Turbo systems for X1/9's

[Dr.Jeff]

I believe it was somewhere in this thread we discussed "detonation" and ECU tuning? So I'll add this here.

The key to tuning for maximum performance without detonation (pre-ignition, knock, etc) is knowing what's going on inside the combustion chamber as combustion exists. The following presentation should be helpful with that:

How to Improve Engine Performance by Optimizing Cylinder Pressure

R&D engineer Gary Patterson at Roush Yates Engines will cover cylinder pressure measurement tools, identifying data in real-time, and actions to take as a result of the data.

www.sema.org

Honestly I'm not sure if you need to be a member of SEMA to join the presentation.

 

[Jonohhh]

I believe it was somewhere in this thread we discussed "detonation" and ECU tuning? So I'll add this here.

The key to tuning for maximum performance without detonation (pre-ignition, knock, etc) is knowing what's going on inside the combustion chamber as combustion exists. The following presentation should be helpful with that:

How to Improve Engine Performance by Optimizing Cylinder Pressure

R&D engineer Gary Patterson at Roush Yates Engines will cover cylinder pressure measurement tools, identifying data in real-time, and actions to take as a result of the data.

www.sema.org

Honestly I'm not sure if you need to me a member of SEMA to join the presentation.

Good link, I'll be sure to check it out on the 19th!

We're actually working on cylinder pressure instrumentation in FSAE through a grant shared between us and the engines lab! Really neat stuff.

We're using the harder and more intricate route of drilling the cylinder wall itself, through the coolant jacket, to install the sensors instead of the more common spark plug sensor route, so it's been quite a time consuming process as you can imagine.

The small town Alabama machine shop said "you want us to do WHAT?"

 

[Bjorn Nilson]

Good link, I'll be sure to check it out on the 19th!

We're actually working on cylinder pressure instrumentation in FSAE through a Denso grant shared between us and the engines lab! Really neat stuff.

We're using the harder and more intricate route of drilling the cylinder wall itself, through the coolant jacket, to install the sensors instead of the more common spark plug sensor route, so it's been quite a time consuming process as you can imagine.

The small town Alabama machine shop said "you want us to do WHAT?"

The problem is not to install the sensor. The problem is to learn the ECU to detect detonations on your specific engine. My tuner struggled for hours to get it working correctly.
FYI the sensor (Bosch motorsport type) is mounted on the head with a bolt between cyl 2 and 3 on an Uno Turbo so that would most likely be the best location also on a X1/9 engine.

 

[Jonohhh]

The problem is not to install the sensor. The problem is to learn the ECU to detect detonations on your specific engine. My tuner struggled for hours to get it working correctly.
FYI the sensor (Bosch motorsport type) is mounted on the head with a bolt between cyl 2 and 3 on an Uno Turbo so that would most likely be the best location also on a X1/9 engine.
View attachment 60647
View attachment 60648

Hi Björn!

I should have clarified a bit more- in my case with FS I'm talking about an actual cylinder pressure transducer that directly measures pressure inside the combustion chamber. Knock events are easily observable based on the Pressure vs Crank Angle plot this sensor allows you easily to generate.


But regarding knock sensors- I absolutely agree. The calibration of the sensor itself is most definitely the most challenging part. Cylinder pressure can be used to calibrate the sensor but it's unreasonable to expect most people on the street to actually aquire cylinder pressure data. The copper pipe trick works well but is rather unscientific at the end of the day.

 

[Dr.Jeff]

Ya, a little confusing. The traditional knock sensing that we're familiar with is what you're referring to @Bjorn Nilson. There's a newer technology being developed that takes a whole different approach. Rather than "listen" for the rattles after knock has begun (like the current systems do), it reads what is going on inside the combustion chamber real time and can detect the potential for knock before it happens. So it requires a completely different controller to work with it...very high tech stuff. The ones I've heard about are mostly ignition based and not fuel (AFR). There's a couple systems already being tested (maybe even available on the market by now?). But it is very expensive. For the most part the OEMs are behind much of it due to the research required. So it is cool that the students have been given the opportunity to play with it. Please keep us posted @Jonohhh.

 

[Dr.Jeff]

The problem is not to install the sensor. The problem is to learn the ECU to detect detonations on your specific engine. My tuner struggled for hours to get it working correctly.
FYI the sensor (Bosch motorsport type) is mounted on the head with a bolt between cyl 2 and 3 on an Uno Turbo so that would most likely be the best location also on a X1/9 engine.
View attachment 60647
View attachment 60648

Bjorn, I have not gone very far with knock sensing yet*. [And I may not actually; after lengthy discussions with some experts I've come to the conclusion you said - "The problem is to learn the ECU to detect detonations on your specific engine. My tuner struggled for hours to get it working correctly." The best advice I've been given is to not add a knock sensor, and instead tune the management a little more conservatively to help prevent the chance of detonation in the first place. At least for a street application; naturally it would be very different for a all out competition application.]

*However I was surprised to learn the UT has the knock sensor on the head. According to what I've read the head is much too noisy and causes a ton of false readings. That just makes tuning it much more difficult...where do you set the threshold between "noise" and "detonation". The standard location is on the block, up near the head but not too close, away from the valve train and other noisy components, and opposite side of the exhaust. All to reduce interfering noises.

 

[Jonohhh]

Ya, a little confusing. The traditional knock sensing that we're familiar with is what you're referring to @Bjorn Nilson. There's a newer technology being developed that takes a whole different approach. Rather than "listen" for the rattles after knock has begun (like the current systems do), it reads what is going on inside the combustion chamber real time and can detect the potential for knock before it happens. So it requires a completely different controller to work with it...very high tech stuff. The ones I've heard about are mostly ignition based and not fuel (AFR). There's a couple systems already being tested (maybe even available on the market by now?). But it is very expensive. For the most part the OEMs are behind much of it due to the research required. So it is cool that the students have been given the opportunity to play with it. Please keep us posted @Jonohhh.

Yeah- I didn't make it quite clear, but I mentioned that because it seems to be the main focus of the presentation. It's most definitely something very interesting that we have the privilege of having access to- but I will say- even if I had all of the resources to fit this instrumentation to any of my personal vehicles...I would absolutely not. It's simply not worth it unless you're an OEM trying to eek every last Joule out of your combustion processes.

Bjorn, I have not gone very far with knock sensing yet*. [And I may not actually; after lengthy discussions with some experts I've come to the conclusion you said - "The problem is to learn the ECU to detect detonations on your specific engine. My tuner struggled for hours to get it working correctly." The best advice I've been given is to not add a knock sensor, and instead tune the management a little more conservatively to help prevent the chance of detonation in the first place. At least for a street application; naturally it would be very different for a all out competition application.]

*However I was surprised to learn the UT has the knock sensor on the head. According to what I've read the head is much too noisy and causes a ton of false readings. That just makes tuning it much more difficult...where do you set the threshold between "noise" and "detonation". The standard location is on the block, up near the head but not too close, away from the valve train and other noisy components, and opposite side of the exhaust. All to reduce interfering noises.

I'm surprised its on the head to be honest...that's quite a noisy environment, especially on a solid lifter motor. The sensor should not be sensitive to a very wide range of frequencies though- you essentially create a knock event on purpose under a few different conditions to determine your primary frequencies (dictated by the resonance of the actual cylinder, piston, and combustion chamber) and then only setup the sensor to react to frequencies in that specific range. In some engines, regular engine harmonics overlap with knock frequencies and it gets a little challenging. I have no clue whatsoever if this is the case for the SOHC and I likely won't find out until I do it myself in a far away future date.

 

[Dr.Jeff]

Recently I saw a TV program* that ran some tests on a engine dyno to compare different exhaust configurations on a turbo'ed engine. The engine was a American V8 with a single turbo, and boost at about 10 PSI. They measured back pressure, turbo boost, and engine output (torque and HP) with a variety of exhaust configurations from the turbo exit on back (not before the turbo). The tests left out a lot of things I would have liked to see, but it was interesting.

Basically they found any added restrictions in exhaust flow had a significant effect in turbo boost, and therefore engine performance. Not surprising really. However the degree of the effect was greater than I would have thought; a relatively small compromise in exhaust system layout had a rather large negative impact.

The primary restrictions that they induced into the exhaust were a pipe diameter less than the diameter of the turbo's outlet, and an assortment of mufflers. Aside from a large pipe alone, the large diameter pipe with a huge, open straight-through muffler was not too bad compared to the other configurations. However I doubt that muffler offered much in the way of sound level reduction (unfortunately they did not measure sound dB's). The worst arrangement was a pipe diameter smaller than the turbo outlet with a relatively standard "performance" muffler. Basically as back pressure went up, turbo boost went down, as did engine output.

While you can't conclude a ton from their tests, it did reinforce the general concept that a turbo will suffer greatly from a poor exhaust system design.

*Engine Masters, season 6, episode 15, "Exhaust for Turbos - The Dos and Don'ts"

 

[Dr.Jeff]

Regarding my last post about the exhaust system tests on a turbo engine.

I really wish they would have measured the sound levels for the various tests. At the beginning they stated this was intended for a mild street engine, and one goal was to determine a good exhaust design to keep it quiet while not impacting performance too much. But without any sound level testing there is no way to judge the trade off between sound and performance.

Their conclusion was to only run a open pipe with no muffler, and let the turbo be your muffler. I know a small turbo at a reasonable boost level can be effective at lowering noise a lot. But it sure would have been the perfect opportunity to test it.

 

[Jonohhh]

For performance, I would personally stick to purely absorptive muffling (glass packs, essentially) where the exhaust flow path is straight through at the nominal pipe diameter...these are essentially not restrictive at all, as you said. The nominal pipe diameter is important because many traditional glass packs I've seen go down to a smaller ID which is undesirable. To get a lot of noise reduction without the use of baffles, the muffler may end up needing to be quite large. I believe magna flow has a few good options for these- although that knowledge of mine is from over 6 years ago so things may have changed these days. In a well designed one, I think you'd be surprised at how much noise reduction they can provide- surely an ample amount for a turbo 1.5.


Another thing too is that exhaust back pressure can cause a significant loss in power even at the same amount of boost...as more pressure between the turbo turbine and exhaust valve equates to more exhaust left in the cylinder after the exhaust valves close, significantly reducing how much air you can fit in the cylinder on the next intake cycle. This is usually dictated by turbine sizing and AR, but adding a restriction downstream of the turbo just adds to this to some degree so i entirely agree that it should be done as well as it can be!

 

[Dr.Jeff]

Another issue with back pressure in a turbo application is heat. The increased resistance increases charge temps exponentially.

 

[quattrorules]

Another issue with back pressure in a turbo application is heat. The increased resistance increases charge temps exponentially.

With my Mustang SVO, I had a problem with the car running hot, and not making proper boost. The Ford dealer had told me it was a bad turbo, but it turned out to be a bad cat. A long crowbar, and some vigorous shaking, and the system was working fine again. Shows the importance of a good flowing exhaust.

 

[Dr.Jeff]

Previously I've noted a TV program that runs engine dyno tests to investigate various parameters of engine modifications. Recently they have done a couple episodes with turbocharged engines. I just saw another such episode, this time looking at water and/or methanol injection.

I thought their results were a bit odd. Basically they found "spraying" (injecting) water had little to no advantage, and could actually hurt performance if too much was sprayed (increased duty cycle of the PWM that controls the injector solenoid). However methanol had a definite advantage, with "the more the better". In their setup, as more meth was injected, more gasoline was removed (by the ECU). In other words, if they basically made the gasoline into nitro funny car fuel then the engine performed better. To me they were essentially turning the pump gas into E85, a gas/alcohol combination. It has been well established for decades that alcohol fuels make more power than gasoline, so no surprise.

They measured intake temp to see how much it was reduced by either substance. The meth reduced temps, which allowed the ignition timing to be advanced some; but they really did not discuss the temp reduction properties in direct relation to any performance increases - only that the intake temp went down (more on this later).

In my opinion, based on what they reported, injecting water and/or meth does not make any sense. I'd rather tune the fuel system for E85 and get the same results without all the hassles of adding a secondary injection system and continually refilling another tank.

A couple questions came to my mind during the episode. At the beginning the goofy guy said they were going to monitor detonation along with intake temp to see the effects of "spraying". But they never mentioned anything about monitoring or recording detonation, so I'm not sure they did that. Also they really did not investigate the effects of ignition timing in relation to temp reduction very much; only one arbitrary increase in timing early on. And without any record of detonation (before and after the change) the effect of that had no real meaning. Another question was how much the performance increases were actually due to the change in fuels (adding a ton of meth to gasoline) vs the intended purpose of the tests - decreased intake temps. As I said, they speculated a little about it but really did not address it directly. That's why I came away thinking the improvement was due to the nature of the fuel itself and not the intent of "spraying". It would have been interesting to see what happens if the meth injection was removed and the meth was added directly to the gasoline tank. I suspect they'd get the same results. Other comparisons that I would like to see are back to back tests with the same engine setup, but comparing a intercooler or E85 to the "spraying" approach. They've tested E85 on a turbo engine before but that was a different setup so the results cannot be directly compared to this session with water/meth injection.

For me the take away was either E85 or a intercooler makes much more sense that adding water and/or meth injection. Or even better, BOTH a intercooler AND E85.

 

[Jonohhh]

So, this is one of the cases where it's heavily dependent on the setup in question.


I would not ever consider of using WMI (water meth) in place of an intercooler on a street car... from both a reliability and performance perspective.

I'm not sure exactly which video/episode is being referenced, but I will say- often, these tests being performed on videos and such exclude many important factors that essentially render the test useless and the conclusion flat out wrong. There is almost always data to back up what is said- but that's sort of like looking at the back of a car that got in a front end collision and determining the car didn't get damaged.

I remember watching an engine masters video a couple of years back where they tested WMI vs an intercooler. The video was alright, but when it came time for data analysis, they barely mentioned the additional timing (and thus power) that could be made by optimizing the ignition timing for the new meth setup. That's literally where 80% of the power is- and it was mostly ignored.

Half way through the video it was revealed that they were running race gas ... which quite literally defeats the entire purpose of the test. Cooler charge air allows you to run more ignition timing with the same margin of safety from detonation- just like running a high octane fuel would allow, but because they're already running race fuel, seemingly don't have ignition timing anywhere near peak, and didn't bother optimizing it, the results were disappointing as expected.

The thing with engine masters or many things similar, is that if you have an vintage car without a bunch of electronics and tunability, the takeaways are absolutely valid. However, for most any modern and complex system, they don't go deep enough into it before coming to a conclusion to actually make the answer correct or relevant for modern applications.

Second: WMI is huge in high power aftermarket applications (hell, bmw has even used it from the factory). But, in 90% of cases it is used alongside with an appropriately sized intercooler.

The advantage here is that your intercooler can do the heavy lifting of reducing charge air temps closer to ambient, and then the WMI, through evaporative cooling, can achieve a temperature far below ambient when desired. This is why WMI can provide such massive gains in these applications.


Without being paired up with an ECU setup to optimize timing and such, the gains will be less significant. I need to read back and see what ECU setup you're thinking about using because that significantly changes what is and isn't beneficial.

 

[Dr.Jeff]

Funny I was just talking about the Engine Masters TV program and E85 fuel. Last night I saw another episode where they compared dyno results with race fuel, E85, and methanol. The abbreviated version of their findings were: for a naturally aspirated engine the "oxygenated" race fuel was their top choice, but it was mentioned (again) that a boosted application would benefit more from E85 or methanol. They discussed various reasons for this but just looking at the tests and data I'd say for a street turbo engine E85 would be a top choice, and for a race only turbo engine meth might be. I say this in part due to performance reasons but also for practical reasons. I might have to research E85 a bit more and see how my project plans may possibly get modified accordingly.

 

[dllubin]

Funny I was just talking about the Engine Masters TV program and E85 fuel. Last night I saw another episode where they compared dyno results with race fuel, E85, and methanol. The abbreviated version of their findings were: for a naturally aspirated engine the "oxygenated" race fuel was their top choice, but it was mentioned (again) that a boosted application would benefit more from E85 or methanol. They discussed various reasons for this but just looking at the tests and data I'd say for a street turbo engine E85 would be a top choice, and for a race only turbo engine meth might be. I say this in part due to performance reasons but also for practical reasons. I might have to research E85 a bit more and see how my project plans may possibly get modified accordingly.

That was a good episode. They did mention that there was also timing optimization done for each fuel but no time to show it during the episode.

 

[Dr.Jeff]

That was a good episode. They did mention that there was also timing optimization done for each fuel but no time to show it during the episode.

I noticed that also. It made me think of the other thread where ignition timing and fuels are being discussed.

Things like this are what keep it from being a great program. The scientist in me wants to see better scientific experimentation methodologies utilized with more controls employed for variables, more data collected, some statistical analysis, and more complete results reporting. But I get it, it's a TV program and like others have said - at least they are doing things on a dyno. So it could be much worse. None the less I always seem to end up with more questions than answers.

Frankly I've never looked into E85 before so I don't really know a lot about using it. It's really not available much in the southwest and I've never owned a flex-fuel vehicle or done a specific engine build that would benefit from it. But the two or three Engine Masters tests involving E85 (plus other sources) have got me thinking about it for my turbo build. The primary goal for that build is to see how practical it might be to add a small, low boost turbo to a relatively stock X engine utilizing a aftermarket standalone ECU. The biggest potential failure for such a engine is detonation destruction and from what I'm seeing E85 will significantly reduce that possibility.

 

[dllubin]

I noticed that also. It made me think of the other thread where ignition timing and fuels are being discussed.

Things like this are what keep it from being a great program. The scientist in me wants to see better scientific experimentation methodologies utilized with more controls employed for variables, more data collected, some statistical analysis, and more complete results reporting. But I get it, it's a TV program and like others have said - at least they are doing things on a dyno. So it could be much worse. None the less I always seem to end up with more questions than answers.

Frankly I've never looked into E85 before so I don't really know a lot about using it. It's really not available much in the southwest and I've never owned a flex-fuel vehicle or done a specific engine build that would benefit from it. But the two or three Engine Masters tests involving E85 (plus other sources) have got me thinking about it for my turbo build. The primary goal for that build is to see how practical it might be to add a small, low boost turbo to a relatively stock X engine utilizing a aftermarket standalone ECU. The biggest potential failure for such a engine is detonation destruction and from what I'm seeing E85 will significantly reduce that possibility.

Although l can drive less than a mile to get hydrogen, I've got no clue where to find E85. The guys on the show mentioned that issue in the western US.

 

[Dr.Jeff]

Although l can drive less than a mile to get hydrogen, I've got no clue where to find E85. The guys on the show mentioned that issue in the western US.

According to the internet there are a couple stations in Vegas that sell it. However I've never verified that. Unfortunately none are near me but for a hobby car that rarely gets driven I suppose it might be ok to fill up occasionally. Although that might be something look into, how well E85 stores in the tank.

 

[Bjorn Nilson]

It seems like E85 is good for applications with a lot of boost as it helps preventing detonation and improves cooling. E85 is commonly used in Sweden in racing, drifting and drag racing. However the pump E85 is not what it used to be. The quality varies from E72 to E82 so it can not longer be used for serious tuning and racing. I prepared my car for E85 but gave up because of the varying quality.
Btw, I recently met a guy at a car meet who had an old super nice Fiat 500 with an Uno T engine. He has 257whp@2.3 bar boost running Premium 98 octane gasoline (just like I do with 1.6 bar boost) so obviously traditional fuel also works well. Maybe it is time to buy a bigger turbo .