Can you make this a more reliable model ?

[Curtis_Waguespack]

So the attached 2015 model is simplified example of a part that is being automated via iLogic as part of a larger design.

I've put a simple form in the part that you can use to adjust a workplane that controls a split feature.

The goal is to change the Offset value, and use a reference sketch to gather a couple of dimensions.

The problem is the way that it will be used, the offset value can be larger than half the diameter, and so the reference sketch gets "lost":

The result is one dimension goes missing and the other stops updating correctly:

Does a more reliable way to create this model, so that the reference dimensions will work, come to mind?  I keep thinking that something will jump out at me, and as of yet nothing has.

Thanks in advance,

Curtis

[Anonymous]

Hi Curtis,

You could change the bounds on your slider. See the values in the red box below.

[Curtis_Waguespack]

Hi Giordanik,

Thanks for the reply. In this case the slider is just to demonstrate the issue. In real life the value is driven from elsewhere.

And unfortunately, this value is not easily bound, as in some instances there will be an offset value and no diameter value, so it's not possible to check one against the other.

( Also, just FYI, the slider bounds restrict the slider, but do not restrict typed in values above or below that. )

Thanks again,
Curtis

[Anonymous]

Hi Curtis,

Thanks for the info. 

Another idea is to make an iLogic rule that would suppress the sketch should it go larger or smaller than ±1.45mm (or the value of your choice). That should take care of the entered-value problem.

[Curtis_Waguespack]

Hi Giordanik,

That would be outstanding! ... but alas we don't have that ability  (unless I've missed something).

But you can go vote for the idea to allow this:

http://forums.autodesk.com/t5/inventor-ideas/feature-suppression-automatic-sketch-suppression/idi-p/...

Thanks again,

Curtis

[Anonymous · Accepted solution]

Hi Curtis,

Well, that's a bummer. See attached part document. I have made one of the profiles an infinitesimally small surface extrusion that should go unnoticed in any assembly. I consider the addition of this worthless feature one of the eggs that needs breaking for your omelet. If you have a better egg to break, let me know.

I have added the rule I talked about, except I moved the EOP instead of suppressing the sketch (because that is strangely impossible, as you pointed out).

[Curtis_Waguespack]

Hi Giordanik,

Well I was kind of looking for a non-illogic solution, so I'll have to deduct 1/10 of a point. But you earn back a full point for the clever use of the EOP.

I'm not sure if I can use this as a solution in the overall workflow or not, just yet. I'll have to put it in place and test it a bit to see. But it is a solution, so I'll accept it along with any others that someone might offer. For now though, I'm going to leave this thread unsolved to see if anyone else has some thoughts.

Thanks again!

Curtis

[-niels-]

Hi Curtis,

This isn't really my usual "cup of tea" but i gave it a shot.

I think this might be more "robust" by using min and max equations, see my attached model.

If you change the "Value" parameter it has hard limits set by using those.

I did recreate the model, so the ilogic is gone and even though it's a simple form i didn't want to figure out how to add it back.

I changed it so the origin plane is now used as the split plane, so the reference sketch remains on a non-moving plane.

As a reference for the min/max i also found this blog post:

https://designandmotion.net/autodesk/inventor/parameters/inventor-parameter-min-max/

Hope this is what you were after....

[admaiora]

Not sure if i have understood, all.

But could this work?

[Anonymous]

Hi Curtis,

I love my wife to bits. That said, you do kind of remind me of when she "lets" me pick where we go to dine out. It's an iterative process, except I get full credit with you. 🙂 Thanks for the feedback.

Alas, I am running Inventor Professional 2015 so I could not see Neils' model. I have never used a max/min function in parameters and I got excited to see a live specimen for a second there.

To make the iLogic adaptive, you could include parameters instead of the outright values (-1.45, 1.45). You could make a custom parameter to be the radius, setting that as the bound for the EOP. Out of curiosity, why would you want to shy away from iLogic? 

[-niels- · Accepted solution]

@Anonymous: let's prolong that excitement for another second...

Here's a screenshot of the parameter window of my part:

It's not all that exciting, i just made sure there was a safety margin equal to the shell thickness.

At the minimum value it should have been "(d6+some offset value)" instead of "0.2" but i forgot to change that and 0.11 would have worked as well.

And like the blog post mentioned, it's kinda backwards...

[Curtis_Waguespack]

---

@Anonymous wrote:

Out of curiosity, why would you want to shy away from iLogic? 

---

This model is part of a larger set of customization, of an assembly. Currently, the user can make changes at the assembly level in a certain order that will put the offset value higher than the diameter of this part. In these cases it doesn't matter to the overall design because this part is not in use at that time, but it creates an error and causes the reference dimensions of the sketch to become upset. If the user then choses to enable this option at the assembly level, then this model no longer works.

Mostly I just felt there was some better way to approach the model that I was missing due to my own tunnel vision of having fixed that sketch  a dozen times during my testing of the configured assembly. And I thought someone else might see a solution more clearly.

So also just for curiosity, and for future reference I suppose.

---

@Anonymous wrote:

To make the iLogic adaptive, you could include parameters instead of the outright values (-1.45, 1.45). You could make a custom parameter to be the radius, setting that as the bound for the EOP.

---

Indeed, that's what I did yesterday after you posted your file. I ended up with the attached version (Inventor 2015), which looks like it would work just fine.

Thanks!

Curtis

[Curtis_Waguespack]

HI admaiora,

Thanks for having a look at this.

It appears you're model works, but I can't tell what you've done.

Thanks again,
Curtis

[Anonymous]

Hi Curtis,

Ah. So you were trying to get a pure modeling solution to the problem. There is value in simplicity. iLogic has thrown a wrench into more than one design for me. 

Something else I thought of - Using equations for arcs and triangles, you might be able to get the values in the sketch. It would be pure "modeling" but a bit complicated and probably not exact, but close. Unfortunately I don't have the time to get into it at the moment. Possibly later I'll post something. 

[Curtis_Waguespack]

Hi -niels-,

Thanks for looking at this as well.

MIN / MAX would work well if I were playing by normal rules, but in this case I expect the Offset to exceed the diameter, I just want the reference sketch to stay together when it does.

It's sort of a "Schrödinger's cat" model that I want to allow to be broken and not broken at the same time.  

Thanks again,

Curits

[Curtis_Waguespack]

---

@Anonymous wrote:

 So you were trying to get a pure modeling solution to the problem.

---

Hi Giordanik,

Yep that was what I was after. And to that point, I've just noticed that if I make the reference sketch a 3D sketch, it doesn't seem to break the same way the 2D sketch does? At least so far. So that might be the simplest solution.

I'm likely going to run with your iLogic EOP solution and the 3D sketch for a while to see if the 3D sketch will really hold up, or if I've just not turned it inside out enough to break yet.

Thanks,
Curtis

[admaiora · Accepted solution]

Hi @Curtis_Waguespack ,

i have removed the Projected Cut Edges and just replaced with classic projected edges (just 2 edges)

[Curtis_Waguespack]

Hi admaiora,

Ahhh! That's interesting. I hadn't caught that the difference in projecting methods behaved differently in this case.

Thanks you,
Curtis