SW: File Size (DRW, ASM) and File Corruption 1

[SDO]

Hi everyone,

1. FILE SIZE: Can anybody tell me if these file sizes are normal?

Assembly:
# of parts ~ 175
Complexity of parts: low to average (size 0.1 – 2.5 MB; screw threads as circular, tilted revolved cuts, screw file size about 2.2MB, say 50 screws total)
Size of the SLDASM file: 18MB!
Drawing (assembly drawing):
# of sheets: 3
# of views: 1 assembly view + 1 assembly exploded + BOM
Size of the DRW: 33MB !!!
(my comment would be, if this is normal for this kind of complexity, what would be the file size of assembly drawings with 1000 parts?! How can you work with that?)
Any drawbacks in using Unfrag or Ecosqueeze?

My system:
SW2005 SP0
System:
DELL Laptop - Intel Pentium M Processor 1.80 GHz
1.00 GB RAM
Nvidia Quadro FX Go700 Video Card
Microsoft Windows XP Service Pack 1

2. File CORRUPTION: Is there any way to recover corrupted DRW’s? Is there a way to prevent or decrease the chances for a file (DRW, ASM, PRT) to get corrupted?

Thanks
Sorin

 

[Heckler]

SDO,

Here is a link that might be of some help thread559-116740 You might want to consider upgrading Windows XP to SP2.0.

Best Regards,

Heckler
Sr. Mechanical Engineer
SW2005 SP 3.1 & Pro/E 2001
Dell Precision 370
P4 3.6 GHz, 1GB RAM
XP Pro SP2.0
NIVIDA Quadro FX 1400
o
_`\(,_
(_)/ (_)

"There is no trouble so great or grave that cannot be much diminished by a nice cup of tea" Bernard-Paul Heroux

 

[MadMango]

I think your file size issue comes from the use of actual threaded screws in the model. If these are helix-based features, the file sizes will be large.

Do you need the detail of the screw pitch? If these are standard pitches, you will gain some performance (& file size decrease) from using cosemtic threads.

[green]"I think there is a world market for maybe five computers."[/green]
Thomas Watson, chairman of IBM, 1943.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[SDO]

Heckler: I will upgrade to SP2.0 by tomorrow; it's worth a try, thanks.
Mad Mango: the threads are not helixes, but circular, tilted cuts, but I agree, a lot more surfaces to process. The reason I used them is lazyness: I had downloaded 3D models of these parts, and I thought what the heck, they are already made, they look good, why not. I'll make a test: I will back up the project, then delete all the hardware that's not with cosmetic threads and see how the size changes. Will let you know, thanks.

Sorin

 

[SDO]

MadMango: I guess the test was conclusive enough: the ASM got down to 4.5MB, the DRW down to 7MB. Heck, I initially hoped that not using helixes but circular cuts would make a big difference, but apparently not. It seems that what counts most is the number of surfaces to process, much more than the shape and equation of the surface. You get a star, thanks.

Thanks to you too Heckler, I think updating the OS is still a good idea and something that I should do.

All the best gentlemen.
Sorin

 

[MadMango]

You can probably keep the "fake threads" in your model, you might just have to approach them in a different way.

1) Were the tilted cuts a linear pattern of individual cuts? If individual cuts, try converting them to Linear Pattern.

2) Try drawing the profile of the threads in the base sketch and doing a Revolve-Boss feature.

3) If you don't need the screws to be parametric, you could export the models as a parasolid, then re-open it and save them as a "dumb" .sldprt files. This will remove all the features in the Feature Tree, but will produce smaller file sizes as well.

[green]"I think there is a world market for maybe five computers."[/green]
Thomas Watson, chairman of IBM, 1943.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[SDO]

MadMango,

Here is again one of those situations where a conclusion was based on a false assumption.
My test and measurements after taking the screws out and the "Eureka" conclusion, was based on the false assumption that the reduction in file size was due to the fact that the screws were detailed with the thread cuts.
I modified the threads and took out the detailed threads, but left the screws in => Some size reduction but not too much. I then took out even the hex slot on the screw heads => not even close to what it was without the screws. I also tested using Unfrag. Here is the situation:

Total # of parts 162
Total # of screws 79
Size ASM param screws 18.0 MB
Size ASM param screws after Unfrag 9.3 MB
Size ASM simplified screws 18.0 MB
Size ASM simplified screws after Unfrag 4.8 MB
Size ASM no screws 4.5 MB
Size DRW param screws 33.0 MB
Size DRW param screws after Unfrag 16.0 MB
Size DRW simplified screws 7.6 MB
Size DRW no screws 7.1 MB
Size DRW no screws after Unfrag 7.1 MB
Size DRW simple screws 7.6 MB

By the way, I also updated to XP SP2.0
Again, do you know any downsides of using Unfrag? What exactly is it taking out (history of commands, features, backup data?). What would be the information you loose that you might still need?

Sorin

 

[SDO]

I forgot to answer your question:

The screws were originally made with a tilted Revolved cut that was then pattern linear.

I'll try the parasolid option too, thanks.

Sorin

 

[SDO]

MadMango,

Here is the result of the last tests:

Replaced screws by Parasolids => 35% reduction in size of the screw files
ASM with Parsolid screws 18.5 MB
DRW w/ Parasolid screws 3 2.6 MB
ASM w/ Parasolids after Unfrag 8.9 MB
ASM w/ Parasolids after Unfrag opened and saved again 17.9 MB
DRW w/ Parasolids after Unfrag 16.0 MB
DRW w/ Parasolids after Unfrag open and saved again 32.0 MB

The numbers appear to say:

1. Using the simplified screws (no threads no hex slot) didn't help much for the ASM but it helped for the DRW (maybe I should have opened and saved the DRW again to see if it changes back to the original size or close)
2. Using the Parasolid screws didn't help
3. Using Unfrag didn't help, unless you don't "touch" the files; just opening and saving will return them to the original size.

Am I missing something on these conclusions?
Sorin

 

[CorBlimeyLimey]

"What exactly is it taking out"

Unfrag hunts down compound files and defragments them. These files are created by Microsoft Word, Microsoft Excel, and other applications that support OLE (Object Linking and Embedding). The files often become fragmented and thus take up an inordinate amount of hard disk space.

Making the best use of this Forum. faq559-716
How to get answers to your SW questions. faq559-1091
Helpful SW websites every user should be aware of. faq559-520

 

[SDO]

Hi CorBlimeyLimey,

Thanks for the info. It sounds pretty harmless, doesn't it, if it's just about defragmenting? I wonder then, how come that companies like SW and Microsoft don't know how to save the files already defragmented right from the start! Well, I will not pretend that I totally understand this. Personally I haven't used Unfrag that much, only tested it a little bit lately, and I didn't notice any downsides, but this didn't mean anything yet.
Thanks again,
Sorin

 

[CorBlimeyLimey]

The downside would be, if a file gets corrupted due to using Unfrag, neither SW or MS will offer support to fix the problem. You are on your own.


Making the best use of this Forum. faq559-716
How to get answers to your SW questions. faq559-1091
Helpful SW websites every user should be aware of. faq559-520

 

[SDO]

I guess backup is then a must.
What do you think about the fact that after Unfragging, if you open and then just save without doing anything else, returns the file to the original size?
- Did I do the test wrong?
- Is Unfrag only designed for before you just store your project, you're not supposed to work with it anymore?

Sorin

 

[MadMango]

That is just how MicroSoft handles OLE objects, it's not a SW error, and you are not doing anything wrong.

As for worrying about file corruption:
thread559-97903
thread559-87953

[green]"I think there is a world market for maybe five computers."[/green]
Thomas Watson, chairman of IBM, 1943.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[SDO]

Gentlemen,

Based on the tests on file sizes presented above: any pointers on how I can reduce the size of my ASM and DRW?

For the number of parts there, it just doesn't seem right!

Sorin

 

[SBaugh]

Save your assembly out as a part then put it into a drawing.

Save a Detached your Drawing.


it may not seem right, but if you have read through and searched out all the threads on this issue, you will see that there isn't much in the way of you doing much about it, simply because MS is to blame for most of the reason SW files are so big.

But if you have a number of configurations, you image quality is set high, you save Edrawings data in your files, these things will cause an increase in file size. Minimizing isn't any easier though. You could delete all the Configs and the file size might drop some but not alot. Because MS is the main culprit.

Reducing size is going to depend on a lot of things.
How do you want to open your files? - If you want to open up the drawing and see the assembly, you will not be able to suppress the assembly to cut file size, or build blocks around each of the parts to cut file size down.

Just some thoughts,

Scott Baugh, CSWP
3DVision Technologies

http://www.3dvisiontech.com

http://www.scottjbaugh.com

faq731-376

 

[SDO]

Scott,

I have very few configurations, my Image quality is right in the middle of the range, I don't save eDRW data.
Saving the assembly as a part could work for some situations, but apparently I can't do cross sections and of course, I can't do "exploded". In this case, I can't use it.

Here is a cute one:
I save the 18MB assembly as a part => 6.6 MB file
I save the same assembly in exploded statet as a part => 26MB (and of course, it collapsed it back)
I opened the file above (26MB) and saved it again => 52 MB !!!

Ok, MS is the culprit, and there isn't much you can do.
I'm just asking (since my working conditions do not exceed the average): what do you do when you have a 1000 parts assembly, working with the average conditions, under MS Win? I know this is being done, so somehow it is possible, and still being able to work at a reasonable speed.

And, if I follow the same prescription, I should be able to work a lot easier when I have a simple assembly like mine now. I must be doing something wrong, cause people are still using MS XP and work with a lot mor components.

Thanks for the input Scott.

Sorin