Waveguide Engineering

Cover Flange

Q: What is the difference between a cover flange and a choke flange?

A cover flange has a flat, plain machined face and seals the joint by direct metal-to-metal contact across the entire mating surface, so it relies on tight bolt torque and good surface flatness to keep leakage and reflection low. A choke flange has a circular groove (a half-wavelength radial choke) machined into its face that creates a virtual short circuit at the waveguide aperture, allowing a small physical gap without RF leakage. In practice the two are used as a mating pair: a cover (flat) flange bolts against a choke flange. The choke side tolerates misalignment and surface imperfections, while the cover side provides the simple, repeatable reference surface. Two cover flanges can also be mated directly when alignment and torque are well controlled, but a flat-to-flat joint is more sensitive to gap and contamination than a choke-to-flat joint.

Q: How flat does a cover flange mating face need to be?

For low-loss RF performance the contacting face is typically lapped flat to within 0.01 to 0.03 mm (10 to 30 micrometers) across the surface, with a surface finish of 0.4 to 0.8 micrometers Ra. The reason is that a flat-to-flat cover joint conducts surface currents across the seam, and any gap behaves like a series inductance plus a radiating slot. At Ka-band (26.5 to 40 GHz) the free-space wavelength is 7.5 to 11 mm, so even a 0.05 mm gap is a measurable fraction of the current path and can raise the joint VSWR above 1.05 to 1.10. Tighter flatness, higher bolt count, and gold or silver plating on the mating face all reduce contact resistance and improve repeatability. When flatness cannot be guaranteed, designers switch to a choke flange or add a conductive gasket.

Q: What bolt pattern and dimensions does a WR-90 cover flange use?

A WR-90 cover flange (X-band, 8.2 to 12.4 GHz) commonly follows the UG-39/U pattern: a square flange roughly 41.4 mm on a side with four mounting holes on the corners and two alignment pin holes. The waveguide aperture is the standard 22.86 mm by 10.16 mm WR-90 broadwall and narrowwall. Bolt holes are typically tapped or clearance for M4 or 8-32 hardware. Different waveguide bands use scaled patterns: WR-28 (Ka-band) uses a smaller UG-599/U style flange about 19.1 mm square, while WR-284 (S-band) uses a much larger CPR or UG pattern with six to eight bolts. Always match the flange standard on both halves of the joint, because cover and choke flanges of the same band share an identical bolt circle and aperture so that either face type can mate.

/KUV-er flanj/

Distinguished by its flat, plain machined face, this is the simplest style of waveguide flange: the contacting surface seals directly against a mating flange by metal-to-metal contact, carrying surface currents and RF power across a bolted joint with no resonant groove of its own. A cover flange is most often bolted to a choke flange so the choke half tolerates a small gap while the flat cover half supplies a clean, repeatable reference surface. To keep joint VSWR below roughly 1.05 to 1.10 the mating face is lapped flat to about 0.01 to 0.03 mm and frequently gold- or silver-plated. Cover flanges follow the same UG and CPR bolt-pattern standards as their choke counterparts so either face type can mate within a given waveguide band.

Category: Waveguide Engineering

Face Flatness: 0.01 to 0.03 mm

Typical Pair: Choke flange

How a Cover Flange Seals a Waveguide Joint

A waveguide carries energy as surface currents flowing along its inner walls. At any joint between two waveguide sections those currents must cross the seam without interruption, because a break in the conducting path forces current to detour around the gap, radiating energy and creating a reactive discontinuity. A cover flange handles this by presenting a wide, flat, precisely machined collar around the waveguide aperture. When two such faces are bolted together, the entire ring of contacting metal short-circuits the seam, so the current crosses the joint almost as if the wall were continuous.

The flat-to-flat contact that makes the cover flange simple is also its main limitation. Real surfaces are never perfectly planar, and the moment a measurable gap opens between faces the joint starts to behave like a short transmission-line stub plus a small radiating slot. This is why a cover flange is so commonly paired with a choke flange: the radial choke groove on the mating half places a high impedance at the physical seam, so the joint stays low-loss even with a few hundredths of a millimeter of gap. The cover face then only has to be flat and clean, not perfectly gap-free.

For lower-frequency, lower-power, or test-bench connections, two cover flanges are mated directly. Success there depends on flatness, surface finish, bolt count, and torque. As frequency climbs into Ka-band and beyond, the tolerance budget tightens sharply because the wavelength shrinks, and designers move toward choke pairing, conductive gaskets, or precision-lapped instrument-grade faces.

Contact Resistance and Plating

Joint loss at a flat cover interface is dominated by constriction resistance at the discrete contact spots where the two surfaces actually touch. Increasing bolt count and torque grows the real contact area, while plating the face with gold (typically 0.5 to 2.5 micrometers over a nickel barrier) or silver lowers contact resistivity and resists the oxide growth that degrades a bare aluminum or brass face over time. A clean, well-plated, well-torqued WR-90 cover joint can hold insertion loss under about 0.03 dB and VSWR under 1.05 across X-band.

Flatness, Gap, and Reflection

Free-space wavelength at the joint:
λ₀ = c / f  (e.g. f = 33 GHz → λ₀ ≈ 9.1 mm)

Gap as a fraction of wavelength:
δ = g / λ₀  (g = mechanical gap between faces)

Approximate reflection from a thin flat-flange gap:
|Γ| ≈ π × (g / λ₀)  →  VSWR = (1 + |Γ|) / (1 − |Γ|)

Where c = 3 × 10⁸ m/s, g = face-to-face gap, λ₀ = free-space wavelength. Example: g = 0.05 mm at 33 GHz gives δ ≈ 0.0055, |Γ| ≈ 0.017, VSWR ≈ 1.035. The same 0.05 mm gap at 90 GHz (λ₀ ≈ 3.3 mm) gives VSWR ≈ 1.10, showing why flatness matters more at higher bands.

Cover Flange Standards by Waveguide Band

Waveguide	Band / Freq	Aperture (mm)	Cover Flange Std	Typical Pair	Notes
WR-284	S, 2.6–3.95 GHz	72.14 × 34.04	CPR-284F / UG-584	Choke or gasket	High power, many bolts
WR-90	X, 8.2–12.4 GHz	22.86 × 10.16	UG-39/U (flat)	UG-40/U choke	Most common lab flange
WR-42	K, 18–26.5 GHz	10.67 × 4.32	UG-595/U	UG-597/U choke	Tight flatness budget
WR-28	Ka, 26.5–40 GHz	7.11 × 3.56	UG-599/U	Choke or precision flat	Lapped faces preferred
WR-10	W, 75–110 GHz	2.54 × 1.27	UG-387/U-M	Precision flat + pins	Alignment pins essential

Common Questions

Frequently Asked Questions

What is the difference between a cover flange and a choke flange?

A cover flange has a flat, plain face and seals by direct metal-to-metal contact across the whole mating surface, so it depends on tight torque and good flatness. A choke flange has a half-wavelength radial groove that creates a virtual short at the aperture, tolerating a small gap without leakage. They are normally used as a pair: a flat cover flange bolted to a choke flange. The choke side absorbs misalignment; the cover side gives a clean, repeatable reference surface.

How flat does a cover flange mating face need to be?

Typically lapped to within 0.01 to 0.03 mm (10 to 30 micrometers) with a 0.4 to 0.8 micrometer Ra finish. A flat-to-flat joint conducts surface currents across the seam, and any gap acts like a series inductance plus a radiating slot. At Ka-band a 0.05 mm gap can push joint VSWR above 1.05 to 1.10. Higher bolt count, tighter flatness, and gold or silver plating all improve repeatability; when flatness is uncertain, switch to a choke flange or add a conductive gasket.

What bolt pattern and dimensions does a WR-90 cover flange use?

A WR-90 cover flange (X-band) usually follows the UG-39/U pattern: about 41.4 mm square with four corner bolt holes plus two alignment pin holes, around the standard 22.86 by 10.16 mm aperture, with M4 or 8-32 hardware. Other bands scale this: WR-28 uses the smaller UG-599/U at roughly 19.1 mm square, and WR-284 uses a larger CPR or UG pattern with six to eight bolts. Cover and choke flanges of the same band share an identical bolt circle and aperture so either face can mate.

Related Terms

← Covariance Matrix Coverage Enhancement →