1. How does melted tin make glass flat?
The method is called the float glass process, or the Pilkington process, after the British company that developed it in the 1950s [1]. A furnace melts sand, soda ash, limestone and other ingredients, along with scrap glass, at about 1,500 °C. The glass is then held steady at about 1,200 °C [1]. From there it flows out as a continuous ribbon onto a bath of molten tin [2].
Tin is a metal that is denser than glass, so the glass floats on it instead of sinking in [3]. Think of a still bathtub. If nobody splashes, the water surface is level, and anything that lies on it rests against that level surface. Molten tin works the same way, only far hotter. The surface of the tin is flat, so the glass becomes flat and parallel on both sides, with an even thickness and no grinding or polishing needed [2].
That is the short answer to the title. The rest of the article is about the details: what the tin bath is like, what sets the thickness, and how the method came to be.
2. What is it like inside the tin bath?
The bath is not a bathtub at all. One description gives it as roughly 3 to 4 meters (10 to 13 feet) wide and about 50 meters (160 feet) long [1]. The glass enters at about 1,100 °C. As it travels along, the temperature is lowered step by step. At about 600 °C the sheet is stiff enough to be lifted off the tin onto rollers [1].
Hot tin has a problem. If it touches oxygen, it forms a crusty oxide. To prevent that, the space above the bath is filled with nitrogen and hydrogen gas, kept at a slightly higher pressure than the outside air so that air does not leak in [1].
After the bath, the glass moves through a long oven called a lehr, roughly 100 meters (330 feet) of it, where it cools slowly. This is called annealing. It is done before the glass is cut into sheets [1]. The slow cooling matters because uneven cooling leaves stress inside glass.
3. What decides how thick the glass sheet is?
If you pour glass onto tin and just let it spread, it does not thin out forever. It settles at a thickness of roughly 7 millimeters. One patent text gives about 6.8 mm for this "equilibrium thickness" [3]. Gravity pulls the glass flat, and the surface tension of the liquids pulls the other way. At that thickness the two balance. I could only read that figure in the patent. A university lecture I found through search reports a similar value of about 6.9 mm, but I could not open its text, so I am not relying on it.
Most window glass is not 7 mm thick, so makers change the balance. Wikipedia explains that the flow of glass onto the bath is controlled by a gate called a tweel, and that the speed of the flow and of the rollers that pull the ribbon forward helps set thickness. Rollers on the sides of the ribbon control its thickness and width [1]. A faster pull stretches the ribbon thinner, much like pulling warm taffy. The patent adds that the equilibrium thickness depends on the glass temperature and on how fast the ribbon is moved forward [3].
The sources I could read do not give a full range of thicknesses or explain the roller arrangement in detail, so I will not guess at those numbers here. A good question for your own research is how the edges of the ribbon are held so that it does not shrink in width as it is stretched.
4. How was window glass flattened before the tin bath?
For centuries, flat glass was a hard thing to make. Until the 1500s, window glass was typically cut from large spun discs. Later, glassmakers blew large cylinders, cut them open, and flattened them. These cylinders limited the size of panes, which is why old windows are divided into small rectangles [1].
In 1848 Henry Bessemer patented a way to roll out a continuous ribbon of glass. It was costly because the surfaces still had to be polished afterward [1]. From the early 1920s, the best glass, called plate glass, was made as a ribbon that passed through a long line of grinders and polishers [1]. Grinding is slow, uses a lot of energy, and wastes glass. A bath of liquid that is already level removes that step.
5. How long did it take to make the float process work?
Alastair Pilkington of Pilkington Brothers had the idea in the early 1950s. One account says it then took seven years of work to prove that it could succeed, at a high cost for what was then a family-owned company [2]. Wikipedia places the development of the first commercial version between 1953 and 1957 at the Cowley Hill Works in St Helens, England, with fully profitable sales first reached in 1960 [1].
These dates are not exactly the same across sources. The dates the sources give for the start, the first sales and profit depend on what each one counts. What they agree on is that it was a long, expensive effort. Once it worked, the process spread. Pilkington's first foreign license went to the Pittsburgh Plate Glass Company in 1962, and makers in Europe, Japan and elsewhere followed. Float plants went up in many countries.
Earlier people had also tried to form glass on molten metal. Some of those attempts received patents but were unworkable at the time [1]. An idea can be right and still have to wait for the furnaces, controls and money that make it practical.
6. How can you notice flat glass in daily life?
Hold a piece of window glass and a piece of glass from an old photo frame or a bottle side by side. Look at the reflection of a straight line, such as a door frame, as it slides across each surface. Does the line bend or ripple on one but not the other? Flat float glass should keep the line straight. Do not use broken or chipped glass, and ask an adult before taking apart a frame.
If you want to follow the story further, the Sources below are a good start. Try to find out how the width of the ribbon is controlled, and how thin the thinnest float glass can be. Those are things I could not confirm from the pages I read.