{"id":2921,"date":"2026-10-13T09:00:00","date_gmt":"2026-10-13T09:00:00","guid":{"rendered":"https:\/\/www.gasyco.com\/staging\/variable-area-flow-meters\/"},"modified":"2026-10-20T14:25:17","modified_gmt":"2026-10-20T14:25:17","slug":"variable-area-flow-meters","status":"publish","type":"post","link":"https:\/\/www.gasyco.com\/en\/variable-area-flow-meters\/","title":{"rendered":"Variable Area Flow Meters: When Gravity Holds the Scale"},"content":{"rendered":"

Over eight articles this series has measured flow through heat, pressure differences, a vibrating tube, gear teeth, sound and vortices, and then put them side by side. One principle was missing, probably the one most engineers saw first: a float in a glass tube that rises with the flow. No power, no electronics. You read it with your eyes.<\/p>\n\n\n\n

The float rises until the gap around it lets the flow pass at exactly the pressure difference that carries its weight.<\/p>\n\n\n\n

Most people call it a rotameter. At the end we explain why we avoid that word.<\/p>\n\n\n\n

One note on scope. We use this technology ourselves, in one place. In our mobile gas analysers a variable area meter with a needle valve sits in the bypass. It sets the flow drawn from the sampling point, while a small share is branched off internally for analysis. Its scale reads litres per minute of air, so for the welding gas mixtures that actually flow through it, the number is an orientation. For that job it is exactly right.<\/p>\n\n\n\n

How does a float measure flow?<\/h2>\n\n\n\n

The tube is slightly conical, narrow at the bottom and wider at the top, and the flow enters from below. Three forces act on the float: its weight, the buoyancy of the medium and the pressure difference across it. In equilibrium, \u0394p \u00d7 A = V \u00d7 (\u03c1F \u2212 \u03c1) \u00d7 g, where A and V are the float’s cross-section and volume, \u03c1F its density and \u03c1 the density of the medium. Nothing on the right depends on flow, so \u0394p stays constant over the whole range.<\/p>\n\n\n\n

What changes is the area. More flow pushes the float higher, where the annular gap is wider, until the flow fits through at that fixed \u0394p. Through the gap, the flow follows the same law as through an orifice: Q = \u03b1 \u00d7 a \u00d7 \u221a(2\u0394p\/\u03c1), with a the gap area.<\/p>\n\n\n\n

The orifice plate keeps the area fixed and lets the pressure difference vary. The variable area meter keeps the pressure difference fixed and lets the area vary.<\/p>\n\n\n\n

The gap area grows almost linearly with height, so the scale is nearly linear. The square-root law that limits an orifice to about 3:1 does not squeeze the bottom of this scale. One manufacturer gives a usual range of 1:12.5, from 8 to 100 percent.<\/p>\n\n\n\n

Which designs are there?<\/h2>\n\n\n\n

The classic version is a glass tube with the scale printed on it. Small sizes use ball floats guided by three flats or ribs inside the tube; larger sizes use shaped floats with a sharp reading edge. For high pressure, high temperature or opaque media a metal cone takes its place and a magnet in the float drives an external pointer. Glass sets the pressure limit: one manufacturer rates a glass tube series from 30 bar in the smallest size down to 2 bar at 2 inches, with a protective shield for gases.<\/p>\n\n\n\n

A needle valve is often built in. The meter then becomes a manual flow setter, as used for purge gas and in our analysers. It is not a controller: nothing corrects the setting when the supply pressure changes.<\/p>\n\n\n\n

Why does the scale only hold for one gas?<\/h2>\n\n\n\n

The scale is printed for one medium, pressure and temperature, for gases usually air at atmospheric pressure. Because the float is far denser than any gas, the flow at a given float height scales with 1\/\u221a\u03c1.<\/p>\n\n\n\n

For a meter scaled for air, one manufacturer’s table gives: for argon, multiply the reading by 0.85, for helium by 2.7, for hydrogen by 3.81. By our own calculation, an argon shielding gas with 18 percent CO2 needs about 0.84, a half-and-half argon-helium mixture about 1.15. At the same float height, the second carries over a third more gas than the first.<\/p>\n\n\n\n

Viscosity is designed out within limits: float shapes keep the scale valid up to a stated viscosity immunity ceiling. Beyond it, the scale must be made for the specific medium.<\/p>\n\n\n\n

Why does pressure matter so much?<\/h2>\n\n\n\n

Density follows pressure and temperature. For standard volume the correction is \u221a(p2\/p1 \u00d7 T1\/T2), with p1 and T1 the scale conditions and p2 and T2 those in the tube, both absolute. By our own calculation, a meter scaled for atmospheric pressure that runs at 1 bar gauge shows only about 71 percent of the true standard flow. At 2 bar gauge the true flow is 1.72 times the reading. Twenty kelvin of warming moves it about 3 percent the other way.<\/p>\n\n\n\n

The pressure that counts is the one inside the tube, and the needle valve decides what it is. With the valve at the outlet, the tube sits at supply pressure and back-pressure changes do not reach it, which is why manufacturers recommend this for gases. When the gas leaves to atmosphere, the valve has to go at the inlet and the tube runs at roughly atmospheric pressure.<\/p>\n\n\n\n

Why does the float bounce on gas?<\/h2>\n\n\n\n

With a gas, the volume between the meter and the nearest throttling point acts as a spring and the float as a mass on it. At low flow and low pressure this system can start to oscillate by itself, and the float bounces instead of settling, a so-called compression oscillation.<\/p>\n\n\n\n

One data sheet for a DN 15 metal cone meter specifies a minimum upstream pressure of 3 to 5.4 bar absolute for gas, depending on the range, or 1 bar absolute with a piston damper. VDI\/VDE 3513 Part 3 lists the remedies: low pressure loss, short pipes to the next throttle, small free volumes, higher operating pressure, and a range restricted from 10\u2013100 to 25\u2013100 percent. Pressure shocks from fast solenoid valves can drive the float into its upper stop hard enough to destroy the meter.<\/p>\n\n\n\n

Why must it hang vertically?<\/h2>\n\n\n\n

Gravity is the reference force. Every other principle in this series would work in orbit. This one would not.<\/p>\n\n\n\n

Tilting changes the axial weight component only with the cosine; by our own calculation, 5 degrees removes 0.4 percent. The real problem is friction. A tilted float leans on its guide or the tube wall, responds less freely and shows hysteresis, which is why manufacturers require vertical mounting with flow from bottom to top. Our first article found orientation sensitivity in capillary thermal sensors too, caused by convection inside the sensor. Here gravity is part of the principle itself. Spring-loaded variable area meters replace weight with a spring and can be mounted in any position, but their scale still depends on density.<\/p>\n\n\n\n

Straight inlet runs matter far less: for gases, one data sheet cites five diameters, after VDI\/VDE 3513 Part 3.<\/p>\n\n\n\n

What does the eye add?<\/h2>\n\n\n\n

The scale sits on the outside of the tube, the reading edge of the float inside, a few millimetres deeper. An eye above or below the float sees that edge against the wrong graduation. A simple geometric example: with the edge 5 mm behind the scale and the eye 15 cm above it at 50 cm distance, the reading shifts by 1.5 mm. On a 100 mm scale that is 1.5 percent of full scale, or 7.5 percent of reading at 20 percent flow. Ball floats are read at the centre, shaped floats at their marked edge.<\/p>\n\n\n\n

The device error comes on top. VDI\/VDE 3513 Part 2 defines an accuracy class as three quarters of the class value in percent of reading plus one quarter in percent of full scale. For class 2.5 that is 2.5 percent of reading at full scale, but 8.1 percent at a tenth of it.<\/p>\n\n\n\n

Strengths<\/h3>\n\n\n\n