Getting Started

(or, How to Keep From Feeling Intimidated by a New Design Project)

Editor’s note: Our engineers here at RL Hudson use Pro/ENGINEER® solid modeling software to design parts, and we know that many of our customers’ engineering departments also rely on Pro/E, which is, after all, the standard by which all other modeling programs are judged. With that in mind, we at Solutions have enlisted two Pro/E experts — Benny Foreman and Bruce Sumpter — to share some design insights with our readers. Both Benny and Bruce served as application engineers and independent consultants for PTC (the company that developed Pro/E) prior to joining our engineering department here at RL Hudson.

sketched curve Benny Foreman graduated from the University of Oklahoma in 1995 with a Bachelor’s Degree in Mechanical Engineering. Since then, he has split his time between working as a manufacturing engineer and as an application engineer designing manufacturing software. Benny has been part of RL Hudson’s engineering team since 2003.
Bruce Sumpter earned a Master’s Degree in Mechanical Engineering from Oklahoma State University in 1985. He now has two decades of experience working with engineering and enterprise-level software solutions. Bruce joined RL Hudson’s engineering team in 2004.

Gentlemen, take it away…

Bruce: For me, the first step in any new design project is to identify the most important features of the part. Usually these are related to its function and to how it connects with other parts.

Benny: I use the same strategy. Even if other aspects of the part change in the future, the connections probably won’t, so I feel safe using those as starting points. I start with the certainties and leave the features that seem most likely to change for later.

Bruce: For complex parts, I like to start with what is called a sketched datum curve, which is essentially a 2-D sketch inside the 3-D part. You can think of the sketched curve as a basic backbone or skeleton for the part.

bad sketch Benny: The alternative to using a sketched curve is to just jump in and start modeling specific features of the part, but that makes it easy to get bogged down early in complex geometry. A curve is a nice simple shape that can be easily modified to add specific features.

Bruce: And as you add features, those are tracked within Pro/E via what is known as a “model tree,” which is essentially an ordering of the separate features that go together to make that part.

Benny: The model tree is kind of like a recipe. I like to use this example: If you’re making gravy, you don’t cook the milk and add the flour and other ingredients afterwards. You prepare the flour mixture first, then add the milk. You get really bad gravy if you prepare it in the wrong order. Just ask my sister!

Bruce: Our first column, and the family insults have already started!

Benny: My bad. But her gravy was terrible!

Bruce: Speaking of family relationships, Pro/E tracks what are known as parent-child relationships among features. In other words, it tracks which features are based on other features. For example, if I model a pocket, then model a boss that’s dimensioned to an edge of that pocket, those two features are now related. If I later move the pocket, the boss will move with it because they’re linked.

good sketch Benny: That’s meant to be a help, and it can be, but new users often link features that don’t need to be – or shouldn’t be – linked. So you really have to think about which features should be related, and which shouldn’t, and get rid of any unnecessary links. If you’re not careful, unwanted links can become like visiting relatives who overstay their welcome!

Bruce: And eat up all the leftovers!

Benny: That too!

Bruce: But you’re right; excessive links can cause big headaches.

Benny: And it’s easy for unnecessary links to stay in. When you create a new feature, Pro/E often links it to some previously created feature. Unless you consciously go in and tell the program not to maintain that relationship, it’s there and can cause confusion later.

Bruce: But you can use the model tree to identify relationships. All you do is select a feature within the tree and drag it up (or down). Doing so will highlight what other features it’s linked to: its parents (or children). But you don’t have to actually move the feature up or down the tree. In fact, you probably don’t want to do that; doing so might very well change your model!

Benny: But if you’ve seen any unnecessary linkages, you can then go in and delete those.

Bruce: Exactly. Another thing I do to cut down on confusion is rename features with a logical name rather than just going with what the program assigns.

Benny: I do that a lot, too, especially on main features, like threaded ends. An average part for us here might have thirty to forty features, so it’s important to be able to find the main features easily.

Bruce: Consolidating the number of features can help, too. Sometimes ten features can actually be one feature if modeled carefully.

Benny: But it’s easy to get in a hurry, and I’ve been guilty of that. Sometimes I just want to make a quick design change, but I have to remind myself to take a step back, get rid of the feature or features I don’t want, then create the new feature. Otherwise you can just be adding lots of unnecessary stuff to the design, which is called “feature creep.”

Bruce: Very true. So, in summary: Start with the main features, and the features least likely to change. Be aware of how the various features of your design are related, and get rid of unnecessary links. And model as efficiently as possible to cut down on future confusion.

Benny: And let us hear from you! If you have a Pro/E question – or a great suggestion for how to get the most out of the program – send it to us at benny&bruce@rlhudson.com. We’ll answer submitted questions in future columns, plus we’ll talk more about a variety of Pro/E issues.

Pro/E® and Pro/ENGINEER® are registered trademarks of Parametric Technology Corporation (PTC).