Going bigger with functional single-wall print structures. Started with smaller fast-to-print architectural models. Went a bit bigger. 
The red columns are printed with 0.5mm layers and 0.8mm walls (using a 0.8mm nozzle). As simple cylinders, the thin walls would be prone to crumpling. Was not sure if the 300mm column (center) would be sufficiently rigid. (It is.)
The blue bands are friction-fit to the join the columns. Seems to work.
#MadScientist
Looks good. Are you experimenting with tall/short layer heights too? 0.5 tall is pretty thick for good cooling but maybe that’s an advantage for bonding strength in this case.
@Ryan_Carlyle You think I did all those test prints only as I like rockets? Well, partly that. 
First, found single-wall prints would slump in sections more curved with overhang. (The tops of the nacelles in particular.) Needed a better duct and stronger part cooling fan.
Second, had problems with inter-layer adhesion on overhangs with 0.6mm x 0.8mm layers. Dropping to 0.5mm layers seems to make quite a difference, so 0.5mm x 0.8mm became my usual starting point with the 0.8mm nozzle.
Have done some prints with the 0.5mm x 1.0mm or 1.2mm layers. Stronger, or course, and seems to do a bit better on overhangs (as we might expect).
Have not yet put on the V6 hotend / duct. Still using the mk8 hotend, with an improved (but could be improved) duct and stronger part cooling fan.
As to cooling this particular print, given the large diameter for one layer, not entirely sure part cooling is needed. So started a print with the part cooling fan completely off. Looks good so far - for the first ~40mm of height. (Might be stronger?)
Is your concern for smaller more detailed prints?
@Preston_Bannister Hmm, well, a general point I try to make is that large nozzles and wide extrusion widths don’t require tall layers like many people seem to automatically assume. (Don’t know whether or not you’re assuming that.) A 1mm wide by 0.1mm or 0.2mm tall layer is perfectly doable. For example, when people using Volcanos drop their print speed to 20mm/s to print giant strands on a printer that can happily do 40mm/s… why not cut the layer height in half and double the speed? You’ll get the same flow rate and approximate print times, but with better Z axis resolution and overhang quality and print cooling.
Some people do claim thicker strands are stronger, but I haven’t seen much in the way of controlled test data for that, where the testing used settings/conditions properly optimized for each layer height. (EG with thin layers you should use less airflow because there’s a higher surface area to volume ratio and the print cools faster.)
Without part cooling (left) the surface is both more shiny and slightly irregular. Not quite as nice a result, but usable.
@Ryan_Carlyle Fair point. With the hotend limiting the extrusion rate for 0.6mm x 0.8mm to about 30mm/s (with skipping at 40mm/s), there should be lots of headroom in the printer XY motion.
Finer resolution does not seem of obvious advantage for the above print, so started a print of one of my more complex test objects with layers dropped to 0.2mm and rate bumped up to 100mm/s.
Let you know in a few hours. 
Well, this was confusing, until I realized the part cooling fan was still off (from the prior exercise). Dropped the temperature when I noticed the print was going badly, but did not notice fan was off until after the print completed. 
Repeating with the part cooling fan on…
Red is 0.2mm layer height and 0.8mm width printed at 100mm/s with a 0.8mm nozzle. Nice finish but some artifacts.
Suggested as @Ryan_Carlyle noted the volume/rate of plastic is about the same for the different products of layer height and rate of movement.
Green is 0.6mm(?) x 1.2mm(?) and a different source (red is Maker Geeks and green is 3D Solutech), so the filament might be a cause (though they seem similar in printing). Had some prints at 0.6mm x 0.8mm and 30mm/s, but they left the state with my grandson.
Not sure about the artifacts. Going to punt on further experiments until I swap out the mk8 hotend for V6.