Flatness is an often misrepresented property of our own intuition. Many of the objects we consider flat, pale in comparison to surfaces manufactured to actually be flat. It’s also a property that our industrialized world relies on to function.
While most of us experience flatness as part of aesthetics and ergonomics, flatness in manufacturing is a critical property of positioning, mating and sealing parts together. The high pressures produced by combustion are contained by two mating flat surfaces aided by a gasket.
Let’s look at a sheet of float glass. The floating process self levels the glass, giving it a relatively flat, uniform thickness.
Let’s say a manufacturer’s specification calls for a 3mm thick sheet of glass. For a sheet to pass a quality check, its thickness is sampled at various points along its length and as long as it is 3mm thick, plus or minus a specified tolerance, the sheet passes.
But what if during the process of moving the floating ribbon of molten glass a subtle disturbance is introduced to the molten metal. Let’s say this disturbance imparts a 0.25 mm wave-like undulation throughout the entire ribbon. Now to eye the cut sheets would appear flat and they would pass the quality check for thickness, but the surface of that sheets of glass is far from flat.
Flatness isn’t derived from how closely a part matches its specified dimension. It is a property completely independent of the part’s gross shape.
If we take a surface and sandwich it between two imaginary parallel planes. The gap between the planes that encompasses the surface is known as a tolerance zone. The smaller this distance the flatter the specification.
On parts that do explicitly define flatness the method of both measuring and producing flatness is determined by how tight of a tolerance zone is required.
Flatness specifications down to around 10 microns or about 4/10,000th of an inch are quite common in machinery.
Those mating and sealing surfaces found in car engines can be found at this level of flatness. Sealing in fluids at this level of flatness requires the use of a gasket.
Field testing flatness at this level is done with a known precise flat edge and a clearance probing tool known as feeler gauges.
Actually measuring the flatness of a surface is a lot more complicated. An obvious solution would be to measure the surface against a flat reference. For example, if a part has a surface parallel to the surface to be measured it could be placed on a surface plate. A surface plate is a flat plate used as the main horizontal reference plane for precision inspection. A height gauge could then be used to probe the top of the surface for flatness relative to the surface plate.
If we first place the part to be measured upon 3 columns with adjustable heights. Then, with a height gauge, run the probe across the surface while looking at the amplitude of the needle, we get a snapshot of the difference between the highest and lowest point on that surface, as it’s fixtured.
Automating the process with the use of a coordinate measuring machine or CMM is a common practice. CMMs are typically computer-controlled and can be programmed to perform the tedious repetitive measurements needed to determine surface flatness.
Going beyond the 10-micron levels of flatness requires the use of surface grinding. This process typically used to produce precision parts, precision fixtures, measurement equipment, and tooling.
Lapping is the process of rubbing two surfaces together with an abrasive between them in order to remove material in a highly controlled manner. In lapping a softer material known as a lap is “charged” with an abrasive. The lap is then used to cut a harder material. The abrasive embeds within the softer material which holds it and permits it to score across and cut the harder working material.
Wringing is the process of sliding two ultra-flat faces together so that their faces lightly bond. When wrung, the faces will adhere tightly to each other, requiring a significant force to separate them.
This technique is used in an optics manufacturing process known as optical contact bonding.
When an optical flat’s polished surface is placed in contact with a surface to be tested, dark and light bands are formed when viewed with monochromatic light. These bands are known as interference fringes and their shape gives a visual representation of the flatness of the surface being tested.
Editor’s Note: New Mind ~ “To everyone wondering why the video cuts off abruptly, I lost the last 2 seconds of my audio file. It’s supposed to end with: At these scales, the definition of flatness quickly becomes moot.”