AnsweredAssumed Answered

Find conflicting mate

Question asked by David Boydston on Jun 3, 2020
Latest reply on Jun 11, 2020 by Jim Sculley

Like • Show 0 Likes0
Comment • 18

Hi,

I have an assembly of a machine which has a gantry.

The gantry sub-assembly consists of many sub-assemblies and sub-sub-sub-assemblies.

When I open the gantry sub-assembly I can use my mate controller and everything works fine.

However when Im at the top level assembly (one level above the gantry assembly) I cant move the gantry anymore. my mate controller has the same mates enabled (theyre not greyed out as all subassemblies are flexible) however if I move the value on the slider it tells me "cannot solve with these values" or something like that.

I looked at the folder under the gantry-subassembly that lists 4 mates in the top-level assembly, but none of these mates are the "problem". I disabled them all and it doesnt solve the proble, I'm assuming theres another mate probably in one of the sub-sub-assemblies that was made in the top level assembly but I dont know how to find this. Is there a way to know what mates are causing this failure?

No one else has this question

Outcomes

18 Replies

Alex Lachance Jun 3, 2020 3:49 PM
Mark Correct
Correct Answer
Hey David,

Sub-assemblies lose their flexibility when inserted into assemblies. In order to bring the flexibility back, you will either need to mate the sub-assembly's environment to the assembly's environment(exemple, origin coincident to origin with axes aligned) and then turn the sub-assembly to flexible or you can fix the sub-assembly and then turn it flexible.

Personally, I prefer mating.

How to set sub-assemblies as flexible:
2020 SOLIDWORKS Help - Flexible Subassemblies

Edit: Excuse me, I jumped the guns on your post, just saw that you do have flexibility activated. Try turning it rigid and then turning it back to flexible. That generally does the trick. It seems that flexible sub-assemblies can only flex within the current session. Once you close SolidWorks and restart it, the flexibility is lost. To restore it, as I said, turn it rigid and bring it back to flexible.
Like • Show 0 Likes0
Actions
- Jim Sculley @ Alex Lachance on Jun 3, 2020 4:01 PM
  Mark Correct
  Correct Answer
  Alex Lachance wrote:
  
  It seems that flexible sub-assemblies can only flex within the current session. Once you close SolidWorks and restart it, the flexibility is lost. To restore it, as I said, turn it rigid and bring it back to flexible.
  Not true. Flexible subassemblies are FAR more complicated than that. They do not lose their flexibility when you close SOLIDWORKS. They lose it for all sorts of other reasons though, most of which are not repeatable and defy common sense. I have seen adding new mates cause supposedly over-defined flexible subassemblies to suddenly behave. I have seen deleting all mates cause well-behaved flexible subassemblies to suddenly say they are overdefined. Throw in a configuration or two and you may as well start flipping a coin to determine what will happen from one rebuild to the next.
  
  This is the madness I deal with on a daily basis designing five, six, seven or more axis custom machines.
  Like • Show 2 Likes2
  Actions
  - Alex Lachance @ Jim Sculley on Jun 3, 2020 4:08 PM
    Mark Correct
    Correct Answer
    Indeed, but the most common cause is generally the resetting of a session.
    
    Other causes include when there's a mate error.
    
    My habit when working with flexible subassemblies is always to ''reset'' them when I start my session. That general avoids the hassles you speak of. A flexible sub-assembly that hasn't been ''reset'', which has a mate added or removed from it, will turn red.
    
    Do you mate around your environment? I know how you feel, it took me a long time to understand them and even longer to have the people I work with understand how they work, but now we barely get any problems when working with them.
    
    Of course, our flexible sub-assemblies are generally ''flexed'' only once in a while, so I suppose you might just be in that case where you need flexible sub-assemblies at all times and that is where it causes you trouble.
    Like • Show 0 Likes0
    Actions
    - Jim Sculley @ Alex Lachance on Jun 4, 2020 8:41 AM
      Mark Correct
      Correct Answer
      Alex Lachance wrote:
      
      Indeed, but the most common cause is generally the resetting of a session.
      
      I don't find this to be the case. It may be for your particular flexible assemblies, but I would be careful not to generalize.
      
      Of course, our flexible sub-assemblies are generally ''flexed'' only once in a while, so I suppose you might just be in that case where you need flexible sub-assemblies at all times and that is where it causes you trouble.
      If I only had to flex them once in a while, I would simply use configurations. These are motion axes. I have to check things like clearances and reachability all along a tool path. To make matters worse, I often have more than one instance of the same flexible subassembly in a top level assembly. I can add the same mates between each instance and top level components and the two subassemblies will behave wildly differently when trying to move them by dragging the same component. Often, one will move and the other will refuse to. The inconsistency is maddening.
      Like • Show 0 Likes0
      Actions
      - Matt Juric @ Jim Sculley on Jun 4, 2020 9:04 AM
        Mark Correct
        Correct Answer
        Jim Sculley wrote:
        
        If I only had to flex them once in a while, I would simply use configurations. These are motion axes. I have to check things like clearances and reachability all along a tool path. T
        That's my general use for Flexible subs as well and for the most part it breaks most of the time. I had an assembly that was essentially a four link system. I had to check clearances with a dozen or so items along it's path. It had a limit in both directions that once passed it started flipping parts all over the place and then just broke and would no longer allow you to move anything. The only "Fix" was to suppress a few mates. manually reposition some of the parts and fix them, then unsupress the other parts. I bet I had to do this 20 times while designing the thing.
        
        I also just ran into a problem with limit mates...which just refused to work. Two configurations with a telescoping cover. Telescoping cover had limit mates. EVERY time I changes configuration those limit mates broke. Rebuild did not fix it despite the fact that it was well within those limits. Worse yet it would randomly select one of the parts and only some random location on that part that I could select and drag and poof...fixed. So every time I changed configurations I would have to select and drag some surface randomly on one of the parts randomly and attempt to drag it until I found the "Chosen one" that worked.
        
        Why limit mates...Flexible assemblies. I had to check clearances along it's entire path with several items so hard mates would not work. Limits worked while dragging and checking....but completely exploded for configurations.
        
        SW really needs to either go back to basics with mates and make a few of them just rock solid or spend the time making what they have rock solid.
        Like • Show 0 Likes0
        Actions
      - Jim Sculley @ Jim Sculley on Jun 4, 2020 9:11 AM
        Mark Correct
        Correct Answer
        Or, stuff like this. Here is a perfectly happy assembly.
        Here it is when added to a top level assembly:
        The mates are all suppressed because they can only solve correctly if the subassembly is flexible. When I make the subassembly flexible, I get this:
        So, a subassembly with no active mates goes completely haywire just by making it flexible.
        Like • Show 0 Likes0
        Actions
        Alex Lachance @ Jim Sculley on Jun 4, 2020 9:28 AM
        Mark Correct
        Correct Answer
        Jim, looking at your tree you are most likely doing that bad practice that I spoke of. There is nothing fixed within that sub-assembly. Is there something that is mated to the origin or to all 3 planes? I'm sure there's a component which you can identify that doesn't really move which could be fixed or mated to the environment.
        
        Once you start doing this, your flexible sub-assemblies will respond better.
        
        Edit: I always use planes as referencial mates for flexible sub-assemblies
        
        Like • Show 0 Likes0
        Actions
        Matt Juric @ Alex Lachance on Jun 4, 2020 9:33 AM
        Mark Correct
        Correct Answer
        Alex Lachance wrote:
        
        Jim, looking at your tree you are most likely doing that bad practice that I spoke of. There is nothing fixed within that sub-assembly. Is there something that is mated to the origin or to all 3 planes? I'm sure there's a component which you can identify that doesn't really move which could be fixed or mated to the environment.
        
        Once you start doing this, your flexible sub-assemblies will respond better.
        I can't speak for Jim but in many cases for me the "Fixed" item may have 2, 3 or more layers of "Flexible", moving assemblies on top of it. So you have a base, that has a SA on it, that has an SA in, That has an SA in it etc etc that all move. The only "fixed" item is the base.
        
        For me I end up working all these SA's as separate steps as if I even attempt to do more than one level of flexible SA's SW just explodes 90 percent of the time.
        
        Very time consuming when you have to check clearances of 2-3 SA's in 5-10 different positions. Lots of iterations versus just being able to have mates that work.
        
        Even with that mates in a single Flexible SA tend to blow up on me....what am I saying, mates blow up on me even without flexible SA's.
        Like • Show 1 Like1
        Actions
        Alex Lachance @ Matt Juric on Jun 4, 2020 9:39 AM
        Mark Correct
        Correct Answer
        Yeah I'm sure multiple different flexible sub-assemblies will give a mess. I'm probably having more success then most people because I can generally lock most movement directions.
        Like • Show 0 Likes0
        Actions
        Matt Juric @ Alex Lachance on Jun 4, 2020 9:53 AM
        Mark Correct
        Correct Answer
        Alex Lachance wrote:
        
        Yeah I'm sure multiple different flexible sub-assemblies will give a mess. I'm probably having more success then most people because I can generally lock most movement directions.
        It's not unusual for me to have a SA on top of another SA and both of those SA's have at least two degrees of freedom. Just picture something simple like an X-Y table on top of an Y-Z table.
        
        Get into something more complex and SW becomes unusable unless you break it into steps and do most of the interference checking in iterations.
        Like • Show 0 Likes0
        Actions
        Jim Sculley @ Alex Lachance on Jun 4, 2020 9:46 AM
        Mark Correct
        Correct Answer
        Fixing components makes no difference. I always fix a single component in an assembly and it is always aligned with the three primary planes. That's not my whole feature tree shown above.
        
        This assembly includes 6 other subassemblies that must be flexible and one of those is flexible two levels deep. Some of these have assembly level features which contribute to the problem. I've been using SOLIDWORKS since before flexible subassemblies existed. They are unreliable/unpredictable now, and they always have been. There was a time, long ago, when in order to have two instances of the same flexible subassembly, you had to use two different configurations. I think it used to work a little better when that restriction was in place. Now, its like herding cats.
        Like • Show 0 Likes0
        Actions
        Jim Sculley @ Jim Sculley on Jun 11, 2020 10:05 AM
        Mark Correct
        Correct Answer
        
        Fixing components makes no difference. I always fix a single component in an assembly and it is always aligned with the three primary planes. That's not my whole feature tree shown above.
        
        This assembly includes 6 other subassemblies that must be flexible and one of those is flexible two levels deep. Some of these have assembly level features which contribute to the problem.
        Here's an example of what I'm talking about. We buy a slide:
        And then modify it:
        The original slide is placed in an assembly as the only part, and then assembly features are used to create the modifications. This gives you a proper BOM structure: modified part is made from purchased part.
        
        I can move the slide to any position and the assembly feature remains correct, because all the sketch relations are tied to features of the moving slide:
        
        Put this in a top level assembly, make it flexible and all hell breaks loose:
        
        Where did those two holes go?
        
        That assembly feature should never be affected by the positioning of the slide in an upper level assembly, but here we are. And sadly, the problem often doesn't show up right away, and it comes and goes, even for the same assembly. I have my fingers crossed that this has improved since 2017 SP5, but I'm not getting my hopes up.
        Like • Show 0 Likes0
        Actions
  - Matt Juric @ Jim Sculley on Jun 3, 2020 4:40 PM
    Mark Correct
    Correct Answer
    SW needs to seriously work on fixing mates. I, like you, have run into all sorts of weirdness with mates. Mates that are clearly correct will sometimes cause the entire model to go red. I've deleted mates and added the exact same mate back and all is well. Mates seem to have zero intelligence with one minor conflict causing everything in the entire tree to go red and yellow. Yeah I get that this screw being mated backwards can thru a series of 4000 steps can cause that screw over their to flip as well...can we please have a bit of intelligence here because even a child playing with lego's knows that doesn't happen in reality.
    
    Also completely locking up an entire assembly from being able to move parts, use flexible assemblies etc because you have a mate error is maddening. So you move an assembly in a manner that causes an error or over travel...sorry, you can't move it back. Geez how hard would it be to simply ignore the mates until they can go back to the right position?
    
    Even some basic ideas seem to be intuitively wrong. A shaft in a hole should not have a direction associated with it. The shaft can not control what side of it is put in first. However if you place a shaft in a hole and then mate the end of that shaft to something it causes problems with the direction of the shaft mate...?!
    
    This is a level of complexity that is not only not necessary but causes a ton of problems. I can't tell you how many assemblies I've mated, saved opened only to find a part flipped or a mate red because the mate shaft direction has decided to flip for some unknown reason.
    
    And angles...no idea what good angle mates are in SW. They appear to randomly change side to side whenever they want. So they over define a concentrically mate and under define angle mates.
    
    When they get done with that maybe they can move back to something more basic....like dimensions loosing attachment every time a small breeze comes by and a painfully slow if not impossible procedure to re-attach them.
    
    Sorry about that I've had a very grating SW weeks with crashes wierdness etc. I had a drawing randomly change the model it was referenced to to another model and completely blow up a seven page weldment drawing. After attempting to fix this for a while I just gave up and went back to an older PDM version.
    
    Oh, and don't get me started on mates while replacing a part....good lord, really, I'm replacing a part I want the mates to mate the same way they were. Don't choose to go the opposite direction and then completely explode the entire assembly based on the part I just replaced.
    
    ok...I'm quitting now...
    Like • Show 1 Like1
    Actions
Walter Fetsch Jun 3, 2020 3:57 PM
Mark Correct
Correct Answer
Keep in mind that you cannot have a flexible sub-subassembly. Flexibility only works one level deep.
Like • Show 0 Likes0
Actions
- Alex Lachance @ Walter Fetsch on Jun 3, 2020 4:00 PM
  Mark Correct
  Correct Answer
  You can have multiple level deep, it just brings another dimension of complexity if the person has a hard time understanding the environment of SolidWorks.
  
  We have flexing sub-assemblies that go 8 level deep.
  Like • Show 0 Likes0
  Actions
Bill McEachern Jun 3, 2020 4:51 PM
Mark Correct
Correct Answer
Nice rant. You are likely to young to remember or more likely never experienced Pro/E when this kind of thing led to the BSD (Blue Screen of Death). SWX has the white SCREEN OF HOPE. Next time I spend money on a CAD system it will be to buy Onshape where this stuff doesn't happen. What about when up do an upgrade and things stop working? The material database goes in the ditch, flow crashes when you apply a BC, simulation loses your results. It's never ending. I love them all. Motion though it is the neglected child, it borders on abuse really.
Like • Show 1 Like1
Actions
- Matt Juric @ Bill McEachern on Jun 3, 2020 5:02 PM
  Mark Correct
  Correct Answer
  Never used Pro-E, but I'm definitely old enough to have had the opportunity.
  
  I used Unigraphics pretty sparingly in 91/92 while using ADesk and never left that family AutoCad, MDT and Inventor until I changed jobs and moved to SW in 2016.
  
  It's stunning to me that for the most part mates in SW are actually less stable than they were in MDT and early Inventor...and then theirs the downhill slide of dimensioning. I guess it must be all the push toward model based manufacturing...."Nobody is using that tired old dimensioning tool anymore so let's just stop making it actually work."
  Like • Show 0 Likes0
  Actions
  - Bill McEachern @ Matt Juric on Jun 4, 2020 12:35 PM
    Mark Correct
    Correct Answer
    SWX 2009 was pretty good. It was the last version with the old motion interface I think. Might have been 2008 but whatever it worked really well. I have noticed the mate stability has deteriorated quite a bit if the last few releases. I doubt that Dassault is going to put much resources into fixing anything. They want you to go to 3DX which is where all their effort is going.
    If you want to do motion analysis with a minimum amount of grief save eveything that acts like a "part" (no relative motion) and re-mate an assembly of those parts for motion analysis. It keeps the mayhem to a minimum. Also the ability to over constrain things is really easy. Most of the time motion handles it fine but on occasion you have to pay attention to the DOF and mate with no over constraints to get robust performance. It is ADAMS underneath it all and the rigid body approximation is rather demanding so you have to pay attention. Good luck.
    Like • Show 0 Likes0
    Actions

Find conflicting mate

Outcomes

Related Content

Recommended Content