For it's free-form production work Solid Shapes™ is looking for a 3D milling machine. On one side there are the traditional 5-axis CNC machines on the other side there are the upcoming CNC robot solutions with 6 or more axes. But what is the right choice for our work? We know that the traditional CNC machines are a proven concept, they have years of proven production work behind them. There are also some disadvantages that come with these machines. For instance they have a fixed size if you want to go bigger you have to buy a new machine. The object you are milling most likely lays flat because of the limited hight the machine can reach. This can be challenging when you want to cut the object from all sides, you’ll have to turn the object over. A traditional machine can’t undercut like a robot can. A robot has a much more flexible arm that can reach a lot more places compared to a traditional machine. On the other hand flexibility of the arm also gives the robot a less rigid and solid structure which resolves in a less accurate end results compared to a traditional CNC machine. But what kind of accuracy does the end product need? Today robots can reach a point and path repeatability of + and - 0,5mm which for most jobs is sufficient. Besides scalability and flexibility a robot is a real jack-of-all-trades, on a robot-head you can fix a lot of different tools and functions besides milling you can let it pick and place, measure, cut, drill and even let it sand and paint the workpiece. Basically a robot can handle the whole workpiece from start to finish. Finally the price of a robot set-up is cheaper than a comparable traditional CNC set-up especially when you take in consideration that you can grow your robot set-up with your business growth. So this are a lot of favores that speak up for the robot set-up.
0-1 for the Robot.
To find out if the CNC robot is the best set-up for Solid Shapes™ we have put the two to the test. We defied the test in two parts. One cutting a three dimensional shape out a solid piece of PU block (940) - how well do the machines work with 3D objects and how well can they finish the product and how much time does it take - and one cutting out shapes in Solid Surface material and fit two parts together like a puzzle - how precise does the machine work and how fast can it cut the material -.
First up was the robot set-up, we hooked up with a robot integrator from the south of the Netherlands. NPD™ modeled a three-dimensional ace like shape with a logo from our friends at Rustic Products™. The IGES file of this model was put through a CAM program and send to the robot. The robot took off with a good start unfortunately when the robot was almost done with roughing one side of the ace we found out that the glue that hold two parts of PU-block together did not hold. Because the model was so well fixed at the bottom the parts stayed together and we could continue the test. It took the robot quite some time to rough the first side so we decided to finish this side first. After a small day milling the one side was finished and it wasn’t as smooth as a I hoped for - this off-course also has to do with the mill-strategy, the tolerance, the step-down, the tool etc. you choose -. More worrying was the fact that everywhere the robot changed direction it left a little bump. By leaving these bumps at turns, bubbles where created at direction points on the model. This was quite a disappointment cause the bubbles where to obviously and could not easily be sand away. The material we used to mill the ace is a hard and unforgivable material and showed us the weakness of the robot, the bumps where probably left because the robot not being ridged enough and having a little movement when moving from direction.
Test two; nesting out two parts of solid surface material and fitting those together. We started of with fixating the solid plate. Pretty soon after the robot started milling we could smell it wasn’t going as planed. The robot didn’t cut through the material, it was actually more melting into the plate - this explained the smell -. Despite advice from the supplier the robot integrator choose the wrong tool - a ball nosed tool - for this kind of job. We tried a second time with a different tool but this also didn’t give us the result we hoped for. When the plate moved we where done for the day.
On to Italy. Prior to flying to Italy we sent our hopefully better glued second block PU to the CNC factory in Italy. Late in the evening I arrived in Milan the next morning we went to the factory. We started of with the same 3D ace model as we did with the robot. Much to my surprise around four hours later at lunch time the ace was almost milled. Coming back from a good Italian lunch with the factory workers the model was perfectly smooth finished. I was stunned by the time difference and smoothness of the final piece.
1 - 1 CNC versus Robot.
Even more surprising was the speed at which the CNC machine cut through the solid surface material. With the blink of an eye it was done. When we tried to fit the two parts together it was so exact we almost couldn’t. With some more pressure we managed and the line between the two parts was as good as invisible.
2 - 1 for the CNC.
With a good feeling of being sure to choose the right machine for our company I flew back to the Netherlands. The guys from the factory promised to send the ace to Solid Shapes™ by post.
Although in theory the Robot has clearly some advantages in practice the traditional CNC proved to be the beter choice for us. An other important item that came out of our test is the fact that (most) robot integrators have no background and history in milling and therefore lack the knowledge to suport and advice it’s customers on milling issues.
Although we will choose for a traditional CNC-setup I’ll keep a close eye on the developments on the robot CNC side. I feel that in the future when things get optimized robots will take a significant share of the CNC market especially in the softer materials.
