I have started the checking process on the CAD model of the lathe. I began with the countershaft hanger assembly, which included some of the first parts modeled. It was a little disappointing to find how many differences there were between the model and the factory drawing. Some reasons for this include:
- The design changed a little bit between 1937 and the 1960's when my lathe was produced
- A number of features were hand-ground after casting
- There is a casting draft angle of about 3 degrees that causes a feature measurement to vary depending on where the measurement is taken
- It is difficult to hand-measure surfaces that are organic rather than straight lines and planes
Below is an image of the cleaned up factory drawing:
The model was checked against the factory drawing by creating a drawing in the CAD system and dimensioning it in a similar way.
Below is an image of the CAD drawing. Fractions were converted to 3-place decimal dimensions and each dimension has a value enclosed in [brackets] that represents the factory drawing value.
I also think that the factory drawing is mislabeled, since it appears to show an earlier version of the hanger (M6-21) but has a drawing number of the later revised design M6-21A.
Anyway, these are a few of the difficulties involved in reverse-engineering a machine that was designed around 1937/1938 and was in production for almost 40 years.
The most valuable resource in reverse-engineering a cast-and-machined part is an original drawing, even if it is from an earlier version of the part. There are roughly 146 parts (not including things like fasteners and bearings) in this lathe and so far I have managed to acquire 28 drawings. If anyone knows where a drawing of an Atlas 618 part that is not already in
Downloads can be found, please let us know.