Adaptivity

[jmckinney3843]

Is there anyone who has had success using part adaptivity within an assembly environment?

I am attempting to use 3D assembly constraints to drive the size of the parts in my assembly. Only half of my model (1 of 2 unique adaptive parts) is working and I was hoping maybe someone else had an adaptive model that worked that they wouldn't mind sharing.

I think something that I am doing different from normal adaptivity is that both of my part sketches were created before the assembly was created. Basically I am saying I didn't use the create in place component option.

[rosterreicher]

Hi,

I'm not sure if this is exactly what your getting at but i'll offer you some input.

the tooling i do usually have 2 main subassemblies. (punch side & die side) some of the parts in each need to be adaptive. Never had much luck making each separately and being adaptive. Fo some things (like the part cavities contour) i'd make a master sketch ipt and use derived for any part that uses that contour. that way if the the contour is changed in the master sketch ipt all things with that change.
Another way i'd do some of this is to have an extra construction assembly. (at least thats what i call it. I'll only put in parts that need to be adaptive with each other. then place the parts into the normal assemblies (punch / die side etc.)
Then if i need to change it i go to the construction iam where they are adaptive.
Another thing about that is that i usually only need 1 part in the construction iam that will go in the norm iam's Basically makes it like a base part. then when that base part is in the norm iam i can base other item within the norm iam off it.

All in all it's a little cumbersome but it seems to work.
IMO the master sketch method seems to work the best. (if the master sketch is a tooth/spine contour you can't change the # of teeth, the parts that are based on it will only have the # teeth that were present when initially created. The angle between teeth will be correct but the # will be wrong)

Anyhow, Hope this is of any help for you.
Rich

[jmstines]

You might want to look into using derived parts instead of adaptive parts. You are less likely to have a part blow up when you change a dim on a different part. It takes a little more planning but if you want to have your parts update when you chage the base component it seems to work better.

You basicly create the base component and then start a new file. Go to the bottom of the menu and click derived part. I am still learning all the ins and outs of this option but I like it.

Jeff

[george.bisanz]

There's actually a whole workflow that's been developed that uses the derived part function, sometimes called skeletal or master sketch modeling. I now use it as much as I can, and if you plan it well, it's bulletproof.

Hopefully the original poster has solved his difficulty with adaptivity, since this thread started back in 2004!

[robert.templeton]

This is largely retelling what the other have said, but it might give a useful idea or two.

In my early assemblies where I used adaptivity, I added assembly constraints to hold the parts in the proper positions (I renamed those constraints so that I knew they were only for the initial setup.) Then I activated the adaptivity on the connecting parts, let them adjust their size and then I turned adaptivity off and suppressed the setup constraints. It worked, but it was tedious and I frequently had the wrong constraints suppressed.

Now, if I am creating an assembly that will be animated or under-constrained, I usually create a dummy assembly as Rich suggested. As an example, I created an assembly of a steering linkage. In my dummy assembly, I fully constrained the spindles and the steering arm. I constrained the tie rods (which were adaptive) to the spindles and the steering arm and let them adjust to the proper length. My real assembly was identical except the spindles were free to rotate and the tie rods were not adaptive. This works well when the adaptive part does not lie in a well-defined plane and it is difficult to calculate the geometry.

If I am creating an assembly and my main concern is that the parts mate with each other, use skeletal modeling (or something like it). I create a part (usually named PARAMETERS) in which I sketch the main parts of the assembly. I put all of the significant parameters for the assembly in this part too. When I create each part, I make a derived part and import the parameters. If I am modeling a wooden box, my parameters might include LENGTH, WIDTH, HEIGHT and THICKNESS. When I extrude the appropriate parts of the sketch, the distance is based on the parameters so that it updates too. To change the size of the box, I edit the values in PARAMETERS and watch the assembly file update. I usually link PARAMETERS to the assembly file so that the values are available at the assembly level.

[george.bisanz]

Robert,

That's a nice summation of both techniques. Each has its place, and experience will often determine which would be best. I've also heard about a new way of creating cross-part relationships, by means of copied surfaces. It works like adaptivity, but is supposedly more robust.

I can't wait to get IV 2008 to try it out!

[jmckinney3843]

There's actually a whole workflow that's been developed that uses the derived part function, sometimes called skeletal or master sketch modeling. I now use it as much as I can, and if you plan it well, it's bulletproof.

Hopefully the original poster has solved his difficulty with adaptivity, since this thread started back in 2004!

It is funny that this thread has been updated recently. I haven't been in the forums for at least a year now and I had forgotten about this post. After much frustration and unexplained behavior resulting from adaptivity techniques using unconstrained parts I finally threw in the towel.

I am however very interested in readin more on the skeletal modeling technique. I have heard this technique is powerful. Does any one have documentation or examples that further detail the fundamentals of this process?

[robert.templeton]

I had not realized that we were responding to an old post; that shows how well I read.

Skeletal modelling is a neat and wonderful way to plan your assembly, lay out the basic framework and still be able to edit it later. Sean Dotson has a tutorial on skeletal modeling at http://www<dot>sdotson<dot>com/freetut/introduction to skeletal modeling<dot>pdf. He has a group of free tutorials at http://www<dot>sdotson<dot>com/tutorials<dot>asp and at http://www<dot>mcadforums<dot>com/fo...m<dot>php?f=41; they are good sites to help you learn new techniques and tricks.

[nick.kuzmik]

Can I get a bit more help on this topic? I'm missing something that is probably fairly obvious.

I'm building an assembly consisting of a cylinder, 1"diam x 1.8" and a 2" x 2" plate. The two pieces are connected by a pair of 1/4-20 screws with thru holes in the plate and tapped holes in the cylinder.

How can I make it so that the spacing of the threaded holes in cylinder.ipt drives the spacing of the thru holes in plate.ipt?

I know this is beyond trivial, but I'm experimenting in this simple case.

[mflayler]

Well if they are holes and bolts I would use the Bolted Connection command to create them. It is a way to create the holes, fasteners and sub assembly right inside your IAM (and it will propagate to your part files). It will require an update when things are changed but can be set to automatically update. (Right Click on Bolted Connection sub assembly and select component --> Automatic Solve)