Running some test I found out my printer was over extruding quite a lot.
A test I did was printing a box with thin walls and measuring the thickness of the walls and compare it to the extrusion width set in Simplify3D. I wound my actual extrusion width was over 125% the width set in simplify3D so I lowered the extrusion multiplier and voilà, now my thin walls are the right size. I thought it would make a small difference, but it actually did a huge one in negative space details!

When trying to run the Make magazine's negative space test found here http://www.thingiverse.com/thing:533472/ I could only remove the first 2 pegs, after the new settings I can remove 4 out of 5 pegs, and I didn't scale the model or anything.

The way I set it to get these results is:

Nozzle diameter: 0,40mm (of course)
Extrusion multiplier: 0,82
Extrusion width: 0,48mm

Hey all,

I have been doing a lot of small testing with my DaVinci and Im having a difficult time finding the proper sizing method to compensate for ABS shrinking and/or shell overlap.  For example, if I model a nut and bolt, I have to upscale the nut in order for it to properly screw onto the bolt.

Im trying to print a ring collar for a 14mm tube with a snug fit.  It is proving to be a challenge; either its too large or too small.
My process, so far, has been simple trial and error of printing out a small piece to see how it fits, then adjust the model accordingly.
My test was modeled at 14.77mm and measured with my caliper to be 14.1mm after it cooled, so there is about a .68mm difference.
I cannot tell if the size difference is due to shrinkage or shell overlap.

Is it all trial and error or are there any general tips/tricks to getting the proper sizing tolerance? 
Thanks!

If you are using Simplify3D the last version upgrade has a feature called 'tolerance adjustments' to do exactly what you need.

Ok, that is understandable.  I tried the same test print with a different filament and it was very loose, so I'll have to make adjustments per filament.
I've read just the pigment in the filament will greatly vary the flow and thickness of the print.

That is correct, and it affects the temperature as well. Unless color is important I always use natural as it has no pigment so the shrinkage and temperature variables will be minimum.

Hey all,

I have been doing a lot of small testing with my DaVinci and Im having a difficult time finding the proper sizing method to compensate for ABS shrinking and/or shell overlap.  For example, if I model a nut and bolt, I have to upscale the nut in order for it to properly screw onto the bolt.

Im trying to print a ring collar for a 14mm tube with a snug fit.  It is proving to be a challenge; either its too large or too small.
My process, so far, has been simple trial and error of printing out a small piece to see how it fits, then adjust the model accordingly.
My test was modeled at 14.77mm and measured with my caliper to be 14.1mm after it cooled, so there is about a .68mm difference.
I cannot tell if the size difference is due to shrinkage or shell overlap.

Is it all trial and error or are there any general tips/tricks to getting the proper sizing tolerance? 
Thanks!

There is no set number as it varies from reel to reel even with the same vendor. What you can do is make a test object. Something smaall with known dimensions and some holes as well. When you get a new spool. print that object and measure. Note the deviation and compensate in your measurements while using that spool.

I have been doing a lot of small testing with my DaVinci and Im having a difficult time finding the proper sizing method to compensate for ABS shrinking and/or shell overlap.  For example, if I model a nut and bolt, I have to upscale the nut in order for it to properly screw onto the bolt.

Im trying to print a ring collar for a 14mm tube with a snug fit.  It is proving to be a challenge; either its too large or too small.
My process, so far, has been simple trial and error of printing out a small piece to see how it fits, then adjust the model accordingly.
My test was modeled at 14.77mm and measured with my caliper to be 14.1mm after it cooled, so there is about a .68mm difference.
I cannot tell if the size difference is due to shrinkage or shell overlap.

Is it all trial and error or are there any general tips/tricks to getting the proper sizing tolerance? 
Thanks!