EX pattern persists in simple loft

Having some difficulty with a simple 1-rail loft of a bent metal strip. Working in Rhino 7.

EX patterns appear in the loft. I used freeform, natural, and flipped a couple of arrows to get them all pointing in the same direction. Also added several templates, thinking this might resist the tendency of the edges to cross over, forming EX;s. No luck. Being careful to click on the same part of each template in setting up the loft. Doesn’t help.

Here is a .jpg screen grab showing the pesky effect:

Here is the 3dm including the loft and its setup.

Thank you for your insights. Michael
EX patterns in loft.3dm (627.6 KB)

Hi Michael - this will work if you get all the curve direction arrows going the same way - the ones at the top like to be flipped.

-Pascal

Many thanks Pascal. Great to hear from you. The loft is improved but still a little flawed.


Here is a jpeg of the loft set up after the top two arrows were flipped. Note that their direction is not identical to those of the other arrows in the array of arrows. Rather, they are at right angles to the majority of arrows. Also please note that there is an arrow akimbo in the lower cusp of the rail curve. After experimentation I left it akimbo because it produces a nice clean loft in the lower cusp.

The upper cusp, where I flipped the two arrows, is pinched flat.

Not understanding the logic of this procedure, nor can I figure out what I should look for in a pattern of arrows in order to get a clean, predictable loft.

Many thanks Pascal. Michael

Hi Michael - with a convoluted curve it is tricky but ‘conceptually’ drive along the path curve and as you get to each shape curve make sure the arrow is a) on a corresponding vertex to the previous curve and b) matches either cloickwise or counter-clockwise from that direction.

There is a button on the Sweep1 dialog that lets you go back in and monkey with the seam points if the preview does not look right. (Actually it looks like the direction is right but the location on the rectngles is wrong)
image

-Pascal

Many thanks Pascal. This was very helpful. Michael