|MadSci Network: Physics|
Well, that's a tough one, and without information on the materials you were testing, I don't know how accurate my answer will be - hopefully this will at least give you some ideas on where to look next.
First, according to my dictionary, a lubricant is somethat that, by definition, reduces friction. So technically, I suppose your experiment was not really strictly looking at lubricants - but that's probably not the answer you were looking for, so I kept looking. :)
The next thing I thought of was the combinations of materials you are studying. As you probably know, materials have different coefficients of friction, depending on the other material in the couple. For example, steel on steel has a kinetic coefficient of friction of 0.57, but steel on copper has a kinetic coefficient of friction of 0.36 (both from this site). That means you will feel more resistance pulling a steel block of a fixed weight over steel than if you pull it over copper.
Now, say hypothetically you did a dry test with teflon on teflon (coefficient of friction of 0.04, according to Machinery's Handbook), and then tried to "lubricate" the surfaces with graphite. If you were able to coat the teflon well enough that you had graphite-graphite contact, your coefficient of friction would increase to 0.1 (also from Machinery's Handbook)
But let's say you already knew this, and were using lubricants that really should have decreased your coefficients of friction. There are a number of possibile explanations that I could think of that might work here. Could the fluid properties have been temperature-dependent? How about velocity-dependent? If your tests were being run too fast, too slow, too hot, or too cold, the properties of the lubricant might not have been what you expected. Could there have been impurities in the lubricant? Maybe some initial wear particles formed at the start of the test became trapped in the lubricant instead of falling away free in the unlubricated test? Last, did you check the measurement equipment? I'm speaking from experience when I say that sometimes that's the last place you think to look when your results don't make sense!
Finally, I came across this report from Georgia Tech, that seems to be aimed at the problem of lubricants forming semisolids and ceasing to lubricate. The work is geared toward work in nano and microstructures - where the forces we are used to dealing with don't always act as expected anyway! Things like surface tension and electrostatic forces are more dominant than things like gravity. Could this be the problem you were having?
Hopefully there's something in here that was helpful to you - good luck with your experiment!
Try the links in the MadSci Library for more information on Physics.