Tee and bullhead loss coefficients

Match the flow paths before comparing coefficients.

Standard tee: branch and straight passage

A standard tee has one common passage, a side branch, and a straight continuation. Converging flow merges into the common passage; diverging flow splits from it. The engine selects different relations for the branch and straight paths, using flow and area ratios. A single fixed tee coefficient cannot describe all of those conditions.

For the round converging partition case Fs + Fst = Fc, the implemented Idelchik 7.9 branch relation is:

ζc,s = 1 + (ws/wc)² − 2 × (Fc/Fs) × (1 − Qs/Qc) × Ks

This coefficient uses common-leg velocity pressure. The velocity ratio is ws/wc = (Qs/Qc)/(Fs/Fc). To express the same loss using branch velocity pressure, divide ζc,s by (ws/wc)² for nonzero branch velocity.

Implemented Ks interpolation points for the 7.9 partition case
Fs/Fc	0.06	0.10	0.20	0.33	0.50
Ks	0	0	0.10	0.20	0.25

The library also implements Idelchik 7.10 for a converging side tap with Fst = Fc, and 7.18–7.20 relations for diverging paths. These cases have different geometric limits. Sparse tables and boundary clamping can produce an estimate outside a fully supported interpolation cell; a displayed number alone does not establish applicability.

Bullhead: two opposed legs

A bullhead has two opposed active legs and a common passage perpendicular to both. It has no straight-through continuation of the common passage. The implemented Idelchik 7-29 model assumes equal leg areas. A large side branch on a standard tee does not make it a bullhead.

For the welded/sheet-metal diverging case, the engine uses ζc,leg = 1 + 0.3 × (wleg/wc)². The converging case uses a separate flow/area table, with implemented leg-to-common area ratios from 0.25 to 1.0.

Applicability: 7-29 is not a proxy for a standard tee. Unequal legs, a continuing main, or geometry outside the source assumptions need another justified model or suitable test data. There is no general guarantee that this substitution overpredicts loss.

Capped cross

The capped-cross model uses the source's capped-port condition, not a standard tee with a renamed branch. The implementation cites ASHRAE 5-35 for converging flow and its direction to ASHRAE 5-23 / Idelchik 7-21 for diverging flow. Check direction, active ports, area assumptions, and warnings before using it.

Using the result

Check each path's coefficient and its reference velocity pressure. Negative junction coefficients can represent energy exchange between streams; do not replace them automatically with zero. For unsupported geometry, use documented manufacturer/test data or an independently justified calculation rather than assuming a nearby source diagram is conservative.

See the applicability and source table and verification scope.