Author Topic: Cumulative error in YZ  (Read 869 times)

0 Members and 1 Guest are viewing this topic.

Offline sandcrab

  • Active Member
  • Posts: 62
    • View Profile
Cumulative error in YZ
« on: February 04, 2018, 12:46:24 PM »
I cast a pair of rotors for a 1/3 scale blower and used Alibre to draw a sample of a half a rotor. I downloaded a copy of Freemill and generated the G code and was able to cut a sample in ABS. I machined both directions, XY and YZ since there is no roughing cut program in Freemill and milled a pair of rotors to test. I mounted a casting and ran on the XY plane and ended up with the correct  profile, but when I loaded the YZ program the tool starts correctly by skimming the surface, but then lowers slightly in Z with each subsequent pass and starts to machine a tapered shape. The DRO's show the same readout as the G code program, but if I stop and return to  zero the Z axis I get a reading that represents the amount the tool plunged beyond what was programmed and what the DRO indicated.
Any thoughts?
I re-tuned all three axis and they were accurate.


« Last Edit: February 04, 2018, 12:50:00 PM by sandcrab »
If you think you can't do it, you're right.

Offline sandcrab

  • Active Member
  • Posts: 62
    • View Profile
Re: Cumulative error in YZ
« Reply #1 on: February 04, 2018, 01:41:49 PM »
These are the two programs. I changed the G17 to G19 but ended up with the same machining error.
« Last Edit: February 04, 2018, 01:44:59 PM by sandcrab »
If you think you can't do it, you're right.

Offline KLJ

  • Active Member
  • Posts: 10
    • View Profile
Re: Cumulative error in YZ
« Reply #2 on: February 25, 2018, 04:55:28 PM »
My first thought is whether the temperature of the ABS work piece rose during machining.  The coefficient of thermal expansion could possibly introduce such an error.  Do you use a spray coolant at a constant temperature?

I cut out parts from Teflon and just recently had a customer contact me claiming that the Teflon piece did not fit the steel carrier.  I told him the ambient temperature in which I had cut the part.  During shipment I believe the package was exposed to sub-zero temperatures and the part simply had not been given time to stabilize at room temperature.

I hope this helps.

Offline sandcrab

  • Active Member
  • Posts: 62
    • View Profile
Re: Cumulative error in YZ
« Reply #3 on: February 25, 2018, 08:49:58 PM »
KLJ, thanks for the reply. The error showed up with cast aluminum rotors. I think I have already melted them, but if not I'll take a photo. Some comments have been made about using the shift key to speed travel and creating runaway action, but I don't know if that is the same issue that I experienced. I have another sample to cut with a new profile and I will see if I get the same results.
If you think you can't do it, you're right.

Offline KLJ

  • Active Member
  • Posts: 10
    • View Profile
Re: Cumulative error in YZ
« Reply #4 on: February 25, 2018, 10:17:37 PM »
What are you using the acronym "ABS" to designate?  Acrylonitrile butadiene styrene can be used as a fugitive pattern material in a precision lost wax casting process (I used to do this).

Offline sandcrab

  • Active Member
  • Posts: 62
    • View Profile
Re: Cumulative error in YZ
« Reply #5 on: February 26, 2018, 12:07:12 AM »
I didn't have any machineable wax to test the code for profiling the rotors, so I just grabbed a block of ABS. As I mentioned there was no problem with the ABS samples, it showed up when I was machining the aluminum castings.
If you think you can't do it, you're right.

Offline KLJ

  • Active Member
  • Posts: 10
    • View Profile
Re: Cumulative error in YZ
« Reply #6 on: February 26, 2018, 06:21:10 AM »
Ok, I understand what you were saying better now.

One of the problems with using large items made from a relatively strong polymer versus a wax as a burn-out material in the refractory is that the polymer will expand as it heats and burns.  With weak refractory this would cause cracking but with stronger material this MIGHT have resulted in a permanent expansion/distortion of the mold which the cast aluminum would then have assumed.

What I am trying to say is that you should be careful in attributing the problem during machining to coding problems alone.  Precision processes rely on each step being controlled and predictable -- try repeating the experiment with wax.

Offline KLJ

  • Active Member
  • Posts: 10
    • View Profile
Re: Cumulative error in YZ
« Reply #7 on: February 26, 2018, 06:47:29 AM »
It was too late for me to edit the previous message, so here is an addendum:

EDIT: I just remembered one experience with a blueprint that was made for me by a customer in Australia on poster board type material.  When he rolled the material for shipment in a tube the differential in radii from one side to the other of the material caused a permanent distortion in the drawing (it was shortened linearly).  If I had scanned that drawing and used it without referring to and double-checking the longitudinal dimensions the error would have cumulative and I might be chasing after errors/wear in the CNC drive.

Alternatively, I also learned that expensive commercial scanners (like those used to copy blueprints) needed correction factors after the machine introduced repeatable linear errors during the feed process into the scanning bed/strip.  It was very expensive for me in terms of time and materials to discover all this.