You have seen radomes without knowing it. The nose of an airliner is one, along with the white golf-ball dome on a hill near an airport, the bulge under a military jet, and the pointed tip of a missile.
Every one is a cover over an antenna. The antenna sends out radio waves and listens for what bounces back, which is how an airliner sees weather ahead of it and how a missile finds what it was aimed at, and it needs the cover because at speed and altitude the weather alone would tear it apart.
So the cover has two jobs. Survive whatever the vehicle flies through, and stay close to invisible to the radio crossing it in both directions. Those two pull against each other, though, and nearly everything radome manufacturers do is a negotiation between them.
Put a straw in a glass of water and it looks broken at the surface, because light changes speed as it crosses into the water and bends at the boundary. Radio waves do that too, whenever they cross into a solid wall.
How much a material slows a radio wave is governed by its relative permittivity or dielectric constant. Low number, gentle bend. High number, hard bend, and a signal that comes out of the wall pointed somewhere you didn’t intend (incidence angle also plays a part in this).
The other property that matters is how much of the signal the material swallows and turns into heat, which engineers call loss tangent. A microwave oven works on exactly that effect. In an oven it’s the whole idea, and in a radome it’s signal you paid for and never got back.
You want both numbers low, and for fast missile work the Navy has stated in its own published requests how low: dielectric constant preferably under 5, loss tangent under 0.05.
If your glasses have an anti-reflective coating, you already own the trick a radome depends on.
A radio wave striking a wall reflects a little off the outside surface and a little off the inside one. Get the thickness right and those two reflections come back exactly out of step and cancel each other, so hardly anything bounces and nearly everything passes through. Get it wrong and they add together instead, and the cover starts behaving like a mirror, which is a poor thing to bolt in front of a radar.
The thickness that cancels them is half of one wavelength, measured inside the material. So the radar’s frequency sets the wall thickness, and the strength requirement has to fit around that number instead of the other way round.
An older patent works through a case at 35 gigahertz. The wall comes out around an eighth of an inch thick, and the thickness has to hold to within a thousandth of an inch anywhere on the part. A thousandth of an inch is twenty-five microns, which is roughly a third the width of a human hair.
Picture holding that on a curved, pointed cone built by stacking sheets of woven fiber and saturating them with resin. Not many shops hold it repeatably, and it rarely comes up early in a program, because thickness tolerance sounds like a problem for the factory floor and those get put off. It decides more contracts than anybody admits on a call.
Holding it means nothing gets eyeballed. Machines braid the fiber, so the pattern comes out the same every time instead of depending on who was at the bench that afternoon. The sheets get cut on computer-controlled tables and placed using lasers that project the outline onto the mold, which ends the argument about where a sheet was supposed to land.
Ask someone who does aerodynamics what shape the nose should be and you get something long and sharp, because sharp moves fast without wasting fuel. Ask someone who does the radio side and the honest answer is a dome, close to half a sphere, since on a dome the radar looks straight out through the wall no matter which way it turns.
The sharp shape wins. It has to, since drag decides how far the vehicle can go.
Which leaves the radar looking through the wall at a slant, at a different slant depending on where it’s pointed, seeing a different effective thickness each time. Most of the clever engineering in this field is an attempt to claw back what the pointed shape took away. Some of it works well. None of it gets everything back.
Because the wall bends the incoming signal slightly, the radar decides the target sits a little to one side of where it truly is. It’s the same reason you can’t spear a fish by aiming straight at it.
Good airborne designs hold that aiming error inside about a sixth of a degree. Which sounds like nothing at all until you stretch it out. A sixth of a degree, ten kilometers away, puts the aim point roughly thirty meters off. A hundred feet.
A steady error is survivable, or correctable at least, which isn’t quite the same thing, since the software can be told to shade its aim by a fixed amount. What you can’t fix is an error that shifts as the radar swings around. A stationary target starts to look like it’s moving, and the guidance chases something that was never there.
All of the above assumes the wall stays the same wall.
Both of those material properties change with temperature. A thickness tuned to the radar on a bench in a quiet shop is no longer tuned at 1,200 degrees Celsius, so a part can pass every electrical test on the ground and slide out of tune during the last seconds of flight. Nothing in the test paperwork would show it. It doesn’t look like a broken part. It looks like a miss.
The temperatures are not exaggerated. Past roughly four times the speed of sound the outer surface runs to 1,500° Celsius, hotter than the melting point of most steels. Published requirements for a material to even be a candidate start can start 1,400° Celsius.
It’s doing that, incidentally, while the vehicle turns hard enough to pull a hundred times the force of gravity. Fighter pilots black out somewhere around nine.
Rain reads like a maintenance concern on paper. At these speeds a raindrop lands like a small stone, and airborne sand and dust wear the surface down more gradually.
What you end up with is a nose that started out smooth and finished rough. Rough skin changes the way air carries heat into the wall, so the structure runs hotter than anyone calculated, and it degrades the radio performance of the outer surface at the same time. Neither of those turns up as a crack when somebody inspects the part.
Ceramics are the traditional answer and mechanically they earn it. Silicon nitride stays strong when hot and handles rain and sudden temperature swings better than most of the alternatives. The trouble is on the radio side, where it bends signals harder than you’d like and swallows more of them as it heats, which is when you can least afford it.
Go the other direction and there’s fused silica, which is essentially very pure glass. It’s the clearest to radio of the common ceramics and it takes steep temperature changes without cracking. Its problem is that you can break it, and breaking isn’t an option on a part pulling a hundred g.
Most of the options sit somewhere between those two. The patent literature has been arguing about which compromise is the right one since the early 1980s.
Composites tend to get written off here, mostly because the word makes people picture resin, and resin makes people picture something that goes soft in an oven. Mentis Sciences has spent seventeen years developing its own process for composite radomes and nose cones, and the material keeps both its strength and its radio performance up near 2,000 degrees Fahrenheit, which is about 1,090 Celsius. That puts a composite in the same conversation as ceramic instead of a tier below it.
Most of what’s worth knowing is about consistency rather than capability. Anybody can build one good part, and the first one is usually the best one a shop will ever make.
So ask how they hold the wall thickness and what they measure it with, and whether the fiber gets placed by hand or positioned by machine. Mentis braids and hand-lays against two computer-controlled cutting tables, with braiding machines and laser alignment systems, which is equipment bought so the tenth part comes out like the first one did.
Ask what has been tested, and how big the test piece was. Small samples tell you about a material. A full-size part tells you whether the process holds together, and Mentis has run its radome material through both, on nose cones for domestic and international missile programs.
Then ask who is answerable for the radio performance once the outer shape is locked. If nobody in the room has an answer, that’s worth settling before any tooling gets cut.
Mentis Sciences builds composite radomes and nose cones at 40 Depot Street in Warner, New Hampshire, alongside its hypersonic materials and composite testing work. The radome fabrication page goes further into the equipment and the process, or (603) 624-9197 reaches somebody who will talk a design through with you. www.mentissciences.com
