Pro-E / PTC sucks

[phoxeoy]

To me, the way I create the model has everything to do with what that part should do and little to do with how it needs to be made.

Now I'm just lost.

 

[phoxeoy]

The drawing is not the end result, the part is. The drawing is a means to communicate either how to make the part or how to inspect it once it is made.

 

[dgs]

That frequently requires different thinking than how to design or model it.

Amazing, how is that?

I recently did a design for a sheet metal drawer cabinet assembly. On the drawer face, we needed a hold for the purchased lock. I needed that hole located relative to the surface that the lock catch was mounted on, inside the cabinet. This dimension came from the lock manufacturer. Let's say it was .50" (I don't remember what it actually was). That did not align with any edge on the sheet metal face. The hole might need to be .75" from the edge of the face to be .50" from the catch surface.


Now, I could create the hole at .75" from the edge of the face,. but my intent is not that the lock be .75" from the face but that it be .50" from the catch surface. I know that the catch surface needs to be 8" from the cabinet center, so I create a datum plane 8" from center and locate my lock hole .50" from the datum.


I've now captured my design intent, but if I show that .50" dim on the drawing, it'll be meaningless. So instead, I create the .75" dim to the edge of the part.


Now in realty I created a skeleton in the cabinet that contained the axis for the lock and the plane for the catch and I copied both down to my face and located my hole directly from the axis. Now I have no dimensions to show in my drawing at all.


Does that make sense?

 

[dgs]

To me, the way I create the model has everything to do with what that part should do and little to do with how it needs to be made.

Now I'm just lost.

When I was referring to the 'part' I meant the actual manufactured item, not the Pro|E part file. Does that make sense?


In other words the drawing, whether created in Pro|E or with pencil on paper, is a means to an end. The end is getting something manufactured. The Pro|E model is simply another means to that same end. The Pro|E model is not a means to getting a drawing made, it's a means to getting a part designed and manufactured.

 

[phoxeoy]

I've now captured my design intent, but if I show that .50" dim on the drawing, it'll be meaningless. So instead, I create the .75" dim to the edge of the part.



Does that make sense?

No, it does not. How can you possible con via your design intent if you just create the .75 dimension in the end. If you hand that drawing to anyone, to them, your design intent is .75 from the edge.

 

[dgs]

I've now captured my design intent, but if I show that .50" dim on the drawing, it'll be meaningless. So instead, I create the .75" dim to the edge of the part.


Does that make sense?

No, it does not. How can you possible con via your design intent if you just create the .75 dimension in the end. If you hand that drawing to anyone, to them, your design intent is .75 from the edge.

No, my design intent is that it be .50" from the catch surface inside the cabinet. It must be or the lock won't work. The .75 from the edge puts the lock hole in the position to achieve the intent of .50 from the catch surface. There's nothing sacred about the .75 dim to the edge, but the .50 dim cannot change.


Lets say the exterior size of the cabinet and therefore the overall size of the drawer face changes. My catch is still 8" from the center, however. Now my .75 dim might be 1.50, but the .50" to the catch surfaceis still the same. If I had modeled my part using the .75" dimension to the edge, the lock hole would be in the wrong place after I modified the overall size. I'd then have to go back an modify the lock hole to put it back where it belongs. If I forgot to do so, the face would be wrong when it was manufactured correctly to the drawing. By using the datum plane (or skeleton), the lock hole is located to what's important - the catch - and the hole will stay where I need it when I modify the overall size.


There's no way to give the guy making the face the .50" dim to the catch that I need, but it is the dimension that's important. Creating the .75 dim tells him where to put the hole in order to achieve the intent of .50" from the catch. What I want is that .75 dim to change as needed to keep the .50 dim that I want.

 

[phoxeoy]

Clearly you feel you have accomplished your end goal. Sounds like a good datum and Gtol could handle the detail side of things. But if I don't have anything on the drawing to tell me that your goal is to achieve the .50 dim; creating a .75 dim from the edge and hoping for no tolerance stack up is good enough then by all means go for it.

 

[dgs]

... creating a .75 dim from the edge and hoping for no tolerance stack up is good enough then by all means go for it.

The tolerance stack up is a whole different discussion. Of course it needs to be done.

But if I don't have anything on the drawing to tell me that your goal is to achieve the .50 dim ...

The guy making the drawer face doesn't need to know that it's important to have the hole .50" from the lock catch. It doesn't matter to him at all as the catch is on a completely different part. The .50" dimension is a design specification for the assembly. Since I can capture it directly in my model, I feel as though I should to ensure that the manufactured assembly captures it as well.


Another design spec might be that the drawer tray hold 75 lbs. I wouldn't put that on my manufacturing drawing because it's my responsibility as the design engineer to design it in such a way that it does hold 75 lbs, not the guy on the shop floor. I do that through material specification,process specification, fastenerspecification and mechanical design. Some of that is then directly communicated on the drawing,some isindirectly communicated and others (like fastener spec) can't be. Then I communicate to the guy in the shop how to make whatI've designed. Now, I can't capture 'must hold 75lbs.' in a single dimension like I can 'must be .50 from lock catch'.


It's the same with the .50 specification. I can't communicate it directly on the drawing, but I can communicate how to make the part so that the .50 spec is satisfied.


I suspect that we could go around and around on this. Suffice to say, that in my mind design intent and communication to manufacturing are not the same thing, and frankly shouldn't be. Youfeel that they must be the same. To each his own. I was hoping to show why I feel as I do, but it's hard to communicate an example with words.

 

[phoxeoy]

Design intent should be important at the shop floor. I would hate to think I located a hole from one surface and the guy on the shop floor decides, "I think I'll index off this other surface today". Mmmmm, why in the world is this part not working....

 

[damormino]

dgs is correct on this point. The whole point of doing all this design in an integrated design system (3D models, assemblies, drawings, CNC programs, etc.) is to capture the all of design intent and interrelations contiguously, and then be able to leverage from that to derive just what is important for any given piece of the product design. That design intent ultimately includes manufacturing intent when all is said and done.

Having said that, we must remember that at some point, 2D drawings will become mostly obsolete. The goal is to remove the importance of paper. Why would one create a paper drawing if the model information could be transmitted to the manufacturing center, and the part produced? That still supports dgs's statements regarding design intent. In this case, the manufactured part is driven by the model, which should really be driven by the assembly.

The danger in all this is, as phoxeoy pointed out, that software systems like Pro/E can lull designers into designing without regard for how parts will be manufactured. SolidWorks is especially poor in this regard.

dgs's example of driving the design with a parameter (the distance to the locking plane) is parametric design. It is one area where Pro/E really does shine compared to other products. Too bad the functionality of much of the user interface requires more effort to achieve this.

 

[gkbeer]

The danger in all this is, as phoxeoy pointed out, that software systems like Pro/E can lull designers into designing without regard for how parts will be manufactured. SolidWorks is especially poor in this regard.

That's nothing new. Drafters where doing the same even before CAD systems became common. It's a people lacking of skills problem.

Edited by: gkbeer

 

[GLEN20]

I'm with you guys, I rarely show dims. I never set tolerances or add notes to the model dims either. Other than the 90mph modeling of purchased parts. My fabricated parts are modeled within the speed limit


The Machine shops get a hard copy (or pdf's) of my model as a "communication tool", but they prefer I send an IGES or STEP files to machine with........ some use the step file to do their own "checks" using Solidworks.

 

[arroyopr]

Having said that, we must remember that at some point, 2D drawings will become mostly obsolete. The goal is to remove the importance of paper. Why would one create a paper drawing if the model information could be transmitted to the manufacturing center, and the part produced? That still supports dgs's statements regarding design intent. In this case, the manufactured part is driven by the model, which should really be driven by the assembly.

Paperless Engineering....It's sound like kaos. Not to mention all the legal battles you're going to lose for not having Offical Doc. on your design, invention,... etc. This have been intented before with no success, what make you think it's going to happen in the next 100's years. Are you forgetting that we have electronics thiefs this days.

 

[dgs]

Having said that, we must remember that at some point, 2D drawings will become mostly obsolete. The goal is to remove the importance of paper. Why would one create a paper drawing if the model information could be transmitted to the manufacturing center, and the part produced? That still supports dg's statements regarding design intent. In this case, the manufactured part is driven by the model, which should really be driven by the assembly.

People have been looking for this as long as I've been using Pro|E - 10 years (Mindripper - that's about 15,000 hours on Pro|E

) I went to the '3D drawings' class at PTC|User. It's actually pretty neat, but as a communication tool it's limited as you need Pro|E or at least Product View to see it.


Adobe's 3D pdf looks more promising as there is full digital rights management built in. DRM is also coming in WF4. It has the ability to incorporate 3D data as well as 2D in one file in a format that nearly universally accepted and used.


The big hurdle to going completely paper less is coming up with ways to communicate non-geometry things like P/N, color, material, finish spec, tolerances, etc in a universal manner. The reasons drawings are great is that they are universally understood as a format that has been around for over 100 years.


Personally, I think PTC ought to focus on how to make drawings easy, easy, easy. If the drawing for a moderately complicated part could be completed in hours instead of days, the push to eliminate them would go away. Frankly, even without drawings, your going to spend a few hours on that same part adding in all the non-geometry stuff and making it clear for the downstream user anyway.
Edited by: dgs

 

[phoxeoy]

Ok, I'm back on the band wagon. I'm currently doing development work on a new STEP standard (AP224). In one file, it will contain feature information, dimension values, tolerances values, geometric tolerance information, notes, material information, datum locations and reference locations. Basically everything.

The focus of this is, as I was trying to point out earlier, without a drawing depicting your design intent you will never be able to get it across to anybody else unless they have Pro/E as well. I know from countless years of doing this, not everybody is using ProE to model, NC code, ...bla bla bla.. So exporting out a IGES or STP or any other dumb geometry file with no drawing is not going to tell anybody how this part should be made. Yeah sure you can make the part but you have no way to know if this hole should be measured from this surface or not.

This is why AP224 is being developed. So you can have a format to export and maintain that information. When I create a model for development, I have to put every bit of information in the model. I can only translate information that is there. This is what I'm getting at here people. If you don't put all this kind of information in your models then I really would like to know how you expect a paperless environment.

You can have all the design intent you want in your models but once you export it out to a different file format, it's gone. This is where the drawing comes in, which still is the only form we have to con via the design.

 

[MarkEngr]

I've been using MCAD software since 1982. I started with CADAM, using a light stylus on a black-and-white screen, with a 9-button function box. It ran on an IBM 370. <?:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" /><O></O>


Shortly after I left college in 1984 and got a job, one of my first tasks was to be part of a team selecting the company's first CAD tools. We spent $3M. We selected Prime Medusa, which was one of the first true solid modelers. This was actually a pretty good package. I used it for about eight years. My next two jobs were running AutoCAD R12 and R13, which was a decent 2D package, but not parametric and very low-end. I still use it occasionally though: it has some handy functionality. <O></O>


In 1997, I started using Pro/E. I found it very difficult to learn, and full of quirks. I wasworking with injection-molded parts, so most of my time was spent workingwith the advanced feature creation and surfacing functions.I got to be pretty good at it after a few years, but it remained a challenge to use. I refer to Pro/E as 'user-hostile software written by anal retentives'. Making drawings was particularly difficult, and I was glad I didn't have to make very many. <O></O>


Then I discovered SolidWorks in 1999. I was so easy to learn and use, yet still pretty powerful. It didn't have the sophisticated surfacing and advanced feature creation functions of Pro/E, but it was a real pleasure to use, and drawings were easy. I liked it so much, I actually bought my own license for $3000. During the last eight years, I have used SolidWorks for a variety of tasks, from complex assemblies of simple components to molded parts. Even today, it lacks the complex surfacing capabilities of Pro/E - but it's core functionalityhas evolved dramatically. <O></O>


Now I have accepted a position where I'm using Pro/E fot the last five months: Wildfire 2. This company has a product with a moderate number of simple parts, and no complex surfacing. This brings out the worst in Pro/E: lots of drawings, and many small assemblies. I have discovered in the last few months that Pro/E has NOT evolved much in the past six years. There is a new cutesy GUI, but the functions - and the core program - are essentially unchanged. Drawing generation has not improved at all, and remains a profoundly weak part of Pro/E. It is still a Unix application: it's running in a curious Windows emulation mode that's neither Windows nor Unix. The new GUI is only partially implemented: some of the functions still have the very same popup menus as Release 16, and the graphic quality is lousy compared to SolidWorks. Command input lines pop up in different places: no consistency in the program operation. Regeneration failure resolution has improved - but only to the extent that info is more accessible. The ability of the core program to deal with these simple failures still pales in comparison to SolidWorks. <O></O>


So I have thousands of hours of experience in both SolidWorks and Pro/E, and tens of thousands of hours of experience running MCAD (especially 3D solidmodeling) over the last 26 years: obviously myopinion is worth something.SolidWorks is a vastly superior product to Pro/E, except in the areas of complex feature creation. It's also much less expensive, has more users, and is growing at a rate PTC can only dream of. And SolidWorks is rapidly evolving into the complex surfacing arena too, as Dassault finally allowed C2 continuity to be included in the 2007 release. Add-ons really aren't needed with SolidWorks, unlike Pro/E (which requires add-on modules at great expense for some of it's core functionality). <O></O>


My advice for ANYONE contemplating buying a solid modeling program and investing their future in it? Forget Pro/E, get SolidWorks. There is no comparison: ease of learning, installed user base, core functionality, cost, Windows compatibility, product quality, support (not that much is needed) all show SolidWorks as the clear winner in every category - unless you're going to do complex surfacing and advanced part modeling such as molded parts. <O></O>


There's a reason why all the best people from PTC pulled up stakes back in the early '90s and started SolidWorks. They wanted to develop a reasonably priced solid modeling MCAD product that was Windows-based, built on a standard kernel (the ParaSolids Kernel), was user-friendly and was supported through a VAR network. They have succeeded admirably, and PTC's decision to stick to Unix and thumb their nose at the rest of the CAD world has proven to be their downfall. SolidWorks now has more seats and happy users, while PTC struggles to get new users to adopt their cantankerous software that makes even the simple production of drawings difficult. <O></O>

Amen, I concur completely

 

[orac]

I don't agree. I like Pro/E. I've used GMS, Anvil, Calmain the past.


Perhaps we can do a benchmark together and you can try to convince me why Pro/E sucks.


Concerning the "sticking to Unix".
if you have developed for X-Windows it is not that straightforward to port it to Windows, e.g. the order in which events are handled is quite different.
If your first customers all use Unix, you can't say to them forget about it, from now on it is Windows only.

Best regards,


John Bijnens

 

[phoxeoy]

Interesting that this discussion is still going. I have not been on this forum for about 4 years now. Not much has changed after all these years, you still have a drawing as the soul master of data even though you can put all the data within the 3D model. Doesn't matter what CAD system you use it falls down to policy and data requirments. I'm a different animal, I deal with governement data. Commercial is different in that; they produce the data, and in what way, that best fits their needs. When the Governement says that a 3D model must have all the data required in the 3D model, it doesn't matter if it a highly complex part or a washer. It must be put in reguardless of how much time it takes. Some CAD systems handle this relationships better and Pro/E just happens to be one of them.

 

[Mindripper]

Does the US government specify that all of the data must be in the 3D model? Have they abandoned the 2D drawing system? Last time I checked, they were still using 2D PDFs of the drawings. Yeah, just like most everyone - but not everybody.


All of the customers and suppliers I have worked with for the past ten years use Windows, and all of them (including the US government) want drawings in PDF format and geometry data in some common format: DXF of 2D drawings, IGES or STEP files of models. Some truly appreciate and make use of native format, especially SolidWorks or Pro/E - although the use of Pro/E is in major decline in my corner of the world.

 

[megaladon]

My 2 cents... i've noticed in here that the people complaining about pro/e have very few posts to their name. my assumption is they are new and don't like the taste of broccoli even though it's good for them.