New 16G Dyno Numbers on C16

[sbengineering]

The blower rotors mix the air and fuel. HOW is that anything like a draw through.

And those big rigs have a "very narrow minded" set up; that way to have all torque low, with no power up top. so again, how is that like a blower.

 

[danl]

PM me if you'd like to go more off topic. Your carburetor(s) atop the blower mix the air and fuel. Else technically its not a carburettor. Also you said you never saw a turbo make torque. I told you of one. Just because you don't like my correct answers doesen't mean I'm not right.

This thread is about the Evo 3 16g. I really don't believe a compressor map from MHI has ever been produced. However even if we use the ones just posted I see no reason somebody with a brain can't make 500 to the wheels with those maps. Especially with charge air cooling and good gas.

 

[tkelly27]

I had thought draw through turbo would be OK as well, however, in high output situations the fuel can actually freeze on it's way to the compressor and ruin the turbo. Even liquid fuel will ruin the compressor blades, just not as fast. Once you are drawing fuel through the inlet of the turbo, you can no longer intercool. You shouldn't really even have low spots as the fuel will puddle.

A centrifugal compressor will be more efficient than a roots style compressor. They each have their own benefits. Why is this even being discussed in this thread?

Rather, let's try and figure out how they got the thing to push 30psi, and how they got the numbers they did while everyone else doesn't. They can't bend the laws of physics, and we know they are limited to tuning with DSMlink. Even if you can't figure out how to make 450 or 500hp, you could still figure out how to help other people break the 400hp mark. Personally, I don't think I would ever need another turbo if I could break 400whp with a 16g.

Here are some questions I have. Why does the power peak, drop off, then come back up again in the 460hp run? In the 449 run it stays flat for nearly 1krpm.

 

[red91gst]

Maybe you shouldn't say your using a actual evo3 16g un altered when your not, then people would not question you. Next time you do this, let someone on the outside go dyno it with to see whats really going on.

IT IS AN ACTUAL ####ING EVO 16G, WITH A PORTED TURBINE HOUSING.

Yet I remind you once again, Curt/Pat's 1g I tuned last fall went 10.33 at 136mph.

You don't understand, I don't have to prove myself. The people that know me, who I work with, and know what I do- know the truth. The rest can go pound sand .

All these threads finish like this... Video the car from when it was made in 1994 until now, video the 16g from the inside during a run in PNC park with 65,000 people watching and one midget sitting in the charge pipe with Al Gore watching the boost gauge.

It happened. Its done. And when it runs 9.80s you will... who the #### am I kidding, you'll still question it! haha

Oh man, I need another beer to take this all in...

 

[red91gst]

Here are some questions I have. Why does the power peak, drop off, then come back up again in the 460hp run? In the 449 run it stays flat for nearly 1krpm.

Boost control

 

[red91gst]

Personally, I don't think I would ever need another turbo if I could break 400whp with a 16g.

EXACTLY my point. 95% of the guys on here haven't even run 11's, something the 16g is easily capable of.

 

[danl]

Yeah lots of guys run WAY to much turbo.

 

[tstkl]

It was 8* out yesterday (-30*F windchill) in Chicago, measured at the MAF. 40*F manifold temp NO problem with my I.C.

Better look at the compressor maps more closely. They rate at 1 psi compressor inlet depression, gain that back and your flow goes up like I said 5-7%.

Also look at the temp compensation formula at the bottom of your typical compressor map, cold temps alone could boost the compressor flow rate up another 5%.

Part selection is not "irrelevant". That short runner manifold alone could have picked up airflow another 3 lb/min on the high end, a fairly typical gain for a 1g-2g DSM, right where the turbo would start slacking off.

no r****d. Regaurdless of turbo, motor, intake manifold, and anything else, a 16g is rated to flow what the compressor map shows. Its rated that, because when tested to its maximum, those were the numbers seen in the labs. Its not like garrett made all their turbo compressor maps based on the 4g63 with the stock intake manifold and maf, and so why the hell would mitsubishi heavy industries? They wanted to know what the turbo was capable in case they ever wanted to apply it to anything else, so why would they create a map that was 4g63 specific. Thats JUST RETARDED. So like I said, based on the airflow of the turbo, it will run out of air to supply 30 psi of boost by 5700 rpms, and Im sure if the OP would like to post said information, it will be very close to that. Because his VE won't be 100%, he might see it drop off slightly higher, but the FACT OF THE MATTER IS, that map can not be extended, so don't bother trying to defy physics with your cracked out theories.

 

[tstkl]

Though to fit an open road in a lab room.

last i checked thats the evo 8/9 compressor wheel/housing map, so accuracy is questionable

 

[pboglio]

no r****d. Regaurdless of turbo, motor, intake manifold, and anything else, a 16g is rated to flow what the compressor map shows. Its rated that, because when tested to its maximum, those were the numbers seen in the labs. Its not like garrett made all their turbo compressor maps based on the 4g63 with the stock intake manifold and maf, and so why the hell would mitsubishi heavy industries? They wanted to know what the turbo was capable in case they ever wanted to apply it to anything else, so why would they create a map that was 4g63 specific. Thats JUST RETARDED. So like I said, based on the airflow of the turbo, it will run out of air to supply 30 psi of boost by 5700 rpms, and Im sure if the OP would like to post said information, it will be very close to that. Because his VE won't be 100%, he might see it drop off slightly higher, but the FACT OF THE MATTER IS, that map can not be extended, so don't bother trying to defy physics with your cracked out theories.

The compressors may be TESTED at their maximum, but they aren't displaying the MAXIMUM on the compressor maps. They stop at 65% efficiency typically. As was mentioned already, that turbo could run at 55% efficiency and with a good I.C. core still pull more mass flow. At 30 psi, 65% comp efficiency, 85% I.C. efficiency, 70*F ambient I get a calculated manifold temp of 107*F. At 55% comp efficiency I get 114*F. Density decrease is minute but massflow rate could increase a decent amount.

From overspeeding you could gain 3 lb/min, removing the MAF another 3 lb/min, and running in cold weather another 3 lb/min. The short runner intake manifold is going to P.R. assist the turbo as long as it can be fed enough massflow, so the problem of maintaining boost pressure near redline also potentially gets solved.

Where am I defying physics?. I've done and datalogged everything I've just discussed EXCEPT running a short runner manifold. I've pulled a recorded 41 lb/min out my little T28 on a 2g maf, and arguably quite a bit more than that when I switched to speed density and yanked the maf. I have no doubt at all an EVOIII 16g could pull 50 lb/min, none.

 

[JayRolla]

Yeah lots of guys run WAY to much turbo.

I think people run too much turbo because its actually easier than getting the most of a small turbo. You can bolt on a 60trim, fmic and fuel mods and get gobs of power. Or spend all the money in the world on cams, intake/exhaust mani's when the stock parts can make good power with a large turbo.

 

[1992awdlaser]

I think people run too much turbo because its actually easier than getting the most of a small turbo. You can bolt on a 60trim, fmic and fuel mods and get gobs of power. Or spend all the money in the world on cams, intake/exhaust mani's when the stock parts can make good power with a large turbo.

I agree 100% with this statement. That is what I first did with this 50trim. Now I am building everything else to get ready for my next turbo. While I am doing this I am going to try to max out the 50 trim. I am aiming for 10's on it with just boost.

 

[DSM90AWD]

The compressors may be TESTED at their maximum, but they aren't displaying the MAXIMUM on the compressor maps. They stop at 65% efficiency typically. As was mentioned already, that turbo could run at 55% efficiency and with a good I.C. core still pull more mass flow. At 30 psi, 65% comp efficiency, 85% I.C. efficiency, 70*F ambient I get a calculated manifold temp of 107*F. At 55% comp efficiency I get 114*F. Density decrease is minute but massflow rate could increase a decent amount

Exactly, great post! You wonder why those 4" FMIC Cores are so popular these days

You ever wonder how all those 14Bs are running stupid amounts of power? Try removing the w/g line and you'll have your answer (can guarantee the airflow line is off the map considerably). Will the turbo last as long as one that's run at a slower wheel speed, no. But to go fast one must make sacrifices

 

[dsm-onster]

The evo4-8 turbo has the identical compressor wheel to the evo3. Down to the hub diameter, the fin thickness, and the blade sweep, depth, etc. Just the scroll is reversed.

It has a larger compressor cover. . . Or MORE flow capability. Also, the coriolis effect would encourage SLIGHTLY more flow, since the scroll is reversed. We're talking about the limit of the evo8 compressor not the smaller cover evo3 compressor with the opposite scroll. Stating that we don't have a map of the evo3 compressor just means that we'll use a compressor map that flows more.

Stretch out the rpm lines at 144K straight from the last angle they show on the graph. The line in purple clearly goes to 50ish% efficiency. 0.32 m^3/sec can be had at that point or 677cfm. Around 46 lbs/min at 1 atmosphere and about 70 degree intake temps (in a dyno shop ? ?).

The red curve is what the likely path the rpm curve would take. Quite a bit less than 46 lbs/min. This is a rpm line, not an efficiency line. Efficiency can be there but not but so much rpms can be. A turbine can only convert but so much work to motion. A turbine can spin a smaller compressor faster (14b at 170K rpms). But only so much work can be used (more rotation inertia evo3 16g 144K rpms).

You must be logged in to view this image or video.

Just because a roots blow operates at 55% efficiency has nothing to do with a turbo compressor. What makes turbos so much better is the fact that they operate to the limit of there rotational speed and still maintain around 60% efficiency. At the edge of the map all the rpm curves eventually converge to a vertical line. It can be seen on many compressor maps. A compressor can flow but so much. In fact, 99% of the time rotational speed on the compressor map is calculated from the efficiency and flow perameters. The less efficiency and the more flow, the higher the rpms. You reach a point of impasse. The higher the flow, the higher the efficiency variable at the point of impasse. See most holset and garrett t series maps for the formula.

And all of this is based on a compressor map with an evo3 16g compressor wheel and larger than evo3 compressor cover.

Could there be some sort of "water tower" effect with the JM fabs race intake manifold? Is the plenum large enough to supply the cylinders their volume demand and then in between peak demand of each cylinder draw, the compressor has enough to resupply the "tank" or "resevoir" like a pump for a water tower? Who knows. We know the evo3 16g can't flow but so much. And 99% of us can't make more than 10 hp per lb/min. We may eventually learn the secrets with our own experience. Until then, we'll keep plugging away.

 

[h8r]

DSM-ONSTER. Thank you for a well thought out and informative post. I always like reading your replies, I gain a lot better understanding on the more technical aspects of turbocharging from you.!

 

[pboglio]

I agree. The ambient side tip velocity of the compressor wheel is going to be 1,200 ft/s or 802 mph at 145,000 rpm. I've always thought that the compressor tips where designed to ride below or around Mach1 at the most for durability and stable operation. So yes 145,000 rpm looks to be the max rotational speed possible based on a back of the envelope calc.

You've illustrated the map perfectly. At 46 lb/min the pressure ratio (P.R.) would be down to 2.4 maybe 2.5. The engine can NOT pull 50 lb/min at this P.R. at 5700 rpm. That is where the intake manifold comes in. It may not increase the massflow rate, but it allows the turbo to operate at a lower pressure ratio where it can whack out high massflow. Then the intake manifold acts to boost the pressure back up during the tuned portion of the rpm band, arguably above 5700-6000 rpm to redline or past. I'd equate it to an inline booster pump. I've run the intake manifold simulators and I've seen values of 6-8 psi generated at the tuned rpms in the intake manifold alone.

 

[dsm-onster]

I agree. The ambient side tip velocity of the compressor wheel is going to be 1,200 ft/s or 802 mph at 145,000 rpm. I've always thought that the compressor tips where designed to ride below or around Mach1 at the most for durability and stable operation. So yes 145,000 rpm looks to be the max rotational speed possible based on a back of the envelope calc.

You've illustrated the map perfectly. At 46 lb/min the pressure ratio (P.R.) would be down to 2.4 maybe 2.5. The engine can NOT pull 50 lb/min at this P.R. at 5700 rpm. That is where the intake manifold comes in. It may not increase the massflow rate, but it allows the turbo to operate at a lower pressure ratio where it can whack out high massflow. Then the intake manifold acts to boost the pressure back up during the tuned portion of the rpm band, arguably above 5700-6000 rpm to redline or past. I'd equate it to an inline booster pump. I've run the intake manifold simulators and I've seen values of 6-8 psi generated at the tuned rpms in the intake manifold alone.

So am I correct to assume that a large plenum can yield a kind of "reservoir effect" and flow to the cylinders can increase for a brief moment (enough to hold to red line)?

A helmholts prepared intake manifold plenum can show pressure throughout the surface? I though that it yields a pressure increase at only the "ends", like at the end of a runner where a tuned runner length is employed. If it's registerable at any surface, 6-8 psi could make an evo3 16g look like it holds to redline. But the pressure just begs more mass flow for it to be maintained, bacause there is no pressure where there is no mass. ???

What I'm saying is that though the helmholts effect increases pressure, it increases flow demand or VE. . . From my understanding.

Unless, I'm overlooking a characteristic or two.

 

[pboglio]

What I'm saying is that when the turbo is maxxing out, the compressor high speedline drops, and so you lose boost at high rpm. BUT, the turbo if it were uncoupled could still keep kicking out more massflow at the lower pressure ratios. The turbo and engine at this point have become mismatched even though there is flow still left in the turbo.

The short runner intake manifold takes advantage of the fact a compressor can kick out more massflow at a lower P.R., although at a lower compressor efficiency. Then to have the engine make use of it the intake manifold boosts the compressor charge to a higher P.R. that the engine CAN use. Its not getting massflow out of thin air. Its basically extending the compressor map to its extreme outer limit.

 

[Black_Bullet]

Wow great posts...

What ive always thought
( perhaps what dsm monster is saying but in a more advanced way of saying it.)
when you increase VE you make it harder for the turbo to supply x amount of psi the higher the rpms/ and engine flow demands go up. Therefore boost drop off in most cases result due to more air mass being pulled at a lower pressure ratio and smims and cams pull more air per psi meaning the turbo is now moving up the compressor map sooner, so you reach the top of your map at a lower rpm that way...
So for instance , say it takes 30 psi and 7k rpm on a 1.8 to max out a turbo or
30 psi at 5k for a 2.3 to max out lbs/ min on that same turbo,
( given both setups were using all the mechanical energy possible from that amount of air flow.)
i thought theyd have roughly the same peak power the turbo can give but just made at different places with different power bands?

The " resourvior" effect mentioned sounds like your saying is a way to keep the turbo dishing out max flow even as you increase rpms instead of air flow drop off? Or in a nutshell, holding x amount of boost for longer...

Im trying to follow,

Pboglio what it sounds like your saying is that you can use max potiential from a turbo if you can continue to increase max flow at a lower pressure ratio to the point were once you do reach the higher pressure ratios its giving more than it wouldve been originally?

Also perhaps our whole assumption on what a 16g can flow was based off of us under misconception to the compressor maps because of the way a "dsm" would use that turbo and not versus other setups.

I mean, i figured itd still only can give the max airflow potiential that that compressor was capable of moving reguardless of were the turbo makes max use of air? I mean obviously one set up/ motor could squeeze more overall out of a particular turbo than another but the limit still becomes how much air can get thru the compressor entry in the first place to provide as much psi as possible as rpms go up before boost drop off.

I can see how colder denser air could have a negligable change on boost holding in higher rpm/ pressure rations.

 

[dsm-onster]

What I'm saying is that when the turbo is maxxing out, the compressor high speedline drops, and so you lose boost at high rpm. BUT, the turbo if it were uncoupled could still keep kicking out more massflow at the lower pressure ratios. The turbo and engine at this point have become mismatched even though there is flow still left in the turbo.

The short runner intake manifold takes advantage of the fact a compressor can kick out more massflow at a lower P.R., although at a lower compressor efficiency. Then to have the engine make use of it the intake manifold boosts the compressor charge to a higher P.R. that the engine CAN use. Its not getting massflow out of thin air. Its basically extending the compressor map to its extreme outer limit.

I see what you are saying. And I completely agree. Certainly an intake manifold upgrade and a set of 272s will push an evo3 16g, 18g, and even a 20g to the end of the map. But, the map does stop. The highest rpm capable by the turbine with that particular compressor inertia decends down the map at an end point almost vertical. There is an end. The top rpm curve eventually rounds to a vertical line.

The top 144krpm curve slices the evo8 map right down. And it looks like at 0.30-0.31 m^3/sec. As shown by the red curve from the above map. You can be loosing boost and still make more power. Of course! But, you can't flow more than what around 144Krpms can give you with that compressor. Your setup can be at 10 psi and it may take the compressor 144K rpms to do that. You'll be all the way to the right and still at that 0.30m^3/sec number.

All compressors do this: all the rpm curves turn vertical as you go right on the map. You can see the likely path of the evo8 compressor 144K curve. I have several DETAILED lecture handouts (50 pages) from a recent fluid dynamics course that specifically delve into turbomachines. It's an intense "read". . . Terrible math But somewere in there, it discusses this from what I remember. I'll gladly email you the package. The Holset compressor speed map of its he431 VGT clearly illustrates my point:

You must be logged in to view this image or video.

It's hard to see, but look how close the rpm lines get to each other at lower PR the farther you go to the right. The lower PR doesn't really help flow when your at very high compressor rpms.

I mean, i figured itd still only can give the max airflow potiential that that compressor was capable of moving reguardless of were the turbo makes max use of air? I mean obviously one set up/ motor could squeeze more overall out of a particular turbo than another but the limit still becomes how much air can get thru the compressor entry in the first place to provide as much psi as possible as rpms go up before boost drop off.

Yes, and yes. That is it. A compressor flows but so much and that is limited by it's rpm. If you could spin it faster, i.e. a bigger wheel that can do more work, then you can raise that number. But the turbine can only do but so much work with a given compressor.

I'll add to that: there is also a limiting factor wrt velocity at the compressor exducer. Pboglio already alluded to this. So a compressor exducer can have a maximum flow capability as well. The evo8 16g can't spin much faster than 144krpms before it develops flow/ durability issues because of mach factor which cause the compressor to exhibit a "wobble" on the initial overspeed episode, or 1 dyno pull. If it does stay together, there is a pressure wave inside the vanes opposing flow.

 

[whitedawg]

I just found a post on the dsmlink forums that said- "I made 423awhp/373awtq, with 43-44 lb/min." Using that ratio (@43lb/min), if you could get 46 or 47 lb/min out of the turbo you could make 460whp. Looks possible to me.

 

[romeen]

No to go off topic but since compressor flow limits and getting more than expected power from a given turbo is being discussed this may be interesting to some.

There is evidence to suggest that pre-compressor water injection changes the compressor's flow characteristics making it in effect behave like a larger turbo. Here is a link for anyone interested in reading more:

Waterinjection :: View topic - Injecting prior to turbo comp' impellers

The thread is very long but has a lot of good info and data (some objective, some anecdotal). But in a nutshell, the pre-compressor injection seems to be able to increase the efficiency of the compressor and cause it to flow more air than the compressor map would otherwise predict. However, despite efforts to avoid it there continues to be problems with damage to (erosion of) the compressor blades.

Potentially, one could use this on a smaller turbo to retain it's desirable spool characteristics while getting the airflow potential of a slightly larger turbo.

I'm not implying that the TPG guys are using this to get these big numbers but who knows (other than them ).

 

[pboglio]

Dsm-Onster,

Excellent post. I would be interested in that technical paper, shoot me a PM. Where we diverge is in the point you made previously. I contend that the high speed curve "would" travel right and downwards and hit .32 m^3/s at some useable pressure ratio a 2.0L motor and SMIM can exploit. This ambiguity in the map is a fine point, but the difference is almost 3 lb/min, instead of 43.7 lb/min choke, it would be 46.6 lb/min choke using standard correction. Now give me a turbo that can pull 46.6 lb/min on or in this case "slightly" off the map, and I personally could boost that up to 50 lbs/min by playing with intake restriction/air temperature alone, assuming all supporting mods.

When I think about it, there really isn't anything stopping the turbo compressor from going beyond 145,000 rpm. Nothing is choked at this point, only the tips are seeing a shockwave from Mach speeds. Not good for turbo but also not choke limiting it either. Yes, efficiency is diving out the window and the compressor wheel is seeing the outer edge of its structural envelope. But there is no speed limiter until the compressor INLET becomes choked, and that nowhere near is happening at 50 lb/min. A bump of another 15,000 rpm would bring a significant bump in compressor flow. Think of the 145,000 rpm speed line as your engines 7000 rpm redline. There are guys including myself who pull 8000 rpm on occasion, why not the turbo so to speak?

 

[danl]

The evo4-8 turbo has the identical compressor wheel to the evo3. Down to the hub diameter, the fin thickness, and the blade sweep, depth, etc. Just the scroll is reversed.

It has a larger compressor cover. . . Or MORE flow capability. Also, the coriolis effect would encourage SLIGHTLY more flow, since the scroll is reversed.

As soon as I read your sentence about the coriolis effect I knew what I had thought all along. Your a moron using words and terms you don't quite grasp. I'm done with this thread. Buncha f uc ki ng morons.

 

[dsm-onster]

Good job, danl.

Now give me a turbo that can pull 46.6 lb/min on or in this case "slightly" off the map, and I personally could boost that up to 50 lbs/min by playing with intake restriction/air temperature alone, assuming all supporting mods.

Remember there is no intake restriction when a compressor map is generated.

When I think about it, there really isn't anything stopping the turbo compressor from going beyond 145,000 rpm. Nothing is choked at this point, only the tips are seeing a shockwave from Mach speeds. Not good for turbo but also not choke limiting it either. Yes, efficiency is diving out the window and the compressor wheel is seeing the outer edge of its structural envelope. But there is no speed limiter until the compressor INLET becomes choked, and that nowhere near is happening at 50 lb/min. A bump of another 15,000 rpm would bring a significant bump in compressor flow. Think of the 145,000 rpm speed line as your engines 7000 rpm redline. There are guys including myself who pull 8000 rpm on occasion, why not the turbo so to speak?

Understandable. But looking at the holset rpm map, there is stoutly diminishing returns as you increase rpms while flowing at a point off the map. I just don't see it. The rpm curves begin to converge onto one line.