Request for sketch critique (take 2)

[Anonymous]

I am dimensioning and constraining this for practice. As jhackney1972 mentioned, there is little practical purpose given the eventual target is frame generator. Currently, I have the following issues

1) unless relaxed constraints are active, modifying some of the dimensions fails

2) the overall length/height (specified as 35mm in the dwg) cannot be set to 35mm without breaking geomery

The strategic question I have is, is it better to set dimensions first, or constraints? 

EDIT: Also, when I finished constraining it, it said it needed 40-50 more constraints. Is that "good enough" Is there an "acceptable number" of missing constraints in a professional environment? Or is the only acceptable number zero? 

EDIT 2:

I had not seen JDMather's responses to my previous post. I will review them and try to replicate them in the coming days. 

[JDMather]

---

@Anonymous wrote:

.... As jhackney1972 mentioned, there is little practical purpose given the eventual target is frame generator. ..... 

---

I disagree with this statement.

My preference would be to do it right, especially since it will be used with Frame Generator.

Unless I misunderstand the design intent - I would expect to slide parts in the direction of the arrows into this part and have them hit the "stops", inner and outer, at the same time.

I would not use my initial 4 attempts.

I spent about another 30 minutes analyzing symmetry and found ways to significantly simplify my sketch.

[Anonymous]

Your analysis is accurate. However, I do not yet see how to use this to simplify the design, unless in the trivial sense that various parts of the "stops" are colinear.

To restate my initial questions; were you able to reduce the number of needed constraints to zero? How many constraints did you end up with?

[Anonymous]

How's this? Still 36unconstrained dimensions but the numerical dimensions all work as advertised.

[Anonymous]

Fully constrained using autoconstrain

[Xun.Zhang]

Hi,

Typically, I would suggest to handle 2D sketch full constraint as below steps:

1. Dimension or constraint with a based point for each possible geometry, such as circle diameter as well as center point, Linear distance as well as location, Angular as well as location, etc.

2. Mirror intent, please full constraint the half before Mirror.

You can also refer below image for a reference, hope it helps...

[JDMather]

---

@Anonymous wrote:

---

@Anonymous wrote:

.... As jhackney1972 mentioned, there is little practical purpose given the eventual target is frame generator. ..... 

---

I disagree with this statement.

---

I guess I got confused about what AutoCAD profile you are using?

When I check the AutoCAD those two lines are aligned - not sure what happened in your file?

[Anonymous]

Oh, now I see exactly what you are saying.

I implemented your suggestion somewhat in 3_8_profile, I think. For now, I started by realizing that I could clear up a lot of mess by simply enforcing symmetry about the diagonal. Please let me know what you think. After I got up to about 25-35 unconstrained dimensions, I used autoconstrain. I left 3_8_profile with about 28 unconstrained dimensions to show how far I can get manually. 

[JDMather]

@Anonymous

See attached.

[Anonymous]

OK, thanks so much. I will study those carefully. For reference, did you use autoconstrain on any/all of them? If so, what is the approximate ratio of your work to autoconstrain?

Also, how long did these take you, if you don't mind my asking?

[JDMather]

In 15 years of using Inventor I have never ever used autoconstain.
About 60 seconds for each.

[Anonymous]

60 seconds from scratch, or based off my work?

Thanks again

Joe

[JDMather]

As @jhackney1972 indicated to you earlier (why many threads for on issue - makes it hard to follow the problem), if the AutoCAD is correct geometry - simply import it, convert to Sketch Block and constrain as a block. Done!

[Anonymous]

Well, I created this thread to ask for critique of my original work - it's fair to say that I copied the original geometry "for practice".

I will do as you suggest regarding the sketch block. However, I must admit I am still perfecting my sketch/dimension/constraint methodology.

[Anonymous]

One last question; 1/2 profile and 3/8 profile are not fully constrained. Is there a point at which you decide that sketches are "sufficiently constrained"? What is your strategy for this?

[JDMather]

If I were creating this myself from scratch - I would be very careful to fully constrain the sketch.

I did this for one of the profiles - I was about 90% done with everything fully constrained.

Then I added another line and Inventor got confused on the constraints and most of the sketch showed unconstrained even though I knew it was.

I continued on to the finish just so I would learn the design intent of the original modeler.

Then I looked at how I might simplify my sketch technique so that I did would not confuse Inventor sketch solver again if I started over.

At this point I had a couple of hours work into something that was of no real benefit to me - so I abandoned the effort.

But if it was for my work - I am quite certain that I could significantly simplify my sketch and have a robust fully defined sketch that would be easy to edit for adjustments or for various sizes.

I always consider my first (or two or three) attempts as "throw-away" attempts that I use only for learning purposes.

Each attempt generally goes much faster than the previous attempt and when I finish I can sit back and declare the final attempt a thing of beauty, a work of art.  Then I come back a year, month, week, day later and think, "What was I thinking, what a piece of rubbish..."  And do it all over again.

[Anonymous]

Thanks so much for the insight.

In the spirit of your own approach, I'm attaching one last attempt (for now). Again, any input you have regarding my use of basic sketch tools would be greatly appreciated.