Common methods of achieving a desired preload during installation are controlling the installation torque, controlling the turn angle of a nut or bolt head after the fastening system is seated, measuring the fastener’s change in length, and using strain gauges on the fastener body.
These methods are listed in order of increasing labor, cost, and accuracy. The most economical method, torque control, is the least accurate for controlling preload.
Refer to Appendix A for explanation and justification of fastener analysis criteria.
4.3.1 Maximum and Minimum Preload
[TFSR 5] Maximum and minimum preload calculations shall account for the preload variation associated with the method of developing initial preload, potential relaxation, creep, and the effects of maximum and minimum expected temperatures.
[Rationale: Many variables can affect the preload that is developed in a threaded fastening system during and after assembly. Proper characterization and use of the full range of possible preloads are necessary to ensure structural integrity is verified.]
Maximum and minimum preloads are calculated as
where Ppi-max and Ppi-min are the maximum and minimum initial preloads, respectively; the other variables are defined in Table 1, Analytical Adjustments of Maximum and Minimum Preloads.
| Symbol | Description | Adjustment used when calculating maximum preload | Adjustment used when calculating minimum preload |
|---|---|---|---|
| Ppr | Short-term relaxation of preload (often attributed to embedment of imperfectly matched surfaces); see Appendix A.3 | Zero | 5% of minimum initial preload for joints with all-metallic clamped parts, test-derived value for joints with any non-metallic parts or coatings)(1) |
| Ppc | Loss of preload from material creep | Zero | Calculated maximum expected preload loss from creep, if applicable |
| PΔt-max, PΔt-min | Change of preload with temperature (differential coefficients of thermal expansion and temperature dependence of elastic moduli)(2) | Calculated maximum increase in preload, PΔt-max, for maximum or minimum expected temperatures | Calculated maximum decrease (as a positive number) in preload, PΔt-min, for maximum or minimum expected temperatures |
(1)For fastened joints with multiple faying surfaces, the assumption of 5 percent relaxation may be non-conservative. In these cases, it is recommended to perform testing to determine the relaxation or analysis that considers the creep-relaxation behavior of the fastening system design. (2)An effective technique to minimize the change of preload with temperature is described in the SAE AIR 1754A, Washer, Thermal Compensating, Metric Series.
Maximum initial preload for strength and fatigue analyses is calculated as
where Ppi-nom is the nominal (mean) preload applicable to installation, is the preload variation, and cmax is a factor that accounts for the maximum value of the controlled installation parameter, as allowed by the specified tolerance. For example, if torque control is used and the effective torque is specified as 40 ± 2 N-m, then cmax = (40 + 2)/40 = 1.05.
For use in separation analysis of separation-critical joints and for fatigue analysis, minimum initial preload is calculated as
where cmin is a factor that accounts for the minimum value of the controlled installation parameter, as allowed by the specified tolerance. For example, if torque control is used and the effective torque is specified as 40 ± 2 N-m, then cmin = (40 – 2)/40 = 0.95.
For use in joint-slip analysis and separation analysis of joints that are not separation-critical, minimum initial preload is calculated as
where nf is the number of fasteners in the joint. See Appendix A.2 for rationale.
In addition to the adjustments noted in Table 1, loss of preload may occur if the joint experiences cyclic slip. To minimize any such loss, the joint should be designed in any one of the following ways:
- Not to slip; or
- With minimal clearance fits achieved by specifying precision fasteners having a closely controlled or “close tolerance” full diameter body in conjunction with precision holes per National Aerospace Standard (NAS) 618, Fastener – Recommended Shank, Hole, and Head-to-shank Fillet Radius, Limits For; or
- With liquid thread fillers that cure properly; or
- With shear pins (dowel pins) or other dedicated shear-transfer devices.
4.3.2 Nominal Preload
[TFSR 6] Calculation of the nominal (mean) initial preload, Ppi-nom , shall be substantiated by tests of a minimum of six sets of the fastening system hardware per Table 2, Nominal Preload Determination, to determine the relationship between initial preload and the parameter controlled during installation (torque, turn-of-nut, turn-angle, or bolt stretch).
[Rationale: When using torque control, the torque-preload relationship has been shown to be sensitive to where lubrication is applied. Significant differences have been observed, for example, when lubricating only under the head compared to under the head and nut. When using turn-of-nut, turn-angle, or bolt stretch designs, the torque-preload relationship has been observed to be configuration dependent; therefore, tests should use hardware and processes that are identical to the flight design.]
| Torque Control(1,2) | Turn-of-Nut or Turn-Angle | Bolt Stretch | |
|---|---|---|---|
| The fastening system hardware has the same diameter and thread form, the same type and number of washers, same materials, and same nut/nut plate/insert as the flight assembly. The clamped part and washer that are adjacent to a non-rotating bolt head or non-rotating nut are not critical and are allowed to vary from the flight assembly. | R | R | R |
| Cleaning, lubricants, and lubrication process are the same as flight assembly. | R | R | NR |
| The fastening system hardware is the same specification (part number) as the flight assembly hardware. | NR | R | R |
| Installation process is the same as flight assembly. | NR | R | R |
| R="Required," NR="Desirable" but not required (1) At least three tests (install, torque, and removal) should be performed on each of the six sets of fastening system hardware for a total of eighteen tests, unless reuse is prohibited in the flight assembly. (2) See NASM 1312-15, Fastener Test Methods, Method 15, Torque-Tension, for guidance for torque-tension testing. | |||
| Method | Separation Critical | Non-Separation Critical | |
| Torque Control | Envelope of statistical basis(1) and variation to satisfy Table 2(2) while using a. Lot-specific testing (testing the procurement lot of fastening system hardware that will be used for spaceflight)(3) or b. Testing equal numbers of sets of fastening system hardware of the same specification as the flight-assembly hardware, from each of at least three procurement lots(3) | a. Statistical basis(1), or b. Greater of (1) 25 percent (if lubricated), (2) 35 percent (if non- lubricated or as- received), (3) Variation from tests to satisfy Table 2(2) | |
| Turn-of- Nut or Turn Angle | Envelope of statistical basis(1) and variation to satisfy Table 2(2) | a. Statistical basis(1), or b. Greater of (1) 25 percent (2) Variation from tests to satisfy Table 2(2) | |
| Bolt Stretch | Envelope of statistical basis(1) and variation to satisfy Table 2(2) | a. Statistical basis(1), or b. Greater of (1) 10 percent (2) Variation from tests to satisfy Table 2(2) | |
| (1)90 percent probability and 95 percent confidence (two-sided distribution), with tests meeting the configuration requirements of Table 2. (2)The variation from the tests to satisfy Table 2 is given by the minimum and maximum values of preload relative to the nominal preload such that Γa-max=Ppi-maxa/Ppi-nom-1 and Γa-min=1-Ppi-mina/Ppi-nom. (3)Each fastener to be installed, torqued, and removed at least three times unless reuse is prohibited for the flight assembly. When reuse is prohibited, only the data from the first cycle is used. | |||
4.3.3 Preload Variation
[TFSR 7] The preload variation, , used to calculate the minimum and maximum initial preload shall be based on the criteria of Table 3, Preload Variation Determination.
[Rationale: Testing has demonstrated that when the same torque is applied, there is variability in the measured preload. To ensure safe operation of fastening system hardware, this variability should be accounted for in the design. A 90 percent probability and 95 percent confidence limit was selected to provide balance, analogous to a B-basis quantity, between conservatisms and limiting the required test matrix to a reasonable number of specimens.]
Table 3 – Preload Variation Determination
For non-separation critical joints, Table 3 provides several methods for determining the preload variation, . The preferred approach is to use a statistical basis with 90 percent probability and
95 percent confidence for a two-sided distribution. However, to avoid excessive testing, it is permissible to use the alternate (not the statistical) method shown in Table 3.
When using the alternate (not the statistical) method for non-separation critical joints, if the test- measured variation exceeds the values stated above, then the test-measured variation should be used or methods for reducing the preload variation should be pursued.
When using the alternate method for torque control, a fastener is considered to be lubricated if a solid-film, grease, or liquid lubricant is applied to the threads and to the turning bearing surfaces of the nut or the bolt head (or to the washer under the nut or the bolt head).
Appendix A.2 shows an example of how to calculate minimum and maximum preloads when preload is generated with torque control.