Has anyone attempted to pass emissions with a half-bridge ported motor?
Long story short, I've finally fully put together my FD from 18 years ago... 18 years ago this was going to be a drag car, so I put a half bridge ported motor in it (Precision Imports built it). Now, I just want to drive it around. Motor now has about 100 miles on it, all in the last 3 months.
I'm running non-sequential twins and all stock emissions equipment (minus the pre-cat). Haltech 1500, tuned on E85 (actually measured out to E67). Have a special lean tune just for emissions which targets 14.7 afr up to 0psi and like 4k rpms. After either of those, it starts adding extra fuel. Boost is set to wastegate (7psi I think) and is very lazy (about 1 psi by 3400rpms) due to back pressure. Brand new spark plugs morning of test.
Failed emissions (IM240 test). HC at 3.5, limit 1.2 and CO 39.6, limit 15. NOx passed at 0.0807 limit 3.
This points to running super rich, which if anything, it should be lean.
Anyhow, should I give up on this motor and see if I can trade someone for a mild street port, or is there something that stands out that I should check before throwing in the towel?

Thanks in advance for any insight

 

Wow.... half bride on twins is wild. That motor is trying to breath through a straw. What fuel are you running when you go test? I don't know anything about emissions testing fortunately but I imagine a different fuel would help here along with a stock cat if you can

The one time I had to do anything emissions related, i had the laptop set up while they were using the sniffer. Just leaned it out in real time until whatever number that was on their screen read 0. Super risky move but here we are 😅😅

 

Things that will likely help.

Get the engine and exhaust hot before the test cycle.

Fit a modern OEM style cat.

Make sure air pump is working.

Injector timing optimised on middle of intake stroke and staging lifted high to stop raw fuel passing straight to the exhaust on overlap from the secondaries or from injection on overlap on primaries.

play with ignition split and timing for richest reading AFR for a given injection quantity then lean out map. You might want to borrow or hire a 4/5 gas instrument to play with it, you may find slightly rich of lambda provides better combustion and lowers HC and still keeps you under the other limits.

Going to hotter thermostat/fan switch points for the test.

You may find a low octane/flex mix gets the best mix to get under targets.

 

Yeah, the only reason I left the twins on was so that I could pass the inspection... turns out, they didn't look at anything at all. Tons of back pressure though, I don't go over 5k rpms. Plan is to switch over to an EFR8474.
I do have a stock cat and I'm running E85 (closer to E67) which should be the cleanest fuel I can run besides pure alcohol.
I can't be in or near the car when they test, plus I'd probably blow it up leaning it out on the fly.

Thanks for the suggestions, these are really helpful.
I do have an oem main cat, but it could be worn out, not sure how to test that.
The injector timing might be really helpful, I have a feeling the overlap is sucking too much fuel out the exhaust.

 

What you need is something like this: Cat If this does not get it done with a good supply of air pumped in during the test, nothing will...

Seriously though, your problem is CO and HC conversion. NOx is very low so you can afford to take steps to lower CO and HC at the expense of increased NOx (which will occur if you get the CO and HC lower through increased air pump-induced oxidation in the cat). Your cat may not have the conversion efficiency left to convert the CO and HC, however. With a well performing oxidation substrate bed you should be able to get it to pass but the cat will get hot and will degrade quickly so you would not want to operate it like that except for smog check.

 

Well that's the nicest cat I've ever seen! It's something for me to think about.
I need to double check my air pump before I do anything else... it might be the issue

 

You can't bribe them?

 

If its not a daily driver, why don't you just get collectors car insurance and not have to worry about emissions at all?

 

I like that some of California smog tests still use a sniffer, since the goal is to run clean. I dislike that there are silly rules like whether the ECU or cat converter or fuel injectors need to be original 1990s tech, when newer tech could get better results without needing to cheat. The way I see it, running clean is running clean and sometimes the spirit of the rules may be more important than the letter of the rules.

Years ago I had access to an emissions sniffer and dyno and got to spend some time experimenting with an FD running E85 and standalone ECU. There are charts online suggesting that the best compromise is around lambda 1.0 and the NOx will skyrocket if the lambda mixture gets leaner than 1.0, but that is not what I remember finding. Running the mixture surprisingly lean would keep reducing CO and HC without much increase of NOx. I remember lowest emissions at surprisingly lean E85 fuel mixtures, something around 1.10 or 1.15 lambda according to the wideband sensor (which would be 16 or 17 AFR on a gauge using the regular gas scale most of us are used to). This was with the factory Mazda cat converter, ACV removed, and air pump configured to send fresh air to the cat converter but not to the lower intake manifold. The engine sounded pretty clean, it didn't sound like it was misfiring or struggling badly. I'm not sure if that would be healthy for a cat converter in the long term, I've heard they want to run near lambda 1.0.

 

You need high combustion temperature to generate NOx, although our plugs cop a flogging actual flame temps aren't that high due to the wonky chamber shape, especially with ethanol blends we don't see the temperatures to generate significant NOx.

The only problem with running that lean is that some testing regimes have minimum CO2 % on load to stop people just diluting exhaust stream with pumps/fans/charger bypass so there may be a limit to how far you go.

 

I do have collectors insurance, and will end up with collectors plates. Here in Colorado you only get exemption if your registration has never lapsed or expired... this car hasn't been registered in 18 years, so it needs to pass at least once.

Colorado apparently doesn't look for anything anymore, they just care that you pass the sniffer. I thought they would check everything since I had to do a state insection prior to emissions... all they checked was that my headlights/turn signals and brakes worked.
This is good info! I'm more worried about blowing up my motor going that lean though. We targeted 14.7 AFR. No idea how far into boost or RPMs they got though.

My CO2 level was 466gpm, but there is no pass/fail condition on the sheet.
This brings up a good question though... I'm running an auto air pump which has a different sized pulley than the manual. I wonder if there is a difference in air being pumped to the cat or not.

 

that is interesting. i've gotten engines to fail with high Nox's, but i'm in CA, and our limit is really low. we only get like 450nox's or something, and a completely stock car will do 150 or so

 

Please run the test. If it fails, scan the exact paperwork you get and provide it here. I (and others) have provided detailed analysis and support based on what exactly failed and by how much.

 

I was reading some of your analysis on other threads last night trying to correlate it to mine. Your insight would be much appreciated.

 

were you able to run a datalog in the ECU?

 

I have zero experience on E85.

On pump gas my experience is same as j9fd3s. I can say that despite our inneficient combustion chamber shape when I ran very lean cruise (16-17afrs) my cruise EGTs were 1000C measured at the exhaust ports and over time melted the Greddy egt probe.

Note, I had a hyped up ignition system so there were zero misfires to cool the egts unlike a stock ignition car trying to run lean.

I had tip in and load very rich for throttle response, so it would drop immediately from 100C to 750-800C on throttle and then slowly rise back up if I kept WOT through 5th.

Friends driving behind my car on freeway complained of stinging and watering eyes. J9fd3s told thats NOx.

 

Btw, melted egt probes and shifting the internal shape of cast turbo manifold divider was the only damage or excess wear I had in 14 years of lean burn cruise (no cat though).

 

If injected air is introduced ahead of the catalyst bed under certain conditions, additional NOx will also be generated. But, getting air to the cat will also increase the amount of CO and HC reduction assuming that the cat is still functioning well enough - at the expense of greatly elevated substrate temps. The Colorado NOx limits seem pretty high, though, so the OP could manage higher NOx to a degree. What would probably work the best is to have the airpump active continuously, with air directed to the both engine port and also to the catalyst through the regular inlet pipe. That way, secondary combustion will be happening all the time for CO and HC reduction, and the extra air to the cat will really get the oxidation substrate cooking.

 

Nope, I didn't even think about using an internal datalog since I always log straight to my laptop.
Next time I will, that would have been very useful

 

so this got me curious, how would an emission test look if tuned for pure mehanol? or e100? cleaner? idk

 

i'm out of practice , i used to be able to match the smog machine, which probably explains a lot... ive smogged a lot of cars in the last ~30 years. California's test is really weird though, so nothing we do really applies to the Colorado test. we just do a steady 15mph and then steady 25mph, and then go over the emissions system like its the Pebble Beach Concourse. i've failed because the downpipe was too shiny before.

but yes, Nox's are like a sharp smell, and eye burning. HC's kind are sort of heavy and sweet, its what sticks to your clothes. the closer you get to optimum the more like a burning pineapple it gets....

 

Sorry for the late response. You're dead by 105 seconds, and it's just getting worse at 120. Any of those square looking CO traces means you're maxing out the range of the chart. Your big CO and HC spikes are coming during acceleration. Airpump and cat don't do too much there, not when the engine out air fuel ratio/lambda is rich and mass flow is high. I mean obviously you should check the air pump and cat, but from what I can see here, it comes down to your tune in the sense that if you don't lean it out during the heavier accelerations you don't have a fighting chance to pass.

Your tune is just too rich. There are a few ways you can go about doing handling this. You can rely on the E85 knock resistance and set the whole Lambda to 1, but only right before the test. You can restrict the boost even further by plugging up the exhaust (especially if you have a canister style with an included silencer plug).

If you're more caution minded, take the fueling (I'd have to see how you have it all set up) and essentially lean out the whole target lambda (the area that's not already set to lambda 1) by 25% steps relative to the current one.

So Lambda of .8 becomes .85 , lambda of .9 becomes .925 , etc. Just across the board. Go run the IM240 test again. See if it passes. Then increment again, instead of 0.8 you've got 0.9, instead of 0.9 you've got .95. Just lean it out in steps until your HC and CO spikes that come during the big accelerations (emission spikes after 5, 90, 155 seconds for example) come down. It could also depend on what kind of tip in fueling you're using. You may have to cut that way down too.

 

Thanks for the reply, I agree that really it comes down to leaning it way out higher in the rpm and higher into boost. I've just decided that it's not worth risking the motor, especially since most of the emissions people don't drive manuals very well, so I'm just going to register it at my dad's house out of the county where I don't need to do emissions.
Thanks to everyone who gave their input, I figured this was a long shot to get a motor like this to pass. Maybe this thread will help someone else out in the future.

 

would adding a switch to the airpump (or using an rx8 electric one) and forcing it to run all the time have an effect in the test?

 

There's so much fuel coming out of the engine during the acceleration from stop that I don't think the air pump is going to do much either way.