Thin walls: when the wall sets the price

When is a wall thin?

A thin wall is not a dimension on the drawing. The same thickness is stiff when the wall is low and flexible when it is tall. What counts is height against thickness, the material, and how far the cutter has to stick out to reach the floor. Aluminium deflects further than steel of the same size, steel takes more cutting force, plastic gives under the clamp.

A geometry check in a CAD suite flags a thin wall, but not what that wall costs in your shop. Subduxion research makes the point that geometry is part of manufacturability, not the whole problem: tool access, tool length, workholding and the machines you have decide the rest. Whether you can mill the wall, and at what price, depends on what the shop does about it. That craft belongs in the estimate, line by line and with its reason.

What goes wrong with a thin wall?

Chatter. The wall springs away from each tooth and leaves waves on the surface. Wikipedia's article on machining vibrations notes that with thin workpieces and long tools, vibration is often the limiting factor: down go the cutting data, up goes the cycle time.

Deflection. Wall and cutter both give, less comes off than the program assumes, and after the pass the wall is thicker at the top than drawn.

Residual stress and distortion. Take a lot of material away and the stress in the block or plate comes free. In 2023 Li and colleagues measured, at the same cutting data, a largest deformation in aluminium plate of just over 2 mm or of about 0.3 mm. The only difference was how much material came off the top and how much off the bottom.

The clamp. A three-jaw chuck pushes a thin-walled bush in at three points. You turn it round, release it, and the bush springs back into a three-lobed shape.

And what goes into the wall later. Wikipedia describes inserts for plastics going in with heat or ultrasound: the plastic melts locally and re-forms round the knurl. A thread insert pressed in cold can split a thin plastic wall. Fitting inserts is a separate operation with its own time.

Which sequence keeps the wall stiff?

The operation sequence is the cheapest measure. Step down, alternating sides, so there is material beside the wall to support it on every pass. The same Wikipedia article describes that strategy for thin-walled workpieces: alternating small passes so the walls do not bend.

Leave support material in place as long as you can and bring the wall to size only in the final pass, with a small depth and width of cut. Keep the tool overhang as short as the pocket allows. Where the material carries a lot of stress, split roughing and finishing and release the part in between.

The manufacturing engineer picks the measures, the cost engineer turns them into lines.

How do you hold a thin-walled part?

The workholding has to hold the part without pressing on the wall: soft jaws machined to the outside shape, a vacuum plate, or a plug or filler in the pocket. For turning it is a mandrel inside the bush, a collet or a six-jaw chuck. Wikipedia's article on chucks gives holding thin-walled tubing with minimum deformation as the main purpose of six- and eight-jaw chucks.

Someone has to make and set up each of these. Extra stock to clamp on also sets the size of the blank, see sizing the blank. If the part comes back, keep the jaws as a dedicated fixture.

Worked example: a housing with a 3 mm and a 1 mm wall

A housing in EN AW-6082, 10 parts, machine at €80 an hour, in two versions: a 3 mm wall and a 1 mm wall. Same blank, so same material. The thicknesses are example values, not a limit. At 1 mm, finishing goes from 15 to 45 minutes, the stepped program costs €30 more, the part is released in between, soft jaws are added and you allow one extra part per 10 instead of per 50.

Two millimetres less wall costs €82 a part here, €41 a millimetre. Scrap sits on its own line too: a scrapped part costs material and machine time, not the setup. Whatever you cannot back up with a line goes in as a risk allowance with its cause.

When do you propose a thicker wall to the customer?

Ask first what the wall does: carry load, hold a seal, take a fit? If it does any of those, the thickness is functional. If the wall is thin to save weight nobody needed, a thicker wall is often cheaper and the part is no worse for it.

Propose it with the quote, with two prices and the difference line by line. If the customer takes the thicker wall, that is a new revision of the drawing, not an agreement buried in an email. An ERP keeps the price after the order, not the reason. Record the measures as lines and the conversation is ready when the repeat order lands.

How to assess a thin wall in a request

Run through these questions before you cost a thin wall as standard milling. Every yes is a line in the estimate or a question for the customer.

• How tall is the wall against its thickness?
• Is there a tight tolerance or flatness on the wall itself?
• Can you reach both sides of the wall in the same setup?
• How do you hold the part without clamping on the wall?
• Will inserts or threads go into the wall?
• Does the part need releasing between roughing and finishing?
• Is the wall thickness functional, or could it be thicker without harming the part?