Continuous Improvement

Some time ago, I publicized an entry about an axis of a pipe, in which, I tried to explain in some way, how to automate the process of marking out that curve. That method saves a lot of time – because the procedure itself involves some clicking. And now a question, can we do better? That template was already stabilized, so we didn’t have to worry about directions of inputs, but in order to mark it out, we had to indicate the edge and surface of the conduit (pipe). Here’s an example of a tuned template, which will require half less work from us, meaning, marking out only the surface of the conduit:

The starting point was to lose the necessity to indicate the edge of the pipe. It can be done using the Boundary command, which will always give ambiguous result in the form of two circles, out of which we would have to pick one, using Near operation (for example a point with 0,0,0 coordinates).

Everything’s wonderful, until we come upon a case, such as this one :

Meaning a symmetrical piece of a pipe, relative to our point used for the Near… When it’s like that, Near becomes “stupefied”… 🙂 And what now? Holders of HD2 licenses have no other option, but to live with it, or decide that due to those cases, it is especially justified to use a template with a higher amount of inputs – more clicking, but also better versatility of a template. Possibly even define the parameters in control of coordinates of our point, which we can publish in the template, and in case of failure – redefine coordinates on a template level.

But can we do it yet differently, have all the described advantages, meaning minimum clicking and versatility? Of course we can, and that’s the case that justifies the use of KWA license…

We will define the edge of the pipe by starting over, in the following way, i.e., instead of Extract (which was “called for” by the previous template), we will create a curve controlled by a rule :


Its syntax looks like this :

If the Near operation “yells” about an error – the program automatically uses a point shifted to a certain degree, and the procedure carries on. When using this type of rules, don’t forget to tick off “do not catch evaluation error” in that rule’s Properties.

And can we improve such template even further? Probably yes, but unfortunately, it needs to be economically justified. Because creation of superb versatile templates is a very creative work (no one yet seen through everything that CATIA has to offer…), and hence, very labor-intensive…



Comments from : 18 december 2016 22:12

Hey, thank you for another interesting post. When you add further improvements to this tool, do you also analyze the influence on update time and file size? I understand the idea of continuous improvement, but this aspect is also interesting for me. In case we need to use a function several times in a single file, it may become important. To improve my efficiency, I prepared a small tool for creating and checking spot welds. First I decided to do it as an UDF, I think it was the most “elegant” way. Later, I achieved exactly the same effect using Power Copy, weight of the template file decreased from 138 KB to 105 KB. This is not a big difference, but since I always use this template dozens times, I decided to use a template prepared as a PC. Sometimes, in practice, I would rather be forced to indicate the orientation of the surface or additional edge if it would save me the file size and update time. Of course I understand and support these “inner need” to improve and squeeze the further functionalities from CATIA :).

Regards, Mateusz




Łukasz 19 december 2016 08:25

To be honest, I have never solved the issues of size of the file in case of templates. I only looked into the change of weight of the file depending on the spline curve of its projections in order to discredit a statement(I will describe it with spline approximation title soon) . In order to decrease the file weight, I would go to description of the procedures by rules (it is just a text that CATIA  interprets each time), but some day it will be necessary to compare the weights of two the same procedures i.e. typed and click-do. But the weight of the template itself is not a problem – it is mostly time saving, which switches off thinking and making errors. I don’t know how about certain OEMS, but I remember times when there was a weight limit for part (MERCEDES) and if it had too big weight PLM cell appeared which checked the model construction methodology. Recently the majority of suppliers provides steps.  So the approach for a model to have small weight is no longer a requirement but common sense, so as not to hurt oneself… and the weight depends on the level of complication of the model – as many constructors as many approaches.

I have noticed one thing – the models of CATIA  V4 specialists have small weight…






Comments: Be first

Recent Posts

I welcome everyone interested in CATIA . I will […]

Each Part in CATIA has its own UUID identification. It’s […]

The matter might seems trivial… it gets a bit […]

In the previous entry, I have described some of […]

And now an entry for Fetishistic B-Rep Dodgers 🙂 […]