Duct Connection Design — Damper Apertures, Velocities, and the Lost 50 Pa

Most AHU pressure-drop conversations stop at the fan curve and the coil dP. The connection between the casing and the ductwork rarely gets the same scrutiny — and it is exactly where quotes quietly pick up 30-50 Pa that nobody accounted for.

The part of the system nobody sizes twice

A duct connection aperture is not just a hole in the casing panel. It has its own free area, its own velocity, and — if a damper sits in or near it — its own resistance curve that depends on blade angle, not just nominal size. Selection software that treats the connection as "same as duct size, no further check" is skipping a step manufacturers' own catalogues do not skip.

The three variables that actually matter at a duct connection:

• Aperture free area — often smaller than the nominal duct cross-section once the flange, corner brackets and any damper frame are subtracted.
• Local velocity — q_v divided by that reduced free area, not the nominal duct area. The same logic as face velocity (the one number that sizes your casing), applied at the boundary instead of inside the unit.
• Damper aperture position — a volume control damper rarely sits fully open in the commissioned state. A blade at 60° open instead of 90° can add a resistance multiple of 2-3x over the fully-open figure for the same nominal size.

Where the 50 Pa actually comes from

Take a supply connection sized for 2.5 m³/s through a 600×600 mm aperture. At full nominal area that is roughly 6.9 m/s face velocity — already above the ~5-6 m/s range most manufacturers use before connection-loss coefficients start climbing steeply. Push the same airflow through a damper commissioned at 70% open rather than fully open, and the local resistance coefficient can move from around ζ = 0.2 to ζ = 0.6-0.8 for that fitting alone.

At 6.9 m/s (dynamic pressure ≈ 29 Pa), that swing in ζ alone is the difference between roughly 6 Pa and 20+ Pa lost at one connection — before adding elbow, transition and screen losses on the same duct run. Multiply by four connections (supply, extract, and often a mixing or bypass branch) sized the same way, and 40-50 Pa of unbudgeted static pressure is a realistic, not worst-case, outcome. That's a fan operating point the original SFP calculation never saw.

Connection state	Approx. ζ	dP at 6.9 m/s (≈29 Pa dynamic)
Damper fully open, generous aperture	0.15-0.25	4-7 Pa
Damper fully open, tight aperture	0.3-0.4	9-12 Pa
Damper at ~70% open (as commissioned)	0.6-0.8	17-23 Pa

Numbers are illustrative order-of-magnitude figures for a single connection fitting, not a substitute for manufacturer loss-coefficient data — but the direction and scale of the effect are consistent across vendors.

Why deck-scoped keys make this worse when translated wrong

In stacked or cross-flow AHU layouts, connection keys (left_supply, right_extract, and so on) are scoped to the physical deck, not to airflow direction. A cross-flow plate heat exchanger swaps the chain across the bridge diagonal, so a naive "left = supply" assumption silently maps the wrong aperture to the wrong airflow — meaning the velocity and dP check above gets run against the wrong duct entirely. Any tool that lets a user or an export script eyeball this mapping instead of deriving it from the actual airflow chain is a source of exactly the kind of error that only surfaces on site, when the measured static pressure doesn't match the quote.

What to check before the quote goes out

1. Compute connection velocity from actual free area (aperture minus frame/damper obstruction), not nominal duct size.
2. Model the damper at its expected commissioned position, not fully open, if the position is already known from the control sequence.
3. Re-derive every connection's airflow direction from the unit's actual internal routing — never assume deck side equals airflow side once a cross-flow component is in the chain.
4. Sum connection losses into the same static pressure budget as the coil and filter dP — a connection loss is not a rounding error once four of them stack up.

Where the tooling matters

AirSelect3D resolves duct aperture-to-airflow mapping directly from the unit's internal chain — including cross-flow bridge swaps — so a left_supply key always means the supply airflow at that deck, regardless of how the components are arranged internally. Connection velocities and aperture free areas come from the same geometry model that drives the 3D viewer and every export, so the number in the pressure budget is the number on the drawing.

See connection velocities and dP recalculate live as you route the ductwork →