Navisworks Federated Models with native DWG's vs. already cached NWC's in terms of file-size

[Shane R]

I wanted to pose the question to see if anyone has seen a huge difference in the overall Federated Model linking the native DWG's versus linking the already converted/cached NWC's in terms of file size? I manage Navisworks Federated Models with hundreds of CAD Models incorporated and had conversations with some of the Modelers that believe that Appending the NWC's files vs the DWG's will bring the file size down and make the NWD distributed easier to work with. While I see how they came up with this thought, as in the Windows Explorer the NWC is much smaller in file size than the native DWG, I thought Navisworks already reads those Cached files regardless of what's directly linked? It would be a huge hassle have to pre-Cache all the DWG's to NWC's prior to opening the Federated Model NWF, where as now with the DWG's loaded the NWF updates the NWC's automatically so I'd like to keep the DWG's linked. I imagine "Refreshing" would be quicker though, but I'm not sure.

Does anyone have any experience on this? Is there a big file size difference between Appending in the DWG converted NWC's vs. the native DWG's? I tried a small scale experiment which resulted in a very minor difference in file size, which would not be worth the extra headache of using NWC's over the native files.

Thanks in advance!

[dgorsman]

No, there shouldn't be a difference. There would be a difference in load time, since there wouldn't be a need to update/refresh any NWCs to match the latest DWG content - they would always be current. But as you've noted with the appropriate Options setting those are auto-generated, and any time saved would be clawed back by needing to generate NWCs.

Refresh time gets tricky, depending on whether you're actively using XREFs or loading all DWGs individually. Because appearances of XREF dependent objects are determined by the host, if you refresh using XREFs to load models *everything* must be XREF'd. If you toggle off XREF loading and load the various DWGs individually, NWCs for each of those files are managed separately and a refresh only targets the ones that are out of date.

As for overall management, you might want to switch up organization a bit. I'm assuming with the size of your project there are multiple "areas" and/or "sub-areas" involved. Federate those individually or in smaller relationally logical groups rather than going straight for the entire project. Then auto-publish those areas together into the overall project model. For the most part, reviewing one "area" won't be influenced by most of the other areas so having them loaded only slows things down. Save the overall project model for the big picture, top-down reviews.

[Shane R]

I appreciate the reply back and it seems for the most part your thought process is similar to mine, that the NWC's vs DWG's would produce roughly the same file-size of a Federated Model NWD. And the problem is not so much the Refresh time there, but instead minimizing the file-size of the NWD that's published regularly for the group. But yes using the NWC's over the DWG's could end up causing a lot of headache for the BIM Manager while the end-user file-size may not even be different than just using the DWG's directly.

[dgorsman]

Reducing the overall NWD is down to a couple of the DWG settings e.g. skipping irrelevant content (not loading frozen/off layers); setting a good faceting factor (can be relatively low if you don't have a lot of curved surfaces); not importing materials or saved views from the models (materials in DWGs are almost as bad as annotation scales for inheritance); merging XREF layers so you don't have redundant content (gotta watch for copy-in-place XREFs in the original models too - I once stripped out *5* copies of the same XREF out of a host file...).

Appearance overrides on saved Navisworks views can bulk up a file quickly as well.