Here's a simple but regular problem I come up against

I am modeling an injection molded plastic part. It could be a single
part that has a planar part line around the outside surface, or it
could be a master part with a split line around the outside, again,
planar. The part is drafted on one side of a datum plane to, say, 1
degree and the other side of the plane is drafted in the other
direction at a different angle, say 3 degrees.

Now if this part has any corners around the outside - say it's a box -
if I put a standard edge round of equal value on all the corners (on
both sides of the part line) then the part line is no longer planar.
This is because the edge rounds, where they intersect dip down in the
corners, because the angle of the adjacent surfaces is different on
different sides of the part line. It's not much, but it can be
noticable.

Anybody know what I'm talking about? What is your work around?
Sometimes I construct the rounds as surfaces, extend them and trim
them to the datum plane and try to merge them, the solidify. Most
recently I constructed one side of the box in surfaces and the other
in solid. They both ended on the part line datum plane. That way all
the edge rounds ran out on the plane. It's a complex way to deal with
a simple problem though.

 

I've had that same problem when I've put rounds on after I did the draft, but not if I rounded the corners first, then did the draft.

David Janes

 

The problem with that is the rounds become conical if you draft them
after. They look different and also machining the tool is more
complicated because the radius of the round changes as you move up the
edge.

 

The problem with that is the rounds become conical if you draft them
after. They look different and also machining the tool is more
complicated because the radius of the round changes as you move up the
edge.

______________

Some interesting problems you bring up, a couple thoughts:
1) The "lifting" you note at the bottom of the round on the drafted surfaces is a product of the "rolling ball" method of round creation; what about trying "normal to spine".
2) How noticeable is this conicality on a 1 degree draft? Obviously, the longer the drafted wall the greater the change from top to bottom, but still on a 1" wall, it's only .015 increase in radius. And the smaller the radius, the more noticeable the change. However, noticeable or not, why wouldn't most people have no or neutral reaction to the fact?
3) Re machining: the simplest way to produce this drafted wall is with a tapered end or ball end mill. Using, as a sweep trajectory, the top or bottom of the drafted wall, naturally produces a conical corner (larger at the bottom than top). If a plain BEM is used instead, then we are talking about very ordinary tool paths with dozens of ways of calcuating the tool path, some better optimized than others, but none that are at all stymied by a corner radius that's bigger at the bottom than top.
4) For alternative ways of producing tapered walls than the draft feature, consider this presentation, carried in Profiles Magazine, given by Dave Lowe at the 2002 PRO/USER conference:
http://www.profilesmagazine.com/p21/tips_low.html
It might actually have something in there that answers your concerns. However, you probably won't think any is as easy as your present extrude-draft-round method.

David Janes

 

The problem with that is the rounds become conical if you draft them
after. They look different and also machining the tool is more
complicated because the radius of the round changes as you move up the
edge.

______________

Some interesting problems you bring up, a couple thoughts:
1) The "lifting" you note at the bottom of the round on the drafted surfaces is a product of the "rolling ball" method of round creation; what about trying "normal to spine".
2) How noticeable is this conicality on a 1 degree draft? Obviously, the longer the drafted wall the greater the change from top to bottom, but still on a 1" wall, it's only .015 increase in radius. And the smaller the radius, the more noticeable the change. However, noticeable or not, why wouldn't most people have no or neutral reaction to the fact?
3) Re machining: the simplest way to produce this drafted wall is with a tapered end or ball end mill. Using, as a sweep trajectory, the top or bottom of the drafted wall, naturally produces a conical corner (larger at the bottom than top). If a plain BEM is used instead, then we are talking about very ordinary tool paths with dozens of ways of calcuating the tool path, some better optimized than others, but none that are at all stymied by a corner radius that's bigger at the bottom than top.
4) For alternative ways of producing tapered walls than the draft feature, consider this presentation, carried in Profiles Magazine, given by Dave Lowe at the 2002 PRO/USER conference:
http://www.profilesmagazine.com/p21/tips_low.html
It might actually have something in there that answers your concerns. However, you probably won't think any is as easy as your present extrude-draft-round method.

David Janes

 

The problem with that is the rounds become conical if you draft them
after. They look different and also machining the tool is more
complicated because the radius of the round changes as you move up the
edge.

______________

Some interesting problems you bring up, a couple thoughts:
1) The "lifting" you note at the bottom of the round on the drafted surfaces is a product of the "rolling ball" method of round creation; what about trying "normal to spine".
2) How noticeable is this conicality on a 1 degree draft? Obviously, the longer the drafted wall the greater the change from top to bottom, but still on a 1" wall, it's only .015 increase in radius. And the smaller the radius, the more noticeable the change. However, noticeable or not, why wouldn't most people have no or neutral reaction to the fact?
3) Re machining: the simplest way to produce this drafted wall is with a tapered end or ball end mill. Using, as a sweep trajectory, the top or bottom of the drafted wall, naturally produces a conical corner (larger at the bottom than top). If a plain BEM is used instead, then we are talking about very ordinary tool paths with dozens of ways of calcuating the tool path, some better optimized than others, but none that are at all stymied by a corner radius that's bigger at the bottom than top.
4) For alternative ways of producing tapered walls than the draft feature, consider this presentation, carried in Profiles Magazine, given by Dave Lowe at the 2002 PRO/USER conference:
http://www.profilesmagazine.com/p21/tips_low.html
It might actually have something in there that answers your concerns. However, you probably won't think any is as easy as your present extrude-draft-round method.

David Janes

 

The problem with that is the rounds become conical if you draft them
after. They look different and also machining the tool is more
complicated because the radius of the round changes as you move up the
edge.

______________

Some interesting problems you bring up, a couple thoughts:
1) The "lifting" you note at the bottom of the round on the drafted surfaces is a product of the "rolling ball" method of round creation; what about trying "normal to spine".
2) How noticeable is this conicality on a 1 degree draft? Obviously, the longer the drafted wall the greater the change from top to bottom, but still on a 1" wall, it's only .015 increase in radius. And the smaller the radius, the more noticeable the change. However, noticeable or not, why wouldn't most people have no or neutral reaction to the fact?
3) Re machining: the simplest way to produce this drafted wall is with a tapered end or ball end mill. Using, as a sweep trajectory, the top or bottom of the drafted wall, naturally produces a conical corner (larger at the bottom than top). If a plain BEM is used instead, then we are talking about very ordinary tool paths with dozens of ways of calcuating the tool path, some better optimized than others, but none that are at all stymied by a corner radius that's bigger at the bottom than top.
4) For alternative ways of producing tapered walls than the draft feature, consider this presentation, carried in Profiles Magazine, given by Dave Lowe at the 2002 PRO/USER conference:
http://www.profilesmagazine.com/p21/tips_low.html
It might actually have something in there that answers your concerns. However, you probably won't think any is as easy as your present extrude-draft-round method.

David Janes

 

The problem with that is the rounds become conical if you draft them
after. They look different and also machining the tool is more
complicated because the radius of the round changes as you move up the
edge.

______________

Some interesting problems you bring up, a couple thoughts:
1) The "lifting" you note at the bottom of the round on the drafted surfaces is a product of the "rolling ball" method of round creation; what about trying "normal to spine".
2) How noticeable is this conicality on a 1 degree draft? Obviously, the longer the drafted wall the greater the change from top to bottom, but still on a 1" wall, it's only .015 increase in radius. And the smaller the radius, the more noticeable the change. However, noticeable or not, why wouldn't most people have no or neutral reaction to the fact?
3) Re machining: the simplest way to produce this drafted wall is with a tapered end or ball end mill. Using, as a sweep trajectory, the top or bottom of the drafted wall, naturally produces a conical corner (larger at the bottom than top). If a plain BEM is used instead, then we are talking about very ordinary tool paths with dozens of ways of calcuating the tool path, some better optimized than others, but none that are at all stymied by a corner radius that's bigger at the bottom than top.
4) For alternative ways of producing tapered walls than the draft feature, consider this presentation, carried in Profiles Magazine, given by Dave Lowe at the 2002 PRO/USER conference:
http://www.profilesmagazine.com/p21/tips_low.html
It might actually have something in there that answers your concerns. However, you probably won't think any is as easy as your present extrude-draft-round method.

David Janes

 

1) You know, I've never used "normal to spine" for a round. I've never
understoof what it meant. I might try it and see what happens.

2) Not very noticeable, but if the draft is greater it is. The example
I gave was a simple one, but at other times I have done things that
were more complicated, using surfaces rather than the draft command.
Of course in that case I can't put the round on before the surafces
are created. In short, the appearance of the part is important
because I work for an industrial design firm and the difference
between a conical and parallel draft can be a big deal to the
designer.

3) I guess you know more than me about this - I just know I've had
toolmakers and model makers tell me they prefer the round to be
constant along the edge. Maybe if the adjacent surfaces are not
actually flat but made from surfaces the situation is different?

4) I will look at this - thanks. You know, you didn't have to post 3
times...