Air Cooled VW and Modern Emissions 1

[blakeclark]

Conventional wisdom has it that VW and Porsche gave up on the air cooled engine because they couldn't meet stricter emission standards. Regardless of whether that's accurate, what would it take to adapt more recent strategies and technologies to clean up an 80 year old design? I'm not just talking about fuel injection, electronic ignition and catalytic converters. Early 80's VW air-cooled had these technologies, but supposedly still ran into an emission wall at some point. I like Vee Dubs and I'm curious to figure out just how close these engines could get to new car specs by applying out-of-the-box thinking. I'd like feedback on my logic and approach.

Air cooled VW engines run hotter than water cooled engines. This seems to be the first major hurdle. Among other things, preignition is more difficult to control. To compensate, VW kept compression ratios low and in the old days ran slightly rich. Neither is good from an emissions standpoint. The other problem with running hot is abysmal tolerances compared to modern designs. Things expand when they heat up, and in a AC engine moving parts are very "far" apart when cold. Lots of nooks and crannies for hydrocarbons to hide and opportunities for fuel to go where it shouldn't.

Possible solutions:
1. Aluminum cylinders. Available aftermarket for a small ransom, builders are able to design in much more precise tolerances for rings and pistons due to much better heat dissipation than cast iron.
2. Liquid propane port injection. To me, this seems like a good way to "beat the heat". LPG injected as a liquid would significantly lower the charge temperature. Combined with propane's higher octane rating should allow for higher compression. Propane supposedly doesn't deposit carbon on the combustion chamber, so no "hot spots". Provided you can control preignition, a higher compression ratio should theoretically reduce operating temperatures (specifically exhaust temperatures). Finally, as a "clean fuel" there's supposedly less to clean up to start with.
3. LPG would also all but eliminate evaporative emissions.
4. Increase the stroke from 69mm to 82mm. Faster moving piston should increase fuel mixing. The longer stroke could also potentially allow for more valve overlap to increase internal EGR.
5. It goes without saying: A closed loop fuel injection system, high output electronic ignition and 3 way cat. (I also understand there's room in the head for two spark plugs per cylinder.)
6. Other ideas: An electric motor (borrowed from a Neighborhood Electric Vehicle) to operate the cooling fan independent of engine speed. Use a start/stop system to eliminate idle emissions, or even a micro-hybrid like GM's eAssist to help with low-end torque allowing cam to be optimized for a narrower power band. (No variable valve timing coming any time soon to aftermarket...)

I know to get to modern tailpipe standards, the devil is in the details. Without the endless funds to sort all of them out that alone would probably keep a solitary engine builder from achieving modern standards. But with the strategies I've proposed (or with whatever else can be dreamed up) how close to new car emissions could Herbie get?

 

[mattsooty]

Hi Blake

Before you could even contemplate putting a vehicle through an emissions test a certain amount of upfront work is required on an engine dyno, fitted with an emissions analyser, to ensure that the engine out emissions are full optimised. This in itself is a significant and costly task.

When the engine is in the car and emissions development has begun one of the biggest challenges, and arguably the most important part of the emission test, is the cold start and catalyst heating phase; with the most critical points being lighting the catalyst off to its operating temperature as quickly as possible. This would necessitate you fitting the catalysts as close as possible to the exhaust ports as possible and/or utilising Secondary Air Injection to ensure as much heat as possible is transferred to the front face of the cat as quickly as possible.

During the afterstart/warm up (open loop fuelling) phases, minimising engine out pollutant emissions becomes crucial since, with cold catalysts, all of these pass straight out of the exhaust and into the sample bag. To enable the closed loop control to happen as quickly as possible the lambda sensor is required to be lit off as quickly as possible. However, after the cold start there is a risk of water condensate thermally shocking the heated cell in the sensor, so full heating power is held off for a period called the Dewpoint Time. If you accepted that longevity will be compromised you could calibrate the heating control so that the sensor is heated as quickly as possible, allowing closed loop control sooner.

Once the catalysts are lit off, and the fuelling is closed loop, the next challenge is the fuelling control whilst the cycle is driven. The catalyst deals with the three pollutants, NOx, CO & HC by way of two reactions: Reduction (NOx - N2 + O2) and Oxidization (HC - H20 + CO2 +CO), it does this with either a deficit or excess of O2 in the feed gas. This is achieve by alternately 'dithering' the fuelling slightly rich (for reduction reactions) or slightly lean (for oxidization); the popular notion that the fuelling is constantly held at stoich. is slightly misleading. The calibration of this dither is very application specific.

One issue that you would find is that to know whether you were making improvements each of the above would need to be calibrated and checked over an emissions cycle, with bagged emissions and also second by second. Which would be pretty bloody expensive!

Also I don't believe that the fact that the engine is air cooled is an issue for emissions per se but more for OBD, which utilises coolant temp as an enable condition for so many diagnostics and coolant temp itself is fairly easy to diagnose.

MS

 

[140Airpower]

Actually, I don't believe either Porsche or VW were forced to abandon air cooling because of emissions -although they might eventually have had to do that. The Beetle succumbed to more modern designs like the Rabbit and Sirocco. Porsche had insurmountable cooling challenges going to an air cooled 4-valve head.

 

[Marquis]

I'm not convinced that airccooled engines died due to exhaust emissions- even Nox.
For good debate we should consider the last of the aircooled Porsche 911s rather than an antiquated VW beetle with it's carburation etc.
The biggest hurdle in meeting modern petrol emissions regs is the cold catalytic converter light off period without a doubt. Yes, the air cooled engine will have higher metal temperatures and therefore slightly higher combustion temperatures which will lead to higher Nox emissions, but by this time the engine is warm and a modern 3 way catalytic converter does a good job in mopping up the Nox, HCs and CO.
if you look at the Porsche M64 engine in the 993 alot of things lend themselves WELL to modern emissions:
Very short ceramically lined exhuast ports. This has been done obviously so that the cooling fins can cope with the thmeral heat rejection but a fortunate side effect is it rejects more heat to the catalysts which aids rapid cat warm up.

Porsche 911 engines have had to be quite savagely oversquare (100 mm bore, 76.4mm stroke) which leads to large ring lands and therefore high crevice volume. However you can;t cite this for a reason for its death as the replacement watercooled 996 engine is watercooled by sports the same oversquare bore-stroke dimensions.
One thing that is correct was that it was very difficult to remain aircooled and migrate to a 4 valve/cylinder layout. This meant that in order to get the air flow into the chamber it was beneficial to have a oversquare chamber for bigger valves. Qversquare chambers aren't as efficient in terms of surface to volume ratio but again this isn't a reason for this aircooled engines death.
Porsche did the right thing with this oversquare chamber by adopting a twin plug layout for faster burn and to migrate knock limit and they pushed the compression ratio to as high as they could (something that Chrysler didn't do initially with their twin plug Hemi).

I still believe a 3 valve layout (2 inlet 1 exh) would have been possible with the aircooled layout-specially if oil spraying was utlised. This would get you to almost as high as a 4 valve specific output.

I like aircooled engines- as when vehicles get old one of the potential sources of problems is the radiator/cooling system and with aircooling you have a more robust cooling system over time.

Aircooled engines do have a certainly noise associated with them- the fins can cause ringing which isn't always perceived as agreeable.
And being constrained to a 2 valve layout does restrict ultimate BMEP and BHp/litre.

In anyway, the 993 aircooled Porsche will remain one of my all time favourites and I have zero intention on 'upgrading' to a 996 or 997 type car.

http://www.auto-scape.com

Sideways To Victory!

 

[patprimmer]

Marquis

I always liked the weight as well as the durability of the air cooled system, and the lack of water leaks from small cracks.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[140Airpower]

Pat, getting rid of water is indeed the surest way to get rid of water leaks. And, cracking is rare in air cooled engines since the worst overheating catastrophe that can befall a liquid cooled engine, loss of coolant, can't happen. But overheating of a correctly functioning system can happen, especially to automotive air cooled engines since the cooling capacity usually has a relatively low absolute limit.

Marquis, your thoughts about a 3-valve head are interesting. IMVHO, the worst problem I've seen for cooling an automotive air cooled engine is getting adequate airflow through the engine at the horizontal fin stratum of the exhaust port seat and combustion chamber top. Having side-by-side ports (like VW and Corvair) reduces this flow by over 50% compared to having ports in line (like Porsche flat 6). Comparing the Corvair flat 6 with side-by-side ports to the Porsche, it appears that the temperature-limited hp potential (not the breathing-limited potential) of the Porsche is at least 40% greater and I would attribute most of the difference to the greater and better placed air flow.

There would be two ways to do a 3-valve arrangement, having both intakes side-by-side with the exhaust in their lee or having the two intakes in line with the exhaust to the side. In the first arrangement, the intakes can be separated to allow airflow to the exhaust. In the second arrangement the stagger between the two intakes and the exhaust could allow a generous path around the exhaust to the inside of the combustion chamber top.

 

[Tmoose]

"cracking is rare in air cooled engines"

I think owners and shops specializing in Type 1 V-dub and 8 and 9 bolt heads Triumph twin MC engines of the 60s and 70s might not use the word "rare" when describing the incidence of cracks forming between valve seats, and even spark plugs, even with stock engines in "normal" service.

http://images.thesamba.com/vw/gallery/pix/767584.jpg

same forum, different engine

http://images.thesamba.com/vw/gallery/pix/767583.jpg

http://a-dingli.tripod.com/porting_and_polishing_vw_type_1_heads_files/image004.gif

http://www.type2.com/rvanness/images/head1.jpg

I guess even big HOnda singles are not immune

http://oldbikehack.blogspot.com/2011/02/what-goes-wrong-with-650-honda-singles.html

 

[140Airpower]

Yes Tmoose, I thought about those cases. Too much concentration of heat in a small area. There are air cooled engines that can sustain softening of the metal without cracking. I think it is a design problem that is avoidable with air cooling.