Very Large File Sizes

Hello there,

I'm working on a file which seems to be growing exponentially in size
every time I add a new feature. Its currently about 700mb, and just
total crap to work with. I have a decent duo core laptop and graphics
card, yet it's still taking 15 minutes to save the file!

Any idea on what kinds of features or rebuild errors may be causing
this, and how I may reduce the size. SP3.0?

Thanks,

Ian H

 

It's hard to say exactly what features or such may be causing this (if it
is) without knowing a little more info. Is this a part or assy? The reason
I ask is I had someone at my last place have an assy file that he modeled a
ton of screws with exact helix sweeps and threads and such. (he was new).

But the first thing, try "save as" to your file and give it a different
name, then check the file size. If you have a lot of configurations that
control things like sweeps, wraps and/or lofts, etc, the size will creep up
on you after load different configurations, but I'm not sure about
exponentially. I've had 70mb files reduce to 15mb after a save as.

Scott

 

You might look at your feature statistics to see if anything jumps out at
you, such as some feature that is taking a whole lot of time, or grows
faster than other features.

WT

 

I have had file sizes shoot up when inserting parts into parts, but
that was somewhat predictable. Unfortunately sometimes even with lean
modeling methods your files can become unwieldy. A few times when I
cant figure out exactly what is causing the problem I save out a
parasolid and work from there. Either that or try to roll it back to
the last place that it was predictable and start from there again with
even leaner modelling methods if possible.

As an aside have had decent luck in reducing file size by suppressing
portions of the feature tree, and saving. I empathize with your pain.
Our Engg dept recently started using Solidworks for certain projects.
Coming from UG they were aghast at the file sizes that Solidworks
generated.

 

You can always use feature statistics to give a hint at where the work
is being done.

Horizontal modeling might also help.

TOP

 

Back when I first saw SW in 1995/96 I was so very enthused by the fact
that it was faster than anything else we had (I had a Pentium 166 with
300mb ram), had extremely small files compared to Anvil or Pro-E and
hardly ever crashed.

Back in the 3Amigo days I was running a 900Mhz machine with 1Gb and
this was cutting edge. But still a 100Mb model was extremely difficult
to work with. Most other things were acceptable.

If I used those same criteria today, the decision would be to look
somewhere else.

You have to wonder how SW can make the statement that they have
improved performance when that P166 couldn't even load a blank part
template today.

TOP

 

If I had a part file ballooning for no reason and a Save As didn't
reduce file size, I'ld take the Save As file and start deleting key
Features or groups of Features and Save As again until I saw the file
size drop back to "Normal".

I'ld to the same with an assembly by deleting part or subassemblies
until "normal" file size returned and then try to figure out what is
going on.

The only thing I ever created which absolutely throttled SolidWorks to
a GLACIAL PACE was straight cylindrical knurling (or double helical
diamond knurling which was even worse), and Diamond Knurl on flat
surfaces. IT ABSOLUTELY KILLED the ability to do darned near anything
in SolidWorks. To show a knurl, I left a very tiny segment of the
desired knurl as a detailed form and the file operations were
acceptable.

Bo