| Established in 1952 | |
Page: 2040 |
The use of a precision ball held against a seat, to check the flow of liquids or gasses, is a common design tactic. A precision ball makes the most accurate and least expensive valve poppet available.
We produce precision balls in a wide variety of sizes and materials that will provide the required performance in systems from life support to lawn sprinklers, from aerospace and industrial hydraulics to fuel cells.
Our standard bearing ball materials are typically hard and wear resistant. In addition we have standard precision balls available in a number of corrosive resistant metals and several plastics. For special applications such as very high temperatures, extremely corrosive environments, or radioactive situations, we will custom manufacture precision balls that will meet these needs also.
If you look at the well-machined surface of a ball seat with the aid of a stereo microscope, it will be wavy and fragmented. These abrupt undulations in the surface will not seal completely, and there will be leakage of the ball seat.
Brinelling or coining a narrow spherical ring in the ball seat of the valve body is a common, inexpensive, way of providing a high quality-sealing surface. The ideal way to achieve this Brinelled ring is by applying a controlled force to a hard steel or tungsten carbide ball while it is sitting in the ball seat.
Although it is not the best practice, it is quite common for the ball to be hit with a hammer to achieve the rather high force required to achieve permanent deformation of the ball seat. The greatest limitation to using a hammer in this application is the uncontrolled or variable force applied by it from part to part. (Caution) Safety equipment should be applied in case metal fragments are ejected from the area of pressure application
Some of the limitations that apply to brinelling a ball seat are: The hardness of the seat must be below 40 HRC, The structure around the seat must be mechanically uniform with no adjacent slots, holes etc. There will be spring back or elastic recovery when the force is removed so the radius of the seat generated by brinelling will be somewhat smaller than the ball used to generate it. This problem can be overcome by using a slightly oversize ball as the coining tool.
All of the problems involved with the coining or brinelling process can be avoided by generating the high quality ball-seating surface through the lapping technique. The standard free abrasive lapping process using abrasive powder mixed with oil or grease to form slurry is a slow, messy process. A better solution is to use our spherical, or ball-lapping tools. They have a heavy concentration of diamond particles embedded in their surface. These diamond embedded lapping balls are usually glued in a shallow conical cup on the end of a cylindrical pin, to facilitate driving it in the seat lapping operation (see Figure #1.)
We can supply the assembled tools to your requirements or just the diamond charged balls for your assembly. If you do the assembly in house, the diamond ball lap and the conical cup on the end of the cylindrical driver should be cleaned with alcohol before applying the glue. A high shear strength epoxy glue such as our EG3000 should be used.
Just before applying the diamond-lapping tool to the seat, it should be dipped into a low viscosity mineral oil. This oil is to prevent the diamond-lapping tool from galling when it comes into contact with the sharp edge of the seat, and it will carry away fragments of the metal removed. The lapping tool is slowly rotated against the seat with a little oscillation, either by hand, or with a flex shaft, or some other very light weight slow moving driver. To improve the roundness of the spherical seat generated by the diamond-lapping tool, a flexible rubber coupling (such as those manufactured by the "Lord Company" ) should be inserted between the rotating driver and the diamond-lapping tool.
Normally the coarsest diamond grit used is 9 micron, the most common is 6 micron, which can usually be used for rough and finish lapping, and 3 micron is the finest commonly used.
When working with seats made of hard and or brittle
materials, the lapping process in one form or
another is the only viable approach. In
addition to its other attributes, the ball seat lapping process
carries almost no risk of personal injury.
| Part No. | Ball Size | Diamond Size | Loose Ball Price |
Ball on Stem Price |
| DCB-62-3 | 1/16" - 0.0625" | 3 micron | $ 8.50 | $ 16.50 |
| DCB-62-6 | 1/16" - 0.0625" | 6 micron | $ 8.50 | $ 16.50 |
| DCB-62-9 | 1/16" - 0.0625" | 9 micron | $ 8.50 | $ 16.50 |
| DCB-93-3 | 3/32" - 0.093" | 3 micron | $ 7.50 | $ 15.00 |
| DCB-93-6 | 3/32" - 0.093" | 6 micron | $ 7.50 | $ 15.00 |
| DCB-93-9 | 3/32" - 0.093" | 9 micron | $ 7.50 | $ 15.00 |
| DCB-12-3 | 1/8" - 0.125" | 3 micron | $ 7.50 | $ 15.00 |
| DCB-12-6 | 1/8" - 0.125" | 6 micron | $ 7.50 | $ 15.00 |
| DCB-12-9 | 1/8" - 0.125" | 9 micron | $ 7.50 | $ 15.00 |
| DCB-16-3 | 5/32" - 0.15625" | 3 micron | $ 7.50 | $ 15.00 |
| DCB-16-6 | 5/32" - 0.15625" | 6 micron | $ 7.50 | $ 15.00 |
| DCB-16-9 | 5/32" - 0.15625" | 9 micron | $ 7.50 | $ 15.00 |
| DCB-18-3 | 3/16" - 0.1875" | 3 micron | $ 7.50 | $ 15.00 |
| DCB-18-6 | 3/16" - 0.1875" | 6 micron | $ 7.50 | $ 15.00 |
| DCB-18-9 | 3/16" - 0.1875" | 9 micron | $ 7.50 | $ 15.00 |
| DCB-21-3 | 7/32" - 0.21875 " | 3 micron | $ 7.50 | $ 15.00 |
| DCB-21-6 | 7/32" - 0.21875 " | 6 micron | $ 7.50 | $ 15.00 |
| DCB-21-9 | 7/32" - 0.21875 " | 9 micron | $ 7.50 | $ 15.00 |
| DCB-25-3 | 1/4" - 0.25" | 3 micron | $ 7.50 | $ 15.00 |
| DCB-25-6 | 1/4" - 0.25" | 6 micron | $ 7.50 | $ 15.00 |
| DCB-25-9 | 1/4" - 0.25" | 9 micron | $ 7.50 | $ 15.00 |
| DCB-28-3 | 9/32" - 0.281" | 3 micron | $ 7.50 | $ 15.00 |
| DCB-28-6 | 9/32" - 0.281" | 6 micron | $ 8.00 | $ 15.50 |
| DCB-28-9 | 9/32" - 0.281" | 9 micron | $ 8.00 | $ 15.50 |
| DCB-31-3 | 5/16" - 0.3125" | 3 micron | $ 8.00 | $ 15.50 |
| DCB-31-6 | 5/16" - 0.3125" | 6 micron | $ 8.00 | $ 15.50 |
| DCB-31-9 | 5/16" - 0.3125" | 9 micron | $ 8.00 | $ 15.50 |
| DCB-34-3 | 11/32" - 0.34375" | 3 micron | $ 8.00 | $ 15.50 |
| DCB-34-6 | 11/32" - 0.34375" | 6 micron | $ 8.00 | $ 15.50 |
| DCB-34-9 | 11/32" - 0.34375" | 9 micron | $ 8.00 | $ 15.50 |
| DCB-37-3 | 3/8" - 0.375" | 3 micron | $ 8.50 | $ 16.00 |
| DCB-37-6 | 3/8" - 0.375" | 6 micron | $ 8.50 | $ 16.00 |
| DCB-37-9 | 3/8" - 0.375" | 9 micron | $ 8.50 | $ 16.00 |
| DCB-40-3 | 13/32" - 0.40625" | 3 micron | $ 8.50 | $ 16.00 |
| DCB-40-6 | 13/32" - 0.40625" | 6 micron | $ 8.50 | $ 16.00 |
| DCB-40-9 | 13/32" - 0.40625" | 9 micron | $ 8.50 | $ 16.00 |
| DCB-50-3 | 1/2" - 0.500" | 3 micron | $ 8.50 | $ 16.00 |
| DCB-50-6 | 1/2" - 0.500" | 6 micron | $ 8.50 | $ 16.00 |
| DCB-50-9 | 1/2" - 0.500" | 9 micron | $ 8.50 | $ 16.00 |
| DCB-62-3 | 5/8" - 0.625" | 3 micron | $ 9.00 | $ 16.50 |
| DCB-62-6 | 5/8" - 0.625" | 6 micron | $ 9.00 | $ 16.50 |
| DCB-62-9 | 5/8" - 0.625" | 9 micron | $ 9.00 | $ 16.50 |
| DCB-75-3 | 3/4" - 0.750" | 3 micron | $ 9.00 | $ 16.50 |
| DCB-75-6 | 3/4" - 0.750" | 6 micron | $ 9.00 | $ 16.50 |
| DCB-75-9 | 3/4" - 0.750" | 9 micron | $ 9.00 | $ 16.50 |
| DCB-1-3 | 1.00" | 3 micron | $ 9.00 | $ 16.50 |
| DCB-1-6 | 1.00" | 6 micron | $ 9.00 | $ 16.50 |
| DCB-1-9 | 1.00" | 9 micron | $ 9.00 | $ 16.50 |
A widely used ball seat design consists of an internal conical surface with a through hole. This design has the advantage that the ball will self center as it moves toward the seat. There are two functional variations of this design.
#1. In the first, the contact of the ball is tangent to the cone. Machining tolerances for this design can be liberal as the ball contact only needs to fall somewhere along the conical surface (see Figure #2.). For this design, to work well, the cone must be machined very round.
#2. The second version of the conical seat has the ball making contact right at the intersection of the cone and the through hole. Using the same conical angle this design yields much higher sealing forces. Therefore, a leak free valve is easier to produce, while still maintaining the good self-centering of the ball and the quick reaction of the valve. The machining tolerances for this design are quite tight, as the ball must end up exactly at the intersection of the cone and the through hole (see Figure #3). This design requires a very round cone and through hole and good concentricity between them.
A design question that must be answered is what should the angle of the conical seat be? There are several factors involved. The steeper the angle the better the balls self-aligning tendency will be. The steepness of the angle must be tempered by the fact that the ball must be released from the seat when the flow reverses. The flatter the seat angle the crisper the ball release will be, but as the included angle increases the sealing quality is reduced. The majority of ball seats have an included angle of 90 degrees, which is 45 degrees per side, or very near it.
The main alternative to the cone with a through hole is to use a flat seating surface with a through hole. This design will give a much higher elastic load on the seat, which in turn will result in a better seal at lower pressures. An additional advantage is that the flat surface and the through hole are both easy surfaces to generate to high levels of quality. This is very important because any errors in flatness, any errors in roundness and any error in squareness will reflect directly into the quality of the sealing surface. The main drawback of this design is that it lacks the self-centering feature of the cone (see Figure #4.)
The ball contact angle on the flat seat design
is determined by the ratio of the through hole diameter to the
ball diameter. The higher up on the ball, that is the steeper
the included angle, the better the seal, but the poorer the
balls release function will be. Design wisdom seems to have
settled on 45 degrees per side or a 90 degrees included angle
(see Figure #5.).
The ultimate in zero leakage ball seat designs is the Tri-Point Seat.
A significant feature of a sphere is that it has a single radius in all directions. This uniquely perfect geometry is used to form the ultimate ball seat. At the intersection of two concave spherical surfaces, of two different diameters, there will occur a perfectly circular land. There are no concentricity errors and there are no squareness errors. The two concave spherical cups are almost always generated by lapping. In practice the two spherical cups are usually 10 to 15 percent larger than and smaller than the valve ball. A very narrow, spherical land (.002 to .005 inch [.05mm to .13mm] wide), that is the same radius as the valve ball is usually lapped on the land at the intersection of the two spherical cups (see Figure #6.). The material used in these seats is usually a high tensile, high hardness metal. Because very high Hertzine Elastic forces result from this design, very low pressures can be sealed to zero leakage. Even Hydrogen and Helium gas can be controlled with this ball seat design.
For the very same reason that it works so well at low pressure, this seat is limited for high-pressure applications. As the pressure on the ball goes up, the elastic limit of the seat material will eventually be exceeded and the sealing land will be brinelled into a broad spherical cup that will no longer seal. A high force may cause cracks in the seat area.
Even with the best achievable geometry and the finest surface finish of the ball and seat, it will still have some leakage. The inhomogeneous crystallographic structure of the metals is huge when compared with the atomic dimensions of gases and liquids. In order to totally eliminate all leakage, enough force must be applied to the ball to achieve mechanical compliance between the contacting surfaces. A small amount of Hertzine Elastic Deformation is required to close the tiny undulations in the surfaces of the ball and the seat.
When the ball check is in the sealing mode, the pressure of the liquid or gas, will be forcing the ball down against the surface of the seat thus increasing the sealing force. The somewhat illogical fact is that very low pressures are much more difficult to achieve a zero leakage seal than higher pressure.
Designing a ball check valve starts with the diameter of the through hole. The diameter of the through hole is determined by the required flow rate, which is in turn dictated by the Reynolds' number, the fluid pressure and the Kinematic viscosity of the solution or gas. Charts and graphs that will give close estimates of the through hole diameter are available in Mechanical Engineering Hand Books. The seat configuration can be chosen by applying logic to the earlier discussions. The ball diameter is dictated by the through hole diameter, the seat configuration and the desired contact angle between the ball and the seat.
The majority of ball-check valves have at least a light spring pressure forcing the ball against the seat and when the ball check forms a relief valve the spring may be substantial. In most spring loaded applications, the force is supplied by a helical compression spring. For small spring loads, the end coil of the spring sits directly on the surface of the ball, on the side opposite the ball seat (see Figure #3.). When applying a heavy force, a spring guide plate is added to the design. This device has a spherical cup that fits over the ball on one side and an extended boss on the opposite side that locates on the inside diameter of the spring.
The mass of a specific diameter ball is fixed. As the design parameters of the valve are changed, to accommodate a larger flow rate, the diameter of the ball must increase. The volume and therefore the mass of a ball varies as four times pi times the radius cubed, over three { 4 * 3.14 * (r^3) / 3 }, so any increase in the ball diameter has a dramatic affect on the mass of the ball. An increase in the mass of the ball will cause the reaction time of the valve to slow down and the increased inertia of the large ball will pound the living daylights out of the seat. By substituting a large, lightweight, spherical tipped poppet for the massive sphere, the good self-alignment and excellent sealing qualities of the ball are preserved with a fast reaction time and only moderate wear of the components (see Figure #8). We can manufacture the spherical poppets for you, or we can precision lap the spherical end of the poppets you supply to a high degree of perfection so that they will form a seal comparable to a high quality ball.
In a closing note, we will look at using balls made of highly resilient materials such as rubbers and plastics. Rubber still finds some applications in low-pressure devices but it has been replaced by much less expensive plastics in most applications.
One common consideration is the specific gravity or mass of the plastic ball material. Polyethylene and polypropylene have specific masses of approximately .98 so they will just float in water. Polyethylene is slightly softer and more resilient while polypropylene is harder and tougher. Both materials have good resistance to organic solvents as well as moderate acids and bases. Both are low cost materials that we stock in most sizes.
For heavy-duty applications Nylon is the hands down favorite. It is fairly hard, very tough i.e. wear resistant and we stock finish ground balls in most sizes. Nylon is resistant to most organic solvents as well as mild acids and bases. Nylon does have a problem in aqueous solutions, as it is slightly hygroscopic and will swell about 1/4 of one percent.
For very high and very low temperatures Teflon is the obvious choice. Teflon has a high density i.e., it is heavy. We stock precision ground Teflon balls in most sizes. Teflon is not cheap, it is several times as expensive as most other plastics. There is another problem, in that Teflon can cold flow under heavy pressure and may not come out of the seat after a long exposure to a high-pressure condition. Care must be exercised in the seat design when using this material. A ball tangent to a rather large conical angle is the normal design.
The simplest test for a ball check valve is to pressurize the valve and measure the leakage. For hydraulic valves it is not uncommon for the test specifications to give a maximum number of drops per minute, and for gas the number of bubbles, at a given pressure that is an acceptable leakage. This method may not be appropriate for several reasons. The leakage may be so slight that it will take a long test period to measure it. The small amount of leakage may get lost in the valve body passages. It may be very expensive to find out that a very complex, completely assembled product must be reworked.
A faster, cleaner, cheaper and more sensitive method is to pressurize the valve with a compressed gas (frequently shop air) at a given pressure, while submerging the valve body in a tank of tap water. The quality of the valve is evaluated by counting the number of bubbles that escape in a given span of time. It should be noted that corrosion prone materials might rust due to submersion in the water bath, so an alternative liquid may be required.
Another very sophisticated approach is to pull a high vacuum down stream of the ball check valve. Then to close a zero leak valve to trap a fixed volume container at vacuum. By measuring the time for the decay of the vacuum to a given pressure, the quality of the ball check valve can be quickly and accurately evaluated. This is probably the fastest and most reliable test that can be performed. After establishing acceptable parameters, this test can be used in highly automated environments. We recommend the JRM Vacuum tester. See: http://www.delucatest.com/access.htm
We produced metal-to-metal ball check valves for the fuel cells of the Apollo Space program. The Tri-Point seat was used in these valves. The system performance was so critical that the leakage was measured with a mass spectrometer. This test had the capability to detect even one molecule leakage of the helium gas used for the test.
Copyright © 1999, revised
2007 Micro Surface Engr. Inc. |
||||||||
|---|---|---|---|---|---|---|---|---|