Magnetic fuel treatment - again...

[mroek]

I know there has been a couple of threads on this subject earlier, but I found it best to start a new topic.

Today a salesman from a company called AutoMEC sent me some documentation to supposedly prove that their magnetic fuel treatment device really works. He wasn't sending me this to get a new customer, he had merely been pointed to a thread on a Norwegian car-related forum where I voiced the opinion that these things are a scam, and doesn't work.

Since AutoMEC is a Norwegian company, their website is in Norwegian. For those of you that can read Norwegian, here's the link:

http://www.mec.as/

There is a sketch that supposedly explains the working principle behind the device:

http://www.mec.as/?CatID=1261

The text to the top picture says: The sketch shows how oxygen (blue) in the air and fuel (green) appears as molecule clusters.

The text to the bottom picture says: The sketch shows how AutoMEC divides the molecule clusters into smaller clusters, thus increasing the reaction area, yielding faster and more optimal combustion.

Further it says that consumption and emissions are reduced, and torque and power is increased.

The device itself consists of a string of two neodymium magnets that are strapped to the fuel line, and a string of three magnets that are strapped to the air supply line.

A picture of an actual installation can be seen here:

http://www.mec.as/?CatID=1274

Now that's the background. Turning to the "evidence":

I've received one dyno test of a large diesel vehicle (a bus), in which the difference with and without the device is very small, but increases after they've applied correction factors for air temperature, diesel density and barometric pressure. This test is useless, in my opinion.

Much more credible is a test conducted by the Swedish laboratory http://www.avl.com/mtc , for those that can read Swedish, the full test document can be downloaded at the bottom of this page:

http://www.mec.as/?CatID=1275

The test document contains the full test procedure, accompanied by pictures and graphs. Translating the full document would be a bit work, so I'll just give you a brief outline:

The test was conducted on one single diesel (turbo) powered car (with an automatic transmission). First they tested it without the device, then with the device, and then without again, referred to by them as an A-B-A test. The tests were done on a rolling road (a chassis dyno), in a controlled environment with a temperature of 22 degrees C, and a humidity of 50%. Prior to each test, they ran a proof test, and then allowed the vehicle to stand still in this environment for 12-36 hours prior to each of the real emission tests, which consisted of three runs each, to allow for calculation of standard deviation and average.

Between the tests, the car was driven for at least 3000 km to ensure that the device would be allowed to perform optimally. The tests in the lab were run on reference fuel, but the driving inbetween the tests were run on regular commercial fuel.

The conclusion of the test is that emissions of CO, HC and NOx is unaffected by the device, but they measured a significant reduction of particle emissions (9%), and a reduction of CO2 and fuel consumption of approximately 3-7%.

For some reason they fail to emphasize that particle emissions fell steadily throughout the test, regardless of the the device. Figure 3 shows this.

At least two large Norwegian transportation company now embraces this "technology", and are installing it in their fleet of vehicles. I find this both alarming and amusing.

I still think this is a scam, even if the test concludes that it actually works!

Edited for spelling and one question:
Would such a device be eligible for the million dollar test?

Regards
-Øyvind

 

[ceptimus]

Immediate thing that springs to mind is that high-power magnets installed near the turbo control system (waste gate) could be causing the turbo to run at a higher boost pressure than it should.

This would give a greater power output, and under certain circumstances might even make the engine more fuel efficient, but this might be at the expense of emissions, shorter engine life, etc.

I don't believe the magnets are doing anything to the fuel or the air, but it's quite possible that they are messing with the sensors on the engine, or the electronic control wiring, or the injectors, etc.

Also the test method seems a bit spurious. How did they allow for variations in atmospheric pressure, temperature and humidity between the tests? Did they change the engine oil and service the vehicle between tests? How many extra hours running had the engine done between the first and last tests?

If these spurious effects can be eliminated, and the magnets still give a performance advantage, then a) I'll be amazed, b) the physics and chemistry books will have to be rewritten, c) manufacturers will all start to fit these wonderful magnets as standard, and d) once it's understood, it will become part of mainstream science.

Pardon me if I don't hold my breath while I wait for that to happen.

Personally, I don't think this product would be eligible for the $1 million challenge, but you'd have to ask Mr Randi. I know he hates all these magnet scams, so it might be...

 

[Mojo]

How did they allow for variations in atmospheric pressure, temperature and humidity between the tests?

In a way that appears to have been highly favourable to their product:

Originally posted by
mroek
the difference with and without the device is very small, but increases after they've applied correction factors for air temperature, diesel density and barometric pressure.

 

[mroek]

Sorry if I wasn't clear: I referred to two tests, the first is easily thrown in the bin for lack of almost every scientific method. That's the one where the corrections made most of the difference.

The second test was more serious, and they had a stable climate (22 degC, 50% humidity) indoors where the tests were conducted. However, as I wrote, the car was driven quite a lot (at least 3000 km) between the tests.

The report doesn't say if the car was serviced, so my guess is that it wasn't.

 

[Ziggurat]

The text to the top picture says: The sketch shows how oxygen (blue) in the air and fuel (green) appears as molecule clusters.

The "before" part of the picture is absurd to begin with. Oxygen in the air sure as hell doesn't clump like that. It's a GAS, every molecule floats about randomly, it simply doesn't clump. If it DID clump, then oxygen would seriously violate the famous ideal gas law, PV=NRT, by some monstrously large number (we're talking an order of magnitude, if their picture is to be taken seriously), and it doesn't. As for the liquid part, well, that's silly too. Liquids don't clump together, liquids ARE a clump. That's what separates them from gases, they're already all bunched up. If they're implying that a magnetic field can cause liquid droplets in air to break apart into smaller droplets, they've got a big barrier to overcome. Not only is gasoline not magnetic (which means it hardly responds to a magnetic field in ANY way), but in order to break it apart, you'd need to affect different gas molecules differently, pulling some in one direction and others in another direction, WITH THE SAME FIELD. That stretches credulity, to put it mildly. Not to mention that magnetic fields can only produce a net force (as opposed to torque) from a field gradient, not from the strength of the field itself. The change in field from one side of the droplet to another are going to be so miniscule that, even if you did have one form of gas that got pulled one way and another that got pulled the opposite direction, there's no chance in hell the forces could overcome the surface tension of the droplet.

Complete and utter hogwash.

 

[mroek]

Thanks for your opinions!

It seems (I asked Randi himself) that the device is eligible for the prize, so I've asked the company to apply. I am not holding my breath, though...

 

[mroek]

I would appreciate some statistics help.
This is the data from the test done at the Swedish lab that I linked to in the first post:

                 Total   Urban  Highway
Without AutoMEC  0.7712  1.0811  0.5901
Without AutoMEC  0.7519  1.0861  0.5563
Without AutoMEC  0.7556  1.0688  0.5718
With AutoMEC     0.7180  0.9962  0.5522
With AutoMEC     0.7168  1.0020  0.5492
With AutoMEC     0.7292  1.0111  0.5622
Without AutoMEC  0.7472  1.0441  0.5733
Without AutoMEC  0.7419  1.0318  0.5748
Without AutoMEC  0.7534  1.0486  0.5816

The first three tests are without AutoMEC installed, then three tests where it is installed, and lastly three tests when it has been removed again.

The numbers represents fuel consumption in liters/10 km, and for each test they tested two different driving cycles, one to simulate urban driving, and one for highway driving. I can't make any sense of the "Total" figure, though, but that can be ignored for now.

As you can see, the tests with AutoMEC installed shows a lower average fuel consumption that the tests without it.

What I'd like to know is how to calculate the probability that these results are not random. I'm guessing it involves standard deviation and confidence intervals, but I'm not a statistician.

I believe this test is flawed, but knowing how to calculate the above would certainly be valuable in the future.

Regards
-Øyvind