Effects of Constraints

[shift1313]

Aardvark, i would like to continue the discussion about constraints in a sketch with you so that I can fully understand.

 

Id like to know more about the implications of using the Fixed constraint in a sketch has outside of the sketch in an assembly.

 

here is your a quote from the other thread

"Placing a fix constraint in a sketch has a huge effect outside of the sketch! If you fix a point and it is not attached logically to the origin then trying to use the origin for mirroring, patterning, extruding, etc is useless. The origin planes are the backbone of your design, they are inviolate and cannot be removed. If you use them to design your parts you will have healthier parts and assemblies.

 

If you have a spherical part that is not constrained to the origin how do you constrain it to anything else? Pretty much your only option is to use a tangent constrain which is a poor choice at best.

 

I can create an assembly made of parts logically designed about their origin axis that can be assembled constraining only to their origin axis that is impossible to break.

 

Fix is just a lazy way of fully constraining a sketch that will cause problems for you later."

 

 

I dont typically use fix in my sketches but I have used it before and havent had any ill reprocutions (yet:)) so I would appreciate more info on the subject to make my designs better in the future. You mention using the origin for mirroring but i typically create a work plane and mirror about that. Also you mentioned using fix in a sketch and not being about to constrain it outside of the sketch?

 

 

also along the same subject but slightly different, when you mirror something, then say you go to make a presentation file of the assembly or disassembly of the file, moving one of the mirrored components moves the other. Is there a way to disassociate them? It's only happened to me once and it was an instance where i copied sub assemblies in a larger assembly.

 

Thanks

[Aardvark]

Why create a new plane when you have one that is more stable already in place?

 

Everything in your model tree is dependent on the features above it. Having a new plane means that that plane has to be associated to something else, right? If you planned your model correctly you could just mirror about the origin and then no matter what changes you make to the part your mirror will still function. Have you ever tried to go back and change the location of a line or other entity in a sketch and had several features created after that one fail? This can be limited if you constrain to the origin planes.

 

Specifically using fix is not the problem but rather a symptom of the problem. A fix constraint has no intelligence or intent. Also they tend to get hidden and hard to find again if you want to change the design.

 

Have you read JD Mather's Autodesk University class "Becoming an Autodesk Inventor® Professor in 90 Minutes" notes? If not I highly recommend it because it has some very good advice. Specifically pages 5-13 discuss sketching hows and whys. Here is the link:

http://home.pct.edu/~jmather/AU2007/MA105-1L%20Mather.pdf

[Aardvark]

I don't have Inventor in front of me right now so I don't know if I can answer your mirrored part question ( I have to use CATIA at my current job). Sorry. There should be a button to break the links to the original parts though because you are essentially creating a derived component when mirroring a part. Perhaps someone else can answer that one, I never used derived components much or mirror component for that matter when I was working with IV.

[shift1313]

david, this is the design i specifically had mirrored components in because everything symetrical. In this instance there was no mirroring in a sketch plane but rather mirroring of components. This worked for the animation as well(except the cylinders that are at an angle to the body). These had to be copies with their own constraints since they were not perpendicular to eachother. I had to revise this file several times to change the base plate width for different width rollers, bearings etc and what i had to do was change the offset of my mirror plane(which was high up in my hiearchy) and all the mirrored components shifted. Im sure this wasnt the best way to do this but it was how i found to work. I have no training on inventor and ive just fumbled through to aclimate myself so I will definately go and read that document you linked me to. THANKS!

 

To answer you question about going back and having something fail, yes i have had things fail because of constraints, if i modify a sketch, a constraint in the assembly that was based on, lets say a face, has given me an error stating geometry which was used no longer exists. In those cases i just delete the bad constraint and redo the constraint to the new face. I dont recall having a feature fail.

 

 

I will have to read through that document later tonight and see how wrong i am doing things:D

[Aardvark]

Look into "flexible" assemblies. Leave the throw length in the cylinder assemblies unconstrained, then in the upper level assembly, RMB the cylinders and make them "flexible" (not Adaptive). You can then constrain the ends of the cylinders independently and they will adjust in length but it will not effect the original definition.

[shift1313]

how would you constrain the cylinder body/piston? I used a pin constraint and when i annimate i change the offset number as my stroke. If i dont constrain this? how will i get the cylinder body/piston together?

[Aardvark]

Just use a centerline mate. Don't worry that the piston can exit the cylinder body this will be controlled in the top level assembly.