Nonmetallic gaskets: Avoiding leaks and blow outs
The compressive stress on a gasket plays a larger role in its ability to maintain pressure than its tensile strength. Why is this significant? It is the reason for many gasketed joint problems!
In a classic scenario, a joint is assembled without controlled bolt loads; that is, without known and controlled bolt torques. The joint withstands a hydro test at 1.5 or 2 times rated pressure, yet leaks or blows out after a period of service at pressures well below the test pressure.
Was this failure due to a loss of gasket tensile strength? Probably not. Gasket tensile strength alone cannot be counted on to hold system pressures. What very likely occurred was loss of compressive stress on the gasket.
Compressive stress is crucial
Why is compressive stress so important, and how does it affect the performance of a given gasket?
A flange or pipe line holds internal pressure because the tensile strength of the flange and pipe material are higher than the forces created by the pressure. Once installed and tested, a pipe system’s pressure capability is predicted based on tensile strength and material thickness. A gasket, on the other hand, holds pressures much higher than it could theoretically hold by virtue of its tensile strength alone.
Tests conducted with 2-in. flanges at ambient temperature prove this point. For example, most compressed fiber gaskets hold 5000-psig internal pressure without rupture. Since the cross-section of the gasket is only 5/8-in. wide and 1/8-in. thick, the outward force can be eight times higher than the maximum possible breaking force in the gasket. With larger diameter gaskets, and therefore larger exposed surfaces, the ratio of outward force to maximum tensile force becomes even greater.
Therefore, 80% to 90% of the pressure capabilities of newly installed gaskets must come from something other than their own tensile strength. That source of blow-out resistance is the friction of the gasket against the flange.
The force to push an object across a surface depends on two factors: coefficient of friction of the materials against each other, and “normal” (perpendicular to the plane of motion) force acting on the materials. For instance, in pushing a heavy box across a floor, the coefficient of friction is determined by the materials of construction and roughness of the box and floor surfaces. The normal force is the weight of the box.
Flange finish affects blow-out resistance of nonmetallic gaskets, since it changes the coefficient of friction of the gasket against the flange. Grease on a gasket face, or an extremely fine or smooth flange surface, reduces this coefficient.
In dealing with gasketed joints, consider the force that pushes the gasket out of the flanges, resulting in a blow out. The outward pressure of the contained fluid produces force on the inside edge of the gasket. The thicker the gasket, the larger the surface area presented to the pressure, resulting in even larger outward force.
This outward pressure is resisted by the gasket’s tensile strength and its friction against the flanges. Force normal to the gasket (“weight of the box”) is supplied by the bolts. Increasing the bolt loading raises the friction force, which results in higher blow-out resistance. Figure 1 illustrates this effect in blow-out tests with a joint installed at two different gasket compressive stresses.
Given the importance of bolt loading and gasket compressive stress, any change in these parameters over time is also a major concern. Many factors can contribute to loss of bolt load, including gasket creep, which typically increases with temperature (Fig. 2) and with the number of temperature fluctuations, bolt creep, fastener thread embedment, vibration, variations in temperature of flanges to bolts, dissimilar metals with dissimilar thermal expansion rates, and more. The bottom line is that compressive stress on a gasket probably declines after installation.
Creep is affected by gasket thickness and material. Thicker gaskets creep more than thinner ones. Figure 3 compares residual gasket stress for two thicknesses. Unfortunately, installers cannot always use thin gaskets. Flange irregularities may call for thicker gaskets, since they are more forgiving of bad flange surfaces.
Because the role of gasket thickness in containing pressure is not intuitively obvious, it’s important to remember that the contained pressure sees the ID of a gasket as a wall on which to push. Outward force on a gasket is proportional to its thickness. So, a thicker gasket loses more compressive stress than a thinner gasket and experiences more outward force. Although thicker gaskets are harder to rupture than thinner ones, the increase in rupture strength is not enough to overcome effects of the higher creep relaxation.
Much of this discussion may seem to be common sense or overly obvious. The somewhat surprising part is the relationship between compressive load and blow-out resistance over time. Any installer who fully understands the importance of compressive load will tremendously reduce gasketed joint failures.
Installation without torque wrenches can have expensive consequences. A contractor installs a series of gaskets in pipe flanges, but chooses not to use torque wrenches. A few joints develop leaks. The customer investigates and discovers that the gasket manufacturer’s literature calls for torque wrenches.
When a few joints leak, even though hundreds are in service, the customer has a strong argument that improper installation may be the problem, and could demand a very costly replacement of all gaskets. This work could include the cost of reinsulation.
“But we used to…”
The reader may well be asking: “Why have we gotten away without using torque wrenches for so many years?”
The answer is that many joints assembled without controlled torque have worked for decades because they use gaskets made of compressed asbestos fiber (CAF). The assumption has been that it was the high residual tensile strength in service that was responsible for this performance level. Although tensile strength may be high enough to hold the pressure in low pressure systems, it is not the main reason asbestos worked without controlled torque.
Asbestos is very good at retaining compressive stress at elevated temperature. This factor means that a gasket that withstood a hydro test would probably work well for years, because the load on the gasket during service remained quite close to the load on the gasket immediately after installation.
Retaining compressive stress
Given that compressive stress is such a crucial factor in joint performance, how can we maximize its retention?
Two approaches are available: Increase the initial compressive stress in the joint or reduce the loss of compressive stress over time. Each of the following tactics addresses one or both of these approaches.
1. Control the bolt load. One of the easiest steps is to determine the torque recommended for the gasket and then use it. This action may mean tightening a bolt to a load many times higher than what is needed to simply pass the hydro test. Use of flat washers and well lubricated hardware minimize friction forces when turning the fastener.
2. Increase the design stress. When designing a joint, build for higher gasket stresses to achieve the needed performance. Current designs frequently use minimum loads required, not optimum. Performance of gasketed joints designed this way would be greatly improved by using the recommended gasket seating stress in place of the minimum.
The main reason code designed flanges work well despite using a minimum stress is probably related to the use of conservative bolt stresses for design. Design is usually limited to 25% of bolt yield, while the installer may tighten to 50% or 60% of yield, thereby achieving a compressive stress above the minimum.
This approach to stress levels should also be applied to the new “gasket constants” design methods. However, the change may not be as significant because in designing for tighter emissions the design gasket stresses go up. (Fig. 4)
3. Use thicker flanges. Using thicker flanges affects both the initial and retained gasket stress. Thicker flanges distribute the bolt load more evenly, and allow use of higher bolt stresses. In addition, thicker flanges create a longer grip length in the bolt (actual working length of the bolt, from the head to the nut, approximately).
A longer grip length means more stored energy in the bolts, because a bolt stretches a certain percentage at a given stress level, like a heavy spring. The actual stretch will be that percentage multiplied by the grip length of the bolt. A longer grip length means more stretch.
4. Use longer bolts. Sometimes chronic problems occur in services where the poor corrosion resistance of high-strength bolts makes the use of low strength bolts necessary. This condition severely limits the gasket stress and bolt stretch.
Using sleeves over bolts that are many inches longer than required, with the washers and nuts placed at the ends of these sleeves, creates a much longer grip length. The gasket stress is not changed, but the bolt stretch is much higher. The joint then retains more stress on the gasket in service, because the joint creep has a smaller effect on the bolt stretch.
The same condition is accomplished by using backer plates over the flanges to effectively increase the bolt grip length. The use of disc spring washers can be considered, since they have an effect similar to increasing the bolt stretch.
5. Avoid over-size bolts. This problem can occur if flanges built for a metal gasket have become pitted or worn and are difficult to seal with the original gasket. If a nonmetallic gasket is substituted, it is more forgiving of those surface irregularities, but less resistant to damage from crush.
To use the softer gasket, the bolt loads must be limited, but that limits the stretch in the bolts. If this joint is operated at elevated temperatures, there may be insufficient stretch in the bolts to handle the joint relaxation.
6. Select the correct gasket. The choice of gasketing also affects gasket stress, by minimizing the stress loss under service conditions. Selecting high quality gasket material that is appropriate for the service conditions increases the load retention of the joint.
Data from ASTM F38 Creep Relaxation and DIN 52913 Stress Relaxation tests help select the gasket needed to improve stress retention at the service temperature.
7. Select the thinnest gasket possible. Gasket thickness also affects load retention, because thicker gaskets lose more compressive load than thinner gaskets (Fig. 5). The only advantage to thicker gaskets is their superior ability to conform to flange irregularities.
Joint performance is improved if, instead of using thicker gaskets to compensate for flange irregularities, the flange irregularities are reduced, so that thinner gaskets can be used. Don’t forget that thicker gaskets are less resistant to blow out. — Rick Dunn, Chief Editor, 630-320-7141, firstname.lastname@example.org
The tensile strength of a gasket is not what keeps it from blowing out.
Compressive stress on the gasket accounts for most of a joint’s pressure capability.
The author is willing to answer technical questions about this article. Mr. Burgess is available at 315-597-3386, e-mail Dave_ Burgess@Garlock.ccmail. compuserve.com.