What will flow better a 2” tube or 2” rectangle with 3/8” radius corners

Well, π × radius2. So 1x1= 1 x 3.1416 = 3.1416

and 2 x 2 = 4

So with out taking time to figure the loss of the radiused corners, I'd say, with all else being equal, the area of the rectangle would be around 25% more.

The rectangle.

There may be a slight loss of flow due to the longer walls of square tube, but not as much as the difference in square area? Take note of the choice of shapes car manufacturers use for intake runners. Square would be more compact, but i see a lot of round intake manifold ports.

intake or exhaust side?
Vacuum or pressure?

All true. Boundary layer separation is real.

He didn't say what he was flowing, how long the tube is, or if there are any bends. I just assumed two pieces of tube, smooth surface, maybe 6" long and flowing something like fuel or water, and figured the area. If air, the same pressure pushing it. Figure the exact percentage including the corner radius should get you closer.

Correct on that

When I port the exhaust on heads I care more about airspeed (fps) than I do cfm. It’s interesting testing an exhaust port with a six inch piece of 2 inch pipe on the outlet.

There is a reason intake manifold runners aren't round.

Kevin

ok, fill us in

I recall reading that air flow through a round pipe will form a vortex that restricts flow. Perhaps that is the issue here.
Since headers and exhaust pipes are normally round maybe this only applies to vacuum induced flow and not pressure, if so then round intake ports may work fine with boost but not too well when normally aspirated.
I am not a fluid dynamics expert, just letting my mind wander late at night.

I'm no expert either but it was explained to me that it's the opposite, a round tube flows better than square or rectangular. A quick google search seems to hold this up, in general at least.

Here is a great article written by Jim McFarland about intake manifold design. https://www.motortrend.com/how-to/the-manifold-tech-engine/
He says the problem with round runners is fuel separation at direction changes.

Edit> That link takes a while to load his pictures and diagrams because of all the stupid ads.

Adding some, Fords tunnel port heads-intake, Mopar W2 oval ports, Chevy heads with the tall narrow pointed top heads. Just the domestic V8 stuff not overseas stuff. Thoughts?

Yep, not sure anybody really knows what all is going on in there. Some of the newer strange looking OEM designs may be because port injection, direct injection, turbo charging and supercharging changed everything they thought they knew about wet and dry flow and fuel separation.

I found an article about racing turbo 6 cyl intakes (Toyota perhaps). They said that round runners would cause air flow to stall at high boost. Why then does every turbo have round tubing between the turbo and the manifold? Why doesn't the same physics apply before and after the plenum?
Maybe this is just an example of what my signature says.

My thinking too, all about boundary layers imo, on a semi square tube the radius corners could break up that layer? On the S/S A cars they made the intakes and heads round to flow more, I guess. Remember the World Cup yachts incorporated a ribbed hull (like a whale belly) to make less drag in the water, they, hulls and tubes all have friction of some amount so it’s where do they want the least friction

What I’m looking at, dang upside down!

What will flow better a 2” tube or 2” rectangle with 3/8” radius corners

Round area for 2" tube is ΠR²: 3.1416in²
Square area L²: 4in², minus the corner radiiof .141" = 3.86", or 23% greater,

Thanks, I thought you would show up, you’re a smart guy!

Maybe the question should be: "What will flow better, a round tube with a 3.1416" cross sectional area OR a square tube with 3.1416" cross sectional area ?" I don't know if anyone has ever tested this
On the other hand, I don't think I've ever seen an exhaust manifold made from square section tubing, but then again I haven't been looking

Actully an intake, cross section is a tad over 2” maybe 2-1/8 x 2-1/8