Engine technology thread.

[Allch Chcar]

Wow, I missed some good stuff in here. So long as I don't go try and explain something again, this should be fun!

[serialk11r]

Hey arghx7 if you see this, I was wondering about the thing you said about direct injection being very reliant on charge motion.

The 2GR-FSE paper has a chart that went something like "tumble ratio" vs. "flow coefficient", and they explained that they used a high flow port on the 2GR. From context, it seems like they're telling us the 2GR has high flow, low tumble ports. It has a little blurb about "consequently the tumble ratio of the high flow test engine was reduced by one third at the valve-lift divided by valve-diameter of 0.3".

Is this to say that the ports are still a compromise between tumble and flow? Is this why you were questioning the viability of increasing revs (intake port flow)?

[Dimman]

Originally Posted by serialk11r

Hey arghx7 if you see this, I was wondering about the thing you said about direct injection being very reliant on charge motion.

The 2GR-FSE paper has a chart that went something like "tumble ratio" vs. "flow coefficient", and they explained that they used a high flow port on the 2GR. From context, it seems like they're telling us the 2GR has high flow, low tumble ports. It has a little blurb about "consequently the tumble ratio of the high flow test engine was reduced by one third at the valve-lift divided by valve-diameter of 0.3".

Is this to say that the ports are still a compromise between tumble and flow? Is this why you were questioning the viability of increasing revs (intake port flow)?

Couple points to consider.

When they say 'tumble' it is probably in the context of low-rpm/intake velocity.

Also it's not tumble specifically that's important, but maintaining charge motion as the combustion chamber squishes approaching ignition.

One reason F1 cars were 'only' 12:1 CR was for room for this motion.

[arghx7]

Originally Posted by serialk11r

Hey arghx7 if you see this, I was wondering about the thing you said about direct injection being very reliant on charge motion.

I work in the auto industry and part of the work I do relates to studying industry engineering trends that relate to engine efficiency. Most direct injection engines currently in production for the North American market rely on charge motion to some degree. The three biggest things are

1) charge motion: tumble or swirl flow, which must be considered 3 dimensionally. Reliance on tumble flow is more common on modern production GDI engines.

2) piston crown design: fuel can spray into a piston bowl to help mixture formation, but there are a lot of problems with relying too much on this.

3) injector placement and nozzle design: the better this is, the less the engine has to rely on the above two factors. BMW relies heavily on the spray for mixture formation, but they use expensive injectors that are located close to the spark plug. They have recently gone to a simpler and less expensive design.

The reality is that all three factors come into play, and there are a bunch of tradeoffs and optimizations involved in any given engine.

The 2GR-FSE paper has a chart that went something like "tumble ratio" vs. "flow coefficient", and they explained that they used a high flow port on the 2GR. From context, it seems like they're telling us the 2GR has high flow, low tumble ports. It has a little blurb about "consequently the tumble ratio of the high flow test engine was reduced by one third at the valve-lift divided by valve-diameter of 0.3".

Yup.

Is this to say that the ports are still a compromise between tumble and flow? Is this why you were questioning the viability of increasing revs (intake port flow)?

ding ding ding. The tradeoff between charge motion and flow is well documented. I can explain further with examples from current mass-produced engines for the North American market. You have to optimize all factors--better charge motion helps combustion efficiency and knock resistance, while more flow means more air to burn. If you really dig into that paper and read between the lines you can see that the engineers were grappling with this problem. All the charts really demonstrate the amount of time and work they spent in an effort to find a balance for that application.

Remember: charge motion, piston crown design, and injector configuration are the big factors here. On a modern direct injected diesel you have a lot of the same kind of tradeoffs with spray penetration, fuel pressure, piston bowl shape, and injection timing. On a diesel though the big deal is particulate vs NOx emission tradeoffs.

Originally Posted by Dimman

Couple points to consider.

When they say 'tumble' it is probably in the context of low-rpm/intake velocity.

Tumble ratio is defined as "the cycle integral of instantaneous tumble ratios at each crank angle, weighted by the instantaneous piston speed." In plain English, the tumble ratio looks at the way charge motion and piston speed change over time during a combustion stroke. It can be broken down further using math based on the geometry of the engine. There is a Volvo study and several Ford studies that go deep into this.

Tumble can also be represented through computational fluid dynamics visualizations. Below are flow visualization diagrams for an experimental E85-capable Ford Ecoboost engine. Imagine you are looking down into the combustion chamber from right above it. Red areas indicate flow that is moving towards the piston, and blue areas indicate flow moving towards the head. The chart is for an intake port with a tumble ratio of 1.89. Each image shows the flow fields at different valve lifts.





If you read about the development of the production Ecoboost 3.5 engine as well as the Nissan direct injected engiens (Juke MR16DDT and M56 VK56VD engines) there is a lot of talk about the shape of the edge of the intake port. The curvature and aspect ratios there have a big effect on these flow dynamics, which then affect mixture formation and combustion efficiency.

Now who knows (except some guys in Japan) what happens when you get in there and start tinkering with the new FA engine. Knowing what we know about direct injection though, it's going to be an uphill climb.

If anyone wants to read up more on this stuff you can PM me.

One reason F1 cars were 'only' 12:1 CR was for room for this motion.

Read up on the development of Mazda's SkyActive gasoline direct injected engines and there is significant discussion of this--the shape of the piston on very high compression engines can start to become counter productive depending on a bunch of other factors.

[serialk11r]

Thanks for the response!
So I take it high rev direct injected engines are a challenge to build because of the compromise between flow and combustion efficiency. Hmm.

[arghx7]

but they do exist. Porsche has them for example. Obviously porsches cost way more and can use more expensive systems

[serialk11r]

Hmmm yea I forgot about Porsche. I failed to think of any non-Ferrari examples.

What does Porsche use that Toyota doesn't?

[SkullWorks]

german engineers

[Homemade WRX]

arghx7; have any of the GDI engines that you know of used a quiesent flow for the ports? I know the diesel world wised up to it for increased VE and use the piston obviously for swirl.

And yeah Justin, I'm working with some of the other vendors that we all know well for bigger supporting mods like ECU options, injectors and I'm naturally getting into my hardcore longblock work.

-Micah
3MI Racing

[serialk11r]

Homemade WRX do you have offset/rod specs yet, or if you're not allowed to say, when will you be?

[mrtodd]

But is it really safe to compare the FA directly to other direct injected engines? I mean, we all know that the D-4S system utilizes two injectors per cylinder, supposedly combining the benefits of both fuel injection methods. Perhaps this has already been discussed, and if so please let me apologize, but if not maybe we can dig in a little deeper here. Do we know how the ECU dictates between using both injectors always or just one at a time, depending on load/throttle/rpm?

EDIT: Sorry, okay I found what I was looking for....

http://www.ft86club.com/forums/showthread.php?t=3172

Looking at the "Lexus IS350 2GR-FSE engine" PDF.

[mrtodd]

But I do have another basic question... Is 'squish' considered to produce charge motion? Or is the term 'charge motion' strictly related to intake port characteristics...?

[arghx7]

You could call tumble, swirl, and squish flows all forms of charge motion. A squish flow is basically a concentrated area of mixture.

[Dimman]

Originally Posted by arghx7

You could call tumble, swirl, and squish flows all forms of charge motion. A squish flow is basically a concentrated area of mixture.

The squish area is basically used to 'push' the charge towards the plug as the piston approaches tdc, right?