Is it good practice to constrain cylindrical parts?

We usally have a fair amount of cylindrical parts in our assemblies.
We usually dont fully constrain cylindrical parts, and let them "spin".
Our VAR said this will speed up rebuild times, as well as design time
since you dont have to do an extra mate. My problem with this is, I
look at my design tree at the end of a design to see if everything is
fully constrained. All cylindrical parts still have a - sign next to
it, so I have to check them to make sure they are defined except for
spining.

Is there a better way to do this? It can be rather time consuming
checking all cylindrical parts to make sure they are indeed only
spining about the axis.

Suggestions/comments welcome. Thanks.

 

I pretty much agree with everything you said. I too leave bolts, etc. free
to spin for the same reasons. I also see the - signs, but have come to the
point where I am consistent enough in my procedures that I pretty well know
what's tied down enough, and what isn't. Once in a while I go pull on
something to check it, but for the most part, I can recognize what still
needs attention.

WT

 

i agree that it saves a little time but if you have to look at some one
elses stuff, or you theirs, trouble shoot a glowing red tree it makes it
easier to muddle thru the mess if stuff is constrained completely. i almost
always throw a // on the planes just to tie it down. i mean u r talking a
split second to do it. if a few // mates is going to break ure system you
need a bigger puter! my .02 ;0)

 

If it is that time consuming you could create a macro to parse the mates and
make sure all cylindrical parts have a coincident and concentric mate. Or
atleast that all components have atleast 2 unsupressed mates.

Corey

 

Just had a mental picture. Not even sure this is viable or desirable, but
what would it take to have a program that "shakes" the model to see if
anything is loose? You could start at the top of the assy, go through each
part and see if it can move. If so, see if the only motion is rotational.
Then, some way to highlight those that go places other than the allowed
cylindrical rotations.

WT

 

good point... i only wish i was that freakin smart......drooooool..... he
hehe......droooool......
<kidding>

 

=) I don't think it has anything to do with smart. Simply LAZY. I hate to
do things over and over when I know I can automate it. Work the computer to
death because that is what you paid for.

Corey

 

Oh it should be viable I don't know off hand how you would determine
cylindrical rotation It must be possible though. You would have to make
sure that you move it back though.

Corey

 

true dat. i fall into the catagory that i know i could 'Learn" how to do it
but i can git er dun as fast and i need it right now! i'll figure out the
better way later....only later never comes!

 

This reminds me of the Degree of Freedom symbols in MDT.

 

Another aspect of this that no one has mentioned yet is that sometimes open
degrees of freedom can mess up dynamic assembly motion. Say you have a
robotic arm with loose screws in it. Moving it around unconstrained may
be too much to solve.

By the same token, if you do constrain the screws rotationally, but
constrain them to parts that the move relative to, then you can be causing
another whole set of problems. Personally, I don't constrain them. Under
normal circumstances I think it's overkill. I only constrain them if I'm
having problems with dynamic assembly motion.

matt

 

What I do is to insert my first part into the assembly, on the
origin and make sure that it is 'fixed'. When each additional
component is added to the assembly, they are either mated
to the first fixed part (usually a center plane), or to another
part of the assembly.

Fasteners and other cyl. parts left spinning.

Now, as long as there are no mates to the assembly itself
(e.g. planes), then all you have to do in 'un-fix' the first
component and drag and rotate the whole assembly around.

You'll be able to verify any mating problems easily. You can
hit undo to bring you back.

If I'm not mistaken, it is a good rule of thumb not to mate
anything to the assemby planes. Maybe there are exceptions
to the rule?

Mike Wilson

 

Hi.

I have always said that I want all part in assemblies to be fully
mated because then yo know what you are working with. And the rotation
mate should be a parallel to one of the three planes. That cannot take
much time to solve.

I think you can use the gravity simulation to check if a part is not
mated correctly.

Have a nice weekend

Klaus