Page 26 of 29

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Fri Mar 01, 2019 8:11 pm
by GrblGru
I have now started programming the Carbide inserts and am making good progress. I hope that I can send you a test version soon. Then we can talk about the details.
You are right. There are still some hardcoded texts that don't appear in the text files. I will change that when I get time.
But I still hope that you can work better with the current Spanish texts now than with the old ones. We can still make corrections later on.

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Fri Mar 01, 2019 11:09 pm
by Leonardo
Thank you!
Fortunately there is always place for improvement... Otherwise the life would be too boring! :)

Have a nice weekend!

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Sat Mar 02, 2019 9:58 pm
by Leonardo
GrblGru wrote:I have now started programming the Carbide inserts and am making good progress.....
Hi GG,
I was investigating about the tool radius compensation and reviewing the GRBL supported G-Code commands I realized that G41 and G42 are not included.
GRBL will work always in G40 as default so, unfortunately, this firmware is not suitable for precision mechanizing with a lathe. Working in G40 you will always have an error equal to the radius of the tool when mechanizing cones, curves, fillets, etc. It will work well only facing and straight turning so, 90º movements only.

I am attaching a picture that explains what I am saying graphically. (The pictures were downloaded from this site (in Spanish): https://www.programacioncnc.es/g41-y-g42-en-torno/

I decided to leave my lathe as is by now because I would need other controller (capable to calculate tool radius compentsation) if I want to be able to mechanize small and precise pieces.

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Tue Mar 05, 2019 2:18 pm
by CrazyBillybob
Leonardo,
Just because Grbl doesn't support the G41 and G42 doesn't mean you can't get proper parts out of your lathe. It means that the cutter radius (diameter) etc can't be compensated in the controller. But if you calculate the compensated path in the design software before you generate the Gcode you can get the same end results using only G40 commands. I'm not sure if GrblGru does that yet or not, but I know software like Fusion360 has that capability. Also I know that GG has always looked at ways to make the software better... So if it doesn't yet, it might somewhere down the line.


CBB

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Tue Mar 05, 2019 6:39 pm
by Leonardo
Hi CBB, thank you for your comments!

You are right. I was playing with a demo version of SprutCAM last weekend and I was able to probe what you are saying.

This software provide many options for tool compensation and, once you select the one fit your requirements best, you can generate the tool path this way and then the G-Code.
I was able to see the generated paths graphically an they look correct. Unfortunately I have not a machine ready to make physical tests yet but it seems to me that this procedure will works fine.
As I commented before I am investigating and collecting all the information I may need BEFORE making any change in the my manual lathe. I just want to be absolutely sure to have all the things clear (software and hardware) as well as having access to them before starting with the conversion procedure.

Regarding the hardware part, I discovered something that I was not aware of like is the stick-slip phenomenon produced when you move a slide at slow speeds. This problem do not arise when you work with ball-screws and linear guide-ways but it happens with the traditional slides. I found a solution what is applying Turcite (what an expensive material!) to the carriage part of the slide... a lot of handwork because you must mating the surfaces using Engineer's blue, hand scraping and so on.

Well, I will continue gathering information! :)
Thank you again.

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Thu Mar 07, 2019 11:12 am
by GrblGru
Leonardo, you are right that the radius compensation provided by GRBL cannot be used.
But as CrazyBillybob already said, you can also calculate this externally. I have been calculating the radius compensation in GrblGru since the beginning.
By the way, I also do this when milling. There are many situations in which this makes sense because the CAM program knows the exact geometry of the tool. GRBL has no information about this.
You can see the effectiveness of the calculation very nicely in the video below. (from 12:11)
The calculated path during finishing ensures that the tool does not destroy the contour at any time.

@CrazyBillybob
Thanks for your help :)


Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Sun Mar 10, 2019 4:56 am
by Leonardo
Yes GG, that is the proper procedure. Having the tools paths compensated GRBL will will not present any problem.
Thank you!

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Sun Mar 10, 2019 5:23 pm
by GrblGru
Leonardo, I have now created a new beta version V3.35.1. It's certainly not bug-free yet, but I think you can use it to do a first test.
You can download it from my dropbox at https://bit.ly/2xCD7Nd_GrblGru_Beta.
I put also a short video on https://www.dropbox.com/s/ffbbsidojr3k0 ... e.mp4?dl=0 for you.

In the current state you can create Carbide inserts. All information like shape, thickness, radii etc. are taken from the ISO designation.
But I have already added some examples. If you want to create your own tools, is best done by copying an existing tool and then changing the ISO name.

Roughing and finishing cycles should already work. I kicked out the little dialog. Now you can only make entries in the Propsbox (bottom left).
I haven't done anything with the shaft yet. But I also wonder if you need it at all.

It would be great if you could test everything once.
Let me know if you still need something.

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Sun Mar 10, 2019 6:01 pm
by Leonardo
Hey GG, that is great!
Unfortunately I am changing my ISP so I will not have access to Internet for 3 days or so... I am using my phone now.
I will be back on this as soon as I have my connection working again.
Thank you very much for your efforts continuing with this development. Really much appreciated!

Re: GrblGru = Free 3D-Simulation for ShapeOko2

Posted: Mon Mar 18, 2019 6:02 pm
by Leonardo
Hi GG,
Finally I have my new Internet service working so I was able to download your beta version and to make some test.

I saw the video you make to show me the process (THANK YOU!) and found something that would like to comment about. It is also related with the need of what you call shaft (I guess that you are talking about the tool holder).
I would propose the following:
Tool holder: the bar to which you screw the carbide insert on.
Tool post: the lathe part that holds the entire tool (tool holder + carbide insert).

The carbide insert tool holders are fabricated for holding the insert in the best position (angle) to work with it following ISO normative (ISO turning in this case) so the tool holder is set perpendicular (standard position) to the lathe axis and the tool (carbide insert) stay in the position to which was designed for (I will call this "working position"). Generally only two side are facing the material and are in working position and the other two are use for exactly locate the insert in the tool holder and remain covered by the tool holder shape.

This angle (Kr) is standard and depend on the carbide insert geometry. It is 95º for the C type, 93º for the D type and so on. By the way, in the software you use the insert center line to set the tool clamping angle. I think that it would be better to use the side of the tool starting with the Kr standard angle as default to save the user all the angle calculations .

Getting back on the demo video you kindly made, I took a screen shot on a moment in which the toll is cutting with the back side so, actually it would be impossible with the carbide insert placed in the tool holder. That shows the need of contemplating the tool holder in the calculations/simulation process.

I also change the tool in your example and placed it in its standard position to show you that, in this way it will not be any collision with the tool holder.

I am attaching some pictures some pictures to show you what I am trying to explain.