Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Non-typical dimensioning stretches

Status
Not open for further replies.

ohiocad

Structural
Apr 6, 2001
193
Have tried to find a lisp, or simple way to do this, other than exploding or typing text. No luck so far...

What we're trying to do, is put a dimension on a object that reads true-to-scale as a "fixed" number, and then be able to stretch the object down to a more "compressed" visual representation while maintaining the actual dimension number... Kind of like a insert once-and-done-non-updating dimension number.

Exploding and stretching works, but is obviously less than ideal as it creates a lot of loose components. Typing in the text also works, but is more time consuming as well as leaving open the possibility of human typing error.

Any lisp routines any have seen that would support this?

For reference.... a part that would be 40' long, and have unique hole spacings grouped together, but the various groups are spaced far apart (say 4 or 6 groups, 5' to 8' apart), meaning that if drawn full size and fit in a viewport, the holes turn to a blob when printed. Have previously done with multiple viewports, zoomed to catch the groupings, but for a few reasons that is no longer feasible.
 
Replies continue below

Recommended for you

What happens if your dimensions are non-associative or can you enter the data in 'text override'?

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik
 
Rather than stretch the object, keep the object full size in Model Space and the dimensions in Paper Space.
 
Appreciate the input, see responses below:

dik,

For the dim associativity, (0, 1, 2) - The results we get are:
0 = (Non-associative) Come in exploded. Number reads correct, and we can then stretch as we're trying to, but the arrowheads, extension lines, etc, are all loose parts, makes a mess of a drawing if only some of the components are moved by people expecting them to be combined objects.
1 = Dims come in correct dimension, associative. But dimension reading updates when stretched.
2 = Associative and tied to geometry. Also updates when stretched.

For entering the "text override", that's what we have been doing when typing in the numbers. Human error when typing hundreds of dimensions makes this less than ideal.


IFRs,

We have tried that, but the problem is trying to get the "compressed" views, making the hole groups larger to see them easier. Only way we've found to do that, while leaving full size, is multiple viewports. (EVen when we do a "compressed view", we still keep the full size version in the file also, so it is always there for reference.
Pspace on a typical 30' part would be 5 viewports, (Left end)(Hole group 1 - 14 holes within 7")(Hole group 2 - 6 holes within 2")(Hole group 3 - 16 holes within 7")(Right end) This gives us the printable view we're looking for to send out for production, but when dimensioning between viewports, the dimensions don't read correctly. They give us Pspace dimensions, instead of the combined viewport model space dimensions.... Then we're back to typing in manually.

A modified version of this was to leave full size, and put the dims in modelspace, then stretch the text so so they are somewhat aligned. Then do the same viewports as above, showing the compressed view, and all the correct (and still associative) dimensions. Two problems with this....
1 - some sheets may have 7 or 8 viewports showing small sections (this we can work around, just time to do it)
2 - Dept manager is not a fan of large numbers of viewports that need to be rebuilt for hundreds of different parts that are all different. (Can't tell the boss what we will and won't do)

Hence, my quest to find a happy variation on dimensioning to keep everyone chipper.

If my explanations are a bit confusing, I can make and upload generic drawings showing something similar.... (proprietary info concerns, etc...)
 
I know what you're talking about. I make broken views all the time.
Create a dimension style that suits the scale of your prints when seen through the model space window.
Place those dimensions in Model space and see them appear through the MVIEW window in your layout. Presto.
If you cannot figure out the rest, like what you need to stretch and move to make the respective pieces visible and aligned, lord help you, or you are using an extremely old version of AutoCAD.
And fix your dim associativity before your employees quit.

 
Spar,

That's actually how we've been doing it so far... Broken views is a good way to explain that, and although it's been tedious, it's worked ok for us. The goal at this point, due to "requests" by management, is to find a way to do a single view, that mimics the look of the broken views.

Was hoping to find a lisp that would substitute text for dimension values, so they could be inserted correct, run the lisp on them to replace with text, so they could be stretched down without changing the values, making it one view instead of broken views.

As far as the associativity..... well... we currently run with dimassoc = 2... but to get what the mgr wants, half want to explode them, and the other half want to type dimensions... Me... well... lets just say I'm pulling out what little hair I have left since I don't like either one.

(Is it still a request when you know it's not a request?)
 
I had a boss who said "don't do anything in AutoCAD that I can't sit down and fix".

There were three strategies to deal with that attitude:
a) Never learn to do anything efficiently in AutoCAD
b) Teach the boss how to use AutoCAD better
c) Be flawless in one's use of AutoCAD, so the boss is never tempted to carry out the threat.

I opted for (c).
It worked splendidly because I learned to do tremendously intricate things with AutoCAD while never allowing the idea to cross the boss's mind to try to take over a project.

In your case, the less the manager every touches your CAD, the better. Give strategy (c) a try.
It may take a while to have sufficient effect, and you need adequately competent coworkers to present a united force of competence.

 
Dimensions between viewports does not work, much to my dismay.
But, viewports are the way to go if you can.
Once you have overwritten a dimension's text you can then use the "Find" commend to change it anytime.
If you do this, you could "automate" it using forms and LISP.
You might be able to feed new text into the dimensions using a link to a worksheet.
 
I wrote a program to achieve this function, the flow of the program is:
1. scale 1:1 mark all dimensions.
2. select all the dimensions which you want to get the TEXT.
3. Blast all the selected dimensions, and get the TEXT words of dimension.
4. -.undo to blast the dimensions
5. replace the TEXT words to all the dimensions.

 
I have a really complex solution with LISP:

Associate a reactor event with your object. I've successfully programmed something like this, but I have to experiment with an additional function that makes the reactor 'permanent' ~ by that I mean when you close the drawing, the reactor's association stays connected to the object. I was working on this because I wanted to devise a way to have Revit style label functionality in AutoCAD. And actually got it working but other priorities forced me to move onto other stuff. You'd need to have an attribute or custom property contained in the block to tell the program what it's 'actual' length is/was.

Some simpler solutions:

If the object (or block) ALWAYS appears at the same scale, you could make a 'dummy' dimension that would be incorporated into the object itself. You could have the text of this object set as an attribute. The attribute could manually be typed in, or could be filled in Automatically with LISP when the block is first inserted (there's ways to insert a block similar to how you'd draw a line: start point, end point & rotation, or whichever other way suited one best).

You could also have a separate 'dummy' dimension added afterwards. It would just be a dynamic blocks (with dynamic grips included). And this dummy dimension could be annotative meaning it will appear like it's supposed to at any given scale. It would require some extra manual manipulation (for example, you can't simply use the stretch command to adjust a dynamic block, you have to use the grips, where an an actual dimension object can be manipulated with the stretch command)... but that's not necessarily a bad thing as it would remind the engineer or drafter that the dimension is in fact a 'dummy' dimension. A long time ago I almost had a major f'up because I forgot that a dimension was fake - luckily it was so out to lunch it was easily caught.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor