How can I drive one parts dimensions based on another part's features or dimensions?
Caveat: There is motion involved so convert entities and inplace mates are out
Use global variables and equations to link dimension. So instead of converting an edge, which essentially has to remain coincident the "derived part" is simply referencing the number value instead of any location.
Is there an efficient way of doing this? Or do I really have to go through each individual dimension and create a global variable for it. Is there a way I can declate something as a global variable or a "modifiable" dimension from the part level?
Maybe you could give us a more high level view of what you are trying to do?
There are several ways to derive parts - you and insert a part into a new part and then have everything the original part has and then added new geometry or change the original geometry by modifying with features. You can also spit parts>out to derive parts. I'm asking because I don't think you want to link every sketch and feature dimension via global variable. Are you trying to create to identical parts but one is slightly different?
Thanks for the replies; I really needed help with this.
Anyway I have attached a very crude example of what I am trying to do. For simplicity I have made only two parts. In the assembly there is a housing (which can rotate) and a piston. The piston is obviously the incorrect size to fit in the groove for the housing. What I would like to do is create some sort of parametric relation between the housing and the piston for all of the places where they interface. I would also like to be able to resize and change the dimensions of one of the components and have the other update. What would be the best way of doing this from a top down approach?
Aparently the server is having issues with my upload but again any help in the mean time would be great
now the server seems to be accepting files example attached
Message was edited by: alex Bradley
You can also insert part A as the first feature of part B, build part B as a separate body or bodies, and finally delete the part A body at the bottom of the feature tree.
That sounds interesting Is there a place where I can find an example or a tutorial of how to insert these features?
alex Bradley wrote: That sounds interesting Is there a place where I can find an example or a tutorial of how to insert these features?
alex Bradley wrote:
I had an image for you, but it wouldn't upload. Anyway, it's easy. Hit the "Insert" drop down menu, look at the center of the list, select "Part...", browse to the file you want, and then finish with any moves or contraints that you want.
You then have all the geometry, sketches, etc. from the first model to reference in your second model. Any changes in the first model are parametric in the second model. Should you open the second model sometime later, SW will look to see if the first model is open. If it is, then any changes in the first model get passed to the second model. If the first model is not open, then the second model uses the geometry it saved.
(perhaps rephrasing Dwight here) So, Depends on who you talk to and their "bent" but sounds like you want to create a multi-body part file then at the end of the Feature Manager and then at the end of your Feature Manager do a "Save Bodies" to save all the bodies out and then back into a main assembly.
You basically have two main choices:
Multibody modeling often is used to separate motion from the "design-as" intent of the parts (because you can't do dynamic motion in a part.) This is an alternative to "top-down" design - i.e. in context design in the assembly. An additional benefit of Master modeling approach is that it is void of the long rebuild times that a top-down complex assembly can exhibit (because one change effects rebuild of all components in the assembly.)
So, i would use multi-body modeling to model all your "qausi-components" as separate bodies, then use the Save bodies command to derive the bodies out of the part as derived parts of the master base part file that you created.
Once all the derived and placed back into the main assembly, you can assign all your mates and drag freely or using with motion simulation, while still being able to modify and refine your design via the master base part.
You can read more about it in the SW help file as well as we have some good multi-body and master model tutorials.
Any particular advantage to deriving out parts, versus using insert part? I know there are a few ways to do the same thing, but with some subtle differences. In fact, I remember seeing a table that described the various options, but I can't find it. Looked in help and searched the forum. I'd like to look at that again.
There was a great video from last year's SW World (2011) that covered the pros and cons of the various methods of saving bodies out in the Master Model technique. And like forgetting a joke's punchline, I can't recall the presentor and can't seem to find links on this website to the old presentations.
Anyone have better specifics and a link to the presentation?
EDIT: I think it might have been the "Master Model for Everyone" presentation by Christohper Castle at SWW2011. As noted in another current discussion, these videos can be found on the open web through a google search.
No part particular advantage, it all depends on what Alex wants to do. If he wants to model all his moveable parts in context of one another, he can do it as one multi-body part and then save bodies to an assembly where he can apply motion. Otherwise, he can build all the parts in context of one another in the assembly (as separate parts, but with in-context refs to the other parts, thru the assembly.
If he just wants to create two parts, and one in succession to the other, then insert a part into a part.
If you are determined to do in-context design, you can create sketches at the assembly level, and control your parts off of these. You will have to think about what you are doing, but an assembly level sketch can account for the current position of everythng.
Man these replies really helped out and made my life so much easier. They didn’t teach us any of this stuff in school it was all just bottom up design. Anyway let me see if I understand my options. The first is creating a layout sketch in the assembly and then create the moving parts off that, but the consequences are that I can’t have part’s features interact with one another. The second is that I create 1 part for design intent and then save the bodies separately and put the parts that the bodies create in the assembly for motion studies etc. The last is that I insert one part into another and model in context at the part level and then put it together at the assembly level.
I tried the third approach as it made the most sense for the way I think; at least for this particular project. One problem I did run into is when I deleted a feature in one model had to propagate that change (by deleting references) to the subsequent models. So although everything updates and works together (which is awesome) it does beg the question if I delete a feature all together will I have to keep chasing the changes around? Would that then make the 1 master part split body approach mark mentioned a better option? Does anyone here favor one approach over another and if so what is the reasoning?
Retrieving data ...