Negative Dimensions in Design Table - Not Consistent 1

[cbrf23]

I have a part with multiple configurations driven by a design table. One dimension, 'T', is a vertical (Y-axis) dimension from the origin to a point in space (sketch is on the right plane). The smaller the part, the higher above origin this point becomes. On the largest parts, the point moves below the origin.

In my table, I've set this up as positive values for the parts where the point is above the origin, and negative values where it is below the origin. When I close the design table, I get a message telling me that the table contains negative values and that solidworks will flip the dimensions to handle this. This is exactly what I want.

What I find strange, is that everytime I open the part or any file that references that file, its a crap shoot as to wether or not Solidworks is flipping the dimensions or just ignoring the negative. Sometimes, the dimension will flip directions appropriately. Other times, the values are plugged in, but the dimension does not flip directions...

I dont understand why it works part of the time and not all the time. Is this a design flaw with solidworks? Am I doing something wrong, or is there some setting that I've missed? Is ther anything I should check that is known to cause this described behavior?

Its become quite time consuming to have to check this dimension every single time I open the file, and everytime I open any assemblies it is used in, to verify that it is dimensioned correctly. I feel like I can't trust solidworks, so when exporting files for use with the machines or to send to the customer, I have to open the freshly created file and measure that point in space, just to verify that solidworks was behaving properly - it isn't consistent, and that is very frighening.

It also raises a legitimate concern, as I am not the only solidworks user here, and anyone else that opens the file may not know to do this and could very very easily send out a bad file and make bad parts!

Does anyone have any insight or advice? I'm waiting for my Solidworks rep to call me back, but any additional help would be appreciated.

I cannot share the file, as it is proprietary work for a customer, but I have included a screenshot of the issue and of the design table...

 

[cbrf23]

It would seem, that the directions will flip just fine if JUST the part itself is open.

If any drawing or assembly or any other file that references that part is open, the dimension will not flip direction unless you have the exact same configuration active in ALL the open files that reference that part. Then, if you open up the design table, and close it, it will flip the direction...

This seems like a pretty serious functionality issue with solidworks. I will be doing more testing on this, and if it turns out to be consistent behavior, I will report the issue.

 

[EEnd]

I remember a post with a similar issue. The suggested workaround was to find / create a sketch feature that the feature to be dimensioned was always on the same side of and dimension to that feature. The values in your design table will have an offset from their current values, but the final result should be more stable.

Eric

 

[handleman]

You probably should not use negative dimensions in design tables. Or equations. As far as I know, SW has no way of knowing whether a dimension has been flipped at some time in the past or not. When SW encounters a negative number, it flips the direction but does not necessarily remember that it was flipped.

As EEnd said, you need to figure out a way to drive that dimension such that it is always a positive number.

-handleman, CSWP (The new, easy test)

 

[cbrf23]

Hi guys. Thank you for responding. The workaround I did was as suggested, I dimensioned an arbitrary line 1mm below then dimensioned the points from that so they are always positive.

 

[mjcole]

I found this to be a bug and told people at SolidWorks but they clainmed it wasn't a bug. ProE does have a setting to maintain a negative value for dims but if you forget and don't type the negative it will switch sides.

Bug as I see it.

1. Add dimension.
2. Flip Direction.
3. Modify value to 0.0
4. Change 0.0 dimension to a value.

The positive value entered will be positive in opposite direction and not maintain the flipped direction.

I would expect that if a positive value is entered after the dim was swapped it would be in the same direction.

I used to use the 1 unit offset back when using Unigraphics.

Michael
____________________________________________________________________
"It's not the size of the Forum that matters, It's the Quality of the Posts"
Michael Cole
Boston, MA
CSWP, CSWI, CSWTS

 

[cbrf23]

Yeah it's definitely a bug in my eyes as well. The work arounds not too bad, but it makes things a little more complicated and my design tables don't match my print, which makes cursory checks on complicated parts take a little longer. Its not a huge issue, but it does seem lazy to me that it's still there, considering when i was researching this issue i learned that support for negative numbers was introduced in sw2009 and it's now 2 major releases later...

 

[handleman]

There's a difference between a bug and something that doesn't work the way you want. If you tell them it's a bug, they will write you off as ignorant - someone who doesn't know what the word "bug" means. If you tell them that the functionality doesn't work well, generates awkward workflow, has limitations, etc. that makes sense. A virus might wipe your hard drive. Would you call that a bug too? No. It's intended functionality.

-handleman, CSWP (The new, easy test)

 

[handleman]

Sorry, meant to include that I also think the current functionality is basically useless - poor execution of the concept. It can help you get sketches to "flip" around the way you want them during editing, but that's about it.

-handleman, CSWP (The new, easy test)

 

[AnnaWood]

Handleman,

That is its intended functionality to just move sketch geometry to the other side of the line.

You will need to put in an Enhancement Request.

Cheers,



Anna Wood
SW2011 SP5, Windows 7 x64

http://www.renderbay.com

http://www.solidmuse.com

http://www.phxswug.com

 

[Updraft]

Let's ask the question, "What does the negative mean?". It is easy to understand, without ambiguity, what a positive dimension means, but what does a negative dimension mean? In order for it to mean something we would have to define direction. This functionality, defining a direction for negative, is not easy to do in a universal application. Keep in mind that the origin and the planes are convenient references, but do not carry enough definition to specify direction.

SWX DOES let us key in a negative value in the dimension box. This effectively flips the dimension to the other side of the fixed reference. In this case the negative can be interpreted relative to the existing situation, i.e., the negative is the opposite direction of the current direction.

If a design table has a negative value then would the dimension flip every time that dimension is rebuilt? I doubt anyone would stand for that.

I LOVE the power of the spreadsheet in the design tables. Simply insert a blank column or row between the SWX part of the DT and the area for the other spreadsheet calcs you want to setup and you have a whole spreadsheet playground to use. Your DT cells can then reference the results of other cells in this playground area or even on another sheet. This is great! So why not use this playground area for your calcs that can generate negative values, but add a suitable biasing value to make the DT cell positive? It is a simple way of having your cake and eating it too.

EEnd and handleman correctly identified it. Negative values should not be used in DT's. You'll need to use an offset reference where the dimension can always be positive.

This is not a bug. If you want to send in an enhancement request I suggest you think through clearly what you are asking for and how that functionality would be interpreted in other, perhaps unintentional, situations.

- - -Updraft