|
TO DRAW OR NOT TO DRAW
I remember when
I first got into CAD. It was 3D wireframe. After I worked for
awhile I said "boy, we will not need drawings any more". Now,
many companies have taken that path. Creating a drawing from a
model is very easy, even when complex it only takes a few
hours. We are now getting models with only the features that
need to be inspected defined. The rest are basically covered by
a profile tolerance. This is called PMI or Product Manufacturing
Information. Now this seems like a good way to cut design costs
and get your parts manufactured. But there are problems.
DESIGN REVIEW
Lately I have
done many drawings from models for a couple of my customers. I
have found that many times there is a problem with the design.
These 3D solid models are not being designed, they seemed to be
created with very little thought, by some CAD jockeys to
basically fit. As you start to detail these parts you can see
that there is not much rhyme or reason to the end result. When
you create a drawing from the model you can see some of the
design problems arise. Many of this parts are very complicated
to detail, most cause by not understanding design intent. Design
Intent, this was a term I was introduced to when we were
considering selling Pro/E. When you designed in Pro/E you
really had to know where you were going before you got there.
You had to know your datum's for sure. But with many systems
you can now design without any constraints. Much of the parts I
have worked with were basically designed with no understanding
of datum's. The drawing defines what is important and what is
not important. Now of course we do get some GD&T (Geometric
Dimensioning and Tolerancing) in a drawing usually as a PDF. We now
have two documents that have to travel together. Some systems
like Catia have this data included with the part, PMI, but very few
programs can read the model and this annotation so we may still have to
have a .pdf that goes with the model. There by creating the
same situation of making sure both pieces of data are reflecting
the latest revisions. Which leads to our next problem.
DATA VERIFICATION
Now that we
don't have a drawing that is the authorizing document, we seem
to have a problem keeping track of the data. When we had a
drawing we had something that was in stone. You could look and
say, "That has to be a .50 fillet" and when it wasn't we had a
problem. Now we have model verification process, that compares
two models. No where does it say that the models are correct,
just that they are the same. Of
course we have a problem keeping the drawing up to date. This
of course is a bit time consuming, but you can check it on paper
or check in at manufacturing after the part is made. Believe me
paper is cheaper. The drawing allows you to scrutinize the
design, allows more eyes on the design. As of now I am really
not sure the procedure for checking and releasing parts to
manufacturing. I am not in the design loop, but I am seeing
what is being delivered to the vendors. It is not pretty. Even
some of the machine shops are sending parts back, saying
that they can not be built. I can not imagine this happening in
the days I was on the board. I remember working on the Manned
Maneuvering Unit at Martin Marietta. I would take the drawings,
after they were checked by the checker for fit and function to
all the manufacturing people, materials and stress analysis for
review and sign off. At Boeing we would have design meetings
with the drawings on the table. After seeing some of the parts,
I can't imagine many of these parts being released. The problem
with drawings is that they are hard to keep up to date. So no
more drawings, but the problems created seems to be worse. So
now we need a CAD model verification program, that does not
assure correctness.
PURPOSE OF DRAWINGS TODAY!
So now we
need CAD software to VERIFY the models and see the differences.
This seems to be the unexpected consequence of eliminating
complete drawings. I have come to the conclusion the elimination
of drawings are having devastating effects on our engineering
process. So what is the purpose of the drawing in this day of
"solid" models?
1. The original
purpose "To inspect the final part."
2. Ease of
design review, anyone can get a pdf and review the part design
and function. To see how the assembly of the parts affect each
other. No need for CAD experience or access to CAD program or
viewer.
3. Absolute
authority for the part. A second check on the model.
4. A review by
the designer on any design functionality the may have been
overlooked. A chance to see if the design can be simplified or
optimized. I have had a manufacturer tell me that he has to
refuse some parts for lack of manufacturability. Even machine
shops have had to refuse some part because they could not be
built.
5. To teach new
drafters, designer and engineers how to design. They need to
have their designs scrutinized by an experienced checker, whose
purpose is to not only check fit, form and function, but to
impart his vast design and engineering knowledge and experience.
6. A history of
the design. Years ago we used the microfiche to review designs
of the past, today you have to have CAD software available to
access the data. I am not sure how we get to the data. With a
drawing you can recreate the part if necessary.
7. I am sure
there are a few more reasons, those above surely would save much
time and money as these parts are being delivered to assembly
and do not fit.
CONCLUSION
As I have said
before, to create a drawing from a model is very easy. Many of
the CAD programs will update the drawings automatically. I
suppose that is the problem, the CAD software really doesn't do
a very good job updating the dimensions. But even then the
problems caused by the lack of drawings, seems to me to worth
the effort.
|