Dimensioning Philosophy

[BigJoe]

Just forget about Shown dims on complex parts and sheet metal parts.

Why not just Show/Erase by Feature and View, select the
dimension you want to keep, and let Pro/E hide the rest? Doesn't
seem any more difficult than creating a dimension in the drawing.

It's not going to happen or parts that or made in the Asm.

I struggle with this one, too. If you build a part in the
assembly and reference other parts you don't have the dimensions built
into the model to show on the drawing. Anyone care to tell us how
you handle this situation?

I
think Design intent and Dimensioning are two diffenrt things. So use
them if youcan. But model your part the way you think best. Is it that
big of a deal anyway?

I disagree for the most part. Dimensioning is the method through
which design intent is communicated. I don't think you can
separate the two so easily. Modeling the part "the way you think
is best" should capture your design intent. The way you define
what "you think is best" in the model is by dimensions and
constraints. It has been a big deal, for me anyway, when I start
doing my tolerance analyses. The numbers for these analyses come from
the drawing. Often I need to make minor changes to the part to
ensure clearances. It turns into a big mess if the driving
dimensions are different from the drawing dimensions because you have
no way of knowing how adjustments to the model are going to affect
other features. If you're using model dims in the drawing, you've
got it all right there in front of you.

 

[dmiller327]

I have read people's position on the issue of driving vs created dimensions over the years. And I have been surprised that my own perspective is in the minority. Even though I have been using pro for 8 years I wonder if Iam missing something.


My experience is that often the most efficient way of modeling a feature, often is not indicative ofdesign intent. Additionally it is not always obvious during the design what features and feature relationships are important to design intent.


Once the part, sub-assembly and/or assembly is complete it is more clear how features should be dimensioned to both be machinist friendly and maintain fit and functionality. I use created dimensions to ensure design intent.


To go back and re-defining features just so they can be used in the final drawing seems an expendeture of time that I can not afford and which delivers little in return.


However, after stating all this I have to wonder if I have some how missed the boat as the most designers favor driving dimensions.

 

[dr_gallup]

To go back and re-defining features just so they can be used in the final drawing seems an expendeture of time that I can not afford and which delivers little in return.

To me the payback is when someone else has to go in and make a change. If they can just click on a dimension in the drawing, modify it, regen and the model behaves as expected without a dozen unintended consequences then it is worth the time up front.
Edited by: dr_gallup

 

[pmack009]

Just jump in and Change the dimension with consequences. You must be making boxes? In my company this will never happen. .
Edited by: pmack009

 

[pmack009]

You build your model so it's easy to change and flexible so other people can work on it down stream but when things get complex you through all the rules out the door and just build the model so you have something to work with. You could take your sweet time and try to make super models with every dimension. Parts here have 100's of futures some times complex plastic parts, get all the dimension in the model that you want in the drawing? Sure no problem if I

 

[BigJoe]

Sheet metal
parts "BigJoe" have you used the sheet metal package? If you have you
would know what I

 

[dr_gallup]

Just jump in and change the dimension without consequences. You must be making boxes? In my company this will never happen. We have model that will fail if a dimension changes .001.

We make many complex models, including insert molded plastic parts with all kinds of drafts & surfaces. But we know how to make robust models. That is the difference between a good modeler and a poor one. Robust models never fail. You should look into Delphi's horizontal modeling practices. We implemented them years before Delphi even thought about it. If you worked for me and your models failed because a dimension changed .001 I would show you the door.

 

[pmack009]

Show and Erase yes use it if you can I do. But it's not always an option. Modelingjust to have the dims in the drawing not me.
Edited by: pmack009

 

[dr_gallup]

Take a deep breath. Relax. Don't you feel better now? If you are under so much pressure, you shouldn't be wasting time in these forums. Sure, I've been in quick fix/ redefine. My fault for saying never.But size isn't really the issue here, we have many parts with tolerances of .000060" (sixty millionths). That has no bearing on robustness of models.

What is important is that we have some parts that have been in production for over 20 years and have been modeled in Pro/E for over 15 years (release 2) and someone who has never worked on them before can make a change without it blowing up in their face. I think proper dimensioning schemes with shown dimensions contribute to that.

 

[Brian_Adkins]

Just another data point from someone who has been around the block...


On modifying parts directly from drawings:


If someone handed mea Pro/E drawing, I would personally find very little value (if any) in aparticular dimension being 'shown' instead of 'created'. I've never worked in an industry where modifying a dimension directly from thedrawing would be an acceptable practice (assuming that 3D CAD system is used). Form, fit & function are too critical to stop even at the part model, let alone the drawing. In almost all cases, assembly fit must be verified at multiple levels. Since we modify our parts in part or assembly mode anyway, allowing the drawing to drive the model really adds little value.


So which do I prefer, created, or shown. Neither really... I guess I'd be in favor of which ever was quickest (aka cheapest) to create.


I've worked at, consulted for, orhave talked to users atcompanies with standards that go both ways on this topic. There were some who mandated all SHOWN dims (which was never actually 100% possible), and there were others who mandated all CREATED dims (no SHOWNs allowed). The latter practice stemmed from an "over-the-wall" mentality where the drafting groups were not allowed to create drawings that could impact the models in any way... only the reverse (aerospace company).


Anyhooo... It's funny that this argument pops up every year or so and that people get so worked up over this.

Here's some older threads on this topic:


http://www.ptcuser.org/exploder/draft/199907/msg00008.html


http://www.ptcuser.org/exploder/draft/200203/msg00064.html


http://www.ptcuser.org/cgi-bin/wilma_glimpse/pro-user?query= Show+vs.+Create+Dimensions&Search=Search&errors=0&am p;am p;am p;am p;maxfiles=50&.cgifields=archive_month&.cgifields=fi lelist


-Brian
Edited by: Brian_Adkins

 

[pedja666]

What if all key dimensions are created in the skeleton?


What about complex injection molded parts?It is all nice to show dimensions when your part is simple.Try to apply it on a complex surface part with 600 hundred features or more.

 

[pmack009]

What about complex injection molded parts?It is all nice to show dimensions when your part is simple.Try to apply it on a complex surface part with 600 hundred features or more.
Edited by: pmack009

 

[donha]

Just my opinion, you know what opinions are worth? The only reason to put a created dimension on a print, is because you simply cannot make the dimension appear on the feature. As a Designer working on Pro/E on a daily basis for 15 years, nothing is more frustrating than getting someones piece of crap drawing where the dimensions are created (driven dimensions) in the drawing. When this happens, it means I have to open up the model and figure out what feature(s) contains the dimension(s) I need to changeto reflect the changes. We do have 400-600 feature parts. If drawn correctly, I should be able to change the dimension(s) on the print and have the model regenerate properly. This would be a good day and I would take note of who modeled the part. If I have to go to the model, find the feature(s) to change, I will take note of the modeler and form an opinion. The opinion,I remind you once again about opinions, would not be a good one of the person who created the drawing. So, go ahead and make my day

To those that advicate putting driven dimensions on prints, perhaps you have your reasons. Those reasons do not apply in my mechanical world.

 

[newjack]

<?:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" />
All right everyone I

 

[BigJoe]

I

 

[BigJoe]

The
friend of my friend in the big telco that I mentioned in my first post
finally fired off an email after reading this thread. This guy
has 15+ years of Pro/E experience in many different industries.
We tried to get him to post his response, but he said he's not even
going to bother to chime in to this discussion (something about talking
with a bunch of bulletheads). His email was colorful but I
thought it was also quite insightful. This is the perspective I
was looking for when I started this thread. Here is what he had
to say:<br style="font-family: arial,helvetica,sans-serif;">


"First of all, the guy is a f**
for trying to modify the part (noticed I said "part") in a drawing.
Lame. You never modify a part in the drawing. The drawing is the LEAST
critical aspect. All tooling is created from the solid model.
Therefore, you modify from the solid model. Modifying a drawing tells
me that he has never designed any parts and is just a drafter. <br style="color: rgb(0, 102, 0);">
Capturing design intent is a
f**king myth. When you design something, between ID changes and
everything in between the design intent literally changes about 10
times. This is fact from all of my projects starting from scratch to
finish. You'd have to be a dickhead to go back and redefine all the
features to show design intent. And for what benefit? So you can do
"show dims"? Haha. That's a waste of f**king time right there. Just
like how you guys dimension every f**king feature. WASTE OF TIME. The
only thing that matters is that the model regenerates cleanly and makes
logical sense if someone were to try and modify it. <br style="color: rgb(0, 102, 0);">
Model cleanly and do drawings with only CTF dims. DONE!"

 

[newjack]

<BR style="COLOR: rgb(0,102,0)">Model cleanly and do drawings with only CTF dims. DONE!"

Thanks BigJoe! This guys points are interesting. Entertaining too!

Goaaaaal...

 

[pmack009]

That my point right on the money big joe and newjack


Fully dims a 200+ featurepart not here we add a note all model geometry to be created from cad model. we will add hole to hole dims thats it.

 

[scottm]

<?:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" /><O>
20-30 even 50 years ago, how did designers & engineers convey the design intent? Was it by using

 

[BigJoe]

<strong style="font-weight: normal;">Using the "SHOW/HIDE" seems like
the right thing to do because its a bit of a PIA. And it if its a
PIA, than it must be the proper way to do it in PRO-E. Creating
dimensions just feels like cheating. Its too easy, so it must be
bad...[/b]

I don't know if you're right scottm, but your logic here is
spot-on. Seems like most of what we've gotta do in Pro/E is a
PItA. Though it's a little better after you get yourself
accustomed to using a good set of mapkeys.



I'm a little suprised to see the amount of emotion behind these opinions on both sides of the philosophy.