In general I like Solid Edge and bought it after comparing it to other programs. As I have mentioned before the things that brought me to it were primarily sheet metal, good MCAD and direct editing.
Recently while talking to another SE user about Mastercam I was asked about hole tolerances. I did have a lot to say and this is one of my biggest pet peeves with SE as it was for this company. You see, what we do with geometry after creation is to then go and actualy produce parts. Oddly enough we think this is the primary reason for the existance of CAD.
This has been brought up before on the forums and basically at this time it is SE’s position that it is not important. This is the problem.
Lets take
a one inch shaft with a 1 x 12TPI.
You see we have in this drawing taken directly from a part created today a 1″ dimension for the major thread size at the threaded end of the shaft. Without taking the time to look the data up all I can say is trust me, the true diameter of the threaded portion of this shaft for manufacture is less than 1″. So now if I apply the correct diameter for a 1″ x 12 tpi thread to the end of the shaft I can no longer use the thread command to apply here and get the joy of those cheesy kinda thread looking things. If I want a threaded shaft in SE, I HAVE to send the actual producer bad information. I don’t have a choice here. Either they get a reduced step on the shaft with no threads or a full sized shaft that recognizes the thread minor but not major dimensions. Same thing is true with threaded holes.
So you get stuff sent to machine shops with notes that say things like “even though we can’t properly depict this pretend here with us for a bit”, or words to that effect anyway. Can you see the potential for scrap and confusion here? There is not one shop involved in manufacturing that has not trashed parts because of this. Same thing with holes. So now you stick a note next to it and hope.
What this does for machinists who work off of 3d geometry is not trivial. I can’t create parts that are true to the design intent and use this same part in a cam plan without problems. So now I have to create TWO sets of parts, one with real dims that does not “show” cheesy threads and one that does show cheesy threads but not correct dims.
The subsequent drawings are handled the same way with two parts included, both with aspects which when combined will give the correct data or one drawing with notes about how it really needs to be done.
This is a PITA for everyone who actualy makes parts. The only reason I inflict cad and cam software upon myself is to earn a living and I bet the same is true for you. So why is it so hard for cad authors to sit down with real producers and fix some of these things? I don’t know. In SE’s case I know in many areas they do spend lots of time with real users to find out what they want and need. Then there is this disconnect with true manufacturing dimensions for threaded parts where they know of user complaints and just flat out don’t think it is important. I am not kidding and this has been the response from SE on the BBS Forums when this topic comes up.
Last time I checked threads were still a method used to put things together with. Oddly enough my customers still expect me to produce parts with threads and I wonder how the software gurus figure I ought to expeditiously do this? On the whole, especially when compared to the problems users of other cad software face, I am satisfied with SE. But it still irritates the heck out of me when cad companies choose not to regard the needs of the people who build things. I know darned well that it can be fixed with not much time on their end. If I could add up all the time we cad and cam users waste each year in dealing with this stupidity every time every day, I am quite sure our expenses far exceed the expense SE would have to spend just once to make it right. Happy customers are the best sales tool you could ask for I should think.
Machined Audio Horn Lens 
You can edit the values in the holes.txt file to suit your manufacturing needs. We change the minor diameters for both our tapped holes and external threaded diameters.
Yes each individual user can do this as a workaround. In your experience how does this effect actual 3D parts for the purpose of CAM use? How does this work for drawings insofar as showing exact manufacturing data derived from these parts?
In my experience, as in the company that prompted this post, it does not happen. It is just another workaround we should not have to do. You add up all the time each user would have to spend to append and amend what should be in there to begin with and it is not trivial.
I know you are offering a solution, and I appreciate the reply and tip. This is the same point brought up on the forums where the general consensus is that this is a problem. This is among those who actually design and then build. My main point however is that we should not even think of having to do this with software whose sole existance is predicated on the basis of design for manufacturing.
I think I am the SE user you are referring to. Our goals are pretty simple: reduce mistakes, reduce time to market, and improve quality. My main concern is with class-fit features, which are modeled and drawn using nominal dimensions. Class-fit tolerancing is widely used in Europe and eliminates guesswork when it comes to specifying fits. The tolerances are usually “plus-plus” or “minus-minus” meaning the actual feature dimension is something other than nominal. But, to get the drawing to show design intent, you must model with nominal, not actual dimensions. This forces the CNC programmer to compensate. We use Mastercam, which has no built-in tools to either recognize or specify class-fit allowances. Many of the parts we do are simple plates with a multitude of class fit features, such as spotfaces, that require circular interpolation. For us, it would be ideal if the Solid Edge model reflected reality rather than nominal. The CNC programmer wouldn’t need to put in an offset, so mistakes would be reduced, programming time decreased, and quality would improve. Sounds like a win all around. It seems like this would be a relatively easy enhancement for SE to implement as the class-fit data already exists in the “SE-LimitsAndFitsTableISO.txt” file supplied with Solid Edge.