I've been watching this thread for a few days and I'm seeing a lot of posts here that tend to suggest that not many of the posters really understand what it takes to accurately and repeatably measure to .0001", one ten-thousandths of an inch with certainty. The total window for all three bearing classes is .0009". In many shops with average machinists, this much measurement error can easily be due to variations in micrometer handling technique and calibration errors.
Let me begin by stating that a dial bore gage is a comparative instrument. It must be set to a reference standard at the nominal dimension of the part you wish to check. The reference standard must, at a minimum, have four times less measurement uncertainty than the tolerance window of the part, preferably ten times.
For measurements +- one thousandths of an inch, a micrometer will work as a reference standard as they are generally capable of accurately measuring to a tolerance window of half a thousandth. It is folly to assume a vernier micrometer is significantly more accurate due to lead error and the heat input of the user's hands, among other things.
If you're going to measure to a ten-thousandths of an inch as you need to do here, a micrometer is completely unsuitable as a reference standard. To set a bore gage to tenth-level accuracy you need a reference with a maximum uncertainty of not greater than .000025", 250 millionths of an inch. This requires reference to either a ring gage or a gage block stack.
Next comes the question of where you're going to take these measurements. Hopefully not in your garage! An average human hair is .005" in diameter and dust is around .0001". Further, the temperature of your work environment and tools will play a massive part in the accuracy of your measurement. Metal expands and contracts with temperature and different metals expand and contract at different rates. The ring gage or gage block master will be most accurate at it's calibration temperature, which is 68F. Any deviation from this standard temperature will cause a measurement error due to differential expansion and contraction of the gage block master and the aluminum engine case.
Example:
Let's say you measure the engine case at 78F, 10F above the ISO Standard Atmosphere. Steel expands at .0000073" per degree F per inch. Aluminum expands at almost twice the rate of steel, .0000127" per degree F per inch. .0000127 - .0000073 = .0000054" differential per degree F per inch.
.0000054 x 10F x 1.417" diameter = .00008". Differential expansion between the steel master and the aluminum engine case will eat up almost .0001" with a mere 10F difference in temperature. Your total window for one size of bearing is only .008mm or .0003"
Then there's heat input to the measuring instrument. Let's say you set it to the master reference and then take a break. For giggles, lets say you warmed the tube of the gage by 5 degrees F over its 10" length setting it. .0000073 x 5F x 10" = .00004" Then you took a break and came back half an hour later to take the measurement. Your gage is now reading almost 1/2 a tenth too small!
The two temperature errors above, when additive, are enough to cause a measurement inaccuracy of half the total tolerance window for each bearing size. The point I'm trying to make here is that measuring to this kind of accuracy is not as easy as buying the tools. There are standards to be considered along with a level of technique that one develops with practice. That's not to say you can't take these measurements, but it simply isn't as easy as buying a tool and using it. There are a number of variables besides the tool that must be controlled to ensure an accurate and repeatable measurement. These little errors can add up quickly and destroy any certainty of measurement.
Unless you have a serious reason to feel the bores may be out-of-round (something a bore gage won't necessarily detect in the case of an odd number of symmetrically located lobes) I would use the bearing class the manufacturer originally used when the engine was built and verify with Plastigauge. Suzuki stamps their engines with the bearing class for each journal right on the case, I suspect Honda probably does something similar.
