几何尺寸与公差论坛

 找回密码
 注册
查看: 678|回复: 0

【转帖】primary and secondary datums of size

[复制链接]
发表于 2009-4-29 21:24:55 | 显示全部楼层 |阅读模式
primary and secondary datums of size
in the book design dimensioning and tolerancing by bruce a. wilson under number of datum references it say:
if the datum features on a part are flat features, then it is necessary to reference three datums to establish the three planes in datum reference frame. ..........
if one of the datumn features is a feature of size, then it isn't always neecessary to reference three datums to estables the three planes in the the datumn reference frame.
the book then gives an example of a part that has a pilot shaft that is identified as datum b. then it goes on saying:
referencing datum axis b as the primary datum establishes two intersecting planes at the axis. referencing a second datumn feature, which is perpendicular to the axis, establishes the third plane in the datum reference frame.
this is in chapter 5 page 128 on 1992 edition of the book. now my questions:
1. does it matter if the axis b is referenced as the primary or secondary datum?
2. the two plane established by axis b are free to rotate around b. don't need another datum to prevent the rotation?
thank you for your response.
find a job or post a job opening
well, my answers are
1. it depends on what you are trying to control.  typically, an axis is an "axis of symmetry" for the part, and establishing it as a primary datum is "natural" for the inspection of the symmetric features.  think of a lathe-turned part, and how simple it is to inspect runout, concentricity, etc. with the part "chucked" in a lathe on the primary datum, or turning on centers that establish the datum.  but, if the axis is the centerline of a bolt hole  in a flat part, you would probably want to make it a secondary or tertiary datum for a positional tolerance.  when establishing what to use as datums, you should think about a) what are you (the designer) trying to control with a tolerance callout, and b) how would you (as an inspector) want to try and measure that tolerance, and c) it never hurts to ask (a colleague, another inspector, the guy on the shop floor) how they might do it.
2.  yes, but not always necessary if there are no features that must have "clocking" controlled.
good response btrueblood - deserves a star.
您需要登录后才可以回帖 登录 | 注册

本版积分规则

QQ|Archiver|小黑屋|几何尺寸与公差论坛

GMT+8, 2024-12-23 01:17 , Processed in 0.034733 second(s), 19 queries .

Powered by Discuz! X3.4 Licensed

© 2001-2023 Discuz! Team.

快速回复 返回顶部 返回列表