The 55 lb number is definitely at 30 psi! If I wrote 25 I had a brain fart or typoed. If you know offhand what thread that was in PM me a link and I'll fix it. 62 lbs/min sounds way too high. In practice, I can't think of a single example where I saw someone move that much air at that boost. Even just looking at the 60-1, which was extremely common a few years back, and flows ~62 lbs, everyone was maxing it out in the mid 30s for boost. Who knows though, we'd have to compare both calculations and see if it was done the same way.

This number was arrived at mathematically, which can leave some room for discrepancy in the real world. I have seen it be quite accurate on many cars, but it does vary by a couple lbs. Some poeple are also able to get better than 10 whp per lb/min. 578 at 28 psi is impressive. Any chance this was on a 2wd dynojet though? Dyno numbers can vary tremendously one from type to another, and even from one unit to another of the same type. Sometimes all the tolerances stack up and make for an interesting case.

What I don't get is, how is it limited to ~50 lbs/min on pump. I understand that the static compression increases as boost does, but a larger turbo which flows more at lower psi's and doesn't turn into a hairdryer at higher boosts should increase flow (in my opinion). A set of nice cams and a high flowing ported head should only increase the efficency of the engine and result in higher airflow numbers. I bet you could see crazy numbers off pump gas if the car was designed with that in mind.

Click to expand...

The limit is not really based on airflow, but on boost. I believe it was in this thread I started to talk about this. You are typically limited on pump gas at around 25 psi. What you get for airflow depends on the mods as you stated. On all of my 2 liter setups, this worked out to 50 lbs. On the 2.3 it was closer to 60 lbs. I usually lowered the boost to 20 psi though since it was pretty much a race car at that point and I wanted to keep it together for the race gas runs it was built for (so much for that idea ). That kept airflow to low-mid 50s. All of your points in your post are valid.

I still don't buy the hairdryer thing though, and I have been catching a lot of flack for this on this forum However, I ran the EVO last night for the first time with the AEM. Now this setup is well documented, I've been running it for over a year with DSMlink. With AEM I can finally confirm my theory. Air temps only rose 9 degrees from idle to the top of 4th gear during a full quartermile pass with the turbo maxed out. That's not enough of an increase to make it even worth considering, IMO. I can get the RPM it hit max flow at if anyone really wants to know, since I was running a boost controller this time (pump gas). I'm going to run it next week on race gas now that the tune is in the ballpark, which means no signal to the WGA and a maxed out turbo for the entire run. It will be interesting to see if the rise in temps is any greater in this configuration. I'll try to remember to post that data when I get it. For reference, this is on the SBR FMIC, same 24x12x3.5 core they use in the DSM kits (I ran it on my 2g as well). So to reiterate my position on this, the turbo may turn into a hairdryer before the IC, but after the IC and at the motor, there doesn't seem to be a problem, with a decent FMIC.

37 lbs/min @ 19 psi on a 50 trim with stock 6 bolt long block (2.73s 70-90 time).

95GSXracer said:

The 55 lb number is definitely at 30 psi! If I wrote 25 I had a brain fart or typoed. If you know offhand what thread that was in PM me a link and I'll fix it. 62 lbs/min sounds way too high. In practice, I can't think of a single example where I saw someone move that much air at that boost. Even just looking at the 60-1, which was extremely common a few years back, and flows ~62 lbs, everyone was maxing it out in the mid 30s for boost. Who knows though, we'd have to compare both calculations and see if it was done the same way.

Click to expand...

Using the calculator below, you can come up with 69 lb/min at 30psi (using 0C inlet temp, 100%VE, 100% compressor efficiency, and 100% IC efficiency).

http://www.not2fast.com/turbo/glossary/turbo_calc.shtml

Obviously, those numbers aren't possible in the real world (yet), but I've found that calculator to match DSMlink's logged lb/min pretty closely and it comes in handy for figuring out how efficient an IC or compressor is. We've always focused on running the best times at the track with C16, but lately we've been playing more with pump gas power. I'm really trying to find some hard data to show that switching from a bolt-on hotside to a T3 (for example) would be a worthwhile gain on pump.

Kevin,

Thanks for that post-- I found your theory-to-practice observations of IAT to be very informative. I don't mean to put you on the spot, but do you have a link to the formulas used to arrive at those flow numbers for 100% VE? I'm also curious if they assume 100% IC efficiency or some other value. If you don't have anything readily available, it's ok. I'm sure there's plenty of internet to be searched on the subject.

Steve,

I just entered "real world" parameters into that calculator, except for 100% VE, and came up with 50 lb/min @ 30 psi, which supports what Kevin posted earlier. I'm sure I could make the numbers slightly more favorable to get closer to 55 lb/min. Thanks for the link.

That calc is interesting. The IC appears to be about 80% at the top of 4th gear, but 75% for IC efficienct is pretty optimistic. Using 60% I still only get 83% though. Which backs up my belief that IC efficiency is way more important than compressor efficiency. I wish I had a temp sensor before the IC. They are relatively cheap, I might as well install one and log it. The only variable I don't have is TC outlet temp, which I could use to derive TC efficiency fully maxed out.

95GSXracer said:

That calc is interesting. The IC appears to be about 80% at the top of 4th gear, but 75% for IC efficienct is pretty optimistic. Using 60% I still only get 83% though. Which backs up my belief that IC efficiency is way more important than compressor efficiency. I wish I had a temp sensor before the IC. They are relatively cheap, I might as well install one and log it. The only variable I don't have is TC outlet temp, which I could use to derive TC efficiency fully maxed out.

Click to expand...

I've got an Autometer dual channel peak/memory thermocouple air temp gauge. I see peak temps of 390F pre-IC and 120-130F post-IC. This is at 25-26psi on a sleeper 20g. Those numbers tell me that the turbo is under 70% efficient and the IC is close to 90% (water/air). Even with a relatively efficient IC, I still get knock on pump/meth and I can't run very aggressive timing. I'm tempted to "downgrade" to an EVO 16g to see if the compressor efficiency will help any for this setup. I agree that IC efficiency is more important, but it'll only go so far in overcoming an inefficient compressor.

If the temps at the IC outlet are reasonable, it does not matter what they are before the IC. The motor will never see it. If you are knocking, it is due to other limit factors, not compressor efficiency.

Not entirely correct. The motor may see the effects of a very inefficient compressor because:

A.) The compressor will have to overcome its own inefficiency and spin faster to maintain flow since alot of its energy is being converted directly to heat.

B.) The energy used to create heat comes from somewhere. Thats where the engine sees the effect on the turbine side. Since the compressor is spinning faster than ideal, the turbine must spin faster as well. The faster the turbine has to spin, the more backpressure you create at the exhuast port. Eventually, exhaust gases don't get fully scavenged from the cylinder and you end up with hot gases mixing with the fresh intake charge. That becomes a breeding ground for knock under any aggresive timing or fuel cuves.

95GSXracer said:

If the temps at the IC outlet are reasonable, it does not matter what they are before the IC. The motor will never see it. If you are knocking, it is due to other limit factors, not compressor efficiency.

Click to expand...

I would just like to say, with my 14b, I didn't see crazy knock in higher boost, with 23ish degrees of timing on pump gas. I had to lean the car out like crazy to get it running well up top, and since all I have is the maft, that left me sitting on a 25 degree timing curve which slightly decreased with hints of knock here and there, until 5k rpms when it would explode with knock. My guess is this was around a 10.5-11:1 afr. I found that the rpm point limited it more, which would probably do to being on the extreme of the compressor maps edge, and hitting 60% or less efficency. Now, I don't have a huge fmic like you do, but my 2.75*25*8 inch core should be able to cool pretty well imo. Intercooler efficency realistically will sit around 85%, and with a turbo going into the 300+ degree margine, and the the ambient temps around 80, thats 33+ degrees. I can't really explain with what was happening there, nor can I disagree with you completely. I just think its a tad bit weird what I was seeing. Clean logs til a certain rpm point, and that point not beeing the point where peak boost is seen. What else COULD it be besides intake temps?

pickens, I don't disagree with your points, but the fact remains that the best time that can be achieved on a turbo is with it completely maxed out, so I have to wonder if all that stuff really matters.

tstkl, what does your timing curve do at 5k rpm? IIRC, the 2g map wouldn't really start to ramp up the timing until 5k rpm. In fact the stock 2g curve doesn't even cross 10 degrees until 5k rpm. Who knows what it really was if you were on a curve that results in 25 degrees! I would have to assume your issue is still not air temp related, but timing related, and timing to this day in all of my experience is the single biggest factor involved in knock.

And, as always, I appreciate when people do disagree with me. I'm not always right, and it keeps me thinking.

This thread !
I hope that someday I can have at least half of the tuning knowledge all of you have.
My hat's off to all of you.

Sorry OP, I just had to commend the fellas.

Kevin,

Don't forget that tstkl has 1G timing maps, which I'm sure you know are a bit more aggressive than the 2G's. I don't think it's all that uncommon for the guys making big power on AFCs and MAF-Ts to get a ridiculous amount of timing up top, especially the 1G guys that have quite a bit to begin with. The artificially lowered airflow signal to compensate for the larger injectors combined with the aggressive stock timing will give you a recipe for some serious knock!

tstkl said:

between kevin's post and yours, I'm leaning towards the e316g. I want something that will make me piss my pants when I floor it at 80 on the freeway, but it just seems so much more practical to have that little bad boy toss me around at low speeds. Also, I'm less likely to kill myself with an evo than I am with a bigger turbo.

What I don't get is, how is it limited to ~50 lbs/min on pump. I understand that the static compression increases as boost does, but a larger turbo which flows more at lower psi's and doesn't turn into a hairdryer at higher boosts should increase flow (in my opinion). A set of nice cams and a high flowing ported head should only increase the efficency of the engine and result in higher airflow numbers. I bet you could see crazy numbers off pump gas if the car was designed with that in mind. Most people just don't waste their time with pump gas when they are running aftermarket cams, on a ported head, with a large turbo, at high boost, with a built bottom end. I only say built bottom end because I don't think the stock internals could survive the power that over 60 lbs/min would make. Also, something like propane injection or water, (which doesn't increase the octane rating of the fuel, but only cools the intake charge) would help drastically. I haven't found anyone using propane in the dsm world, but when i was into 3kgts, I found a couple site selling the pro-pain kit, which advertized sub zero temperatures due to the compressed combustable propane.

Then again, the static compression probably has a larger effect on it than the intake air temps.

Click to expand...

If you read all of Kevin's post you see the part where he hints to boost being the true limitation on pump. That being said, a 2.0 with all the mods is going to have a hard time getting much past 50lb/min on pump (w/o meth/alky). With cams, smim, and a big turbine housing I was seeing 47-48 lb/min in cool weather. Certainly there was some more to be has with bigger cams and a good head (i was on a stock 2g head and 264/272 cams) but you're not talking 10 lb/min more.

Guys running a good form of Meth/alky injection can run a good amount more boost and see the big airflow numbers from it.


Also, I have a problem with "larger turbos flow more at low psi" as they don't. Certainly they have the capability to flow more at a given psi, but they will flow the same on the same motor at the same psi, unless you either: a) outflow one of the compressors, or b) run a larger hotside to improve VE.

donmagicjuan said:

Don't forget that tstkl has 1G timing maps, which I'm sure you know are a bit more aggressive than the 2G's. I don't think it's all that uncommon for the guys making big power on AFCs and MAF-Ts to get a ridiculous amount of timing up top, especially the 1G guys that have quite a bit to begin with. The artificially lowered airflow signal to compensate for the larger injectors combined with the aggressive stock timing will give you a recipe for some serious knock!

Click to expand...

That's the right answer. I've seen 1G's peak at close to 26-27 degrees of timing if the airflow subtraction was aggressive enough on a 1G map. I'm sure in the "olden" days when tuners were correcting ND660's on an AFC or MAFT that numbers were even higher.

andymoraitis said:

That's the right answer. I've seen 1G's peak at close to 26-27 degrees of timing if the airflow subtraction was aggressive enough on a 1G map

Click to expand...

If you datalog timing vs RPM and compare to the 1990 Timing Map you can predict at what airflow your ECU thinks you are running (post-MAFT compensation).

Can also compare to the 91-94 timing Maps ) which are slightly more aggressive > 5K RPM.

This is one of the main reasons people swich to a DSMLink / Modified EPROM so that large adjustments of airflow outside the ECU do not put you on a more unfavorable map. Also allows you to "tune out" that big timing spike at 6K

Both of which will allow you to better control knock onset which will allow for more boost on pump

I was using the 2g curve as an example, I assumed everyone already knew the 1g curve ramped up faster. I could have been more clear I guess.

1Gs have the most ridiculous stock timing curve ever. I always use the 2g curve as an example of how to do it. I even run the 2g curve on my EVO. I also wouldn't say that the 1g curve is more aggressive, since it peaks around the same point as the 2g map, I'd say it's more retarded. And I don't mean retarded as in less advance, I mean retarded as in special needs. They were probably just trying to kill power at high rpm, much like how the EVO maps drop the timing to 9 degrees peak as soon as you get over what the stock turbo can do, to help control thier warranty liability.

andymoraitis said:

That's the right answer. I've seen 1G's peak at close to 26-27 degrees of timing if the airflow subtraction was aggressive enough on a 1G map. I'm sure in the "olden" days when tuners were correcting ND660's on an AFC or MAFT that numbers were even higher.

Click to expand...

With 550s, 5 knob AFC, and hacked MAF I've seen 28 degrees. Just one of the reasons I ran so well on the 14b with race gas. I've seen guys with 660s and AFCs on stock 1G ECUs hit 30+ degrees, you can imagine how well they ran on C16 (and how much knock they saw on pump). It truly is a whole new era with DSMlink, AEM....

Interesting. I always found that trying to run timing that high actually reduced my mph, even when it didn't knock. Funny how people get away with different amounts of timing on different cars. Before DSMlink came out I used to retard timing with the 1g CAS, but then idle and cruise suffer slightly. Beats blwoing shit up though. Man, that brings back memories...

You're definitely right on with that statement, Kevin. While race gas may allow you to get away with some ridiculous timing, more doesn't necessarily mean better. The goal is to obtain peak cylinder pressure precisely at the moment in crankshaft rotation where there exists the greatest mechanical advantage during expansion, which works out to ~14 degrees ATDC. If you time ignition too early, you can miss this mark just as easily as with timing that is too retarded, except now you add the increased risk of detonation.

Exactly. I found that on my stock 2.0, I saw the best performance from 23 degrees of advance on high octane. Any more than that and the trap speeds slowed down. Interestingly enough, being on the low side (around 19-20) wasn't as detrimental as being too high.

The stroker seems to be another story entirely however. It loves 12 and we're about to try 8 and see if we can squeeze 28 psi out of pump gas with a large sidemount since the 28/12 combo destroys the clutch.

Another interesting point about that calculator is the water injection section. It gives you the possibility to choose what order you inject water, turbo, Ic. One of the options is pre-turbo. That's not something you ever see on a DSM. I've seen some people swear by it, with no noticeable compressor wear.

Regarding that setup, I can speak from experience with steam turbines, which I believe would translate appropriately to a turbocharger, but I'm not 100% on it.

These turbines were driven by dry, saturated steam, meaning that any moisture from the boiler had to be separated out before entering the turbine. If boiler water level rose too high, moisture carryover into the turbines would occur, which was known to cause impingement damage to the turbine blades. Granted, a turbocharger compressor is not nearly as intricate in design as a steam turbine, but I believe the same damage would be possible if the volume of water injected was sufficiently large.

Anyone with experience running this type of setup who has evidence to the contrary can feel free to prove me wrong here, but I know I won't be running a pre-turbo injection system anytime soon.

thekellbeast said:

Another interesting point about that calculator is the water injection section. It gives you the possibility to choose what order you inject water, turbo, Ic. One of the options is pre-turbo. That's not something you ever see on a DSM. I've seen some people swear by it, with no noticeable compressor wear.

Click to expand...

running it pre turbo will mess it up pretty bad. There have been countless wrx owners in europe who tried this and had the turbine wheel all chewed up.

ok, obviously you guys missed my point. The 25 degrees was while tuning my 1000cc injectors with a eprom chip burned by my main man, dave. He kinda just guessed the first time around though, since going from 450 to 1000 is kinda a big jump. This resulted in me subtracting 30% airflow up top and getting a timing curve that looked like this:







edit: found some logs, heres the timing/rpm: (knock included for kicks and giggles)
27.00º,2468.75rpm,0.00 12:16:00
27.00º,2562.50rpm,1.00
25.00º,2687.50rpm,0.00
25.00º,2812.50rpm,0.00
22.00º,3187.50rpm,1.00
22.00º,3406.25rpm,0.00
24.00º,3593.75rpm,0.00
23.00º,3843.75rpm,0.00
24.00º,4031.25rpm,0.00
23.00º,4250.00rpm,0.00
23.00º,4375.00rpm,0.00
23.00º,4625.00rpm,0.00
22.00º,4750.00rpm,1.00
21.00º,4937.50rpm,0.00
21.00º,5062.50rpm,0.00
21.00º,5250.00rpm,4.00
18.00º,5375.00rpm,7.00
18.00º,5656.25rpm,7.00
18.00º,5718.75rpm,10.00
17.00º,5875.00rpm,11.00
18.00º,6125.00rpm,10.00
18.00º,6187.50rpm,17.00
18.00º,6375.00rpm,16.00
19.00º,6562.50rpm,16.00
19.00º,6625.00rpm,15.00 12:16:09

yeah, apparently my memory sucks, but I wasn't too off.
now that the car is at a rock steady 10.5:1 afr and the maft is at 0% correction, my timing map is extremely week, but that might be due to my ecu seeing over 2600 hz. I really want dsmlink or something so I can crank up the timing.

as for that "tune out the spike at 6k rpms",... some of us just see that as a breif increase of testasterone, and we would like to turn that spike, into a consistant upward curve starting at 6k and going up through redline.

I loved the above timing curve, because I thought if I could figure out why I knocked so much at higher rpms, and fix it, the car would go crazy fast. The current timing curve might barely see 18 degrees if Im lucky.