One Die Instead of Two: Forming Sheet Against Rubber

**US 2,190,659**, *Apparatus for forming sheet metal*, **Henry E. Guerin** of Beverly Hills, California, assignor to the **Douglas Aircraft Company** of Santa Monica; granted **20 February 1940**, long expired. The drawing on this page is the patent's own sheet 1. Its front page is worth reading for the dates alone, because they tell you how long this took: the original application was filed **9 September 1935** and issued as **US 2,055,077** on 22 September 1936; a continuation followed in April 1936; and **this** application, a division, was filed **26 January 1939**. Five years of filings for one idea. When a later rung in some batch says an invention 'appeared in 1940', this is what that usually conceals. The patent has **one claim**. The specification says what it is for in the first paragraph: forming sheet material into predetermined shape, *particularly to an improved method and apparatus whereby sheet metal may be economically and expeditiously formed into desired shapes*, and it is explicit that the case it cares about is **small quantities of a great many different parts** — *the airplane building industry is necessarily to employ metal members which are made from sheet metal*, in *a relatively few shapes of the same size and configuration*. That is the economic argument in the patent's own words, and it is the point of the whole rung: *if such ribs are made in accordance with standard practice of making die and punch blocks, there is a considerable investment in such parts since it is well known that a very accurate three-dimensional relationship must be adhered to.* Guerin removes half of that investment.

The form block is the *only* shaped tool in this process, and the whole saving is that it does not have to be hardened steel. Make a simple one: a plate about 80 × 60 mm with a rounded profile on one long edge, plus a cut-out or two. Options, and try more than one if you can: - **MDF or plywood**, cut on the bandsaw and filed. Good for a handful of parts. It will crush at the edges eventually, and you will see exactly where. - **Cast polyurethane resin**, poured into a mould. Tougher than wood, cheap, and it takes fine detail. - **Aluminium plate**, filed and sanded. Effectively permanent at this scale. Whatever you use, the **edge radius matters more than anything else**. A sharp edge on a form block cuts the sheet rather than forming it — the rubber presses the metal hard onto that edge and there is nothing to stop it. Break every edge to at least 1.5 times the sheet thickness and polish it; a scratch on the block prints on every part. Make the block **thicker than the flange is tall**, so the rubber can wrap the sheet all the way down and still be clear of the bed. Finally, fix the blank to the block so it cannot move. Two small dowel pins through matching holes in the blank is the classical answer and it is what the patent's figures show. A blank that shifts under the pad makes a part that is the right shape in the wrong place.

Build the retainer first, because an unconfined pad does nothing. A steel box open on one face, deep enough to hold a stack of rubber sheet 25–40 mm thick, with walls stiff enough not to bulge. Weld it, or bolt it up out of plate. It sits open-face-down on the press ram. Stack natural rubber sheet — about 50–60 Shore A — into the box until it stands a little proud of the rim. Several thinner sheets work as well as one thick block and are easier to buy. Then, on the press bed: form block, blank pinned to it, and the retainer lowered over the top. Press **slowly**, and watch the rubber. It should bulge only where the part is, and the box should not visibly deform. If the box is bulging, it is the box that is being formed and not the metal. Start with 0.8 mm annealed aluminium and a shallow flange. Step 5 says why: a 12 tonne press covers about 77 mm square at the pressure a soft aluminium flange needs, and nothing at all at the pressure a sharp detail needs. **Keep the part small and the detail soft**, and the process works on a bench press; ask for more and you need an aircraft factory. Look at the result. The face against the block is crisp; the face against the rubber is slightly soft and carries no tool marks at all, which is the process's signature and the reason it was liked for visible skins. Measure the flange height round the part — it will vary, and where it varies tells you where the rubber ran out of pressure. Keep clear of the press and never reach under a loaded ram. A confined rubber pad stores real energy, and if the retainer lets go it lets go sideways.

Cut a second form block with one **concave** edge and one **convex** edge of the same radius, and form a flange along both in one hit. They will not behave the same, and the difference is the most useful thing in this rung. **The concave edge** gives a *shrink flange*. Its metal has to fit into a smaller circumference than it started in, so it is in compression, and it **buckles** — a series of small waves along the flange, or one big one. **The convex edge** gives a *stretch flange*. Its metal has to reach round a larger circumference, so it is in tension, and it **cracks** — starting at the free edge, usually exactly where rung 1's sheared edge left its roughest fracture zone. Step 5 computes how much shrink or stretch each radius demands. Compare it against the cupping and *n* figures from rung 6, and you can predict which of your blocks is formable before you press it. Three fixes, all of which the aircraft industry used: 1. **Anneal.** A soft temper takes far more of both. Form soft, then age or heat-treat if the alloy allows it. 2. **Deburr and polish the blank edge.** A stretch flange cracks from the edge, so the quality of rung 1's cut is the dominant variable — this is the most common fix and the cheapest. 3. **Split the job.** Rubber-pad form what you can, then finish the awkward flange by hand over a stake with a mallet, shrinking the waves out. That combination — machine for the bulk, hand for the difficult edge — is exactly how it was done.

**Attribution.** US 2,190,659, *Apparatus for forming sheet metal*, Henry E. Guerin, assignor to Douglas Aircraft Company; original application 9 September 1935 (issued as **US 2,055,077**, 22 September 1936), continuation April 1936, this divisional application 26 January 1939, granted 20 February 1940. Both expired. The drawing on this page is the patent's own. **Why an aircraft company.** A car maker stamps one body panel a million times and can afford matched steel dies. An aircraft maker in the 1930s needed hundreds of *different* ribs, frames, clips and stiffeners, in runs of dozens. The tooling, not the metal, was the cost — and half of it was the die that mirrors the punch. Guerin's process deletes that half and, more importantly, lets the remaining half be made in wood by a pattern maker rather than in tool steel by a die sinker. When a design changed, you made a new block in an afternoon. The process ran right through the war and is still in use; later variants — Verson-Wheelon with a rubber bag and hydraulic pressure behind it, and fluid-cell presses — are the same idea with the pressure raised so that deeper parts become possible. **The principle worth keeping.** Replace one precise tool with an **indifferent medium under pressure**. The rubber does not need to know the shape; it simply has nowhere else to go. The same move turns up in hydroforming, in explosive forming, in vacuum bagging a composite and in an isostatic press, and every time it converts a tooling problem into a pressure problem. **Honest limits, and they are severe at bench scale.** Pressure has to be applied over the whole plan area at once, so press force scales with part area — step 5's table is the reason a small shop can do a 77 mm flange and not a 300 mm one. Depth is limited: this is a flanging and beading process, not a deep-drawing one, because the rubber cannot hold a blank down and draw it in the way a blank holder does. The rubber wears and takes a permanent set, so pads are consumables. And the face against the pad is never as crisp as a matched-die part — which was a feature for aircraft skins and is a defect if you wanted detail on both sides.