He encontrado en una pagina llena de modelos CAD gratuitos una geometría 3D que me llamo mucho la atención Una cabeza humana (una mujer) bastante genérica y en formato STP, listo para poderlo abrir con programas paramétricos, como CATIA, SolidWorks, etc… Un buen ejemplo de lo que seria capaz de hacer, encima con la dificultad añadida de tener que voltear la pieza y seguir el mecanizado por la parte de atrás. Técnicamente algo de dificultad, pero me gusta ir superando limites con esta maquina.
I found a page full of free CAD 3D models, one of them really struck me in STP format, human head (a woman) ready with parametric order to open programs such as CATIA, SolidWorks, etc … a good example of what would be able to do over with the added difficulty of having to flip the piece and follow the machining the back. Technically some difficulty, but I like to be surpassing limits with this machine.
Similar to the previous entry, the propeller, but then discovered some shortcomings such kind of flipped the part I did not like me. To begin, I will not make the programming of the two phases in the same G program, I’ll employ my work to have a stock geometrically well done (which is not complicated, it is a hub of parallel sides) and with two programs I’ll get the origin each time. We will see how it goes.
In this first part , I will make the front of the head, which is the same process as the other parts, a roughing and then finishing with spherical tools. I’m close to controll feed rates , and depths of machining, depending on the cutter diameter and type of material. For ureol, that’s what I’m using, and two lips tool Ø8mm as roughing, I’ll machining with 700mm/min feed rate and 5mm depth. With all this, I’ll need about 40 min . only for this phase.
The surface to leave will have an «offset» 2mm with respect to the final surface , so I will make a rough with a spheric tool Ø10mm, and then make a finishing pass with a Ø4.5mm spherical tool, with a path every 0.5 mm. The more close this distance, more definition, now you will see the results:
Step 1: Stock brackets :
This operation is the most elaborate , and «cumbersome» , mainly because for every piece have to improvise the brackets: Yes, I still need more universal system to tie pieces, but do not have time . So for every time I have to think about how tie it, if they are oriented along the axes of the machine, and is capable of maintaining such alignment when I rotate it. In addition , of course, has to be robust .
This is what I came up with :
Thus, it’s easy to flip, then the head left in horizontal position, and aligned with the longitudinal axis of the screws. However the accuracy is not important as I think back to find the origin again, and supposedly I can trust measures of the limit sensor.
Step 2: Start roughing:
Nothing, more of the same, I will program the operation and catia, which is very clever, makes me all the paths based on my conditions:
Mental note, I must to find a better solution to fix the aspirator clamping, So i could orientate the tube, if not all room will be full of dust…
This is the final result of the first roughing phase:
It’s starting to look like a face, isn’t it? the truth is it is great to see it working.
Step 3: Close roughing:
Change to spherical cutter 10mm, and we continue:
Well, well, I like how it looks …
Step 4: Finish:
Changing to the spheric tool to 4.5mm and let’s go, this phase will take half an hour. Between one thing and another, I’ve made my motors work hard, especially the Y axis, it gets warmer, but runs smoothly. However, I’m not going to have such long continuous phases …
At this stage, it’s very clear the good definition of the machine, and the finish surface that is able to show, overall for how cheap it was and materials has been done, obviously. To leave the surface smooth, I would have to sand a little to remove the rows, or it could have had paths every 0.2 mm, and 3 hours machining.. but of course, it’s good like this …
Here is the final result:
The result has left me pleasantly surprised. The good definition and accuracy that can be achieved, the truth is that they are very good.
Of course, there are errors that can be solved, such as trademarks of the cheeks, which was my fault but easily solvable. The fact is that there are areas where the tool can get off at Z a lot, and the tool was inserted too much into the machine at a distance too short of support, so when trying to get down to the base, the nut touch the model, marking the workpiece. Had taken the tool 5mm more of his position, IT would not have this problem, now I know for next time …
When I have the second part, you will see the final result.