Promotions or Wave

[Bishbosh]

NX5.0.2.2.
I have a Nitrile washer when assembled into a 'cupped' tube with the force of a spring behind it would deform to suit the internal rad of the tube. The assembly of the spring when under load I can control with deformed part. In the past (NX3) I would use promotions for the washer. Is this still the preferred method?

Andy

 

[hudson888]

Deformed part would go close to being the most elegant method. I'll fess up that I'm not absolutely certain of what happens when the Nitrile washer in your case deforms, but I think you probably, mean what we refer to as an O-ring in the cheap seats.

So if you can describe the deformation of the washer using parameters in that file then I would do so and deploy it as a deformable part to describe that condition.

Now if you want to define the deformed section as ovoid, then you would use an ellipse. And you would point out correctly that ellipses aren't parametric, but as long as you're only deforming it in a plane I have seen somebody create a very cunning method of revolving a sketched circle about an axis to describe the ring and then tilting the sketched axis relative to the axis of revolution so that the section described an ellipse. Fun and Games!

Otherwise you have some features that you want to add to the model as assembled, and you can use either a promotion or a wave linking method to do this. Most I suspect would argue some quite vehemently that the wave link being the more modern antecedent of promotions is the only permissible method. I think that I'd probably go for the promotion though there are those elements of simplicity in having the parts list co-operate in reflecting your intent that tilt the balance in that direction for mine.

Before I finish on this I'd like to ask something without meaning to be flippant at all because several people have asked about this kind of thing in the past. I wonder why it is necessary to bother. For what it is worth we usually simply show a slight interference. I wonder do others simply feel differently about the importance of it or whether there is another reason that I'm absolutely missing.

Best Regards

Hudson

 

[Bishbosh]

The Nitrile washer is just a plain washer stamped out of a sheet with square corners all round. The reason I need to show deformation, is this assembly complete with other seals similarly deformed will create a closed volume that I need to define.

 

[hudson888]

Thanks for the quick reply. So if it is a flat washer then you'll probably be able to define a deformation easily enough.

Cheers

Hudson

 

[SiW979]

Hudson

When you say elipses are not parametric, do you mean in sketches? I'm would just like to clarify what you mean?

Best regards

Simon (NX4.0.4.2 MP4 - TCEng 9.1.3.6.c)

http://www.jcb.com

 

[JohnRBaker]

Sketches are really the only way that you CAN actually create a 'parametric' Ellipse.

Actually that's not true, but it's generally easier to go that route than it is to control the orientation and location of the curve of intersection of a cylinder and a plane ;-)

John R. Baker, P.E.
Product 'Evangelist'
NX Design
Siemens PLM Software Inc.
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.org/museum/

 

[SiW979]

John

During my sketch learning period I stuggled with constraining elipses, in fact I never fully could, it's only within the last year or so that I started to sue reference geometry to constrain a portion of the elipse and then use the mirror function within the sketcher. See attached. If I try and constrain the full elipse with center lines then I always seem to get one constraint which when I try to resolve it results in either conflicting constraints or over constraints.

Best regards

Simon (NX4.0.4.2 MP4 - TCEng 9.1.3.6.c)

www.jcb.com

 

[JohnRBaker]

Simon,

Look at the attached example. The 'trick' is the Parallel constraint between the Line representing the Semi_Major Axis and the Ellipse itself.

John R. Baker, P.E.
Product 'Evangelist'
NX Design
Siemens PLM Software Inc.
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.org/museum/

 

[hudson888]

Simon,

What I was posting was for an ellipse curve. It is parametric insofar as you can alter the values but cannot assign an expression to those values.

I guess I should have tested again because I referred to something that I saw somebody do a while back perhaps due to limitations in the system of the time that may have since improved. I though tilting the plane of the sketch was a very cool work around just the sort of tip nobody would have thought of.

You example clearly proved otherwise.

John,

You example was even better I duplicated it instantaneously. The only dodge with it is that if you double click on the ellipse curve it still presents you with the curve parameters which are actually useless if you are overriding them with the dimensions that you added. I'm pleased to have learned something inadvertently from this post even though it turned out to be a bit off topic. it makes the forum more rewarding doesn't it . Anyway I thought the ellipse in sketcher may be a candidate to be slightly better integrated. I don't want that to sound like sour grapes or biting the hand, looking a gift horse etc..., but maybe having to guess at this trick shouldn't be necessary.

Thanks guys,

Hudson

 

[JohnRBaker]

Hudson,

While I called it a 'trick', that was being a bit theatrical ;-)

But you can try this yourself. Open a a new sketch and create as your first object an ellipse and then enter the assign Constraints function so that you can see the DOF (Degrees Of Freedom) symbols (there should be 3 of them, one each for the major- and minor-radius, and one indicating that the ellipse has the freedom to 'rotate'). What that's telling you is that there is an angular sense about the center of the ellipse which also indicates that this angular sense has to be relative to something, in this case a sense of direction or axis. It is this axis (which is aligned with the major-radius) that I was constraining parallel to the line that was part of the 2-curve set that was being used to constrain the size of the ellipse. Granted, that might have been a bit subtle, but we were providing as much feedback as we could using mechanism available to us, the DOF symbols.

Anyway, while I agree that we can alway use some enhancements and improvements, which we continue to work on with each new release, there is a large amount of functionality already there if you just can take the time to learn where and what it is and how it all can be used to lock down your design criteria.

John R. Baker, P.E.
Product 'Evangelist'
NX Design
Siemens PLM Software Inc.
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.org/museum/

 

[hudson888]

Thanks John,

you could have saved the trouble of describing it all as I said I was able to duplicate your example instantaneously. That is it took me all of two seconds.

As I also commented it seems almost churlish to pick fault when you have provided a handy piece of information. I just thought that if it were an easy enhancement then it would be a feather in your cap to provide. The error I made in the first instance was one of assumption. I created an ellipse in a sketch and assigned expression values for the semimajor and semiminor, using them in the appropriate places within the ellipse creation dialog. The numbers were duly substituted, but when the expressions are edited the values are obviously not associative since nothing happens. At that point I made the assumption that was all there was and would have given up. It is great that you showed me where the functionality and that it works really well, but of course others will make my mistake and won't know how well and easily these things can be done.

Best Regards

Hudson

 

[JohnRBaker]

Yes, the fact that Ellipses, in and of themselves, whether they are part of a sketch or not, are NOT parametric is a pet peeve of mine. The last time we did a major rework of the ellipse the same developer also implemented the helix as part of the same project. When he was done we ended up with an Helix that was a parametric feature and an ellipse that was not. The whole thing came down to a spec based on some requirements that we had gotten from a new customer. What they had asked for was a 'parametric' helix and an ellipse that they could easily 'edit', and the person who wrote the specs took that literally. So we got a parametric expression-driven helix and an ellipse that you can 'edit' to change its size. When I confronted the guy who wrote the spec (the developer was just doing what the specs described) and asked how in the heck did we end up with two different behaviors considering it was part of the same project implemented by the same developer based on requirements from the same customer. Anyway, his response was that if the customer had wanted a 'parametric' ellipse then that's what he SHOULD have asked for. My argument was that if we had made BOTH of them parametric that we would have still fulfilled the customers requirements, but he insisted that he was told to implement WHAT the customer had asked for and that it was not his responsiblity to do anything more than that. Sometimes COMMON sense is the most UNCOMMON of qualities in an organization.

John R. Baker, P.E.
Product 'Evangelist'
NX Design
Siemens PLM Software Inc.
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.org/museum/

 

[hudson888]

John,

Who'd have thought that ellipses had that much in common with helices. Even more that they both have something in common with an old joke about a guy with a genie and a million ducks, and an episode of the X-Files, be careful what you wish for . Human nature we couldn't live without it!

Thanks for the info.

Hudson

 

[SiW979]

I love this forum, the information available and the things I've learnt are invaluable, I think before long I'll be spending as much time on here as you two (John/Hudson) Which brings me to a cheeky question......Do you pair ever get any 'real' work done? ;-)

Best regards

Simon (NX4.0.4.2 MP4 - TCEng 9.1.3.6.c)

http://www.jcb.com

 

[hudson888]

Simon,

Work has been a bit slow so I'm entertaining a hobby. I like solving problems. Like you I suspect I know that John probably considers it complimentary to if not part of his job to post here so we all hope that never stops. ..

Cheers

Hudson

 

[JohnRBaker]

Simon,

Actually, this IS work (at least part of it). The question should really be, "Do you ever get anything done OTHER than work?"

John R. Baker, P.E.
Product 'Evangelist'
NX Design
Siemens PLM Software Inc.
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.org/museum/