ECE 495 — Electromagnetic Warfare and Survivability

Block 3: IR EW and Signature Management
Lesson 26: The Aircraft RF Signature

Juan Jurado, Col, USAF, Ph.D.
Permanent Professor and Head
Electrical and Computer Engineering
Fall 2026

←→ navigate · C draw · B board · S speaker view · F fullscreen · ? all shortcuts

Block 3 · Lesson 26 · Where We Left Off

Where We Left Off

L25 gave you the physics: $\sigma = A_{\text{geo}} \times \Gamma \times D$, plates spike, corners betray, and the median prices the pattern.

Today we walk the airframe. Where do the echoes actually come from — and what did the designer do about each one?

Today

This is L22’s move on the RF side: find the contributors, rank them, and hand each a verdict.

Block 3 · Lesson 26 · Objectives

Lesson Objectives

By the end of this lesson, you will be able to:

Name the major RF signature contributors on an aircraft and the physics that makes each loud
Assign each contributor its design verdict: shape, mask, treat, or manage
Explain planform alignment and why flying wings take it to the limit
Argue the loudest-first budget discipline using $\sigma_{\text{tot}}$ and the fourth-root law

Fix the loudest scatterer — the rest is rounding error.

Block 3 · Lesson 26 · Part One

Walking the Airframe

Five families of echo, and the design verdict each one earns. Organized by what you do about it, not by why it scatters.

Block 3 · Lesson 26 · Contributors

Where the Echoes Hide

Contributor	Why it is loud (L25)	Verdict
Inlets, exhausts, cockpit	cavity multi-bounce	mask
Radome and antennas	built to radiate	treat
Wing/body speculars, junctions	plates and corners	shape
Edges and surface waves	diffraction	shape + treat
Gaps, seams, panels, stores	small corners everywhere	manage

Five families of echo, four verdicts — the rest of the lesson is one frame per verdict.

Block 3 · Lesson 26 · Verdict: Shape

Shape It Away: Send the Energy Elsewhere

Shape

Mask

Treat

Manage

Shaping attacks $D$ — the echo goes somewhere, just not home:

Slope and blend every surface — no flat plate or right angle faces the threat
Planform alignment: force every edge and seam into a few shared orientations
Aligned edges stack their spikes into a handful of narrow butterfly lobes — pointed away from the nose
The flying wing is the endgame: few edges, few lobes, no tail-body corners

You cannot delete the energy — you choose where to send it.

Block 3 · Lesson 26 · Verdict: Mask

Mask It: Hide the Things That Cannot Be Shaped

Shape

Mask

Treat

Manage

Some contributors are cavities by function — you cannot slope an engine face:

Inlets: serpentine ducts deny line-of-sight to the compressor; top-mounted inlets shield from below
Screens and vanes block the cavity where a duct cannot bend enough
The cockpit is a cavity full of corners — a conductive coat keeps the wave out
Exhausts: shielded, flattened nozzles — the same hardware fixes IR (L24) and RF

Block 3 · Lesson 26 · Verdict: Treat

Treat It: the Parts That Must Radiate

Shape

Mask

Treat

Manage

The radar, antennas, and edges cannot be deleted — so we compromise:

An antenna built to radiate is built to echo — flush, conformal apertures limit it
Frequency-selective radomes: transparent in the ship’s own band, a shaped wall to everyone else
Edges keep their diffraction — edge treatments and RAM soak it up (that is L27)
Until then: point the loud face away and radiate on your terms (EMCON, also L27)

Block 3 · Lesson 26 · Verdict: Manage

Manage It: Stealth Is a Lifestyle

Shape

Mask

Treat

Manage

The last dBs are not designed in — they are maintained in:

Every gap and panel edge is a small corner; serrations push returns off-normal
External stores are RCS catastrophes — carry internal or stay home
Surface condition counts: chipped coatings and proud fasteners own the nose sector
This is why LO fleets live with signature checks — the test never ends

Block 3 · Lesson 26 · Design Principles

The Design Commandments

Sixty years of LO design compress into a short list:

No right angles to the threat

corners retroreflect (L25)

Align the edges

few orientations, few lobes, none in the nose sector

Hide the cavities

inlets, cockpits and bays never show their insides

Curve continuously

early designs faceted because 1970s codes only priced flat panels; modern curvature kills the facet spikes too

Sweat the details

seams, fasteners and stores set the floor the shaping bought

Block 3 · Lesson 26 · Part Two

Pricing the Signature

Shaping chose where the energy goes. The budget decides where the money goes — and the fourth root sets the exchange rate.

Block 3 · Lesson 26 · The Budget

The Budget: Loudest First

Contributors add in linear space — and the sum has a tyrant:

The signature budget

$$\sigma_{\text{tot}} \;\approx\; \sum_i \sigma_i \qquad\qquad R_{\max} \;\propto\; \sigma_{\text{tot}}^{1/4}$$

One $+10$ dBsm corner beside a $-20$ dBsm airframe is the signature
So LO design is a budget: every contributor under the line, loudest first
The fourth root prices it: $-20$ dB of median RCS keeps a third of the reach
This budget logic is exactly how Project 3 should argue

Fix the loudest scatterer — the rest is rounding error.

Block 3 · Lesson 26 · Where This Is Headed

Where This Is Headed: Buying the Last dBs

Shaping did the heavy lifting. What it cannot fix, materials and discipline must:

RAM: absorb what shaping could not redirect — edges, seams, inlet lips
Polarization tricks: exploit how edges and wires favor one polarization
EMCON: the loudest thing on a stealth aircraft is its own transmitter
L27 closes the RF story; L28 fuses every signature into one target

Block 3 · Lesson 26 · Quick Exercise

Quick Exercise / Type-along: Price the Budget

Type this with me — three contributors, one wallet.

% L26 type-along: itemize the budget, then price the fix.
sig_dB = [2 -7 -15];             % three contributors [dBsm]
sig    = 10.^(sig_dB/10);        % linear [m^2] -- sums add here
tot    = sum(sig);
fprintf('Total  %.3f m^2 = %.2f dBsm\n', tot, 10*log10(tot));
[~, k] = max(sig);               % the tyrant of the budget
fprintf('Loudest is #%d (%+.0f dBsm) = %.1f%% of the sum\n', ...
        k, sig_dB(k), 100*sig(k)/tot);
cut = 10*log10(2^4);             % R ~ sigma^(1/4)
fprintf('Halving range needs %.1f dB off the TOTAL\n', cut);
new    = sig;
new(k) = new(k)*10^(-cut/10);    % spend it all on the loudest
fprintf('Fixing only #%d: %.2f dBsm, a %.1f dB cut\n', ...
        k, 10*log10(sum(new)), 10*log10(tot/sum(new)));

You spent the whole halving cut on #1 — did the total drop that far? Why not?

What did that buy in range — and where does the next dollar go?

Block 3 · Lesson 26 · Quick Problem

Quick Problem: Nose-Sector Budget

A notional airframe’s nose-sector budget, one line per contributor family, with a serpentine duct on the table.

Engine inlet and duct: $-10$ dBsm
Radome and antennas: $-16$ dBsm
Leading edges: $-20$ dBsm
Gaps, seams, and fasteners: $-23$ dBsm
A serpentine duct and inlet screen would cut the inlet line by 10 dB

(a) The total, in m$^2$ and dBsm.
(b) The loudest line’s share of the total, in percent.
(c) Fit only the duct: the new total in dBsm, and how many dB it fell.

(a) 0.140 m$^2$ = $-8.5$ dBsm · (b) 71 % · (c) $-13.0$ dBsm, a 4.5 dB drop

Two to three minutes, calculators or a MATLAB console; then click. Worked path for (a): convert each line with ten to the dBsm over ten — 0.100, 0.025, 0.010 and 0.005 square metres — and add in linear units, never in dB: 0.140 square metres, -8.5 dBsm. (b) The inlet is 0.100 of 0.140, 71 percent: one line of four owns nearly three-quarters of the signature. (c) The duct takes the inlet from 0.100 to 0.010, so the sum is 0.050 square metres, -13.0 dBsm — the total fell 4.5 dB, not 10. The so-what: the three lines you did not fix sum to 0.040 square metres, -14.0 dBsm, a floor no inlet work can go below, and the radome now owns half the new total. The 4.5 dB buys 0.77 of the range, ten to the minus 4.5 over 40, and the next dollar goes to the radome, not to more duct.

Today’s Interactive Demo: Planform Alignment

Companion script: L26_PlanformAlignment.m

Live demo

Planform Alignment

usafa-ece.github.io/ece-495-ew

Demo hand-off; drive it from here. Start on the 4th-gen fighter: thirteen edges in a dozen orientations, each throwing an L25-style specular lobe at its normal and an edge-diffraction lobe 6 dB down at the mirror. Read the three tiles out loud — 23.1 % of the compass loud, +7.0 dBsm in the nose sector, 24 spike directions. Then the Hopeless Diamond and the flying wing: four spike directions each, 2.0 % and 2.9 % loud, the nose sector down in the fuzz at -20.3 dBsm — but the wing's lobes are the tallest on the board at 21.9 dBsm. Ask them why, and make them say it: the energy is concentrated, not deleted. Then drag the diamond's tail point off the centreline so they watch a flat transom appear, fire straight aft and, by the mirror, straight up the nose. Last, raise the diffuse scatter from -25 to -15 dBsm on the diamond: no edge has moved, the nose sector climbs to about -10 and the verdict goes amber — say it plainly, that is L27's problem now, not the planform's. Close by pointing at L27 — RAM, polarization and EMCON, the dBs shaping left behind. Companion script runs on their own.

Block 3 · Lesson 26 · Wrap-Up

Wrap-Up

Key ideas

Five families — cavities, antennas, speculars, edges, seams — each with a verdict
Shape redirects; mask hides cavities; treat tames radiators; manage holds the floor
Planform alignment stacks edge spikes into few lobes pointed away from the nose
$\sigma_{\text{tot}} \approx \sum \sigma_i$: the loudest scatterer owns the signature — budget and fix loudest-first

Next lesson

L27: RAM, Polarization, and EMCON — materials and discipline buy the last dBs
Pre-flight quiz for L27 due before class