Quick Poll: Am I Crazy???

[jeanchile]

Hello all,

We use Frame Generator a lot to create frames for our equipment supports as well as access platforms, etc. Most of the time these platforms are relatively small (20'-0" square or less, usually). I am working on a design that has multiple pieces of equipment and the access platforms are all going to be tied together into one larger frame.

I created the skeleton idea for what we need using sketches and I was thinking I would use FG for the whole thing. The sketch looks like this:

I occurred to me after I got this far (it's not complete yet but it's close) that I may be completely nuts.

Has anyone ever done anything this big in Frame Generator before?

We need to create AISC/NISD compliant fabrication drawings of each piece and we usually use some other company's software to do this but I HATE that program and I want to get rid of it completely. My ability to do that depends on whether IV can handle something this large and remain stable. Any thoughts, workflows, etc?

[jeanchile]

Sorry Chris, I've been out of town a lot lately and am just catching up on these emails.

In answer to your questions...

1.) That model was actually done using IV 2010 so there was no functionality difference between our programs at that point. I've included some connection screenshots and stuff anyway. On this project we utilized a single shear plate connection for pretty much everything (the reactions were quite low). The connections we use with Inventor are not anything to brag about. They are a simple table driven iPart that we manually place into demoted FG sub-assemblies. Sometime we use the skeleton file dimensions linked to each girder so the connections move when the skeleton file moves but most of the time we don't even bother. By the time we get to the point of adding the connection material the design doesn't change very much and we have documented procedures on how to revise something so things like manually moving the connection material doesn't get missed.

2.) The way we create the drawings/fabrication details is no "holy-crap" kind of thing either but we have created a pretty streamlined process by using the programmable keys on our 3D mice. Each "fabrication" drawing we do has the same annotation on it so we "run through" the 10 buttons on our mice to detail out a piece and it's gotten pretty damned efficient. When we're doing fabrication drawings of structural members (e.g. simple beams and columns) we are spitting our drawings on an average of about 1 every 6 minutes. Our buttons are programmed like this:

   1.) Place View (we use a standard 3 view projection)

   2.) Center Mark (for hole centerlines that didn't show up using automatic centerlines)

   3.) Centerline Bisector (for beam and column centerlines)

   4.) Parts List Placement

   5.) Dimension (for all standard dimensions, we use ordinate dimension tool for "running" dimensions)

   6.) Balloon Tool

   7.) Welding Symbols

   8.) Hole and Thread Annotation

   9.) Leader Text (for notes, "no paint" information, etc.)

   10.) Custom routine that collapses the browser tree, prints a PDF with the revision info, saves the file, then closes it.

All we have to do is create a new drawing and run through the buttons and we end up with drawings that have everything they need. This makes training a new hire VERY simple when it comes to drawings. We've also reduced the amount of shop questions (e.g. What's this weld supposed to be? I don't have a piece mark for this piece. What's this hole size? etc.) by over 56%! It's pretty rare when we forget to put something on a fabrication drawing and it's all because of those 3D mice. There have been some concessions when it comes to things like AISC/NISD audits and fabricator standards because IV is incapable of creating drawings according to AISC published standards without a TON of extra work. We just got our fabricators/auditors to overlook the stuff because we showed them how our program couldn't do it.

3.) We don't create a "line" diagram for our drawings anymore. We tried several ways of doing this without any success and said screw it, it's time to adapt. Now our design process uses a skeleton file and FG to create the initial layout of the steel. We populate the members using FG based on company standards and then check that with other software or we use the output of the analysis software to populate the members. We just put the "sticks" in the model and don't do anything with connections and stuff until much later in the process after most of the design is finalized.

As far as your idea goes, I wish you luck but we have never figured out a way to have sketch geometry contain any "intelligence". We tested a way of using a customized round bar CC family that had beam size iproperties. We used that CC family in our prelim design phase and we inserted rods that showed up on drawings as lines but when ballooned they said (W8x18). We ended up abandoning that process and just using the WF family instead. Our platforms aren't usually that big and putting a view on the drawing that didn't show hidden lines was enough for us.

Hope this helps. Good luck!

[jeanchile]

More pics

[jeanchile]

One more.

[cadman777]

Jean,

Thanx for the excellent summary w/comments.

You've helped me to re-assess, choose a definate direction, and gauge where I'm currently at w/this process.

Much obliged for sharing your hard-won experience.

And thanx for the excellent pics!

Cheers