By Peter C. Williams, Chief Scientist, Swagelok
Containing leaks is a vital consideration in the design of fluid system components for gas service. Leaks may pose risks to safety, the environment and energy conservation. Selecting components requires an understanding of how leaks occur and what variables determine the rate of leakage. It is also helpful to know how materials—metals, elastomers and plastics—behave under stress and, when assembled together, form a seal. Some seals are simply more effective than others because they are designed for a known set of fluids at known pressures. Further, seals may be improved through developments in materials technology, such as low-temperature carburization of stainless steels.
Kinetic Theory Of Leaks There are many types of leaks, with the two most common being permeation and path leaks. Permeation leaks occur through solid materials, such as fluoropolymers, by way of absorption, diffusion and desorption, whereas path leaks occur through an opening, crack or pore, and typically relate to metal-to-metal or non-elastomeric seals. In this article, we will be concerned exclusively with understanding and preventing path leaks.
Path leaks may be either molecular or viscous, depending on the geometric size of the leak relative to the molecular mean free path of collision of the leaking gas. Close consideration of the size and shape of the leak path hole can lead to some surprising observations. Mean free path is defined as the distance a molecule travels between one collision with its neighbor and the next collision with its neighbor. For example, the mean free path for helium at standard temperature and pressure for experimental measurements is about 0.25 microns. If the leak hole is large relative to the mean free path, the gas flows through the hole with laminar parabolic Poiseuille distribution (fluid flowing in parallel layers without turbulence or disruption between layers).
When the hole size is significantly smaller than the mean free path of the gas molecules, the leak is said to be molecular rather than viscous. The leak assumes a Knudsen diffusion, which means the molecules collide predominantly with the walls of the leak path rather than each other.
Additional effects include surface absorption, surface diffusion and other molecular interactions. Notwithstanding, the size of the hole has the overwhelmingly greatest effect. Successful mitigation of potential leaks requires designs that effectively minimize micron-scale holes, metal-to-metal seal defects and path leaks.
Fluid System Component Design
 Figure 1 |
 Figure 2 |
A critical question is how to eliminate or minimize these gaps and pores between mating surfaces. In some cases, elastomers or plastics may be employed. Under pressure, they deform or flow into pores, stopping up leakage pathways. In other cases, metals are preferred because of their strength, hardness, fatigue resistance, etc. Metals also plastically deform under pressure. Ideally, the design enables one surface to become knife-like in applying sharp, concentrated stress at a particular line on the second surface. Both fittings and valves employ such methods in metal-to-metal seals.
 Just as in Figure 2, the bonnet of a diaphragm valve fits onto the valve body, trapping the diaphragm in between. |
Let’s look at cross sections of three designs in which metal-to-metal surfaces create effective seals because of sharp, concentrated line stresses. In Figure 2, the bonnet of a diaphragm valve fits onto the valve body, trapping the diaphragm in between. Note the multiple points for concentrated stress.
In the Sno-Trik fitting design (Figure 3), the end of the tube is shaped into a cone before it is pulled up. Note that the tube’s cone-shaped tip and the receiving angle in the female side of the fitting are deliberately mismatched to enhance concentrated stress at the tip and heavy deformation of the metal.
In a VCR fitting (Figure 4), two raised circumferential ridges create concentrated stresses on the gasket between them. The ridges are hard, while the gasket is soft. A differential in the hardness between the two mating surfaces enhances metal deformation, and increases the filling in of pores and gaps.
 Figure 3 |
A difference in hardness between two mating metal surfaces is an effective approach to closing gaps and pores that could become leakage pathways. However, in most applications, a soft metal component is undesirable. Stainless steel and alloy tubing, for example, is commonly required with a hardness up to 90 HRB (200 HV). Designing a tube fitting or other component with a significant hardness advantage over a base material of 90 HRB is challenging and may require advanced materials, technology and treatments.
New Materials & Technology
 Figure 4 |
LTCSS occurs at low temperatures (below 800ÂşK) over long periods of time such that large quantities of carbon are diffused into the surface structure of the metal. Normally, when carbon enters the crystal structure of stainless steel, chromium moves in from surrounding areas, forming chromium carbides, which negatively affect both the achievable hardening effect and corrosion resistance. However, in the case of LTCSS, carbides do not form.
Figure 5 shows X-ray diffraction peaks at the surface of a treated specimen and at several micron depth intervals. Note that the metal remains austenitic throughout, in spite of the carbon absorption. There is no phase change between the surface and the core untreated portion, which means the treated material resists peeling or delamination.
 Figure 5 |
In tests for electrochemical pitting potential, treated 316L specimens averaged 945 mV compared with 344 mV for untreated material. In polarized crevice-corrosion tests, treated specimens showed no damage after a week, whereas untreated specimens showed damage within hours. Treated 316L in these tests show crevice-corrosion behavior comparable to that of aerospace titanium-base alloys or Hastelloy™ C22, a nickel-base alloy.
Because LTCSS may be performed on finished components without distortion or dimension change, including components with complex shapes and deep structures, it is ideal for treating small parts, such as ferrules, valve bodies and bonnets, which are principally responsible for seals in fluid system components. Treated specimens are many times harder than other standard metals used in component design; the corrosion-resistant qualities are often enhanced; and their austenitic ductility is substantially maintained.
When creating a seal between two metal surfaces, the harder metal plastically deforms the softer metal, closing gaps and pores. With low-temperature carburization, this differential hardness is achievable to a great degree. In addition, ductility enhances this effect between two metals of different hardnesses. The harder metal, which is also ductile, deforms along with the softer metal, while it smoothes out gaps and pores. Also, a ductile surface is not brittle, and does not allow new gaps and pores to open on its surface as it deforms.
Conclusion Containing leaks in fluid system components is a complex science and art. However, even a basic understanding of the issues enables one to make intelligent choices in selection. Look for designs that create sharp, concentrated stresses over a very narrow sealing area. Localized strains permit the sealing materials to plastically deform and fill in the gaps created when flat surfaces are assembled to form a seal. A difference in hardness between the mating materials is especially useful in causing materials to plastically deform. In the years to come, new materials and treatments, such as LTCSS, will further enhance the ability to contain leaks.
This article is based on a presentation given at NGV 2008 – 11th IANGV Conference and Exhibition. For more information, visit www.iangv.org.