Displaying Constraints in Assembly Navigator 4

[Finglas]

Is there any way of displaying constraints in the assembly navigator for individual components? i know they are all listed under the descriptive part name but i only want to see the constraints for individual parts.

Also know i can display constraints in graphics window but not gone on this either.

 

[phillpd]

John,

What we do understand is that attempting to interrogate Assembly Constraints is a complete PITA

Yes, with Mating Conditions we had to periodically endure the "circular reference", but they were so much easier to create/suppress/interrogate. And you had the dynamic DOF.

NX 6.0.5.3
Windows XP32

 

[JohnRBaker]

Note that the on-screen display of the DOF symbols has been reintroduced with NX 7.5.

As for your basic issue, this is the nature of the beast.

Perhaps a little thought-excersize might help.

Let's say you have 3 blocks, call them 'A', 'B' and 'C', and you create an assembly and add these 3 blocks as components. Now position 'B' relative to 'A' and then 'C' relative to 'B'. Based on your questions, I suspect that what you're expecting is that when you ask 'B' how it was constrained, that only 'A' would be mentioned, correct? Now if 'A' were 'fixed', perhaps there might be some argument for that. However, what if 'C' were fixed, even though it was last component added and the last one positioned? What if BOTH 'A' and 'C' were fixed, what would you expect to see when you ask how 'B' was constrained? What if NEITHER 'A' nor 'C' is fixed, now what should the system report?

And what happens when we add another component to the mix, block 'D', also positioned relative to 'B'? What should be reported when we start to edit the 'fixed' status of any or all of the 3 blocks now constrained to 'B'?

The truth is that ALL relationships are part of the definition of what is constraining 'B'.

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

http://www.siemens.com/plm

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

To an Engineer, the glass is twice as big as it needs to be.

 

[cubalibre000]

Hi John,
maybe I don't understand the nature of your 'Assembly constraints', but you don't understand NX users needs.
This request is asked three times without absolution by users.
Maybe you aren't a designer that work every day with assembly or big assembly where you need to understand which mates you have to edit because the project is changed or because is made by others coworkers.
I don't think that is how D-Cubed solver works, because is used by Solid Edge and SolidWorks and they show different.
Catià, Pro/E and Inventor show different.
At the final Catià, Pro/E, NX3, SolidWorks, Solid Edge and Inventor show only the mates used for the positioning of the component selected.

Please ask to the users what do they think about this visualization ?
I'm not alone.
Please add this filter in NX8.

Thank you...

Using NX 7.0.1.7 MP1 and TC8.1

 

[JohnRBaker]

I'm sorry that you don't understand, but the system is showing you EXACTLY what you're asking it to. That's it, PERIOD. As I stated, EVERY constraint contributes equally to positioning a component. The software has no way to read the mind of the user to determine WHICH of the positioning constraints is of interest IF THE USER ONLY SELECTS A SINGLE COMPONENT. If you KNOW all of the components for which you wish to see the constraints related to them, then when the 'Show and Hide Constraints' dialog is open, select the 'Between Components' option as well as the 'Filter Navigator' option and then SELECT ALL THE RELEVANT COMPONENTS. Now when you look at the constraints in Assembly Navigator, ONLY the ones which are shared by the selected components will be displayed. I don't know how to make the description of how the current behavior works more explicit or clear.

Note that for NX 8.0 we are going make it easier for you to see ALL of the constraints related to any ONE component. However, that will NOT change the fact that ALL constraints related to a Component will still be shown since if you're ONLY looking at one component, we don't know which constraints YOU think are important until you provide some additional information, such as the OTHER component(s) which you wish to include in the query, then the system can limit the list to only the constraints which are shared by these multiple components. But if you only indicate a single component, we have to list ALL of them.

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

http://www.siemens.com/plm

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

To an Engineer, the glass is twice as big as it needs to be.

 

[phillpd]

John,

Read what you just typed ...

"...then when the 'Show and Hide Constraints' dialog is open, select the 'Between Components' option as well as the 'Filter Navigator' option and then SELECT ALL THE RELEVANT COMPONENTS..."

That's just far too much effort, is in no way intuitive, and simply does not improve on the useability of interrogating Mating Conditions.

Creating Assembly Constraints is so much better then Mating Conditions. You did a fantastic job there. It's just afterwards that they're a nightmare.

NX 6.0.5.3
Windows XP32

 

[JohnRBaker]

Nowhere did I say that the 'Show and Hide Constraints' in anyway "...improved on the useability of interrogating Mating Conditions." The point of what I was saying is that considering the way Assembly Constraints WORK, the use of 'Show and Hide' is the best that you can expect. There is NO way to query a Component positioned using Assembly Constraints and expect to get back the type of information that one did when that Component was positioned using Mating Conditions. The mechanism is completely different. It uses different rules, it makes different assumptions, it creates different relationships. It would be like running a diagnostic on your fuel injected car looking for the size of the jets used in the carburetor. You can only expect to get an answer based on the what data is available.

The reason that Assembly Constraints are better behaved than were Mating Conditions, when creating and editing them, is because:

a) There is NO longer any dependence on the order in which constraints are assigned.

b) It does NOT make any difference which Component was picked first and which Component was picked second when assigning a constraint.

c) It NO longer matters if more than one Component has a fixed constraint (with Mating Conditions, one component was ALWAYS considered fixed, but you had little or no control over which one it was). In fact, there is no need for ANY Component to be fixed.

d) Any constraint can now be deleted and redefined anywhere in the assembly without having to consider when or where it appears in the positioning scheme.

Now the point I've been trying to make is that the above items are the things which we are NO longer having to be concerned with since this sort of stuff is NO longer even being saved with the Assembly. Therefore, when you start to ask questions, like what constraint is used to 'mate' one Component TO another, there is nothing that we can report since we NEVER saved any 'FROM/TO' information in the first place. Granted, I can ask for the constraints applied to this Component, but since the ORDER in which they were applied or knowledge of which Component was the 'FROM' and which Component was the 'TO' is NO longer part of the data needed to position and solve the Constraint scheme, there is nothing to report except what we do. In this case ALL constraints since ALL of them have an equal chance of being responsible for the current position of the Component within the Assembly. I can't get info about data which NEVER existed.

Anyway, I think we've beat this horse near enough to death for now. Since we are so close to NX 8.0 Beta and since we have made significant improvements in how and what we're able to report with respect to Assembly Constraints and we expect that further tweaks will be made in the functionality based on both discussions like this (please don't think that this exchange has been limited to only myself, as others involved in Assembly Constraints have been following this thread and it has caused us to look closer at what we ARE able to do) and as a result of what the Beta testers are going to be saying to us as they actually work with the new functionality being implemented, that we let things proceed and then revisit this later to see what we've learned and how what we're doing has impacted your concerns, OK?

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

http://www.siemens.com/plm

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

To an Engineer, the glass is twice as big as it needs to be.

 

[DMiller01]

I would like to respectfully summarize my perception and that of other posters to this thread. Any benefits of the current assembly constraint method is outweighted by the difficulty required to maintain a robust assembly model with reliable positional control. The current system is a web of inter-tangled constraints. Every additional component (node) increases the complexity of this web. Better interrogation methods are not going to change this.

The last thing Sieman's wants, is to redo their current assembly constraint method, but I predict that at some point they will have to. Till then I'll try to cope and hope to share constructive advise with others on this forum.

I am using 7.5.2.5 NATIVE on Dell with windows XP OS

 

[eex23]

sometimes happens you need to free one component in the assembly. I mean, to suppress constraints making only certain component fixed. In new assembly constraints it is difficult to find these constraints, because you see all the constraints concerned with desired component. "show and hide constraints" does not help, because usually you dont even remember to what component you constrained desired component.

http://www.eng-tips.com/viewthread.cfm?qid=290637&page=6

 

[JonSelby]

A couple of tips that I stumbled across which I find really useful (note NX6) ...

1. When you open up your "Assembly Constraints" dialog, by 'dragging' any part, you can see how that part is constrained.
Its a good visual indication of any DOF left unresolved.

2. In your "Assembly Navigator" (which I keep pinned open), add the "Position" column. A black dot means its totally constrained. A half black half white is partially constrained. A white dot is unconstrained.
To do so right click a blank space -> columns -> configure and you'll work the rest out!

3. Do re-read the second post in this thread. That dependency window is genuinely useful if you keep it pinned open.

Being really picky I've one silly question ... If you have two entirely unconstrained parts and you apply a constraint BETWEEN them, upon clicking 'Apply' does the first selected part move to the second, or vice versa or is it random? I know it doesn't change the resultant relationship, but occasionally the order can cause parts to disappear from view and you have to go off and find them.

Jon S

Medical Design Engineer - Glos. UK
NX 6.0.4.3 / TCE V10.0.3.8.6

 

[JohnRBaker]

Per your last question, as far as what you will see, it's 50/50.

What happens is that when the constraint equations are solved, in this particular case since all that we know is that a constraint has been added to both components and neither of them have any other constraints applied to them, the solution will give two valid results. And since the system has to do something, one of them is honored and the other is ignored. As to which is which or how a user would even know what to expect, while there obviously is some algorithmic scheme which is controlling this, the variables involved are not items which the user has any control over or to even be in a position to know about so that he/she could predict what will happen.

That being said, if this is a critical issue with you, than we recommend that you first apply a 'Fixed' constraint on the component that you wish to NOT move. Now this is not as big of a deal as it was in Mating Conditions where the FIRST component selected in the FIRST Mate created, determined for all times (unless that component or the Mate was subsequently deleted) which component was 'Fixed' (even if we never reported that to be the case). Now with Assembly Constraints you always know which component(s) have been 'Fixed' and you can always delete that constraint and assign it to another or even leave you model totally unfixed, it's your choice. So we don't think it would be unusual at all for people to create temporary constraints which they have every intention of deleting later on, since there is no longer any downside to doing that, given that the order of creation nor what component was selected when, no longer has any effect on the final position of the components in your assembly.

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

http://www.siemens.com/plm

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

To an Engineer, the glass is twice as big as it needs to be.

 

[cowski]

The last thing Sieman's wants, is to redo their current assembly constraint method, but I predict that at some point they will have to.

Actually, they recently did and that's what's giving the long time users problems - thing don't behave quite like they used to. In the long run I believe the new way will be better, we just have to get used to it and Siemens has to find the most efficient way to interrogate/organize the information displayed.

 

[RobLN]

As a newer user to NX (NX 4 onwards) I can't really see where the problem is with the display / interrogation of constraints as it is currently (NX6)

As a general tip for working with constraints, keep the assembly navigator open on the side & pinned all the times with both preview & dependenceis panes expanded - with the magnifying glass permanently orange. (it not a painful process if its always there when you need it...)

I find that if you have an assembly on screen and you're wanting to check the status of the constraints you assigned, keep the 3D display in wire frame and then highlight constraints from the dependenceis listing. As long as the wireframe display is set then as you click on the constraint (and is goes blue in the list) you get instant feedbak on the geometry/constraint in orange on the 3D view. Used in conjunction with the constraint default text description (relating it to other parts) I find the system easy to use.

I can only compare this current constraining to the one used in older versions of Solid Edge (more like the old mating style conditions) and using that was far more frustrating in my opinion...



Solid Edge; I-Deas 7 to 12; NX4, NX5 & currently NX6.

 

[SiW979]

I've been using NX for a few years now, and have been using NX7.5 for a few months and I must say that I have no problem with the assmbly constraints now and can manage them perfectly well in some very large assemblies. In fact they are so much better than the mating conditions.

What does still fry my noodle though is that there is no EASY way to create mating conditions across assmeblies without WAVE linking etc. A real PITA for sure.

Best regards

Simon NX4.0.4.2 MP10 - TCEng 9.1.3.6.c - (NX7.5 native)

http://www.jcb.com

Life shouldn't be measured by the number of breaths you take, but by the number of times when it's taken away...

 

[cubalibre000]

Hi JCBCad and RobLN,
different from NX3 there are this problems :
1) With the new assembly constraints, you can add infinite constraints and only after you close the mask you see that there are infinite icon errors.
In NX3 if you add a no supported mate or over meted constraints, NX3 tell you and doesn't add mates.

2) In NX3 you can see mates used for the component to fully constraints.
In NX7 you see all mates of all connected components, very difficult to understand.

3) Grouped differents NX3 constraints into same constraints name, for example tangent.

4) When you delete, edit or change via RMB button a mate in the details tab, the details tab become empty and you have to re-select the component.

5) If you edit a distance mate because has lost reference or edit to change the references, it create a new distance mate.

6) In NX3, if you change a distance between two hole in a component, NX3 advise you that the mate 'X' is not resolved because.....
In NX7 no, so you can work for four hours without any advise (NX is alone in this case).

Maybe are more powerful in the solver, but there are lot of unsatisfied users.
There are difference from who use NX for 8/10 hours every day (me) to who use NX for some hours.

Thank you...

Using NX 7.0.1.7 MP1 and TC8.1

 

[RobLN]

Hi Cubalibre00

Sorry to see that you have so many frustrations with the assembly constraints. I guess it all comes down to what you're used to...

Here's some tips:

1. just add constraints as you go part by part. If you just add one incorrect (overconstrained) new item it can/will invalidate any number of related constraints these will then display red or orange X icons (or both!) against problem constraints. However this is only ever as a result of incorrect constraint specification. Just ensure that you keep an eye on your constraint list as you go. If you just keep on adding constraints regardless it will quickly become a big mess

2. seems simple enough to me. just keep the magnifying glass icon orange. each constraint is named so you know clearly what all the constraints do rather than a select few.

3. Why bother ? they're grouped by component. order on type is irrelevant.

4. Agreed that is a PITA. NX / siemens should fix that.

5. Hmm not seen that. Just double click on either the constaint in the 3D view or on the navigator list and change the value as needed. If it cocks up its easy to just redo it if needed.

6. Again in my experience its generally resolved well by the software which is nicer than being told it failed...

In general I wouldn't try to get everthing fully constrained (full black dots everywhere) until most deign work is frozen. otherwise you'll just spend you're life doing constraints.

Solid Edge; I-Deas 7 to 12; NX4, NX5 & currently NX6.

 

[JohnRBaker]

In the next release of NX we are adding, in addition to being able to see and manage your constraints in the Assembly Navigator, an additional Constraint Navigator which will provide a varied set of ways to control what and how to display constraints and interact with them. For example, there will be options to display the constraints based on constraint type, such as having all Concentric together, all Fixed together, etc.

Or you will be able to list be status, meaning that it will easier to find all the ones which have failed or have warning associated with.

Also you will be able list by component, similar to what you can do today with Dependencies except this will allow you see the constraints associated with more than one component at a time.

You will also be able to group by Component status, which will make it very easy to find which components have conflicting or inconsistent constraints, which are unconstrained, partially constrainted, fully constrained, as well as which components have suppressed constraints or which components have had their constraints overridden by an Arrangement. It also be very easy the find which components in sub-assembly have had their positions overridden from a higher level in the assembly.

While we recognize that there are still some issues that we will need to address, we feel that the addition of this dedicated Constraint Navigator and all of the new options it will offer the user in terms of seeing how his constraints are being appstraightforward.lied and their various status levels, that this will make the management of large and complex assemblies much easier and striaghtforward.

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

http://www.siemens.com/plm

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

To an Engineer, the glass is twice as big as it needs to be.

 

[st4r]

Sorry for bringing this up, just saw it, but I have to agree with Cubalibre. At least to some extent. It can really be a frustrating experience to work with assembly constraints. But I got used to it (also to some extent), well I had to, since none asked me do I like them I'm also using nx on daily basis (8+h).

Cheers

 

[JohnRBaker]

All we can ask, is that until you've had a chance to see what we've done with the next version of NX and it's new dedicated 'Constraint Navigator', could you please hold off on making any more requests to somehow "make it behave like it used to".

I know this is a bit of a tease, but in all honesty, what else do you expect us to say ;-)

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

http://www.siemens.com/plm

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

To an Engineer, the glass is twice as big as it needs to be.

 

[phillpd]

Well, I certainly hope that for the production release of NX8.0 (currently beta testing) that the "degree of freedom" becomes dynamic, or at least, when I re-issue the command, it clears the previous one.

NX 6.0.5.3 (NX 8 Beta)
Windows XP32