cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Want the oppurtunity to discuss enhancements to PTC products? Join a working group! X

EXTERNAL: Decimal Rounding

BrianMartin
11-Garnet

EXTERNAL: Decimal Rounding

I normally wouldn't chime in to everyone like this but bug or not, I can't believe anyone would want to go back to creating all dimensions in a drawing. I'm not trying to re-ignite the whole created versus shown discussion but I think most people agree that creating dimensions is the way of the past. If you're trying to cut metal from your model, if you're trying to go paperless, or if you just like living in the 21st century, your dimensions come from the model. Work with PTC to document the bug and develop a work around if necessary but don't throw out all the good of shown dimensions and model based engineering for such a thing.

In fact, it concerns me that a company such as Lockheed is even considering such a thing. I know my employer works with Lockheed and it would give us pause to trust work coming from you if we knew that was your policy.



This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.
12 REPLIES 12

I'll side with the created dim people vs. the shown dim at least for parts that have 2D drawings.
As many have said before, the dimensions that appear on a drawing may not be in the same layout, etc... as they are desired in features / sketches.

Especially when getting 3D models from customers, I'll create a verification drawing using the same views, etc... and dimensions the same as the customer's 2D drawing just to make sure their model matches their design.

One of the biggest problems I have with created dims in a 2D drawing is being able to select edges that are not the outline of the part and even edges that are not shown on the drawing (hidden). It is a real bear when you are dimensioning a part width / length that may have small rounded / chamfered edges.

I think the drawing selection filter for created dims should only let you select outlines (white) or centerlines / axes. No tangent edges or hidden lines should be able to be selected.

It would also be very useful if the hole callouts, etc... that are used in a pattern would prefix a nX for the number of instances in the pattern on the hole note.

For most mechanical assembly parts (automotive / aerospace / medical) get rid of decimal place-based tolerancing, get all dims to 4 decimal places (inch) or three decimal place (metric) regardless of tolerance and I'll start using model dims in drawings as well.


Christopher F. Gosnell

FPD Company
124 Hidden Valley Road
McMurray, PA 15317

I agree. I just can’t get past the idea that someone who would change a drawing dimension (which is a deliberate act) would somehow not simply open the model and change it there instead. Couple this with a release process, where drawings are supposedly reviewed prior to being locked during the release process and it moves even further afield. And don’t even get me started on the ease with which created dims can be overwritten using @O, destroying the very parametric functionality CAD programs should all strive for.

can't help myself

I love it!


BenLoosli
23-Emerald II
(To:BrianMartin)

Couple of old ones

First of all, that is NOT Lockheed’s policy. That is MY personal belief that there is no value add by forcing the modelers to try to anticipate what changes in design direction might happen during the design process. My geometry comes from the model, but WHY do the dimensions have to come from how the model was built. If you want to annotate the 3D model, you can do that as well without being tied to HOW the model is structured. This is still model BASED engineering. Please explain what is the magic that comes with shown dimensions. I have been using Pro for 20 years, and making parts without drawings MANY times in that period. Do you not understand that created dimensions DO come from the model? If the model is updated, the created dimension updates also? You say that we shouldn’t throw out all the good of shown dimensions, what good is that? Give me some examples of what you can do with shown dimensions that can’t be done with created dimensions. My preference is to build models based as much on the first three datums in the part which leads to stable geometry and features.

Model based engineering does not limit us to having to work one way. My way will also get you all the data you need.

This is like a religious thing, if I don’t toe the line somehow I’m bad, my design is bad, my engineering is bad? I often even refer to myself as a heretic but that has usually been at least somewhat in jest. Now you want to burn me and my employer at the stake for questioning the conventional wisdom?

It gives me pause that your employer would consider such a closed attitude to another idea.

Rob Reifsnyder
Mechanical Design Engineer/ Producibility Engineer / Components Engineer / Pro/E SME / Pro/E Librarian
[LM_Logo_Tag_RGB_NoR_r06]

Rob,

A few reasons that I can state right off and I will leave it up to you to determine how important they are to you:


1. Why recreate what you have painstakingly already done? You need dimensions to drive feature geometry in the sketches, why not make it meaningful to begin with?

2. Able to change part geometry right in the drawing. This can only be done with driving dimensions.

3. If you do not model the design intent correct, you cannot be sure that some other feature did not move when changing another features dimension.

4. Proper tolerance analysis within Creo.

Timothy Wells
Vice President - Engineering

[cid:image004.jpg@01CEB54D.72C5D4C0]

KG Technologies, Inc.
429 E. Cotati Ave.
Cotati, CA, USA 94931

Please see below for my comments.

Rob Reifsnyder
Mechanical Design Engineer/ Producibility Engineer / Components Engineer / Pro/E SME / Pro/E Librarian
[LM_Logo_Tag_RGB_NoR_r06]
StephenW
23-Emerald II
(To:BrianMartin)

I am a firm believer that the created vs. shown dimension debate is simply the most entertaining and aggravating argument that has ever been on the exploder. It rears its ugly head from time to time.

When someone says that “my way is the only right way”, I have a hard time believing that they have seen enough to know all the possibilities. Not saying their way may not be the absolutely best way for their job and their company and all things related to them, but in the last few years, I have seen some crazy stuff I would have never have thought was the best way or even a ½ way acceptable way and it turns out, I might have been wrong (no admission of guilt!!!).

When someone says “use ALL the tools at your disposal to the best of your ability”, I have a tendency to believe they have seen a few different approaches and have come to the realization that more than one way can be right and it doesn’t really matter how you get the job done as long as you get it done right (and really, really fast because that’s what my boss wants).

FYI, I use both created and shown dimensions. I show all the dimensions I can and usually redefine my model so the shown dimensions properly reflect on the drawing. I create the dimensions that don’t make sense from the design intent standpoint or are for manufacturing. And every once in a while, I’ll sketch a dimension using lines, just because I’m cool like that!!


Rob

Well said…couldn’t agree more.

Regards,

Neal Hanratty
Engineering Systems and Standards Manager
Terex Materials Processing Group

T +44 (0)28 8241 8769
F +44 (0)28 8225 2740
M +44 (0)75 3495 2739
E neal.hanratty@terex.com<">mailto:neal.hanratty@terex.com>

#2 & #3 has me a little curious. I’m trying to see it from your viewpoint, so bear with me here. You said:


2. I SPECIFICALLY don’t want a part changed in the drawing. If I could I would restrict changes (after release) to be done in assembly so the actual impact of your change could be seen. Actually, I wouldn’t go quite that far, but the principle remains. I don’t want parts changed in the drawing.

a. Is your system set up that a user could open the drawing and make modifications that would propagate into the model (due to life cycle state) but that user couldn’t simply open the model and make modifications there? If both the drawing and the model are released, they cannot change it anyway (except locally). If it’s at some sort of In Work state, that user could simply open the model and make the change after finding out it’s a created dimension.

b. Also, even in drawing mode the process of modifying a shown dimension is a deliberate act. Unless I’m missing something, this is not an act that someone can stumble through to accidentally change the model from the drawing interface: double click shown dimension > pick ‘Nominal Value’ cell for dimension value > enter new value > pick OK button > regenerate. (Alternately: single click shown dimension > RMB > Modify Nominal Value > enter new value > Enter key > regenerate.) If someone is claiming they’re accidentally modifying a shown dimension they’re shoveling a lot of BS.

3. You’re right, but again, if I use FRONT, TOP and RIGHT that isn’t a problem. Also, your assumption is that whoever changes the model after me is less skilled than I am and doesn’t see that change.

a. You reference everything to FRONT, TOP AND RIGHT instead of referencing everything to set datums? So when someone wants to change the distance from [-A-] displayed as a created dimension they have calculate the difference, go into the model, find a different dimension and apply that difference? Somehow I feel that I’m reading this explanation as a good thing yet I can’t figure out how it is a good thing.

b. Yes, always assume the next user will struggle unless it is the most straight-forward methodology possible. Actually, even then.


BenLoosli
23-Emerald II
(To:BrianMartin)

Many comments in this thread have changed the original as being a discussion of how Pro/E incorrectly rounds numbers to the shown/created dimensions flames.

As has been stated, there are many different ways to accomplish the task at hand using a tool like a CAD system. The tool does not create the geometry, datums or manufacturing operations. These are all created by a human with foresight, forethought and planning as to what is the best method to design or manufacture a part. The CAD system is merely the tool that the person uses to make a mathematically correct model and drawings, if needed. Each person who is trained to use their CAD tool picks up different aspects as they learn the system. Many have been using the CAD system for years, but how many have received new training on the features in a new release? I started using CAD in 1978 at school using an Applicon 880 system. I have also used Computervision CADDS 4/4x with no formal training, then moved to Unigraphics UGii V3 and after 14 years moved to Pro/Engineer 2000i2. Extensive training when I learned UGII and 2000i2/2001 but nothing since then beyond a ½ day PTC Wildfire update class.

One of the things I learned when I had to convince management to spend the $$s to retrain 35 Unigraphics people on Pro/E was that management had to buy into the concept that training is not an expense but an investment! If they treat it like an expense, it is too easy to get cut from the budget for some office supplies the HR department needs. Treat training as an investment for both the people and the company. The person benefits in that he learns a new tool. The company benefits from the person being productive with that tool in a much shorter time frame than if he had to learn it himself or use a self-training library. We used both at one company and could see the difference in user proficiency. We also tried different training scenarios on the Pro/E training. We did 12 days training minimum and most had 17 days with a handful more getting 22 days depending on their job function. We tried staggering the training, 5 days of class, 3 weeks work, then more training to 12 days straight training. It all made a difference in the amount of information the user retained and how well the performed on Pro/E after the training.


Top Tags