Spring compresion in motion simulation in NX5 3

[akrah]

Hi,

I am trying to simulate the motion of some components in an assembly. The assembly includes a spring as well.
Is it possible to simulate the compression of the spring in motion simulation.

Looking forward to any input in this regard

 

[hudson888]

Koen11,

No simulation files are involved just positioning constraints or mating conditions, deformable parts and a key frame animation to iterate the default expression FrameNumber. Its all in the files that were posted.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[Koen11]

Thank You for help, I think we are getting closer to the solution The last thing is - how to "merge" the motion from motion simulation, with the motion of the spring so that I can record a movie with both motions at the same time? I managed to save the motion simulation as a sequence so now i can view every step (frame) in the Modeling application. Now, is there any way to update spring position (deformation) in every step? Or is there a completely different way to achieve the result that me and akrah are looking for?
Best Regards, Koen

 

[hudson888]

You Don't need motion simulation at all it is just done with the expression FrameNumber and a changing distance mating condition.

When you set up the animation tick on "update expressions" all it does and in capable of doing is to increment the expression called FrameNumber by 1 for each successive frame of the animation.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[Koen11]

Yes you are right in this case, but the actual question was related to engine assembly which has got some more complex relations impossible to create with assembly constraints/mating conditions so motion of main parts (all excluding springs) have to be "imported" somehow from motion simulation application.
Best Regards, Koen

 

[hudson888]

Take a look at the videos page of my website. The engine there is also reasonably complex. The motion is nevertheless achieved by the use of mating conditions and expressions. The only trick involved is to articulate it by rotating the flywheel or crankshaft. If you try to do it by reciprocating the pistons they always get stuck at top dead centre (i.e. either end of the stroke).

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[Page3]

Hello Hudson 888,

I have taken apart your example, and can recreate everything. Accept this last expression... I have attached a screenshot of the item in question.
I am running NX5, and the screenshot is in assemblies.

this expression should be the last piece to the puzzle,
any help would be greatly appreciated

 

[hudson888]

Page3,

Please take a look at my post above dated 11 Mar 09 19:20 since I found after much going back and forth with the other posters that it is likely that what works with mating conditions in this instance will not operate correctly when you employ positioning constraints instead. I have yet to hear anything from other posters to indicate that if done differently it can be yet be achieved.

My example works when I use mating conditions. Then when I change the customer default settings in NX-5 and allow it to automatically convert the mating conditions it still works. However when I try to construct the same example from scratch then it will not allow me to build the expression that links the length of the deformed spring with the measurement between the two ends.

You appear to have hit the same hurdle my earlier posted example (8 Mar 09 20:00) using mating conditions may be better suited to your needs if you're not already heavily vested in positioning constraints. Sorry if you feel you were mislead it appears that I had assumed that the functionality would be equivalent and indeed I was surprised when I discovered what appeared to be a deficiency.

I have suggested that if this is the case then I would obviously request that this functionality be catered for and a fix or an enhancement to bring the level of positioning constraints capability into line with what I'm able to do with mating conditions. I can only suggest that I would be more than happy to welcome anybody who proves me wrong on this. Maybe it is I who am using positioning constraints wrongly? Thus far all is quiet.

P.S You're better to attach Jpegs to your posts in the future since the bitmap format is larger and therefore less efficient.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[Page3]

Hudson,

I just found the mystery expression, If you right click the spring and select 'deform' a window pops up and you can select the deformable part created when first defining the spring. then attach it to the distance measurement.

Now that I can deform the spring, I can apply it to the big project

I have attached the assembly I made while following along with yours. It is very very simple, just to show the springs deformation.


thank you Hudson for responding so quickly, and for providing such a solid example to use in the first place

 

[hudson888]

Page3,

Yes I'd very much expect that example to work despite what I said earlier. I hope we get to the bottom of that other thing but for what you're doing I can see it working quite easily.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[akrah]

Hello Hudson,

I used positioning constraint in my assembly and it works perfectly. The problem I had regarding not being able to deform the spring during the animation was that I defined the spring to be deformed in a specific range in the third step of defining the deformable part.
Thanks to you for your help in this regard.
By the way, I am still not able to animate the deformation in motion simulation. Is there any further help on the way

 

[hudson888]

akrah,

I don't know whether motion simulation actually has any other means by which it would be better to deviate the spring. In other words you might constrain all the rest of the assembly with the motion solver for whatever reason, and yet still employ a measurement between the elements at either end of the spring to drive the deformation.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[Koen11]

I think its not possible in motion simulation module from the simple reason - expressions from basic assembly (like spring height or measurments) are not being updated in motion solver.

Hudson,

Nice website, I wish I had such job after I graduate
Anyway back on topic, I managed to animate all gears in my assembly using angular constraints and framenumber expression, but I have a problem with cam followers touching cam ring (see attachment). Its not a problem in motion simulation (Im using curve on curve constraint) but in assembly, I can't set a constrain betweet cam follower surface and cam ring surface, because I can select only one face on cam ring, and it consists of 2 faces. Do you have any advices how do to this?

Best Regards, Koen

 

[cowski]

Disclaimer: I have never applied constraints to a cam and follower(s). If the problem is that the cam consists of 2 faces, here is what I would try: you have 2 tracks on the cam, can you split that into 2 cams so that each side has its own track? If so, you are still left with the problem that each track will be made of multiple faces. To minimize this you can remodel the outside shape of your cam track with spline(s) (or use join curves on the geometry you have - essentially the same idea) so that it is only 1 face.

 

[Koen11]

cowski,
Your solution works perfectly! I extracted the edge of my cam, joined curves, extruded new shape and moved to reference set. I dont know how to thank you guys (hudson888 and cowski), you were really helpful
I'll post a link to the movie when I put everything together.

Best Regards, Koen

 

[hudson888]

I would have suggested the same. For an animation you'd create a piece of construction geometry for mating (constraining) the CAM contact and hide it out of view on another layer and/or with reference sets so that you get around that single face requirement.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[eex23]

Hi all,

I read this thread accurately. I experienced the same problems with springs both with animation or as deformable part in the assembly. Everything started when i had to design switch with torsion spring. To create torsion spring is very easy even with nx2 (now we are on nx4 and preparing to move on nx6). The main problem is to position and assemble it to the assembly. Note, that in most cases in the assembly spring must be compresd, and this compresion depends on your assembly. So if your spring is defined as deformable part, before defining assembly constraints for the spring you must have a certain value as parameter for deformation (e.g. measurment value, or already existing positioning value).
Then you must define, how realistic will your spring be. Then the spring compress, it tries to roll around its axis, so torsion ends will became shorter/longer. This can be easily calculated.
But then you combine both deformable part and physics of the spring, simple torsion spring becomes to a bunch of curves, expressions and of course errors
I have some discusions with mr John R. Baker about deformable parts (not here, but in bbs.notes forum). And its clear, that deformable components (especialy springs) is one of the most complex parts of the nx.
How to create torsion spring it is another story. Because i had lot of problems with tangency of the curves (e.g. it is quite complex to add tangent line to the one of the end of the helix curve. It is quite difficult to describe it here, but if you try to create smth like in the attached picture, you will see).

 

[hudson888]

Oh boy have you got some stuff to look forward to if you thing that's complex! There's plenty more where that came from .

They can be a little tricky though I don't think that there is much doubt about it. The fact is that most people approximate in one way or another when they undertake to create such a spring and the subtle changes as it deforms are frequently overlooked. We could spend time arguing as to whether the axis of the coils ought to be aligned with the axis of the shaft, or whether the arms at either end lie on parallel planes or tend to twist as the spring flexes. The differences are small and nobody much notices until that is you need to construct a CAD model and you can't quite hit on the best thing to do first time around.

The fact remains it usually only has to look relatively correct and support a drawing, while for deformations in an assembly you can approximate. Mathematically it would indeed be awkward, but if you can model the spring with deformable parameters that duplicate the boundary conditions in your assembly then you can deform it. If on the other hand the boundary conditions can vary from one application to the next then you can't guarantee that the deformation will be identical.

There are however a lot of constructional tricks that you could use to develop associative sections that might work with a certain amount of fuss and linked geometry in order to arrive at measured expressions that you could get quite close to working. I suspect it is a lot of trouble to go to for one of these springs but it ought to be possible.

I think this example looks doable but time consuming.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[RobLN]

Hi Eex23

I saw your post regarding the torsion spring and I am still having problems with elements of the torsion spring not updating when using the deform command. Would it be possible to see the .prt file of the torsion spring to see how it's done?

Regards