AnsweredAssumed Answered

Definitions of the "Standard Planes"?

Question asked by John Willett on Dec 30, 2014
Latest reply on Feb 19, 2015 by Jared Conway

Like • Show 1 Like1
Comment • 19

I now have two books each on CAD and on FEA with SolidWorks (I am installing SW Premium, not the full Simulation package). None of these tomes appears to define the "Standard Planes."

The only place I've actually encountered these planes is in CSV numerical output tables (made for me by another user) of nodal displacements from an FEA analysis of an assembly. I was getting inexplicable results with "Top Plane" specified for the output and was told then, "The 'Front' plane (at the assembly level, if you are using an assembly) is supposed to be the referenced 'global' coordinate system for a collection of parts. Likewise the 'Front' plane is supposed to be the referenced 'global' coordinate system for a single part." Changing the output to "Front Plane" did the trick, although I have no idea what was really going on.

Can somebody please explain the meaning of these planes and how the relate to the global coordinate system (whatever that is) in a part or an assembly? -- John Willett

No one else has this question

Outcomes

19 Replies

John Willett Jan 27, 2015 8:27 AM

Perhaps this looks like a dumb question. Now that I'm running SW Premium 2014, of course I see the planes (Front = X-Y; Top = X-Z; Right = Y-Z) that show up as soon as you open a new sketch. I suppose these are the so-called default "Standard Planes." I still don't know, however, how these planes relate to the default "Datum Planes" (Tertiary, Primary, Secondary), whatever they are. (The preceding terminology came from a book of SW tutorials by Planchard.) Are they the identical, respectively as listed? If so, why have two different naming systems?

Maybe a better question is, "Why are these planes called 'default'?" Can other options than the defaults be defined for these planes? If so, what might they do for me?

More importantly for my purposes, I don't know how selections like "Selected reference : Front Plane," etc. relate to the coordinate systems for displacements/stresses when used to dump to a CSV table of nodal results from a simulation. Is this a question only for the Simulation section, or is it more general within SolidWorks? -- John Willett

Actions

Robert D. @ John Willett on Jan 23, 2015 3:08 PM
Mark Correct
Correct Answer
These planes are set in the global coordinate system as follows. 0,0,0 is where the reference planes reference off of. You can make another plane off of these reference planes if you want to but i wouldn't at first. When you go into Assembly: The first part you add to the assembly will default to god knows where. After you insert the first part right click on it and select float. Then add mates front to front, to to top, right to right. When you add all new parts everything will agree. I am not sure why SW does not have its planes set up that way in assembly but trust me the way above works best. As far as displacement and stresses part of your question I have no clue. Good luck. Hope this helps
Like • Show 1 Like1
Actions
- Glenn Schroeder @ Robert D. on Jan 23, 2015 3:46 PM
  Mark Correct
  Correct Answer
  Robert D wrote:
  
  The first part you add to the assembly will default to god knows where. After you insert the first part right click on it and select float. Then add mates front to front, to to top, right to right. When you add all new parts everything will agree. I am not sure why SW does not have its planes set up that way in assembly but trust me the way above works best.
  
  Robert,
  
  When placing the first (or any) component if you will select your component, then click on the green check mark in the Insert Component's PropertyManager without clicking in the graphics area then this component will be placed with it's three primary planes aligned with those of the assembly. Then there is no need to Float the part and apply Mates.
  Like • Show 1 Like1
  Actions

Jared Conway @ John Willett on Jan 27, 2015 2:16 AM

what is the actual workflow they are trying to get you to follow for the example?

Actions

John Willett @ Jared Conway on Jan 27, 2015 8:36 AM

>>what is the actual workflow they are trying to get you to follow for the example?<<

Hi, Jared -- This question is not related to a tutorial. It relates to a discussion that we had at Numerical Node/Displacement List Incomplete . Part of the answer (about using the "Selected reference : Front Plane") came from Anthony Botting in that thread, but I've never found the whole answer. -- John Willett

Actions

Jared Conway @ John Willett on Jan 29, 2015 3:22 AM

I guess i'm still not sure what your question is or what workflow that you're trying to follow or what you're trying to get out of the software

Actions

John Willett @ Jared Conway on Feb 5, 2015 12:53 PM

Gurus, please correct any errors in what follows:

Here's what I believe to be a complete answer to my underlying question, just determined by experimentation with simulation on a simple SW model since I could not find it in this forum or anywhere in the SW literature. I'm sure most users already knows this stuff, but I hope it will help any who are as confused as I was.

1) SolidWorks has a "global coordinate system" that is always indicated by the little red/green/blue Cartesian axes in the lower-left corner of the graphics area. It rotates exactly as one would expect with changes among the "standard views" and is apparently immutable. Sketches can be oriented with respect to it by selecting one of the "standard planes," (Front = [X, Y]; Top = [X, -Z]; Right = [-Z, Y]. Here I'm using the form [x', y'] to represent the global axes **and their directions** corresponding to "right" and "up" when the specified plane is viewed face-on).

2) After a simulation analysis you can define plots (e.g., UX on a specified surface) **after** specifying one of the standard plane (perhaps any user-defined plane?) as "reference geometry." Once this is done, you can also "list selected" results from this plot (e.g., [Node, UX, X, Y, Z] to a CSV file for ingestion into Excel or another spreadsheet for further analysis. (You can also list **all** components of a parameter, e.g., [Node, UX, UY, UZ, "value"], without reference to any plot and without any specified surface, but again a reference plane must be specified, either implicitly or explicitly.)

3) Now here's the key point (from my point of view at least): The specified reference geometry results in a simple coordinate transformation for the displacement components from the simulation (and probably those of other vector fields as well). Effectively, the chosen reference plane has its own **local** coordinate system (call it [x', y', z'], and call the results given in it [ux', uy', uz']) into which the displacement components are transformed as follows:

Front Plane -- [ux', uy', uz'] -> [UX, UY, UZ]

Top Plane -- [ux', uy', uz'] -> [UX, -UZ, UY] (Here I'm assuming there is a global coordinate system in which [UX, UY, UZ] are still defined.)

Right Plane -- [ux', uy', uz'] -> [-UZ, UY, UX].

All of this makes perfect sense once you are expecting it. It's not clear, however, why these transformations are made in the first place, especially since the [X, Y, Z] node locations in the same listings do **not** undergo a similar transformation. In every case these node locations apparently continue to refer to the global coordinate system, independent of the reference geometry that's transforming the displacement components. If there's a reason that SW chose to make these transformations in this way, I'd love to know what it is!

Again, comments and/or corrections are most welcome. -- John Willett

Actions

Jared Conway @ John Willett on Feb 6, 2015 7:40 PM

i think best to post an example

i see where you are going but not what you're expecting it to be

i think you may also be reading into something that doesn't actually exist which is that you must select a reference plane. this is for convenience only.

Actions

John Willett @ Jared Conway on Feb 6, 2015 8:43 PM

>>i think you may also be reading into something that doesn't actually exist which is that you must select a reference plane. this is for convenience only.<<

Thanks, Jared, for looking it over. I gather you didn't find any glaring errors. Correct?

About posting an example, the runs, plots, and listings that I made for this experiment are too complicated and obscure to be of much general interest. Suffice it to say that I was able to convince myself that the transformations reported above (which were all I really wanted to understand) are actually occurring.

About why I looked into this, I'm an ex-physicist, hence I like to understand what's going on physically, particularly when it affects results that I'm about to further analyze. I stumbled on this issue because the colleague I was working with (before I owned my own "seat") accidentally selected some reference other than "Front Plane" and sent me displacement data that I couldn't make any sense out of. Eventually Anthony Botting told us that "Front Plane" was what we should use, but nobody explained exactly why or what the other options meant. I still have no idea why this selection is provided by SW in the first place. Have you ever used anything other than "Front Plane" in your simulations?

About **having** to select a reference plane, I did notice that (at least in some cases) that section of the PropertyManager is not initially expanded, so I suppose there is a default selection -- probably "Front Plane." I did not verify this. Have you? -- John Willett

Actions

Bill McEachern @ John Willett on Feb 6, 2015 10:53 PM

I didn't read the whole thread. But I might be able add something to this. A SWX assembly file has as many coordinate systems as there are parts in it and potentially more. The so called global direction is that defined by the highest level assembly - the assembly. All the nodes are define by this reference system. They are not going to change and no function exists that I am aware in the software that will derive a nodal transformation (on the Un displaced original locations). If you want to get the displacements in a different reference frame then the option exists to define the reference and the displacements can be output in this frame. in most codes you have directions 1, 2, & 3 ( x,y,z). When you pick the reference you define the new direction 1 and the rest fall in line with the right hand rule. Same thing happens in SWX but you need to know secret decoder ring system which must exist when you start having planes and other geometric entities that have potential ambiguities as opposed to s direction vector. You seem to have it now figured out so there you have.

Actions

John Willett @ Bill McEachern on Feb 8, 2015 2:42 PM
Mark Correct
Correct Answer
>>A SWX assembly file has as many coordinate systems as there are parts in it and potentially more. The so called global direction is that defined by the highest level assembly - the assembly.<<

Bill -- Maybe I see what you're driving at here: Suppose you want to show displacements on a particular face of a particular part (in or out of an assembly). You might want to see these displacements in directions [x', y', z'] **relative to that face,** as opposed to the global coordinate system. Correct?

For a single part, that face might be parallel to one of the standard planes, which would be simple, but in an assembly this probably would not be the case. Then you might use a plane constructed coincident with the face of interest as the "reference geometry" for your output (or use the axes belonging to the part, if that's possible). Correct?

That might still leave open the orientation of [x', y', z']. Presumably [x', y'] would lie in the chosen plane, but at what rotation angle? And would z' point into or out of the face? To address these questions, would you have to define a local coordinate system in relation to the new plane, or would SW take care of this for you somehow?

Does anybody actually do this, and if so, could they explain why and how? -- John Willett
Like • Show 1 Like1
Actions

John Willett @ Bill McEachern on Feb 10, 2015 8:10 PM
Mark Correct
Correct Answer
>>If you want to get the displacements in a different reference frame then the option exists to define the reference and the displacements can be output in this frame.<<

OK, here's a specific example:
The tetrahedron has three sides on the principal planes and the fourth side on oblique "Plane1," which is normal to the current viewpoint. If Plane1 is specified as the "reference geometry" for the three displacement-component plots, what are the transformed directions for [UX, UY, UZ]? That is, exactly how is the coordinate system [x', y', z'] oriented with respect to Plane1 (I'm not asking for the mathematical transformations here)? -- John Willett
Like • Show 0 Likes0
Actions

John Willett @ Bill McEachern on Feb 10, 2015 8:47 PM

>>...exactly how is the coordinate system [x', y', z'] oriented with respect to Plane1...<<

Is the answer to this question as simple as the following quote from SW Web Help?

Display local coordinate system reference triad

When checked, the local coordinate system reference triad is displayed at the lower right corner of the graphics area. The triad is displayed only for plots where you define a reference plane, axis, or coordinate system.

(Note that this image is projected from the opposite side of Plane1 than the previous one, which of course shows no local coordinate system.) This would suggest that, when viewed perpendicularly, the local coordinate system of any plane, no matter its orientation to the global coordinate system, is [x' = right, y' = up, z' = out of the screen]. -- John Willett

Actions

Jared Conway @ John Willett on Feb 13, 2015 1:48 PM
Mark Correct
Correct Answer
i think this is what bill and I were saying previously.

maybe to synthesize the question better, are you trying to figure out what direction 1 and direction 2's orientation are?
Like • Show 0 Likes0
Actions
- John Willett @ Jared Conway on Feb 13, 2015 3:22 PM
  Mark Correct
  Correct Answer
  >>maybe to synthesize the question better, are you trying to figure out what direction 1 and direction 2's orientation are?<<
  
  Jared -- This may just be a matter of terminology. I'm not familiar with the usage, "direction 1," etc. If these refer to the Cartesian directions in a local coordinate system (e.g., looking normal to an oblique plane), the answer is probably yes.
  
  I still would not understand which "direction #" was what I'm calling x' (right), which was y' (up), and which was z' (out of the screen). Is there some reason for yet another naming system for directions? Can you give me a reference that explains it? -- John Willett
  Like • Show 0 Likes0
  Actions
  - Jared Conway @ John Willett on Feb 17, 2015 5:50 PM
    Mark Correct
    Correct Answer
    rhyme or reason > nope, its always just been that way
    
    i'm pretty sure the trick i use is to go normal to and dir 1 is up, dir 2 is to the right. but haven't had to worry about that in ages
    
    if you feel the terminology is lacking, i'd suggest submitting an enhancement request
    Like • Show 0 Likes0
    Actions
- John Willett @ Jared Conway on Feb 19, 2015 11:41 AM
  Mark Correct
  Correct Answer
  Jared -- It seems you are right; your [Direction 1, Direction 2, Direction 3] are identical to my [x', y', z']. See quotation from KB S -020957:
  
  "To know what the x, y, and z directions are, you can define a force relative to a plane you have created and apply forces in direction 1, direction 2, and direction 3, which correspond to x, y, and z, respectively. The force vectors displayed on the screen will tell you which direction is x, y, and z."
  Moderator -- Maybe it's time to mark this thread as answered, even if not by just one post. -- John Willett
  Like • Show 1 Like1
  Actions
  - Jared Conway @ John Willett on Feb 19, 2015 11:52 AM
    Mark Correct
    Correct Answer
    thanks for the confirmation.
    
    I think as the OP, you actually have the power to mark the one that you think is the right answer
    Like • Show 0 Likes0
    Actions

VENKATESH S. Jan 23, 2015 10:52 PM
Mark Correct
Correct Answer
FRONT PLANE = XY
TOP PLANE = YZ
RIGHT PLANE = ZX,
These are the 3 planes which are linked between the global coordinate system X,Y,Z.
To change sketch/something which referenced the plane from one to another RT.click on any plane -> Click edit sketch plane -> Sel another plane from flyout feature manager design tree.
Thanks and Regards
Venkatesh S
Application Engineer
E G S Computers India Pvt. Ltd.
http://www.egsindia.com | http://www.egs.co.in | http://www.egsindia.blogspot.in
Like • Show 2 Likes2
Actions

Incoming Links

Re: Reference geometry of directional components

Definitions of the "Standard Planes"?

Outcomes

Related Content

Recommended Content

Incoming Links