The fundamental flaw of the DISI?

[Mazdasloww3]

Saw a comment the other day from someone who claimed they knew a Ford powertrain engineer, and claimed they said that Ford didn't use the DISI (yes the Ecoboost is based on it, but not the same) because it was essentially a parts bin engine thrown together because Mazda really wanted to use direct injection. Mazda pretty much just threw a hole for the injector in the existing MZR head, which has more of a swirl flow and didn't have the proper tumble flow needed for DI to have good air/fuel mix in the combustion chamber. Anyone heard of this?

Looking into it, it makes a lot of sense. Compare the DISI head to the Ecoboost and other DI heads like even the Skyactiv, and a key feature you notice is the narrower valve angle - a big part of generating more of a tumble flow. Also notice smaller valves, which also increases tumble flow, as well as a seemingly steeper included angle of the injector with more of a guide machined into the cylinder head for the injector. It would also go to explain why the pistons in the DISI have an uncharacteristically wide and indirect fuel bowl compared to the much more targeted fuel bowls seen in most other DI pistons with an offset injector.

The reply mentioned the improper flow causing fuel washing down the cylinder walls, stripping lubrication, and causing fuel dilution. This is a known issue with the DISI, and in addition, wall-wetting causes carbon to build up on hot spots, such as the rings. However, I believe that clearly the biggest issue in all of this, is the inconsistent and incomplete combustion from the poor air fuel mixture, leading to this engine's seemingly high tendency, relatively speaking, to lean out and torch pistons, and being more prone detonation and hotspots.

Throw in the poor flowing intake manifold (ford changed the manifold for a reason), normal DI issues, especially when you throw a turbo on top, along with this platform's other quirks, and it all adds up. Although makes sense for an innovative engine like this. The DISI had to walk so the Ecoboost and Skyactiv could run.

Unfortunately, I'm not too sure what can be done with this information at this point in the speed's life cycle. A custom head especially with R&D required seems out of the question. The Ecoboost head has that integrated exhaust manifold, and a different HPFP design, but the water jacket passages do look similar at the least, so who knows. Curious to hear what y'all think of this or if this has been looked into before.

 

[SyntheticAtmosphere]

What you have heard has merit for sure. At this stage, we will have to live with it.
I do think that Ford effectively stiffed Mazda with the bill...Mazda gets the bad press while Ford totes the Ecoboost as a FoMoCo winner. However the Ecobang is bogged down with the design being Ford-ified, LOL (Head gaskets & overheating, piston issues).
Judging by the way both manufacturers quickly dropped our version and moved on proves it was just a means to an end. The SkyActive in my wifes CX5 is an absolute rock solid appliance that Mazda was able to clean sheet with the information from my Speed6.
Most early adopters scabbed DI onto existing engines and failed miserably, then proceeded to bandaid the designs with more failure...Ford/GM/Chryco, LOL. Clean sheeting new technology has led to some fantastic failure stories throughout history...pick your 70s euro-trash, LOL.

I like the 'ol DISI. Its a link in the chain and there are many cars I can find that are better after 20 years. When coupled to the 6spd and AWD in my Speed6 it creates a great GT sedan.

 

[Mazdasloww3]

What you have heard has merit for sure. At this stage, we will have to live with it.
I do think that Ford effectively stiffed Mazda with the bill...Mazda gets the bad press while Ford totes the Ecoboost as a FoMoCo winner. However the Ecobang is bogged down with the design being Ford-ified, LOL (Head gaskets & overheating, piston issues).
Judging by the way both manufacturers quickly dropped our version and moved on proves it was just a means to an end. The SkyActive in my wifes CX5 is an absolute rock solid appliance that Mazda was able to clean sheet with the information from my Speed6.
Most early adopters scabbed DI onto existing engines and failed miserably, then proceeded to bandaid the designs with more failure...Ford/GM/Chryco, LOL. Clean sheeting new technology has led to some fantastic failure stories throughout history...pick your 70s euro-trash, LOL.

I like the 'ol DISI. Its a link in the chain and there are many cars I can find that are better after 20 years. When coupled to the 6spd and AWD in my Speed6 it creates a great GT sedan.

I agree with you, it's necessary and pretty expected for trying out such a new technology.

The Skyactiv is amazing. I found a 2013 Sky 3 for my brother, and I'm always shocked by the gas mileage it achieves. I read an interview where a Mazda engineer mentioned that a ton of manufacturers reached out and asked if they'd be open to sharing technology after they released the Skyactiv. I like how Mazda innovates with their engineering.

Yeah Ford's decision to go open deck when these blocks are so strong and almost never have head gasket issues was extremely questionable, especially on a turbo engine. I don't even think they fixed that for a while/ever. Ford will always be Ford lol.

I like the DISI too for what it is. Hard not to get nervous after reading through all the posts of people who blow these left and right. As finicky as these are though, I think there's enough knowledge out there that the risk can be greatly mitigated if you're conscious about it.

 

[automaticonlyspeed3]

I don't know too much about engine designs to say the least, but from what I know about this platform..

Most failures are...

-Lspi --(this might be what your referring to)

-full Throttle in 6th gear..

-full throttle in low rpm especially in higher gear..

-bad tunes and pushing the car too hard.

-too much torque

-fuel system.. (leaking or lack of fuel etc)

-----

Using the right oil helps.

HPFP upgrades helps with fueling issues.

They fixed the pistons in the second gen..


a normal stock speed3 can EASILY go 100k miles if driven correctly.. a modded one can do that too as well if done correctly

If the engine was that flawed, I feel like people would be blowing them up 20k- 40k in.. (some did user error of coarse)

Edited

 

[Mazdasloww3]

I don't know too much about engine designs to say the least, but from what I know about this platform..

Most failures are...

-Lspi --(this might be what your referring to)

-full Throttle in 6th gear..

-full throttle in low rpm especially in higher gear..

-bad tunes and pushing the car too hard.

-too much torque

-fuel system.. (leaking or lack of fuel etc)

-----

Using the right oil helps.

HPFP upgrades helps with fueling issues.

They fixed the pistons in the second gen..


a normal stock speed3 can EASILY go 100k miles if driven correctly.. a modded one can do that too as well if done correctly

If the engine was that flawed, I feel like people would be blowing them up 20k- 40k in.. (some did user error of coarse)

Edited

Yes LSPI is the main issue that arises from this poor mixture. I guess I was referring to the gen2 piston design in my post, I wasn't even aware the pistons were changed in the gen 2, apparently the first gen pistons don't have a fuel bowl at all, interesting. It makes sense the fuel bowl is so non directional then; the injector placement and spray pattern into a fuel bowl wasn't a design consideration when the DISI was first made.

Yes one does have to be mindful driving the DISI to avoid this condition, as you mentioned, you can find some users reporting premature engine failure a couple years after first released (I do imagine most were related to tunes/modifications). With how common Turbo GDI engines are nowadays, OEMs have to be mindful of how to avoid this with the average consumer. Obviously automatic transmission shift logic helps a lot, but even then, there are still plenty of enthusiast cars with manuals that need to be stout enough to handle some uninformed teenager romping on it (or maybe granny lugging it - these engines can feel great due to the torque at low speed high load). I'm sure all the more modern TGDI cars are frequently running around with more uninformed drivers, generic tunes, and mods with no tune without nearly as high of failure rates.

And of course, OEMs main goal in reducing LSPI is for emissions and fuel economy purposes. If they are more reliably able to reduce the risk of detonation by engine design, they can run leaner air fuel mixtures (more fuel from running richer cools cylinder), more timing, more boost, and higher compression ratios, which improves thermal efficiency and emissions, by far the biggest consideration in engine design nowadays.

Looking it into it more, the integrated exhaust manifold in the Ecoboost (along with essentially every other modern TGDI engine) may help benefit this as well. The obvious benefit is faster spool and reduced time to catalytic converter warmup, along with easier packaging and lower materials cost. However, the integrated manifold is also lowers EGTs via coolant routing in the water jacket (goes to further explain EB cooling issues.) The twin scroll designs most likely have a decent scavenging effect as well, especially without utilizing equal length runners, which OEMs usually don't do with traditional headers, which could further decrease cylinder temps by consistent and efficient evacuation of the hot exhaust gasses. Too bad UEL sounds so cool.

Edit: Thinking of it, probably gives more of an explanation to Ford's choice for an open deck. I'm sure the engineers determined the extra cooling of the cylinders was beneficial/necessary for their design goals. Or maybe it was simply to reduce manufacturing cost, I'd like to believe that knowing Ford. Lower temps doesn't mean much when your head gasket fails.

And yeah, as I said, I like the DISI. I'm not trying to say it's some completely terrible lost cause of an engine at all lol. I just think it's interesting to gain more insight into why exactly it has the characteristics it does.