Required Upgrades for E-85?

[WildHorse]

Weird, I get close to 200 on E85 and I'm boosted, assuming you are too. I run can bus flex, so I can do whatever but normally I stick to e85 because I go to the track fairly often.

he's NA.

Sponsored

 

[GL95]

he's NA.

How could gas mileage possibly be that bad then, even on e85 that's 10 mpg

 

[K4fxd]

I have a heavy foot.

 

[K4fxd]

Also lots of twisty roads with large elevation changes.If I'm on the freeway I get much better mileage.

 

[engineermike]

Something else worth mentioning is my truck is my Roush f-150. I’ve converted it to flex using only an 18v bap and tuning. This truck runs the oem f-150 pump (at 18 volts) and Roush supplied 47# injectors. I’ve even raised the boost 2 psi to 13-14. The fuel pressure holds flat and injectors are only at 50% duty cycle (75% of what Bosch allows).

And the thing starts beautifully in cold weather. Some commercial tuner’s rep once berated customers for wanting e85 tunes that start in cold weather. Ford actually gives us all the data we need to make it work but hpt doesn’t define the parameters to allow you to use it.

 

[Angrey]

My N/A stock fuel system car on E85 gets about 150 miles to a tank. About 125 before I start getting panicky about finding an E85 pump. A hell of a lot less if I'm playing. I do run a flex tune to mitigate some of the stress, but I hate pumping 93 when I want E85.

I hate to imagine what double the power would do to the range of the car.

Not as much as you would think for "cruising" but yeah, totally crushes economy for anything other than long cruises.

For long cruises, it also depends wildly on how high you run the RPM (and ultimately the air load). If you stay out of it and do 70 with decent gearing it gets nearly the same range as when I was N/A E85. But yeah, if you wanna run 90mph the whole way, it gets chopped hard.

The longest crossing for me is 120 miles and like you, I'm getting pretty nervous by the end of it but when I fill up, I realize I wasn't all that close.

It's nice having the backup option to slap some 93 in it if I get into a real jam.

I suppose I could run straight 93 and add back 30% more range, but it would take a LOT of discipline to stay out of it and while it technically should be fine, MOTEC advised not to dog it on 93. Just use it for a pinch.

 

[engineermike]

I'm not a tuner, but I've been building software for 24 years (starting on Irix running on SGI boxes, plain ol' Unix, and then on to Linux, then to the containers in the cloud) and I've done my share of math. The reason many tuners don't do more is because the devil is in the details and that last 5 to 10% requires more time to execute than the 90% (or so they thought it was the ninety percent). I have a saying about building software and outsiders looking in. Everything is easy when you aren't the one doing it (if I had a nickel for every dummy who said that, particularly when it involved a path/tree traversal problem). If it was easy and lucrative there would be many many more "tuners" giving it go. The guys on forums, who definitely know their crap, don't hop into it because they aren't bold enough to exit and start their own company. True knowledge can breed fear. There is a lot of what ifs. How to deal with customers? How to deal with all the possible scenarios? Growth? Failure? Etc. So what we are left with in many cases are the dumb a$$es who really weren't all that knowledgeable, but had the balls to take the leap.

Another comment on this...the Ford OEM calibration is the result of a few decades and literally hundreds of engineers on teams piecing it all together, with each engineer or team specializing in just one thing. I can imagine they have at least one turbomachinery engineer, control systems, software, mechanical, calibration, combustion modelling, the list goes on and on. Then you have aftermarket typically non-engineers working alone trying to figure out how to adapt all that to some aftermarket parts...blindfolded with one hand tied behind their backs. I say that because hptuners only lets you see probably 10-20% of the parameters, while PCMTec lets you see probably 95% of them. But the other half of the OS/strategy is the logic and coding and neither PCMTec nor HPTuners lets you see that, not that most could make any sense of it even if you could see it. There are literally millions of lines of code and here we are playing guess-n-check trying to figure out what the logic does by changing the constants and seeing what happens. The only reason I've been able to get as far as I did is my engineering background, my friend being a control systems and IT engineer, and one OEM-level calibration specialist I met by chance.

Even still, some of the stuff in commercial aftermarket tuning isn't that hard to get right but it's still generally not good.

 

[lo-fi]

Someone on the forum once said that the people who have the ability to be really good at it can make way more money with much less risk and headache in other industries.

True. I can't argue with that logic. I've never taken an interest in tuning because I know I'll never be able to get a handle on the whole pie. With my background I understand what that means with regards to what I can accomplish and the types problems that might arise which I can't fix (and that bothers me a lot). I'm in a world where things must be close to perfect, not much margin for error, but I get to see the whole pie and not just a slice. Plus, I need a break from the keyboard. I've been at this a long time.

 

[engineermike]

Just FYI, this is a stock boss lambda vs blk spark table:

This table adds almost 13 deg timing as the engine goes rich. Conversely, if it goes lean, it removes almost 13 deg timing. Most aftermarket tunes zero this table out because it's less math for them to worry with.

This is how I limit boost as a function of learned ethanol:

This table converts the learned ethanol % into a factor from -1 to 1, just like octane modifier. It then uses this factor to weight the load limit tables. -1 uses the low load limit table, 0 uses the nominal table, and 1 uses the high table. Below is how I have the low and high limit tables configured:

When learned ethanol is less than 50%, the -1 factor causes it to use the above table which limits load to 1.0 (about 0 psi boost) at low rpm and 1.5 (about 10 psi boost) at mid and high rpm.

This table changes the load limit to 3.0 when it learns 60% ethanol or more.

Both tables still limit load when the charge temp rises excessively, though I could probably increase this a bit when high ethanol is learned.

 

[WildHorse]

Just FYI

So have you applied any of this to your Mustang calibration ?

 

[engineermike]

So have you applied any of this to your Mustang calibration ?

Most of it. I'd done torque-by-drive mode a couple of years ago, but ultimately put it back. I am doing torque-by-gear in the car to soften 2nd gear street launches. I definitely incorporate timing retard-by-lambda in the car; did that long ago. The only thing I did in the truck but not the car was boost-by-ethanol % because I spent a ton of time tuning the car at 16 psi on pump gas before going flex.

Something I forgot to mention earlier that I've yet to see a commercial tuner do is to vary the system voltage as a function of fuel flow. You can safely raise system voltage to 15.2 volts at only high fuel flows, which acts as a pseudo-bap if you don't have one, and can get more voltage from a bap if you do. This was not that difficult to make work. And yes, I've done this in my car and truck. There are dozens of these things I've learned over the past few years so I'm forgetting a ton of them I'm sure.

 

[WildHorse]

Most of it.

But the 1000000 question. You ever figure out the proper tune for IMRC delete ?

 

[engineermike]

But the 1000000 question. You ever figure out the proper tune for IMRC delete ?

I’ve done a few. It’s not as intuitive as you’d think, especially gen2, but not awful either.

 

[WildHorse]

I’ve done a few. It’s not as intuitive as you’d think, especially gen2, but not awful either.

For instance / examples on a gen2 ? Or are you gonna take that info to yer grave >

 

[engineermike]

For instance / examples on a gen2 ? Or are you gonna take that info to yer grave >

I did it a couple years ago remotely for a guy with a gen2. I initially tried flipping all the switches hpt gives us that appeared to delete them, but encountered various other issues. In one case, it mostly worked but disabled knock advance. I wound up making it work by transplanting all the tuning data from the imrc-open mapped points to the equivalent-locationed imrc-closed mapped points, which worked in all aspects. Coincidentally, the same guy contacted me just a couple of weeks ago because he had discovered a simpler way. That’s his info though, so I’ll leave it up to him if he wants to share.

Sponsored