Somebody put me out of my misery please

My question is about associativity and why when I iterate (ie. even to change the nature of a defining spline minutely), I must delete my surfaces and solids, 'remove links' and rebuild my object from the driving curves.

In order to illustrate this, I have made a very simple model - we can call it the 'butter dish lid' (attached)

It uses a series of intersecting surfaces, which I trim back to one-another and stitch into a solid

(The method is exactly the same as I use to create enormously complex free form shapes, which I can then feature with holes, bosses, and rad blends and then shell)

So the problem ... I expect to be able to utilise 'associativity' in my 'drive curves' to affect the surfaces, in turn modifying the surface intersects (and of course trims), thus in turn redefining the perimeter of the 'butter dish lid'. (... Ultimately this would naturally update a 'stitched solid', several feature holes perhaps, some bosses and then rad blends and ultimately the shell, let's say).

Should just work, right !?

Model is attached - You'll see that if you adjust one or more of the splines (in the plan view), the surfaces will update - as will their intersections to one another ... Great !

One more note - I have not put a cover surface on the bottom of the butter dish lid, so you'll be unable to stitch it into a solid until you do. This may be where the problem lies.

If you do add the necessary cover surface - (and go back to the top view) - you will see that you can no longer adjust the driving spines and have the driven surfaces and intersects recalculate. (A given spline point will just nonchalantly return to where they were before you tried to move it).

Further, if you stitch your 6 surfaces into a solid - you will then encounter error dialogs (and the adjustment will fail as before)

At this point, in a live project, knowing I must adjust a driving spline, I would normally dump the surfs / solids and remove all links before rebuilding - but this simply CANNOT be right. I waste days rebuilding and rebuilding and rebuilding.

So what am I doing wrong ?

Help Meeee !

OK - even simpler example

(The more an example of a fail can be distilled - the easier it is for the dev teem to fix, right ?)

Here is a square polygon with a cover surface

It's split by a surface (which is extruded from spline A)

(Hide half of the split cover)

Then extrude a surface (call it B) from the split edge of the remaining half of the cover

Associativity DOES NOT EXIST between spline A and extrude B

This is fundamental !

I have it on mine. I used the surface to split the other surface, not the curve. Grab the interior point and move it laterally. You will see the surface split moves too.

Edited by user Monday, February 5, 2018 5:55:28 AM(UTC)  | Reason: Not specified

I cannot seem to get my mind wrapped around the concept of unresolved links in Shark. I resolve them before I save but other than that, clueless.

And yes, I still have not helped you in any way. I will try one fresh here and see if I get the same result.

Try this one.

Still has "unresolved links" after a change. But the more that I think of it, this is nothing more than a Creo, Catia or SolidWorks regen. Just in Shark it's an explicit thing.

The trick I found on this one was I could not extrude the blue surface from the split edge. I needed to add an surface intersect to get the curve at the split and used that as the drive curve for the blue. I layered the items.

There is a spline driving the large surface. Modify that one and the splits/blue surface go along for the ride.

Edited by user Monday, February 5, 2018 6:40:37 AM(UTC)  | Reason: Not specified

It appears so, Shark needs an explicit "something" to reference when driving an extrude. I am in the middle of working a funky shaped "butter lid" as well. I have the shape modifiable with the same technique just used. But, now I am seeing that to have a solid that will accept a draft feature, the technique will need to get a bit more complex. Ahhhhh, the work-arounds, my friend....the CadLand entry fee.

I will upload it when it's ready.

OK - going back to the 'butter dish lid' - I can now get it to update a stitched solid

- thank you Dean, I needed a different perspective

(Unfortunately, on updating a driving spline - I then had issues with 'unresolved links' - but edit>resolve links did fix it - woohoo !)

every time I update a driving spline, I get this nonsense

thankfully if I edit>resolve links, it goes away

.. but can't it check for integrity when it rebuilds it ?

I accidentally rear ended a car at a stop light on my way to work this morning. A "little person" climbed out of the car and came over to me. I said "Are you OK?". He said, "Well, I'm not happy." I said, "Well, which one are you?"

Here's my take on son of 'Butter Dish Lid'

It's a shelled solid

You should be able to adjust the driving splines and have the solid regenerate

Notes: unsubtle moves will result in no-regenerate (unsurprisingly)

... any regen will cause 'unresolved link disease' ... but this may be resolved by edit>resolve links

What a ball ache to get this to work - thanks to Dean for the 'intersect surfaces' work-around !

So Tim ... listening ?

1. - 'Surface Split' edges ARE usable as drive curves for downstream extrudes etc ... BUT they need to maintain ASSOCIATIVITY

... Tim & co ... is this possible ??

2. - On regenerating, a part must check itself for UNRESOLVED LINKS and FIX ITSELF (if possible).

Can we do this ?

Thanks !!

Jol