The correct way to use the Mylar strip is to measure from the center-width locations of the boundary line around the circumference of the now-deformed ellipse. Measuring from inside-to-inside or outside-to-outside is wrong! With a fuzzy ellipse boundary line (old stencil, poor gridding technique, etc.), it is not hard to make measurement errors more severe than just measuring the outside-to-outside dimensions of a crisp circle/ellipse. The width of the line forming the boundary of an etched circle is about 0.008″ which is also the thickness of the lines of the Mylar strip commonly used to measure the deformed ellipse manually (the diverging railroad tracks). If you are measuring inside-to-inside of a ellipse that was formed after starting with a 0.100″ diameter circle that was stretched 20% in one direction, you’ll measure the major axis as 0.120″-0.008″, or 0.112 inch, which is 12% on the major strain axis. You’ll also run the risk of measuring the minor strain wrong at 0.100”-0.008”, or 0.092 inch, which corresponds to a minor strain of -8%. Similarly, if you are measuring the outside-to-outside dimensions, you’ll wind up with 28% on the major strain axis and + 8% on the minor axis. A poor technique can turn a correct (20%, 0%) reading into anything between (28%,8%) and (12%,-8%)! Unless you are measuring from exactly the center- width position on the line making up the circumference of the circle/ellipse, you can get vastly different results, confusing the strain analysis interpretation.

Circle grid