Ah, that is true. You're not being a PITA, by the way, this isn't my job to post here I just do it in the hopes of learning something.
With some testing I found that -2 Thicknessses results in an OD of 27.973 in vs your 28. This is using 12ga.
I made a sketch inside of the transition file with a skeletal representation of the profile of one of the transitions (drawn to OD). I put a driven dimension to the id and named it "id". Setting the transition then to id (the parameter) gave me 28.001 on the OD.
I also cam up with this nightmare formula for calculating it directly, but IV doesn't understand the inverse tangent. *(I'm sure I facked it all up anyways):
id = cos(tan^-1(h/((l-d)/2)))*T
where h = height, l = length, d = diameter, T = thickness
As I said - you need to suppress the Delete Face feature in the feature tree to get the flat pattern to work. Here is an iPart which has two entries in the table. One of them - for producing the Flat Pattern only. But if you don't mind the small notch in the corner - then you don't need an iPart.
On my computer your part doesn't completely flatten all the way and I can't figure out why. Does yours go completely flat?
My apologies Igor (on two fronts). One for the fact that I read that part of your post but by the time I was free to open it I had forgotten, and two for replying so late. I have been out of town.
Thank you (and all the others) for the wonderful examples and usable information. Utilizing a "mixed bag" approach to the items posted here I was able to produce a file the "bend guy" was looking for and it worked great.
Now, I just need to figure out a way for me and my cohorts to convince him to do it the "normal" way.
Here's the picture of the transition fit together. They said it went together like a "swiss watch". I think that means it was okay.