Is there a proper way to develop a 3D model from drafted 2D data?

* I realize this is a basic question, but I really have never been shown how.

Woah! That's a BIG question!! :eek:

Standards. There are standards (which you should stick to- otherwise your fellow engineers won't have a clue what you are talking about!).

Then there are projected views- usually 3rd Angle. Put your cup on the table, remove the perspective, then look straight on at it. Then look at it from above- 3rd angle. If you look at it from above, then tip it over, then look at it (from the same position), that's 1st Angle. If you get my drift?

Then there are measurements, which ideally should be in metric- but America (and a little of England too, I have to admit, but we are getting much better), are still stuck in a time warp! ;)

Books. You need to read books. Books are good.

Shark will allow you to do this (convert from 3D to 2D), but it is hard work (I know I shouldn't say that). Programs like Solidworks are far more refined. However, the new 2D tools in SFX7 do look really promising. So there is hope yet!

TTRW: Good advice. i do know how to draft by hand and i do know how to model in 3D (but not at expert level). The problem is that I am new to 2D CAD itself and want to know if there's a way to easily (in Shark) take my orthographic projections and create a 3D model DIRECTLY from it (e.g. using front elevation, end view and plan.)

Agreed about the metric system.. I can't think of any other country besides the US that uses miles/feet/inches still.. I didn't know that the UK still did..?

STEVE M: Could be imported .dxf file, but also one created within Shark. What's an easy way to draft and then quickly model from that data a 3D image? I hope my question isn't too wordy and makes sense.. I'm trying to explain it as best as possible.. :)

I find that using a 2D drawing as only a source of dimensions and drawing directly in 3D is much quicker than trying to wrestle a 2D drawing into a 3D one.

I've recently bought viaCAD. My (limited) previous CAD experience has been 2D only. I'd also like to learn how to take a three-view 2D drawing and make it into a 3D. My ultimate objective is to trace the top, front and side views of PDF views of an experimental airplane I built, then adjust each view so they are the same scale and locate common points on each of the views in the hopes of getting a rough 3D approximation of the actual airplane.

Because of all the compound curves on the airplane and my lack of any experience in 3D modeling, I can't believe it would be easier to model the airplane in 3D first (but if that's the case, then I'm willing to take a run at it).

Thanks.

Bob

----

If you are inclined to experiment with boat or ship hulls (for modeling in 3d practice, but not structural designs... that would be really ambitious...), visit www.delftship.net and go to the download secion. You may need to log in in an account in order to download user-submitted models.

Just grab a simple one, like a yach, or even better, a supertanker (it won't have canopies, rudders, sails etc.

Then, turn on stations, buttocks, waterlines, making sure you have added stations at some random or convenient-for-you distances based on the size of the model in its real-world units. For example, if the hull is 400 feet long, maybe place stations at 40 foot intervals. Again, this is just to speed things up. Add waterlines, too, every 9 or 10 feet to simulate deck-to-overhead (floor to ceiling) spacing.

Now, making sure that the network of points is turned on, export the model in DXF.

In ViaCAD, import that model and expand the layers tree. You can for now turn off waterlines and buttocks. If you want, turn of "lines" as well, leaving just the stations in view. Note that the stations will have port/starboard (left/right) halves.


switch the model to a front view and notice the curves of the hull. Imagine trying to "fair" those lines by hand, without hydrostatics feedback to tell you whether the hull shape was efficient or not at a given waterline.

Now, use the skin tool and select a forward and an aft station. Doing this on port and starboard, you end up with sideshell or hull plating at this location. Now, skin the port and starboard station lines for forward and for aft boundaries of the sideshell surface you created. Also, if you turn on the lines layer and "break" or segment the lines intersecting with/at the stations, you have accurate boundaries for the deck plating or top boundary. You could skin from surface to surface, too, using a right and a left top edge of surfaces.

Now, you have a (if done correctly) a "watertight" section of a boat or ship hull -- well, until you cut holes for pipes or ladderways, or if you made straight-on skin covers.

The hard part to surface is near the bow, especially if a sonar dome is on the model. For even MORE fun (or therapy, in my case), I draw/extrude along the curved hull a number of sideshell stiffeners (T-beams) from top to the keel. I have to extend lines the depth/distance of the web and sweep that with a careful selection of the sweep/extrude features. I also have to do this for other structural elements that form a grid of minor steel to support the flat or larger plates. It goes on, too, because i'll have to draw pipes, extruded or swept from circles, along a path. I'll have to draw inclined latters.

This is why i like ViaCAD. I could not do this in version 14 of TurboCAD nearly as easily, and TC14 didn't offer easy surface/solids. VC 2D/3D and VCP both do. I gave up kick-butt icon pallets and whiz-bang buttons once i found VC 2D/3D giving me far more than i otherwise had. Then, I upgraded to Pro.

Say, Zumer,

I really like your explanation. I think it's probably the best, concise presentation of the matter.

As i got ready to look at your attachment, thoughts came to mind. I thought, "it would be nice if ViaCAD had semi-parametric facilities to handle:

-- volume, area, product weight constraints

-- stability

Let's say you make a mug. We've all seen many of those that have tall, narrow bases with diameters sized to be held comfortably but which flare to hold more volume at the upper portion. To me, those are poor and unstable. I prefer, wider-based, in-tapering mugs with leak-proof/minimally-leaking lids. Of course, the ones with taller, narrow bases fit in vehicles' cup holders and such.

If the weight, moment, and density of the liquid such as coffee (a milk shake would be less a leak problem) are accounted for, it might be possible for a good seal (rubber?) could be designed into the lid to hold it tight to the mug inner surface.


I should have looked at Shark's parametric features more closely before the beta timed out.

Zumer I agree with ZeroLengthCurve good, concise answer.

So is it protocol to physically overlay 2D CAD data (orthographic views) and create a 3D model directly from those lines? I know this sounds rather lame, but I just want to know if experienced industrial designers, etc do this, or do they model directly by looking at hand-sketches and sort of "sculpt" from there?

2D data is useful for dimensions and relationships, and you get that by glancing at a hand-drawn sketch or print. When the features and relationships are established, you connect them together with surfaces, blends, or chamfers, or integrate the features into a shaped enclosure. Sculptors work to an overall concept refined with details. Designing with CAD, you create details and the relationships between them, then fit an envelope around them, and/or design structures that connect them. Form follows function. Functional items are usually composed of primitive shapes, blocks, spheres/ellipses and conics, so the fastest workflow is achieved by being able to recognise which a shape is derived from. A 2D drawing only describes a shape from one viewpoint. Move away from that viewpoint by swinging your model around in space and it doesn't look like the 3D shape it was meant to convey any more, ie change the viewpoint and the shape isn't "overlaid" anymore. 2D shapes are conceptual guides, not templates.