous to a hot wire/film MAF sensor, sort of.) So instead of a discrete lean/rich, a wideband sensor can detect exhaust oxygen levels equivalent to ~10:1 on and output a signal accordingly. But with either type a zirconia sensor does the same thing: measures the oxygen content of the exhaust gas, allowing the ECU to calculate and adjust the AFR.

You can see how leaded gasoline is death for oxygen sensors, as the element is quickly plated with lead oxides and rendered inoperative. Zirconia is simply the common name for zirconium dioxide (usually with a bit of yttrium to stablize the crystal structure) rather like alumina (spark plug porcelain) is the common name for aluminum oxide. Alumina and zirconia look nearly identical in their raw forms.

i got two words for you.....galvanic battery.

[cheapracer]

Wow the news for me just got better today.

As mention my Mate builds and sells CIP's but last week he started on a new CIP for his own testing lab and I will have access to it. For those who don't know what CIP does, in simple terms it forges at extreme pressures but unlike what you know as a forge which presses from a single direction, it doesn't leave a grain because it forges uniaxial - it sits in a mould in fluid and that fluid is extremely pressurised (up to 600mpa for this coming lab rat).

And then at dinner tonight he told me he ordered one of these - - a 5 axis high speed mill. By the way, Lola have one according to Loxin (but they don't have a CIP).

So in summary I will be able to extreme pressure forge alloy magnesium compounds to be stronger or lighter than a billet example and then finish mill them to any complex shape. I will take orders :-)

Andy, his is the only CIP system in the world to use seperate tanks and seperate primary pressure staging pumps to keep the contamination out of the valves etc. besides that he runs the oil on the outside of a sheath that holds the water/oil which bears the mould. It's that oil on the outside of the sheath thats pressurised and since its of low volume, he also has one of if not the fastest pressure rising time in the world too.

[J. Edlund]

Zirconia can be used in thermal barrier coatings on exhaust systems for instance. Don't think that there are that many applications for zirconia in a race car otherwise.

As for the lambda sensor thing, it's true that lambda sensor is affected by other gases than just oxygen, but this is more of an error in the reading rather than a function of the sensor. If we place one lambda sensor in front of a catalyst and a second after, the catalyst also has an oxygen reserve that will affect sensor reading. Modern engines also adjust the air fuel ratio with time, between say lambda 0,98 and 1,02 with an average value slightly below lambda 1. These changes causes the reading of the first sensor to go up and down (from rich to lean and back again), with say, one cycle per second or so. The output from the second lambda sensor is on the other hand more or less constant. If the second sensors show an output similar to that of the first, the catalyst isn't functioning properly.

[gruntguru]

Originally posted by J. Edlund
If we place one lambda sensor in front of a catalyst and a second after, the catalyst also has an oxygen reserve that will affect sensor reading. Modern engines also adjust the air fuel ratio with time, between say lambda 0,98 and 1,02 with an average value slightly below lambda 1. These changes causes the reading of the first sensor to go up and down (from rich to lean and back again), with say, one cycle per second or so. The output from the second lambda sensor is on the other hand more or less constant. If the second sensors show an output similar to that of the first, the catalyst isn't functioning properly.

Have you tried this?

I'm not sure about the oxygen reserve you mention. Whatever it is - the effect will be transient and I still contend that "steady state, the sensors pre and post cat will read the same, except in abnormal conditions eg misfire)"

[J. Edlund]

Originally posted by gruntguru
Have you tried this?

I'm not sure about the oxygen reserve you mention. Whatever it is - the effect will be transient and I still contend that "steady state, the sensors pre and post cat will read the same, except in abnormal conditions eg misfire)"

Yes, I have tried it.

Blue is pre cat sensor, red is post cat sensor, green is lambda integrator value and purple is engine power output. X axle shows time in seconds, sensor output is in mV, engine power in kW and lambdra integrator is dimensionless.

[cheapracer]

Originally posted by J. Edlund
Zirconia can be used in thermal barrier coatings on exhaust systems for instance. Don't think that there are that many applications for zirconia in a race car otherwise.

.

I believe the balls in roller bearings?? I stand to be corrected.

It's hardness is near to diamond after sintering.

[McGuire]

Originally posted by gruntguru
Have you tried this?

I'm not sure about the oxygen reserve you mention. Whatever it is - the effect will be transient and I still contend that "steady state, the sensors pre and post cat will read the same, except in abnormal conditions eg misfire)"

In terms of voltage, depending on sampling rate/fenceposting the two sensors will appear to read the same. See Edlund's plot. That is normal operation. Just as Dr. J says, the converter employs significant O2 storage and that is in effect what the downstream O2 sensor and catalyst monitor function in the ECU are measuring. (The downstream sensor has no metering function, it's only watching the cat.)

When some strict entry conditions are met (P+VT, >70C, closed loop etc etc) once or twice per drive cycle the cat monitor starts a running tabulation of the ratio of the signal strengths between the pre and post cat O2 sensors (it's comparing cross counts, to grossly oversimplify). A high ratio indicates high cat efficiency; low ratio approaching 1:1 indicates failing cat. Meanwhile there are also minimum threshholds for post-cat sensor activity to be met or the ECU will trip the MIL and store a DTC event. Typically the ECU will want to see say, <250mv and >650mv per monitor cycle and/or >2 cross counts or so per 30 seconds out of the downstream O2S in CL, just to make sure it isn't dead or poisoned. Of course the window and specific methodology varies among manufacturers, etc.

So if anything, I might be a little concerned that the post-cat O2S response is a bit too flat in the plot above. However, just from the plot I can't say there is something wrong. Could well be an artifact of the test regime, shop's exhaust system or dyno stack perhaps. It is not at all unusual for the monitor cycle to fail to complete in one driving cycle. It will keep filling on an exponentially weighted moving average but the ratio numerator will only increment once so it can take at least two and perhaps three to six drive cycles to trip the MIL. LEV applications may employ three HO2S, one in the front of the cat where degradation tends to occur first.

[McGuire]

Originally posted by penske1414


i got two words for you.....galvanic battery.

Two words for you: frog legs.

[gruntguru]

Originally posted by J. Edlund


Yes, I have tried it.

Blue is pre cat sensor, red is post cat sensor, green is lambda integrator value and purple is engine power output. X axle shows time in seconds, sensor output is in mV, engine power in kW and lambdra integrator is dimensionless.

Unfortunately your graph is not for steady-state engine operation. I omitted one detail from my original quote (sort of took it as understood) - the system needs to be in open loop. The dynamics when in closed loop operation are obviously very different pre and post cat (Although I agree with McGuire - the post cat sensor looks too flat.)

My original quote
"1. If you fit 2 lambda sensors - 1 immediately before and 1 immediately after an oxidising catalyst and run the engine at - say stoichiometric AFR, you will find both sensors will have the same reading - about 0.5 volt. If you analyse the gases at the same locations you will find a small oxygen concentration just before the cat and a lower concentration after the cat."

[McGuire]

I dunno, from what we can see in the graph it looks pretty steady-state. The engine is producing a fairly constant 7 kW (9 hp) so it's idling as far as we know. I would defer to Dr. J, it's his chart.

[gruntguru]

Originally posted by McGuire
I dunno, from what we can see in the graph it looks pretty steady-state. The engine is producing a fairly constant 7 kW (9 hp) so it's idling as far as we know. I would defer to Dr. J, it's his chart.

My problem is that its in closed-loop, so the AFR is cycling up and down - hardly steady state if you are an oxygen sensor.

[McGuire]

Originally posted by gruntguru
My problem is that its in closed-loop, so the AFR is cycling up and down - hardly steady state if you are an oxygen sensor.

What you are looking at is normal O2 sensor operation. That's how it works.

[gruntguru]

Originally posted by McGuire


What you are looking at is normal O2 sensor operation. That's how it works.

I know. This is closed loop operation. For the pre-post comparison I was talking about you would need to be in open loop - ie unplug the sensor from the ECU.

[cheapracer]

.

So when's the wedding Guys - are we invited?

.

[Tony Matthews]

Originally posted by cheapracer
.

So when's the wedding Guys - are we invited?

.

[phantom II]

You introduced them. What I would like to know is if they learned anything from one another, either on this thread or the other. It would help me if they had the balls to answer this question. I have enjoyed everyone's input as I did 3 years ago on the same subject and I'm sure that this is why we are all here.

Originally posted by cheapracer
.

So when's the wedding Guys - are we invited?

.

[McGuire]

Originally posted by phantom II
You introduced them. What I would like to know is if they learned anything from one another, either on this thread or the other. It would help me if they had the balls to answer this question. I have enjoyed everyone's input as I did 3 years ago on the same subject and I'm sure that this is why we are all here.

I should be charging him $800/hr for all I have taught him so far.

[McGuire]

Originally posted by gruntguru
I know. This is closed loop operation. For the pre-post comparison I was talking about you would need to be in open loop - ie unplug the sensor from the ECU.

It really and truly doesn't matter one iota. As long as the AFR is stoichiometric as you stipulated, the interval average of the O2 sensor's voltage output will be Lambda. That's a given for a properly working O2S.

Meanwhile, the integral of the Lambda integrator will also be equivalent to Lambda, he said redundantly. Look at the chart. Note that each time the O2S value rises (rich signal) the Lambda integrator value falls (lean command) and vice versa.

If you uncouple the O2S from the ECU and operate the sytem in open loop, as long as the AFR remains stoichiometric, as you stipulated, the O2S will report stoichiometic value as well. Unless the sensor is broken or something, of course. All that changes is the correction rate. Obviously some means of active fuel trim will have to be employed to maintain an exact 14.7:1 AFR, and to whatever extent it adjusts AFR on the fly the O2S is going to fall below and rise above exact Lambda at some rate and within a predetermined range, just as in the chart above.

[phantom II]

I don't know enough to know if that is true or not. They replaced all the O2 sensors on my ZR1 last week. I was getting about 10mpg. Jees, the thing uses so much gas compared to my Z06. One thing though that has improved drastically, is the suspension. The softer springs are what probably did it and also the fancy shocks. This has caused different shift points because it is easier to maintain traction I think. I wonder how different the shift points would be on a stock Z06 with this suspension?
Have you thought of writing an engine book from A to B? And how come you haven't presented your own dyno videos to prove your points?

Originally posted by McGuire


I should be charging him $800/hr for all I have taught him so far.

[J. Edlund]

Originally posted by cheapracer


I believe the balls in roller bearings?? I stand to be corrected.

It's hardness is near to diamond after sintering.

Rollers in bearings are usually made from silicon nitride.

Originally posted by McGuire
I dunno, from what we can see in the graph it looks pretty steady-state. The engine is producing a fairly constant 7 kW (9 hp) so it's idling as far as we know. I would defer to Dr. J, it's his chart.

Engine (GM L850, 210 hp) is run with constant speed and constant load, load/speed is selected to match the power output at low speed driving. Data is logged through the use of a development version OEM ECU (Trionic 8).

Note that the time period is rather short, only 10 seconds, using a longer time frame the output from the post cat sensor is not quite as constant as seen here. Time is in seconds since engine start up, so the engine has only been running a little over a minute before the data in the graph was logged.

[McGuire]

Originally posted by J. Edlund


Engine (GM L850, 210 hp) is run with constant speed and constant load, load/speed is selected to match the power output at low speed driving. Data is logged through the use of a development version OEM ECU (Trionic 8).

Note that the time period is rather short, only 10 seconds, using a longer time frame the output from the post cat sensor is not quite as constant as seen here. Time is in seconds since engine start up, so the engine has only been running a little over a minute before the data in the graph was logged.

Cool, thanks for the info. Is Trionic 8 the same PCM as the E37, an enhancement of the E37, or completely different?

[gruntguru]

Originally posted by McGuire


It really and truly doesn't matter one iota. As long as the AFR is stoichiometric as you stipulated, the interval average of the O2 sensor's voltage output will be Lambda. That's a given for a properly working O2S.

So have you integrated it yet? As it stands the graph is not suitable to prove or disprove my point concerning lambda sensors responding to AFR.

[gruntguru]

Originally posted by phantom II
You introduced them. What I would like to know is if they learned anything from one another, either on this thread or the other. It would help me if they had the balls to answer this question. I have enjoyed everyone's input as I did 3 years ago on the same subject and I'm sure that this is why we are all here.

I've learned heaps from McGuire - Wouldn't pay $800/hr though.

I doubt McGuire has learned anything from me - he doesn't strike me as the receptive type.

[gruntguru]

Originally posted by McGuire
If you uncouple the O2S from the ECU and operate the sytem in open loop, as long as the AFR remains stoichiometric, as you stipulated, the O2S will report stoichiometic value as well.

So if both sensors report stoichiometric in spite of the pre-cat sensor seeing a higher oxygen concentration than the post-cat - they must be responding to AFR.

[gruntguru]

Originally posted by McGuire
Obviously some means of active fuel trim will have to be employed to maintain an exact 14.7:1 AFR, and to whatever extent it adjusts AFR on the fly the O2S is going to fall below and rise above exact Lambda at some rate and within a predetermined range, just as in the chart above.

You obviously haven't actually done anything like this.

If the engine is at steady state, manual adjustment of the mixture will produce a steady lambda sensor reading for long enough for purpose of this exercise.

[gruntguru]

Originally posted by phantom II
It would help me if they had the balls to answer this question. I have enjoyed everyone's input as I did 3 years ago on the same subject and I'm sure that this is why we are all here.

Careful - one of us might be a lady.

And yes - this is more fun than logic. Right up there with sex in fact.

[McGuire]

Originally posted by gruntguru
You obviously haven't actually done anything like this.

If the engine is at steady state, manual adjustment of the mixture will produce a steady lambda sensor reading for long enough for purpose of this exercise.

Only a few thousand times.

Sure, if the AFR is Lambda the sensor will read Lambda too. I hope you finally have that nailed down.

Meanwhile you are way missing how the system works. The pre-cat sensor is SUPPOSED to oscillate. This is called FAR perturbation. At the one end the three-way cat is better at NOx reduction, at the other end HC/CO. But it doesn't hurt AFR management due to the high response rate. Interval average remains at Lambda. Actually, this is a first principle of any feedback loop -- its authority is determined by the transit rate; stability is overrated.

I think you have been tripped up by the fact that the pre and post cat O2 sensor values are interpreted two different ways. However, in cat monitoring the ECU is not really comparing the two signal voltages but the difference in signal activity between them.

So sure, depending on its sampling and averaging rates, if you put a DC voltmeter on the two O2S outputs (don't actually, it can damage the sensor) you may well see around the "same" .5 volts... and be totally oblivious as to how this whole deal operates. These values are trivial.

[gruntguru]

Originally posted by McGuire
Meanwhile you are way missing how the system works.

I know exactly how it works. My comments and suggested experiment were about the lambda sensor - not a lambda feedback circuit in operation.

If you open the loop, lock the throttle at say 2000 rpm the sensor will read a steady value rather than oscillate. It will hold a steady reading without needing mixture adjustment for long enough to compare pre and post-cat sensor readings.

[gruntguru]

Originally posted by McGuire
Sure, if the AFR is Lambda the sensor will read Lambda too. I hope you finally have that nailed down.

You are the one who insisted the sensor responded only to oxygen when I said it responded to AFR! Check Post #3.

quote:Originally posted by gruntguru

"Lambda sensor" is a more accurate name than "oxygen sensor" as its response is more closely related to AFR than exhaust oxygen content.


McGuire:
No, an O2 sensor works by directly measuring the oxygen content in the exhaust.

[phantom II]

Is this what is referred to as circular dialogue? I'm getting dizzy.

Originally posted by gruntguru
You are the one who insisted the sensor responded only to oxygen when I said it responded to AFR! Check Post #3.

McGuire:
No, an O2 sensor works by directly measuring the oxygen content in the exhaust.

[cheapracer]

Originally posted by gruntguru


I doubt McGuire has learned anything from me - he doesn't strike me as the receptive type.

Well I guess that clears up which one of you will be biting the pillow on wedding night then.

[Tony Matthews]

Originally posted by cheapracer


Well I guess that clears up which one of you will be biting the pillow on wedding night then.

[gruntguru]

Originally posted by cheapracer
Well I guess that clears up which one of you will be biting the pillow on wedding night then.

This wedding thing is not in the bag yet - I need more information.

[gruntguru]

Originally posted by phantom II
Is this what is referred to as circular dialogue? I'm getting dizzy.

Oh you've noticed that too! I'm starting to realise why McGuire's favorite word is "circular"

[McGuire]

Originally posted by gruntguru
I know exactly how it works. My comments and suggested experiment were about the lambda sensor - not a lambda feedback circuit in operation.

If you open the loop, lock the throttle at say 2000 rpm the sensor will read a steady value rather than oscillate. It will hold a steady reading without needing mixture adjustment for long enough to compare pre and post-cat sensor readings.

No, you have only managed to invent an experiment with a pointless methodology producing a meaningless result. Congratulations, you scored a twofer.

For the purposes of your test there is no reason to decouple the O2S from the ECU to drive the system open loop. If AFR remains at stoich as you stipulated, you have not changed the effective output of the sensor whatsoever. If the mixture is stoich a properly operating O2 sensor will indicate stoich as well. You only want to run open loop because due to your ignorance of the system's operation, you were incapable of recognizing that in the engine depicted, the sensor is already outputting stoich value, simply varying the signal at its normal rate.

The only thing decoupling the O2S will accomplish is to eliminate the feedback circuit's normal cycling function. Now the three-way catalyst can no longer oxidate properly, thus skewing the post-cat O2S output out of spec and rendering its measurement false/meaningless.

The fact is that the upstream and downstream O2 sensors produce output signals of two markedly different amplitudes and frequencies, just as J. Edlund's graphic illustrates. That's how the system works. If you can devise a test that shows it working differently, your test is wrong.

[cheapracer]

Originally posted by McGuire


No, you have only managed to invent an experiment with a pointless methodology producing a meaningless result. .

I guess then that puts him in contention to get Goverment grants.