Autodesk Inventor

[#1]

JD,

You said:

"I would use 2D sketches to create and control (might not even use 3D sketch at all for that)."

Would you care to explain how you'd do that?

[#2]

I can't help but feel that I must be missing something?  (see attached)

(this could have been done all 2D sketches if I had created a couple more workplanes)

Even when I have complex bends I use 2D sketch to control the 3D sketches.

http://home.pct.edu/~jmather/content/DSG322/Inventor%20Tutorials/Inventor%2011%20Tutorial%207.pdf

[#3]

JD,

Thanks for the pdf. I learned something in it: You don't have to extrude surfaces to intersect the sketches.

Your attachment was too new for my program. I have IV2010. So I couldn't view it.

My biggest problem is that projected entites from sketches break when I resize the model.

The best work-around I've come up with is to use work points at vertices and other critical location points.

That way, when the model changes, the points update, and they maintain their projected statuts as both derived entities and as part of parent ipt file.

The other biggest problem I have is w/frame generator.
Assembly features always break (e.g., holes & cuts) when the underlying wireframe changes dimensionally.
I learned to use frame generator only when I'm certain that the parts won't get modified with assembly features.

Otherwise, I'm asking for a load of work when the customer changes things.

I wonder how others have dealt with these 2 kinds of defects in the software?

[#4]

---

cadman777 wrote:

JD,

Thanks for the pdf. I learned something in it: You don't have to extrude surfaces to intersect the sketches.

Your attachment was too new for my program. I have IV2010.

---

I am aware that it is no longer necessary to extrude the surfaces - those tutorials were written years ago and I haven't had time to update.  In my class we do the problem differently.

Too bad you couldn't open my 2013 file - you are doing wayyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy  (if you didn't notice I added more yyyyyyyyyyyyy than my usual response) too much work.  Your problem is really fairy trivial as far as I can tell.  I think there is an easier more robust way.

 All dimensions are in Sketch1 & 2.

[#5]

JD,

Thanks for the screen shot.

If I understand your screen shot correctly, then it looks like you did it the way I usually do it.

Here's how I interpret your screen shot:

The green lines are in Sketches 3 & 4 as "projected" lines, and the red lines are in the 3d sketch connecting to the 2d end points.

If that's true, then here's what I normally encounter:

The green lines will "break" (= disconnect associativity) from their parent sketch entities.

I use the parent sketch to make planar "patches" and "thicken" them into separate "sheetmetal" parts (= derive the parent sketch into individual sheetmetal files). When I change the parent sketch, the patch breaks b/c the projected entities break inside the parent sketch (= the derived ipt).

So, the project I'm currently working on provided an opportunity to try a new method using mostly 3d sketches for the main shell, in an attempt to keep the sketches from breaking.

I used projected points derived from work-points, instead of projecting sketch entities (which break). It seems to work better than my "normal" method, even with 2d sketches and work-plane placement.

The caveat to this method is, the wireframe ends up being cluttered with work-points (and work-planes, in the case of using 3d sketches).

One more notable:

I began with the top plane sketch, per my original (unlike yours), b/c in "real life", that "flange" (heel outline) doesn't change. In other words, I built the model like we build the part in real life.

The way I make the wireframes is by placing the "knowns" into sketches, and working around them, which is why it seems so complicated. If I don't do it that way, I get very confused when I'm "deep" in a model. It's a very simple way to keep things mentally-categorized.

My "method" may not represent "good modeling practice" from a software training perspective, but it is the most efficient way I've discovered, due to "job-site reality".

Sometimes, like in this case, we end up adding "realities" into the wireframe after the model is all finished. That happened in this case 2x. One time the model totally changed, which you'll see in the attached sketch. Those little changes (i.e., adding an additional "flange" and changing from a sloped "skirt" to vertical) cost the customer 8 hours of modeling & drg scrubbing time. Had I done my model like I normally do (which is same in principle to how you did yours), I would have wasted an extra day on it.

Customer changes present quite a challenge when doing my projects. I have to "anticipate" those changes and build them into my original wireframe, which is what makes me push for unbreakable projected sketch entities.

So, I'm just trying to find a way to keep the wireframe projected entities from breaking when things change, since that's a normal part of my project-work.

Having said that, let me ask you this:

Do your green lines break associativity when you change the dims on sketchs 1 & 2, after you have your model all built in the iam file?

If not, then if you were me, what method would you use to prevent my wireframe/model from breaking?

Find attached an example similar to what I normally would do to make this hopper. Please note that it's a quick revision of the latest "evolution" in my original 3d wireframe, so it's very cluttered, due to many customer changes. Also, I omitted the myriad of details, such as manways door, flg profiles, &tc.

[#6]

Find attached a screen shot of the finished model.

[#7 · Accepted solution]

---

cadman777 wrote:

JD,

Thanks for the screen shot.

If I understand your screen shot correctly, then it looks like you did it the way I usually do it.

Here's how I interpret your screen shot:

The green lines are in Sketches 3 & 4 as "projected" lines, and the red lines are in the 3d sketch connecting to the 2d end points.

If that's true, then here's what I normally encounter:

The green lines will "break" (= disconnect associativity) from their parent sketch entities.

I used projected points derived from work-points, instead of projecting sketch entities (which break). It seems to work better than my "normal" method, even with 2d sketches and work-plane placement.

If not, then if you were me, what method would you use to prevent my wireframe/model from breaking? 

---

I normally only project points if I expect there will be changes that would cause projected lines to "break".
Sketch the lines myself.  I rarely project lines or part edges unless I know that geometry isn't likely to change.

In cases where I project part edges (doesn't look like the case here) I turn it into construction and sketch my own lines overtop the projected lines.

Much easier to fix if things change too much that projected points go sick.

I mostly avoid 3D sketches accept as controlled by 2D sketches.  If the work is all planar then I rarely use 3D sketches even though it is pretty easy to connect dots between the 2D sketches.

[#8]

Hi cadman777,

I did not read the entire discussion, but I wanted to point you to this link in case it helps:

http://forums.autodesk.com/t5/Autodesk-Inventor/Combining-2d-and-3d-sketches/m-p/3735908#M456488

It's common for users to want to use 3D sketches to create wireframes for FrameGenerator models, but it's almost never needed, or ever as robust as using solids or surfaces from 2D Sketches.

I hope this helps.
Best of luck to you in all of your Inventor pursuits,
Curtis
http://inventortrenches.blogspot.com

[#9]

Curtis,

Thanks for the info.

The Wiki site looks useful.

I'll try it some time.

[#10]

JD,

Thanks for the affirmation.

"I normally only project points if I expect there will be changes that would cause projected lines to "break".
Sketch the lines myself.  I rarely project lines or part edges unless I know that geometry isn't likely to change.

In cases where I project part edges (doesn't look like the case here) I turn it into construction and sketch my own lines overtop the projected lines.

Much easier to fix if things change too much that projected points go sick.

I mostly avoid 3D sketches accept as controlled by 2D sketches.  If the work is all planar then I rarely use 3D sketches even though it is pretty easy to connect dots between the 2D sketches."

This is what I'm glad you said.

Any idea if Adsk will fix this issue in the future so IV projected sketch entities don't break when the parent sketch changes?