Dry Pressing and Cold Isostatic Pressing

Two Ways to Squeeze a Powder into a Shape

Guide 1 laid out the ceramic problem: you cannot melt-and-cast or forge a ceramic the way you would a metal, because its ionic-covalent bonds would sooner shatter than flow. So you do it backwards — you build the shape first, out of loose powder, and only then fire it to weld the particles together. That first step, giving the powder a shape, is forming, and this rung walks through its methods. The two simplest, and the whole subject of this guide, both work by the same brute idea: squeeze a dry-ish powder until it holds together. They are dry (die) pressing and cold isostatic pressing.

The two share almost all their DNA but differ in one decisive place: where the pressure comes from. In a die you push from a single direction, squeezing the powder between two hard punches. In cold isostatic pressing you wrap the powder in a rubber mould and squeeze it with fluid from every side at once. That one difference — one direction versus all directions — turns out to decide everything downstream: how uniform the packing is, which shapes you can make, and which defects you inherit. Hold onto it; it is the hinge the whole guide turns on.

Dry Pressing: Fast, Cheap, and Everywhere

Uniaxial die pressing is the plainest recipe in ceramics: pour a free-flowing powder into a hardened steel die, squeeze it between an upper and a lower punch, release, and eject the compact. It is also the fastest — a modern automatic press stamps out tens to hundreds of parts a minute. That speed, plus a rigid die that sets the shape to a precise dimension, is why a huge share of the world's ceramics is made this way: floor and wall tiles, the little carbide and alumina inserts in cutting tools, spark-plug bodies, grinding wheels, structural refractory shapes, capacitor and electronic discs. When a part is small, squat, and wanted by the million, you press it.

1. Fill the die cavity with free-flowing granulate, letting it settle evenly to the same packing everywhere.
2. Lower the top punch and press — either single-action (bottom punch fixed) or double-action (both punches move).
3. As pressure rises the granules slide, then deform, then fracture, and the pores collapse until the green density reaches about 55 to 60 percent of theoretical.
4. Release the load; the compact springs back slightly as the elastic strain in the powder recovers.
5. Eject the green body by driving it out with the bottom punch, and repeat — seconds per part.

One subtlety hides in step 1. The powder you pour is not the fine primary powder from the last rung — a one-micron powder is fluffy, clings to itself, and will not flow evenly into a die. Instead it is spray-dried into free-flowing granules, little spheres of maybe 50 to 200 micron, each a cluster of primary particles glued together by a binder and a plasticizer. Those granules pour like fine sand. Then, under the punch, the compact densifies in three acts: first the granules slide and rearrange, then they deform against one another, and finally they fracture so the primary particles repack into a denser bed. Typical pressures of 20 to 150 MPa carry the whole sequence.

The Wall-Friction Problem: Where Density Gradients Come From

The great weakness of die pressing is written into its geometry. When you push the punch down, the powder does not pass the pressure faithfully to the far end — the powder column drags against the rigid die wall, and that die-wall friction steadily bleeds pressure away with depth. So the powder near the moving punch feels the full load and packs dense, while the powder far from the punch feels only a fraction and stays loose. The upshot is a density gradient baked right into the green body: dense at one end, slack at the other, and no way to see it by eye.

The falloff is roughly exponential. The axial pressure at depth z drops as P(z) = P0 times exp(-4 mu K z / D), where mu is the wall-friction coefficient, K the ratio of sideways to axial stress the powder transmits, and D the die diameter. Put numbers on it: take a lumped wall factor 4 mu K near 0.5, and a powder column as tall as it is wide (z/D = 1). The far end then feels only exp(-0.5), about 0.6 of the top pressure — a 40 percent loss across a single part-height, and it gets exponentially worse the taller the column. That single equation is why die pressing is kept to squat shapes and never used to press a long thin rod.

UNIAXIAL DIE PRESSING                COLD ISOSTATIC PRESSING (CIP)
pressure from the punch, downward    fluid pressure from ALL sides

         | P  (top punch)                    fluid at P = 100-400 MPa
         v                              -->  +-----------+  <--
    +===========+  <- rigid steel die        |  #######  |
    | ######### |   DENSE near the punch -->  |  #######  |  <--
    | ######### |      |                      |  #######  |
    | ::::::::: |      | friction bleeds  -->  |  #######  |  <--
    | ......... |      v pressure w/ depth    |  #######  |
    | .  .  .   |   LOOSE, far from punch -->  +-----------+  <--
    +===========+                             fluid at P

  density DROPS down the column        density UNIFORM everywhere
  gradient  ->  warps on firing        no wall  ->  no gradient

Two tricks tame the gradient without leaving the die. First, keep the part squat — a low height-to-diameter ratio, so no powder ever sits far from a punch. Second, double-action pressing: drive both punches inward at once, so pressure enters from top and bottom together, halving the effective column and parking the loose, weak plane in the middle instead of at a face. Why fuss over a gradient you cannot even see? Because it is a time bomb. Sintering shrinks a body 15 to 20 percent linearly as it densifies, and denser regions shrink less than loose ones. So a green body that is 60 percent dense on top and 54 percent on the bottom pulls unevenly in the furnace — differential shrinkage that bows the part into a warp, or tears it with a crack. The gradient you could not see green becomes the warp you cannot miss fired.

Cold Isostatic Pressing: Pressure from Every Side

If wall friction is the disease, taking the wall away is the cure. In cold isostatic pressing, or CIP, you seal the powder inside a flexible rubber or elastomer mould, submerge the sealed mould in a fluid — usually oil or water — and pressurize the fluid, typically to 100 to 400 MPa and sometimes up to about 700. Now the pressure presses inward equally from every direction (that is what isostatic means), so there is no wall to drag against and no privileged axis. The green density comes out higher than a die can reach and, far more importantly, uniform — no gradient, so green-body uniformity is excellent and the part shrinks evenly when it is fired.

Because no rigid die is dictating a squat shape, CIP frees you to make the very things die pressing cannot: long thin rods, tubes and crucibles, big billets and sputtering targets, and parts with a high aspect ratio a die would never fill evenly. It comes in two flavours. In wet-bag CIP the filled mould is loaded into the pressure vessel and lifted out after each cycle — flexible, good for large or one-off parts, but slow. In dry-bag CIP the mould is built permanently into the vessel and only the powder is loaded and unloaded, which is fast and automatable; it is how spark-plug insulators are pressed by the hundreds of millions.

CIP is not a free lunch, and it pays to be honest about the bill. A flexible mould cannot hold a precise dimension the way a steel die can, so CIP parts come out with rough surfaces and loose tolerances, and usually need green machining — turning or milling the soft compact before firing — or diamond grinding after. Wet-bag cycles are slow, and a high-pressure vessel is expensive kit. So the choice is a familiar engineering trade: reach for the die when you want cheap, fast, precise, squat parts by the million; reach for CIP when you need uniform density, a shape a die simply cannot make, or a part big or slender enough that a gradient would ruin it.

The Green Body, and What Pressing Cannot Do

Whichever press you use, you walk away holding the same thing: a green body at roughly 55 to 65 percent of theoretical density, keeping its shape on friction and the little binder you mixed in, with just enough green strength to be handled without crumbling. Before it can be fired, that binder must be gently burned out first — a slow, low-temperature hold — or it will blister and crack the part; but binder burnout and the firing itself belong to later rungs. For now, the green body is simply the baton the forming step hands to the furnace, and everything the furnace can do is already fixed by how good that baton is.

Pressing is unbeatable at what it is good at, but it has hard limits — and those limits are exactly why the rest of this rung exists. It needs a free-flowing granulate, so it cannot lay down a thin flat sheet; that job goes to tape casting (guide 3). It presses from the outside, so it cannot form a complex hollow shape like a teapot or a sink; that goes to slip casting (guide 3). It cannot pull a continuous tube or a honeycomb — that is extrusion — nor mould a small, intricate, precision part — that is injection molding (guide 4). And the newest routes build a shape layer by layer with no die at all, the ceramic wing of additive manufacturing.

Notice, though, what every method in this rung shares. Each one is only a way to arrange powder into a shape and a green density, and each is judged by the same yardstick: how high, and how uniform, is the packing it leaves behind. That is why guide 5 steps back from the individual methods to the green body itself — its density, its uniformity, and the unforgiving rule this guide has already shown in miniature. The density gradient you press in today is the warp, the crack, or the weak spot you fire in tomorrow. Master the green body, and you have mastered forming.