I have a very simple configured part (din rail of different lengths)
But if you use the temporary axis at the center of thes slot to drive holes in an assembly - those holse move randomly when a different configuration is selected.
I am ready to give up on configurations and want to have each configuration saved to its own static file. Is there an easy way to do this?
Do I need to write a macro?
Giving up on configs would be giving up one of the core features which makes SW so versatile.
Instead of using the temp axes, create and use reference axes. They appear to be more stable and can be more easily selected via the Feature Manager.
The problem of mating holes to a slot has been raised here before, and I believe several methods were given.
I'm not seeing that here. I took your rail part and added it to an assy with another part. Then I put a hole in that other part that had a pierce relation between the center of the hole and the center temp axis of the first full slot at the start end of the rail. Then I went to the rail part and changed the active config. Back to the assy - no change. Back and forth several times, no difference.
Not sure what version you are on - I'm running 2009 SP4.1.
Another thought for you, though. As stated, it would be a real shame to give up on configs. We use them a lot and don't see many problems. Might I suggest that instead of driving the hole with the slot, you position the rail to the hole. Therefore you put the holes in the plate with dims and mate the rail to one of the holes to position the rail. That way you don't have to have the in-context relation which should make your problem go away, but yet the holes and the rail are still tied together. If you want to reposition the rail, change the hole dim.
WT
Try this assy - the holes wer created relative to the 3" in config.
Changing back to 3" puts them where they belong, changin to anything else moves them around...
Rleations are pierce for one point, and coincident for the other
Kevin,
Your problems are not related to configurations, or to any behavior of the software (SolidWorks), the base cause of the problem is an improper modeling technique, unfortunately, SW allows many ways to do the same thing, but not all of them are the proper and correct way of doing them.
If I get some free time (at this time I do not have any, but I will do my best), and you send me some information regarding the objective of your model (see below) I will do for you a model of what you are trying to do with the proper estructure, so you can see onother way to skin the same cat but a little more solid and completelly stable.
The information I need from you is the follwing:
I assume that the rail will change to whatever dimensions you need (any of the variables), however, regarding the fastener threaded holes on the back plate, are those allways two holes or you need those to be also a vairble number?. How about the position of those holes, are any conditions there? in other words, do you want the ability to crete those holes on any position as they relate to the slots (for example every slot or every 2, 3 or 4 slots), and also want to change the number of holes? If you explain to me a little more in detail the purpose of the back plate, I may be able to do for you an assembly with all the functionality you may need without using all those mates and all those configurations, and which will be able to be changed into whatever configuration you may need, by just entering some parameters in a configuration table.
By the way, next time you post and assembly, you can PACK it inside a ZIP file directly from inside the application, that produce a stronger way to transfer files which are linked to each other.
While there may be better wasy of getting the job done, this behaviour is a bug. There is no way that in context holes should appear off the part when a different length of din rail is chosen.
- the two holes in this model are only shown to illustrate the bug, thier location is not special
- in practice the number of holes will vary depending on the overall length of the rail,
- the back plate and the rail are never defined together - the back plate is much larger and supports many other components, the rail is added later wherever it fits
- holes are typically inserted by hand using hole wizard by a designer, usually 3-5 slot spacing between holes
My goal is a stable din rail model so the user can pick the length, and insert holes using hole wizard. You design table approach is worth looking at, but may not be applicable to the way we do business. I'd appreciate any input you have though - if it doesn't aply here, it may be usefull else where.
My company's main objection to configurations is the problem of stability, in a multi-user environment, where not all users share teh same skill, they offer far more ways of blowing up assemblies that simple part models.
Kevin,
The behaviour you are experiment on your model is not a bug, as mentioned, in my opinion is the result of improper modeling technique, let me explain...
If I am correct (I do not know the internal code of SW), the internal ID of a face change for each different configuration as you create more or less number of slots, in your model, the position of the holes is related to the axis of the face, so as you change the configuration, the relation is loss.
Attached is a ZIP file containing an assembly of your Rail and Backplate, I modeled this Top-Down with the correct sketches and relations, therefore it works considering a couple of important "limitations" (defective and/or incomplete functionality, in my language) in SW, which I will explain at the end of this post.
To see these assembly working you will need to do the following:
1.- Open the assembly (fully resolved)
2.- Open the "Din Rail Phoenix Contact 35mm Wide" part "in-context" in his own window.
3.- Activate a particular configuration (any).
4.- Execute a Rebuild and leaving this part open (you can "save" but do not "close"), change to the assembly window.
5.- In the assembly window, open the "Components properties" window of the "Din Rail Phoenix Contact 35mm Wide" part and select the same configuration as the one you have "active" on the part on his own window, and press OK,
As the Components Properties Window close, you will see the backplate changing to reflect the new configuration, including the fastener holes on the correct position and the correct number of holes (as specified inside the design table for the "Din Rail Phoenix Contact 35mm Wide" part, which I modified a little).
You can analyze how I builded the model and what is different from your method, notice the way the assembly is "mated", also take a look at the changes made to the design table (which can be improved to be fully automated with a "configurator" data input, but at this time I do not have the time to do it).
The SW "limitations" mentioned before are:
1.- SW fail to pass to the assembly model the correct values for variables inside the particular configuration selected inside the assembly if the part is not open on his own window and it have active the same configuration (in fact the part shall be open and with the particular configuration "active" before that configuration is selected on the part inside the assembly for SW to "pass" the correct value to the assembly equations or relations.
2.- The equations module of SW fail to "Extract" values from inside the configuration selected for a part inside an assembly if that configuration is not open on his own window.
I believe that I may found a workaround to this problem (the "limitations") by using a little different technique to "link" the values for variables inside configurations to the "bridge" and the "external" parts (in this case the plate) in order to avoid the need to do the aforementioned process (1 to 5). I will try this other approach and if it works I will let you know. For the moment, the attached assembly (which is correctly modeled as a "Hybrid" model) works by doing the indicated process, since SW have the aforementioned problem (which I reported to the VAR back in 2006). So if you want this to work without the need for that process, report the problem to SW and wait or have another 500 people to report the same.
The technique can be used to place many Rails on a single plate, all you need to do is to properly structure the "Bridge" geometry (inside the sketches on the "Ref. Geometry" part. As you see, there are other ways to skin a cat...
Hope this is of some help to you.
I am unable to check SW at this time, so I can't see what you have done - but the behavior I have seen is definnately not accepable.
While I agree using temporary axis as reference points is not good - The "normal" error would be that the new holes would appear on some other tempory axis - for example if I assign a mate to Axis1 and Axis5, which happen to be in holes 1 and 2 - then when the part is rebuilt with a different length, the new locations of the temporary axis will be undetermined, but should still remain in one of the slots.
If you look at the original assembly, you see that they wind up completely outside the envelope of teh DIN rail - this is not well defined behaviour.
Unfortunately I cannot use a top down design approach, these parts are used in coombinations where the numer anr type of compnents are undetermined in advance.
However - I am interested in learning more about top down design for teh actual components on the din rail, that is often a present combinatino of x sets of parts A, B and C
I agree that the behavior you are experiment is not normal, since the "relation" is completelly lost, but this is a problem with SW when you have multiple configurations (since in reality each configuration is a complete new set of data inside the SW DB's). Aslo, the problem I descrive forcing the process explained in my previous post is the result of a SW "limitation" (they call "limitation" when a funtionality is imcomplete or defective).
Take a look of the assembly I post there and see how it is done, the same tecnique can be used for an undetermined number of rails and on undetermined positions, is a matter of how you estructure the sketches and equations in order to generate the features (fastener holes) using "dynamic" patterns of a single "seed" hole. I was planning to show you that (that is why you will see inside the model a reference to "Rail-01" as the "seed" element. But unfortunatelly I only have a couple of hours to do the example, however if you know how to model top-down you will be able to figure how it can be done, the objective of the model was to show you how to model using a "bridge" to pass data from one part (the Rail) to another (the Plate) in a solid way where the relations can not be loss (with the prevously mentioned SW "limitation").
Kevin,
Check out this thread.
Extracting Configurations
Regards,