Tuning for Fuel Economy

[utc_pyro]

Recently have been thinking if going to a bigger turbo actually hurt my highway gas mileage (off boost of course). The possible reason is, smaller turbo spools easier so starts building boost sooner, helping overcome pumping loss inducing manifold vacuum.

 

Most of your pumping losses are across the throttle body, and you use the to control power output. If the turbo spools up this will INCREASE the required pressure differential across the throttle body to maintain the same torque output. You'll also expend more energy by increasing exhaust back pressure (spinning the turbo) and all the inefficiencies extracting that power and putting it into the intake air.

 

You'd actually have better fuel economy by removing the turbo completely, or at least bypassing it. Rigging up a way to force the wastegate and BOV full open at cruse would allow you to effectively do this. You'd just need a vacuum pump, a two port WGA and a solenoid. Apply vaccume to the lower port and the BOV and force them open, then you'll have less total plumbling losses.

[Flinkly]

Most of your pumping losses are across the throttle body, and you use the to control power output. If the turbo spools up this will INCREASE the required pressure differential across the throttle body to maintain the same torque output. You'll also expend more energy by increasing exhaust back pressure (spinning the turbo) and all the inefficiencies extracting that power and putting it into the intake air.

 

You'd actually have better fuel economy by removing the turbo completely, or at least bypassing it. Rigging up a way to force the wastegate and BOV full open at cruse would allow you to effectively do this. You'd just need a vacuum pump, a two port WGA and a solenoid. Apply vaccume to the lower port and the BOV and force them open, then you'll have less total plumbling losses.

interesting idea...

[covertrussian]

your first point is more of a thought experiment. could come up with hypothesis both for and against. most are probably right (to an extent), but with your current system it appears that the "against" column carries more weight.

 

and the second point is "fact", at least for the Subaru OEM SS vs. TS systems, although the amount of savings to be had is still under debate.

 

First point is really what I'm trying to get to, I brought up the TS because it was the easiest find related to turbo + pumping losses (google is polluted with too many unrelated links).

 

TS is known to spool faster, so it's like having a smaller turbo, but having separated pulses does have it's own benefits (though as I found out by going ELH, it didn't really matter at least on my setup).

 

 

Throwing another wrench into this, I just attached a backpressure gauge to my manifold's O2 sensor bung. At 3,000 RPM and ~12in/hg (my typical highway RPM and load), I saw 3psi of backpressure in the manifold just by maintaining speed.

 

Most of your pumping losses are across the throttle body, and you use the to control power output. If the turbo spools up this will INCREASE the required pressure differential across the throttle body to maintain the same torque output. You'll also expend more energy by increasing exhaust back pressure (spinning the turbo) and all the inefficiencies extracting that power and putting it into the intake air.

 

Right most of the pumping losses come from the throttle body it self, but turbo helps shove air past even the smallest throttle body openings, reducing manifold vacuum, which in theory should reduce that pumping loss.

 

Now on the other side of the coin, say your throttle is open 25% going 70mph. A smaller turbo could be spooling enough to keep in in positive boost area at slight touch of the throttle, this would lead to closing the throttle more, and increase in pumping losses.

 

Same scenario but with bigger turbo, 25% throttle wont keep the turbo spooling enough to keep you going 70mph, so you might actually need to use more throttle (reducing pumping losses).

 

Purposefully reducing torque is something I'm actually gonna test after testing a catback. Idea is, by reducing torque by reducing timing, might force a bigger throttle opening, which leads to reduced pumping losses.

 

 

You'd actually have better fuel economy by removing the turbo completely, or at least bypassing it. Rigging up a way to force the wastegate and BOV full open at cruse would allow you to effectively do this. You'd just need a vacuum pump, a two port WGA and a solenoid. Apply vaccume to the lower port and the BOV and force them open, then you'll have less total plumbling losses.

 

My experience and logs disagree with this actually. My Infiniti is not a stock turbo car, which means it doesn't have low compression or other turbo related crutches. One week, NA, 16mpg city, next week slapped a T28, 18mpg city.

 

On a more related test to our cars, on my FXT and LGT I disconnected the wastegate, car barely built any boost by redline, and my gas mileage was no different then with it connected.

 

Turbos really do help efficiency, some say they atomize the air (I think what they mean by this is they tumble the air, so it mixes better with fuel). But I think turbos help the most by forcing the air into the manifold even at cruising speeds (turbos are still spinning), which works the same way as ram-air setups do.

[covertrussian]

Part of my, "make her breathe better" initiative, I switched to a 3" turbo-back exhaust.

 

The new Downpipe is very similar to the old one, just 3". With this downpipe and stock exhaust, I was getting the following backpressures at full throttle: 23psi preturbo, 7.5psi post turbo precat, 4.5psi post cat. Full throttle doesn't matter for MPG so I tested highway RPM's and engine loads: about 3psi pre-turbo at 2900rpm with ~10in/hg engine load.

http://legacygt.com/forums/attachment.php?attachmentid=253813&stc=1&d=1502161090

 

Next I took out the stock exhaust and replaced it with a custom built 3" single exit setup. This reduced the backpressure by 3.5psi, pre-turbo manifold PSI was 19.5ish, 4.5psi post turbo pre-cat, post cat 1.0psi. Highway EGBP was reduced to 2psi pre-turbo, at 2900rpm, so a 1/3 reduction.

http://legacygt.com/forums/attachment.php?attachmentid=254303&stc=1&d=1503120834

http://legacygt.com/forums/attachment.php?attachmentid=254301&stc=1&d=1503120834

 

 

With all that out of the way, now onto the tune and gas mileage.

 

Just to recap...:

I've been getting mid 29's at best for the last couple months, now that I think about it ever since I played around with different springs/struts (between wagon springs & spec B setup, but I wont blame them yet). Then when I installed my custom intake I got 30.97, but due to it not being rain friendly removed it, MPG went back to mid 29's.

 

Installed the 3" exhaust, got 29.72mpg with AVCS 5 and 40ign Hwy tune, basically no real gain . Though the car did feel less powerful overall until boost, this is a sign of the timing map being either under-advanced or over-advanced.

 

Next, I figured that maybe with a free flowing exhaust the engine wants AVCS 10 again. Usually I would run 38* of peak hwy timing with AVCS being at 10*, but with a new exhaust mod, it's possible that the engine needs less timing now. Thus for this run I used 36* of timing, this got me 29.50mpg, once again exhaust = no real gain. Interestingly enough, this tune felt much more powerful before boost, this map has a good bit less timing around 10in/hg to 0psi.

 

 

Since exhaust didn't help at all, I started looking for other items that would get me back up to those glorious mid 30 mpg's :

 

I've been hearing an odd clicking noise when starting and stopping the car, from the front left wheel. It was the wheel bearing, replaced it, replaced the engine oil, installed the stock 130k mile front struts/springs (have been running AT wagon struts/springs), flashed the AVCS 5* and 40* IGN map, and got 30.85mpg. Though I will admit, for 25 miles there was some heavy traffic which required pulse & glide instead of just cruise control.

 

 

TL;DR: Exhaust did nothing to help gas mileage, just like EQL header . I think others saw an MPG increase with exhaust or header because their tunes not optimal for their old setups. When they installed the header or exhaust, it made the tune be more tuned for their setup. Guess this is why I test all mods myself, and then post my findings here for others to hopefully use this information.

 

I'll try one more tune test on this setup, before I start fighting with the intake related mods.

Edited September 12, 2017 by covertrussian

[Flinkly]

going to stick with the custom 3"? i've always wanted to clip the stock exhaust before the Y and do a single 2.5" back. it's too bad the old sti muffler is too big, always thought about sourcing one of those to tuck up underneath. guess it'll be a ProXS if i do it.

 

So when are you going to test some custom bellypans? i assume you've got the stock front tray and the two underbody pieces.

 

or a lower grill block...

[covertrussian]

going to stick with the custom 3"? i've always wanted to clip the stock exhaust before the Y and do a single 2.5" back. it's too bad the old sti muffler is too big, always thought about sourcing one of those to tuck up underneath. guess it'll be a ProXS if i do it.

 

I wonder if a single 2.5" would flow better then the stock duals. Though with the STI muffler it definitely should, I think the biggest restriction is the stock LGT muffler design vs the resonator and y-pipe.

 

STI Muffler will fit, I'm running one right now, you just have to do some persuasion .

 

 

 

So when are you going to test some custom bellypans? i assume you've got the stock front tray and the two underbody pieces.

 

or a lower grill block...

 

Correct on having the stock bellypan and two underbodies, though one is a little torn from when I ran over the dear in 2014. I have two spare underbody panels from my parted out legacy though, just too lazy to install .

 

Since I now have a S204 lip, I might be removing the stock bellypan. While the belly pan helps cover up the turbulent header and suspension parts that stuck below the front bumper, it sticks quite a bit down, increasing the frontal area.

http://legacygt.com/forums/attachment.php?attachmentid=255214&stc=1&d=1505151387

 

S204 lip on the other hand covers up everything you need, without further increasing the frontal area.

http://legacygt.com/forums/attachment.php?attachmentid=255213&stc=1&d=1505151387

 

 

Also never on grill blocks and other ugly modifications, my grill block will be an FMIC

[covertrussian]

While I've been quiet here, I've actually been busy working on a couple theories and brainstorming on how to utilize these theories on our cars. Once again I'm avoiding unrealistic driving styles (below speed limit, pulse & glide, drafting, etc.) and hideous aero modifications. This is another thought experiment & brain dump post, feedback is greatly appreciated.

 

All of tech papers that I read have one thing in common, reducing pumping losses is the main goal, ie keeping the throttle open more which reduces manifold vacuum. There are a few mainstream ways of accomplishing this with new engines: Atkinson Cycle, cylinder deactivation, turbo charging and downsizing, variable valve lift/timing technologies, but most of those are not applicable or retrofit-able into our existing motors. Hence this post will have the theories that apply to our motors, that I've been thinking on and trying to test

 

 

 

Reducing Throttling Loses

 

Throttling loses are Pumping Loses that happen from having a nearly closed throttle body, which results in higher manifold vacuum that pistons have to overcome. This is a big part of why diesels are much more efficient then gas engines, they don't have throttle bodies!

 

If you have a boost/vacuum gauge you know that when you reduce manifold vacuum, via throttle increase, the car tries to accelerate past the desired speed. This is the crux of the matter, we have to find a way for the car to not accelerate as we open the throttle more, sounds counter intuitive right?

 

I couldn't find any concrete vacuum figures that we should try to stick to, the more we increase throttle the more fuel we use, thus you have to find a happy medium. I don't think it would be efficient to drive around everywhere with manifold vacuum being at 0psi all of the time. My goal right now is to have the manifold vacuum be at ~10 in/hg during flat road cruising, currently it's at 15-20 in/hg.

 

 

 

How can we Realistically Accomplish this?

The end goal here is to make our cars use more throttle, to maintain the same speed, but not be so torqueless that we have to downshift or be in boost all of the time.

 

Wastegate Duty Cycles - This one is first, because it's the easiest thing to modify without needing to know how to tune timing or AVCS. Stock tune asks for the same wastegate duty cycles from closed throttle to 100% open throttle. This means it's keeping the waste stays closed longer, this reduces turbo lag and of course that means your in boost much easier.

 

Zeroing out wastegate tables until 90% throttle will help us make the turbo be more laggy, making the stock turbo act like a bigger turbo. Which is good since is our goal is to reduce the amount of time we spend in boost and we want to use more throttle without going into boost.

 

Variable Valve Timing - AVCS - Active Valve Control System allows us to modify the mechanical timing of the camshafts, by doing so we are changing dynamic compression ratio, cam overlap, and EGR.

 

AVCS also controls how late or early our intake valves open and close.

 

Higher AVCS opens the intake valves earlier, increasing overlap and EGR effect, it also closes the intake valves sooner, which increases dynamic compression ratio because it bleeds less of the mixture back through the intake valves during the compression stroke.

 

This is where you can't have your cake and eat it, with increased EGR you need more ignition timing due to poor combustion tendencies recirculated exhaust gasses, but increased compression from higher AVCS plus increased ignition timing will lead to detonation. Another drawback is, higher overlap numbers increase the odds of fuel going straight out the exhaust especially with free flowing exhausts.

 

Lower AVCS opens the intake valves later, reducing overlap & EGR and closes valves later in the compression stroke. This reduces the dynamic compression ratio since more of the mixture is bleed back into the intake manifold, which has the advantage of further mixing the air and fuel.

 

This is the route that Subaru took with our 05-06 cars, they run 0* AVCS at highway speeds. In theory this should make less power due to lower compression ratios, but in practice I've found the car to be much more darty (too much torque) at highway speeds. Every time I've changed AVCS away from 0* to 5* or 10*, I've gained gas mileage.

 

EGR - Increasing EGR fills the cylinder dead space that would otherwise be filled with Air and Fuel, this reduces torque output and requires more throttle to maintain the same speed. Since our EGR is part of the AVCS system, we must deal with all of the early valve closing drawbacks if we want more EGR.

 

Gearing & Transmissions - Going from dual mass flywheel to a single mass flywheel on my 5 speed cost me 2mpg in the city. With the heavier dual mass, I could be in 4th gear at 30mph without lugging the motor, but with a lighter single mass flywheel, I now have to be in 3rd which has way too much power (you know when the car jerks as you let go of throttle). This of course means I'm in higher manifold vacuums more now.

 

Another one to consider is a 6 speed transmission. These usually have two benefits, they reduce the frictional losses by keeping the RPM's down, and once again they reduce available torque requiring more throttle to be used. Around town it should also help, since most 6 speeds have closer ratio gears you can go to a higher gear without lugging the motor. Main downside to STI/Spec B 6 speeds is added weight, which around town will hurt the economy slightly.

 

Perhaps the easiest one is Taller Tires, which Increase gearing which reducing the RPM. Downside of this is, the more tire weight does shift to the outside of the tire, which increases the power required to rotate them, which is bad.

 

Performance Mods - Power adder mods like intakes and exhausts all increase torque, thus in theory they should hurt gas mileage, but since these mods are geared for higher RPM gains, they usually result in torque loss while cruising, which in turn forces us to open the throttle more. Another benefit a power adders can do is, it can reduce engine lugging in higher gears, which allows you to stay in higher gear without needing to downshift.

 

Ignition Timing - Advancing timing makes torque, to a point (MBT = Minimum advance for Best Torque), after that point it reduces torque. Timing Retard reduces torque, unless your past MBT (which I've found myself to be every time I've done an intake/exhaust efficiency mod). In theory we should be able to reduce or increase torque, which would control how much throttle is needed.

 

Downsides to over Retarded timing, it will increase EGT's, because it increases the amount of air and fuel escaping unburnt into the exhaust. Another downside is the function of our imperfect air to fuel reading devices, aka the Oxygen Sensors (which only read left over oxygen content). With all that unconsumed air escaping, Oxygen Sensors will read lean, leaner then 14.7, which ECU will compensate for and enrich the mixture, which is not what we want.

 

Downsides to over Advanced timing, cylinder peak pressures increases, which increases pumping losses (covered in detail below), and detonation. It also will burn up more oxygen, which will make our oxygen sensors register a rich condition and ECU will lean out the mixture, jury is still out on if this is a bad thing or not.

 

 

Reducing Ignition Timing Pumping Losses

 

This one you don't hear much about, but it makes sense when you think about it: As you increase ignition timing, you increase peak cylinder pressure, which requires more power to overcome these pressures, this is what Ignition Timing based pumping losses are. To explain it in a more visual way, think of a turkey injector without the needle; you have the outside cover which is your cylinder, and inside plunger, which is analogies to a piston. First keep the needle end uncovered and press the plunger in, it should move in fairly easily, but if you cover the needle side with your finger, it will become much harder to push the plunger in. Running the proper amount of timing for the setup will basically delay when the needle end gets capped, thus allowing the plunger to move freely longer.

 

 

 

How can we Realistically Accomplish this?

The end goal here is, make your engine more efficient at ingesting, mixing, and burning the air and fuel mixtures, requiring you to use the less ignition advance. The less efficient your motor is, the more timing it needs to get a complete burn, but the more timing you run the more resistance the pistons have to overcome with every stroke.

 

Just about all of these will be mechanical modifications, with the exception of AVCS, which we can control within the tune.

 

AVCS - As mentioned in the throttle section, increasing AVCS increases the dynamic compression ratio and overlap/EGR, while decreasing AVCS reduces dynamic compression ratio and decreases/eliminates overlap and EGR.

 

Since we have a fairly low static compression motor, which requires more timing due to slower flame fronts. Increasing AVCS will help us increase the dynamic compression ratio and thus reduce the ignition advance needed to a full burn. But since increasing AVCS also increases overlap and EGR, we need to run more timing to to make up for the slower flame front. This is where you have to find the happy medium of AVCS vs EGR vs timing advance. This happy medium will also change as you do mods that increase efficiency of the motor.

 

I've found 10* of AVCS advance and 45* of peak cruising timing to be the best combination so far on my VF46, stock intake and cat-back (with better flowing aftermarket Uppipe and downpipe). 16G on the other hand seems to be happier with 5-10* AVCS and 38-40* of peak cruise timing.

 

Intake & Turbo Inlet - Just about all factory intakes have a vacuum that happens between the turbo (throttle body if NA) and the entry point. This is due to restrictive filters, sharp turns, accordion style pipes for flexibility, and finally undersized pipping. In my experience, a properly designed intake's MBT is usually a few degrees less then the stock intake, which is good.

 

TGV Delete - This one technically falls under the intake, in theory this will reduce the timing advance since it increases the CFM by quite a bit. But it's really hard to say if it will help during cruise conditions, since your airflow demand is very low. It might even make things worse due to reduction of tumbling airflow.

 

Exhaust - Factory exhausts are very restrictive, from very fine meshed catalytic converters to badly designed pipes and bends, this increases the back pressure on the turbo, and back pressure that the engine has to fight to exhaust it's gasses. In our case, doing a catless up pipe and an aftermarket downpipe greatly reduces the need for timing, my old FXT wanted 8* less timing at WOT then what the stock tune called for (to be fair, I didn't fine tune the stock tune first, but I can't imagine a factory turbo car being that over advanced).

 

Headers - To my surprise, equal length header didn't help me at all, this might be because I'm running fairly low AVCS advances (5-10*). I lost power with the equal length header, until I added 2* of timing and it got me back to where I was before, having to add timing is a step in the wrong direction. I think my header's tubing is simply oversized for my setup.

 

Bigger Turbo - Bigger turbos will not only keep you out of boost easier (which as mentioned earlier should be avoided), they will also deliver cooler and denser air to the engine, which will require less timing to reach MBT. When I went from the VF46 to the Big 16G, I was over advanced for the much more efficient setup, this resulted in gas mileage and power loss. After reducing the cruise timing by 8* and WOT timing by 5*, not only did I regain the gas mileage, I also gained about 20whp at similar boost levels.

 

Spark Plugs - Since our cars already come with fine tipped platinum or iridium plugs, this is not as important. But if your car runs copper plugs, switching to the fine tipped plugs will help with the burn efficiency. On my FWD SR20 I recently gained 1.5mpg city by simply simply switching out the copper plugs for Iridium, copper plugs had only 1,100 miles on them.

 

I personally love to side gap my spark plugs, a process where you cut the ground electrode shorter to better open the spark to the flame front. It worked wonders on my copper plug setup, but the jury is still out if it will work on fine tipped iridium/platinum plugs.

 

 

In conclusion... This is what I've been up to lately, brainstorming on how to reduce my cruising timing advance, this is why I started doing all of these airflow modifications too. I've also been trying to increase my low end torque (sub 2,000 rpm), so that I don't lug the motor and have to downshift as much, and trying to reduce my 2,000-3,000rpm torque so that I have to use more throttle.

 

I've also been trying to test these theories, but I only have so much free time to just waste 3 hours driving .

Edited September 13, 2017 by covertrussian

[Flinkly]

TMIC/FMIC is something else to think about. could mock up a TMIC delete tube and stay out of boost and test it's restriction/change.

 

also can pull in "hot" air (WAI). not that you don't know, but warmer air is less dense, so need more for the same amount of fuel, so throttle needs to be further open.

 

there are also crazier things, like switching to a STI intake for it's higher and more even flow. i assume this is out since it's ~$500 for used parts and quite a bit of work to swap.

 

 

So any plans on where to start? TGV deletes are easy. can just remove the butterfly and test, and if you really want to go for the gusto, can grind out the rib (anr/or "port" match).

 

I assume coatings are out (intake for flow, heads for heat), and head modifications.

 

and who said grill blocks had to be ugly? i'm planning to do a lower grill opening block that's honeycomb front, solid back.

Edited September 13, 2017 by Flinkly

[covertrussian]

TMIC/FMIC is something else to think about. could mock up a TMIC delete tube and stay out of boost and test it's restriction/change.

 

also can pull in "hot" air (WAI). not that you don't know, but warmer air is less dense, so need more for the same amount of fuel, so throttle needs to be further open.

 

Interesting thing about WAI's, I've tried removing the snorkel, which would leave the airbox to suck in hotter engine bay air, didn't help at all. I wonder if WAI concept no longer applies to newer ECU's.

 

I think FMIC will function similarly to a CAI too, so that should in theory make things worse (due to more dense air). But in practice, might be similar to my experience with CAI's, where it actually helps MPG a good bit. Also my FMIC G20 car gets excellent gas mileage.

 

I'm leaving FMIC testing for later, I have a piping kit for the legacy, I just don't want to cut up the car that much yet .

 

there are also crazier things, like switching to a STI intake for it's higher and more even flow. i assume this is out since it's ~$500 for used parts and quite a bit of work to swap.

 

I would be interested in trying a better flowing intake manifold actually, especially if I go FMIC and rotate it to face forward, but that's one of those that will never pay for itself. Though our manifolds are already pretty well designed, all cylinders get equal amount of air, much better then my SR20 manifold, where the inlet is right by cylinder 4.

 

So any plans on where to start? TGV deletes are easy. can just remove the butterfly and test, and if you really want to go for the gusto, can grind out the rib (anr/or "port" match).

 

I already built TGV deletes, just need to install them now. I'm wanting to do a little more exhaust testing (proof of concept stuff), a little more CAI testing (need to build a CAI that doesn't get wet, like my current prototype does, and airbox!)

 

I assume coatings are out (intake for flow, heads for heat), and head modifications.

 

I'm not sure coatings really do all that much. I'm well versed in house insulation, there are companies that sell "ceramic" coatings to better insulate the house, they are all snake oil. Maybe with hotter engine bay things it's different, but R value is R value (and ceramic has none )

 

and who said grill blocks had to be ugly? i'm planning to do a lower grill opening block that's honeycomb front, solid back.

 

I would love to see what you come up with! I wonder how much of a block it needs to be to actually benefit you. Can it be like 1" recessed?

[Flinkly]

Interesting thing about WAI's, I've tried removing the snorkel, which would leave the airbox to suck in hotter engine bay air, didn't help at all. I wonder if WAI concept no longer applies to newer ECU's.

 

The 2014 Malibu had an extreme WAI system, where it literally exchanged heat with the post-cat exhaust. would be awesome to have a system that could swap between that and "regular" air vs. boost pressure (mechanical vacuum/pressure actuator).

 

but i do remember your work in regards to the earlier tests. and would be quite the system to do an extreme WAI.

 

 

 

I would love to see what you come up with! I wonder how much of a block it needs to be to actually benefit you. Can it be like 1" recessed?

 

my light thinking is that the majority of the savings from a grill block would come regardless of recess. sure, the deeper it is the more turbulent you make air near it, but it's crazy turbulent already regardless, and the more you get to go around rather than through, the better.

 

i figured i'd just cover up at the inside lip of the opening, which seems like 6" deep. the recess will "fill" with air and the rest will go around.

[covertrussian]

That's pretty cool, do you have any more info on the malibu system? I couldn't find much of anything for it.

 

About the grill block, that makes sense, to me it felt like if it's not perfectly flush it's pointless (because it wouldn't push the forward hitting air left and right).

[utc_pyro]

TMIC/FMIC is something else to think about. could mock up a TMIC delete tube and stay out of boost and test it's restriction/change.

 

also can pull in "hot" air (WAI). not that you don't know, but warmer air is less dense, so need more for the same amount of fuel, so throttle needs to be further open.

 

Perrin actually did this:

http://blog.perrinperformance.com/your-sti-doesnt-need-an-intercooler-right/

 

also can pull in "hot" air (WAI). not that you don't know, but warmer air is less dense, so need more for the same amount of fuel, so throttle needs to be further open.

 

I'm working on duct feeding a Grimmspeed cold air intake to solve some urbin driving issues (duct sucks in air from upper grill, like the later Subaru WRC cars). It'd be easy to rig up a duct/fan to suck in air from the headers at highway speeds. Not sure if the plumbing loss gain would outweigh the alternator drag though.

 

Actually with the above considerations, you could do this with a air-water intercooler and a valve. Feed it engine coolant at low load and highway speeds, and switch to a low temperature circuit when you want more power.

Edited September 18, 2017 by utc_pyro

[covertrussian]

Perrin actually did this:

http://blog.perrinperformance.com/your-sti-doesnt-need-an-intercooler-right/

 

That's a real cool find! No wonder 80's turbos got such a bad rep, lots of them were not intercooled. This does make me wonder if they tested with the hood closed with TMIC.

 

Another interesting thing is, where we get wastegate feed from, ie straight from turbo. In SR20 world we know that if you get wastegate feed from turbo your boost will taper towards the redline, for that reason we use a feed from somewhere very close to the throttle plate. This would spin the turbo more to compensate for the IC based pressure loss. But, we also don't try to push 18+ psi from a tiny turbo (cough vf40), we usually do 15psi tops on T25/T28 setups.

 

I'm thinking of moving my Legacy's wastegate feed to be post intercooler for this reason, especially if I go fmic.

Edited September 18, 2017 by covertrussian

[utc_pyro]

Another interesting thing is, where we get wastegate feed from, ie straight from turbo. In SR20 world we know that if you get wastegate feed from turbo your boost will taper towards the redline, for that reason we use a feed from somewhere very close to the throttle plate. This would spin the turbo more to compensate for the IC based pressure loss. But, we also don't try to push 18+ psi from a tiny turbo (cough vf40), we usually do 15psi tops on T25/T28 setups.

 

I'm thinking of moving my Legacy's wastegate feed to be post intercooler for this reason, especially if I go fmic.

 

Interesting point... That probably contributes to some of the boost taper when using a MBC on these cars as well. Though you should be able to compensate with WGDC at higher RPM if you have boost control enabled.

[covertrussian]

Since my 16G came with a 13psi gate I never bothered hooking up the stock EBC, and part of me is glad, less things to worry about retuning every time I did a major mod (like inmtake or exhaust).

 

I'm running an MBC (simple ball and spring design) on my SR20, it holds 9-10psi to redline, but it does bounce up an down through the RPM range. But no downslope like we see on our Subaru's.

[NSFW]

How much intake/exhaust overlap is there if you fully advance the intake cams? I'm guessing that would give you maximum EGR and thus minimum pumping losses, since the cylinders would be filling with more exhaust and less intake air... No idea what that would do to emissions, or how well the engine would run, but it might be interesting to experiment with it.

 

I'm pretty sure that WGDC is a non-issue here, since it only really makes a difference when the boost level starts to exceed whatever the wastegate spring provides. With any less boost, the WG spring is holding the WG shut, so it doesn't matter what the solenoid is doing.

 

If you want to make it harder to get into boost, a weaker wastegate spring would sorta help, but really I think the best way to minimize boost is to modify the nut behind the wheel. Just use less throttle.

[covertrussian]

How much intake/exhaust overlap is there if you fully advance the intake cams? I'm guessing that would give you maximum EGR and thus minimum pumping losses, since the cylinders would be filling with more exhaust and less intake air... No idea what that would do to emissions, or how well the engine would run, but it might be interesting to experiment with it.

 

I believe the our max AVCS is 40* since 07's run it:

http://legacygt.com/forums/attachment.php?attachmentid=239740&stc=1&d=1477486376

 

This is where you can't have your cake and eat it, sure your getting a lot of EGR, which requires lots of timing, but your compression ratio is also going up, which requires less timing because this much overlap starts detonating a lot. On top of that you start blowing your fresh air and fuel out into the exhaust manifold.

 

I think the latter starts being the biggest issue once you start opening up the exhaust. My car didn't mind 20* of AVCS when I had a stock downpipe, when I tried to run 20* on stage 2 downpipe, I even reduced timing to be 35* (from 40*) and it still got a pretty a pretty bad 24mpg, where AVCS 10* with 40* got 28.40 just a week later.

 

I'm not done playing with higher overlaps yet though. Back when I tested this stuff originally I didn't have as a holistic understanding of it all. So this time around I might have more success.

 

I'm pretty sure that WGDC is a non-issue here, since it only really makes a difference when the boost level starts to exceed whatever the wastegate spring provides. With any less boost, the WG spring is holding the WG shut, so it doesn't matter what the solenoid is doing.

 

On most internal gates (especially oem), once boost starts expanding the diaphragm, it will start opening the gate, it's not a on/off switch like fancier gates can be (Synapse comes to mind).

 

MBC or EBC can keep the gate shut longer because the gate's diaphragm wont see any pressure until they overcome the internal springs of those setups (for MBC), or duty cycles for EBC.

 

This is why zeroing out the WGDC table up to a certain throttle position works for us, it really reduced the throw head into seat on my car when simply using cruise control. This was really annoying to passengers too, so win win.

 

 

If you want to make it harder to get into boost, a weaker wastegate spring would sorta help, but really I think the best way to minimize boost is to modify the nut behind the wheel. Just use less throttle.

 

This is only true in the city and exactly why I started this thread too though. While I agree, you as a driver can greatly effect gas mileage, this doesn't really help with making the engine more mechanically efficient. On my highway gas mileage tests, I remove as much of the driver and driver techniques from the equation. I use cruise control, I avoid accelerating into boost, drafting, pulse & glide, and all other driver induced variables.

 

With that in mind, my gas mileage was around 25mpg when I first got the car, with occasional 27-28's. Right now the car is around 29mpg with occasional 31mpg. My goal is higher then that, I wont name any numbers not to get any hopes up, but I will say my FWD Infiniti G20's goal was ~300hp and 35mpg. Yesterday that G20 got 39mpg on sticky summer race tires, driving back from the race track, oh and she's putting down around 290bhp (250whp).

 

I'm gonna post about it soon, because two weeks ago G20 got 32mpg.

[xt2005bonbon]

I have not read all the past few posts you guys made. But I've been doing a few trips to Ocean city in MD from the DC area lately. Mostly flat roads with almost no noticeable tail/head winds. With a family of four (two little kids) and a trunk full of stuff, I was pleasantly surprised to hit 26mpg with a mostly 80mph average speed!! I can tell when I cruise around 80, you just need to blip a tidy bit the throttle and it quickly accelerates and reaches target cruise speed. Sweet spot.

 

Also noticed that when I went 50 mph for quite a long stretch of road, I was hitting like close to 32 mpg!!!

 

stage 2, catless up pipe, hiflo catted dp, 3 inch turbo back. avo tmic, avo wastegate actuator, grimspeed ebcs, stage 1 exedy clutch with lwfw and lwcp. and cobb cold air intake.

Edited September 18, 2017 by xt2005bonbon

[covertrussian]

No worries, I post a lot, with lots of text, this should hopefully make it easier for people to find via google. When I would google for this information, I would get basic non technical threads, that's what sparked it.

 

50mph is the sweet spot on most cars, low wind resistance, and good distance coverage, going below 40mph would get you worse MPG most likely.

 

At 80mph, your probably at 3300rpm, so that makes sense, these motors have a lot of power at that range.

[xt2005bonbon]

yep. 3.3k rpm around there. The turbo likes that spot and I do too

[covertrussian]

While my Infiniti G20 is now my dedicated race car, I don't really care about fuel economy nearly as much, but since I have a fully adjustable ECU and live tuning, I'm starting to test proof of concepts on this car before trying to apply them to the Legacy or Outback. The car also has all the all the mods that I would do on the legacy (3" Intake, FMIC, 3" DP, 3" exhaust, 6 speed etc.).

 

I mostly do city tuning on this car, my LGT is much more comfortable for highway driving, but recently I've done some highway testing and it's been an eye opening experience.

 

Spark Plugs

In May car needed new spark plugs badly, I had some new BKR6E's laying around. I gapped them to 0.035", which is on a higher end for 10psi, but I didn't get any spark blowout. Did some city testing, got 24.18mpg, not bad was seemed down from last year. Keep in mind my LGT is getting around 19mpg with same conditions.

 

After the first hot day autocross, I saw my post intercooler IAT's being at 130*F, coupled with higher compression (9.5:1), I got some colder plugs, BKRE7IEX and I also gapped them to 0.035". Since my city driving is so short and hard on the car, I was afraid that colder plugs would foul more and do worse, Instead I was greeted with 25.76mpg city.

 

Lesson here is, finer tipped spark plugs are much more efficient, they reduce what's called spark quenching, which is heat loss into the big center electrode. Our EJ255/7's come with fine tipped plugs already, but if you have an older Subaru get those finer tipped Iridiums in!

http://legacygt.com/forums/attachment.php?attachmentid=255522&stc=1&d=1505827044

 

Exhaust

Exhaust Mods overview: The car has a very short runner 4-1 exhaust manifold, to a GT28R (with a GT2860r "Disco Potato" compressor housing, because 3"). This goes to a ported stock exit pipe (j-pipe), then 3" stainless steel down pipe, 3" 400 Cel Ceramic Thunderbolt cat, 5" can Borla resonator, 3" stainless steel tubing to a 3" bored out 07 STI Muffler. This setup has about 3.5psi of backpressure pre-cat and 2.5psi of backpressure without the STI muffler. This setup was putting down about 235-240whp.

 

I thought I would be clever and show off the Legacy with G20's great MPG's. Took it on a highway trip, got 32.02mpg, last year got 35.94mpg with similar tune and setup, (just side gapped copper spark plugs). At least I got some data logs to compare to the Legacy . This was with 225/45/17 cooper RS3-S summer tires.

 

Bummed at this I parked the car, the exhaust was hitting the body anyway so I needed to reweld it and I needed the 5" resonator to fit onto the Legacy. I cut up that setup, saw that the ceramic cat was shot (cracked core), thus rebuilt the cat-back portion. New setup has 6" dynomax resonator, 3" 300 cel metallic Thunderbolt cat, same 3" tubing, and loudened STI muffler (same as I did with legacy). Pre-cat this setup registered 2psi worth of backpressure by redline.

 

First thing I noticed is no power increase instead had a nasty spark plug breakup (car missing at higher RPM's). This was odd, I only reduced the backpressure by 1.5psi... I gapped the spark plugs down to 0.030", went for a test drive, and the car was very torque steer happy in 2nd gear (that's new), checked the power and she's putting down 250whp! Now why couldn't my Legacy see such gains from shedding 3psi of back pressure. I think the exhaust made a bigger difference on G20 because I have fixed overlap cams with ~6* of overlap, while legacy has AVCS which lets you tune out the badness of restrictive exhausts.

 

Highway Testing the New Setup

I didn't have time to do a proper highway A-B-A test and tune before the long trip to the race track, so left the tune be. I started driving to the race track, this leg of the trip is mostly downhill. On the way there I got 27.17mpg, which was not too surprising since my other cars get worse MPG on this leg of the trip, plus the new 235/45/17 Yakohoma AD08r's are stickier, poke out more (bad for aero), and it was night driving (which hurts the MPG on all my cars). Biggest thing that I noticed was, car felt too powerful & darty, was staying at a constant vacuum range (constant throttle) instead of varying the throttle a lot.

 

I figured maybe with the new efficient exhaust I needed less timing, I pulled 2* of timing, setting me down to 21* peak timing, previous years gas mileage favored 23*. This is basically half of what the Legacy is running... Highway driving around at the destination, I noticed the gas gauge wasn't moving down as quickly, and this is with me going ~75mph instead of 70mph like the night before.

http://legacygt.com/forums/attachment.php?attachmentid=255524&stc=1&d=1505827044

 

Another thing I noticed was my ECU wasn't as good at keeping the car at 14.7, it was ranging from 14.0-14.8 for the most part. I set the target AFR to be 15.0, this made the car be much better at keeping it at 14.5-15.3 AFR.

http://legacygt.com/forums/attachment.php?attachmentid=255525&stc=1&d=1505827044

 

After racing most of the day (about 60min track time at full throttle), 250 miles at 75mph, I still got 21.93mpg, which is pretty good if you ask me!

 

On the way back I stopped at the same Shell gas station as before, this time to get gas for the return trip home. Usually I see better fuel economy on all my cars on this leg of the trip, it might because it's during the day, or because it's driving back up the mountain from here. I drove 140miles and filled up 3.55 gallons, this translates to 39.58mpg!!!. The car was once again constantly changing throttle positions (can be seeing on the boost/vacuum gauge), basically it would open throttle enough to go to 0-5 in/hg then close it to 10-15 in/hg, instead of sticking to one constant vacuum range. This makes the car basically pulse & glide on it's own with cruise control.

 

Here is the cell count for 32mpg vs 39mpg run, while 32mpg one was an A-B-A trip and 39mpg was an A-B, the log was taken from same mile markers on the same highway. This shows how many times the engine was sitting in each vacuum range. If you look closely enough you can see that the lower graph (39mpg one) spent more time at lower vacuum levels overall, which should indeed lower the pumping losses.

http://legacygt.com/forums/attachment.php?attachmentid=255523&stc=1&d=1505827044

 

 

How can I apply this information to the Legacy?

Even though I cut out 3psi of back pressure, the Legacy still is happy with 0.035" gap, which means my new exhaust didn't increase the efficiency enough yet. My catted downpipe is still providing the bulk of the back pressure, I need to test out my other down pipe, and maybe a catless one to see if I can pin point the issue.

 

Once the Legacy starts missfiring/breaking up at 0.035" gap, I'll know that I'm flowing much better, and maybe then the new exhaust will see some good MPG gains.

 

TL;DR: Cooper to Iridium increased my gas mileage by 1.58mpg. Reducing the exhaust back pressure on my G20 required a smaller spark plug gap and less timing, which resulted in a major MPG bump.

Edited September 19, 2017 by covertrussian
Rewording

[Flinkly]

That's pretty cool, do you have any more info on the malibu system? I couldn't find much of anything for it.

 

man, i was so wrong on this one. it does take heat from the exhaust, but it uses it to heat the coolant during warmup for shorter warmup time (and to heat passengers faster, lol). it also is the 2016 Malibu.

 

http://media.chevrolet.com/media/us/en/chevrolet/news.detail.html/content/Pages/news/us/en/2015/mar/0325-malibu.html

[covertrussian]

That's still pretty cool! Such tech would help me a lot with my 4 mile trips.

[Flinkly]

That's still pretty cool! Such tech would help me a lot with my 4 mile trips.

 

oh yeah, still super cool. a different solution to the same problem as the Prius thermos (it "locks" down hot coolant on shutdown for use on next startup).

[covertrussian]

oh yeah, still super cool. a different solution to the same problem as the Prius thermos (it "locks" down hot coolant on shutdown for use on next startup).

 

Though I will say a traditional thermostat does a good job too. My Legacy has the constant running fans, but only when it was below 40F, I ran it like that for about a year until ECU finally burned enough for dealer to cover under recall.

 

What really surprised me was the lack of gas mileage increase going from fans always running, to fans never running in the winter. I think it's because the radiator side of thermostat would stay closed and it didn't matter until the car was fully warmed up.