Tips for Testing KUKA Callibration

Hey there! Our new KUKA robot is currently being installed and configured as we speak. Before the team of engineers that are setting it up leave, I would love some feedback and any suggestions on how we can ensure it is being setup, configured, and calibrated properly. We want to make sure that everything is being calibrated perfectly, before the engineers wrap up installation :smile:

I'm hoping someone might have some suggestions, sample files and tests we could run to be able to quickly confirm everything is setup and dialed in perfectly. Any tips on what kind of things to be on the lookout for? Known issues, common problems, etc... Here are some of the questions we have:
- What are the best ways to ensure the 7 axis robot, liner rail, rotary table, and parts change is setup properly?
- Are there any test files that we can use to help us easily test and validate the calibration?
- Are there any other software or tools we should be using?
- What things should we be looking out for in order to ensure the calibration is setup perfectly? (Repeatability, Accuracy, etc...)
- What would be some great projects to test each of the areas?
- What else are we not thinking of?

Robot Cell:
- KUKA KR-150 r3100 on a
- Tools: Mill, Hot Wire, (Extruder, Chainsaw, Plasma Cutter, Coming Soon)
- Wireless Touch Sensor
- Robot sits on a 40ft Linear Rail
- 48" Rotary Table
- ATI Quick Connect Head and Plates
- 10 Part Tool Garage Changer
- Laser Part Setter
- Laser Fence

Software:
- ZBrushGrasshopper
- PowerMill
- Grasshopper
- ArtCam

Thanks again!

Any tests you should do, other than a basic repeatability test will be specific to YOUR particular application. You should really not be asking us those things, those specifications should have already been established long before installation (actually, before purchase).

Some basic things with the 7th axis is to tell it to go to say 10,000 mm and make sure it is actually at that position when it finishes. Other than that, robots are pretty well calibrated when they come off the pallet.

You could do a bar bell test in multiple orientations.

From the questions you have asked up to this point, I think you are greatly under estimating the complexity of integrating the software into functioning programs.

You have a very good robot however your application has multiple tools that are dramatically different in operation. It is hard enough to program and successfully run parts with 5 axis using a gantry type machine operating in a Cartesian coordinate system in a production environment.

Your system and the robot in particular will be very challenging. You are trying to do this with a traditional robot manipulator arm system, functioning in a polar coordinate system, making the velocity and positioning calculations very complex and doing this with a spectrum of attachments.

Most applications that utilize arm robots such as yours are in a production environment so that the huge front end resource charges can be amortized over many parts. From the sound of your description up to this point, it sounds like you are planning on doing one of type projects.

This will be extremely challenging and complex. So much so that you will be mostly on your own in developing the system once the installation team leaves.

I would suggest that you start with simple two dimensional plane type objects with some 3 dimensional features to get your feet wet and then duplicate these in various axis orientations. This will allow you you to verify the integrity of axis rotations and how that changes the operational characteristics of your robotic cell.

OP was ThinkTank Group

machtool said:

OP was ThinkTank Group

Click to expand...

That's an obscure post, even for you.