ECE 495 — Electromagnetic Warfare and Survivability
Block 3: IR EW and Signature Management Lesson 27: RAM, Polarization, and EMCON
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
Take attendance; laptops out for the type-along later.
Block 3 · Lesson 27 · Where We Left Off
Where We Left Off
L26 handed out the verdicts: shape what you can, mask the cavities, manage the seams — and left the edges and antennas to be treated .
Today we pay those debts. How do materials absorb what shaping could not redirect — and what is left to hide after that?
Today
Absorbers on the edges shaping could not fix, the tells
polarization leaves behind, and the one emission no coating
can touch.
The loudest thing on a stealth aircraft is its own transmitter.
One minute of recall. Hold the bumper until someone guesses.
Block 3 · Lesson 27 · Objectives
Lesson Objectives
By the end of this lesson, you will be able to:
Explain how RAM attacks the reflectivity knob $\Gamma$ and where it goes on an airframe
Use quarter-wave cancellation to size a resonant absorber and its notch
Describe how polarization betrays edges and wires, and the design response
Argue EMCON from the intercept asymmetry: one-way $1/R^2$ beats two-way $\sigma/R^4$
Read them out; return here at wrap-up.
Block 3 · Lesson 27 · Part One
Absorbing What Shaping Could Not Redirect
Shaping bought the orders of magnitude. Materials buy the stubborn last decibels — and they charge for every one of them.
Shift from geometry to materials.
Block 3 · Lesson 27 · RAM
RAM: Buying Down $\Gamma$
Shaping chose where energy goes ; RAM decides how much survives the trip .
In L25's rule, RAM attacks $\Gamma$ — convert the wave to heat instead of echo
It goes where shaping ran out: edges , inlet lips, seams, and cavity throats
Absorbers trade bandwidth, thickness, and weight — you may pick two
RAM finishes the budget: shaping bought the orders of magnitude, RAM buys the stubborn last dBs
RAM is a finishing tool , never a substitute for good planform.
Cold-call: which knob did L26 spend? Which one is left?
Block 3 · Lesson 27 · Light Rule
The Resonant Trick: Quarter-Wave Cancellation
Put a resistive sheet a quarter wavelength above the metal skin:
Quarter-wave cancellation
$$d \;=\; \frac{\lambda_0}{4} \;=\; \frac{c}{4 f_0}$$
Front and rear reflections return $180^\circ$ apart — and cancel
Off frequency, the phase geometry breaks and the echo comes back
The catch is in the formula: $d$ is fixed hardware, so the null lives at one $\lambda_0$
At 10 GHz, $d$ = 7.5 mm; at 1 GHz, 75 mm — low-band RAM gets thick fast
A resonant absorber is a promise kept at exactly one frequency.
Let them do the 1 GHz arithmetic out loud before advancing.
Block 3 · Lesson 27 · The Notch
One Screen, One Promise
A 7.5 mm spacer nulls at 10 GHz and holds $-10$ dB from 6.3 to 13.7 GHz — a notch sitting squarely on $X$-band.
−10 dB (90% absorbed)
2
4
6
8
10
12
14
16
18 GHz
null 10 GHz · protected 6.3 – 13.7
X-band
One layer, 47% of 2–18 GHz protected. Double the spacer and what happens?
Take a show of hands on the question before advancing — most rooms say "twice the coverage".
Block 3 · Lesson 27 · The Notch
Double the Depth, Move the Promise
15 mm halves $f_0$ to 5 GHz , and the third harmonic at 15 GHz slides in from above. Two windows, each half as wide — and $X$-band handed straight back.
−10 dB (90% absorbed)
2
4
6
8
10
12
14
16
18 GHz
null 5 GHz · 3.1 – 6.9
null 15 GHz · 13.1 – 16.9
X-band
Two windows now, where there was one. Read the two sub-bands off the plot — is that more protection than the 7.5 mm screen bought, or the same amount moved somewhere else?
Twice the thickness. And X-band is bare.
Let the band walk before you say anything.
Block 3 · Lesson 27 · Bandwidth
The Broadband Problem
The threat does not transmit at your design frequency as a courtesy:
An IADS spans VHF to $K_u$ — a single notch defends one band and betrays the rest
Layered (Jaumann) absorbers stack graded sheets: wider notch, more thickness
Magnetic RAM (iron-loaded coatings) absorbs broadband and thin — but heavy
Structural RAM builds the absorber into the skins and edge wedges themselves
Every choice re-runs L26's budget: dBs per pound, per inch, per maintenance hour
At VHF the quarter-wave depth is meters . That is not a materials problem any longer — it is a regime problem, and materials cannot fix it.
Cold-call on the VHF case; it is the one that breaks the pattern.
Block 3 · Lesson 27 · Polarization
Polarization: the Tell in the Echo
The echo carries orientation information — and edges talk too much.
Long conductors
Scatter strongest in the polarization parallel to them
Corners
Flip circular polarization — radars use that to reject rain
Cross-pol channel
Hunts the LO features the co-pol channel misses
Design response: no long straight conductors; orient and serrate the discontinuities — planform alignment's quiet cousin .
Ask where a real airframe hides a long straight conductor.
Block 3 · Lesson 27 · Part Two
The Signature You Choose to Radiate
Every trick so far shrinks an echo somebody else caused. The last one is entirely yours.
Second beat of the lesson.
Block 3 · Lesson 27 · EMCON
The Signature You Choose to Radiate
After shaping, masking, and RAM, one megawatt-class contributor remains.
The fire-control radar, the datalinks, the altimeter — every emission is a beacon with your name on it
B2's ES lesson returns: their receivers are patient, passive, and everywhere
No coating absorbs a wave you chose to transmit
So the last LO discipline is emissions control : radiate seldom, radiate smart
Everything before this slide was about an echo. This one is about a transmission .
Name the three emitters on a real jet and let them add more.
Block 3 · Lesson 27 · Light Rule
The Asymmetry: Why the Listener Wins
The physics is rigged in the receiver's favor:
Intercept asymmetry
$$\underbrace{P_\text{RWR} \propto \frac{1}{R^{2}}}_{\text{they hear you}} \qquad\qquad \underbrace{P_\text{echo} \propto \frac{\sigma}{R^{4}}}_{\text{you see them}}$$
Your signal reaches the enemy RWR one way ; your echo pays the round trip and their small $\sigma$
Result: they detect your radar far beyond where your radar detects them
LPI fights back — manage power, hop frequencies, spread the waveform thin
Or refuse the duel: stay passive, let the IRST (L23) and offboard cues do the finding
Whoever transmits first, telegraphs first.
Cold-call set, if there is time — for each, name the L27 tool and its cost:
a seam glowing only at the seeker band; a VHF early-warning radar seeing what
X-band cannot; a cross-pol channel flagging a straight wing-root fairing;
mission planners wanting the fire-control radar searching all the way in.
The last two are the only active exercise this lesson has for polarization
and for EMCON, so work at least those.
Block 3 · Lesson 27 · Looking Ahead
Where This Is Headed: Every Other Signature
The RF story is told. But the aircraft is still seen, heard, and felt .
Visual : contrails, glint, and paint schemes — the oldest sensor is the human eye
Acoustic : engines and airflow announce low-altitude penetrators
EO/IR : L22's five components never left — they fuse with everything above
L28 closes the loop: multi-spectral fusion , and the rule that you are only as quiet as your loudest band
Thirty seconds. Do not start L28 here.
Block 3 · Lesson 27 · Quick Exercise
Quick Exercise / Type-along: Move the Notch
Type this with me — the 15 mm screen, twice the $X$-band depth.
c = 3e8; Z0 = 377; Rs = 377; % free space, matched sheet
d = 15e-3; % spacer depth [m]
f = linspace(2e9, 18e9, 3201); % the band the threat picks
f0 = c/(4*d); % quarter-wave null
nulls = f0*(1:2:7); nulls = nulls(nulls <= 18e9);
Zs = 1j*Z0*tan(2*pi*f/c*d); % shorted air spacer
Zin = (Rs*Zs)./(Rs + Zs); % resistive sheet in parallel
G = 20*log10(abs((Zin - Z0)./(Zin + Z0)));
b = G <= -10; % the 90%-absorbed window
lo = f(diff([0 b]) == 1)/1e9;
hi = f(diff([b 0]) == -1)/1e9;
fprintf('d=%g mm: nulls at %sGHz\n', d*1e3, sprintf('%g ',nulls/1e9));
fprintf('protected: %sGHz\n', sprintf('%.1f-%.1f ', [lo; hi]));
fprintf('at 8 GHz: d = c/(4f) = %.3f mm\n', c/(4*8e9)*1e3);
Twice the depth. Why did the protected fraction not go up?
HW3 sizes one for 6 GHz. Thicker or thinner than the 8 GHz answer — and why?
Type it line by line with the class, then take answers on both beats.
Block 3 · Lesson 27 · Quick Problem
Quick Problem: Size the Screen
A Salisbury screen tuned to a 4 GHz acquisition radar, checked against the threat band and against a VHF early-warning radar.
Matched sheet, $R_s = 377\ \Omega/\text{sq}$, over an air spacer on the metal skin
Design frequency $f_0 = 4$ GHz
Threat band: 2–18 GHz
VHF early-warning radar: $150$ MHz
$c = 3\times10^8$ m/s
(a) The spacer depth $d$, in mm.
(b) The screen’s nulls inside the 2–18 GHz threat band.
(c) The depth a 150 MHz screen would need, and how many times deeper than (a) that is.
(a) $d = 18.75$ mm · (b) 4 and 12 GHz · (c) 500 mm, 26.7 times deeper
Two to three minutes, calculators or a MATLAB console; then click.
(a) d is c over 4 f0: 3e8 over 1.6e10 is 0.01875 m, 18.75 mm. (b) The
nulls sit at the odd harmonics, 4, 12 and 20 GHz, so inside 2 to
18 GHz only 4 and 12; 20 is past the top. The even multiples between
them, 8 and 16 GHz, are the worst case — the spacer is a half
and a full wavelength deep there, and the screen reflects in full.
(c) 3e8 over 6e8 is 0.5 m, 500 mm: 26.7 times deeper, which is simply
the frequency ratio, 4 GHz over 150 MHz. The so-what: depth goes as
one over f0, so a notch that costs under 2 cm at 4 GHz costs half a
metre at VHF, and even at 4 GHz it protects only two narrow windows
of a 16 GHz-wide threat band — RAM is a narrowband promise.
Today's Interactive Demo: the Quarter-Wave Notch
Companion script: L27_QuarterWaveRAM.m
Live demo
Quarter-Wave RAM
usafa-ece.github.io/ece-495-ew
Demo hand-off; drive it from here. Two tabs. Tab one is the Salisbury
screen: spacer depth and sheet resistance on sliders, three presets
along the bottom, and three readouts — the fundamental null, the
protected sub-bands, and the share of 2–18 GHz under the line. Open on
the X-band preset, then hit Double it and have them read the coverage
tile aloud before and after. Thick is worth one click too. Then
mistune the sheet — drag Rs to
either end and the notch stops reaching the line at all, which is the
half of the design the formula does not show. Tab two is the L11 view
returning: one-way against two-way, and the detect-before-detected band
between the two thresholds. Companion script runs on their own.
Block 3 · Lesson 27 · Wrap-Up
Wrap-Up
Key ideas
RAM attacks $\Gamma$: quarter-wave notches, layered and magnetic for bandwidth, structural for edges
$d = \lambda_0/4$ is a one-frequency promise — broadband absorption costs thickness and weight
Polarization betrays straight edges and corners; discipline the geometry
EMCON : one-way $1/R^2$ beats two-way $\sigma/R^4$ — radiate seldom and smart, or go passive
Next lesson
L28: Multi-Spectral Signatures — the loudest band sets the detection range
Work the L28 pre-flight before class — you could be the one up front
Materials and discipline buy the last dBs that shaping left on the table.
Close the loop on the objectives slide.