We have an rma that had a bushing that was over reamed causing the x carriage to rotate about the Y thus causing less than par prints
There have been several more incidents of parts being over broached on the production floor. This is one of the reasons we try to only broach if the parts need it and do drag tests on the bushing parts in cluster.
I went over the broaching process with my weekend crew this weekend and with our day crew on last Thursday, so we should see less of this problem. Did we also not start a process just recently where they would be broached on the line, so that the assembly was with the part?
@david.hall yes you are correct, the broaching on the assembly line is only temporary due to the interference in the x double bearing holder (T6933)
once the ECO is complete there should be no need to broach on the line anymore
The broaching on the line is being done by replacing the lower x rod with a rough rod and moving the carriage back and forth, so it is broaching parallel to the axis of travel,
when broaching by hand, there is no guarantee that you are broaching along the axis of travel. I think it is still possible this problem could arise with the assembly broaching method, but i think it is much more likely that it happened when broaching by hand
Im inspecting a machine for high x backlash measurements that @franklin brought over.
I noticed that the xcarriage had a lot of rotational play about the y axis, Upon removing the lower igus bushings, I could see that these bushings were extremely over broached.
The bushing slides on a 12mm rod, and igus has a maximum internal diameter spec at 12.04mm
Prior to swapping the bushings the measured x backlash was about .41mm. This was giving shifted circular edges, like so:
I installed fresh 12mm bushings, with no broaching and the axis was much tighter as this x carriage printed part is prior to T6933, and has the bushing/part interference. Now the x backlash measured .96mm. I then replaced the lower x rod with a rough rod and ran the carriage back and forth ~5-6 times by homing the x, turning motors off, and pushing the carriage to the right (pressing in the middle of the carriage where the belt attaches), then homing x again and repeating. This broached the bushings along the axis of travel. I reinstalled the x lower rod, and now the x backlash measures 0.26mm.
I then re ran the calibration print and got true circles.
Based on this it appears that broaching these by hand is causing excess play in the bushings on the x carriage, that excess play introduces rotational play along the y axis of the x carriage. This then causes high x backlash measurements and will over compensate when printing circular edges or changing direction in X.
I feel that no broaching should be done by and should only be done with the rough rod installed in the machine such that the broaching happens along the axis of travel.
@tutley , I agree with the broaching understanding that you measured out. Can we get the part number of the part you are testing, so that I know if we are dealing with just one part (x carriage). I want to make sure that this is the only part being broached currently. "I feel that no broaching should be done by and should only be done with the rough rod installed in the machine such that the broaching happens along the axis of travel."
We will need to see the volume of bushings requiring broaching and then see how this can be done for current issues.
@bowman , Rasy, got with me this morning and asked about why parts can't be broached ( en masse). That his work station is getting so slow due to having to do all parts with the method downstairs. I wanted you to know that we will see a slow down in production now due to the method needed to get the parts broached correctly. We need to get the testing phase of this process done quickly, as we have already tested the new parts for drag. We are going to slow final assembly down considerably now.
I have also seen an increase in issues with backlash today, it is definitely slowing things down for me as well. We are broaching them on the machine, as is Rasy, but we are still seeing problems.
@MikeR do we have any more of the PP-GP0392 that were redesigned on the change orders that we can test on these trouble machines at all?
The others were installed on lago machines and installed fine... but they weren't having this problem before so we can't tell if it really fixed the problem or not?
If we don't have and @david.hall can we get some?