Solutions for DI Engines?

[gpshumway]

Originally Posted By: A_Harman
I disagree with most of this. GDI engines may have first come to the public eye in racing at the LeMans 24-Hour, but that is not where its development started. Mitsubishi introduced wall-guided GDI engines into the Japanese market in the late '80's and early '90's, but they were not terribly successful due to problems with excess soot from the combustion process.


Mitsu may have been the first with automotive GDI engines, but the first GDI engine I'm aware of in mass production was the Daimler engine in the Messerschmitt Bf 109 WWII fighter plane. http://www.battleofbritain1940.net/0013.html

At any rate, the deposit issues seem to break down into two categories, combustion chamber and intake valve. A high temperature detergent like Shell's V-Power or Redline SI-1 will likely help with the former, but regular solvent based intake cleaning is likely the only thing which will help the latter.

You can get a reasonable idea of problem and how solvable it is by following BG Products' Ford Taurus SHO (GDI) on their blog: http://www.bgfueltest.com/

BG is the OEM supplier of intake cleaning solutions to VW/Audi among others. Wynn's is another OEM supplier, both have a reusable injection tool which uses shop air to inject cleaner into the intake of a car. These products and tools are marketed primarily to professional mechanics, but I've seen them for sale to the DIY mechanic on the internet, ebay is of course always a good source.

On the more DIY side, I've used a variety of induction cleaners on my MPFI cars in the past with good results. My particular problem child was a 1998 Mercury Mystique (RIP) with the 2.5l Duratec V6. This engine is unique in that the fuel injectors act on only 6 of the 12 intake valves with the variable length manifold feeding the other intake valves from separate, shorter runners. Because of this, the Duratec V6 sees intake valve deposits similar to a DI engine due to the lack of detergent/fuel flowing over the secondary intake valves. The Contour Enthusiast Group (contour.org) had lots of threads on intake cleaning, but the database is getting unreliable due to age and lack of interest. The consensus there is solvent cleaners work reasonably well, but only if you don't let the deposits get too thick. Many Contour/Mystique owners (myself included) did the cleaning once a year (15k miles for me). A thread showing the problem:
http://www.contour.org/ceg-vb/showthread.php?2993-Cleaning-UIM-LIM-dirty-pictures-enclosed

Amsoil Power Foam, Berryman's B12, GM Top Engine Cleaner (now discontinued in aerosol), Mopar Combustion Chamber Cleaner (rebadged Wynn's) and the 3M DIY intake cleaner kit have all worked well for me. The 3M kit is my favorite (though expensive) because of it's convenient self-spraying can and hose which routes the cleaner to a central location in front of the throttle body. Seafoam also makes an aerosol can with a long tube for routing to the throttle body, but I find the harsher solvent cleaners to be more effective than Seafoam. The other cleaners mentioned above require finding a convenient vacuum line.

3M video tutorial on their kit (DVD comes with the kit):
http://www.youtube.com/watch?v=q6G0zb4QcIs&feature=relmfu

MFR Links
http://www.amsoil.com/storefront/apf.aspx
http://www.bgprod.com/products/fuelair.html
http://www.wynnsusa.com/fuel.aspx

I hope that helps!

 

[A_Harman]

Originally Posted By: gpshumway
Originally Posted By: A_Harman
I disagree with most of this. GDI engines may have first come to the public eye in racing at the LeMans 24-Hour, but that is not where its development started. Mitsubishi introduced wall-guided GDI engines into the Japanese market in the late '80's and early '90's, but they were not terribly successful due to problems with excess soot from the combustion process.


Mitsu may have been the first with automotive GDI engines, but the first GDI engine I'm aware of in mass production was the Daimler engine in the Messerschmitt Bf 109 WWII fighter plane. http://www.battleofbritain1940.net/0013.html



I was thinking of electronically controlled GDI engines. An earlier mechanically controlled GDI automotive engine that I can think of is the Mercedes LeMans racers (W196?) from ~1955. These had 3L direct-injection straight-eights. If I recall correctly, these engines were all designed with injection timing on the intake stroke only. They made no attempt at charge stratification by injecting on the compression stroke.

 

[gogozy]

sorry not a solution here but i wonder additional additive such as sea foam in the fuel and crank case will make the matter worse? isn't additive help break down oil and think them out, and carry more heavy stuff from oil in the vapor, and may cause more deposit on the valve? so i also wondering about oil with high Noack Volatility, shall DI engine avoid these oil with high Noack Volatility reading?


my personally view of DI is just a transitional product... like MMT as substitute for lead in the gasoline... the carbon deposit and loss of thermal efficiency will quickly rob the MGP and power off the engine as soon as it leave the dealer lot. and IF not even sure if Hyundai agree the carbon deposit is warranty claim (i had seen door handle colour fade as "regular wear and tear" so it's not an warranty on Hyundai so, better ask first)

 

[A_Harman]

I see a lot of new cylinder head designs from manufacturers all around the world in my job. All of the ones I have seen in the last 3 years have direct injection. I think DI is here to stay. The problems with intake valve carboning will be solved, but there may be a few bumps along the way.

 

[demarpaint]

Originally Posted By: A_Harman

I disagree with most of this. GDI engines may have first come to the public eye in racing at the LeMans 24-Hour, but that is not where its development started. Mitsubishi introduced wall-guided GDI engines into the Japanese market in the late '80's and early '90's, but they were not terribly successful due to problems with excess soot from the combustion process.


Fast forward to 2011, and they still haven't perfected them, that scares the carp out of me.

 

[gpshumway]

Originally Posted By: A_Harman


I was thinking of electronically controlled GDI engines. An earlier mechanically controlled GDI automotive engine that I can think of is the Mercedes LeMans racers (W196?) from ~1955. These had 3L direct-injection straight-eights. If I recall correctly, these engines were all designed with injection timing on the intake stroke only. They made no attempt at charge stratification by injecting on the compression stroke.


How silly of me to forget the Merc LeMans engines and the contemporary 300SL engine. Mercedes likely built on lessons from the BF109 engine, no?

As to stratified charge, I thought that was primarily in Europe, where looser NOx emissions requirements along with special catalysts allow for lean-burn mode. I know the VW/Audi GDI engines run stoic here in the US, but have lean-burn in Europe.

Even while running stoichiometric mixtures, evaporative cooling during the compression stroke should help prevent knock and allow higher compression, right?

Are some manufacturers running lean mixtures here in the US? Possibly in milder form than their European brethren? Are some makers using a non-homogeneous, stoic mixture? Doesn't that result in both higher NOx and HC emissions relative to a homogeneous mix?

Thanks for the input.

 

[gpshumway]

Originally Posted By: demarpaint

Fast forward to 2011, and they still haven't perfected them, that scares the carp out of me.


Well, "perfected" is subjective. MPI engines need detergents which carbureted cars didn't, now DI engines need intake cleanings which MPI engines didn't.

The biggest problem with DI engines IMHO is manufacturers not being straight with consumers about the extra maintenance they require. VW/Audi and BMW shouldn't be recommending 10k mile oil changes and all manufacturers should be recommending frequent intake cleaning.

IIRC Hyundai recommends 3k OCIs on their DI engines, which is a step in the right direction. Making DI tolerant oil like RLI's formulation more available and required in DI engines would be a huge leap in the right direction.

 

[Jeff_in_VABch]

Originally Posted By: A_Harman
I see a lot of new cylinder head designs from manufacturers all around the world in my job. All of the ones I have seen in the last 3 years have direct injection. I think DI is here to stay. The problems with intake valve carboning will be solved, but there may be a few bumps along the way.


I agree, DI is certainly here to stay.

This is mainly much ado about nothing in the first place. IF, we KNEW there would always be massive intake valve buildup of whatever material, we could just reduce stem size and/or undercut the stem above the valve head and configure the intake port to maintain a consistent flow as obstruction increases throughout engine life. However, if this were a real problem we would most likely develop a solution to eliminate the deposited material. I.E. is this is oil related due to valve stem leakage because of stem seals leakage or oil composition, rectify that. If this is combustion gas issue, change the valve overlap period or exhaust port scavenging timing (by exhaust & port design) to rectify it.

This topic is mostly a bunch of armchair mechanics discussing nothing.

Or please feel free to full a spark plug from your DI engine and show me the problem using a cheap snake scope camera you can get from Home Depot or anywhere nowdays.

 

[demarpaint]

Originally Posted By: gpshumway
Originally Posted By: demarpaint

Fast forward to 2011, and they still haven't perfected them, that scares the carp out of me.


Well, "perfected" is subjective. MPI engines need detergents which carbureted cars didn't, now DI engines need intake cleanings which MPI engines didn't.

The biggest problem with DI engines IMHO is manufacturers not being straight with consumers about the extra maintenance they require. VW/Audi and BMW shouldn't be recommending 10k mile oil changes and all manufacturers should be recommending frequent intake cleaning.

IIRC Hyundai recommends 3k OCIs on their DI engines, which is a step in the right direction. Making DI tolerant oil like RLI's formulation more available and required in DI engines would be a huge leap in the right direction.


Ok so "perfected" was a poor choice of words. How about this, get DI engines to be a little easier on oil in the fuel dilution category, and make adding a bottle of a good FI cleaner once a year all it takes to keep them running like an MPI engine would. I just read Hyundai is suggesting 3,000 mile OCI for their DI engines. Maybe it's just me but I see it as a step backwards.

I'm certain it is here to stay, and will be awesome when the bugs are worked out. Until then let the buyer beware.

 

[440Magnum]

Originally Posted By: A_Harman
Originally Posted By: ProfPS
You have to understand that DI came out of the racing industry.


I disagree with most of this. GDI engines may have first come to the public eye in racing at the LeMans 24-Hour, but that is not where its development started.


Quite true. Curtiss-Wright produced many gasoline direct-injected R-3350 airplane radials in the 1950s, for example. All of the engines in the late Lockheed Super Constellations and Douglas DC-7Cs were Wright direct-injection turbo-compound R-3350s. The system was purely mechanical/pneumatic, but it worked and worked well. As with many "modern" inventions, technology has just made something very old producible for the masses at reasonable cost, not introduced a fundamentally new idea.

PS- a gasoline direct-injected 1950s vintage R-3350 won the Reno National Championship Air Races in 2006.

 

[A_Harman]

Originally Posted By: gpshumway
Originally Posted By: A_Harman


I was thinking of electronically controlled GDI engines. An earlier mechanically controlled GDI automotive engine that I can think of is the Mercedes LeMans racers (W196?) from ~1955. These had 3L direct-injection straight-eights. If I recall correctly, these engines were all designed with injection timing on the intake stroke only. They made no attempt at charge stratification by injecting on the compression stroke.


How silly of me to forget the Merc LeMans engines and the contemporary 300SL engine. Mercedes likely built on lessons from the BF109 engine, no?

As to stratified charge, I thought that was primarily in Europe, where looser NOx emissions requirements along with special catalysts allow for lean-burn mode. I know the VW/Audi GDI engines run stoic here in the US, but have lean-burn in Europe.

Even while running stoichiometric mixtures, evaporative cooling during the compression stroke should help prevent knock and allow higher compression, right?

Are some manufacturers running lean mixtures here in the US? Possibly in milder form than their European brethren? Are some makers using a non-homogeneous, stoic mixture? Doesn't that result in both higher NOx and HC emissions relative to a homogeneous mix?

Thanks for the input.


It wouldn't surprise me if some of the old DB601 engine technology made its way into the Mercedes racers of the postwar era. The engineers that designed those engines had to do something, after all.

Yes, I think the European's are going more aggressively for lean burn due to the higher limits of NOx. True stratified charge, lean-burn combustion in gasoline engines would probably need new emissions control technology, such as lean NOx cats. If I were Emperor of the EPA, I would decree higher NOx limits across the board so that engines could be made more efficient.

Higher compression ratio due to charge cooling during the compression stroke is one of the advantages of DI, and another benefit is improvement of air capacity due to injecting fuel directly in the cylinder.

 

[Old Mustang Guy]

Interesting topic. For my D.I. (my wife's 09 Acadia) I'm running 5k oci's on synthetic (PP or QSUD) and top tier fuel. Otherwise I'm just choosing not to worry about it.

 

[BMWTurboDzl]

Use a low saps oil and halve your oci.

 

[INDYMAC]

I don't have a DI application. But after reading this thread I would probably look for a fully formulated low saps PCMO or HDEO with highest solvency possible in SN/GF-5 or CJ-4. I might even try a maintenace dose of ARX in it...an annual vacuum intake valve cleaning with BG cleaner would be on my menu also.

Oil analysis would definitely determine my OCI's, and not the manufacturer.

 

[440Magnum]

Originally Posted By: INDYMAC
I don't have a DI application. But after reading this thread I would probably look for a fully formulated low saps PCMO or HDEO with highest solvency possible in SN/GF-5 or CJ-4. I might even try a maintenace dose of ARX in it...an annual vacuum intake valve cleaning with BG cleaner would be on my menu also.

Oil analysis would definitely determine my OCI's, and not the manufacturer.


Not sure oil analysis would tell you anything. DI engines aren't any harder on the oil IN the crankcase than any other engine. Their problem is that there's no fuel in the intake plenum to "wash" the intake valves. Shell can put all the nitrogen detergents in the fuel they want, but if the fuel never gets to even SEE the deposits (the valve stems are never touched by fuel) then its all for naught. Same for ARX- not going to do a darn thing for the backsides of the intake valves.

But yes, a low volatility oil is probably going to help, as would periodic intake cleanings (although that carries its own risks).

 

[JHZR2]

But the claims of some of the MFRs is that it is thermally stable and may survive combustion and loop around through the PCV.

 

[demarpaint]

I think the solution to the problem lies in an Inverse Oiler until the bugs are ironed out. Time will tell.

 

[Alex38]

I just won't be buying any vehicles with DI until they have been around longer and the manufacturers come up with a way to avoid this mess. I find it curious that Honda, which builds some of the best engines made, has not embraced DI with it's lineup in the way that some other companies have.

 

[Bill in Utah]

I think what we will see is a throttle body injector (or some type of injector) placed in the intake track that will at times throw some fuel in BEFORE the valves that washes the oil/deposits off the valves.

When being used with the DI system that can back off not to make the intake charge richer.

Bill

 

[demarpaint]

Originally Posted By: Alex38
I just won't be buying any vehicles with DI until they have been around longer and the manufacturers come up with a way to avoid this mess. I find it curious that Honda, which builds some of the best engines made, has not embraced DI with it's lineup in the way that some other companies have.


I think Honda is onto something, and if/when they bite the bullet they'll have it right.