• PT Youtube
  • PT Facebook
  • PT Linkedin
  • PT Twitter
4/25/2011 | 2 MINUTE READ

The Practical Processor

Facebook Share Icon LinkedIn Share Icon Twitter Share Icon Share by EMail icon Print Icon

Your processing questions answered.

Q: Some of our molders say they can’t accept a surface-finish spec using the traditional Roughness Average (RA) because they can’t measure it or meet it. We have found SPI mold-finish ratings in SPI standard AQ-103, but this seems to be an inspection spec. Is there a reason RA surface finish can’t be applied to, or measured on, a molded part? If not, is there some standard that can measured numerically instead of the AQ-103 standard, which is only pass/fail? I am concerned about sealing a rubber O-ring. The RA surface finish is what I know works, based on Parker O-ring guidelines. Is the SPI rating appropriate for sealing surfaces?  

 —Electronics OEM

A1:Typically RMA or RA is used to measure surface finish of cut metal. I see no reason why it could not be measured on a plastic part. The SPI finish is what is used for measurement of plastic parts. I would expect more variation due to the variability in the molding process, versus the controlled process of machining and polishing. I went looking online for a few minutes and this is what I found:

SPI    SPE        Method                                    RMS (µ) inch
 A-1    #1         Grade #3 Diamond Buff                0.5-1.0
 A-2                 Grade #6  Diamond Buff
 A-3    #2         Grade #15 Diamond Buff              1.0-2.0
 B-1                600 Grit Paper
 B-2                400 Grit Paper
 B-3    #3        320 Grit Paper                               7.0-7.5
 C-1                600 Stone
 C-2                400 Stone
 C-3                320 Stone
  —    #4         280 Stone                                    12.0-15.0
 D-1                Dry Blast Glass Bead #11
 D-2    #5        Dry Blast #240 Oxide                  26.0-32.0
 D-3    #6        Dry Blast #24 Oxide                    160.0-190.0

Granted, you would need to convert RMS to RA. Can I vouch for total accuracy of the above data? No, I cannot. The SPI specs correlate to RMS readings in the chart. In my opinion, choose a RMS value you want, find the corresponding SPI finish then spec the mold to that finish. The molded part should match the finish of the mold. Exceptions would be knit lines, improper venting, or if the part is not packed out correctly.

Tim Atwood, technology manager
MSI Mold Builders Cedar Rapids, IA
(319) 848-7001


A2: Measuring textured plastic means taking a sample across a defined flat section of the surface and averaging the peaks and valleys. Unfortunately, few plastic pieces with tight tolerances have a flat expanse of an inch to evaluate. If you measure via contact, you are lightly scraping a probe/needle along the surface and risk digging into soft materials. If you measure optically, variations in surface texture affect the accuracy of the readings. There are some metrology technologies claimed to handle these tasks, but the investment is heavy. Check out: zemetrics.com; kla-tencor.com

John Berg, director of marketing,
MGS Mfg. Group, Germantown, WI
(262) 255-5790 • mgstech.com