High-Purity Gas Panels Part 8: Flow Restrictors

Restrictors2Figure 1: Two types of  gas flow restrictors

In this week’s entry in our blog series exploring the design of high-purity gas panel design for semiconductor fabs, I’ll be covering components called flow restrictors or flow limiters.  The subject of last week’s entry was pressure regulators and their utility in proportionally controlling the pressure in a gas line.  Flow restrictors share a similar basic function, in that they limit the flow rate of gas; however, unlike pressure regulators, flow restrictors are typically not adjustable or controllable.  Despite being simple monolithic fittings with no moving parts, they are invaluable for maintaining the safety and efficiency of a semiconductor process.

Flow Restrictor Principles

Flow restrictors are simple devices that reduce limit the flow of gas in a line.  Like the check valves covered in part 6, flow restrictors are not usually under the control of the user or control system.  All flow restrictors function simply by reducing the cross sectional area of the gas line.   In simple mathematical models, gas acceleration and de-pressurization is followed by a return to the original state when the cross section widens out again, but flow restrictors have two factors working for them that keeps the pressure low after the constriction:

  1. At low flow velocities, the gas will lose energy in the restrictor due to frictional effects in the orifice and the development of turbulent flow on the outlet.  Thus, the gas will be unable to return to its original pressure when the pipe expands back to its original size.
  2. At higher flow velocities, the gas can reach a state called “choked flow”, in which it reaches a near sonic velocity in the constricted segment.  At this point, “a decrease in downstream pressure can only propagate at the speed of sound and therefore cannot pass back through the orifice”1.  In other words, once the gas reaches this state the upstream flow will remain at a constant pressure no matter how the pressure drops on the downstream side.  The flow rate only depends on the upstream pressure.  Thus, flow restrictors usually include the pressure as part of the specification.

Flow Restrictor Design

The simplest flow restrictors reduce the cross sectional area by employing a simple plate with a hole laser-drilled through the center of it (see the upper restrictor in the Figure 1).  The size of the hole requires careful calculation, however, as the flow profile for any given diameter will change depending on the density of the gas, the inlet pressure, and the desired outlet pressure and flow rate.  Manufacturers often use laser drills, due to the difficulty in drilling such small, precise holes.  Manufacturers offer drilled orifices set within a standard pipe fitting with threaded ends, or as single gaskets that install between the face seals on a VCR® coupling.

Another option uses a porous media, or “frit”, to reduce the cross-sectional area (the lower restrictor in the Figure 1).  Instead of just one straight hole, frits channel the gas through hundreds of pores in a central plug.  Although generally more expensive than a simple laser-drilled plate, these offer a number of advantages.  The pores in the frit reduce the gas velocity as it travels through, reducing the erosion of the restrictor over its lifetime as compared to a drilled orifice; this will keep the restrictor accurate and consistent over a longer lifespan than its competitor.  Additionally, with so many different pathways for the gas, it’s much less likely that a piece of debris will fully clog a frit than a single drilled orifice.

A frit type flow restrictor and a filter, both of which use porous media, can be distinguished by their performance and function.  A frit type flow restrictor produces a relatively large pressure drop by design, but does not remove particles well; on the contrary, it is designed not to do so in order to avoid clogging.  On the other hand, a filter by nature efficiently removes particles, but minimizes the pressure drop due to its much larger surface area.

Flow Restrictor Placement

The most common place that semiconductor equipment engineers will install a flow restrictor is directly proceeding the gas cylinder outlets in the facility’s gas cabinets.  Gas cylinders often store compressed gases at pressures of 3000 pounds per square inch (PSI).  This pressure, in which the flow were left unlimited, could be a safety hazard difficult to design against.  So, by placing a restrictor before the first leg of the gas line, the system can be designed to accommodate a valve or regulator failure, because the flow restrictor limits the amount of gas or chemical that can escape.  Furthermore, by adding additional flow restrictors along the gas lines, engineers can step down the flow rate to values very close to what the process ultimately requires, thereby limiting the amount of gas that can escape if there is a leak, valve failure, regulator failure, or some other unintended problem.  The final pressure regulator and mass flow controller (MFC) can also operate over a much small range, granting the user fine control over the incoming gases, so the point of use flow restrictors are much smaller than those at the gas cylinder.

Flow Restrictor Specification Options

As they are tied into the same system as the gas lines and gas valves we’ve previously discussed, flow restrictors require that the designing mechanical engineer specify some of the same options: fitting type (VCR, Swagelok, or buttweld); fitting sex (male or female); valve diameter (¼ inch is common at point of use); and construction material (316L VIM VAR stainless steel is common).

Apart from those standard options, flow restrictors require careful selection based on the pressure and flow rate for each gas, due to differences in density and viscosity.  Typically, the manufacturers use nitrogen gas (N2) as a baseline to establish the maximum flow rate (from 30-50,000 standard cubic centimeters per minute), inlet pressure (10-300 psi) and outlet pressure (0.2-25 psi).  Armed with the known gas viscosity and target flow rate, an engineer can calculate the equivalent N2 flow rate and use that to select the correct size of flow restrictor.

Using Flow Restrictors in Semiconductor Processes

For all their simplicity in design, flow restrictors can be quite useful for a mechanical engineer or semiconductor process engineer.  They play a vital role in reducing the wear and tear on other components, increasing the accuracy of the semiconductor process, and ensuring the safety of the operators.  All that is required is a little forethought and calculation on the part of the designer, to make sure that each restrictor is appropriately sized and specified for its ultimate use.

References Cited:

1: Downie, N. A. Industrial gases. London New York: Blackie Academic & Professional, 1997. 13-15.

By | 2017-05-04T15:46:10+00:00 September 3rd, 2015|Mechanical Engineering, Safety, Semiconductor|0 Comments

Leave a Reply

%d bloggers like this: