re: rE:Re:RE: programming by hand

ballendo

Thanks for answering this question for me, couldnt have said it better
myself! I would just like to add that I stated the encgineers woud manually
dimension the print with the correct tolerance. I tried to make the point
that I rarely ever receive a print, only the solid model to work from so the
MANUALLY entered dimensions and tolerances contained in the print are not
being conveyed by the model. Our ProE system is set up for a global
tolerance of plus minus 0.0050" so as ballendo points out, some of the parts
will not fall within the tolerance band (defined by the print) if cut from
the geometry (and tolerancing) contained within the solid model. If given
the print to work from, the tolerancing contained on the print would be
accommodated.


"Typically, they would draw something at 3 inches and then they would
manually dimension the print with a 3 inch plus 0 minus .010
tolerance While some will say that this is not a problem, it
is when you are machining the parts from the model files
sometimes weeks before drawing files are actually created."

While Im at it, i would also like to state that I see the engineers build
their models with with critical features refrenced from other features or
datums and then when they make the print, they will once again MANUALLY
dimension a feature with a tight tolerance refrenced from a different
feature or datum than the one refrenced in the model. While this is not a
problem if you are working from the drawing, it can be a real pain if you
are working from the model geometry. I see this in molded parts where we are
providing the mold maker with the geometry files and then giving him a print
to supplement the model files. When questioned about this practice, they
state the fact that they would have to re-construct their model from scratch
to have it represent the actual part they want and they just dont have time
to do that..... Ands upper level management wonders why we have so many
problems with incomming inspection rejecting parts, cost of rework, missed
delivery schedules, assembly problems etc. etc.

there is never enough time to do the job right, but there is always enough
time for re-work and re-design!

Bill Griffin


Message: 17
Date: Sun, 14 Jan 2001 00:02:10 -0000
From: ballendo@...
Subject: Re: rE:Re:RE: programming by hand

Smoke asked:

>It appears to me your saying a drawing for a part should not have any
>tolerance at all Bill.
>Is that what you mean to convey...or did I mis-read something there?

From an engineering standpoint there are very good reasons for using

>a nominal dimension and then adding a plus or minus (as required)
>tolerance on the part.

Smoke,

I'm not Bill, but I believe you misunderstood his point.

He is saying that using a 3 inch dimension (contained in the
electronic file of the drawing) as the BASIS for CUTTING the part
WILL result in bad parts IF the part tolerance (in the TEXT
DIMENSIONS of the CAD DWG) is not FIGURED INTO THE CUTTING PATH.

Using the numbers given; 3 inches +0/-.010 we have a tolerance range
of 2.990 to 3.000. If the CNC programmer USES the "drawn" line (which
is 3.000) to generate a toolpath, standard machining variations will
mean that a lot of parts will not meet tolerance (they'll most likely
be too big).

The 'generally considered' CORRECT way to do this tool path will
split the tolerance given, and program a 2.995 length. This way the
machining variations will fall on either side of the "programmed"
path, and MOST(if not all) parts will meet spec.

It is a VERY GOOD point to have made on this forum! I'm glad he
brought this up...

Hope this helps.

Ballendo