AUTOMATION FOR DIAGNOSTICS
As the tests performed with clinical analyzers become more specific and more sensitive, the requirements placed on the fluid-handling components and assemblies used in such analyzers become more stringent. More-accurate dispense volumes, low dead volume, reduced carryover, and minimal pressure drop are expected of the fluid path traveled by reagents and samples.
These requirements can impose several design constraints on analyzers, including the use of low-surface-energy materials, the use of semirigid or rigid tubing, and the use of precision fluid connections. Also, in order to keep the cost per test low, many clinical analyzers are designed for fluid sample volumes below 20 µl. This allows less reagent to be used, which lowers the cost per test. But lower sample volumes also place greater demands on the performance capability of fluid-handling components and assemblies, as volumetric errors that used to be considered insignificant can now be problematic.
A new technology developed for fluid-connection fittings used in clinical analyzers is designed to ensure that fittings are properly tightened and thus support an optimum level of fluid-handling-system performance.
Design Constraints
Fluid carryover is often a major concern for designers of clinical analyzers. Carryover can be loosely defined as an occasion of a fluid moving through a flow path picking up some remnant of another fluid that had previously flowed along the same path. Two principal problems that can arise when the earlier fluid is carried over into the present fluid are dilution and contamination.
When a fluid traversing the fluid path picks up some volume of a preceding fluid, it becomes diluted; it no longer has the same concentration. This can be problematic when multiple reagents travel through one fluid path.
Contamination is another problem that can arise from fluid carryover. When one fluid picks up some of another, it is no longer a homogeneous fluid. If blood from one patient gathers in some from another patient, a misdiagnosis can result, as the test may be run on a mixed sample.
Materials with low surface energy, such as PTFE, FEP, PEEK, ETFE, and MFA, are used for the flow path in an effort to reduce fluid carryover. If a fluid flowing through a piece of tubing has a higher surface energy than the tubing itself, it will usually stay bound to itself, and not bind to the inner-diameter surface of the tubing. This reduces the likelihood of carryover, as the next fluid moving down the flow path may not encounter any of the previous fluid to pick up.
This is true for most fluids, but not for those with high protein content. With repeated protein flows, the tube wall will eventually become coated, and the tubing surface energy will change to that of the protein. In applications necessarily involving protein-laden solutions traversing the fluid path, the only way to have the tubing surface energy remain constant is to remove the protein from the tube wall by chemical means.
The surface finish of the tubing and discontinuities in the fluid path also can determine whether a fluid will remain bound to itself, and thus are additional factors that can potentially affect carryover.
Semirigid or rigid tubing can be used instead of flexible tubing to reduce the potential for the tubing bore to expand and contract and thus lead to errors. A tube’s inner diameter can increase when a fluid is pumped through it. This is caused by the increase in internal pressure that occurs when flow starts, and by an increase in pressure caused by a restriction. When this happens, the volume of fluid per length of tubing also increases. When the fluid flow stops, thus lowering the pressure, the tube diameter then decreases, causing a small amount of fluid to move out of the tube. If one end of the fluid path is open, as with an aspirate/dispense probe, errors in fluid dispense volumes can be the consequence.
Semirigid tubing is made, for example, from PTFE, FEP, ETFE, and MFA, while PEEK, stainless steel, and polysulfone are materials of construction for rigid tubing.
Semirigid and rigid tubing do not easily lend themselves to the device most commonly used to connect soft tubing, a barbed connector. Barbed connectors are often employed to connect tubing made of such soft materials as silicone, polyurethane, Norprene, and Tygon. These materials are easily deformed to go over the barb; then, returning to their original shape, they seal the tube against the barb. Semirigid and rigid tubing do not deform enough to go over a barb easily, if at all.
Also, many semirigid tubes will cold-flow over time when pressed over a barb. As this happens, the material begins to take the shape of the barb in a nonstressed state, losing the interference fit created with the barb. This can result in a leak.
Barbed connections can also bring about an increase in dead volume, as the area where the top of the barb meets the tube inner wall creates a step change in the diameter of the fluid path. This sudden change in flow-path diameter provides an opportunity for a certain volume of fluid to remain behind (dead volume), and not be swept out as the flow passes.
Rigid-Tubing Connection
The two most common ways to make a connection between semirigid or rigid tubing and a port are to flare the tube or use a ferrule. These types of connections are called precision fluid connections because they minimize the amount of dead volume at the connection and also reduce discontinuities in the fluid path.
) |
Figure 1. (click to enlarge) Section view of a flat-bottom port. |
A flared connection is made by forming the tube end into a shape similar to that of a trumpet bell. This flared tube is then put into a flat-bottom port (see Figure 1). A threaded fitting is used to compress the face of the flare against the face of the port. With a flared connection, either the face of the flared material or the base of the flat-bottom port deforms against the other part to create a seal (see Figure 2). Semirigid tubing is most commonly used to make a flared connection.
) |
Figure 2. (click to enlarge) Connections with flared tubing (left) and with a cone ferrule (right). |
A ferruled connection is made by pulling a piece of tubing through a ferrule, and cutting the tubing flush with the end of it. This assembly is then placed into a flat-bottom port and a fitting used to force the ferrule or tubing end to seal against the face of the port (see Figure 2). Depending on the type of ferrule used, the fitting is sometimes used to seal the tubing’s outside diameter to the ferrule inner diameter, as well. Both semirigid and rigid tubing can be used with ferrules.
Connecting flared and ferruled tubing assemblies to a flat-bottom port requires maintaining a delicate balance between tightening enough to seal and not tightening so much as to damage components or the fluid path. A seal is created by turning the threaded fitting until the fitting places enough pressure on the compressible feature of the tubing assembly to seal it against the bottom of the port.
Potential Connection Problems
Many analyzer manufacturers have included in their assembly procedures for precision fluid connections the instruction to tighten the fittings until finger tight. This is a subjective specification whose interpretation will vary from person to person. What is considered finger tight by one assembler may be loose to someone else. This can lead to several problems.
If insufficient force is exerted against the compressible component, the connection will leak. This is obviously a problem. However, it will usually show up as soon as the system is first pressurized and a leak occurs. Simply tightening the fitting a little more should fix the leak. (This is less simple if it is not in an accessible location.)
Another problem with undertightening a fitting is that it may provide a seal at the manufacturing site that is lost when the unit is shipped and used in the field, circumstances that subject it to thermal cycling and vibration. Real-world conditions can cause fittings to loosen over time and cause leaks that require an expensive field service visit to correct.
The other cause of problems that can be introduced when making a precision fluid connection is overtightening of the fitting. Some of the problems this can lead to are obvious, while others are more subtle.
Among the obvious problems is stripping the threads on the fitting or, worse, the port. As some ports are made of PTFE, which is not a very strong material, the threads can strip fairly easily, which can result in the need for an expensive replacement. If the threaded hole on a port strips, it must be either repaired or replaced in order to enable a seal to be created.
Repairing a stripped thread is usually accomplished by adding an insert thread or by reforming the thread using a larger thread size. Both techniques are labor-intensive, and can require disassembly of the part to permit machining. Also, adding an insert thread can mean that noninert materials may come into contact with fluids that could attack them. Making the thread a larger size can lead to compatibility issues in the field; because a different fitting thread size will be required for that port, a custom assembly will be necessary when a replacement is called for.
Replacing the port is costly because of the disassembly and reassembly time, as well as the cost of the new component itself.
) |
Figure 3. (click to enlarge) Flared (left) and ferruled connections that have been overtightened. |
The subtle problems associated with overtightening a fitting usually involve the fluid path and can occur with both flared and ferruled connections (see Figure 3).
As can be seen in the flared assembly at the left side of the figure, when the fitting was overtightened, the O-ring compressed the tubing such that its inner diameter was reduced by more than 30%. In the cone-ferruled assembly also pictured, the inner diameter of the tubing has again been reduced by about 30%. The tubing has pulled away from the face of the ferrule, as well. This results in dead volume, which can lead to carryover problems. The reduction in the inner diameter of the tubing has created a restriction in the fluid path, which can lead to a pressure drop and decreased flow rates. Other potential side effects of a sharp reduction in fluid-path diameter are the generation of microbubbles by turbulence and a sharp drop in fluid pressure on the low-pressure side of the restriction.
How can people know whether they have tightened a connection correctly when they do it by hand? Unfortunately, they have generally not known. The best ways to achieve correctly tightened connections have included using a torque wrench and restricting tightening of all fittings to a select few properly trained people. However, each of these approaches has some drawbacks.
While a torque wrench would indeed provide repeatable torque, the user would need to always have ready access to it. Also, it should be calibrated, and it will need a special adapter to fit around the tubing coming out of the fitting. And the space constraints encountered in many instruments can make use of a torque wrench difficult.
Training selected people to apply the proper amount of torque for creating a good seal can be done, and having these experts on staff can be useful. But this strategy can be problematic if the number of systems built is large, or if many fluid connections must be made in each system. It can also be a problem for maintaining instruments in the field, as service people might not all be adequately trained. If a fitting is undertightened but does not leak until some time after the service engineer has left, a costly return service visit will be necessary.
A Technological Solution
) |
Figure 4. (click to enlarge) Section view of a flat-bottom port. |
To solve these problems, engineers at Diba Industries Inc. (Danbury, CT) developed the Click-N-Seal torque-limiting fluid-connection fitting (see Figure 4). This fitting is designed to provide positive feedback when the correct sealing torque is reached. In addition, the design limits the amount of torque that can be applied to the threads. This technology greatly reduces the likelihood of a fitting being over- or undertightened.
The fitting is composed of two main sections: the body (shown in gray in the cross section), which is made from polycarbonate, and the cap (shown in white), which is made from acetal. The body section has threads on the bottom portion, a bore throughout its length to accommodate the tubing, and levers with protrusions in the top portion. The cap section provides clearance for the top part of the body, and has several abutments that are designed to interfere with the protrusions on the levers of the body.
As the cap is rotated, the abutments come into contact with the protrusions on the levers of the body. Torque thus is transmitted to the fitting body, and both components turn in unison. As the threaded section begins to compress the flare or ferrule against the base of the port, the levers begin to deflect; meanwhile, the cap continues to turn. The faces of the abutments and the protrusions on the levers have mating angles engineered for smooth deflection of the levers. The torque applied to the fitting body is determined by the amount of interference between the lever protrusions and the abutments on the cap, by the rigidity of the lever, and by the friction between the abutments and the protrusions. When the abutment slides over the lever protrusion completely, the lever snaps back to its previous position with an audible click. Further turning of the cap does not apply any more torque to the threads because the levers deflect at the same torque as was previously applied.
The levers in the body section of the fitting were designed so that the forces required to deflect them are within the elastic range of polycarbonate. When deflected during use, the levers do not permanently deform; they return to their original position, with their original stiffness. This allows the fitting to be used multiple times. Tests performed on the fitting have shown that it remains within the torque range required to seal a flared or ferruled connection even after 100 uses.
The backsides of the abutment and lever protrusions were designed for a more positive interference, so that greater torque is transmitted to the fitting body in the reverse direction before the levers deflect. This allows the fitting to be removed after it has been tightened.
Testing conducted by the technology developer determined that flared connections made from semirigid tubing are sealed when the torque applied to the fitting is between 13 and 30 oz.•in. Below 13 oz.•in, a seal may not be created, while above 30 oz.•in the fluid path may become restricted. This torque range is also applicable for sealing most ferrules as well. The fluid-connection fitting described here was designed to produce a nominal torque of 20 oz.•in.
The version of the fitting currently available was designed to operate in most applications now encountered in the IVD marketplace. However, its torque is not adjustable. If a high amount of torque is needed for a particular sealing application, this fitting will not be appropriate.
 |
Gary Helstern is vice president of engineering at Diba Industries, Inc. (Danbury, CT). He can be reached at gary.helstern |
Also, care must be taken with this fitting when the port is made from a soft material such as PTFE. The number of threads engaged between the fitting and a PTFE or other soft port is crucial for creating a seal without stripping the threads of the port. Without sufficient thread engagement, the PTFE can strip out, eliminating one of the fitting’s major benefits. Depending on the quality of the threads in a soft port material, a minimum of 10 threads engaged between the fitting and the port may be required to use the fitting repeatedly without stripping the threads.
Conclusion
As the capability requirements for fluid-handling components and assemblies become more stringent, several design techniques will likely have to be employed to meet them. Designers may have to use low-surface-energy materials, semirigid or rigid tubing, and precision fluid connections.
Making precision fluid connections repeatable and reliable has been problematic in the past, since finger-tightness is a subjective condition. The use of torque wrenches is not alone sufficient, and training all people who might possibly be called upon to install precision fluid connections can be impractical. But by incorporating positive feedback and a torque-limiting feature into the design of its fluid-connection fitting, a fluid-handling-component manufacturer has introduced technology that eliminates many of the problems associated with installing precision fluid connections.
