Engraving the Maker’s Mark on the Engine’s Hub Caps (April 2020)
The ‘Allchin’ maker’s mark was cast into the hind hub caps on the original traction engine and the same should be achieved on the model. My Grandad had acquired some brass pressing of the correct size (see the insert image), that could be attached to the hubcaps to achieve this result; however, in my opinion, this would look a bit naff, especially if the hubs were to be left polished. I therefore opted to engrave the detail into the hub caps on my small CNC milling machine.
In the last year I had to upgrade my mill with reciprocating ball screws, which are now very reasonably priced and readily available on Ebay. I machined the ends of the screws to suit the ways of the mill and modified the space where the bronze nuts were located to house the ball nuts. This modification was amazingly straightforward to do and could be covered in a future article. Before the upgrade, the mill suffered with about 0.3mm of backlash in X & Y and with the ball screws fitted I could not measure any backlash with the measuring equipment I owned!
With the CNC mill now being capable of engraving the tiny detail required on the hub caps, a 3D model was created in Autodesk Fusion360. The working drawings for the Allchin provided a guide and some photos found online provided more detail; however, the arrangement was essentially achieved by eye using a font that looked like the original. The overall size of the design is 25.4mm in diameter so the details of the individual letters were incredibly fine!
Design For Manufacture
The design was laid out in a sketch in Fusion with each letter forming a single entity and the ‘dots’ and other details drawn manually. I planned to do the engraving with a 0.2mm (diameter at the point) 20 degree inclusive angle D bit cutter/engraving cutter, however to achieve the best result the model needed to take into account the possible runout in the chuck and the other possible out of tolerances associated with cheap Ebay cutters. For example, the minimum slot width that could be achieved was 0.2mm (plus a bit); thus, any detail in the design that was less than that would not be machined. I therefore had to move some of the letters around to ensure that the gaps were greater than 0.2mm, plus the thickness of the font had to be reduced to achieve sufficient gaps inside, for example, the letter ‘A’.
The ‘Allchin’ maker’s mark was cast into the hind hub caps on the original traction engine and the same should be achieved on the model. My Grandad had acquired some brass pressing of the correct size (see the insert image), that could be attached to the hubcaps to achieve this result; however, in my opinion, this would look a bit naff, especially if the hubs were to be left polished. I therefore opted to engrave the detail into the hub caps on my small CNC milling machine.
In the last year I had to upgrade my mill with reciprocating ball screws, which are now very reasonably priced and readily available on Ebay. I machined the ends of the screws to suit the ways of the mill and modified the space where the bronze nuts were located to house the ball nuts. This modification was amazingly straightforward to do and could be covered in a future article. Before the upgrade, the mill suffered with about 0.3mm of backlash in X & Y and with the ball screws fitted I could not measure any backlash with the measuring equipment I owned!
With the CNC mill now being capable of engraving the tiny detail required on the hub caps, a 3D model was created in Autodesk Fusion360. The working drawings for the Allchin provided a guide and some photos found online provided more detail; however, the arrangement was essentially achieved by eye using a font that looked like the original. The overall size of the design is 25.4mm in diameter so the details of the individual letters were incredibly fine!
Design For Manufacture
The design was laid out in a sketch in Fusion with each letter forming a single entity and the ‘dots’ and other details drawn manually. I planned to do the engraving with a 0.2mm (diameter at the point) 20 degree inclusive angle D bit cutter/engraving cutter, however to achieve the best result the model needed to take into account the possible runout in the chuck and the other possible out of tolerances associated with cheap Ebay cutters. For example, the minimum slot width that could be achieved was 0.2mm (plus a bit); thus, any detail in the design that was less than that would not be machined. I therefore had to move some of the letters around to ensure that the gaps were greater than 0.2mm, plus the thickness of the font had to be reduced to achieve sufficient gaps inside, for example, the letter ‘A’.
Cutting Speed Testing
My mill runs comfortably at 2500 rpm; it can do 4200rpm but I do not have the heart to run it at this speed for any extended length of time! Therefore, I needed to conduct a series of speeds and feeds tests to inform the CAM program. I fixtured a piece of scrap brass, set the 0.3mm depth of cut, and ran the engraving tool at 2500rpm through a series of straight slots at ever increasing feed rates until I broke the cutter. It broke (repeatedly) at 80mm/min so I opted for 40mm/min to be safe for the actual engraving to give some scope reaching the end of the job with the tool intact.
My mill runs comfortably at 2500 rpm; it can do 4200rpm but I do not have the heart to run it at this speed for any extended length of time! Therefore, I needed to conduct a series of speeds and feeds tests to inform the CAM program. I fixtured a piece of scrap brass, set the 0.3mm depth of cut, and ran the engraving tool at 2500rpm through a series of straight slots at ever increasing feed rates until I broke the cutter. It broke (repeatedly) at 80mm/min so I opted for 40mm/min to be safe for the actual engraving to give some scope reaching the end of the job with the tool intact.
Computer Aided Manufacture (CAM) Development - Fusion360 has a built in CAM function, which was loaded with the 3D model, tool features, cutting speeds etc, and a CAM sequence developed using an adaptive cutting pattern. A simulation was ran at 40mm/min and the program suggested a 1hr 10min cycle time, which is ages for such a tiny job! I tested the CNC recipe by isolating a small portion of the job (again, a simple operation in Fusion) and only creating the tool paths for that and then exporting the test code to the CNC machine. I ran the cycle on some scape brass and found that run out in the spindle created a larger effective tool diameter; thus, some of the finer details were milled away, plus the federate needed to be reduced further to achieve an acceptable surface finish. Fusion has the option to ‘leave material’ in a CAM sequence as a distance value from the modelled features. I used this option to tell the program to leave some stock around the features; thus, accounting for the tool run out and reduced the feed to 25mm/min.
The result was a cycle time of over 2 hours for each hub cap, which was a function of the tool not really running fast enough (10,000rpm would have been better!) and, therefore, having to have a very low feed rate. However, that was the recipe that was right for my machine, so that was what was used!
The result was a cycle time of over 2 hours for each hub cap, which was a function of the tool not really running fast enough (10,000rpm would have been better!) and, therefore, having to have a very low feed rate. However, that was the recipe that was right for my machine, so that was what was used!
The Result
The actual bronze hubcaps were fixtured on the mill and clocked so the X0 Y0 reference was at its centre and tool point touching the surface at Z0 and the program was started and ran with success. The detail of the model was fully preserved and the surface finish was great.
The actual bronze hubcaps were fixtured on the mill and clocked so the X0 Y0 reference was at its centre and tool point touching the surface at Z0 and the program was started and ran with success. The detail of the model was fully preserved and the surface finish was great.