Random ramblings

Thursday March 31, 2005

Where's all the oil gone?

Watched Mythbusters yesterday, and they did an experiment to determine the relative fuel consumption of an SUV/people-carrier type thing with the air-conditioning on versus the same SUV/people-carrier with the extra aerodynamic of drag of open windows. They used a laptop hooked up to the SUV's ECU to get an MPG figure for same SUV first with air-con on full blast, then with air-con off and all windows open - while driving round a track at a constant speed. The resulting MPG figure was, to this European, shocking - about 11.7MPG for air-conditioning versus ~11.3MPG for open-windows. Now, apparently US gallons are smaller than English ones3 - as far as I can ascertain 1 US gallon is 0.83 Imperial gallons, so those figures would be 14.1 and 13.6 imperial MPG, or about 20 litres/100km!!

That's just a stunningly low MPG figure, even with windows-open or air-con running! Just for comparison, here are (rough) figures for some european/japanese gas guzzlers:

• Aprilia RS250 motorbike
  • Engine: 249cc two stroke V-twin (Suzuki RGV250 engine).
  • Year: 1999, original engine design dates from 1980s.
  • Transmission: 6 speed, wet clutch in low-viscosity gear oil

  It has an insatiable thirst for petrol due to the fact that it is a two-stroke and hence spits an ungodly amount of the petrol it consumes out of the exhaust only partially-burnt1 (It is otherwise a very efficient engine). Does about 140 to 170km on about 16l of petrol, depending on how it is ridden. However, it is still a bike, and bikes generally are more efficient, being lighter and less loaded down with useless crud like heavy seats, acres of plastic, spare tyres, tool boxes and vanity mirrors.. (well, real bikes at least ;) ).

  • l/100km: ~11.3 (worst) to 9.1 (best)
  • MPG (imperial gallons): 25 to 31
  • MPG (US): 20.75 to 25.73

  Note that four stroke bikes, particularly modern ones, get much better MPG, 50MPG is easy for a four stroke bike, 70 or higher MPG isn't uncommon. But the RS250 still does well enough to last a whole afternoon at the track on one tank - about 90 to 100 miles running mostly at full scream - and still have enough left in reserve to make it to the petrol station in Naas. Oh, it consumes 1l of two stroke oil every 300 odd miles too.

• Mini Cooper
  • Engine: 1.3l inline four, pushrod driven valves, fuel-injected2
  • Year: 2000. Engine design dates from 1952:
  • Transmission: 4 speed + reverse, wet clutch in engine oil
  • l/100km: 10.4
  • MPG (i): 27 (urban cycle)
  • MPG (us): 22

  Very very old engine design - the original "Series A" engine, as found in various British Leyland / Austin / Morris cars over the years since 1952 and the MG Metro in the 1970s/1980s. Engine is mostly the same as the original 803cc A series engine, only minor, incremental updates to the basic engine block, some changes to the valve-train and head over the years (in the 1970s mostly) and carbs replaced with fuel-injection in the early 1990s2

. Age of the design is reflected in the relatively small amount of power it puts out, only 50 or 60bhp or so (the 250cc RS250 puts out as much if not more). It still manages about 27MPG though, on the not terribly free-flowing urban Dublin cycle, so that's probably the worst it can get.
• Rolls Royce Silver Shadow
  • Engine: 6.8l V8
  • Year: 1976
  • Transmission: Automatic
  • l/100km: 20.2 to 16.2
  • MPG (i): 15 to 17
  • MPG (us): 11.62 to 14.11

  My father has a weird interest in old cars, particularly luxury ones, hence this and the next entry. He claims he has gotten 17MPG out of this beast on long steady journeys, despite its monstrous engine.

• Jaguar Mark VIII
  • Engine: 3.4l inline 6 cylinder
  • Year: 1958
  • Transmission: Automatic
  • l/100km: 23.5 (worst) to 18.8 (best)
  • MPG (i): 12 to 15
  • MPG (us): 10 to 12.45

  Possibly the most immensely thirsty vehicle I have ever driven or been driven in (again, because of my father's obsession with cars he drooled over as a kid), at least on this side of the atlantic. Despite a rather small engine, compared to the Rolls Royce at least, this thing is so thirsty it has two large fuel tanks, with a switch to select which tank to slurp from. My father claims it usually does about 12MPG, but that he can get 15MPG out of it on a very good day.

• The US-spec SUV/People carrier
• Year: New apparently (so 2003->2004?)
• Make/Model/Engine size: unknown
• l/100km: 20.8 to 20.0
• MPG (i): 14
• MPG (us): 11.7

Bear in mind they ran with air-con turned all the way up then with air-con off but windows down @ steady 90km/h around a track. Air-con isn't terribly common over here (least not in Ireland), but we do drive with the windows down occassionally, though usually to get rid of condensation from the cold rather than excess heat. ;)

For further comparison, the Subura Impreza WRX (2.0 litre) is listed on subaru.ie as having urban cycle fuel efficiency of 15.5 l/100km (18MPG, 15 US MPG), and 11.1 l/100km (25.4MPG, 21 US MPG) for the "combined" cycle. The more sensible Subaru Legacy 2 litre is listed as 10.5 (26MPG, 22 US MPG) and 7.5 l/100km (37MPG, 31 US MPG). Note that I didn't quote the "extra urban" cycle.. (5.9 l/100km, 47MPG, 39 US MPG), and note further that Legacy isn't that economical for a modern car :).

I actually used to feel a bit guilty when riding my RS250 smoker, guilty at the trail of oil and half-burnt hydrocarbons I knew it was leaving behind (quite literally4), but it only does a thousand or two k every year. I also know that my Mini, with its 1950s era engine, isn't as efficient as modern cars, but it's such a joy to drive compared to modern cars that I couldn't bear to get rid of it. However, I think now I can feel a lot less guilty, safe in the knowledge that there's an entire continent whose modern cars aren't half as efficient, and busily working away to rid the world of the problem of oil pollution, by burning it as fast as possible... ;)

If there are any automotive engineers reading: Exactly how do you make a modern vehicle slurp more fuel than a 6.8 litre, 2.5 tonne, 1970's Rolls Royce, and barely less than a big 1950's "sports saloon" Jaguar?

1. Because of this, two stroke bikes over 50cc are no longer for sale for street use in Europe, due to emissions. Aprilia had developed a direct injection two stroke engine which solves most of the emissions problem, but they don't seem to have done anything with it. Lotus apparently have a pneumatic-valve two-stroke testbed engine, but I've not heard of this technology making its way to commercial engines.

2. Again, due to emissions regulations.

3. Yet another reason to despise imperial measures. Yards, feet, gallons, barrels, bushels, etc.. stopped being a sensible idea about 300 years ago.

4. I've had people on bike ride outs tell me that they knew they had been catching up on me on a road because they could smell the two stroke oil.. ( Mar 31 2005, 03:56:43 AM IST ) Permalink Comments [0]

Tuesday March 01, 2005

CDDL and patents

There appears to be some confusion in the greater "Open Source" community regarding CDDL and patents. It seems pretty simple really, at least to my legally-untrained eyes (corrolary: please do not take anything in this post as gospel truth or as anything but my personal, armchair-IP-lawyer, IANAL opinion). The licence essentially says "If you want to join the party, you have to bring your patents with you" (wrt to any contributions to the work concerned). This is roughly analogous to the GPLs once contraversial "viral" feature where it requires all derived works to also be licenced under the GPL (or compatible licence). The CDDL simply extends this to patents.

Further the CDDL has a patent MAD clause (see 6.2 of the CDDL), if a user should sue a contributor for patent infringement in their contribution to the CDDL covered work, the litigating party loses all rights granted to them under the CDDL by contributors. Or to put it another way: the community of users of and contributors to a CDDL project are each bound, at pain of losing all rights granted to them by the CDDL to that software, to not sue each other for patent infringement regarding contributions to that software. Ie, to use the software of a CDDL project, you have to leave your patent-litigation club at the door.

Given the recent news which makes it look ever more certain that the EU will adopt a new patent directive mandating software patentability, and that if the EU somehow does not adopt that position, much of the rest the world does allow it, the question of whether and how open-source software projects should tackle patent risk likely will need to be addressed. Which makes the CDDL's introduction of patent pooling and patent-MAD to open-source software licencing very interesting at least, and IMHO a very good idea .