Project E
 

Preface:

At career fairs and interviews, my fuel economy project seems to gain a lot of interest. I'm an Electrical Engineering major so I've been emphasizing the gains that are to be had from working on the electronic and control side of things instead of aero gains. So far this has worked well, but I want to go higher. I am currently in my fourth year of the fuel economy project. Currently I attribute 80% of my gains to my technique, 10% to tires, and 10% to my airdam.

Summery:

I'm going to pursue improved technique via improved sensor readout, reduced parasitic loads, and aerodynamic mods with an EE twist: Sensors everywhere. I had a precatlyst melt down and I don't know why, as such I'm going to have sensors in places that may get hotter with restricted cooling air. The upside of this is that I can optimize things more and free up more fuel economy. I ultimately plan to have an automatic shutter that opens preemptively before the fans kick on such as if the engine is under heavy load without a large positive change in RPM, logic being that the only time that happens is on a hill. Some of things I have planned may not be strictly needed, but I'm going to use this to show off my skills as an EE major.

Limitations:

• No LED or HID in Halogen housings.
• No defeating emission controls.
• No P&G (worn CV joints and poor fine motor skills).
• Nearly no money. I predict many consumer electronics will be torn down for free parts.
  

Road map:

16/3/27:

NOW. I'm going to adjust my tire pressure to known values.

16/3/28

I'll pick up a fuel pressure tester and start working on implementing the injector balance test process outlined in my Prizm's Factor service manual.

16/3/31

I hope to have datalogs from my commute so I can see how underutilized the cooling system is.

16/4/2

I will start work on thermal monitoring. Need to figure out how to monitor existing sensors and add new ones where I don't have them. I plan to use donated I2C sensors from some dead HP laptops. Will allow me to determine how I can keep various parts at operating temperature. Eventually I'll likely have a sensor network on this car.
Points of interest:

• Transmission, likely too cold.
• Exhaust manifold: Tends to crack due to heat and stress points due to the precatlyst on the California model, aftermarket replacements have a vented heat shield to reduce operating temperature.
• Engine Oil: I'm not sure how the engine oil temperature is regulated. I have managed to degrade conventional 10W30 enough to force an early oil change before. I suspect a leaky injector at play here, but I'm playing things safe since this is my only car and the most efficient car I have access to.
• Fuel tank: Exhaust runs close to my fuel tank.
• Battery: Has a cooling air supply sourced from driver's corner lower grill. Heat degrades lead acid batteries.
• Alternator: Right next to exhaust.
• Radiator outlet temperature vs inlet: Will allow me to track thermal headroom.

16/5/15

The project will be on hold first due to finals and then while I'm on an undergraduate research project.

16/8/10

The project will be able to resume with the ending of research project. I will attempt to set a record trip on a 110F day somewhere around this time. I know for a fact that under the right conditions, I can break 60 MPG with a geo prizm. At this point, the car will have nearly 300,000 miles on it with 40,000 of them being miles I have put on the car. Being a high mileage, high gas mileage car seems to attract a lot of questions.

Stuff that I have already done:

• I have installed my MPGunio and made a workable approximation of how ECU software behaves so that I can abuse every quirk in the software.
• I have reattached my airdam.
• I have switched to LRR tires.
• I have a massive spreadsheet I use to keep track of maintenance items by counting how many miles since the last time each maintenance item has been done. This ends up being key to keeping the car running well.
• I have made a timing diagram for the data output of the ECU (TDCL) so that I can code an ardunio sketch to decode it so I don't need to use a laptop for it. Took an afternoon of decoding poorly translated Russian forum posts to get the details right.

http://i.imgur.com/qhknE5g.png

Update: Can't find my TDCL cable right now. Can't datalog sensor data without that. Busy with coursework and organizing the junk room into a work space.

Wow! Impressive!

Found my TDCL cable. Tested MAP and TPS and they are accurate enough for the ECU's uses. One of my injectors sounds funny and I'm down to 37 with normal driving, 39 with efficient driving. My oil is 500 miles old and heavily fuel diluted, I'm actually consiering putting in some 15W-40 HDEO to keep my bearings intact. Project E will focus on getting fuel dilution down for the conceivable future. If I get the chance to swap things in during the process, I will, but otherwise don't expect much until I can get fuel dilution down.

I do plan to get my hands on Aerodynamics of Road Vehicles so I can reference it later, but that is a low priority at this time. I can pick up a copy on amazon for $80 if I really need to.

For purposes of 'career fairs and interviews' would it be preferable to have a flamboyant aerodynamic gizmo bolted to the trunk lid, or keep it on the down low so improvement appears to come from the other mods?

I could see it going either way.

I have a 4.312 final drive that gets me 3000 rpm at 60 MPH. If I put a boat tail on the car, it wouldn't net much without other tweaks. I'm thinking that hot air intake may be needed to help light throttle running. I can't bump start properly due to poor fine motor skills, all the P&G I get is with the engine on while doing hills. I go up a 7% grade and down a 5 % grade, up and down a pair of 5% grades, and then up and down another pairs of 5% grades on my drive to class. The 7% grade has me getting at best 18 MPG to hold my speed so I need the heat dissipation there, but then going down the grades I have the temp gauge drop like a rock. With a grill block to handle the Engine on Coasting, I can't handle the uphill parts.

I ask again, visible aero mods or invisible aero mods?

Seal the engine compartment and have movable shutters on the exits, either under the car or in the wheel wells. Doesn't the Prism have a low, oval air intake? A duct that shape that sticks out 3-4" in front, maybe with a bellmouth, that transitions to the rectangle of the radiator itself would be optimal. The exhaust side needs to be 'pulled' by the air moving past it, that might be the back edge of a step.

It's probably all packaged pretty tightly.

There will be visible aero mods. My logic is that less engine load == less heat==more need for heat dissipation management.

The prizm has an ovalish intake, but it has an obstruction going across the opening horizontal.


And now for something completely different:

Tried to test fuel pressure but the gauge doesn't work. Replaced fuel filter because the FPR sounded nasty and the pump was pulling enough current to make a paperclip too hot to hold while it was jumpering the pump on. Old filter was clogged up. With the new filter, the FPR sounds much nicer and the pump is drawing significantly less current.

I'll check the oil after a long drive and see if I'm able to burn off the gas in the oil. According to TDCL, my ECU is leaning out my fuel trims right now. Chances are that I'll have to recalibrate my MPGunio

I won't be much help on the instrumentation. I have a mechanical speedometer, a (modern luxury) electric fuel gauge and two warning lights. Oh and an infrared thermometer and air pressure gauge.

The reason I ask is that the benefit from something singular and obvious like a boat tail, would give about as much benefit as a diverter, full bellypan, wheel spats and a diffuser in the rear.

A boattail is a maybe. For short term, I'm thinking rear wheel skirts and smooth hubcaps. I have a little over a month until I have to abandon this project until August.

Small update: I've been looking at the datalogs I have and here is some useful ECU data about closed loop behavior:
The ECU needs to see a coolant temperature of at least 40C (If you are accelerating onto a highway, this takes about 30 seconds.)
The ECU also needs to see the oxygen sensor sending a rich signal at least once. Idling in the drive doesn't get the O2 up to operating temperature. 7 seconds of driving with a decent load matches about a min of idling. Should benefit from a heated O2, not sure how it would go over with the smog test guys though.
The ECU seems to switch over to open loop if it goes over 80kPa and cycles out if it goes below 70kPa (refresh rate of datastream is too slow to know for sure).
The ECU will run closed loop out to at least 4000 RPM and 80kPa with slow RPM change and a steady throttle, but it is more picky the closer you get to 4000 RPM, it wants to see 80C or more. Not sure about upper temperature behavior, I went up a 7% grade at 17 MPG in 5th and the coolant temperature didn't break 95C, however I was running at 90-95 kPa.
Movement of the throttle position sensor will cause the ECU to dump in a bit of fuel like an accelerator pump.
The ECU does not like to see the engine running lean.
Side notes:
TDCL injection time appears to be per fireing, not duty cycle, meaning that you need to look at RPM to determine how much fuel is actually being used. Also can't go to zero when in DFCO.

I'm working on an eco light as part of a subsection of the section 'Abusing the Toyota Diagnostic Connector [TDC] for fun and profit'. I think that the diagnostic connector has an output line that just happens to be at a certain voltage when the car is in feedback. It isn't super accurate, but it should help with efficient acceleration. The electronics required to give good enough results for most cases amounts to a N channel mosfet with an appropriate Vgs, a handful of resistors, and a green LED. Simple, easy to read, and you don't need to pull the dash to try it out since the TDC has a ground, a battery positive, and the ECU output lines right there.

When you get 'bored' with that.....can you stop by and see why my sunroof quit working???? ;):thumbup::D

And the warm days appear to be here again.
Got the best tank of the year tonight.

Drove into a 30 MPH headwind (44 MPH gusts) at 55 MPH and managed to get 28 MPG. Not bad. If anyone near me in % over EPA ranking thinks they are safe, I haven't even fixed my fuel pressure regulator (something that can cost 10-15% on this engine) or taken off my power steering belt yet.

Thoughts on a projector retrofit, allowing you to use HIDs or LEDs more safely? The savings are not huge, but also non-zero.

http://i.imgur.com/wewbW3a.jpg

Not outright permitted by the NHTSA so it is questionable. If I get in an accident, I don't want to have to worry about the person who hit me worming out of paying due to a technicality.

https://en.wikipedia.org/wiki/Everything_which_is_not_forbidden_is_allowed

Further depondent sayeth not.

Posting to draw attention to the headlight in my avatar pic.

That doesn't matter if you aren't a dealer selling the car for the first time. FMVSS 111 requires an exterior driver's side mirror, but most state regulations do not, and plenty of people here (myself included) do without the mirror perfectly legally. Similarly, the headlights have to be certified to FMVSS 108 standards "before the first purchase of the vehicle in good faith for purposes other than resale;" what you do with the car after you've bought it is governed by the state regulations where the vehicle is registered.

I didn't say that I wasn't going to do it. Questionable is just that: I'm going to have to do more research before I make a choice (stuff like finding acceptable material to make the lens out of, making sure that my electrical system won't murder the current regulator on the assembly, and that my auto insurance doesn't care about headlight mods-- they specifically asked about at least engine swaps and some other things I can't remember). I'll likely spend more time on the design phase than anything else.

Still need to deal with fueling issues before anything else though.

My current list of mods that I expect will give the best FE boost for the money and time spent:

Duct work between front of car and radiator.
Grill block sized to reflect the cooling improvement.
Side skirts.
An indicator to keep track when the ECU is in feedback mode.
Corrugated plastic hubcaps.