Strain and Short-Term Creep Behavior of Thermoplastics in a Luer Taper Fitting Application
Medical Plastics and Biomaterials Magazine
MPB Article Index
Originally published January 1998
TECHNICAL PAPER SERIES
The luer fitting or luer-lock fitting is among the most widely used connectors in the medical industry. Its purpose is to "connect two medical devices in a liquid-leak-proof and mechanically secure manner."1 Applications for these male and female tapered, interlocking fittings include, but are not limited to, syringes, needles, stopcocks, IV sets, and diagnostic and therapeutic catheters. The rigid connectors are available in a vast range of metals and thermoplastics, the selection of which is determined by the end use. When metal fittings are too heavy or overengineered for a specific application, thermoplastic materials are typically chosen.
One consideration in selecting a thermoplastic luer-lock material is the extent to which the fitting will strain and/or creep during its lifetime. Some luer-lock components are employed in short-term, single-use applications, whereas other fittings are assembled and disassembled several times over a period of weeks. For a female luer-lock fitting in the latter situation, the repeated cycle of stress — strain — stress relaxation — stress removal — stress reapplied is often capable of progressively distorting the connector to a point at which it is no longer functional—exhibiting cracking, leaking, or improper mating. Moreover, because of the tapered design of the fittings, the hoop stress and resulting strain on the component may exceed the maximum allowable working stress (~50% of the yield value2) or critical strain of the material, and the part will fail in service.
Although a considerable quantity of strain and creep data exist, predictions of a material's suitability based on such information may not be valid for specific end uses due to the idealized nature of the testing or test specimen.3,4 It is therefore of great interest to establish, through experimentation, the expected performance of different thermoplastics in this type of application.
Figure 1. Hoop stress and percent strain on day 28 after one-time torque of 6.0 kg-cm was applied on day 0.
One approach toward this end is to apply interference press-fit equations to the controlled engagement of luer-lock fittings. Specifically, the initial hoop stress and strain present in the female component can be calculated from the degree to which the two components are interlocked. The degree of interlocking is controlled through the level of torque (load) that is used to mate the components. Finally, a representation of short-term creep behavior can be approximated from the level of strain (as a function of time) remaining after the torque is removed.
EXPERIMENTAL
Materials. Female luer locks (FLLs) were molded from four commonly used medical-grade resins that have similar mold-shrinkage factors: polycarbonate (PC), acrylonitrile butadiene styrene (ABS), acrylic/PC alloy (ACR/PC), and rigid polyvinyl chloride (R-PVC). The physical properties of these materials can be found in Table I. A winged FLL design, typical for the medical industry, was selected for this study. The male luer lock (MLL) used as the mating component was manufactured from polyphthalate carbonate (PPC) because of its toughness and strength. An attempt to find a suitable metal MLL was unsuccessful, because each of the metal MLL candidates either cut the FLLs or had a short thread land.
Material | PC | ABS | ACR/PC | R-PVC | PPC |
---|---|---|---|---|---|
Tensile modulus (MPa) | 2414 | 2310 | 2414 | 1690 | 2109 |
Tensile strength-yield (MPa) | 63 | 39 | 61 | 39 | 66 |
Tensile elongation-yield (%) | 6 | 3 | 4.7 | 4.7 | — |
Melt-flow rate (dg/min) | 15 | 5.5 | 3.9 | 1.4 | — |
Appearance | clear | white | white | clear | — |
Process melt temperature (°C) | 287 | 203 | 248 | 164 | — |
Mold temperature (°C) | 93 | 38 | 82 | 43 | — |
Table I. Material properties and processing conditions.
Equipment and Process. The FLLs were molded in a four-cavity shuttle mold (two cavities per shuttle half) on a 24.5-tn vertical injection machine with a 42.5-g shot capacity. The mold cavities and core pins, designed for processing polycarbonate, were used for all four materials. The runner was 6.35 mm full-round, and the parts had a single subgate of 1.14 mm diam in one wing. Processing parameters for each material were based on the material vendor's recommendations and on manufacturing experience. Table I lists the melt temperatures and mold temperatures for each material.
Additional equipment used in this study included a torque measurement device, an optical comparator, and a luer taper gauge based on ANSI/HIMA standard MD70.1-1983.
Procedure. The FLLs were molded in accordance with GMPs and separated into four material sets within three test groups. There were 10 FLLs per material per test group. The three test groups were defined by the level of torque and the frequency at which the connectors were assembled to a corresponding MLL, as follows:
One-time torque—assembled once on day 0 at a specified torque (6.0 kg-cm), disassembled and measured on day 28, measured on day 35.
Incremental torque—assembled on day 0 at a specified torque (6.0 kg-cm); disassembled, measured, and reassembled on days 7, 14, and 21 at respective torque levels of 7.0, 7.5, and 8.0 kg-cm; disassembled and measured on day 28; measured on day 35.
Repetitive torque—assembled on day 0 at a specified torque (6.0 kg-cm); disassembled, measured, and reassembled on days 7, 14, and 21 at a constant torque level of 6.0 kg-cm; disassembled and measured on day 28; measured on day 35.
The starting-point torque of 6.0 kg-cm was determined through experimentation to be a torque that could be obtained by assembling the components manually, but without great effort. The higher torques were introduced to better differentiate the properties of the four resins and to mimic the stresses applied by medical personnel when using tools to connect the fittings.
Figure 2. Hoop stress versus time for incremental-torque testing. Torque targets on days 0, 7, 14, and 21 are 6.0, 7.0, 7.5, and 8.0 kg-cm, respectively.
Three stations were used for the schedule outlined above: a torque measurement device, a luer taper gauge, and an optical comparator. Each test function was performed by the same person for each test period in order to minimize experimental error. The mated fittings were tested and stored in a controlled environment at 22°C and 50% RH. It must be emphasized that the data generated in this study were collected in the absence of typical medical-environment fluids such as alcohol, lipids, water, and saline. Including these and other fluids in a similar study would doubtless produce substantially different data.
At the beginning of the study, the luer tapers on all of the FLLs were measured with an ANSI/HIMA standard MD70.1-1983 gauge and with the optical comparator in order to determine a gauge position baseline (the distance that the tapered gauge enters the FLL) for each part. Each subsequent assembly/disassembly of the individual components generated a new gauge position to which the baseline could be compared. These changes in position provided a means by which the hoop stress and strain of the FLL could be calculated.
RESULTS AND DISCUSSION
The equation for hoop stress in a press-fit application when the shaft (MLL) and the hub (FLL) are of essentially the same material is
where Di and Do are the inner and outer diameters of the FLL, Ds is the outer diameter of the MLL taper, E is the Young's modulus of the FLL material, and is the interference factor.2 Another hoop-stress equation—which does not assume that the materials are the same—was used to validate this approach.5
The equation for hoop strain in this application is:2
In applying these equations, it was assumed that the PPC MLL would distort to a small degree compared with the FLL, and that the gauge-measurement test accurately represented the extent of the components' engagement. It was further assumed that a portion of the deformation in the inner diameter of the FLL was localized and was not proportional through the part's wall; hence, the calculated stress and strain results may be higher than actual. Before the equations could be used, the change in the inner diameter of the FLL had to be determined. For example, the PC FLLs in the one-time-torque group showed an average dimensional change of 0.167 mm. Since the angle of the FLL taper is 1.72°, the dimensional change () is consequently 0.005 mm per side or 0.01 mm per diameter . Finally, the nose diameter (Ds) of the MLL taper is 3.866 mm, and thus the distorted diameter of the FLL (Di) equals Ds — or 3.856 mm.
Material | PC | ABS | ACR/PC | R-PVC |
---|---|---|---|---|
Gauge change (mm) | 0.167 | 0.493 | 0.276 | 0.318 |
Diameter change (mm) | 0.010 | 0.030 | 0.017 | 0.019 |
Distorted diameter (mm) | 3.856 | 3.836 | 3.849 | 3.847 |
Strain (%) | 0.26 | 0.77 | 0.43 | 0.50 |
Hoop stress (MPa) | 4.7 | 13.4 | 7.8 | 6.3 |
Table II. Data for one-time-torque test group at 28 days.
Tables II, III, and IV display the data and calculated results for the three test groups. The progression of the hoop stresses and strains over the 28-day torque period is presented in Figures 1 through 5.
Material | PC | ABS | ACR/PC | R-PVC | |
Day 7 | |||||
Gauge change (mm) | 0.142 | 0.432 | 0.226 | 0.296 | |
Diameter change (mm) | 0.009 | 0.026 | 0.014 | 0.018 | |
Distorted diameter (mm) | 3.857 | 3.840 | 3.852 | 3.848 | |
Strain (%) | 0.22 | 0.68 | 0.35 | 0.46 | |
Hoop stress (MPa) | 4.0 | 11.7 | 6.4 | 5.9 | |
Day 14 | |||||
Gauge change (mm) | 0.240 | 0.602 | 0.371 | 0.526 | |
Diameter change (mm) | 0.014 | 0.036 | 0.022 | 0.032 | |
Distorted diameter (mm) | 3.852 | 3.830 | 3.844 | 3.834 | |
Strain (%) | 0.37 | 0.94 | 0.58 | 0.82 | |
Hoop stress (MPa) | 6.8 | 16.3 | 10.5 | 10.4 | |
Day 21 | |||||
Gauge change (mm) | 0.237 | 0.590 | 0.450 | 0.567 | |
Diameter change (mm) | 0.014 | 0.035 | 0.027 | 0.034 | |
Distorted diameter (mm) | 3.852 | 3.831 | 3.839 | 3.832 | |
Strain (%) | 0.37 | 0.93 | 0.70 | 0.89 | |
Hoop stress (MPa) | 6.7 | 16.0 | 12.7 | 11.2 | |
Day 28 | |||||
Gauge change (mm) | 0.278 | 0.618 | 0.485 | 0.609 | |
Diameter change (mm) | 0.017 | 0.037 | 0.029 | 0.037 | |
Distorted diameter (mm) | 3.849 | 3.829 | 3.837 | 3.829 | |
Strain (%) | 0.44 | 0.97 | 0.76 | 0.96 |
Table III. Data for incremental-torque test group.
Material | PC | ABS | ACR/PC | R-PVC | |
Day 7 | |||||
Gauge change (mm) | 0.120 | 0.448 | 0.159 | 0.177 | |
Diameter change (mm) | 0.007 | 0.027 | 0.010 | 0.011 | |
Distorted diameter (mm) | 3.859 | 3.839 | 3.856 | 3.855 | |
Strain (%) | 0.19 | 0.70 | 0.25 | 0.28 | |
Hoop stress (MPa) | 3.4 | 12.1 | 4.5 | 3.5 | |
Day 14 | |||||
Gauge change (mm) | 0.119 | 0.605 | 0.241 | 0.263 | |
Diameter change (mm) | 0.007 | 0.036 | 0.014 | 0.016 | |
Distorted diameter (mm) | 3.859 | 3.830 | 3.852 | 3.850 | |
Strain (%) | 0.19 | 0.95 | 0.38 | 0.41 | |
Hoop stress (MPa) | 3.4 | 16.4 | 6.8 | 5.2 | |
Day 21 | |||||
Gauge change (mm) | 0.146 | 0.633 | 0.291 | 0.318 | |
Diameter change (mm) | 0.009 | 0.038 | 0.017 | 0.019 | |
Distorted diameter (mm) | 3.857 | 3.828 | 3.849 | 3.847 | |
Strain (%) | 0.23 | 0.99 | 0.45 | 0.50 | |
Hoop stress (MPa) | 4.1 | 17.1 | 8.2 | 6.3 | |
Day 28 | |||||
Gauge change (mm) | 0.187 | 0.676 | 0.337 | 0.362 | |
Diameter change (mm) | 0.011 | 0.041 | 0.020 | 0.022 | |
Distorted diameter (mm) | 3.855 | 3.825 | 3.846 | 3.844 | |
Strain (%) | 0.29 | 1.06 | 0.53 | 0.57 |
Table IV. Data for repetitive-torque test group.
If one examines Figures 1, 3, and 5, two trends become obvious: PC distorts less than the other materials, and ABS distorts more. The behavior of the PC is not surprising, given that this material is well documented for exhibiting exceptional strength and stiffness. Characteristics such as molecular regularity and stiffness, and intermo- lecular attraction resulting from the proximity of resonant phenyl groups from adjacent chains, contribute to these properties.6 On the other hand, ABS is not known for high strength and toughness, but rather for an overall balance of properties. In fact, PC is sometimes alloyed with ABS to enhance the strength, stiffness, and toughness of ABS resins.7
Figure 3. Strain versus time for incremental-torque testing. Torque targets on days 0, 7, 14, and 21 are 6.0, 7.0, 7.5, and 8.0 kg-cm, respectively.
Figure 4. Hoop stress versus time for repetitive-torque testing. Torque targets on days 0, 7, 14, and 21 are all 6.0 kg-cm.
Figure 5. Strain versus time for repetitive-torque testing. Torque targets on days 0, 7, 14, and 21 are all 6.0 kg-cm.
The ACR/PC also behaved as expected. This material's performance was slightly inferior to that of PC; its acrylic and modifier components probably contributed to the resulting increased strain. Regarding the R-PVC fittings, they performed admirably until they were tested at the highest torques (7.5 and 8.0 kg-cm). At these loads, 60% of the specimens fractured in a brittle manner, primarily at the weld line during the torque assembly procedure. The mode of failure for these components was hoop stress, or a combination of hoop stress and torsion. A weak weld line most likely contributed to these failures. It has been reported that a thin wall in an FLL application can result in weld-line failures, whereas a thicker, stronger wall can remedy this problem.8
An additional observation for the three test groups is that crazing was detected in the clear samples (PC and R-PVC), even at the lower strain levels. Crazing was not detected by microscopy (10x) in the white components (ACR/PC and ABS). Although each of the crazed components was capable of sustaining loads, crazing will ultimately lead to failure.4 One may also consider a theory that proposes that preferential yield formation behaves as a toughness enhancer—more specifically, that crazing acts as a stress relief that extends the time prior to failure.3
In relative terms, each of the materials behaved predictably in this application with respect to its vendor-reported stress/strain curve. Assuming the toughness of each material to be approximated by the area beneath its respective stress/strain curve, the poly-mers would rank as follows, in order of descending toughness: PC, ACR/PC, R-PVC, and ABS. The data obtained in this study are for the most part consistent with these rankings.
Material | PC | ABS | ACR/PC | R-PVC |
---|---|---|---|---|
One-time torque | ||||
Gauge change (mm) | 0.148 | 0.408 | 0.236 | 0.297 |
Diameter change (mm) | 0.009 | 0.025 | 0.014 | 0.018 |
Distorted diameter (mm) | 3.857 | 3.841 | 3.852 | 3.848 |
Strain (%) | 0.23 | 0.64 | 0.37 | 0.46 |
Incremental torque | ||||
Gauge change (mm) | 0.205 | 0.599 | 0.389 | 0.515 |
Diameter change (mm) | 0.012 | 0.036 | 0.023 | 0.031 |
Distorted diameter (mm) | 3.854 | 3.830 | 3.843 | 3.835 |
Strain (%) | 0.31 | 0.94 | 0.60 | 0.91 |
Repetitive torque | ||||
Gauge change (mm) | 0.123 | 0.639 | 0.249 | 0.252 |
Diameter change (mm) | 0.007 | 0.038 | 0.015 | 0.015 |
Distorted diameter (mm) | 3.859 | 3.828 | 3.851 | 3.851 |
Strain (%) | 0.18 | 0.99 | 0.39 | 0.39 |
Table V. Data for 7-day strain-recovery period.
Table V and Figures 6 through 8 depict the unrecovered strains of the materials in each test group 7 days after the removal of the hoop stress. These strain data represent the short-term creep behavior of the test specimens, with short-term being defined as 650+ (cumulative or consecutive) hours of applied load and 168 hours of recovery.
Figure 6. Total and unrecovered strains versus material for one-time-torque testing. Percent change is noted in parentheses.
Figure 7. Total and unrecovered strains versus material for incremental-torque testing. Percent change is noted in parentheses.
Figure 8. Total and unrecovered strains versus material for repetitive-torque testing. Percent change is noted in parentheses.
Again, there is a clear pattern for the ABS: the inability of the material to recover from the strain indicates that the resin's critical strain level has been approached and also that the ABS may not be suitable in this combination of stress level and part wall thickness. Supporting the former proposition is the fact that the greatest strain recovery of the ABS in this study takes place in the lowest-torque test group, and that very little recovery occurs in the higher-torque test groups. Similarly, R-PVC failed to recover at the highest torque levels, and, like ABS, may be questionable at this level of stress. PC and ACR/PC typically showed the greatest elastic behavior in each test group.
It is interesting that the recovery of the resins (apart from ABS) in the one-time-torque test group was lower as a percentage than that for the repetitive-torque test group. This behavior perhaps supports the assertion that the rate of stress relaxation—and therefore of strain recovery—is greater at higher elongations that approach the yield point.4 On the other hand, the influence of the test-group parameters (torque, frequency of assembly and disassembly, time) could be dictating this response. For instance, the FLLs in the one-time test group were exposed to one torque for 28 days, whereas the FLLs in the repetitive test group had the opportunity to relax three times (albeit for less than 1 hour) between disassembly and assembly, and the longest application of any single torque was for 7 days. Accordingly, the former group may have sustained a greater amount of creep as opposed to the temporary, recoverable strain that was exhibited by the latter.
CONCLUSION
Several conclusions can be drawn from the results of this study:
The ABS luer taper fittings displayed the highest levels of hoop stress and strain in each of the test groups. PC, followed by ACR/PC, demonstrated the greatest resistance to strain. R-PVC showed adequate strength except at the highest torque levels (7.5 and 8 kg-cm), where some of the specimens failed in a brittle manner.
ABS exhibited the worst strain recovery of the materials in both the incremental- and repetitive-torque test groups. R-PVC failed to recover only in the highest-torque applications. PC and ACR/PC demonstrated moderately good elastic behavior in each test group.
A relationship between the strain/recovery performance and the tensile strength at yield values of each material was apparent.
The study offers a method that quantitavively approximates strain levels in female luer fittings.
ACKNOWLEDGMENTS
The authors thank Prakash Kolluri, Bayer Polymers Div., for his thoughts on press-fit applications. Additional thanks go to Bart Inlow and Mike Zawilinski for their contributions to this undertaking.
REFERENCES
1. American National Standard for Medical Materiel—Luer Taper Fittings—Performance, ANSI/HIMA MD70.1-1983, New York, American National Standards Institute, 1984.
2. Plastics Design Manual, Monsanto Technical Bulletin, St. Louis, Monsanto Co., 1994.
3. Birley AW, Haworth B, and Batchelor J, Physics of Plastics—Processing, Properties, and Materials Engineering, Munich, Hanser Verlag, 1992.
4. Nielsen LE, and Landel RF, Mechanical Properties of Polymers and Composites, New York, Marcel Dekker, 1994.
5. Malloy RA, Plastic Part Design for Injection Molding: An Introduction, Munich, Hanser Verlag, 1994.
6. Deanin RD, Polymer Structure, Properties, and Applications, Boston, Cahners, 1972.
7. Margolis JM, Engineering Thermoplastics—Properties and Applications, New York, Marcel Dekker, 1985.
8. Stubstad J, Irradiation of IV Sets: A Ten-Year Case Study, Preventing Plastic Part Failure, Madison, WI, University of Wisconsin, 1990.
Joseph Lane is a project engineer at Sabin Corp. (Bloomington, IN). Prior to joining Sabin, he worked in resin development and technical service at Nylon Corp. of America. He holds BS and MS degrees in plastics engineering from the University of Massachusetts at Lowell. A process engineer at Sabin since 1990, Delbert C. Miller has worked in the injection molding field since 1972. His experience includes both management and technical responsibilities, with a primary focus in the processing area.
SPE TECHNICAL PAPER SERIES APPLIED MEDICAL POLYMERS
Medical Plastics and Biomaterials is pleased to present the thirteenth installment of its SPE Technical Paper Series: Applied Medical Polymers, a collaborative effort of Canon Communications and the Society of Plastics Engineers Medical Plastics Division. The purpose of this effort is to provide early publication of technical papers dealing with important aspects of the application of polymeric materials to uses in the medical device and general health-care areas and to disseminate this information to the readers of Medical Plastics and Biomaterials, including the membership of the SPE Medical Plastics Division. The papers will have been subjected to comprehensive peer review and approval by prominent scientists and technical specialists in the field. Papers appearing in this series will be subsequently published by the Society of Plastics Engineers as the Journal of Applied Medical Polymers and be available from the SPE by subscription.
CALL FOR PAPERS
Authors are invited to submit both papers reporting original research in the development of medical polymer applications and review articles on medical polymer science and technology for publication in Medical Plastics and Biomaterials' SPE Technical Paper Series: Applied Medical Polymers. An Information for Authors sheet detailing required manuscript style can be obtained by calling the MPB editorial office at 310/392-5509. Manuscripts should be forwarded to the SPE Series Scientific Editor at the following address: Robert C. Portnoy, PhD, c/o Canon Communications llc, 3340 Ocean Park Blvd., Ste. 1000, Santa Monica, CA 90405.
Scientific Editor Robert C. Portnoy, PhD
Exxon Chemical Co.
Advisory Board
Len Czuba
Baxter Healthcare
Joel L. Williams, PhD
Lecon Woo, PhD
Baxter Healthcare
Copyright ©1998 Medical Plastics and Biomaterials
About the Authors
You May Also Like