I had this idea to scrape in a pair of 45-45-90 triangles, and get the angles perfect without a 45/90 reference. Is this already well known? Or is there a better way? Here's the idea:
- First, the faces need to be flat and parallel, and the edges need to be square to the faces.
- Now, clamp them together on the long edge. Measure parallelism of the opposing edges. When you have a parallelogram, the triangles are equal to each other.
- Flip one triangle over and repeat. When you have parallelograms in both orientations, the triangles are equal and symmetric (meaning the acute angles are equal, e.g. 44-44-92).
- Measure squareness or spot one triangle against the other. When the corners match, along with the parallelism above, you have two perfect 45-45-90 triangles.
Of course there's some strategy involved, to get all the measurements correct in a reasonable number of cycles.
I ordered some triangle castings from Gary Martin at Martin Model. They have been sitting on the shelf for a few months and I finally had some time the last few weeks to work on them. They came together reasonably quickly and I'm pleased with the result. I wish I had something accurate enough to measure the angles, I'm tempted to send them to a calibration lab (but I probably won't bother).
Here's some pics. First the final product:
Measuring parallelism:
Measuring squareness. This only works after you have parallelism. Flip the pair of triangles around, the readings should be identical. The setup in the pic gave the most repeatable results, but I still trust the measurement less than spotting.
Spotting for squareness. Blue one triangle, and stand them up against each other on the surface plate. There was some back-and-forth getting it to spot this well, while maintaining all the other measurements.
- First, the faces need to be flat and parallel, and the edges need to be square to the faces.
- Now, clamp them together on the long edge. Measure parallelism of the opposing edges. When you have a parallelogram, the triangles are equal to each other.
- Flip one triangle over and repeat. When you have parallelograms in both orientations, the triangles are equal and symmetric (meaning the acute angles are equal, e.g. 44-44-92).
- Measure squareness or spot one triangle against the other. When the corners match, along with the parallelism above, you have two perfect 45-45-90 triangles.
Of course there's some strategy involved, to get all the measurements correct in a reasonable number of cycles.
I ordered some triangle castings from Gary Martin at Martin Model. They have been sitting on the shelf for a few months and I finally had some time the last few weeks to work on them. They came together reasonably quickly and I'm pleased with the result. I wish I had something accurate enough to measure the angles, I'm tempted to send them to a calibration lab (but I probably won't bother).
Here's some pics. First the final product:
Measuring parallelism:
Measuring squareness. This only works after you have parallelism. Flip the pair of triangles around, the readings should be identical. The setup in the pic gave the most repeatable results, but I still trust the measurement less than spotting.
Spotting for squareness. Blue one triangle, and stand them up against each other on the surface plate. There was some back-and-forth getting it to spot this well, while maintaining all the other measurements.