Mixing tee calculator

Type the two pipe diameters, velocities and temperatures. You get the mixed outlet temperature from the energy balance, the outlet velocity and Reynolds numbers, the momentum ratio and which flow pattern the branch jet makes: wall jet, deflecting jet or impinging jet. Free, no sign-up, runs in your browser.

Add the outlet standard deviation from your CFD run to see how mixed the outlet really is. It opens on the mixing-tee challenge of Skill-Lync's Fluent course; the second tab checks the outlet temperatures in your report and names the usual slips.

Energy balance · Igarashi, Kamide et al. 2004 · OECD/NEA Vattenfall benchmark · Danckwerts 1952 · Ansys Fluent 2024 R2 · Updated 20 Nov 2026

The outlet is the main pipe downstream, same diameter. Constant-density air is Fluent's default material and what the challenge solves; pick the ideal gas if you switched density to ideal-gas. The ratio box sets the branch velocity, or choose "Velocity" and type it.

Sketch of the wall, deflecting and impinging jet patterns in a tee (after Kamide et al. 2009, not to scale)wall jetdeflecting jetimpinging jet
Mixed outlet T30.32 °C303.47 K
Momentum ratio M_R0.636Deflecting jet
Outlet velocity4.504 m/sbranch 33.4% of mass
Kamide momentum ratio against branch-to-main velocity ratio for this tee, with the wall, deflecting and impinging jet bandswall jetdeflecting jetimpinging jet0.010.111012345678velocity ratio Vb / VmM_R (log scale)
  • The branch jet reaches into the middle of the main pipe before the main flow turns it downstream, so the mixing layer sits away from the walls.

What is the mixing tee challenge?

It is the first steady-state project of Skill-Lync's Fluent course ("HVAC Simulation Inside a Mixing TEE", the lesson "Simulating Flow Through a Mixing T Joint Using ANSYS"), and 619 public student reports cover it, 506 of them titled "Week 1 - Mixing Tee". Hot air at 36 °C enters the straight run of a tee at 3 m/s; cold air at 19 °C enters the side branch. The brief asks for steady simulations of a short and a long tee, each at a "momentum ratio" of 2 and 4, a choice of turbulence model (k-epsilon or k-omega SST), and a comparison of how well the two streams mix, judged by the average and the standard deviation of temperature at the outlet.

The brief gives no diameters, but 19 reports print the inlet areas Fluent reports, 901.3138 mm² and 225.9823 mm²: a 33.88 mm main pipe and a 16.96 mm branch, area ratio 3.99. The calculator opens on that geometry.

What outlet temperature should you get?

Energy in equals energy out. With adiabatic walls, one fluid and a constant specific heat, the outlet temperature is the mass-weighted mean of the two inlets, and the specific heat cancels:

With constant density the density cancels too, and mass conservation gives the outlet velocity:

BranchShare of flowToutVout
6 m/s33.4%30.32 °C
303.47 K
4.50 m/s
12 m/s50.1%27.49 °C
300.64 K
6.01 m/s

So 30.32 °C (303.47 K) with the branch at 6 m/s and 27.49 °C (300.64 K) at 12 m/s. The second is almost exactly the plain average of 36 and 19, 27.5 °C, and that is not a coincidence: the branch has a quarter of the main pipe's area (1/3.99), so at four times the velocity it carries the same flow. The Reynolds numbers are 6,957 in the main pipe, 6,967 and 13,935 in the branch and 10,446 and 13,935 at the outlet: all above 4,000, so a turbulence model is right. Fluent's default air has a constant density; switch it to ideal gas and the colder, denser branch carries more mass, which moves the answers to 30.11 and 27.25 °C. Use the one that matches your material settings.

A converged steady run with adiabatic walls should reproduce the balance to a few hundredths of a kelvin: the outlet values in the reports cluster at 30.3 °C and 27.5 °C. When it does not, the usual causes are an unconverged outlet monitor, backflow at the outlet, or the averaging below.

What is the momentum ratio of a mixing tee?

Three different numbers go by that name. For the course's two cases they come out at 2, 1.00 and 0.64, and at 4, 4.01 and 0.16:

  • Velocity ratio Vb/Vm. This is what the challenge means: 387 of the 402 reports that state the branch velocity set it to 6 and 12 m/s for "momentum ratio" 2 and 4.
  • Momentum flux ratio, the momentum each pipe carries in per second, branch over main:
  • With a quarter of the area, the course's "momentum ratio 2" is a momentum flux ratio of 1.00, and "4" is 4.01. A momentum flux ratio of 2 would need the branch at 8.47 m/s.
  • Kamide's momentum ratio, the one T-junction studies use to predict the flow pattern (Igarashi, Kamide, Tanaka and Kimura 2004, from the Japan Nuclear Cycle Development Institute's water tests on a 150 mm main pipe with a 50 mm branch; Kamide et al. 2009 use the same three patterns). It puts the main flow on top, and the main pipe's momentum through the branch's projected width DmDb:

Large MR means a strong main flow that bends the branch jet over at once; small MR means a strong jet that crosses the pipe. They classify three patterns, for MR from 0.05 to 32:

MRPatternWhat the branch jet does
above 1.35Wall jetThe main flow bends the branch jet over at once: it runs along the pipe wall on the branch side, so the boundary between hot and cold fluid lies along that wall downstream of the tee.
0.35 to 1.35Deflecting jetThe branch jet reaches into the middle of the main pipe before the main flow turns it downstream, so the mixing layer sits away from the walls.
below 0.35Impinging jetThe branch jet crosses the main pipe and hits the opposite wall before it turns, spreading the branch fluid across the pipe close to the tee.

The limits were measured with a branch one third of the main diameter. The authors say their own tests cannot show the effect of the diameter ratio; the JSME guideline on high-cycle thermal fatigue of piping adopted the classification after visualisation tests at other ratios. The course tee's branch is half the main diameter, so read a case close to a limit as on the boundary, not on one side of it.

In the course tee, MR = Dm / ((π/4) Db (Vb/Vm)²), so the pattern changes at velocity ratios of 1.37 (wall to deflecting) and 2.70 (deflecting to impinging). The two cases land on opposite sides: 6 m/s is a deflecting jet (MR = 0.64), 12 m/s an impinging jet (MR = 0.16). At 12 m/s the cold jet crosses to the far wall and spreads there instead of turning in the middle, and in the conclusions that compare the two in so many words, 16 reports find it mixes better and 2 the opposite. The chart in the calculator plots MR against the velocity ratio for whatever geometry you type. None of the 619 reports names the three patterns or Kamide's ratio.

Area-weighted or mass-weighted outlet temperature?

Mass-weighted. Fluent's Surface Integrals panel offers both:

The mixed temperature is what the outflowing mass carries, so each face counts by its mass flow. Picture an outlet with one half at 22 °C moving at 6 m/s and the other half at 34 °C at 3 m/s: the area-weighted average is 28.0 °C, the mass-weighted one 26.0 °C, and only the second closes the energy balance. In the course case the outlet is close to mixed, so the two differ by little, but 327 reports use the area-weighted average and 1 the mass-weighted one. Report the mass-weighted value next to the energy balance; then the gap between them is your solution error, not your post-processing.

Is the outlet standard deviation a good measure of mixing?

It is the right quantity, read against the wrong scale. 423 reports compare the standard deviation σ of temperature at the outlet between cases (the 144 that print one in text: median 1.15 K, middle 80% 0.45 to 2 K). But σ depends on how much there is to mix. If nothing mixed at all, part of the outlet at Tb and the rest at Tm, the standard deviation would be σ0, set by the branch's share x of the mass flow. Danckwerts' intensity of segregation compares the two:

I = 1 is completely segregated, I = 0 perfectly mixed. Some papers use the ratio of standard deviations, σ/σ0, instead of the variances; the calculator gives both. In the course, σ0 = 8.02 K at 6 m/s and 8.50 K at 12 m/s, because the 12 m/s case brings a 50:50 flow split, which has the largest σ0 of any split. A σ of 1.15 K means σ/σ0 = 0.143 at 6 m/s but 0.135 at 12 m/s: the same raw σ is slightly better mixing in the second case. Type your σ in the calculator to get σ/σ0 and I for your run. Two cautions: check whether your σ is area- or mass-weighted, the same question as for the mean, and a σ from a coarse mesh is smoothed by numerical diffusion, so check it in the mesh study like the mean.

The same 36 °C and 19 °C: the Vattenfall T-junction benchmark

The challenge's two temperatures are those of the best-known T-junction experiment. The OECD/NEA-Vattenfall benchmark (run at Älvkarleby, Sweden, in November 2008; CFD results compared in 2010) mixed water at the same two temperatures, a constant ΔT of 17 K, but the other way round: cold 19 °C water in a 140 mm horizontal main pipe at 9 L/s, hot 36 °C water from a 100 mm branch on top at 6 L/s. That is 0.585 m/s in the main pipe and 0.764 m/s in the branch, a mixed temperature of 25.80 °C, and MR = 1.04: a deflecting jet by Kamide's map, with the caution that its branch is 0.71 of the main diameter. Pick it in the calculator's "Start from" list. The benchmark measured temperature fluctuations downstream for LES codes to predict, which is the thermal-striping question this page leaves out.

Which slips does the check tab catch?

Skill-Lync read all 619 reports on 20 Nov 2026. Most outlet temperatures sit at the energy balance; the ones that do not come from a short list, and the Check tab tests your number against each:

  • 273 instead of 273.15. Inlets typed as 309 K and 292 K, or a correct °C answer converted with 273: 0.15 K low, 303.32 K for the 6 m/s case.
  • Velocities without areas. (3 × 36 + 6 × 19) / 9 = 24.67 °C treats the two pipes as the same size.
  • Diameters instead of areas, which happens to give 27.49 °C for 6 m/s, close to the right answer for 12 m/s.
  • The plain average, 27.5 °C, right only when the two mass flows are equal (the 12 m/s case, nearly).
  • Swapped velocities, the main pipe at 6 m/s: 34.11 °C.
  • Density model mismatch: an ideal-gas hand calculation (30.11 °C) against a constant-density run, or the reverse.
  • Area-weighted outlet values, above.

What this calculator does not cover

  • Thermal striping and fatigue. The flow pattern tells you where temperature fluctuations hit the wall; how large they are and what they do to the pipe needs a transient (LES) simulation or a test, and a fatigue assessment.
  • How long a pipe it takes to mix. That depends on the pattern, the Reynolds number and the turbulence; the page gives no correlation for it rather than an unsourced one.
  • Different fluids or heat loss. The balance assumes one fluid, constant specific heat and adiabatic walls. Mixing two different fluids needs each one's cp, and a heated or cooled wall adds its heat flow.
  • Tees with a branch larger than the main pipe, angled branches or other shapes. Kamide's limits come from 90° tees with the branch a third of the main diameter.
Learn CFD

The mixing tee is the first case where a simulation has to close an energy balance you can check by hand, and the first where a turbulence model matters. It is the opening project of Skill-Lync's Fluent course.

Frequently asked questions

What outlet temperature should a mixing tee give?

The mass-weighted mix of the two inlets: T = (ṁ_m T_m + ṁ_b T_b) / (ṁ_m + ṁ_b), for one fluid with constant specific heat and adiabatic walls. In Skill-Lync's Fluent challenge (hot air 36 °C at 3 m/s in a 33.88 mm main pipe, cold air 19 °C in a 16.96 mm branch) that is 30.32 °C (303.47 K) with the branch at 6 m/s and 27.49 °C (300.64 K) at 12 m/s. A converged steady run should land within a few hundredths of a kelvin of it.

What is the momentum ratio of a mixing tee?

In the T-junction literature (Igarashi, Kamide et al. 2004; Kamide et al. 2009, Nuclear Engineering and Design) it is M_R = M_m / M_b, with M_m = D_m D_b ρ_m V_m² for the main pipe and M_b = (π/4) D_b² ρ_b V_b² for the branch. Above 1.35 the branch jet bends over and runs along the near wall (wall jet), between 0.35 and 1.35 it turns in the middle of the pipe (deflecting jet), and below 0.35 it hits the opposite wall (impinging jet). Course briefs often use "momentum ratio" for the plain velocity ratio V_b / V_m instead, which is a different number.

Is momentum ratio 2 the same as a velocity ratio of 2?

No. In the Skill-Lync challenge geometry a branch at twice the main velocity has a branch/main momentum flux ratio of 1.00 and a Kamide momentum ratio of 0.64. A momentum-flux ratio of 2 would need the branch at 8.47 m/s, and a Kamide M_R of 2 at 3.38 m/s. 387 of the 402 reports that state the branch velocity read "momentum ratio 2 and 4" as 6 and 12 m/s.

Should I use the area-weighted or mass-weighted average temperature at the outlet?

Mass-weighted. The mixed temperature is the one the outflowing mass carries, so each face is weighted by its mass flow, not its area. At a well-mixed outlet the two agree; where the outlet is still hot on one side and cold on the other, the faster part carries more of the flow and the area-weighted average is off. Fluent's Surface Integrals panel offers both.

How do I know if a mixing tee mixes well?

Compare the standard deviation of temperature at the outlet with what it would be if nothing had mixed: σ₀ = ΔT √(x(1 − x)), where x is the branch's share of the mass flow. The squared ratio is Danckwerts' intensity of segregation, 1 when fully segregated and 0 when perfectly mixed. A raw σ from two cases with different flow shares is not a fair comparison; σ/σ₀ is.

Why is my mixing tee outlet temperature 0.15 K off?

Kelvin with 273 instead of 273.15. Typing the inlets as 309 K and 292 K puts every temperature in the run 0.15 K low, and converting a correct answer with 273 does the same on paper.

Related on Skill-Lync

Sources

  • M. Igarashi, H. Kamide, M. Tanaka, N. Kimura, "Study on temperature fluctuation characteristics for high cycle thermal fatigue in a mixing tee", Transactions of the JSME Series B 70(700) (2004) 3150-3157, Eq. (2) and the pattern limits MR > 1.35 wall jet, 0.35 to 1.35 deflecting jet, < 0.35 impinging jet; restated in H. Kamide et al., JNC Technical Review 27 (2005) 41, Eqs. (2) and (3), with the test geometry (Dm 150 mm, Db 50 mm) and the note on diameter ratio.
  • H. Kamide, M. Igarashi, S. Kawashima, N. Kimura, K. Hayashi, "Study on mixing behavior in a tee piping and numerical analyses for evaluation of thermal striping", Nuclear Engineering and Design 239 (2009) 58-67: the three patterns by momentum ratio, main-to-branch diameter ratio 3.
  • B. L. Smith, J. H. Mahaffy, K. Angele, "A CFD benchmarking exercise based on flow mixing in a T-junction", NURETH-14 (2011), Table 3: 19 °C, 140 mm, 9.0 L/s main and 36 °C, 100 mm, 6.0 L/s branch (benchmark report NEA/CSNI/R(2011)5).
  • P. V. Danckwerts, "The definition and measurement of some characteristics of mixtures", Applied Scientific Research A 3 (1952) 279-296: intensity of segregation, as restated by Aref et al., Reviews of Modern Physics 89 (2017) 025007, Eq. (48), and Schikarski et al. 2019 (arXiv 1909.11868), where σ²max is the variance of completely segregated streams.
  • Ansys Fluent 2024 R2 Theory Guide, §25.3.1 "Computing Surface Integrals": area-weighted average Eq. (25-12), mass-weighted average Eq. (25-18). User's Guide §8.1 and §8.7.1: default air 1.225 kg/m³, cp 1,006.43 J/(kg·K), μ 1.7894 × 10⁻⁵ kg/(m·s), molecular weight 28.966.
  • Skill-Lync, all 619 public mixing-tee reports, read 20 Nov 2026: branch velocities, outlet temperatures and how they were averaged, standard deviations, inlet areas.
  • Skill-Lync, scripts/verify/mixing-tee.py: re-implements the balance, the three ratios and the pattern limits, checks closed forms, the segregated-variance formula on a cell field, the course, benchmark and slip cases, and writes the numbers this page checks itself against at build time.
Free counselling

From a mixing tee to a CFD career.

Closing an energy balance, picking a turbulence model and proving a mesh is fine enough are the everyday work of a CFD engineer. A Skill-Lync counsellor can map the course that fits the role you want.

Talk to a CFD course counsellor

Free. We call you back, usually the same day.

By submitting you agree to our Terms and Privacy Policy.

Talk to a career counsellor

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.