The sketch on plane modelling method, producing a linear history tree was
the preference for earlier Master Series Releases. It is possibly are
quicker way to model if you're always moving forward, never returning to
features already created to modify or re-order them.
There were also some issues with the software in earlier releases, with it's
handling of migrating geometry. If there was a "Branch" of history tree
that needed to be removed but contained reference geometry then the features
could be deleted but the reference geometry would "migrate" to the next
lower node in the main stem. Once migrated, and with all relations removed,
they would then corrupt and be un-deleteable.
With these earlier issues fixed and the "real world" scenario being that the
part history is constantly replayed, edited and re-ordered, the Boolean
modelling method, producing a "bushy" tree became favoured (in general).
Although perhaps taking slightly longer to initially create geometry,
editing an area is much quicker since returning to an operation halfway down
the tree doesn't require all the lower operations to be replayed, just the
operations with it's branch.
With regards to reference features, these can be treated as the backbone of
the part. Properly constructed they allow the design intent of the component
to be captured and easily modified. It is often preferred to create
reference geometry for a new feature rather than sketching on an already
existing face, to avoid editing issues later if that face moves, or becomes
"out-of-date".
How best to use them is a big subject, you need to break the question down
to avoid an even longer answer.作者: Diego 时间: 2002-11-18 13:58
布尔运算在i-deas里的确是一大优势，它的历史树往往成一棵树状有许多分叉，与proe，sw不太一样。灵活性更强，更便于修改。
总是感觉i-deas太深奥，再加上参考资料太少，很难琢磨透
（ccf兄，可以把这段文字翻译成中文吗？俺是英盲……呵呵）作者: opengl 时间: 2002-11-18 16:10
i-deas的历史树确实好管理，又直观。::y::y::y作者: ccf 时间: 2002-11-18 22:34
Diego,
I am very slow typing chinese. It i time you learn English, good to your future.
