Adaptive not working the same when objects use padding

[Alex.S]

Hi,

I'm modelling some Armco, and I have the armco barrier and posts in separate Linear generators, I'm running into an issue that, even though I add padding so that the armco posts take the same space as each segement of the barrier, they don't act the same.

Both have 50% adapative, and I've added padding so the width of both adds upto 10m, however they behave differently. The strange thing is with different larger armco posts it seems to work okay.

I've disabled the end and start pieces to make sure there is nothing interfering.


Bit confused here.

[Alex.S]

To show this a little better, I did another test where in a transform option I set the fixed width of the start, default and end post segments to the same as the armco and removed the padding, and yet it still does not behave the same?

Red is the split between the posts and the blue is where the armco splits. They should be at the same point though... :/

[Paul Roberts]

Hi,

Thanks for contacting us. I can't reproduce that behaviour yet here but can I just check that the length of the armco and the posts are using exactly the same segment length on the X axis? For example you cannot have an railing segment that's 300cm on the X with 3 posts segments with a size (after padding) of 100cm, This is because the adaptive calculation for these will be different if, for example, the spline was 500cm you'd get two railing segments but 5 post segments and the alignment would be incorrect. To get around this you could put 3 posts in a compose operator before wiring them to the default input, in that way they will be the same length for the purposes of calculating the adaptive distribution.



 I've attached a sample file, perhaps you could take a look and see if they differ from the approach you are using. If this doesn't help please feel free to send me your problematic .max file and I'll happily take a closer look.

Many thanks,

Paul

[Alex.S]

Hi,

Yes I made sure that after padding the widths are identical.

I have multiple posts in a single object for optimization purposes, I've created a very basic test here which shows the crawl in the middle post. Not sure if this is exactly the same issue as I have... I am testing a little bit more.

[Alex.S]

I'll try composing four posts as you do in that sample rather than having them pre-grouped.

[Alex.S]

Works via compose alright, it's when you try to setup the geometry with multiple poles already there things start going weird... I'll convert things to this method.

[Alex.S]

Well, works mostly, there is still a disconnect at certain spline lengths, despite the combined lengths being exactly the same (after padding)

[Alex.S]

Got it working all the time, user error sorry, there was some transform acting up I missed and lost a good few hours on