Mixing gasses

[willem79]

Dear all,

I've been asked to solve a problem involving fluid dynamics and I wonder if anyone could help me find a way to approach this problem.
Situation:
Two gasflows of different flow and temperature are mixed using a T-piece.
Known or previously calculated values:
flows, temperatures of both streams (multiple scenarios).
and density
Question:
How to calculate distance traveled for the mixed stream until gas temperature is homogenous?

 

[dcasto]

without a mixer, I've seen to gases stay seperated for 100's of feet. Literally, the top of the pipe was hot and the bottom cold.

 

[willem79]

Thanks for your reply dcasto,
I'm also afraid that this is one of those non lineair problems with a strong dependance of starting conditions.

In other words, withouth a static mixer theres no telling what will happen?

Greets and thanks,

 

[moltenmetal]

Is flow turbulent? If flow is turbulent, how exactly do you get density-driven stratification in anything other than a very large diameter pipe?

 

[dcasto]

i've seen it in action on ethylene gas. We would split the stream, sending 40% trough a heater to raise the temperature to 160 F and the remaining was at 65 F. When the stream came together at 900 psi (above critiicle point, the top would be hot because the lower density, and the bottom cold. The gas would then be measured by an orifice meter, Reynolds number was over 1,000,000.

I've seen ethylene climb over N2 during purges of pipelines, this was at low reynolds numbers or static conditions.

 

[willem79]

Here's some data i received.

pipe diameters all DN600 (600mm)
both streams are of same composition (smoke)
largest stream comes in the T-piece horizontally
(T= approx 200 degC).
other stream (massflow about 1/3rd) comes in from top (branch)
(T = approx 500 degC).
all speeds in excess of 15 m/s... so they are turbulent

combined stream enters a smokestack after about 10m.

The client is afraid for hotspots downstream of the T-piece so special steel grade is used. Question to me is what length of special pipe is needed downstream of the T-piee until connection to normal grade steel.

Intuitively you would say injecting the hot stream perpendicular at these speeds would create such turbulence that temperature difference is immediately nullified... but how can you say for shure... is there a way to play it safe (except for changing all the piping domwstream into special grade which is expensive)?

 

[dcasto]

I'd just put a mixer in. a couple of baffle plates will work.