Thanks for the ideas, guys.
The trouble I see with using a drill blank or a piece of drill rod is that the tolerance will not be anywhere near what it should be. Thread wires should be within 0.00002" of nominal size, while drill blanks are usually around +0, -0.0003 .
I'd welcome a link to the sewing machine needle size charts; I did some searching but couldn't really find anything that gave decimal or mm equivalents. Also, don;t sewing machine needles usually have a flat side, or other deviation from true cylinder? Might make them hard to use....
Gauge pins might be an option, but the best I have been able to find are Class X with a 0.00004" tolerance, and they are about $13 each from MSC. I can buy a whole new set of the Pee Dee wires from Fisher Machine for about 2/3 of that cost......
Hmm, so after writing all that, I discovered that both the Flynn wires and the Pee Dee ones are only specified to 0.001" accuracy.
Why did I think they needed to be more accurate? Well, Machinery's Handbook says,
"A set of three measuring wires should have the same diameter within 0.0002 inch. To measure the pitch diameter of a screw-thread gage to an accuracyof 0.0001 inch by means of wires, it is necessary to know the wire diameters to 0.00002 inch. If the diameters of the wires are known only to an accuracy of 0.0001 inch, an accuracy better than 0.0003 inch in the measurement of pitch diameter cannot be expected."
The basic premise is that the error in the measurement works out to about 3X the error in the diameter of the wires.
I guess the consensus is that a 0.003" measurement error is OK, based on the Pee Dee wires and the Flynn ones being spec'd only to 0.001" accuracy.