In my opinion, stick with 2001 Plus, the last good working version.
We have had a painfull upgrade to 2004, it also cost about 1000 dollars per
workstation to upgrade video and ram to cope with the 2004 version, in the
meantime we hit penalty clauses in the contract we were just starting due to
the consistent crashing and inability to cope with large assemblies.
Just one persons view, others have had no problems, depends on whether you
just work in simple details/weldments or need large assy capability.
We are still not satisfied with its performance.
We discovered that resolving large models in a large assy caused it to
crash, and then found that the model we had tried to resolve was now three
times the size.
Keep backups handy.
I haven't used 2001 Plus in a long time and can't answer your comparison
test. However I can tell you that contrary to some users experience my
large assembly mode experience with 2004 sp0 was good.
My machine was a Dell 360 w/ 2.4 GHz P4, 3.00 GB RAM, XP PRO SP1, and
(QUADRO NVS W/AGP8X W/64 MEG RAM) game card.
My large assembly is:
TOTAL COMPONENTS: 15285
It loaded in the following times:
Lightweight 56 SEC
Resolving from the LW load 88 SEC
To load fully resolved 3:13 MINUTES
I hope this helps.
Does this assembly have any errors, in-context parts, flexible
sub-assemblies, sub-assemblies with multiple configurations, mates with
invalid faces or over defined mates? Also are there many undermated parts?
Tom Chasteen wrote:
I no longer have convenient access to that assembly, but have the following:
Total components 13801
total parts 11382
unique parts 120
unique subs 86
suppressed comp 172
Lightweight comp 13629
top level mates 24
top level comp 8
max depth 6
AMD 1.8 Ghz processor w/ 512 meg ram, Quadro4 750 xgl
To load lightweight = 58 seconds
to fully resolve after lightweight load 92 seconds
The assembly has NO errors. The assembly has many in context parts,
features, and component arrays. Most parts are fully mated. Nothing is
fixed. There are no mirrored parts in the assembly anywhere. There are
still a few mates with invalid faces in the assembly. These occur when I've
changed something and not caught up the mates. I delete them whenever I
find them and remate to component in the lowest assembly possible. There
are no over defined mates in the assembly. (However, there are ten mates
that I feel should show as over defined and they do not. they are
intentional. I added a point-line coincident mate to assist SW in being
able to solve an already fully defined mate. SW needed the assistance.)
There are still some underdefined mates. I'm fixing most of them as I get
to it. This is a different assembly from the previously mentioned assembly
and I'm running it on a different computer. The assembly size is 6.4 meg
after ecosqueeze (I do not remove the display list or the preview image).
The assembly includes structural steel, fasteners, motors, gearboxes,
cotterpins, chain, etc., etc.
I had 2000+ fasteners this includes bolts, washers , nuts, and cap screws.
All fasteners were copies of the toolbox part and all unnecessary
configurations were deleted. Nuts are drilled out to prevent interference
with the bolts. I have sheetmetal self drilling self tapping (SDST screws),
they were down loaded from McMaster-Carr. I also have square beveled
washers. One of the down loaded parts came in as a 2 meg file. I remade
the part and it shrank by 6+ times.
There are no flexible sub-assemblies
When you say the parts count seems quite low for assembly that size, I take
it that you are complimenting my design skills!
My assembly now is:
Total components: 14547
total parts: 11964
unique parts: 120
Unique Subs: 86
Lt Wt Comp 14375
Top level mates 24
Top level comp 8
Max Depth 6
Time to load Lt Wt 37 sec
Time to resolve 100 sec
Time to load resolved 174 sec
I changed a couple of features and added some more assemblies
It appears that changing a couple of features (it affected 282 parts) has
reduced the Load LW time.
I really don't understand all of the variables. But, I can change the
assembly width by 1 foot and its height by 6 inches and the width change
will add over 550 parts to the assembly (required by new width) and the
height change will adjust the dimension and change length on over 50 parts.
The assembly still handles (O.K.)
I have sheet metal parts with six flanges and numerous slots and cutouts.
There must be something that I'm doing or not doing that greatly affects the
assembly resolve times. If we can figure out some of these variables, maybe
SW can fix the ones that are really slowing down the process.
Getting apples to apples comparisions isn't necessarily easy; however,
how many users do you think have your system's resources or better?
I think if a survey were conducted, your hardware would leave a large
percentage of computers running SolidWorks in the dust...
Unfortunately, programmers may assume that most users have the latest
& fastest, but the reality is far different.
Per O. Hoel
That's why I'm trying to send as much info as possible. Some people are
finding that SW 2004 is a real DOG on large assemblies. I've found that it
outperforms any other version I've used. Maybe I'm just getting better at
it?? I pay a lot more attention to how I set them up than I did before.
I wouldn't say that I'm as good or better at large assemblies than anyone
else, so there must be some differences in how we set them up, parts design,
or something. I would like to find out what it is, because I don't see
assembly sizes decreasing in the near future.
It's been along time since I too have ran 2001+ but I know that when I
switched form 01+ to 03 I did't notice any slow downs. I wasn't doing
large assemblies then, just advanced injection molded parts, but was
often working w/ 15+Mb parts. Now I am working in 04 and have noticed
a slight drop in speed, but it is more than made up for w/ the
stability it has brought alone. I have two other guys here working on
04 one of which is happy w/ it other than some minor bugs (and it cut
his crashes from about 10/day down to 1-2/day) and one guy who hates
it w/ a passion. I believe most of his problems are w/ currupt parts
though, but I ain't about to argue that point w/ him cause it's a
waste of my time.
Just my $.02 but I know that if I were trying to accomplish the things
I am now w/ '01+ that I would be having a much more difficult time
doing it. It's all a judgement call though.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.